Recht en Menselijk Gedrag

Faculteit der Rechtsgeleerdheid
Amsterdam College of Law

Algemene Rechtsleer
Bachelor Rechtsgeleerdheid
Recht en menselijk gedrag

Reader
2022-2023 / Semester 2 blok 1
3011RMX1VY
Inhoudsopgave reader Recht en menselijk gedrag 2022-2023
WEEK 1 – Inleiding
Studiewijzer
Studiewijzer Empirisch juridisch onderzoek: Wat is het en hoe bestudeer ik het? 3
Pleasance, Balmer en Denvir 2017

P. Pleasence, N. J. Balmer, & C. Denvir, ‘Wrong about rights: public knowledge of key areas of 19
consumer, housing and employment law in England and Wales.’ The Modern Law Review 2017,
vol. 80, afl. 5, pp. 836-859.
Rijnhout, Giesen, van Aken en van Hoof 2016

R. Rijnhout, I. Giesen, M. A. G. van Aken, & J. T. C. van Hoof, ‘Agressie van jeugdige 43
voetballers en de rol van het recht: kennis en (on) begrip van de (spel) regels empirisch getoetst.’
Aansprakelijkheid, Verzekering en Schade 2016, afl. 5, pp. 171-179.
WEEK 2 – Implementatie van het recht I: Afschrikkende werking van straffen
Wermink, Blokland, Nieuwbeerta en Tollenaar 2009

H. Wermink, A. Blokland, P. Nieuwbeerta, & N. Tollenaar, ‘Recidive na werkstraffen en na 55
gevangenisstraffen.’ Tijdschrift voor Criminologie 2009, vol. 51, afl. 3, pp. 211-227.
Thornton, Gunningham en Kagan 2005

D. Thornton, N. A. Gunningham, & R. A. Kagan, ‘General Deterrence and Corporate 73
Environmental Behavior.’ Law & Policy 2005, vol. 27, afl. 2, pp. 262-288.
WEEK 3 – Implementatie van het recht II: Naleving in de sociale context
Anderson 2000
E. Anderson, Code of the street: Decency, violence, and the moral life of the inner city. New 103
York, NY: WW Norton & Company, pp. 9-11, 32-34, en 67-90.
Kerr Lawrence 2015

N. Kerr Lawrence, ‘Highlighting the injunctive norm to reduce phone-related distracted driving.’ 133
Social Influence, 2015, vol. 10, afl. 2, pp. 109-118.
WEEK 4 – Implementatie van het recht III: Handhavingsstijlen
de Winter 2020
P. de Winter, ‘Handhaving van de bijstandsregels’ in: T. Kampen, M. Sebrechts, T. Knijn and E. 145
Tonkens (red.), Streng maar onrechtvaardig, Amsterdam: van Gennep 2020, pp. 115-129.
Mascini en Van Wijk 2008

P. Mascini & E. van Wijk, ‘'Vis ruikt nou eenmaal zo': Responsive regulation door de Voedsel en 159
Waren Authoriteit.’ Tijdschrift voor Criminologie, 2008, vol. 51, afl. 2, pp. 114-129.
WEEK 5 – Geschilbeslechting I: Mobilisatie van rechten
Alexander en Prasad 2014

C. S. Alexander & A. Prasad, ‘Bottom-Up Workplace Law Enforcement.’ Indiana Law Journal, 177
2014, vol. 89, afl. 3, pp. 1069-1118.
Fritz-Mauer 2022
M. Fritz-Mauer, ‘Naming, blaming, and just plain giving up.’ SSRN, 3 May 2022, 227
http://dx.doi.org/10.2139/ssrn.4029827. Geraadpleegd op 1 december 2022.
WEEK 6 – Geschilbeslechting II: Presentatie voor de rechter
Conley, O’Barr en Lind 1979

J. M. Conley, W. M. O'Barr, & E. A. Lind, ‘The Power of Language: Presentational Style in the 279
Courtroom.’ Duke Law Journal, 1979, afl. 6, pp. 1375-1399.
Wermink, de Keijser en Schuyt 2012

H. Wermink, J. de Keijser, & P. Schuyt, ‘Verschillen in straftoemeting in soortgelijke zaken. Een 305
kwantitatief onderzoek naar de rol van specifieke kenmerken van de dader.’ Nederlands
Juristenblad, 2012, afl. 11, pp. 726-733.
WEEK 7 – Geschilbeslechting III: Beleving door rechtzoekenden
Relis 2006
T. Relis, ‘It's not about the money: a theory of misconceptions of plaintiffs' litigation aims.’ 311
University of Pittsburgh Law Review 2006, vol. 68, pp. 701-746.
Robbennolt 2003
J. K. Robbennolt, ‘Apologies and legal settlement: An empirical examination.’ Michigan Law 361
Review 2003, vol. 102, afl. 3, pp. 460-491, 501-502, en 504-516.
Week 1
1
2
Studiewijzer Empirisch juridisch onderzoek: Wat is het en hoe bestudeer ik het?
Drie typen van rechtswetenschappelijk onderzoek
Stel je voor dat je als wetgevingsjurist wordt gevraagd om nieuwe regels te ontwikkelen om het
verbod van mobiele telefoons tijdens het fietsen beter te kunnen handhaven. Volgens een van jouw
collega’s is het probleem dat overtreders niet streng genoeg gestraft worden. Deze collega stelt
voor om hardleerse overtreders, die voor een derde keer betrapt worden terwijl ze appen op de fiets,
voor een half jaar de toegang tot hun mobiele abonnement te ontnemen. Een dergelijk voorstel
roept verschillende vragen op.
Ten eerste kan je je afvragen of er wel al wettelijke basis is voor een dergelijke sanctie en of dit
juridisch wel mag. Een dergelijke sanctie zou een directe inbreuk maken op een contractuele relatie
tussen de overtreders en hun telefoonaanbieder. Bovendien zou het een grote aantasting vormen
van hun persoonlijke vrijheid, en hen sterk belemmeren in hun dagelijks leven. Om te onderzoeken
of deze sanctie juridisch gezien wel mag, dien je positiefrechtelijk onderzoek te doen. Je gaat kijken
in geldende bronnen van het recht welke bepalingen er relevant zijn voor de voorgestelde sanctie
voor appende fietsers, en op welke manier deze dienen te worden geïnterpreteerd. Bij een dergelijk
onderzoek zal je dus waarschijnlijk kijken in geldende regels (zoals die in het bestuursrecht,
constitutioneel recht, eigendomsrecht, contractenrecht en mensenrechtenbepalingen) en de
relevante jurisprudentie die deze regels verder interpreteert.
Bij een dergelijk voorstel kan je je ook afvragen of een dergelijk voorstel wel rechtvaardig en
wenselijk is. In dit geval stel je een vraag die niet doctrinair is (wat juridisch toelaatbaar is), maar
normatief (wat goed en slecht is). Om deze vraag te beantwoorden is filosofisch onderzoek vereist.
Bij een dergelijk onderzoek ga je het voorstel beoordelen aan de hand van rechtsfilosofische ideeën,
zoals bijvoorbeeld hoe het belang van persoonlijke vrijheid afweegt tegen het beschermen van
publieke veiligheid. Bij dergelijk onderzoek kijk je naar wat filosofische denkers eerder schreven
over vergelijkbare dilemma’s en analyseer je hoe verschillende normatieve standpunten wel en niet
op het voorgestelde plan passen.
Tenslotte kan je je ook afvragen in hoeverre de voorgestelde sanctie wel effectief gaat zijn in het
tegengaan van digitale afleiding in het fietsverkeer. In dit geval stel je een vraag die niet
positiefrechtelijk (is het juridisch toelaatbaar) of filosofisch (is het wenselijk) is, maar empirisch:
als wij het denken of handelen van mensen onderzoeken, wat gebeurt er daar dan aan als zij strenger
gestraft kunnen worden wanneer zij de regels overtreden? Om deze vraag te kunnen beantwoorden
is empirisch onderzoek nodig. In dit geval ga je in de werkelijkheid observeren hoe strengere
straffen van invloed zijn op de mate waarin fietsers appen – of, probeer je daar inzicht in te krijgen
door mensen bijvoorbeeld vragen te stellen over hun eigen gedrag (hoe vaak appen zij op de fiets?
Hoe ervaren zij de sancties?), of door hun gedrag te bestuderen in een experiment (hoe beïnvloeden
hogere sancties hun naleving van bepaalde regels, vergeleken met een controlegroep waar
dergelijke sancties niet voor gelden?). In zulk onderzoek wordt gebruik gemaakt van
onderzoeksmethoden afkomstig uit de sociale wetenschap (zoals interviews, enquêtes of
(veld)experimenten). In empirisch juridisch onderzoek probeer je juridische vragen dus te
3
beantwoorden aan de hand van empirisch onderzoek, waarin observatie van waarneembare
uitkomsten (zoals voorkeuren of gedrag) centraal staat.
Positiefrechtelijk
Filosofisch Empirisch
Figuur 1: Drie typen van juridisch onderzoek
Er zijn dus drie typen van juridisch onderzoek:

• Positiefrechtelijk onderzoek: onderzoek dat een analyse maakt van wat de geldende
juridische regels van het positieve recht zijn op basis van rechtsbronnen en de interpretatie
daarvan.
• Filosofisch onderzoek: onderzoek dat met rechtsfilosofische bronnen de rechtvaardigheid
en wenselijkheid van (aspecten van) rechtsregels, of de toepassing daarvan, analyseert.
• Empirisch onderzoek: onderzoek dat met gebruik van sociaalwetenschappelijke methoden
vaststelt hoe het recht werkt in de praktische werkelijkheid, en hoe dit door
mensen/actoren/betrokkenen ervaren en geïnterpreteerd wordt.
Zoals Figuur 1 duidelijk maakt is er geen absoluut onderscheid tussen de drie typen van juridisch
onderzoek. Een filosofische analyse naar de wenselijkheid van sancties om het app-verbod te
handhaven zal ook empirisch onderzoek moeten omvatten naar de effectiviteit van die sancties. En
een empirische analyse naar de effectiviteit van een bepaalde regel, zoals bijvoorbeeld het
ontzeggen van een mobiel abonnement voor recidiverende telefoongebruikers op de fiets, zal eerst
positiefrechtelijk moeten analyseren wat zo’n regel precies zal inhouden binnen de bestaande
rechtskaders en bij empirische onderzoek zullen altijd filosofische aannames en waarden impliciet
of expliciet een rol spelen.
Typen van Empirisch Onderzoek
Er zijn verschillende typen van empirisch juridisch onderzoek. In algemene zin kunnen we een
onderscheid maken tussen kwalitatief en kwantitatief onderzoek. Kwantitatief onderzoek vertaalt
de empirische werkelijkheid naar patronen en cijfers die zich lenen voor statistische analyse. We
kunnen hierbij denken aan enquêtes, experimenten, en analyses van overheidsdata. Vaak gaat het
hier om grotere datasets die observaties van een groot aantal personen of casus bevatten.
Kwantitatief onderzoek is uitermate geschikt om algemene patronen aan te tonen, en om de
4
statistische relatie tussen gemeten variabelen te doorgronden en toevalligheden uit te sluiten.
Daarom is kwantitatief onderzoek bij uitstek geschikt om theorie te toetsen. Kwalitatief onderzoek
poogt de empirische realiteit te begrijpen door een nauw contact met die werkelijkheid. We kunnen
hierbij denken aan onder andere participerende observatie en diepte-interviews. De onderzoeker
probeert het fenomeen of het perspectief van de betrokkenen in alle rijkheid te illustreren, op een
manier die zich het beste vertaalt in geschreven taal. Kwalitatief onderzoek is uitermate geschikt
om een inzicht te krijgen in de empirische werkelijkheid vanuit het perspectief van de onderzochte
personen. Het biedt nieuwe perspectieven op bestaande wetenschappelijke ideeën en kan daarom
goed dienen om theorie te ontwikkelen.
Onder in Tabel 1 staat een overzicht van een aantal belangrijke typen van empirisch juridisch
onderzoek.
Type Omschrijving
Bestudering door observatie van processen in de werkelijkheid
Participerende observatie
tijdens participatie in die processen.
Data verzameld door het (persoonlijk) stellen van vragen aan
Interviews
respondenten/informanten.
Data verzameld door middel van vragenlijsten ingevuld door
Enquêtes
respondenten.
Analyse van overheidsdata Analyse van data verzameld door de overheid.
Testen van het effect van een of meer interventie (s) /
Laboratorium-experimenten variabele(n) in een (meestal artificiële) gecontroleerde
onderzoeksomgeving.
Testen van het effect van een (of meer) interventie (s) /
variabele(n) in de dagelijkse praktijk, door die interventie daar
Veldexperimenten in werkelijkheid uit te voeren, en het effect ervan te vergelijken
met een (vergelijkbare) controlegroep voor wie deze interventie
niet heeft plaatsgevonden.
Vergelijken van data tussen twee (of meer) in werkelijkheid
voorkomende situaties die van elkaar verschillen op één (of een
Natuurlijke experimenten beperkt aantal) kenmerk(en) / variabele(n) (maar waarbij niet
voor alle verschillen tussen deze situaties gecontroleerd kan
worden).
Tabel 1: Typen van empirisch onderzoek
Stel dat we empirisch zouden willen onderzoeken welke invloed boetes hebben op de mate waarin
mensen appen op de fiets in een bepaalde woonwijk. We zouden dat met elk van deze methoden
kunnen doen, als volgt:
• Participerende observatie: de onderzoeker zou kunnen meelopen met BOAs wanneer ze in
deze buurt overtreders aanspreken op hun gedrag en boetes uitdelen. Tijdens deze interactie
zou de onderzoeker kunnen optekenen hoe de BOAs hun werk doen, hoe ze met de burgers
communiceren over de boetes, en hoe dat de burgers daar weer op reageren.
• Interviews: de onderzoeker zou kunnen proberen om toegang te krijgen tot inwoners van
de buurt om ze te interviewen (liefst in persoon, maar kan ook telefonisch of virtueel). De
5
onderzoeker zou hen dan kunnen vragen wat ze denken over deze regels, hoe ze hun eigen
gedrag met hun telefoon beleven (waaronder ook op de fiets), waarom ze zelf denken dat
ze zich wel en niet aan de regels houden, wat ze denken over de boetes, en wat er breder in
hun leven speelt dat wellicht relevant is voor de naleving van deze regels.
• Enquêtes: de onderzoeker zou ook een schriftelijke vragenlijst kunnen ontwikkelen, met
vragen over het gebruik van de telefoon tijdens het fietsen, hoe ze de pakkans en strafmaat
van de sancties voor overtredingen zien, en andere factoren die mogelijk van invloed op
hun naleving zouden kunnen zijn. Deze enquête zou in persoon, online, of per telefoon
kunnen worden afgenomen onder een representatieve steekproef van bewoners uit de buurt.
• Analyse van overheidsdata: de onderzoeker zou deze vraag ook kunnen bestuderen door
toegang te verkrijgen tot de data die geregistreerd wordt door de politie of door BOAs,
bijvoorbeeld over regelovertredingen, overtreders, en uitgedeelde boetes. De onderzoeker
zou door middel van statistische analyse van deze data kunnen proberen te achterhalen of
er verbanden bestaan tussen naleving en de kenmerken van overtreders, de locatie waar de
overtreding plaatsvindt, en de hoogte van de sanctie.
• Laboratorium-experimenten: het is logistiek niet makkelijk om een laboratorium-
experiment over appen op de fiets op te stellen, omdat er in een laboratorium meestal
simpelweg geen ruimte is om te fietsen. Wat wel kan (en ook gedaan wordt), is om met een
experiment te bestuderen hoe sancties van invloed zijn op (verboden) telefoongebruik in
andere situaties. Zo zou de onderzoeker bijvoorbeeld aan deelnemers in het laboratorium
kunnen vragen om tegen betaling lastige historische kennisvragen te beantwoorden, zonder
daarbij hun telefoon te gebruiken, maar de deelnemers kunnen dit wel degelijk doen. Het
effect van straffen op naleving zou hier getest kunnen worden door drie groepen met elkaar
te vergelijken: een groep waar er een hoge boete is; een groep waar de boete laag is; en
tenslotte een controlegroep, waar er geen boete is.
• Veldexperimenten: de onderzoeker zou ook een veldexperiment kunnen opzetten, waarin
het effect van boetes in de praktijk onderzocht wordt. In twee vergelijkbare buurten (op
voldoende afstand van elkaar) wordt eerst door middel van observatie geteld hoe veel
mensen appen op de fiets op een bepaald tijdstip in de week. Vervolgens wordt een paar
dagen later op een van deze twee plekken een bord opgehangen waarop staat dat fietsers
een boete riskeren wanneer zij appen op de fiets. Op de andere plek wordt geen bord
geplaatst. Er wordt vervolgens opnieuw op hetzelfde tijdstip in de week geteld op de beide
locaties. Het aantal mensen dat appt op de fiets wordt vergeleken tussen de twee de locaties
en met de meting die voorafgaand het experiment gedaan werd. Bij een veldexperiment
wordt het effect van een interventie in de werkelijkheid vergeleken met een controlegroep
zonder deze interventie. De interventie wordt door de onderzoekers zelf georganiseerd en
toegepast zodat er zo weinig mogelijk verschillen zijn in de samenstelling en
omstandigheden van de controlegroep en de interventiegroep.
• Natuurlijk experiment: tenslotte zou de onderzoeker onder sommige omstandigheden een
natuurlijk experiment kunnen verrichten. Dit zou bijvoorbeeld mogelijk zijn als er op een
bepaald moment in deze buurt een extra handhavingscampagne zou zijn, waardoor BOAs
vaker controleren op overtredingen, en de pakkans sterk omhoog gaat. De studie zou dan
het aantal overtredingen voor, tijdens en na de campagne kunnen registreren, zowel in deze
buurt als in een vergelijkbare buurt waar op dezelfde tijdstippen geen campagne plaatsvindt.
Door de niveaus van naleving op deze tijdstippen te vergelijken zou de onderzoeker kunnen
6
analyseren wat het effect van een dergelijke extra intensieve handhaving is. Bij een
natuurlijk experiment bestudeert de onderzoeker het effect van een interventie zonder dat
de onderzoeker zelf directe invloed heeft op deze interventie, maar door gebruik te maken
van het feit dat de interventie reeds heeft plaatsgevonden en er een vergelijking te maken
is tussen de groep waar deze plaatsvond en een zo vergelijkbaar mogelijke andere groep
zonder de interventie.
Voor- en nadelen van methoden: validiteit, 1 generaliseerbaarheid, en causaliteit
Empirisch juridisch onderzoek probeert de werkelijkheid waarin het recht functioneert in kaart
brengen, te beschrijven, te verklaren, en soms zelfs te voorspellen. Dit is geen eenvoudige opgave.
Het onderzoek is inherent beperkt in hoe goed het in staat is om deze werkelijkheid weer te geven.
Hierbij hebben verschillende typen van empirisch onderzoek elk hun eigen voor- en nadelen –
zowel voor wat betreft de deugdelijkheid van het onderzoek, de haalbaarheid en betaalbaarheid van
de soort studie, en de relevantie voor theorie en beleid. In Appendix 1 zie je een overzicht van een
aantal belangrijke voor- en nadelen die aan elke methode verbonden zijn. In algemene zin kunnen
we deze voor- en nadelen samenvatten tot drie kernelementen van de kwaliteit van empirisch
onderzoek: validiteit, generaliseerbaarheid, en causaliteit. Deze voor- en nadelen staan nauw
verbonden aan het doel van het onderzoek. Zo zal onderzoek dat probeert de attitudes van rechters
te begrijpen bij het beoordelen van beslissingen juist veel hebben aan subjectieve informatie uit
diepte interviews, die dan het beste betrouwbare beeld geven, hetgeen weer moeilijker is wat betreft
generaliseerbaarheid en causaliteit.
De validiteit van empirisch onderzoek duidt op de mate waarin het onderzoek een valide en
realistisch beeld geeft van hetgeen dat onderzocht wordt. Bij het onderzoek naar de invloed van
sancties op naleving van het appverbod gaat het dan om vragen zoals: hebben de onderzoekers
tijdens de etnografie wel echt geobserveerd op de locaties waar er overtredingen plaatsvinden?
Heeft hun aanwezigheid de interactie tussen BOAs en burgers niet mogelijk beïnvloed? Zijn
mensen wel eerlijk wanneer zij in een interview of vragenlijst vragen krijgen over hun eigen
overtredingen? Omvatten data van de overheid of uit natuurlijke of veldexperimenten het
daadwerkelijke aantal overtredingen, of registreren zij slechts diegenen die gepakt zijn? Het
antwoord op deze vragen is cruciaal om in te schatten wat het onderzoek werkelijk laat zien, en in
welke mate het een valide beeld geeft van het fenomeen zoals het in de werkelijkheid voorkomt.
De generaliseerbaarheid van empirisch onderzoek duidt op de mate waarin wat het onderzoek laat
zien bij de onderzochte personen, gedragingen, en omstandigheden ook iets kan zeggen over
andere relevante situaties, buiten degene die direct zijn onderzocht. Bij het onderzoek over het
appverbod kunnen we ons bijvoorbeeld afvragen: Zeggen de data uit de onderzochte buurt (uit
etnografie, interviews, overheidsinformatie, veldexperimenten of natuurlijke experimenten) wel
iets over de manier waarop sancties in andere gebieden zouden werken? In hoeverre zeggen
laboratoriumdata over illegaal telefoongebruik bij het beantwoorden van kennisvragen iets over
appen op de fiets in het dagelijks leven buiten het laboratorium?
1
De term validiteit zoals die in deze cursus gehanteerd wordt duidt specifiek op wat binnen de methodenleer de
interne validiteit genoemd wordt, en omvat niet de externe validiteit (= generaliseerbaarheid).
7
De causaliteit van empirisch onderzoek duidt op de mate waarin de studie conclusies trekt over
oorzaak en gevolg, en de mate waarin de gebruikte data in staat is om dit echt te kunnen bewijzen.
Echt causaal bewijs over de relatie tussen A en B bestaat alleen bij experimentele studies. Hier
kunnen de onderzoekers het onderzoek zo vormgeven dat deze relatie geïsoleerd kan worden,
terwijl alle andere (mogelijk verstorende) factoren constant worden gehouden. Het laboratorium-
experiment over telefoongebruik is hiervan een voorbeeld: vergelijkbare groepen studenten
verrichten allen dezelfde taak, waarbij het enige verschil de boete is (hoog, laag, of geen). Door het
telefoongebruik tussen deze groepen te vergelijken kan zeer nauwkeurig worden bestudeerd hoe
boetes de naleving van het telefoonverbod beïnvloeden. Om een dergelijke causaliteit te kunnen
aantonen is het essentieel dat de groepen en omstandigheden die in het onderzoek met elkaar
vergeleken worden verder zo vergelijkbaar mogelijk zijn. Zowel bij het veldexperiment met het
verbodsbord als bij het natuurlijke experiment over de handhavingscampagne kun je je
bijvoorbeeld afvragen of de twee buurten vergelijkbaar zijn qua demografie: indien er in een van
de twee buurten bijvoorbeeld meer jongeren wonen (die vaker appen dan ouderen), zou dit de
vergelijking minder zuiver maken, en dus de meting van het effect van de sancties verstoren.
Causaliteit kan (meestal) niet worden vastgesteld bij enquêtes en overheidsdata. Bij deze methoden
ontbreekt de precieze controle over verstorende factoren, zodat alleen samenhang, ofwel correlatie,
kan worden vastgesteld: of er een relatie is tussen B (bijvoorbeeld naleving) en A (bijvoorbeeld
boetes), waar als er meer B is er meer A is of als er minder B is er minder A is, en vice versa (meer
A meer B en minder A minder A), of omgekeerd (meer A minder B en minder A meer B). Een
dergelijke correlatie is echter geen direct bewijs dat A (boetes) leidt tot B (naleving), omdat deze
relatie ook omgekeerd zou kunnen zijn (B leidt tot A: naleving beïnvloedt het aantal boetes), of via
andere factoren kan lopen (A leidt niet tot B of omgekeerd, maar beiden zijn het gevolg van C:
bijvoorbeeld, het aantal controles). Door middel van statistiek kan wel enig inzicht worden
verkregen in de relatie tussen A en B, omdat sommige analyses kunnen controleren voor andere
factoren die hun samenhang kunnen beïnvloeden (in het geval van de relatie tussen boetes en
naleving bijvoorbeeld het aantal controles, of de leeftijd van fietsers en het gedrag van hun
vrienden).
Hoe bestudeer je empirisch onderzoek
Net als het lezen van juridische teksten (zoals bijvoorbeeld jurisprudentie van de Hoge Raad) vergt
het lezen van empirisch onderzoek wat oefening. In het begin zullen empirische teksten qua
opbouw, taalgebruik, terminologie en logica wat vreemd en lastig zijn. Om je te helpen de
empirische studies in deze cursus (en hopelijk ook empirisch werk dat later op je pad komt) te leren
lezen, bieden we hier een aantal handvaten voor het lezen van deze teksten.
Het is belangrijk stukken eerst te bestuderen op de inhoud (wat is de onderzoeksvraag en wat is het
antwoord op die vraag), dan op de gebruikte methoden (hoe wordt de vraag beantwoord) en dan op
de implicaties die de studie heeft voor juridische en maatschappelijke vraagstukken. Zoals je in de
Leidraad van het vak kunt zien betreffen dit drie van de vier leerdoelen van het vak. Het is nodig
dat je elk artikel goed kunt bestuderen op deze drie aspecten (inhoud, methoden en implicaties). De
volgorde van bestudering, zoals onder weergegeven is als volgt, en elk wordt onder in meer detail
uitgewerkt:
8
A. Wat is de hoofdvraag?
B. Wat zijn de belangrijkste gebruikte theorieën concepten (en soms hypothesen)?
C. Wat zijn de antwoorden op de hoofdvraag?
D. Welke methoden gebruikt de studie en welke type en selectie van data gebruikt het?
E. Wat zijn de voor- en nadelen van deze methoden in termen van validiteit,
generaliseerbaarheid en causaliteit?
F. Wat zijn de implicaties van die antwoorden voor wetenschap en maatschappij?
A. Hoofdvraag
Om empirisch onderzoek goed te lezen is het ten eerste van belang om te weten welke empirische
vraag het onderzoek wil beantwoorden en waarom deze vraag relevant is. Bij veel studies kan je
de hoofdvraag ofwel in de titel, het abstract (samenvatting) bovenaan (als de studie die heeft) of in
de inleiding vinden. Daar zal je ook vaak informatie vinden over het bredere onderwerp waar deze
vraag relevant voor is. Bij de bestudering van de artikelen van dit vak is het raadzaam om hiervan
aantekeningen te maken: Schrijf duidelijk op wat de hoofdvraag van de studie is.
B. Hoofdconcepten
De volgende stap is om een inzicht te krijgen in wat de belangrijkste begrippen en termen, ofwel
hoofdconcepten, in het onderzoek zijn. Vaak worden die genoemd in de abstract en soms de titel
en verder besproken in de inleiding. Daarnaast is er in kwantitatieve studies (en sommige
kwalitatieve) meestal een theoretische sectie waarin de belangrijkste theorieën en eerdere studies
over het onderwerp besproken worden. Schrijf bij het lezen hiervan op wat de belangrijkste
concepten zijn die worden besproken, en hoe deze omschreven of gedefinieerd worden. Richt je
hierbij op die concepten die direct gerelateerd zijn aan de hoofdvraag (of vragen) van de studie.
Probeer te begrijpen hoe deze concepten gerelateerd zijn aan elkaar en aan de hoofdvraag. Als je
vanuit de tekst niet zeker bent wat bepaalde concepten betekenen, of hoe belangrijk ze zijn voor
het onderzoek, vraag hier dan naar in de werkgroep of tijdens het responsiecollege. In sommige
(met name kwantitatieve) stukken formuleren de auteurs expliciete hypothesen, waarin ze vanuit
de bestaande inzichten mogelijke te verwachte antwoorden op de hoofdvraag beredeneren. Het is
nuttig om deze hypothesen te noteren en ze goed te lezen, en te zien of je ze snapt en begrijpt hoe
ze zich verhouden tot de hoofdvraag. Let op: voor sommige Engelstalige concepten is niet altijd
een eenduidige Nederlandse vertaling, het is dan niet erg om het Engels aan te houden.
C. Antwoorden op de hoofdvraag
Nu je weet wat de hoofdvraag is en wat de hoofdconcepten zijn kan je op zoek gaan naar het
antwoord op de hoofdvraag. In sommige stukken kan je de belangrijkste antwoorden op de
hoofdvraag al meteen in het abstract of in de inleiding lezen. Maar om dit echt te snappen moet je
meestal verder in het stuk lezen. In kwantitatieve stukken is er vaak een resultatensectie die wordt
gevolgd door een discussie en/of conclusie. De resultatensectie omvat de statistische data en de
analyse daarvan. Er wordt in dit vak niet van je verwacht dat je de statistiek begrijpt. Je moet een
dergelijke resultatensectie dan ook diagonaal lezen, en zoeken naar zinnen die in taal (en niet in
cijfers) uitleggen wat de uitkomsten van de statistische analyse zijn. Soms zal dit maar heel beperkt
inde resultatensectie zijn en indien dat zo is kan je de resultatensectie zelf overslaan en gelijk naar
de discussiesectie gaan. Wat de resultaten betekenen voor de hoofdvraag wordt meestal samengevat
in de discussie. Hier vind je ook vaak of de hypothesen of verwachtingen bevestigd worden of niet.
9
Maak aantekening van wat deze antwoorden zijn, en noteer welke hypothesen wel en niet bevestigd
worden. In kwalitatief onderzoek volgt na de inleiding en een theoretische beschouwing vaak een
langere bespreking van de empirische bevindingen, waarin ook de antwoorden op de hoofdvraag
naar voren komen. Vaak worden die vervolgens ook samengevat in de conclusie van het stuk.
D. Gebruikte methoden
Als het goed is heb je nu een inzicht in wat de antwoorden zijn die het onderzoek op de hoofdvraag
geeft. De volgende stap is om in de studie op te zoeken op welke manier men tot deze antwoorden
is gekomen. Om dit te doen zoek je in de studie naar een sectie waar de onderzoeksmethoden
worden uitgelegd. In kwantitatieve stukken wordt vaak al in het abstract of de inleiding omschreven
welk soort methoden er gebruikt wordt. Dit wordt vaak verder uitgewerkt in een aparte methode-
sectie. Bij kwalitatief onderzoek worden de methoden vaak (soms wat summier) omschreven in de
inleiding, en soms ook in een aparte methodesectie. Ook kan je soms uit de omschrijving van de
kwalitatieve data aspecten van de gebruikte methode afleiden.
Als je eenmaal gevonden hebt waar in het stuk de methoden omschreven staan, maak dan
aantekeningen over welke methode(n) er in het onderzoek gebruikt zijn (zie Tabel 1), en welke
data er gebruikt worden. Als het gaat om data die gedurende het onderzoek verzameld wordt over
bepaalde personen (zoals bijvoorbeeld bij etnografie, interviews, enquêtes of experimenten),
probeer dan vast te stellen wat de groep van onderzochte personen is, en hoe deze is geselecteerd.
Als het om andere data gaat (zoals reeds bestaande gegevens van de overheid), probeer dan te
kijken wat deze omvatten (zoals welk type van gedragingen, welke jurisdictie, welke jaren, etc.).
E. Voor- en nadelen (inclusief beperkingen) van de gebruikte methoden

Om de voor- en nadelen, en ook de beperkingen, van de onderzochte methoden te bestuderen dien
je je te richten op de drie aspecten die we boven hebben genoemd: validiteit, generaliseerbaarheid
en causaliteit. In Appendix 2 vind je een overzicht van het soort vragen dat je hierover bij elk type
studie kunt stellen. We leggen hieronder in algemenere zin uit waar je bij elk van deze aspecten op
moet letten.
Bij het bestuderen van de validiteit van het onderzoek kijk je naar of de empirische data die is
vergaard een valide en realistisch beeld geeft van het onderwerp dat onderzocht wordt. Als het
onderzoek bestaat uit antwoorden op gestelde vragen (in interviews of enquêtes) dan is een
belangrijke vraag in hoeverre de onderzochte personen in staat of bereid zijn om deze vragen te
beantwoorden. Denk hierbij aan of ze wel de juiste kennis hebben om deze vragen te beantwoorden,
en of de vragen niet over gevoelige kwesties gaan waarover zij mogelijk sociaalwenselijke (te
rooskleurige) antwoorden zullen geven. Bij data die bestaat uit geobserveerd of gerapporteerd
gedrag (zoals etnografie, overheidsdata, of experimenten) ligt validiteit in de mate waarin deze data
een nauwkeurig beeld geven van het gedrag zoals dat zich in de werkelijkheid afspeelt. Zo zullen
criminaliteitscijfers van de politie zelden een compleet beeld geven van alle misdaden en
overtredingen die er plaatsvinden, omdat maar een beperkt gedeelte betrapt of aangegeven wordt.
En bij etnografie of experimenten kan het gedrag van de onderzochte personen beïnvloed worden
door het feit dat ze weten dat ze bestudeerd worden, en zich daarom anders gaan gedragen dan
wanneer zij niet geobserveerd zouden worden.
10
Bij het bestuderen van de generaliseerbaarheid van het onderzoek kijk je naar de mate waarin wat
het onderzoek laat zien bij de onderzochte personen, gedragingen, en omstandigheden ook iets kan
zeggen over andere relevante situaties, buiten die direct zijn onderzocht. Om dit te bepalen moet je
bekijken over welke andere personen, situaties, en gedragingen de auteurs iets proberen te zeggen
op basis van hun onderzoek (bijvoorbeeld over mensen in het algemeen; over andere omgevingen
of landen; over andere typen overtredingen, enzovoort). Vervolgens stel je de vraag in hoeverre het
onderzochte daarmee vergelijkbaar is. De eerste vraag die je hierbij kan stellen is of de onderzochte
groep personen een goede afspiegeling vormt van de bredere populatie waar de studie iets over wil
zeggen. Zo zal een survey onder studenten aan de UvA over hun telefoongebruik tijdens het fietsen
maar zeer beperkt te generaliseren zijn naar de Amsterdamse bevolking, laat staan naar alle
Nederlanders of mensen in het algemeen. De tweede vraag die je hierbij kan stellen is of de
onderzochte gedragingen voldoende lijken op de gedragingen in de werkelijkheid waar de studie
iets over wil zeggen. Zo zal telefoongebruik tijdens een laboratorium-onderzoek maar in zeer
beperkte mate iets zeggen over telefoongebruik op de fiets in het dagelijks verkeer.
Bij het bestuderen van de causaliteit van het onderzoek dien je eerst te bekijken of het onderzoek
causale claims of argumenten maakt. Kijk hierbij in het abstract, de resultaten en conclusies en
zoek naar uitspraken die een oorzakelijk verband tussen de onderzochte concepten suggereren (bv.
beïnvloedt, veroorzaakt, doordat, omdat, komt door). Kijk vervolgens of de studie experimentele
methoden gebruikt waarmee de relatie tussen oorzaak en gevolg onderzocht wordt, door een situatie
waarin de (vermeende) oorzaak (b.v. een boete) aanwezig is te vergelijken met een gelijke situatie
waarin de oorzaak niet aanwezig is (b.v. geen boete). Indien dit het geval is, bekijk dan of de twee
groepen die met elkaar vergeleken worden zo goed mogelijk op elkaar lijken (op alles behalve de
vermeende oorzakelijk). Wanneer het effect van een verbodsbord op appen op de fiets bestudeerd
wordt moet dit bijvoorbeeld gedaan worden in buurten en op tijdstippen en omstandigheden die zo
dicht mogelijk op elkaar lijken. Indien het onderzoek geen experimentele methoden gebruikt kan
het geen echte causaliteit aantonen (en zijn causale claims dus onjuist). Kijk bij dit soort onderzoek
naar hoe de auteurs dan tot hun causale claims en argumenten komen. Een eerste vorm, die je met
name in kwalitatief onderzoek zal vinden, is een verhalende vorm (eerst kwam A en toen kwam B),
waarbij de relatie tussen A en B vanuit een narratief wordt besproken en soms door logisch
redeneren wordt verondersteld. Een onderzoek kan hiermee tot theorieën over mogelijke causale
verbanden komen, maar kan deze met zulke data niet bewijzen. Een tweede vorm, die veel bij
analyses van enquêtes en overheidsdata wordt gebruikt, is om door middel van statistische analyses
de samenhang van twee of meer variabelen te duiden (bijvoorbeeld door correlaties of
regressieanalyse). Zoals eerder al vermeld werd, kunnen causale verbanden met zulke data niet
bewezen worden. Waar je bij dergelijke studies wel op kunt letten is of men bij het bestuderen van
de samenhang van twee variabelen (b.v. de hoogte van de boete en de mate van naleving) ook
rekening houdt met relevante variabelen die deze relatie zouden kunnen beïnvloeden (bijvoorbeeld
de leeftijd van fietsers en de mate van naleving onder hun vrienden). Je kunt dit bestuderen door te
bekijken of dergelijke variabelen (als controlevariabelen) worden meegenomen in de statistische
analyse, zodat hun invloed op de relatie tussen de andere concepten uitgefilterd wordt.
Door een studie op deze drie aspecten te beoordelen kun je de sterke en zwakke punten van haar
methode beoordelen. Dit is ook essentieel om de vragen te kunnen beantwoorden die je hierover
gaat krijgen in de werkgroep en op het tentamen. Het is zaak om deze voor- en nadelen te bezien
vanuit het doel van de studie. Stel dus eerst vast wat die doelen zijn, zie onder A. en bedenk dan
11
als je algemeen hebt vastgesteld wat de voor- en nadelen zijn in hoe verre die het beantwoorden
van de hoofdvragen positief of negatief beïnvloeden.
F. Implicaties van de studie

Tenslotte dien je te bestuderen wat de implicaties van de studie zijn voor relevante
wetenschappelijke en maatschappelijke vragen. Dit doe je op drie manieren. Allereerst kijk je
zowel in het abstract (als dat er is), de inleiding, de discussie (als die er is) en de conclusie wat de
studie zelf zegt over haar juridische en maatschappelijke implicaties. Kijk hierbij naar wat de studie
verwijst (bijvoorbeeld in de introductie) als het bredere probleem waaraan zij wil bijdragen. Kijk
ook hoe de studie zelf de bijdrage van het onderzoek hieraan verwoordt (bijvoorbeeld in de
discussie of conclusie). Welke juridische of maatschappelijke problemen denken de auteurs
hiermee te kunnen oplossen, en hoe? Denk daarnaast zelf na voor welke andere kwesties deze studie
relevant zou kunnen zijn. Zijn er andere juridische of maatschappelijke kwesties waar deze studie
informatief voor zou kunnen zijn? Schrijf beiden (de implicaties die de studie zelf vermeldt en die
die je zelf bedacht hebt) kort op.
Betrek vervolgens je analyse van de voor- en nadelen van de gebruikte methoden (zie punt E
hierboven) in deze overweging. Geeft de studie ook een voldoende betrouwbaar, generaliseerbaar
en causaal beeld om bij te kunnen dragen aan deze kwesties? Studies die lager scoren op validiteit
hebben wellicht minder bij te dragen aan het begrip van het probleem zoals het in de werkelijkheid
bestaat. Studies met een lagere generaliseerbaarheid kunnen relevante implicaties opleveren voor
de specifieke situatie die onderzocht is, maar deze zijn mogelijk minder goed bruikbaar buiten deze
context. Studies die geen causale claims maken, of die hun causale claims niet deugdelijk
onderbouwen, zijn minder geschikt om de oorzaken van juridische en maatschappelijke problemen
te identificeren. Ze zijn bovendien ook minder bruikbaar als basis voor interventies om zulke
problemen op te lossen, omdat ze niet met zekerheid kunnen vaststellen dat de vermeende oorzaak
ook werkelijk het verwachte gevolg zal hebben.
12
Appendix 1: Voor- en nadelen (en beperkingen) van verschillende typen empirisch juridisch
onderzoek
Type Voordelen Nadelen en Beperkingen

Bewerkelijk qua uitvoering en
analyse. Dit beperkt observaties vaak
tot een gering aantal contexten. Niet
Biedt een nauwkeurig beeld van geschikt voor theorietoetsing. Niet
complexe realiteit en kan zowel geschikt voor statistische analyse.
Participerende
processen (hoe zaken gebeuren) als Leidt tot complexe informatie die niet
observatie
perspectieven daarover blootleggen. altijd goed te vertalen is naar beleid.
Geschikt voor theorievorming. En ook een risico dat de onderzoeker
zich te veel gaat vereenzelvigen met
de onderzochte groep aanmeet (going
native).
Biedt een goed inzicht in de
Redelijk bewerkelijk qua uitvoering.
opvattingen en perspectieven van de
Biedt een subjectief perspectief, kan
geïnterviewde personen alsook in hoe
leiden tot sociaalwenselijke
zij denken dat processen in de
antwoorden, en geïnterviewden
werkelijkheid plaatsvinden. Kan onder
hebben niet altijd de kennis/capaciteit
een diverse groep gedaan worden.
Interviews om de vragen te beantwoorden.
Doordat vertrouwen met de
Minder geschikt voor theorietoetsing.
geïnterviewden kan worden
Minder geschikt voor statistische
ontwikkeld en door vergelijking van
analyse. Kan leiden tot complexe
antwoorden kan de validiteit van de
informatie die niet altijd goed te
antwoorden worden vergroot.
vertalen is naar beleid.
Geschikt voor theorievorming.
Biedt een subjectief perspectief, kan
leiden tot sociaalwenselijke
Kan een zeer groot aantal respondenten antwoorden, en respondenten hebben
bereiken in precies de relevante niet altijd de kennis/capaciteit om de
doelgroepen. Biedt een inzicht in de vragen te beantwoorden. Kan ook
perspectieven en ervaring van deze leiden tot oppervlakkige antwoorden
Enquêtes respondenten. Kan een redelijk groot of foutieve antwoorden doordat
aantal variabelen meenemen. Geschikt respondenten te snel en oppervlakkig
voor statistische inferentie. Kan een lezen en onzorgvuldig antwoorden.
redelijk eenvoudig beeld geven dat Minder geschikt voor inductieve
beleid kan ondersteunen. theorievorming. Biedt een beeld van
de correlatie van variabelen niet de
precieze causaliteit.
Indien beschikbaar en toegankelijk, Geeft vaak geen informatie op
kan het de mogelijkheid geven tot individueel niveau of over
analyse van een grote hoeveelheid individuele perspectieven. Geeft
databronnen over specifieke alleen informatie over variabelen
Analyse van maatschappelijke fenomenen. waar overheidsdata op bestaan.
overheidsdata Weergave van (meestal) zeer groot Kunnen vertekeningen bevatten (wat
aantal personen of gebeurtenissen. wel en niet wordt opgenomen in de
Worden vaak periodiek herhaald of data en wat wel en niet beschikbaar
bijgewerkt waardoor data over langere is). Wordt vaak voor andere doelen
periode beschikbaar is. Leent zich voor verzameld dan de onderzoeker voor
13
statistische inferentie. Kan een lange ogen heeft en sluit daarom niet altijd
termijn beeld geven. Leidt tot precies aan.
conclusies die bruikbaar kunnen zijn
voor beleid. Geschikt voor
theorietoetsing.
Past vaak niet goed bij de
werkelijkheid buiten het
Kan zeer precies het effect van een
laboratorium. Kan
variabele of interventie vaststellen.
kortetermijneffecten laten zien die
Controleert voor de invloed van andere
Lab- niet zouden bestaan indien de
factoren die in de werkelijkheid hier
Experimenten interventie breed wordt uitgezet in
niet van kunnen worden losgekoppeld.
beleid. Kan maar een beperkt aantal
Leidt tot eenvoudige conclusies voor
variabelen meenemen. Is vaak weinig
beleid. Geschikt voor theorietoetsing.
informatief over
langetermijneffecten.
Door de nauwe focus op de
gemanipuleerde en gemeten
variabelen is er maar beperkte
Kan redelijk precies het effect van een aandacht voor de invloed van de
variabele of interventie vaststellen in context waarin het fenomeen
Veld- de context waarin het fenomeen in de voorkomt. Is maar beperkt praktisch
Experimenten werkelijkheid voorkomt. Leidt tot en ethisch toepasbaar. Kan
redelijk eenvoudige conclusies voor kortetermijneffecten laten zien die
beleid. Geschikt voor theorietoetsing. niet zouden bestaan indien de
interventie breed wordt uitgezet in
beleid. Kan maar een beperkt aantal
variabelen meenemen.
Door de focus op de gemeten
variabelen is er maar weinig aandacht
voor de invloed van de context waarin
het fenomeen voorkomt.
Kan tamelijk goed het effect van een
Vergelijking tussen situaties is
variabele of interventie vaststellen in
moeilijk vanwege mogelijke
Natuurlijke de context waarin het fenomeen in de
verschillen in tijd (voor vs na), beleid
experimenten werkelijkheid voorkomt. Leidt tot
of andere relevante contextuele
redelijk eenvoudige conclusies voor
factoren. Wordt beperkt door de
beleid. Geschikt voor theorietoetsing.
beschikbare data en keuzes in welke
data wel en niet worden
meegenomen. Kan maar een beperkt
aantal variabelen meenemen.
14
Appendix 2: Vragen over kwaliteit van methoden voor verschillende typen onderzoek
Validiteit Generaliseerbaarheid Causaliteit

Participerende Wat is precies geobserveerd? Over welk bredere Maakt de studie causale
observatie In hoeverre heeft de vraagstuk pretendeert de claims (waar er impliciet
onderzoeker echt kunnen studie iets te zeggen? In of expliciet een relatie
observeren in de dagelijks hoeverre biedt de tussen oorzaak en gevolg
werkelijkheid? Hoe heeft de vergaarde data een goed gelegd wordt)? Op welke
onderzoeker keuzes gemaakt beeld of voorbeeld van dit wijze worden deze
wat wel en niet te observeren bredere vraagstuk? Op causale claims door de
en wat wel en niet te welke wijze poogt de data en de analyse
analyseren en beschrijven? In studie te generaliseren daarvan ondersteund? In
hoeverre gaat de studie zelf in van de eigen data naar dit hoeverre gaat de studie
op de limieten in haar bredere thema? In zelf in op haar limieten in
validiteit? hoeverre gaat de studie causaliteit?
zelf in op de limieten in
deze generalisatie?
Interviews Wie wordt waarover Over welk bredere Maakt de studie causale
geïnterviewd? In hoeverre vraagstuk pretendeert de claims (waar er impliciet
hebben de geïnterviewden de studie iets te zeggen? In of expliciet een relatie
kennis en bereidheid om hoeverre biedt de tussen oorzaak en gevolg
antwoord te geven op de vergaarde data een goed gelegd wordt)? Op welke
gestelde vragen? In hoeverre beeld of voorbeeld van dit wijze worden deze
heeft de onderzoeker bredere vraagstuk? Op causale claims door de
voldoende vertrouwen kunnen welke wijze poogt de data en de analyse
wekken om gevoelige vragen studie te generaliseren daarvan ondersteund? In
te stellen? In hoeverre heeft de van de eigen data naar dit hoeverre gaat de studie
onderzoeker vragen zo bredere thema? In zelf in op haar limieten in
opgesteld dat ze geen hoeverre gaat de studie causaliteit?
sociaalwenselijke antwoorden zelf in op de limieten in
uitlokken? In hoeverre gaat de deze generalisatie?
studie zelf in op de limieten in Hoeveel respondenten
haar validiteit? zijn geïnterviewd? Wat is
er bekend over hoe
representatief deze groep
is?
Surveys Wie wordt waarover Over welk bredere Maakt de studie causale
ondervraagd? In hoeverre vraagstuk pretendeert de claims (waar er impliciet
hebben de respondenten de studie iets te zeggen? In of expliciet een relatie
kennis en bereidheid om hoeverre biedt de tussen oorzaak en gevolg
antwoord te geven op de vergaarde data een goed gelegd wordt)? Op welke
gestelde vragen? In hoeverre beeld of voorbeeld van dit wijze worden deze
heeft de onderzoeker vragen bredere vraagstuk? Op causale claims door de
zo opgesteld dat ze geen welke wijze poogt de data en de analyse
sociaalwenselijke antwoorden studie te generaliseren daarvan ondersteund?
uitlokken? In hoeverre gaat de van de eigen data naar dit Maakt de studie gebruik
studie zelf in op de limieten in bredere thema? In van statistische methoden
haar validiteit? hoeverre gaat de studie om te controleren voor
zelf in op de limieten in relevante variabelen die
deze generalisatie? de relatie tussen de
Idem respons en hoofdvariabelen kunnen
representativiteit beïnvloeden? In hoeverre
gaat de studie zelf in op
haar limieten in
causaliteit?
15
Analyse van Welke overheidsdata wordt Over welk bredere Maakt de studie causale
overheidsdata precies gebruikt? Voor welk vraagstuk pretendeert de claims (waar er impliciet
doel worden die verzameld? In studie iets te zeggen? In of expliciet een relatie
hoeverre is bekend of deze hoeverre biedt de tussen oorzaak en gevolg
data een compleet en getrouw vergaarde data een goed gelegd wordt)? Op welke
beeld geven van wat ze meten? beeld of voorbeeld van dit wijze worden deze
In hoeverre is deze data bredere vraagstuk? Op causale claims door de
compleet? In hoeverre gaat de welke wijze poogt de data en de analyse
studie zelf in op de limieten in studie te generaliseren daarvan ondersteund?
haar validiteit? van de eigen data naar dit Maakt de studie gebruik
bredere thema? In van statistische methoden
hoeverre gaat de studie om te controleren voor
zelf in op de limieten in relevante variabelen die
deze generalisatie? de relatie tussen de
hoofdvariabelen kunnen
beïnvloeden? In hoeverre
gaat de studie zelf in op
haar limieten in
causaliteit?
Laboratorium- Wat wordt precies gemeten in Over welk bredere Maakt de studie causale
Experimenten het experiment? In hoeverre vraagstuk pretendeert de claims (waar er impliciet
zijn de onderzochte personen studie iets te zeggen? In of expliciet een relatie
in staat en bereid op hoeverre biedt de tussen oorzaak en gevolg
deugdelijke wijze aan het vergaarde data een goed gelegd wordt)?In
experiment deel nemen? In beeld of voorbeeld van dit hoeverre heeft het
hoeverre lokt het experiment bredere vraagstuk? In experiment echt twee of
sociaalwenselijk gedrag uit? In hoeverre zijn de meer vergelijkbare
hoeverre geeft het gedrag van bevindingen uit het lab groepen die in alles gelijk
de deelnemers hetgeen weer ook vertaalbaar naar de zijn behalve de
dat de onderzoekers daarover werkelijkheid buiten het kerninterventievariabele?
veronderstellen? In hoeverre lab? Op welke wijze In hoeverre wordt
zou de data anders zijn als het poogt de studie te gecontroleerd op
experiment langer was generaliseren van de relevante verschillen
doorgegaan en de eigen data naar dit tussen de bestudeerde
onderzochten gewend zouden bredere thema? In groepen? In hoeverre is er
zijn geraakt aan de hoeverre gaat de studie een echte controlegroep?
bestudeerde interventie? In zelf in op de limieten in In hoeverre gaat de studie
hoeverre gaat de studie zelf in deze generalisatie? zelf in op haar limieten in
op de limieten in haar causaliteit?
validiteit?
Veldexperimenten Wat wordt precies gemeten in Over welk bredere Maakt de studie causale
het experiment? In hoeverre vraagstuk pretendeert de claims (waar er impliciet
gedragen de mensen zich in studie iets te zeggen? In of expliciet een relatie
het experiment ook zoals de hoeverre biedt de tussen oorzaak en gevolg
onderzoekers denken dat ze vergaarde data een goed gelegd wordt)? In
zich gedragen? In hoeverre beeld of voorbeeld van dit hoeverre heeft het
wordt het gedrag in het veld bredere vraagstuk? In experiment echt twee of
beïnvloed door de hoeverre zijn de meer vergelijkbare
onderzoekers? In hoeverre zou bevindingen uit de groepen die in alles gelijk
de data anders zijn als het veldstudie ook zijn behalve de
experiment langer was vertaalbaar naar de kerninterventievariabele?
doorgegaan en de werkelijkheid buiten deze In hoeverre wordt
onderzochten gewend zouden studie? Op welke wijze gecontroleerd op
zijn geraakt aan de poogt de studie te relevante verschillen
bestudeerde interventie? In generaliseren van de tussen de bestudeerde
hoeverre gaat de studie zelf in eigen data naar dit groepen? In hoeverre is er
op de limieten in haar bredere thema? In een echte controlegroep?
validiteit? hoeverre gaat de studie In hoeverre gaat de studie
zelf in op de limieten in zelf in op haar limieten in
deze generalisatie? causaliteit?
16
Natuurlijke Welke data wordt precies Over welk bredere Maakt de studie causale
experimenten gebruikt om de situatie voor en vraagstuk pretendeert de claims (waar er impliciet
na de studie iets te zeggen? In of expliciet een relatie
interventie/kernverandering te hoeverre biedt de tussen oorzaak en gevolg
meten? In hoeverre is bekend vergaarde data een goed gelegd wordt)? In
of deze data een compleet en beeld of voorbeeld van dit hoeverre heeft het
getrouw beeld geven van wat bredere vraagstuk? In experiment echt twee of
ze meten? In hoeverre gaat de hoeverre zijn de meer vergelijkbare
studie zelf in op de limieten in bevindingen uit de groepen die in alles gelijk
haar validiteit? veldstudie ook zijn behalve de
vertaalbaar naar de kerninterventievariabele?
werkelijkheid buiten deze In hoeverre wordt
studie? Op welke wijze gecontroleerd op
poogt de studie te relevante verschillen
generaliseren van de tussen de bestudeerde
eigen data naar dit groepen? In hoeverre is er
bredere thema? In een echte controlegroep?
hoeverre gaat de studie In hoeverre gaat de studie
zelf in op de limieten in zelf in op haar limieten in
deze generalisatie? causaliteit?
17
18
bs_bs_banner
14682230, 2017, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/1468-2230.12290 by Uva Universiteitsbibliotheek, Wiley Online Library on [28/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Wrong about Rights: Public Knowledge of Key Areas
of Consumer, Housing and Employment Law in
England and Wales
Pascoe Pleasence,∗ Nigel J. Balmer,∗∗ and Catrina Denvir∗∗∗
Over many decades, processes of juridification have brought about huge growth in legal rights,
responsibilities and protections, yet citizens appear to poorly understand this ‘law thick’ world.
This impacts citizens’ capacity to ‘name, blame and claim’ in the legal domain at a time of retreat
from public funding of civil legal services. This article examines public knowledge of rights in
key areas relating to consumer, housing and employment law. Drawing on data from the 2010–
2012 English and Welsh Civil and Social Justice Survey, the article uses responses to a series of
hypothetical scenarios to explore public knowledge of rights and characteristics associated with
knowledge. Our findings highlight a substantial deficit in individuals’ understanding of legal
rights and responsibilities – even among those for whom particular rights and responsibilities
have specific bearing. We also consider what these findings mean for public legal education and
the efficiency, efficacy and legitimacy of the law.
INTRODUCTION
We live in a ‘law-thick’ world1 in which our daily activities are played

out within a complex and extensive legal framework. Successive waves of
‘juridification’ – defined by Habermas as ‘the tendency towards an increase in
formal (or positive, written) law that can be observed in a modern society’2
– have seen law become ‘ubiquitous in social life’.3 An increasing range of
legal rights, responsibilities and protections have come to apply in the spheres
of education, employment, children and families, health, housing, welfare
∗
Professor Empirical Legal Studies, Faculty of Laws, University College London.
∗∗
Reader in Law and Social Statistics, Faculty of Laws, University College London.
∗∗∗
Director of the Centre for Legal Innovation, Ulster University. We are very grateful to the Legal
Education Foundation for funding this work and to the 3,911 respondents who gave their time to
take part in our survey.
1 G. K. Hadfield, ‘Higher Demand, Lower Supply? A Comparative Assessment of the Legal
Resource Landscape for Ordinary Americans’ (2010) 37 Fordham Urban Law Journal 129, 133.
2 J. Habermas, The Theory of Communicative Action, Volume 2: Lifeworld and System (Cambridge:
Polity Press, 1987) 357. He then went on to distinguish between ‘the expansion of law, that is the
legal regulation of new, hitherto informally regulated social matters, from the increasing density
of law, that is the specialised breakdown of global statements of the legally relevant facts into more
detailed statements.’ Habermas argued that there have been four epochs of juridification, with the
most recent accompanying the expansion of the welfare state. Veitch et al (2007) argue that we
are now in the midst of a fifth epoch, accompanying the decline of the welfare state, comprising
‘the re-embedding of private law mechanisms – contract and property law in particular – within
formerly public, State-owned areas’ and the expansion of supranational law. See further, S. Veitch,
E. Christodoulidis and L. Farmer Jurisprudence: Themes and Concepts (Oxford: Routledge, 2012).
3 W. Twining, Blackstone’s Tower: The English Law School (London: Sweet and Maxwell, 1994) 16.
"
C 2017 The Author. The Modern Law Review "
C 2017 The Modern Law Review Limited. (2017) 80(5) MLR 836–859
Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
19
Pascoe Pleasence, Nigel J. Balmer and Catrina Denvir
benefits, consumer goods and services, and our environment.4 Yet, it appears
that, at least in some contexts, ‘the vast majority of the population systemati-
cally mispredicts . . . the content of the law’.5 A steadily growing number of
studies of the public’s understanding of law point to a substantial knowledge
deficit.6
The deficit appears greater in some areas of law than others, in part a
function of salience.7 After all, there is less reason for individuals to possess
knowledge with no clear bearing on their lives. So, for example, Casebourne et
al’s study of employee awareness of employment rights in the United Kingdom
demonstrated that those with dependent children were ‘understandably’ more
likely than others to know a lot or fair amount about the detail of the parental
right to request flexible working (40 per cent vs 27 per cent).8 There is also
less reason for people to have a grasp of the intricacies of the law than broader
legal principles, and again this is reflected in findings to date.9
Furthermore, there is evidence of a lesser deficit for those with an interest
in or, more predictably, relevant experience of the law. So, Baker and Emery10
found that students just about to start a course on family law had slightly
4 See, for example, G. Teubner, Juridification of Social Spheres: A Comparative Analysis in the Areas of
Labour, Corporate, Antitrust and Social Welfare Law (Berlin: Walter de Gruyter, 1987); G. Howells,
and S. Weatherill, Consumer Protection Law (Aldershot: Ashgate, 2nd ed, 2005); T. Goriely,
‘Making the Welfare State Work: Changing Conceptions of Legal Remedies within the British
Welfare State’ in F. Regan, A. Paterson, T. Goriely and D. Fleming (eds), The Transformation of
Legal Aid (Oxford: OUP, 1998); H. Gibson, ‘Home-School Agreements: explaining the growth
of “juridification” and contractualism in schools’ (2013) 39 Oxford Review of Education 780; K.
Veitch, ‘Juridification, Medicalisation, and the Impact of EU Law: Patient Mobility and the
Allocation of Scarce NHS Resources’ (2012) 20 Medical Law Review 362; Veitch, Christodoulidis
and Farmer, n 2 above.
5 K. Williams, Litigants in Person: A Literature Review, (London: Ministry of Justice, 2011).
6 See, for example, C. F. Cortese, ‘A Study in Knowledge and Attitudes Towards the Law’ (1966)
2 Rocky Mountain Social Science Journal 192; M. Williams and J. Hall, ‘Knowledge of the Law
in Texas: Socioeconomic and Ethnic Differences’ (1972) 7 Law and Society Review 99; L. V.
E. Saunders, ‘Collective Ignorance: Public Knowledge of Family Law’ (1975) 24 The Family
Coordinator 69; P. T. Kim, ‘Norms, Learning, and Law: Exploring the Influences on Workers’
Legal Knowledge’ (1999) 2 University of Illinois Law Review 447; J. M. Darley, C. A. Sanderson and
P. S. LaMantia, ‘Community Standards for Defining Attempt: Inconsistencies with the Model
Penal Code’ (1996) 39 American Behavioral Scientist 405; A. Barlow, S. Duncan, G. James and
A. Park, Cohabitation, Marriage and the Law: Social Change and Legal Reform in the 21st Century
(Oxford: Hart, 2005); M. Militello, D. Schimmel and H. J. Eberwein, ‘If They Knew, They
Would Change’ (2009) 93 NASSP Bulletin 27; R. Panades, R. Corney, C. Ayles, J. Reynolds and
F. Hovsepian, Informing Unmarried Parents About their Legal Rights at Birth Registration (London:
One Plus One, 2007); L. J. Parle, Measuring young people’s legal capability (London: Independent
Academic Research Studies and PLEnet, 2009); P. Pleasence and N. J. Balmer, ‘Ignorance in
Bliss: Modeling Knowledge of Rights in Marriage and Cohabitation’ (2012) 46 Law and Society
Review 297; C. Denvir, N. J. Balmer, and P. Pleasence, ‘When legal rights are not a reality:
do individuals know their rights and how can we tell?’ (2013) 35 Journal of Social Welfare and
Family Law 139; L. A. Baker and R. E. Emery, ‘When Every Relationship is Above Average –
Perceptions and Expectations of Divorce at the Time of Marriage’ (1993) 17 Law and Human
Behavior 439.
7 See, for example, Saunders, ibid; J. Casebourne, J. Regan, F. Neathey and S. Tuohy, Employment
Rights at Work: A Survey of Employees 2005 (London: Department of Trade and Industry, 2006);
Pleasence and Balmer ibid.
8 Casebourne et al, ibid.
9 ibid.
10 Baker and Emory, n 6 above.
"
C 2017 The Modern Law Review Limited.
(2017) 80(5) MLR 836–859 837
20
Wrong about Rights
higher levels of knowledge of divorce statute law than did the general public.11
As would be expected, those who had completed the course had higher levels
of knowledge still; although Baker and Emery found that, ‘in absolute terms
. . . even these students’ perceptions were highly inaccurate’.12,13
Beyond the substance of the law, Williams’s14 review of research into litigants
in person found ‘most research’ to suggest that litigants in person experience a
variety of knowledge related disadvantages; having trouble ‘identifying facts rel-
evant to the case’, ‘understanding evidential requirements’ and ‘understanding
the nature of proceedings’.
All this is of concern, as poor understanding of law and process may pre-
vent people from acting to protect their rights (or discharge responsibilities),15
prevent people acting to protect against the likelihood of particular eventual-
ities16,17 and militate against good outcomes.18 Williams’s review also points
to the poor knowledge of litigants in person creating an ‘extra burden . . . for
court staff and judges’.
Evidently, how people understand the world impacts on behaviour. For
example, Felstiner, Abel and Sarat’s influential framework for understanding the
emergence and transformation of disputes sets out how ‘injurious experiences’
only become transformed into disputes through being recognised (‘naming’),
attributed to another (‘blaming’), and communicated to that other along with
a request ‘for some remedy’ (‘claiming’) that is refused.19 Whether law is
properly understood directly links to whether there is legal framing (as opposed
to, say, a moral or social framing), and what form it takes, in this process
of naming, blaming and claiming. And, as Pleasence, Balmer and Reimers
have shown,20 legal framing is a significant influence on people’s behaviour
11 Answering 64% of questions correctly, as compared to 60%.

12 Baker and Emory, n 6 above, 445.
13 Their scores averaged 70%.
14 Williams, n 5 above.
15 See, for example, P. Bowal, ‘A Study Of Lay Knowledge Of Law In Canada’ (1999) 9 Indiana
International And Comparative Law Review 121; P. Pleasence and N. J. Balmer, How People Resolve
‘Legal’ Problems (London: Legal Services Board, 2014).
16 Kim, n 6 above; N. Meager, Awareness, Knowledge and Exercise of Individual Employment Rights
(London: Department of Trade and Industry, 2002).
17 For example, in the context of the ‘common law marriage myth’, which involves the systematic
over-estimation of the rights of cohabitees against one another, whereas divorce acts as a ‘redis-
tributive’ process, ‘financial division on the breakdown from cohabitation sustains the financial
power dynamics in the relationship’ (J. Lewis, R. Tennant and J. Taylor, ‘Financial Arrangements
on the Breakdown of Cohabitation: Influences and Disadvantage’ in J. Miles and R. Probert
(eds), Sharing Lives, Dividing Assets: An Inter-Disciplinary Study (Oxford: Hart, 2009) 179). Thus,
poor legal knowledge will lead to some financially vulnerable cohabitants to ‘underinsure against
the financial cost’ of relationship breakdown (W. S. Hannon ‘Sticky expectations: Responses
to persistent over-optimism in marriage, employment contracts and credit card use’ (2009) 84
Notre Dame Law Review 759), ‘militate against amicable resolution of disputes’ (S. Dowding,
‘Self-Determination or Judicial Imposition?’ in Miles and Probert (eds), ibid 207), and create
‘needless emotional harm’ (Williams, n 5 above, 735).
18 Pleasence and Balmer, n 15 above.
19 W. L. F. Felstiner, R. L. Abel and A. Sarat, ‘The emergence and transformation of disputes:
naming, blaming, claiming . . . .’ (1981) 15 Law and Society Review 635.
20 P. Pleasence, N. J. Balmer and S. Reimers, ‘What really drives advice seeking behaviour? Looking
beyond the subject of legal disputes’ (2011) 1 Onati Socio-Legal Series 6.
"
838 (2017) 80(5) MLR 836–859
21
in resolving ‘justiciable’21 problems; notably in relation to whether advice is

sought from lawyers, who Felstiner, Abel and Sarat described as ‘the most
important . . . agents of dispute resolution . . . [as a] result of the lawyer’s
central role as gatekeeper to legal institutions and facilitator of a wide range of
personal and economic transactions’.22
In the United Kingdom, both traditional legal services and the wider advice
sector play critical roles in enabling citizens to understand the many dimensions
of everyday justiciable problems, and to facilitate legal framing and action if
appropriate. Thus, current concerns about poor public legal literacy are exac-
erbated by concerns about the retreat of public expenditure on legal services,
including basic forms of legal help. For example, the net civil legal aid budget
in England and Wales fell by 15 per cent (£133 million) between 2012–13 and
2015–16,23 following implementation of the Legal Aid, Sentencing and Pun-
ishment of Offenders (LASPO) Act 2012;24 placing increasing reliance on the
abilities of individual citizens to understand and navigate the law-thick world
alone. This significant challenge was recognised in the Public Bill Committee
debates regarding the passage of the LASPO Bill, where it was observed that
changes to legal aid demanded greater public knowledge of rights if individuals
are expected to self-help and self-represent more often.25
BELIEFS ABOUT LAW, ATTITUDES AND SOCIAL NORMS
Importantly, holding erroneous beliefs about the law is not simply a matter of
chance ignorance. We have previously argued that legal reality and the public’s
perception of legality – both in the case of cohabitation and marriage – are
each coherent and distinct, with the latter fuelled and entrenched by attitudes
and social norms. As Lewis et al have observed, ‘people’s perceptions of their
rights and duties are learned in a social context’.26
In the employment context, Kim found that workers’ beliefs were ‘system-
atically erroneous’, yet ‘remarkably similar’ between US states, ‘despite wide
variations in the states’ laws’.27 She concluded that respondents ‘assumed that
the requirements of the law coincide with their beliefs as to how employers
should behave and, therefore, answer the legal questions according to their own
notions of fairness’.28
21 A 15th century word, notably defined by Genn, as a matter that raises legal issues, whether or not
these are recognised as being legal and whether or not any action taken to deal with the matter
involves the use of any part of the civil justice system (H. Genn, Paths to justice: What people do
and think about going to law, (Oxford: Hart, 1999)).
22 Felstiner, Abel and Sarat, n 19 above, 645.
23 Legal Aid Agency, Annual Report and Accounts (London: HMSO, 2014, 2016).
24 Legal Aid, Sentencing and Punishment of Offenders (LASPO) Act 2012.
25 Public Bill Committee 2011, Public Bill Committee Transcript of Discussions on the
Legal Aid and Sentencing of Offenders Act at http://services.parliament.uk/bills/2010-
11/legalaidsentencingandpunishmentofoffenders/stages.html (last accessed 28 October 2016).
26 Lewis, Tennant and Taylor, n 17 above.
27 Kim, n 6 above, 447 and 45.
28 ibid, 490.
"
(2017) 80(5) MLR 836–859 839
22
Wrong about Rights
Similarly, Darley et al’s study of people’s understanding of aspects of criminal

law in Wisconsin, Texas, South Dakota, and North Dakota pointed to little
variation between states in citizens’ beliefs about the law, despite substantial
differences in the actual law.29 Drawing on the ‘false consensus effect’30 they
concluded that people ‘assume that the state, in its moral wisdom, shares their
personal views’.31 Even in the case of Texans’ relatively good understanding
of Texas’s irregular law concerning the use of deadly force to protect property,
differences in beliefs about the state of the law ‘disappeared completely when
the relevant attitudes of the citizens [were] covaried out’.32
Further findings from the English and Welsh Civil and Social Justice Panel
Survey (CSJPS – data from which is also used for the current study) relat-
ing to a series of questions exploring a hypothetical family scenario made
evident
symmetry of error in people’s beliefs about marriage and cohabitation law, where
beliefs about both cohabitation and marriage law err from their (often opposing)
correct legal positions to rest more closely in line with social attitudes.33
People appeared to tend to assume the law concurred with what they thought
it ought to be. For example, in the case of a childless couple who had lived
together for 10 years, with the female partner having looked after the home
and not worked since they started living together and the male partner having
a good salary and sizeable savings, 52 per cent of respondents wrongly believed
that a financially dependent cohabitant would have had a good legal claim,
with 65 per cent believing the same in the case of a spouse, with most of
the remainder wrongly believing that such a spouse would not have a claim.
This reflects the attitudes of the public as indicated by the 2006 British Social
Attitudes Survey more than the actual state of the law.34
BELIEFS ABOUT LAW AND REFERENCE TO TIME
Another aspect of public beliefs about some legal rights is that they are be-
lieved to crystallise over time. So, for example, it has been found that survey
respondents’ beliefs about the rights of spouses and cohabitees against the
estate of an intestate deceased partner depend upon relationship duration.35
So, the percentage of respondents believing a cohabitee would automatically
inherit went from seven per cent for short relationship durations to an asymp-
tote of 27 per cent, with the increase fairly gradual as duration increased. In
29 J. M. Darley, K. M. Carlsmith and P. H. Robinson, ‘The Ex Ante Function of the Criminal Law’
(2001) 35 Law and Society Review 165.
30 L. Ross, D. Greene and P. House, ‘The False Consensus Effect: An Egocentric Bias in Social
Perception and Attribution Processes’ (1977) 13 Journal of Experimental Social Psychology 279.
31 Darley, Carlsmith and Robinson n 29 above, 168.
32 ibid, 178.
33 Pleasance and Balmer, n 6 above.
34 Barlow et al, n 6 above.
"
840 (2017) 80(5) MLR 836–859
23
the case of spouses, the asymptote was 44 per cent. In both these examples
(cohabitees and spouses), the passing of time can be taken to be associated
with the accretion of expectation and dependency on the part of cohab-
itees/spouses, and thus is reflected in a public expectation that the law will
protect the vulnerable. The absence of such a public expectation in the case
of rights linked to parenthood, such as child support and decisions about chil-
dren’s medical treatment, meant that beliefs were independent of relationship
duration.
CHALLENGES TO DISLODGING ERRONEOUS BELIEFS
One consequence of the tendency of beliefs about the law to align with social
attitudes is that, in line with cognitive dissonance theory,36 erroneous beliefs
may prove resistant to being dislodged. Cognitive dissonance theory37 predicts
that when a person’s cognitions (which include knowledge, opinions and be-
liefs) are at-odds, or dissonant, this is ‘psychologically uncomfortable’ and will
motivate the person to ‘reduce the dissonance and achieve consonance’. Cog-
nitive dissonance can therefore lead people to avoid or disregard ‘information
that would likely increase the dissonance’.38
For example, in the employment context, Kim argued that erroneous beliefs
based on attitudes are likely to be ‘resistant to change’, as a fairness norm ‘over-
shadows the influence of most . . . experiential factors’.39 Likewise, Ellickson,
in his study of the law’s place in ranching disputes in Shasta County, California,
found that ‘the cattlemen resist absorbing information that is inconsistent with
their folklore’.40 Thus, even repeated experience of insurance companies and
courts following different principles did not dislodge the belief that, in the
event of road collisions in ‘open range’, ‘the motorist buys the cow’.
The impact of cognitive dissonance is also compounded by common ‘in-
difference to law’ resulting from phenomena such as optimism bias;41 though
there is evidence that interest and objectivity may be raised during periods of
life transition.42 For example, Gagné et al pointed to more objective thinking
about the quality and risks of relationships at ‘choice points in the relationship
or major life transitions’, with greater ‘motivation to maintain . . . positive
views’ once decision making has concluded.43
36 Kim, n 6 above; H. Mahar, Why Are There So Few Prenuptial Agreements? (Cambridge, MT:
John M. Olin Center for Law, Economics, and Business, 2003) at http://www.law.harvard.edu/
programs/olin_center (last accessed 16 October 2016); Pleasence and Balmer, ibid.
37 L. Festinger A Theory of Cognitive Dissonance (Stanford, CA: Stanford University Press, 1957) 3.
38 ibid.
39 Kim, n 6 above, 447-448.
40 R. C. Ellickson, Order Without Law (Cambridge, MA: Harvard University Press, 1991) 115.
42 See, for example, N. D. Weinstein, ‘Unrealistic Optimism About Future Life Events’ (1980) 39
Journal of Personality and Social Psychology 806.
43 F. M. Gagné and J. E. Lydon, ‘Bias and Accuracy in Close Relationships: An Integrative Review’
(2004) 8 Personality and Social Psychology Review 322, 328.
"
(2017) 80(5) MLR 836–859 841
24
Wrong about Rights
AIMS AND HYPOTHESIS
In this paper we build on the existing literature, using data from both waves
of the English and Welsh Civil and Social Justice Panel Survey (CSJPS) to
investigate more fully the profiles of those who have good and poor knowledge
of their legal rights and responsibilities. We do so with a view to determining
what this might say about the level of legal understanding among individuals
in England and Wales, as well as to identify which groups, if any, have specific
education needs. While the scenarios pertain to English and Welsh law and the
respondents were resident in England and Wales, findings have significance for
public legal education and legal service delivery beyond this jurisdiction.
Based on Williams,44 it is firstly hypothesised that individuals will answer the
hypothetical questions incorrectly more often than not. Secondly, based on the
work of a number of different researchers,45 it is hypothesised that in spite of
a general lack of knowledge, knowledge improvements will be seen where the
particular area of law is likely to be of greater salience to the individual. Given
that knowledge is often acquired in a social context, it is thirdly hypothesised
that household effects will be evident in our findings. That is to say that levels
of knowledge will be similar within households. Finally, it is hypothesised that
knowledge is likely to be higher where the legal position sits more closely in
line with social norms.
METHODOLOGY
Dataset
Data in this study was drawn from the CSJPS, a large scale survey of the general
population’s experience of 97 types of legal problem (concerning consumer is-
sues, employment, neighbours, owned housing, rented housing, money, debt,
welfare benefits, education, clinical negligence, relationship breakdown, do-
mestic violence and care proceedings)46 and strategies used to resolve them.
The survey was a substantial development of the English and Welsh Civil and
Social Justice Survey (CSJS), which was first conducted in 2001,47 then again
in 200448 and on a continuous basis between 2006 and 2009.49 The CSJS was
itself a substantial development of the Paths to Justice survey.50
44 Williams and Hall, n 6 above.

45 See Saunders, n 6 above; Casebourne et al, n 7 above; Pleasence and Balmer, n 6 above.
46 Problems were identified by asking a variant of the following question in relation to each of the
13 categories of legal problem included in the surveys: ‘[have you/has your partner] had any
(other) problems or disputes of the type shown on this card since [18 months]?
47 P. Pleasence, A. Buck, N. J. Balmer, A. O’Grady, H. Genn and M. Smith, Causes of Action: Civil
Law and Social Justice (Norwich: The Stationery Office, 2004).
48 P. Pleasence Causes of Action: Civil Law and Social Justice (Norwich: TSO, 2nd ed, 2006).
49 P. Pleasence, N. J. Balmer, A. Patel, and C. Denvir, Civil Justice in England and Wales 2009:
Report of the 2006-09 English and Welsh Civil and Social Justice Survey (London: Legal Services
Commission, 2010).
50 Genn, n 21 above.
"
842 (2017) 80(5) MLR 836–859
25
Two waves of the CSJPS were conducted prior to the survey’s replacement
by the Justiciable Problems Resolution Survey in 2012. Wave 1 interviews
were conducted between June and October 2010. Wave 2 interviews were
conducted eighteen months later, concluding in May 2012. The first wave
of the survey included 3,806 respondents (aged 16+), drawn from a random
selection of 2,316 residential household addresses across 194 postcode sectors
of England and Wales. The household response rate was 61 per cent, and
the cumulative eligible response rate was 54 per cent. The second wave in-
cluded 3,911 respondents, 2,604 of whom had also been interviewed at wave
1. Of the remainder, 148 were resident in a household surveyed at wave 1,
but not interviewed until wave 2, 96 were new residents in a household sur-
veyed at wave 1, and 1,063 were new respondents from new households. For
the longitudinal sample, the household response rate was 75 per cent and
the cumulative eligible response rate 70 per cent.51 For the cross-sectional
sample the household response rate was 53 per cent and cumulative eligible
response rate was 43 per cent. Wave 1 interviews took an average of 37 min-
utes,52 and wave 2 interviews an average of 35 minutes. Across both waves of
the survey, the sample was broadly representative of the residential household
population of England and Wales, which comprises around 98 per cent of the
total population.
The 2010 CSJPS included questions designed to explore knowledge of rights
and awareness of problem resolution options in relation to hypothetical legal
scenarios. Respondents were asked a series of ‘Yes/No’ questions about legal
rights and responsibilities for one of three hypothetical scenarios, relating to
housing, employment or a consumer transaction. All respondents were also
asked about a relationship breakdown scenario and this has been reported on
elsewhere.53,54
For each of the first three hypothetical scenarios, respondents were ran-
domised into three sub-groups, with a time-related aspect of the scenarios
altered for each sub-group. The purpose of this variation was to allow for
51 The individual level response rate was a very high 93 per cent.
52 An initial longer form of the questionnaire (asking about more questions in detail) averaged 42
minutes (n=762), with the final questionnaire averaging 35 minutes (n=3,044).
53 Pleasence and Balmer n 6 above.
54 The use of hypothetical scenarios, or vignettes, has a number of advantages in the examination
of beliefs. See, K. D. Watson, M. J. Polonsky and M. R. Hyman, ‘Designing Vignette Studies
in Marketing’ (2002) 10 Australasian Marketing Journal 41 which summarises the benefits of
vignette approaches to include enhancement of respondent involvement, greater realism in the
survey context, enhancement of measurement reliability and construct validity. However, the
vignette approach is far from limitation free (Denvir et al, n 6 above). Risks of ambiguity and
misinterpretation remain. Moreover, as is the case more generally, individuals can feel compelled
to respond to knowledge-based questions in a manner other than ‘don’t know’ (R. Nadeau, and
R. G. Niemi, ‘Educated Guesses: The Process of Answering Factual Knowledge Questions in
Surveys’ (1995) 59 The Public Opinion Quarterly 323), even in the face of considerable uncertainty
(D. Chong, ‘How People Think, Reason, and Feel about Rights and Liberties’ (1993) 37 American
Journal Of Political Science 867). Responses can therefore sometimes constitute ‘wild guesses’. Also,
people without knowledge may sometimes answer fact-based questions in a manner no different
to how they would answer attitudinal questions (R. Tourangeau and K. Rasinski, ‘Cognitive
Processes Underlying Context Effects in Attitude Measurement’ (1998) 103 Psychological Bulletin
299).
"
(2017) 80(5) MLR 836–859 843
26
Wrong about Rights
examination of the extent to which people regard legal rights and responsibili-
ties as time-dependent. While we report on this aspect of the scenarios below,
in most instances we focus on knowledge across respondents as a whole, while
accounting for the experimental conditions as far as possible.
In the housing scenario, ‘Alisha’ (the protagonist) agrees to rent a house55
from a landlord who lets out a number of properties and lives elsewhere. Six
weeks after moving in, she discovers that the bath has been leaking, causing
the house to become damp. She asks the landlord to repair the leak. Without
providing any notice, the landlord visits the house one afternoon and, after
knocking on the door, lets himself in to inspect the leak. At this point in the
scenario, respondents were asked whether the landlord is entitled to enter the
house in this way and whether the landlord is legally obliged to repair the leak.
Respondents were then told that the landlord refuses to repair the leak, and
that, three months after moving in she herself pays for the repair to be done and
deducts the cost from the next rent payment. She does not tell the landlord that
she is going to do this, but encloses a note with the rent payment explaining
what she has done. After the next rent becomes due, the landlord calls Alisha
and says that she must leave the house in 28 days’ time. The landlord says she
is in breach of the tenancy agreement by not paying the rent in full. At this
point in the scenario, respondents were asked whether Alisha has breached her
tenancy agreement by not paying her rent in full and whether, if she refuses to
leave, the landlord is able to evict her without first obtaining a Court Order.
Respondents were then told that, after 28 days have passed, two employees of
the landlord arrive at the house and say they have been sent by the landlord
to help Alisha move out. Respondents were asked if the two employees have
the legal right to enter the property to remove Alisha’s belongings. Finally,
respondents were told that before the 28th day the landlord obtained a Court
Order stating that Alisha must leave the house by the 28th day. Respondents
were then asked whether the two employees now have the legal right to enter
the property to remove Alisha’s belongings after 28 days have passed.56
In the employment scenario, Alisha (aged 59) has been working 48 hours
per week as an employee of Zap Computers, earning £5.50 per hour.57 Her
manager says he needs her to increase her hours to 50 hours per week, but
Alisha does not want to work the extra hours. Her manager shows her part
of her contract that says she can be asked to work up to 50 hours per week.
Respondents were then asked whether Alisha has to work 50 hours per week,
whether her salary is above or below the (2010) national minimum wage, and
whether the minimum wage varies by age. Respondents were then told that
55 In the housing scenarios, respondents were told that Alisha had a lease for six months, one year
or two years.
56 The ‘correct’ answers to the questions were that (1) the landlord cannot enter the property in
the way described, (2) the landlord is legally obliged to repair the leak, (3) Alisha breaches her
tenancy agreement by not paying her rent in full, (4) the landlord cannot evict Alisha without
first obtaining a Court Order, (5) the two employees do not have a right to enter the property,
and (6) the two employees still have no right to enter the property following the Court Order
being obtained.
57 In the case of the employment scenario, respondents were told that Alisha had been in her job
for six months, one year or two years.
"
844 (2017) 80(5) MLR 836–859
27
Alisha had been asking to see the main terms of her contract since she started
working at Zap Computers, and asked whether she has a legal right to see
these. One month later she is told that she is going to lose her job. At this
point in the scenario, respondents were asked whether Alisha is covered by the
full range of (time-dependent) unfair dismissal laws. Finally, respondents were
told that Zap Computers’s personnel manager explains that ZAP is reducing
the number of technicians it employs, and that ALISHA is going to be made
redundant. The personnel manager tells her it is only fair that ‘the older staff go
first’. Respondents were asked whether Zap Computers is allowed to consider
Alisha’s age in deciding who is to be made redundant.58
In the consumer scenario, Alisha buys a new ‘off the shelf’ sofa from a local
discount shop SOFAS4U. She agrees a delivery date that is in two weeks’
time. No delivery is made on the scheduled delivery date. When Alisha calls
SOFAS4U, the shop says they forgot to send the sofa out. Respondents were
asked whether Alisha has the right to cancel the order and get a refund. They
were then told that the sofa is delivered the next day. However, after Alisha
receives the sofa she decides she does not want it.59 She has not yet unpacked
the sofa, but SOFAS4U tell her they do not accept returns or offer refunds. At
this point in the scenario respondents were asked whether SOFAS4U have to
take the sofa back and provide a refund, and whether the situation would be
different had Alisha bought the sofa from SOFAS4U’s website instead of their
shop. Respondents were then told that Alisha keeps the sofa, but when she
unpacks it the next day, she discovers a minor defect that SOFAS4U should
be able to repair easily. She calls SOFAS4U and asks them if they will arrange
for a replacement or repair. SOFAS4U say she should get in touch with the
manufacturer and not them. Respondents were then asked if SOFAS4U are
legally obliged to replace the sofa. Finally, respondents were told that the sofa
then collapses when two of Alisha’s friends sit down on it at the same time.
Alisha calls SOFAS4U and the manufacturer, but finds that they have both
gone bust. The sofa had cost £400 and Alisha paid with a credit card. She
decides to call the credit card company to see if they will pay for the repair or
give her a refund. The credit card company says her problem has nothing to do
with them. Respondents were asked whether the credit card company is right
that Alisha’s problem is nothing to do with them.60
58 The ‘correct’ answers to the questions were that (1) Alisha does not have to work for 50 hours
per week, (2) Alisha’s salary is below the national minimum wage, (3) the national minimum
wage does vary by age, (4) Alisha does have a legal right to see the main terms of her employment
contract, (5) Alisha is covered by the full range of unfair dismissal laws in the case of 1 and 2 years’
employment, but not in the case of 6 months’ employment (though the position is different
today), and (6) Zap Computers cannot consider Alisha’s age in deciding who is to be made
redundant.
59 In the case of the consumer scenario, respondents were told that Alisha decided to return goods
after one day, three days or seven days.
60 The ‘correct’ answers to the questions were (1) Alisha does not have the legal right to cancel the
order, (2) SOFAS4U are not legally obliged to take the sofa back and provide a refund if bought
from their shop, (3) SOFAS4U are legally obliged to take the sofa back and provide a refund if
bought from their website, (4) SOFAS4U are not legally obliged to replace the faulty sofa, and
(5) the credit card company is also responsible.
"
(2017) 80(5) MLR 836–859 845
28
Wrong about Rights
In total, 1005 respondents were asked questions about the housing scenario,
966 about the employment scenario and 982 about the consumer scenario.
In the next section we describe the patterns of answers provided to the
scenario based legal rights questions. Then in the following section we set out
the results of a series of statistical analyses undertaken to explore the drivers of
knowledge as exhibited through the scenario questions.
Analysis
First we used simple descriptive statistics to look at how respondents answered

the housing, employment and consumer scenarios and the number of correct
answers they gave. We then compare this to the scores they would have ob-
tained had they relied on chance (guessing) alone. We explore which problems
respondents had particular trouble or ease answering, before turning to explore
what factors were associated with higher/lower scores.
We then fitted three multilevel binary logistic regression models designed to
predict score on the basis of a range of predictor variables linked to legal literacy.
These included: age (16 to 24 years old, 25 to 34 years old, 35 to 44 years
old, 45 to 59 years old, 60 to 74 years old, 75+ years old); educational quali-
fications (apprenticeship/none, GCSE/Other, Post GCSE/pre degree, degree,
unknown); professional status (routine manual/other, technical/semi-routine
manual, other management/clerical, professional/senior management, legal
professional); and whether respondents were recent migrants61 (within the past
10 years). The models also included a variable indicating whether respondents
had recent personal experience of problems of the types being asked about, a
variable indicating whether they had been at risk of experiencing problems of
the types being asked about62 (apart from in the case of the consumer model,
as no simple differentiation was possible), and a variable indicating whether
they felt legal problems of the type included in the CSJPS should be resolved
‘within . . . family or community’ (rather than ‘by using lawyers or courts’).63
Finally, variables were included to reflect the experimental structure of the hy-
pothetical scenarios64 and the household structure of the sample.65 The results
section presents findings without the need for specialised statistical knowledge.
61 This recognises that recent migrants will generally have had less experience and opportunity to
learn about the detail of English law.
62 For the rented housing scenario, risk was taken to be present for all those living in rented
accommodation. For the employment scenario, risk was taken to be present for all those currently
in employment.
63 This attitude variable was included as a proxy for practical interest in the form of the law.
64 As detailed above, there were three time conditions included within the hypothetical scenarios
(related to the length of the lease, the length of employment and the time elapsed since the
sale). The hypothetical scenarios were therefore constructed as ‘factorial vignettes’ rather than
‘constant variable value vignettes’. Indeed, in the case of the employment scenario, the correct
answers varied by condition. Thus, multilevel models were used to establish whether answers
varied by condition.
65 Since in some cases more than one household member answered the same questions, we explored
whether responses tended to be similar between household members.
"
846 (2017) 80(5) MLR 836–859
29
40
35 Housing
Employment
Percentage of respondents
30 Consumer
25
20
15
10
5
0
0 1 2 3 4 5 6
Number of correct answers
Figure 1: Number of correct answers to the six housing and employment questions and five consumer
scenario questions66
However full statistical model outputs can be found in Table 3 in the statistical
appendix.
RESULTS
General knowledge and chancing answers
Overall, respondents answered 59 per cent of the fact-based scenario questions

correctly, though there was a statistically significant difference in the scores
for the different scenarios.67 So, while housing scenario respondents answered
4.3 (SD = 1.2) (71 per cent) of their six questions correctly on average, and
employment scenario respondents 4.0 (SD = 1.3) (66 per cent), consumer
scenario respondents answered just 1.72 (SD = 1.0) (34 per cent) of their 5
questions correctly;68 an interesting initial finding given that consumer rights
have been found to be among those that CSJPS respondents are most confident
about.69
Figure 1 illustrates the distribution of correct answers to the housing, em-
ployment and consumer scenarios. As can be seen, the distributions are skewed
towards a greater number of correct answers in the case of the housing and
66 So a score of six was not possible for the consumer scenario. In total, 1,005 respondents answered
rented housing, 966 employment and 982 consumer questions. For comparison, if respondents
answered the questions randomly (ie, a binomial distribution of five trials for consumer and six
for employment/housing and a probability of 0.5), we would expect 3.1% scoring 0, 15.6% 1,
31.3% 2, 31.3% 3, 15.6% 4 and 3.1% 5 for consumer and 1.6% scoring 0, 9.4% 1, 23.4% 2,
31.2% 3, 23.4% 4, 9.4% 5 and 1.6% 6 for housing or employment.
67 F=921.9, p<0.001, after adjusting consumer scores to be equivalent to other scores.
68 34.9% if answers to a 6th question asked only of some respondents are included.
69 Denvir et al, n 6 above.
"
(2017) 80(5) MLR 836–859 847
30
Wrong about Rights
employment scenarios and, in stark contrast, towards a lesser number in the

case of the consumer scenario. In the case of the housing scenario, 77 per cent
of respondents managed to answer four or more questions correctly, with 49
per cent answering at least five correctly and 13 per cent all six correctly. For
the employment scenario the figures were slightly lower at 66 per cent, 37 per
cent and 12 per cent.
Given that providing random ‘Yes/No’ answers to the questions could be
expected to yield four or more correct answers 34 per cent of the time, five
or more correct answers 11 per cent of the time and perfect six scores 1.6 per
cent of the time, the performance of respondents in respect of the housing and
employment scenarios looks creditable. However, the performance in respect
of the consumer scenario looks anything but. For this scenario, just 20 per
cent of respondents provided three out of five or more correct answers, three
per cent provided four or more and just 0.3 per cent scored perfect fives. This
compares to the 50 per cent, 19 per cent and three per cent that could be
expected to be delivered by chance.
It should however be noted that respondents did not always offer yes or no
answers to the questions; sometimes simply stating that they did not know. In
fact, just 59 per cent of respondents answered all six housing questions decisively,
44 per cent all six employment questions decisively and 53 per cent all five
consumer questions decisively.70 Thus the above comparison with chance may
be a little unfair.
Alternatively, a comparison between the scores of those who answered all
five consumer questions decisively and chance sees like competing with like.
However, as can be seen from Figure 2 even among only those who answered
the consumer scenario questions decisively scores were substantially below
those of chance. So, which questions proved so troublesome, and what were
the drivers of higher scores?
Findings in detail
Table 1 sets out the overall structure of the scenarios and respondents’ success
in answering each of the scenario questions. Across all scenarios/questions,
respondents answered correctly 59 per cent of the time. As can be seen, there
was a considerable range of scores across the 18 legal rights questions asked
about the scenarios. The highest score – with 95 per cent of respondents
answering correctly – was associated with the employment scenario question
concerning whether an employee has a legal right to see the main terms of
70 In the rented housing scenario, ‘don’t know’ responses varied from 6.1% to 16.5%, with the
highest percentage for the question exploring the landlord’s ability to evict Alisha without a
court order. In the employment scenario, ‘don’t know’ responses varied from 2.9% to 24.0%,
with the highest percentage for the question regarding whether Alisha was covered by the full
range of unfair dismissal laws. In the consumer scenario, ‘don’t know’ responses varied from
11.2% to 25.4%, with the highest percentage for the question asking whether the sofa company
would have to provide a refund if the sofa was bought from a website. Unless decisive answers
are referred to, analysis treats ‘don’t know’ responses as being incorrect.
"
848 (2017) 80(5) MLR 836–859
31
Figure 2: Number of correct answers (consumer scenario), for all answers and restricted to only
those answering questions decisively, compared to what might be expected by chance [Colour figure
can be viewed at wileyonlinelibrary.com]
Table 1. Respondents’ success in answering questions relating to each of the three scenarios (overall
and for specific questions)
Scenario Specific question (percentage correct)
Housing scenario Is the landlord entitled to enter the house? (77.4%)

overall (71.0% Is the landlord legally obliged to repair the leak? (91.4)
correct) Has Alisha breached her tenancy agreement by not paying the rent in full (66.9%)
Can the landlord be able to evict Alisha without a Court Order? (73.9%)
Do the employees have the right to enter the property after 28 days? (84.1%)
Do the employees have this right if a Court Order has been obtained? (32.5%)
Employment scenario Does Alisha have to work 50 hours per week? (53.6%)
overall (66.2% Is Alisha’s salary above or below the national minimum wage? (56.5%)
correct) Does the national minimum wage vary by age? (57.6%)
Does Alisha have the legal right to see the main terms of her contract? (95.0%)
Is Alisha covered by the full range of unfair dismissal laws? (52.4%)
Can Alisha’s age be considered in deciding to make her redundant? (82.1%)
Consumer scenario Does Alisha have the right to cancel the order for late delivery? (15.0%)
overall (34.4% Do SOFAS4U have to take the sofa back and provide a refund? (34.2%)
correct) If bought on the website, would SOFAS4U have to provide a refund? (52.9%)
Is SOFAS4U legally obliged to replace the faulty sofa? (12.1%)
Is the problem nothing to do with the credit card company (57.8%)
their employment contract. Then followed the housing scenario questions con-
cerning whether a landlord is legally obliged to repair a leaking bath (91 per
cent) and whether employees of a landlord are allowed to effect an eviction
without a Court Order (84 per cent), and the employment scenario ques-
tion concerning whether age can be a consideration in redundancy decisions
(82 per cent).
"
(2017) 80(5) MLR 836–859 849
32
Wrong about Rights
40
35
Exposed to
30 risk of
problem
25
20
15 Not exposed
to risk of
10 problem
0
0 1 2 3 4 5 6
Figure 3: Number of correct answers to the housing scenario questions on the basis of whether or
not respondents were living in rented housing
The lowest scores – with just 12 per cent and 15 per cent of respondents
answering correctly, respectively – were associated with the consumer scenario
questions concerning whether a shop is legally obliged to replace a faulty item
and whether an order made in a shop can automatically be cancelled for late
delivery. The lowest housing scenario score was associated with the question
concerning whether a landlord’s employees are allowed to effect an eviction
following the grant of a Court Order (33 per cent).
The lowest employment scenario score was associated with the question
concerning when employees are covered by the full range of unfair dismissal
laws. This was a question where the answer varied with the length of employ-
ment, and those respondents asked about an employment of just six months
least often provided correct answers (23 per cent). Those asked about em-
ployments of one year and two years provided correct answers far more often
(61 per cent and 73 per cent, respectively).
Knowledge where it is needed
There is, of course, a big difference between knowledge deficits among those
who are unlikely to require specific knowledge and those who will potentially
require it. Figure 3 sets out the different housing scenario scores of those
respondents living in the rented sector – and therefore more at risk of problems
of the type depicted in the scenario – and those living elsewhere.
As can be seen, the pattern of scores looks quite similar for each of these
groups. Indeed, there was no difference in the proportion of correct responses
provided by respondents living in the rented sector (4.3 out of six) and living
elsewhere (4.3 out of six). Looking at those who reported having actually
experienced similar housing problems, they correctly answered a slightly higher
proportion of questions (4.5 out of six), but the difference was relatively small
and not significant, owing to just 21 respondents falling into this group.
"
850 (2017) 80(5) MLR 836–859
33
35
30 Exposed to
risk of
25 problem
20
15 Not exposed
to risk of
10 problem
0
0 1 2 3 4 5 6
Figure 4: Number of correct answers to the employment scenario questions on the basis of whether
or not respondents were in employment
Figure 4 sets out the different employment scenario scores of those re-
spondents in employment71 – and therefore, again, more at risk of such
problems – and those not in employment. As can be seen, in this instance
markedly different patterns can be seen, with those in employment scoring
significantly higher. On average, those in employment answered 4.3 out of six
questions correctly, compared to just 3.6 for those not in employment, a sta-
tistically significant difference.72 However, there was no significant difference
in the scores of the 26 respondents who reported having experienced similar
employment problems.
For the consumer scenario, there was no means to identify respondents
substantially more at risk of experiencing similar problems, but, as Figure 5
illustrates, a comparison of the scores of those who actually reported problems
concerning faulty goods and those who had not, suggested no difference. Both
groups answered 1.7 of the five questions correctly.
Factors associated with knowledge
To identify factors associated with higher levels of knowledge of legal rights and
responsibilities, controlling for other matters, we employed multilevel logistic
regression to predict respondents’ scores in relation to each of the housing,
employment and consumer scenario questions.73 Table 2 shows the factors
associated with the number of questions answered correctly for each of the
three scenarios.
As shown, age was found to be significantly associated with legal knowledge
across all three scenarios. The oldest respondents consistently obtained relatively
71 This included self-employed respondents.

72 T=-7.88, p<0.001.
73 Model outputs set out in the statistical appendix.
"
(2017) 80(5) MLR 836–859 851
34
Wrong about Rights
45
40
Experienced
35 related
30 problem
25
20
No related
15
problem
10 experienced
5
0
0 1 2 3 4 5
Figure 5: Number of correct answers to the consumer scenario questions on the basis of whether or
not respondents reported having experienced a recent problem concerning faulty goods
Table 2. Factors associated with the number of questions answered correctly for each scenario (full
details of the statistical models can be found in the statistical appendix)
Whether Statistically Significant in Model
Variable Housing Employment Consumer

√ √ √
Age √
Education qualifications √ √
Professional status
Recent migrant
Experience of similar problem √
Risk of similar problem ∗
√
Attitude to problem resolution √
Scenario structure √ √
Household effect
low scores, while middle-aged respondents generally performed relatively well.

However, as Figure 6 illustrates,74 there were some differences in the patterns of
association between scenarios. For example, the youngest respondents scored
significantly lower than those in other age groups in relation to the housing
scenario, but performed relatively well in relation to the other scenarios.75
Professional status was significantly associated with knowledge scores in
relation to two scenarios, the housing and consumer scenarios, but not the
employment scenario. In the case of both the housing and consumer scenarios,
74 Figure 6 is based on simulations from the models, controlling for other variables.
75 There was also some indication that they youngest respondents (16-18 year olds) scored worse
still in the housing scenario. However, findings should be treated with some caution since the
number of 16-18 year olds answering housing questions was relatively small (40 respondents).
"
852 (2017) 80(5) MLR 836–859
35
0.8
0.7
Proportion of correct answers

0.6
Housing
0.5
Employment
0.4 Consumer
0.3
0.2
0.1
0
16-24 25-34 35-44 45-59 60-74 75+
Age
Figure 6: Association between age and legal knowledge for each of the three scenarios, derived from
the statistical models and controlling for other variables
legal professionals scored higher than others76 and, more subtly, knowledge
appeared to increase with professional seniority.
Educational qualifications were found to be significantly associated with
legal knowledge in the case of only one scenario, the rented housing scenario.
Here, a lack of qualifications was associated with significantly higher scores.
There was little evidence of qualifications, or lack thereof, relating to success
in answering the employment or consumer scenario questions.
Whether respondents had migrated to the UK within the past ten years was
not found to be significantly associated with legal knowledge in relation to any
scenario. While this group scored lower than others across all three scenarios77
the small number of recent migrants included in the models78 was insufficient
to confirm the significance of differences of the magnitude observed.
Turning to respondents’ experience of problems of the types asked about,
again no significant associations were found with legal knowledge, although,
again, numbers were small.79 However, a significant association was found
between respondents’ exposure to the risk of employment problems and legal
knowledge. Here, simulation from the model suggested that, controlling for
other factors, those at risk of employment problems would be expected to
76 Incidentally, while small numbers of legal professionals scored higher better than other respondents
in housing and consumer scenarios, they were far from perfect. Only 7 of 18 answered all rented
housing questions correctly, while none answered all of the employment or consumer questions
correctly (of 8 and 12 respectively).
77 On the basis of simulations from the models, controlling for other variables.
78 There were just 28 recent migrants in the housing model, 36 in the employment model and 31
in the consumer model.
79 There were just 21 respondents who had experienced similar problems in the housing model
and 26 in the employment model. In both these cases respondents scored higher than those who
had not experienced problems (on the basis of simulations from the models, controlling for other
variables), but the findings were not significant. For the consumer model, the model suggested
that, controlling for other factors, those who had experienced similar problems scored slightly
lower than others.
"
(2017) 80(5) MLR 836–859 853
36
Wrong about Rights
answer 65 per cent of questions correctly, compared to 60 per cent for other
respondents. This association was not mirrored in the case of the housing
scenario.
Elsewhere, respondents who disagreed that, in general, problems should be
resolved within the family or community, rather than by using lawyers or
courts, were found to score significantly higher in the housing scenario than
those who were neutral on the matter.80
Finally, turning to methodology related variables, the experimental elements
of the hypothetical scenarios were found to be significantly associated with legal
knowledge scores in the case of the employment scenario, with respondents
scoring lower when presented with a scenario in which Alisha had been em-
ployed for six months, as opposed to one or two years. This reflects the fact
that the correct answers were different for this condition, and a similar answer
pattern to those associated with the other conditions yielded a lower score.
Also, there were significant household effects in relation to housing and em-
ployment scenario scores. This indicated that where more than one household
member answered the questions, scores were likely to be related. No household
effect was observed in relation to the consumer scenario.
CONCLUSION
Our findings shed important and extensive new light on the public’s under-
standing of legal rights and responsibilities. In line with previous studies,81 our
results make evident a substantial public legal knowledge deficit across England
and Wales. Overall, respondents answered only 59 per cent of our fact-based
scenario questions correctly; only moderately eclipsing chance. Also in line
with previous studies, our results indicate that this deficit is greater in some
areas of law than others.82
So, while modest levels of public understanding were evident in respect of
rented housing (71 per cent correct) and employment law (66 per cent correct),
profound ignorance was observed in the case of consumer law. On average,
consumer scenario respondents answered just 34 per cent of five consumer
law questions correctly; a score substantially worse than would be expected
by chance! Indeed, just 12 per cent and 15 per cent, respectively, correctly
answered questions concerning whether a shop is legally obliged to replace a
faulty item and whether an order made in a shop can automatically be cancelled
for late delivery.
80 In percentage terms, simulated from the statistical model and controlling for other variables, those
who held no strong views would be expected to answer 68% correct on average, compared to
71% for those who ‘agreed’ and 72% per cent for those who ‘disagreed’.
81 See, for example, Cortese, n 6 above; Williams and Hall, n 6 above; Saunders, n 6 above; Baker
and Emery n 6 above; Darley et al, n 6 above; Kim, n 6 above, Darley et al, n 29 above; Barlow
et al, n 6 above; Militello et al, n 6 above; Panades et al, n 6 above; Parle n 6 above; Pleasence
and Balmer, n 6 above; Denvir et al, n 6 above.
82 For previous studies, see: Saunders, ibid; Casebourne et al, n 7 above; Pleasence and Balmer, ibid.
"
854 (2017) 80(5) MLR 836–859
37
Despite this profound ignorance, consumer issues represent the most com-
mon form of justiciable issue and those about which CSJPS respondents ex-
pressed greatest confidence in their legal knowledge. Indeed, 66 per cent of
those who reported consumer problems in the 2010 and 2012 CSJPS professed
to understand their rights either ‘mostly’ or ‘completely’. However, when we
compared the legal knowledge scores (as derived from hypothetical scenario
questions) of those who reported consumer problems with those who did not,
we found no difference. Moreover, when we compared the scores of those
who claimed ‘complete’ understanding of their consumer rights with others,
we found that they scored worse still. People think they know their consumer
rights, but they do not.
So, what is going on here? Building on earlier findings concerning public
understanding of family law,83 which suggested that people’s beliefs are influ-
enced by social norms and personal moral viewpoints,84 we here go further and
suggest that the profound mismatch between people’s actual and professed un-
derstanding of the law in the case of consumer law is likely strongly influenced
by the practice norms of retailers. Respondents’ beliefs about consumer law,
while strikingly wrong, are also strikingly in line with retail practice, where
cancellations of orders for late (or even on-time) delivery are routinely ac-
cepted, refunds are consistently provided for ‘mistake’ purchases and defective
products are ordinarily replaced with new ones.
Happily, in the case of employment law, it appeared that those at risk of
problems (ie those in employment) had greater levels of relevant legal knowl-
edge than others (72 per cent versus 60 per cent), although this difference
could not be seen among those who actually reported employment problems.
But numbers were small and definitive conclusions not possible. In the case of
rented housing law, those at risk of problems appeared not to have any knowl-
edge advantage, although in this case there was some suggestion that those who
had faced problems may have been slightly more knowledgeable. However,
numbers were again small and definitive conclusions not possible.
Our findings concerning the social patterning of legal knowledge indicated
that the oldest CSJPS respondents were more likely to perform poorly, while
middle-aged respondents were more likely to perform relatively well in an-
swering the hypothetical scenario questions. In the case of the housing and
consumer scenarios, knowledge seemed to increase somewhat along with pro-
fessional standing (with lawyers, perhaps unsurprisingly, scoring highest). There
was also a slight suggestion that knowledge was related to time spent living in
the UK, though numbers were too small (in the case of recent migrants) to
draw conclusions.
In highlighting a substantial deficit in the public’s understanding of legal
rights and responsibilities – even among those for whom particular rights and
responsibilities have specific bearing – the implications of our findings are
profound. With law ubiquitous in everyday life, the ubiquity of ignorance of
law is of evident concern.
83 Pleasence and Balmer, ibid.
84 See also, Ross, Greene and House, n 30 above; Kim, n 6 above; Barlow et al, n 6 above, Lewis,
Tennant and Taylor, n 17 above.
"
(2017) 80(5) MLR 836–859 855
38
Wrong about Rights
Knowledge of legal rights and responsibilities provides a framework that in-

forms expectations and enables the management of risk in social and economic
interactions.85 It is also critical to any role law may have in guiding social
behaviour and, as the Public Legal Education and Support Task Force noted,
contributing to the ‘agendas of government’.86 It is also relevant to the rule
of law. Unequal knowledge of law weakens equality under law and there is,
anyway, an arbitrariness to unknown law, whatever its democratic origins.87 As
Lord Diplock observed in 1975:
The acceptance of the rule of law as a constitutional principle requires that a

citizen, before committing himself to any course of action, should be able to know
in advance what are the legal principles which flow from it.88
Of course, Lord Diplock here identifies that citizens of a democratic state

cannot be assumed to be passive in their understanding of law. Citizens have a
responsibility to inform themselves of laws relevant to their activities. However,
given that citizens have only limited comprehension of the extent to which law
impinges upon contemporary life89 and that, as Lord Bingham has observed,
‘legislative hyperactivity has become a permanent feature of our governance’ –
with nearly 5,000 pages of primary legislation and 11,500 pages of subordinate
legislation enacted in 2006 alone90 – the responsibility of citizens to inform
themselves about the law must be quite limited.91
Thus, while the legal illiteracy of the public may not be something that
threatens the fabric of democracy, it does nonetheless pose important questions
about the role of law and responsibility of government to enable citizens to
understand and engage with it. Moreover, our findings of social patterning
in levels of legal literacy indicate that supplementary assistance is important
for some sections of the community if understanding of law is not to act as a
substantial barrier to engagement with it.
The above suggests that state powers should be mindful of public under-
standing of the law, and take reasonable steps to support it. Plainly, these steps
extend to the national curriculum (which addresses foundational aspects of our
constitutional framework and legal system within the citizenship programmes
of study) and broader public legal education (PLE) efforts concerning the cur-
rent state of the law.92 But there is also a particular need for PLE initiatives in
85 Meager, n 16 above.
86 Public Legal Education and Support Task Force, Developing Capable Citizens: The Role of Public
Legal Education (London: PLEAS Task Force, 2007) 29.
87 M. Galanter, ‘Why the Haves Come Out Ahead: Speculations on the Limits of Legal Change’
(1974) 9 Law and Society Review 95.
88 Black-Clawson International Ltd v Papierwerke Waldhof-Aschaffenburg AG [1975] AC 591, 638D.
89 P. Pleasence, N. J. Balmer and S. Reimers, ‘Horses for Courses? People’s Characterisation of
Justiciable Problems and the use of Lawyers’ in Legal Services Board (ed), The Future of Legal
Services: Emerging Thinking (London: Legal Services Board, 2010); Pleasence, Balmer and Reimers,
n 20 above.
90 And this ignores law emanating from the European Union and elsewhere.
91 T. Bingham, The Rule of Law (London: Allen Lane, 2010)
92 These are generally quite narrowly focused. See, for example, the Living Together campaign in
relation to cohabitation law (A. Barlow, C. Burgoyne, E. Clery and J. Smithson, ‘In Practice:
"
856 (2017) 80(5) MLR 836–859
39
the context of legislative change. After all, what government would want to
enact legislation that is not known about by those to whom it applies? In the
case of new law, the challenge of supporting legal literacy is made greater by
the likelihood that beliefs about law are likely to be ‘resistant to change”.93
The importance of PLE has also been heightened by the substantial cuts to
legal aid brought about by the LASPO Act 2012 which have impacted not just
on help provided to those engaged in formal process, but also on basic help
provided to those seeking to understand the the nature of problems they face
and their resolution options. As was recognised in the Public Bill Committee
debates regarding the passage of the LASPO Bill, there is now a far greater public
reliance on self-help in the resolution of legal disputes, including representation
before courts and tribunals.94 Yet, while the government earmarked a ‘£2m
package of support aimed at avoiding expensive and confrontational courtroom
battles’, to include additional funding of the legal and advice sectors to ‘increase
legal and practical support for litigants in person in civil and family courts’,95
it has
imposed upon itself no duty to promote knowledge of rights, develop just-in-time

legal information, share the third sector’s burden of equipping citizens to better
handle their problems alone or for that matter, inform itself as to the need for
public legal education interventions.96
Thus, we can perhaps expect fewer disputes to progress to formal legal process
or resolve ‘in the shadow of the law’.97 In the case of those disputes that come
before legal institutions with parties ignorant of the law, we might expect less
effective argument and greater burden on the institutions concerned.98
Finally, in relation to our findings concerning consumer law, it might be
thought that people’s systematic lack of understanding is unproblematic in
practice, as it evidences norms of behaviour (eg, ‘goodwill’ returns policies)
that offer greater protection than the law requires. However, over-optimism
in relation to legal protection can lead people to take on far greater risk than
they would want, or be able, to deal with. And it is not just consumer law
that people err on to their own advantage. Our findings also highlighted that
32 per cent of CSJPS respondents believed a tenant has the legal right to deduct
from rent payments the cost of services a landlord is obliged to, but does not,
The Living Together Campaign – The Impact on Cohabitants’ (2007) Family Law 166) and
the recommendations of the Vulnerable Worker Enforcement Forum (Department for Busi-
ness, Enterprise and Regulatory Reform, Vulnerable Worker Enforcement Forum: Final Report and
Government Conclusions (London: BERR, 2008)).
93 Kim, n 6 above.
94 Public Bill Committee 2011, n 25 above.
95 Ministry of Justice press release, More Support for Separating Couples and Parents, 23 October
2014.
96 Denvir et al, n 6 above 156.
97 R. H. Mnookin, and L. Kornhauser, ‘Bargaining in the Shadow of the Law: The Case of Divorce’
(1979) 88 The Yale Law Journal 950.
98 Williams, n 5 above, Pleasence and Balmer, n 6 above.
"
(2017) 80(5) MLR 836–859 857
40
Wrong about Rights
provide. And similar erroneous optimism has also been documented in other
areas, such as family law and employment law.99
Over-optimism as regards some aspects of the law is also mirrored by pes-
simism as regards others; as in the case of the 46 per cent of CSJPS respondents
who were unaware of their right to limit their working hours to 48 per week.
Here, again, the impact of people’s reliance on their understanding of the law
can be highly detrimental and long lasting.
In conclusion, public ignorance of law is ubiquitous, can act to undermine
efforts to navigate the legal framework of everyday life, impacts on the outcome
of legal issues and imposes burdens on legal institutions. It strikes at law’s
efficacy, efficiency and legitimacy. It is therefore not a matter of simple academic
concern, but one of practical and constitutional significance – and our findings
suggest there is much that needs to be done to address it.
STATISTICAL APPENDIX
The three statistical models fitted in relation to the hypothetical scenarios

were logistic regression models (for proportions), implemented using MLwiN
statistical software (Rasbash et al, 2009). Logistic regression is a common model
type where the response variable (eg, respondent’s score out of six in the rented
housing scenario) is a proportion (ie, r correct out of n questions). A multilevel
model was used (with two levels) since the data was hierarchical, with scores
nested within households. Multilevel models can be used to correctly account
for this type of data structure.
The model is used to assess whether particular factors (eg, age group) are
associated with an increase or decrease in score. How to interpret logistic
regression is set out in a range of statistical texts, though all of the key findings
are summarised in simple terms in the text above, without statistical output or
jargon.
Model output is set out in Table 3 below.
99 Pleasence and Balmer, ibid, Kim, n 6 above.
"
858 (2017) 80(5) MLR 836–859
41
Table 3. Multi-level logistic regression models predicting score on the hypothetical scenarios on the basis of a range of characteristics
C
"
Housing Employment Consumer
Variable Level N Estimate SE N Estimate SE N Estimate SE
Constant 1.06 0.19 0.42 0.18 −0.67 0.17

Problems should be resolved within family/ Neutral 255 0.00 − 233 0.00 − 247 0.00 −
community Agree 602 0.14 0.08 604 0.02 0.07 603 0.07
(2017) 80(5) MLR 836–859

−0.04
Disagree 148 0.23 0.11 129 0.14 0.11 132 −0.10 0.11
Age group 16-24 113 −0.31 0.12 117 −0.07 0.12 111 −0.21 0.12
25–34 127 −0.04 0.11 119 −0.24 0.11 124 −0.28 0.11
35–44 170 0.08 0.10 162 −0.14 0.10 161 −0.19 0.10
45–59 271 0.00 − 235 0.00 − 261 0.00 −
60–74 204 0.05 0.10 220 −0.23 0.10 217 −0.03 0.09
2017 The Author. The Modern Law Review "

75+ 117 −0.20 0.12 106 −0.79 0.12 101 −0.64 0.12
Profession Routine manual/other 157 0.00 − 170 0.00 − 155 0.00 −
Technical/semi-routine manual 262 0.10 0.10 239 0.16 0.09 254 0.02 0.10
Other management/clerical 226 0.18 0.11 207 0.18 0.10 227 0.24 0.11
Professional/senior management 268 0.23 0.12 267 0.07 0.11 254 0.26 0.12
42
Legal professional 18 1.00 0.31 8 −0.28 0.35 12 0.61 0.29
Unknown 74 −0.09 0.15 75 −0.02 0.14 80 −0.02 0.15
Qualifications Apprenticeship/none 56 0.00 − 63 0.00 − 59 0.00 −
GCSE/Other 219 −0.47 0.16 244 −0.04 0.14 244 0.07 0.15
Post GCSE/pre degree 236 −0.42 0.17 213 0.06 0.14 218 0.20 0.15
Degree 209 −0.41 0.17 193 0.14 0.15 209 0.16 0.16
Unknown 285 −0.50 0.16 253 −0.21 0.13 252 −0.07 0.14

Recent migrant No 977 0.00 − 930 0.00 − 951 0.00 −
Yes 28 −0.28 0.20 36 −0.17 0.17 31 −0.21 0.19
Experienced relevant civil justice problem No 984 0.00 − 940 0.00 − 913 0.00 −
Yes 21 0.33 0.24 26 0.26 0.20 69 −0.07 0.12
Exposure to relevant circumstances No 743 0.00 − 484 0.00 − − − −
Yes 262 0.10 0.08 482 0.23 0.08 − − −
Experimental group (time) 6 months (1 day for consumer) 323 0.00 − 318 0.00 − 343 0.00 −
1 year (3 days for consumer) 358 0.13 0.08 332 0.41 0.07 310 0.01 0.07
2 years (7 days for consumer) 324 0.08 0.08 316 0.44 0.07 329 0.02 0.07
Household effects 0.15 0.05 0.09 0.04 0.003 0.003
859
Dr. R. Rijnhout, Prof. dr. I. Giesen, dr. A. van Artikelen
Hoof & Prof. dr. M.A.G. van Aken1
Agressie van jeugdige voetballers en de rol van

het recht: kennis en (on)begrip van de (spel)regels
empirisch getoetst
AV&S 2016/34 ten aan te hollen,5 en uiteindelijk draaien alle hoofden rich-
ting de KNVB.6
Op en rond de voetbalvelden vindt regelmatig on- De KNVB ziet hier uiteraard voor zichzelf een taak wegge-
wenselijk gedrag plaats. Om dit probleem op te los- legd, en ziet een kracht(iger) normatief kader als één van de
sen, zet de KNVB o.a. in op het creëren van inzicht oplossingen voor onregelmatigheden bij amateurvoetbal-
en begrip voor voetbalregels, met als onderliggende wedstrijden:
veronderstelling dat dit tot minder normafwijkend
gedrag zal leiden. Maar is het wel zo dat meer re- “Het is onze primaire verantwoordelijkheid duidelijke
gelkennis zorgt voor meer begrip voor die regels en normen te stellen en degenen die buiten de oevers tre-
voor respectvol gedrag? En hoe staat het met de ken- den zwaar en direct te straffen (repressief). Dit doet de
nis van de regels (inclusief de regels van het civiele KNVB met zijn tuchtorganen, waarbij de districtsbestu-
aansprakelijkheidsrecht en het strafrecht) bij onze ren onder andere bevoegd zijn om verenigingen uit de
(jeugd)voetballers? Wij hebben deze vragen empi- competitie te nemen. Tegelijkertijd helpt het opstellen
risch onderzocht en berichten hier over de uitkom- van dit ‘normatieve kader’ onze verenigingsbesturen om
sten van dat onderzoek. de grenzen te kunnen bewaken. Om het zo maar te stel-
len: met deze acties herstellen we dus de orde.”7
1. Inleiding
Duidelijke normen stellen en normhandhavend optreden,
1.1 Aanleiding zou dus leiden tot ‘ordeherstel’. Onduidelijk is wat hier pre-
De laatste jaren wordt er regelmatig verslag gedaan in de cies met ordeherstel wordt bedoeld, maar gezien de aard
media van allerlei onwenselijk en soms strafwaardig gevan de tekst zal waarschijnlijk worden gedoeld op een situ-
drag door spelers (en toeschouwers) op (en rond) de Ne- atie waarin spelers sportief en respectvol gedrag vertonen.
derlandse (amateur)voetbalvelden. Regelmatig ontsporen Het is echter niet zo dat de KNVB louter inzet op repressief
voetbalwedstrijden in schelden zelfs knokpartijen, zo is optreden. Naast een duidelijk normenkader en normhand-
nog steeds in de krant te lezen.2 De dood van grensrechter havend optreden, wordt ook aandacht besteed aan preven-
Richard Nieuwenhuijzen in 2011 vormde hierbij een diepte- tieve acties, zoals het ondersteunen van initiatieven uit de
punt. Ondanks dat het aantal excessen3 op de amateurvoet- praktijk om het voetbal aantrekkelijk te houden.8 Er is dus
balvelden lijkt af te nemen4 , is het probleem van negatief gekozen voor een combinatie van middelen om gewenst
of agressief voetbalgedrag nog niet oplost. De maatschappij voetbalgedrag te stimuleren.
verafschuwt wat er langs en op onze velden gebeurt, maar Eén van die middelen is de invoering van een spelregelbe-
weet tegelijk niet goed hoe deze uitwas aan te pakken. Het wijs voor jeugdspelers. De KNVB doet dat op basis van de
OM krijgt een rol toebedeeld, maar is niet in staat om alle volgende aanname:
onruststokers aan te pakken, de politiek lijkt achter de fei-
“Een spelregelbewijs voor alle jeugdspelers in Nederland
(…) maakt voetballers van jongs af aan vertrouwd met de
regels en geeft hen meer inzicht in – én begrip voor – de
1 Rianka Rijnhout en Ivo Giesen zijn als universitair hoofddocent, resp. beslissingen van scheidsrechters.”9
hoogleraar verbonden aan het Utrecht Center for Accountability and Li-
ability Law (Ucall) en het Molengraaff Instituut voor Privaatrecht van de
Universiteit Utrecht (UU); Anne van Hoof en Marcel van Aken zijn resp. De KNVB wil dus meer inzicht en meer begrip creëren voor
universitair docent en hoogleraar bij de afd. Ontwikkelingspsychologie voetbalregels, met als onderliggende (maar niet geëxpli-
van de UU. Dit artikel kwam tot stand onder auspiciën van het project ‘Vio-
lent behaviour among young people on the football field’ dat gesubsidieerd
citeerde) veronderstelling dat meer inzicht (en begrip) tot
werd door het Strategic theme ‘Youth & Identity’ van de UU en dat bestond minder normafwijkend gedrag zal leiden, hetgeen dan
uit een samenwerking tussen USBO, de afd. Ontwikkelingspsychologie en
Ucall. Citeerwijze: R. Rijnhout, I. Giesen, A. van Hoof & M.A.G. van Aken,
‘Agressie van jeugdige voetballers en de rol van het recht: kennis en (on) 5 Dat wil echter geenszins zeggen dat er niets gebeurd in Den Haag op dit
begrip van de (spel)regels empirisch getoetst’, AV&S 2016/34, afl. 5. vlak, zie bijvoorbeeld het Actieplan ‘Naar een veiliger sportklimaat’ en de
2 Zie bijvoorbeeld het bericht van 15 maart 2016 in het Algemeen Dagblad, Nadere uitwerking daarvan: Kamerstukken II, 2010/11, 30234, 36 en 55.
w w w.ad.nl/ad/nl/38261/Nieuws/article/detail/4263179/2016/03/15/ 6 Zie in algemene zin over de werking van maatregelen tegen geweld rond-
KNVB-Meer-geweld-in-amateurvoetbal.dhtml. om de velden, maar dan specifiek in het betaalde voetbal: H. Ferwerda
3 Met excessen wordt bedoeld buitensporig fysiek geweld, ernstige bedrei- e.a., Raak geschoten?, WODC: Den Haag 2014. Het onderzoek waarover wij
gingen en het bespugen van anderen, zie D. Romijn et al., VSK Monitor 2015. hierna rapporteren, betreft jeugdspelers in het amateurvoetbal.
Voortgangsrapportage Actieplan “Naar een veiliger sportklimaat”, Utrecht: 7 KNVB, Tegen geweld voor sportiviteit, 2013.
Mulier Instituut 2015, p. 49. 8 KNVB, Tegen geweld voor sportiviteit, 2013.
4 Romijn et al. 2015, p. 49-50. 9 KNVB, Tegen geweld voor sportiviteit, 2013.
AV&S 2016/34 Afl. 5 - oktober 2016 171

43
T2_AV&S_1605_bw_V02.indd 171 10/14/2016 3:30:57 PM

AGRESSIE VAN JEUGDIGE VOETBALLERS EN DE ROL VAN HET RECHT Artikelen
uiteraard één van de middelen zou zijn voor het beoogde 2. Wettelijk en zelfregulerend kader
‘ordeherstel’. Maar die aanvliegroute roept vragen op: is
het wel zo dat (meer) regelkennis zorgt voor meer begrip 2.1 Wettelijk kader: strafrecht en onrechtmatige
voor die regels? En zorgt regelkennis vervolgens ook voor daadsrecht
sportief en respectvol gedrag? En daaraan voorafgaand:
hoe staat het eigenlijk met de kennis van de regels bij onze 2.1.1 De doelen van beide rechtsgebieden
(jeugd)voetballers? Wij doelen daarbij niet alleen op ken- Als het gaat om mogelijke reacties op ongewenst gedrag op
nis over spelregels, maar ook over de regels voortvloeiend de velden, komen er wettelijk gezien twee normenstelsels
uit het civiele aansprakelijkheidsrecht en het strafrecht. Het voor toepassing in aanmerking: het strafrecht of het pri-
zijn deze vragen die wij empirisch onderzocht hebben en vaatrecht, meer bepaald het onrechtmatige daadsrecht. Het
die in deze bijdrage centraal staan. onderscheid tussen het onrechtmatige daads- en het straf-
recht wordt zichtbaar bij een vergelijking tussen de doelen
1.2 Het empirisch onderzoek en verdere structuur van beide rechtsgebieden. Het strafrecht kent namelijk een
De twee eerstgenoemde auteurs (juristen) van deze bijdra- ander vertrekpunt dan het privaatrecht. Van oudsher ligt in
ge hebben in samenwerking met de derde en vierde auteur het strafrecht de nadruk op, zoals De Hullu het omschrijft:
(ontwikkelingspsychologen), de afgelopen jaren een (be- ‘vergelding, generale preventie (normbevestiging en af-
perkt opgezet) empirisch onderzoek gedaan naar positief schrikking) en speciale preventie (beveiliging en resociali-
en negatief voetbalgedrag. Er zijn daartoe in het voetbalsei- satie) (…).’11 Tegenwoordig wordt echter ook gewicht toe-
zoen 2014-2015 (onder andere) verschillende vragenlijsten gekend aan het belang van reparatie.12 In zoverre lijkt het
afgenomen bij jeugdspelers van enkele Utrechtse voetbal- strafrecht enigszins naar het privaatrecht toe te schuiven
verenigingen; het ging daarbij om de spelers van elftallen want ‘reparatie’ is privaatrechtelijk een belangrijke functie.
in de A- en C-jeugd. Deze vragenlijsten zijn in het najaar van Maar het strafrecht blijft zich onderscheiden van andere
2015 nader geanalyseerd. Wij lichten dit nader toe in para- rechtsgebieden doordat er ook nog een sterke symbolische
graaf 3 hierna. functie van uitgaat.13 Het signaal dat immers uitgaat van de
In deze bijdrage worden de resultaten van dit voorlopige on- inzet van het strafrecht richting de samenleving als geheel
derzoek naar de invloed van regelkennis op het begrip voor is dat bepaald gedrag niet getolereerd wordt.
de regels en het voetbalgedrag gepresenteerd. Bij de opzet Het onrechtmatige daadsrecht is met name een vergoe-
van het onderzoek is gekozen voor een brede aanpak. Niet dings- en een handhavingssysteem in een particuliere set-
alleen is onderzocht in welke mate kennis over en begrip ting, tussen twee private partijen, dus losstaand van enige
voor voetbalregels van de KNVB bijdragen aan positief voet- maatschappelijke reactie op bepaald gedrag.14 Het reageert
balgedrag, maar ook is gevraagd naar de kennis van straf- als er ‘verkeerd gedrag’ is vertoond,15 en als dat is vastge-
rechtelijke en civielrechtelijke aansprakelijkheidsregels en steld, wordt er een schadevergoeding toegekend. Volgens
de invloed die daarvan uitgaat op het voetbalgedrag. De Engelhard en Van Maanen berust het aansprakelijkheids-
reden daarvoor is dat bij de aanpak van allerlei maatschap- recht op ‘specifieke tot gedragsregels verheven normen die
pelijke problemen in toenemende mate een beroep lijkt te tot doel hebben bepaalde geïndividualiseerde belangen te
worden gedaan op de inzet – en de preventieve werking – beschermen.’16 Het doel daarvan is om de mogelijkheid van
van deze rechtsregels, terwijl veelal nog onbekend is welke herstel te bieden, zodat ‘aanspraken worden gewaarborgd
werking, if any, uitgaat van de regels van deze rechtsgebie- (gehandhaafd) en eigenrichting kan worden voorkomen’.17
den.10 Hoewel er een verband kan worden gelegd tussen handha-
ving van regels aan de ene kant en preventie van bepaald
De structuur van deze bijdrage is als volgt. Wij starten met gedrag door regels aan de andere kant, is een onderscheid
een beschrijving van het juridische kader als het gaat om tussen beide begrippen ook mogelijk en gebruikelijk.18 Het
mogelijke juridische reacties op agressie en deviant gedrag
op en rond het voetbalveld (§ 2). Vervolgens worden de op-
11 J. de Hullu, Materieel strafrecht, Kluwer: Deventer 2012, p. 5.
zet en methodologie van het empirische onderzoek nader 12 De Hullu 2012, p. 5.
omschreven (§ 3). De resultaten hiervan worden gepresen- 13 Zie hierover uitgebreider R. Rijnhout, E. Sikkema & W.S. de Zanger, ‘Straf-
teerd in § 4. Afgesloten wordt met een discussie naar aanlei- rechtelijke en civielrechtelijke aansprakelijkheid van de overheid als
wegbeheerder bij dodelijke ongevallen; de zaak Stichtse Vecht’, RM Themis
ding van de verkregen resultaten (§ 5). 2014, p. 20-25, en aldaar genoemde literatuur.
14 T. Hartlief, ‘Keuzevrijheid in het personenschaderecht’, NJB 2004, p. 1833.
Vgl. Mon. BW B34 (Lindenbergh), p. 7.
15 Mon. BW A15 (Engelhard/Van Maanen), p. 47 e.v. De hiernavolgende pas-
sage bouwt voort op R. Rijnhout, Schadevergoeding voor derden in perso-
10 Zie over het nog niet ontrafelde ‘mysterie’ van de wel of niet aanwezigheid nenschadezaken (dissertatie Utrecht), BJu: Den Haag 2012, p. 335-337.
van preventieve werking als het om het aansprakelijkheidsrecht gaat, o.a. 16 Mon. BW A15 (Engelhard/Van Maanen), p. 12.
I. Giesen, Attributie, juridische causaliteit en preventieve werking, in: W.H. 17 Mon. BW A15 (Engelhard/Van Maanen), p. 12. In het bovenstaande wordt
van Boom, I. Giesen, A.J. Verheij (red.), Capita Civilologie, Boom Juridische over ‘privaatrechtelijke handhaving’ gesproken in de zin van vestiging van
uitgeverij, Den Haag 2013, p. 485-516; G. van Dijck, Aansprakelijkheid en aansprakelijkheid aan de hand van de klassieke meetlat van het onrecht-
defensief gedrag: nuancering gewenst, in: W.H. van Boom, I. Giesen, A.J. matige daadsrecht (in ruime zin, waaronder wij ook verstaan de risicoaan-
Verheij (red.), Capita Civilologie, Boom Juridische uitgeverij, Den Haag sprakelijkheden en bijzondere aansprakelijkheidsregels).
2013, p. 459-484; Asser-Vranken, Algemeen Deel***: een vervolg, Kluwer: 18 Zie echter W.H. van Boom, Efficacious Enforcement In Contract And Tort
Deventer 2005, p. 48 e.v. Zie ook noot 21 hierna. (oratie Rotterdam), Den Haag: BJu 2006.
172 Afl. 5 - oktober 2016 AV&S 2016/34

44
T2_AV&S_1605_bw_V02.indd 172 10/14/2016 3:30:57 PM

Artikelen AGRESSIE VAN JEUGDIGE VOETBALLERS EN DE ROL VAN HET RECHT
onderscheid dat Engelhard in dat verband maakt, is tref- De feiten van deze zaak waren de volgende. Een speler van
fend: handhaving kan gaan over handhaving van privaat- Sparta zet met hoge snelheid een vliegende tackle in met
rechtelijke normen, maar ook over handhaving door pri- een gestrekt been op een halve meter hoogte. Hij raakt
vaatrechtelijke normen.19 Die laatste vorm van handhaving daarbij een speler van Go Ahead Eagles op het been, met
gaat ook over preventie van gedrag door middel van de inzet een gecompliceerde beenbreuk als gevolg. Die beenbreuk
van het aansprakelijkheidsrecht. Het feit dat een norm zegt zorgt ervoor dat de voetballer van Go Ahead Eagles vroeg-
dat x of y niet mag, leidt ertoe, zo is de claim, dat zulk ge- tijdig zijn carrière moet beëindigen. Vaststaat dat de speler
drag niet wordt vertoond teneinde te voorkomen dat schade van Sparta voetbalschoenen met metalen noppen droeg.
wordt toegebracht die vervolgens vergoed zou moeten wor- Ook wist hij dat dit type voetbalgedrag een risico op ern-
den. En hoewel er juist over die laatste variant in de litera- stig letsel zou opleveren. De scheidsrechter heeft bovendien
tuur de meeste discussie bestaat,20 gaan er desondanks in verklaard dat hij de actie van de Spartaspeler beschouwt als
de literatuur ook steeds meer stemmen op om het civiele een ‘aanval’.
aansprakelijkheidsrecht als preventief middel in te zetten.21 In de zaak die voorligt bij de Hoge Raad rijzen twee vragen.
In het licht daarvan zou dan betoogd kunnen worden dat Ten eerste de vraag of het gedrag van de speler van Sparta
het stafrechtelijke en civielrechtelijke aansprakelijkheids- gekwalificeerd kan worden als het opzettelijk toebrengen
recht ook een gezamenlijke functie hebben, namelijk om van zwaar lichamelijk letsel. In dit verband stelt de Hoge
ongewenst gedrag te voorkomen (preventie). Het is dan ook Raad voorop dat het feit dat sprake is van een sport- of spel-
die functie die hierna centraal staat.22 Maar eerst zullen we situatie er niet toe leidt dat andere maatstaven gelden voor
een nadere blik werpen op de specifieke regels als het om de beoordeling van het opzetvereiste.25 De vraag of hieraan
sport- en spelsituaties gaat. is voldaan moet worden vastgesteld aan hand van de regu-
liere vereisten van – in dit geval: voorwaardelijke – opzet.
2.1.2 Strafrechtelijke regels in sportsituaties Het hof oordeelde in hoger beroep dat voorwaardelijke op-
De concrete vraag naar wanneer strafrechtelijke aanspra- zet niet alleen kan worden afgeleid uit verklaringen, maar
kelijkheid kan bestaan, valt uiteen in vier deelvragen. Ten ook uit de aard van de gedraging en de omstandigheden
eerste moet het ten laste gelegde kunnen worden bewe- waaronder de gedraging is verricht.26 De Hoge Raad over-
zen. Vervolgens moet voldaan zijn aan de vereisten van een weegt dat het hof geen blijk heeft gegeven van een onjuiste
strafrechtelijke delictsomschrijving en moet het bewezen- rechtsopvatting door te oordelen dat in het licht van de om-
verklaarde feit wederrechtelijk zijn (is het feit strafbaar?). standigheden van het geval in dit geval sprake is van voor-
Ten derde moet de verdachte strafbaar zijn (is er sprake van waardelijke opzet, omdat sprake is van een flagrante over-
verwijtbaarheid?). En als is voldaan aan deze vereisten, dan treding van de voetbalregels.
rijst de (vierde) vraag welke straf of maatregel opgelegd De tweede vraag betreft die naar de wederrechtelijkheid.
moet worden.23 Voor de strafrechtelijke aansprakelijkheid Net zoals in het privaatrecht, zoals we nog zullen zien, geldt
voor gedrag op en rond het voetbalveld zijn in het bijzonder ook in het strafrecht dat niet alle gedragingen binnen een
de vereisten van ‘wederrechtelijkheid’ en ‘verwijtbaarheid’ sport- of spelsituatie die gevaarlijk zijn ook strafrechtelijke
van belang. Die twee vereisten worden besproken aan de aansprakelijkheid opleveren. ‘It’s all in the game’, zo wordt
hand van het arrest van de Hoge Raad 22 april 2008 dat over ook wel gesteld.27 Door de vrijwillige deelname aan bijvoor-
een zaak van overmatige agressie in het voetbal handelde.24 beeld een voetbalwedstrijd, wordt geaccepteerd dat zich
bepaalde risico’s kunnen materialiseren als gevolg van ge-
19 E.F.D. Engelhard, ‘Handhaving van en door het privaatrecht’, in: E.F.D.
drag dat nog valt binnen de sfeer van datgene dat geoorloofd
Engelhard, I. Giesen e.a. (red.), Handhaving van en door het privaatrecht, is.28 De Hoge Raad overweegt in dat verband:
BJu: Den Haag 2009, p. 12-13. Zie over dat onderscheid in dezelfde bundel
ook I. Giesen, Handhaving in, via, door en met het privaatrecht: waar staan
we nu?, p. 307-329.
“De deelnemers van een sport, zoals voetbal, hebben
20 J.S. Kortman, Tort law industry (oratie UVA), Amsterdam: Vossiuspers UVA immers tot op zekere hoogte gevaarlijke gedragingen
2008 en J.S. Kortman en C.H. Sieburgh, ‘Rechtshandhaving door privaat- waartoe het spel uitlokt over en weer van elkaar te ver-
recht’, in: Toezicht op de financiële markt en de energiemarkt. Staatsge-
heimen en het strafproces. Rechtshandhaving door het privaatrecht, Den
wachten, terwijl bij een door duidelijke spelregels afge-
Haag: Bju 2009. bakende sport die spelregels mede van belang zijn voor
21 Zie o.a. M. Faure en T. Hartlief, Nieuwe risico’s en vragen van aansprakelijk- het bepalen van de grenzen van wederrechtelijkheid.”29
heid en verzekering, Deventer: Kluwer, p. 8; W.H. van Boom, ‘Effectuerend
handhaven in het privaatrecht’, NJB 2007, p. 982-991. Door Engelhard en Van
Maanen wordt preventie als ‘nevenfunctie’ genoemd, Mon. BW A15, p. 13- Overwogen wordt vervolgens echter dat bij dusdanige
15. Zie echter ook J.S. Kortman en C.H. Sieburgh, ‘Rechtshandhaving door schending van de spelregel en gevaarlijk handelen geen
privaatrecht’, in: Toezicht op de financiële markt en de energiemarkt. Staats-
geheimen en het strafproces. Rechtshandhaving door het privaatrecht, Den
sprake kan zijn van ontbreken van wederrechtelijkheid.30
Haag: Bju 2009; Giesen 2009; Giesen 2013, m.n. p. 506-508; Y. Bacharias,
S.D. Lindenbergh en P. Mascini, ‘Preventie van arbeidsuitval’, in: Van Boom,
I. Giesen & A.J. Verheij (red), Capita civilologie. Handboek empirie en pri- 25 HR 22 april 2008, NJ 2008/375, m.nt. N. Keijzer, r.o. 4.3. Vgl. HR 28 juni 2011,
vaatrecht, Den Haag: BJu 2013, p. 585-603. ECLI:NL:HR:2011:BQ4203, r.o. 2.3.
22 Het onrechtmatige daadsrecht onderscheidt zich echter in zoverre van het 26 Hof Den Haag 29 mei 2006, NJFS 2006/225.
strafrecht dat het geen punitieve doel heeft. 27 Kelk en De Jong 2013, p. 190.
23 C. Kelk en F. de Jong, Studieboek materieel strafrecht, Deventer: Kluwer 28 Kelk en De Jong 2013, p. 190.
2013, p. 67. 29 HR 22 april 2008, NJ 2008/375, m.nt. N. Keijzer, r.o. 4.5.
24 HR 22 april 2008, NJ 2008/375, m.nt. N. Keijzer. 30 HR 22 april 2008, NJ 2008/375, m.nt. N. Keijzer, r.o. 4.5.
AV&S 2016/34 Afl. 5 - oktober 2016 173

45
T2_AV&S_1605_bw_V02.indd 173 10/14/2016 3:30:57 PM

Concrete voorbeelden waarin de regels van de Hoge Raad zorgvuldigheid. De rechtvaardiging daarvoor is dat, zoals
worden toegepast, zijn te vinden in de lagere rechtspraak.31 de Hoge Raad het verwoordt:
Een voorbeeld van een veroordeling, in dit geval op grond
van het delict ‘zware mishandeling’, biedt de zaak waarin “deelnemers aan die sport of dat spel in redelijkheid tot
een voetballer een tegenspeler eerst een duw gaf, waardoor op zekere hoogte gevaarlijke, slecht gecoördineerde, ver-
die tegenspeler op de grond komt te liggen, en de voetbal- keerd getimede of onvoldoende doordachte handelingen
ler de tegenspeler vervolgens vol in het gezicht trapt.32 Er of gedragen waartoe de activiteit uitlokt of die daarin
wordt echter niet altijd tot zware mishandeling geconclu- besloten liggen, van elkaar [moeten] (…) verwachten.”36
deerd. Een struikelblok daarbij lijkt vooral het opzetvereiste
te zijn. Met name als het letsel het gevolg is van een tackle, Deze norm betekent de facto dat alleen bij opzettelijk of met
lijkt het moeilijk vast te stellen dat die tackle opzettelijk is bewuste roekeloosheid veroorzaakt letsel sprake zal zijn
ingezet om daarmee ernstig letsel te veroorzaken. Het oor- van een onrechtmatige daad als de gedraging zich binnen
deel daarover is sterk met de feiten verweven, waarbij ten de sport- of spelsituatie voordoet.
eerste een rol lijkt te spelen hoe de scheidsrechter de situ- Een voorbeeld hiervan biedt het Natrap-arrest.37 De feiten
atie heeft waargenomen, ten tweede of er redelijkerwijs op van dit arrest waren de volgende. Tijdens een voetbalwed-
de bal gespeeld kan zijn en, ten slotte, onder welke omstan- strijd schopt Dekker, terwijl de bal niet meer in de buurt
digheden de gedraging is uitgevoerd (sportieve wedstrijd of is, Van der Heijde. Van der Heijde loopt daardoor ernstig
grimmige sfeer).33 knieletsel op. De Hoge Raad overweegt dat het hof in hoger
Als is vastgesteld dat de verdachte strafbaar is, moet de beroep voldoende duidelijk heeft gemaakt dat het in deze
straf worden bepaald. Het doel van het strafrecht (leedtoe- casus niet gaat om een gevaar dat eigen is aan het voetbal-
voeging en – tot op zekere hoogte – preventie) komt tot op spel.38 De gedraging bevindt zich duidelijk buiten de regels
zekere hoogte overeen met het doel van het tuchtrecht van van het spel en is abnormaal gevaarlijk.39 Overwogen wordt
de KNVB (zoals wij nog zullen zien in paragraaf 2.2.). Ver- dat de enkele overtreding van de spelregels niet direct de
meldenswaardig is daarom dat er twee uitspraken te vinden conclusie rechtvaardigt dat sprake is van een onrechtmatige
zijn waarin de rechter bij de straftoemeting rekening houdt gedraging in de zin van art. 6:162 BW, maar ‘de overtreding
met reeds door de KNVB opgelegde disciplinaire maatrege- van een spelregel [is wel] een factor die meeweegt bij de be-
len.34 oordeling van de onrechtmatigheid’, aldus de Hoge Raad.40
Wederom zijn in de lagere rechtspraak voorbeelden te vin-
2.1.3 Regels van buitencontractuele aansprakelijkheid in den van abnormaal voetbalgedrag. Eén daarvan is de situ-
sportsituaties atie waarin een keeper met één of twee gestrekte benen
Het onrechtmatige daadsrecht wordt (grotendeels) be- inkomt op een veldspeler, die de bal al op het doel heeft
heerst door art. 6:162 BW, zeker in de gevallen waar het geschoten en scoort. De veldspeler loopt een beenbreuk op
hier om gaat.35 De vereisten die in dat verband gelden, zijn met blijvend letsel als gevolg. Het hof in de betreffende zaak
een onrechtmatige gedraging, toerekenbaarheid, schade, oordeelt dat de gedraging plaatsvond na het schot van de
causaliteit en relativiteit. Als aan die vereisten is voldaan, is veldspeler en het schot op deze manier niet meer gestopt
toekenning van een schadevergoeding in beginsel mogelijk, had kunnen worden, de gedraging niet meer had kunnen
waardoor tevens private handhaving van de norm plaats- voorkomen dat er gescoord werd en de keeper had kunnen
vindt, tenzij de aangesproken partij een rechtvaardigings- afzien van zijn actie.41 Op dergelijk gedrag hoeft de veldspe-
grond kan inroepen. ler redelijkerwijs niet bedacht te zijn, en dus is er sprake van
Vooral het vereiste ‘onrechtmatige gedraging’ is relevant in een onrechtmatige gedraging.42
gevallen waarin letsel veroorzaakt werd door het voetbal- De verhoogde drempel voor aansprakelijkheidsrecht geldt
spel. De Hoge Raad heeft namelijk een verhoogde drempel overigens niet alleen tijdens een wedstrijd. Ook tijdens
voor aansprakelijkheidsrecht gecreëerd voor sport- en spel-
situaties, door lichtere eisen te stellen aan de te betrachten
36 HR 28 maart 2003, NJ 2003/719, m.nt. C.J.H. Brunner (Broere/Kegel), r.o.
3.7. Vgl. HR 19 oktober 1990, NJ 1992/621 (Heck en Nijgh), r.o. 3.3; HR 28
31 In één uitspraak wordt de verdachte veroordeeld wegens poging tot dood- juni 1991, NJ 1992/622, m.nt. C.J.H. Brunner (Dekker/Van der Heide), r.o. 3.3;
slag, Rb. Dordrecht 29 maart 2007, ECLI:NL:RBROD:2007:BA1971. De be- HR 28 maart 2003, NJ 2003/718 (Hettinga/De Vries), r.o. 3.6; HR 20 februari
treffende speler gaf het slachtoffer eerst een kopstoot om hem vervolgens 2004, VR 2004, 81 (Verberne/Megens), r.o. 3.4.
vol in zijn gezicht te trappen. 37 HR 28 juni 1991, NJ 1992/622, m.nt. C.J.H. Brunner (Dekker/Van der Heide).
32 Rb. Dordrecht 3 mei 2007, ECLI:NL:RBDOR:2007:BA4370. 38 HR 28 juni 1991, NJ 1992/622, m.nt. C.J.H. Brunner (Dekker/Van der Heide),
33 Hof Den Bosch 4 februari 2009, ECLI:NL:GHSHE:2009:BH3371 (vrijspraak); r.o. 3.3.
Hof Leeuwarden 29 oktober 2009, ECLI:NL:GHLEE:2009:BK1553; Rb. Den 39 In deze zin oordeelde het hof en de Hoge Raad laat dit oordeel in stand.
Haag 29 maart 2010, ECLI:NL:RBSGR:2010:BL9377 (veroordeling). 40 HR 28 juni 1991, NJ 1992/622, m.nt. C.J.H. Brunner (Dekker/Van der Heide),
34 Rb. Groningen 3 juli 2008, ECLI:NL:RBGRO:2008:BD6244; Rb. Den Haag 29 r.o. 3.3. Over deze uitspraak zie o.a. N.K. van Mullem, ‘Aansprakelijkheid bij
maart 2010, ECLI:NL:RBSGR:2010:BL9377, r.o. 6.3. sport en spel’, AA 2003, p. 167-168.
35 Titel 6.3 BW, waarvan art. 6:162 BW deel uit maakt, kent verschillende 41 Hof Amsterdam 15 november 2011, ECLI:NL:GHAMS:2011:BX7845, r.o.
rechtsgronden voor aansprakelijkheid, zoals de aansprakelijkheid voor het 3.13. Een soortgelijk geval speelde in Rb. Haarlem 11 februari 2009,
gedrag van anderen en productaansprakelijkheid. Voor de analyse die hier ECLI:NL:RBHAA:2009:7675.
wordt gemaakt, is vooral art. 6:162 BW van belang. Over hetgeen hierna 42 Andere voorbeelden aansprakelijkheid en voetbalgedrag bieden de uit-
besproken wordt (aansprakelijkheid bij sport en spel) verscheen recent een spraken Rb. Arnhem 23 april 2008, ECLI:NL:RBARN:2008:BD1779; Rb. Den
zeer uitvoerige bijdrage van Paijmans, waarnaar wij graag verwijzen: B.M. Haag 6 februari 2008, ECLI:NL:RBSGR:2008:BC3860; Rb. Haarlem 11 febru-
Paijmans, Sport en spel. In de verlenging?, Letsel & Schade 2015/4/368. ari 2009, ECLI:NL:RBHAA:2009:7675.
174 Afl. 5 - oktober 2016 AV&S 2016/34

46
T2_AV&S_1605_bw_V02.indd 174 10/14/2016 3:30:57 PM

trainingen, gedurende pauzes of luwe momenten tijdens spelers om die regel na te volgen; maar daarbij verdient dan
een wedstrijd (bijvoorbeeld: bij tennis tussen de games in), wel nog opmerking dat de spelers de regels behoren te ken-
korte tijd na het einde van een spel, of bij aanschouwen van nen. De vervolgvraag is alleen of deze regels, en de kennis
andere deelnemers, geldt die aangescherpte norm.43 Zolang daarvan, ook leiden tot vermindering van onwenselijk voet-
iemand nog niet zijn hoedanigheid als deelnemer aan het balgedrag. Die vraag kan ook worden gesteld ten aanzien
spel heeft verloren, geldt dus een verhoogde drempel voor van het strafrecht en het onrechtmatige daadsrecht, welke
aansprakelijkheid. ook (mede) de preventie van ongewenst gedrag als doel lij-
ken te hebben. Het is dan ook die vraag die het vertrekpunt
2.2 Zelfregulerend kader: tuchtrecht en maatregelen vormde voor ons empirisch onderzoek.
door de KNVB
Het beleid van de KNVB is gericht op zowel repressie als pre- 3. De onderzoeksopzet
ventie van deviant gedrag op de velden zoals al bleek in pa-
ragraaf 1.1. Door middel van het opstellen van een normatief Als onderdeel van een omvangrijker onderzoek,49 hebben
kader zou de orde kunnen worden hersteld en preventieve wij onderzoek gedaan naar de relatie tussen regelkennis
en stimulerende acties zouden de verenigingen moeten on- en voetbalgedrag bij jeugdvoetballers. Bij drie voetbalver-
dersteunen die het voetbal aantrekkelijk willen houden.44 enigingen in Utrecht zijn aan jeugdteams vragenlijsten uit-
Tot op zekere hoogte lopen de doelen van zelfregulering gedeeld. Van elke voetbalvereniging zijn vier jeugdteams
door de KNVB dus parallel met die van het strafrecht,45 met geselecteerd, twee A-teams en twee C-teams, met dien ver-
dien verstande dat degene die handhavend optreedt, een stande dat bij één van deze drie verenigingen drie C-teams
andere entiteit is, en de verwachting bestaat dat de KNVB zijn geselecteerd. In totaal hebben dus dertien voetbalteams
ook bij lichtere vergrijpen handhavend zal (kunnen) optre- geparticipeerd in het onderzoek. In het najaar van 2014 is
den, anders dan het OM (vanwege het opportuniteitsbegin- de eerste vragenlijst afgenomen, begin 2015 is een tweede
sel). vragenlijst afgenomen en een paar maanden later, in april
Voorbeelden van repressieve acties, in 2013 geïntroduceerd, 2015, de derde vragenlijst. Die derde vragenlijst is de lijst
zijn het directer straffen (bij een eerste gele kaart al een die in dit onderzoek centraal staat. Deze vragenlijst bestaat
tijdstraf) 46 en de bevoegdheid van de tuchtorganen van de uit twee delen. Het eerste deel betrof een onderzoek onder
KNVB om voetballers die zich schuldig maken aan misstan- de jeugdspelers naar hun kennis van regels en sancties, de
den, direct en zwaar te straffen. Tot slot mogen districts- mate van zekerheid over die regelkennis, en hun begrip voor
besturen verenigingen uit de competitie halen als daartoe de regels. Het tweede deel van de vragenlijst betrof een ont-
aanleiding bestaat.47 wikkelingspsychologisch onderzoek naar zogenoemd pro-
Preventieve maatregelen die in 2013 zijn genomen, betreffen sociaal en/of agressief voetbalgedrag.
het verplichte spelregelbewijs voor alle B-junioren, de inzet In het eerste deel zijn in de vragenlijst zes situatieschetsen
van extra waarnemers bij recidiverende clubs, preventieve opgenomen, waarin één van de spelers een overtreding
maatregelen bij verenigingen die bovengemiddeld negatief maakt. Een deel van deze casus was ontleend aan de vra-
scoren op het punt van sportiviteit en respect, de verplichte gen die worden gesteld in het kader van het spelregelsbe-
cursus ‘sport en agressie’ voor misdragende voetballers bij wijs (casus 1, 3 en 550 ). De andere casus zijn ontleend aan de
zware overtredingen, de publicatie van tuchtrechtuitspra- rechtspraak (casus 2, 4, en 651).
ken en de oproep aan verenigingen om gedragsregels te Het eerste item op de vragenlijst (item A) betrof de vraag
publiceren. Ten aanzien van die laatste maatregel merkt de naar de kennis van de sanctie die moet volgen op bepaald
KNVB op dat de bekendheid met regels ‘een voorwaarde [is] voetbalgedrag. Zes keuze-opties zijn gegeven, waarbij werd
om misdragingen te kunnen benoemen en bestraffen.’48 In- aangegeven dat meerdere antwoorden mogelijk zijn:
dien een regel onbekend is, wordt het immers lastig voor de
43 HR 19 oktober 1990, NJ 1992/621 (Heck en Nijgh), r.o. 3.3; HR 28 maart

2003, NJ 2003/718 (Hettinga/De Vries), r.o. 3.6-3.7; HR 28 maart 2003, NJ
2003/719, m.nt. C.J.H. Brunner (Broere/Kegel), r.o. 3.7; HR 20 februari 2004,
VR 2004, 81 (Verberne/Megens), r.o. 3.3.
44 KNVB, Tegen geweld voor sportiviteit, 2013.
45 Zie over de relatie tussen zelfregulering en civielrechtelijke aansprakelijk-
heid o.a. I. Giesen, Alternatieve regelgeving en privaatrecht, Kluwer: Deven-
ter 2007, m.n. p. 68-81, en specifiek voor sportsituaties L. Westhoff, ‘Zelf- 49 Zie hiervoor noot 1.
regulering: een belangrijke bron voor civiele aansprakelijkheid binnen de 50 Casus 1 betrof een handsbal door een verdediger die een doelpunt voor-
sport’, Tijdschrift voor Sport & Recht 2013-4, p. 70 e.v., alsmede R. van Kleef, komt, casus 3 ging over een aanvaller die een keeper belet de bal in het
‘De invloed van regelgeving van sportorganisaties op het civiele aanspra- spel te brengen, en casus 5 zag op het geval dat een verdediger schuin van
kelijkheidsrecht’, Ars Aequi 2016, p. 545 e.v., met verdere verwijzingen (o.a. achteren een sliding inzet en daarbij de aanvaller verwondt.
naar het proefschrift van dezelfde auteur). 51 Casus 2 betreft een keeper die met gestrekte benen een aanvaller raakt
46 KNVB, Tegen geweld voor sportiviteit, 2013. nadat die al op doel heeft geschoten (ontleend aan Rb. Haarlem 11 febru-
47 KNVB, Tegen geweld voor sportiviteit, 2013. ari 2009, ECLI:NL:RBHAA:2009:7675, casus 4 ging over een speler die een
48 KNVB, Tegen geweld voor sportiviteit, 2013. Andere maatregelen die in tegenspeler naar de grond duwt en hem dan in het gezicht trapt (ontleend
2013 zijn geïntroduceerd, zijn de verplichting dat de scheidsrechter na de aan Rb. Dordrecht 3 mei 2007, ECLI:NL:RBDOR:2007:BA4370), en casus 6
wedstrijd samen met beide aanvoerders het wedstrijdformulier onderte- zag op het geval dat een omspeelde verdediger in zijn draai tegen de kuit
kent en de introductie van een hulplijn voor noodgevallen en een Meldpunt van de aanvaller trapt en deze daarbij verwondt (ontleend aan Hof Leeu-
wanordelijkheden. warden 29 oktober 2009, ECLI:NL:GHLEE:2009:BK1553).
AV&S 2016/34 Afl. 5 - oktober 2016 175

47
T2_AV&S_1605_bw_V02.indd 175 10/14/2016 3:30:57 PM

– Agressie gericht op het eigen team;

1. Er volgt geen sanctie
– Agressie gericht op officials en omstanders.
2. Er volgt een vrije trap of strafschop, maar de speler
krijgt geen kaart
Uit analyse bleek dat in deze steekproef alle vormen van
3. Er volgt een vrije trap of strafschop, en de speler
gedrag zich van elkaar onderscheiden zoals verwacht, be-
krijgt een gele kaart
halve de agressie gericht op officials en omstanders. Deze
4. Er volgt een vrije trap of strafschop, de speler krijgt
laatste vorm van gedrag vormt samen met agressie gericht
een rode kaart en wordt geschorst voor een aantal
op het eigen team één aspect van gedrag. Door de beper-
wedstrijden
kingen van de steekproef is ervoor gekozen toch ook deze
5. De speler moet de kosten van de geblesseerde te-
vijfde vorm van gedrag afzonderlijk te analyseren. Alle
genspeler vergoeden
vormen van gedrag bleken betrouwbaar gemeten, hetgeen
6. De speler krijgt een werkstraf of gevangenisstraf
aangeeft dat de scores niet afhingen van toevallige fluc-
tuaties.
In situatie 1, 3 en 5 was er één goed antwoord mogelijk en In de analyse is vervolgens onderzocht of er een relatie be-
betrof het antwoord altijd een sanctie vanuit de KNVB (res- staat tussen kennis van de sancties (zoals die werd geïnven-
pectievelijk antwoord 2 of 3). Situatie 2, 4 en 6 betroffen ca- tariseerd in het eerste gedeelte van de vragenlijst) en het
sussen waarin, naast antwoord 4, antwoord 5 en/of 6 ook voetbalgedrag van de spelers.
goed was. In twee casus (1 en 3) was geen sprake van schade De vragenlijst is ingevuld door 114 respondenten (N=114),
en daarin werd keuzeoptie 5 dus weggelaten. waarvan 44.7% A-jeugd en 56.3% C-jeugd. Voor de vraag naar
Vervolgens werd aan elke deelnemer gevraagd (item B) hoe de generaliseerbaarheid van dit onderzoek naar de Utrechtse
zeker hij was van zijn antwoord, waarbij de volgende op- A- en C-jeugd of zelfs de landelijke A- en C-jeugd, moet on-
ties werden gegeven: heel zeker (gecodeerd als 1), zeker derscheid gemaakt worden naar twee typen resultaten. Bij de
(=2), niet zeker niet onzeker (=3), onzeker (=4) en heel onze- resultaten over de verdeling van kennis of gedrag, dat willen
ker (=5). Ten slotte werd de vraag gesteld of de betreffende zeggen hoe vaak iets voorkomt of hoe sterk iets voorkomt,
respondent het terecht vindt dat de sanctie wordt opge- kunnen geen generaliserende uitspraken naar Utrechtse of
legd, waarbij ook weer een keuze werd gegeven tussen vijf Nederlandse A- en C-jeugd worden gedaan. Als het gaat om
opties: heel terecht (=1), terecht (=2), niet terecht niet on- verbanden, bijvoorbeeld een verband tussen kennis en ge-
terecht (=3), onterecht (=4) en heel onterecht (=5) (item C). drag, dan geven de resultaten een indruk over de verbanden
Het tweede deel van de vragenlijst betrof, zoals gezegd, die kunnen worden verwacht in de populatie (de Utrechtse
een onderzoek naar pro-sociaal en/of agressief voetbalge- of zelfs landelijke A- en C-jeugd). Of die verbanden dan ook
drag. Pro-sociaal gedrag is vrijwillig gedrag dat bedoeld bij de hele populatie zijn terug te vinden en met name hoe
is om ten goede te komen aan een ander individu of een sterk, hangt af van hoe specifiek de kenmerken zijn van de
groep van individuen of om hen te helpen. Voorbeelden steekproef in dit onderzoek. Doorgaans geldt hoe sterker het
zijn het helpen van een geblesseerde tegenstander (pro- verband in een selectieve steekproef als deze, hoe groter de
sociaal gedrag richting de tegenstander), en het felicite- kans op replicatie van het verband in de populatie.
ren van een teamgenoot (pro-sociaal gedrag richting het
eigen team). Agressie of antisociaal gedrag is vrijwillig 4. De verkregen onderzoeksresultaten
gedrag met de bedoeling anderen te benadelen of schade
toe te brengen. Voorbeelden zijn een tegenstander pro- 4.1 Regelkennis bij jeugdvoetballers en hun (zelf)
beren te blesseren (agressie gericht op de tegenstander), vertrouwen daarover
ruzie maken met een teamgenoot (agressie gericht op Uit de analyses blijkt dat de jeugdvoetballers de vraag over
het eigen team), en schelden tegen de scheidsrechter, of de regelkennis (vraag A) gemeten door middel van zes ver-
schelden tegen de supporters (agressie gericht op officials schillende situaties, doorgaans verschillend beantwoorden
en omstanders). 52 per situatie: hoewel elke vraag regelkennis toetst, laten de
De vragenlijst behelsde een zelfrapportage (self-report), jongens geen systematische kennis zien over de verschil-
waarin de respondenten moesten aangeven in welke mate lende situaties heen. Het voor elk individu ‘optellen’ van zijn
zij bepaald gedrag vertonen op het voetbalveld. Het volgen- scores over de casus heen (hoeveel van de zes vragen heeft
de gedrag werd gemeten: elke jeugdvoetballer goed?) geeft daarmee geen valide en
betrouwbare score. Daarom is ervoor gekozen om de ana-
– Pro-sociaal gedrag richting de tegenstander; lyses voor elk van de zes situaties apart uit te voeren. Als
– Pro-sociaal gedrag richting het eigen team; een respondent de situatie goed heeft beantwoord volgens
– Agressie gericht op de tegenstander; de antwoordsleutel, krijgt hij een score 1. Als hij het fout
heeft, krijgt hij een score 0. Dat betekent dat de respondent
52 De eerste vier voorbeelden van antisociaal gedrag zijn ontleend aan M.
voor die situaties waarop meerdere goede antwoorden te
Kavussanu, N. Stanger, I. D. Boardley, ‘The prosocial and antisocial beha- geven zijn (bijvoorbeeld: een schorsing én compensatie) pas
viour in sport scale: Further evidence of construct validity and reliability’, de score 1 krijgt als hij beide antwoorden heeft omcirkeld.
Journal of Sport Sciences 2013, vol. 31, nr. 11, p. 1208. De laatste categorie,
agressie gericht op officials en omstanders, heeft Van Hoof (één van de au-
Deze analyse leverde het volgende, in Tabel 1A verzamelde
teurs van dit stuk) zelf ontwikkeld en toegevoegd. beeld op:
176 Afl. 5 - oktober 2016 AV&S 2016/34

48
T2_AV&S_1605_bw_V02.indd 176 10/14/2016 3:30:57 PM

Tabel 1. Aantallen (N) en percentages (%) goed en fout per afhangen van het totaal aantal goede antwoorden, maar bij-
situatie voor voetbalregels en sancties (Tabel 1A), en voor voorbeeld van een minimaal aantal goede antwoorden per
enkel de voetbalregels (Tabel 1B) categorie van spelregels.54
1A: Goed en fout voor regels 1B: Goed en fout Totaal: N

Als vervolgens naar een relatie wordt gezocht tussen de
en sancties voor voetbalregels
juiste antwoorden en de mate van zekerheid, respectievelijk
Situatie Goed: Fout: N (%) Goed: Fout:
de onjuiste antwoorden en de mate van zekerheid worden
N (%) N (%) N (%)
niet altijd statistisch significante verschillen vastgesteld.55
1 29 (25.4) 85 (74.6) 29 (25.4) 85 (74.6) 114 En als dat wel het geval is, zijn die verschillen tegengesteld
2 2 (1.8) 112 (98.2) 55 (48.2) 59 (51.8) 114 van aard.56 De groep die de situatie goed heeft beantwoord,
3 43 (37.7) 71 (62.3) 43 (37.7) 71 (62.3) 114 is gemiddeld genomen niet zekerder van het antwoord dan
de groep die de situatie niet goed heeft ingeschat. Met an-
4 3 (2.6) 111 (97.4) 64 (56.1) 50 (43.9) 114
dere woorden: de overtuiging van het eigen gelijk omtrent
5 48 (42.1) 66 (57.9) 48 (42.1) 66 (57.9) 114 de voetbalregels is niet recht evenredig met de juiste kennis
6 0 (0) 114 (100 ) 35 (30.7) 79 (69.3) 114 van de regels (en sancties). Ook hier zou inzicht bij de voet-
ballers in het gegeven dat overtuiging van het eigen gelijk
Opvallend aan deze analyse van de regelkennis is dat rela- niet altijd betekent dat het gelijk ook aan de eigen zijde is,
tief weinig jeugdvoetballers de regels goed kennen. Zeker kunnen helpen. Dat inzicht zou kunnen leiden tot een min-
in situaties 2, 4 en 6 – daarin waren meerdere antwoorden der heftige verdediging van het “eigen gelijk” op en rond het
mogelijk – wordt erg laag gescoord (0 tot 2.6% komt slechts voetbalveld. Overigens is er in deze data geen ondersteu-
tot het juiste antwoord). Het aantal jongens met een goed ning voor de relatie zeker/heel zeker zijn van je antwoord en
antwoord is per situatie te klein om de leeftijdsgroepen (A- gedrag. Anders gezegd, vanuit deze data kun je niet zeggen
en C-jeugd) te vergelijken. dat degenen die overtuigd zijn van hun gelijk, ook agressie-
In een tweede analyse is vervolgens wederom gekeken naar ver gedrag vertonen op het veld.
de situatie goed/fout, maar dan alleen op het niveau van de
kennis van voetbalregels. Met andere woorden, in situaties 4.2 Relatie tussen het begrip voor een sanctie en de
2, 4 en 6 werd een antwoord (ook al) goed gerekend als ie- regelkennis
mand voor die situatie (alleen) antwoord 4 (de voetbalspel- In item C is de vraag gesteld hoe terecht de respondent de
regel die hier geldt) had omcirkeld en niet ook nog antwoord sanctie vindt die hij in item A heeft omcirkeld. Het idee be-
5 en/of 6 (over de sancties). De verschillen tussen de spelers stond dat spelers die de regels kennen, daarvoor ook meer
kwamen hiermee beter naar voren, zo blijkt uit Tabel 1B. begrip zullen kunnen opbrengen. Aangezien de responden-
Toch is ook wat betreft de pure voetbalregels het aantal res- ten heel slecht scoren op hun kennis over het civielrechte-
pondenten dat deze goed heeft beoordeeld, niet hoog: het lijke en strafrechtelijke aansprakelijkheidsrecht, wordt hier
varieert tussen de 25 en 56%. Wanneer bij deze analyses de alleen ingegaan op hun begrip voor de voetbalregels.
leeftijdsgroepen (A-jeugd ten opzichte van C-jeugd) onder- Er blijken bij drie van de zes situaties statisch significante
ling worden vergeleken, levert dat geen statistisch signifi- verschillen te bestaan, namelijk in situatie 2, 4 en 6. In situ-
cante verschillen op.53 aties 2, 4 en 6 vindt de groep die de situatie goed heeft be-
antwoord, gemiddeld genomen de sanctie vaker terecht dan
Item B van de vragenlijst ging vervolgens over de mate
waarin de respondenten zeker zijn van hun antwoord op
item A. Zoals gezegd, een relatief (zeer) klein deel van de
respondenten kent het juiste antwoord op de vragen. Toch
is een meerderheid van de respondenten (over de situaties
variërend van 70.2 tot 92.1%) zeker of heel zeker van het ei-
54 Voorzichtigheid is geboden: wij hebben maar zes verschillende regels be-
gen antwoord op item A en denkt dus de kennis in huis te trokken in ons onderzoek.
hebben. Dat maakt dat feedback nodig is, waardoor de spe- 55 In situaties 3, 5 en 6 zijn geen statistisch significante verschillen vastge-
lers inzicht krijgen in hun gebrekkige kennis van de regels. steld; situatie 3: t(112)= 0.46, p = 0.65 (2-tailed); situatie 5: t(112)= -0.75,
p = 0.46 (2-tailed); situatie 6: t(111)= 1.04, p = 0.30 (2-tailed). Bij situaties
Die feedback kan eerder gegeven worden door bijvoorbeeld 1, 2, en 4 werden statistisch significante verschillen gevonden, maar niet
eerder dan met de B-jeugd te starten met het regelbewijs, systematisch in dezelfde richting, zie volgende noot.
waardoor hiaten in regelkennis eerder bekend worden. Ook 56 Situatie 1: t(111)= -2.06, p = 0.042 (2-tailed). De groep die de voetbalregel
goed heeft, is gemiddeld genomen minder zeker van zijn antwoord dan
kan het criterium voor het behalen van het bewijs worden de groep die de voetbalregel fout heeft (resp. gemiddeld 1.66 (N = 29, SD =
verhoogd. Verder blijkt uit onze onderzoeksresultaten dat 0.67) en 1.38 (N = 84, SD = 0.60); waarbij score 1 = heel zeker). Situatie 2:
spelregels niet één dimensie vertegenwoordigen, waardoor t(110)= 4.13, p < 0.001 (2-tailed). De groep die de voetbalregel goed heeft,
is gemiddeld genomen zekerder van zijn antwoord dan de groep die de
het behalen van het spelregelbewijs niet alleen zou mogen voetbalregel fout heeft (resp. gemiddeld 1.67 (N = 55, SD = 0.77) en 2.33 (N =
57, SD = 0.91); waarbij score 1 = heel zeker). Situatie 4: t(111)= 3.14, p = 0.002
53 Situatie 1: χ2(1) = 0.73, p = 0.20, N= 114; Situatie 2: χ2(1) = 0.02, p = 0.44, N= (2-tailed). De groep die de voetbalregel goed heeft, is gemiddeld genomen
114; Situatie 3: χ2(1) = 1.15, p = 0.14, N= 114; Situatie 4: χ2(1) = 0.06, p = 0.41, zekerder van zijn antwoord dan de groep die de voetbalregel fout heeft
N= 114; Situatie 5: χ2(1) = 0.34, p = 0.28, N= 114; Situatie 6: χ2(1) = 1.86, p = (resp. gemiddeld 1.49 (N = 63, SD = 0.69) en 1.98 (N = 50, SD = 0.96); waarbij
0.09, N= 114; alles 1-zijdig getoetst. score 1 = heel zeker).
AV&S 2016/34 Afl. 5 - oktober 2016 177

49
T2_AV&S_1605_bw_V02.indd 177 10/14/2016 3:30:57 PM

de groep die de voetbalregel fout heeft toegepast.57 Voor nis van de voetbalregels en sancties op overtreding daarvan
situatie 1, 3 en 5 is dit verschil niet vastgesteld. Het is open gerapporteerd voetbalgedrag nauwelijks aanwezig is, en
vallend dat dit onderscheid zich alleen voordoet in situaties voor zover deze aanwezig is, is deze niet in de verwachte
waarin eigenlijk meer sancties gelden dan alleen de voet- richting. De resultaten van dit onderzoek geven geen aan-
balsancties. Een verklaring daarvoor kan wellicht gevonden leiding om heel optimistisch te zijn over het idee dat meer
worden in de ernst van de gebeurtenis in die gevallen 2, 4 en kennis van de voetbalregels zou leiden tot minder agressie
6, welke ernst (blijkbaar) beter wordt ingeschat (en dat dan op de velden. Definitief uitsluitsel daarover vereist echter
ook met meer begrip) door degenen die de vraag goed (met een vollediger meting van de kennis van voetbalregels.60
een zwaardere sanctie) beantwoorden.
Het lijkt erop dat jeugdvoetbalspelers die kennis hebben 5. Conclusies, beperkingen en discussiepunten
van de voetbalregels die kunnen leiden tot ernstige letsel-
schadesituaties (rode-kaart-situaties) eerder begrip ervoor Het hiervoor beschreven onderzoek en de resultaten daar-
hebben dat de betreffende sanctie wordt opgelegd, dan spe- van zijn slechts indicatief. Het biedt een eerste indruk van
lers die deze kennis niet hebben. kennis en gedrag van A- en C-jeugd in een bepaalde regio.
4.3 Leidt kennis van regels tot positiever Uit de verrichte analyse wordt helder dat er bij jeugdspelers
voetbalgedrag? die oordelen over voetbalsituaties vrijwel geen kennis be-
De vraag die overblijft, is in hoeverre regelkennis samen- staat over de regels en sancties van het civiele en strafrech-
hangt met positiever voetbalgedrag. Zoals hierboven al telijke aansprakelijkheidsrecht. De preventieve werking die
omschreven, is onderscheid gemaakt tussen vijf vormen van het recht op dit terrein zou uitgaan, mag dus niet wor-
van gedrag, waarover de spelers zelf gerapporteerd hebben den overschat. Daarmee zeggen wij overigens verder niets
(pro-sociaal gedrag richting de tegenstander; pro-sociaal over de juridisch-inhoudelijke juistheid van de regels zoals
gedrag richting het eigen team; agressie gericht op de te- beschreven in paragraaf 2; wij zullen hier dus ook niet zeg-
genstander; agressie gericht op het team; agressie gericht gen of het terecht is dat in sport- en spelsituaties minder
op officials en omstanders). Er is vervolgens (alleen) een snel aansprakelijkheid wordt aangenomen. Wij menen wel
analyse gemaakt op het niveau van de (kennis van de) voet- dat er weinig kennis van sancties is, en dus de preventieve
balregels, omdat er te weinig variatie bestaat wat betreft de werking van civiele of strafrechtelijke) aansprakelijkheids-
juistheid van de antwoorden op de voorgelegde situaties als recht mogelijk daardoor niet aanwezig is. Dit zou ervoor
het gaat om de combinatie van voetbalregels met de regels pleiten om naast de toetsing van de regelkennis van voet-
over compensatie en/of straffen. ballers, hen ook op de hoogte te brengen van mogelijke an-
Uit die analyses komt naar voren dat voor de zes situaties dere sancties.
geen statistisch significante verschillen bestaan qua agres- Hiermee levert het onderzoek een belangrijke bijdrage aan
sie tussen de groep die de sanctie goed heeft ingevuld en de verdere theorievorming op het terrein van het aanspra-
de groep die deze fout heeft.58 Voor het pro-sociale gedrag kelijkheidsrecht waar het de vermeende preventieve wer-
blijkt in één situatie een statistisch significant verschil te king (of het gebrek daaraan) van het aansprakelijkheids-
bestaan, namelijk situatie 3, maar dat verschil is in onver- recht betreft.61
wachte richting (degenen die de regel fout hebben vertonen
gemiddeld genomen meer pro-sociaal gedrag richting de Inhoudelijk komt uit de analyses ook naar voren dat ook de
tegenstander dan degenen die de regel goed hebben).59 De spelregelkennis van de bevraagde jeugdspelers niet om-
conclusie is dus dat in dit onderzoek de relatie tussen ken- vangrijk is. Hoewel deze resultaten niet generaliseerbaar
zijn naar de totale populatie, baart ons dit wel zorgen. Die
57 Situatie 2: t(101)= 2.40, p = 0.018 (2-tailed). De groep die de voetbalregel
zorgen kunnen echter in zoverre worden genuanceerd dat,
goed heeft, vindt gemiddeld genomen de sanctie terechter dan de groep hoewel wel meer begrip lijkt te bestaan voor sancties die
die de voetbalregel fout heeft (resp. gemiddeld 1.63 (N = 48, SD = 0.53) en gelden in rode-kaart-situaties als er meer kennis is, uit dit
1.95 (N = 55, SD = 0.78); waarbij score 1 = heel terecht, 2 = terecht, etc.). Situ-
atie 4: t(105)= 2.82, p = 0.006 (2-tailed). De groep die de voetbalregel goed
onderzoek niet bleek dat kennis van regels tot positiever
heeft, vindt gemiddeld genomen de sanctie terechter dan de groep die de voetbalgedrag leidt. Er is daarmee dus niet direct veel ver-
voetbalregel fout heeft (resp. gemiddeld 1.29 (N = 58, SD = 0.50) en 1.67 (N loren, concluderen wij licht cynisch, met het gegeven dat
= 49, SD = 0.88); waarbij score 1 = heel terecht, 2 = terecht, etc.). Situatie 6:
t(101)= 2.45, p = 0.016 (2-tailed). De groep die de voetbalregel goed heeft,
regelkennis ontbreekt. Op basis van deze voorlopige resul-
vindt gemiddeld genomen de sanctie terechter dan de groep die de voet- taten zou je kunnen concluderen dat de rol van regels in het
balregel fout heeft (resp. gemiddeld 1.48 (N = 31, SD = 0.63) en 1.88 (N = 72, kader van stimuleren van positief voetbalgedrag – en op een
SD = 0.79); waarbij score 1 = heel terecht, 2 = terecht, etc.).
58 Situatie 1: F(3,104) = 1.53, p = 0.210; Situatie 2: F(3,104) = 1.49, p = 0.223;
hoger niveau: ordeherstel – niet mag en moet worden over-
Situatie 3: F(3,104) = 0.20, p = 0.996; Situatie 4: F(3,104) = 2.14, p = 0.10; schat. Het gedrag zou in plaats daarvan weleens gestuurd
Situatie 5: F(3,104) = 0.57, p = 0.982; Situatie 6: F(3,104) = 0.80, p = 0.497.
59 Situatie 1: F(2,105) = 1.38, p = 0.255; Situatie 2: F(2,105) = 0.33, p = 0.723; Si-
tuatie 3: F(2,105) = 3.44, p = 0.036. Dit verschil wordt veroorzaakt door pro- 60 Een strategie gericht op meer kennis van de regels kan om andere rede-
sociaal gedrag richting de tegenstander F(1,106) = 5.13, p = 0.026, waarbij nen uiteraard zeer nuttig en/of gewenst zijn. Of meer kennis van (bijvoor-
de groep die de regel fout heeft gemiddeld hoger scoort dan de groep die de beeld) de strafrechtelijke regels wel effect zou hebben gehad, hebben we
regel goed heeft (resp. gemiddeld 3.33 en 2.96; waarbij hoe hoger de score, niet kunnen toetsen; daarvoor zou grootschaliger vervolgonderzoek nodig
hoe meer het gedrag wordt vertoond). Situatie 4: F(2,105) = 1.45, p = 0.239; zijn.
Situatie 5: F(2,105) = 0.21, p = 0.808; Situatie 6: F(2,105) = 0.15, p = 0.862. 61 Zie nogmaals de literatuur genoemd in paragraaf 1.2.
178 Afl. 5 - oktober 2016 AV&S 2016/34

50
T2_AV&S_1605_bw_V02.indd 178 10/14/2016 3:30:57 PM

kunnen worden door de (in hogere mate aanwezige) over-

tuiging van het eigen gelijk in plaats van werkelijke kennis
van de regels. Vervolgonderzoek is echter nodig om te toet-
sen of deze resultaten worden gerepliceerd.
Dit onderzoek is op meerdere manieren beperkt in het trek-

ken van harde conclusies. Ten eerste is het eerste deel van
de besproken vragenlijst voor het eerst gebruikt, wat bete-
kent dat er meer onderzoek moet worden gedaan naar de
validiteit en betrouwbaarheid daarvan. Verder kunnen bij
de resultaten over de verdeling van kennis of gedrag geen
generaliserende uitspraken naar Utrechtse of Nederlandse
A- en C-jeugd worden gedaan. Als het gaat om verbanden,
dan geven de resultaten slechts een indruk daarover in de
populatie. De indruk die het onderzoek oproept, is echter
wel waardevol: de rol van regels in het kader van het stimu-
leren van ordeherstel mag en moet niet worden overschat
door de KNVB.
AV&S 2016/34 Afl. 5 - oktober 2016 179

51
T2_AV&S_1605_bw_V02.indd 179 10/14/2016 3:30:57 PM

52
Week 2
53
54
ARTIKELEN
Recidive na werkstraffen en na
gevangenisstraffen
Een gematchte vergelijking*
Hilde Wermink, Arjan Blokland, Paul Nieuwbeerta & Nikolaj Tollenaar
Het doel van dit artikel is het vergelijken van de recidive van werkgestrafte met de
recidive van gevangenisgestrafte volwassen daders in Nederland. We maken gebruik
van longitudinale, justitiële gegevens om de recidive over een maximumperiode van
acht jaar van daders die een werkstraf kregen opgelegd, te vergelijken met die van
daders die een gevangenisstraf ondergingen. Om rekening te houden met mogelijke
selectie-effecten wordt gebruikgemaakt van ‘propensity score matching’ en ‘matching
by variable’. De resultaten laten zien dat daders na een werkstraf minder recidiveren
dan na een gevangenisstraf.
Binnen het huidige palet aan strafmodaliteiten dat de Nederlandse rechter ter
beschikking staat, neemt de werkstraf een belangrijke plaats in. Cruciaal kenmerk
van de werkstraf is dat – anders dan bij detentie – geen sprake is van vrijheidsbe-
neming. Ter uitvoering van een werkstraf verricht de veroordeelde naast zijn of haar
reguliere dagbesteding onbetaalde arbeid binnen non-profitorganisaties. De afge-
lopen tien jaar verdubbelde het aantal aan volwassen daders opgelegde werkstraffen
van 14.485 in 1997 naar 32.590 in 2007 (CBS, 2008).
Werkstraffen werden eind jaren tachtig van de vorige eeuw geïntroduceerd in het
Nederlandse strafrecht. Bij de invoering in 1989 konden werkstraffen uitsluitend
worden opgelegd ter vervanging van een onvoorwaardelijke gevangenisstraf van
maximaal zes maanden.1 Sinds 2001, na de invoering van de Wet Taakstraffen,
kunnen werkstraffen met een maximale duur van 240 uur ook als zelfstandige
hoofdstraf worden opgelegd. De voornaamste doelstellingen van de nieuwe werk-
straffen waren terugdringen van de (duurdere) korte onvoorwaardelijke vrijheids-
straf, bijdragen aan de humanisering van de strafrechtspleging en verminderen van
recidive (Kamerstukken II 1986/87, 20 074, nr. 6, p. 2-3). In dit artikel richten wij
ons op het laatstgenoemde doel. De vraag die we in dit artikel trachten te beant-
woorden, is in hoeverre werkstraffen, gelet op de recidive in de navolgende periode,
een goed alternatief zijn voor gevangenisstraffen.
* Deze studie betreft een samenwerking tussen het NSCR en het WODC. Onze dank gaat uit naar het
WODC voor het beschikbaar stellen van de OBJD-gegevens en voor het verstrekken van een
onderzoekssubsidie. Verder willen we Gerben Bruinsma, Daniel Nagin en Stefan Bogaerts bedanken
voor hun ondersteuning en hulp bij de totstandkoming van dit onderzoek.
1 Zie de Wet op onbetaalde arbeid ten algemenen nutte (ATAN).
Tijdschrift voor Criminologie 2009 (51) 3 211
55
Directe vergelijkingen van recidivecijfers laten zien dat de recidive na een werkstraf
lager is dan na een gevangenisstraf. Recent onderzoek van het WODC bijvoorbeeld
laat zien dat van de daders die veroordeeld zijn tot een werkstraf, ongeveer 50 pro-
cent recidiveert (Wartna e.a., 2005b), terwijl van de daders die uitstromen uit een
penitentiaire inrichting, ruim 70 procent opnieuw met justitie in aanraking komt
(Wartna e.a., 2005a). In dit soort directe vergelijkingen wordt echter geen rekening
gehouden met selectie-effecten. Rechters leggen in de meeste gevallen een werkstraf
op als daders voor de eerste maal zijn veroordeeld en/of als daders veroordeeld zijn
voor het plegen van minder zware criminele activiteiten (Tak, 2001). Deze selec-
tieprocessen zorgen ervoor dat gevangenisgestraften al op voorhand een hoger
risico hebben op herhaald crimineel gedrag dan personen die tot een werkstraf wor-
den veroordeeld. Om daadwerkelijk aan te kunnen geven in welke mate strafrech-
telijke sancties de recidive van de daders beïnvloeden, is het noodzakelijk om voor
dergelijke selectieprocessen te controleren.
Slechts één gerandomiseerd experiment (Killias e.a., 2000) en een handvol mat-
ching-studies, gebaseerd op observationele data, vergelijken recidive na werkstraf
en na gevangenisstraf (Muiluvuori, 2001; Nirel e.a., 1997; Bol & Overwater, 1986;
Spaans, 1994). Door het kleine aantal observaties en het regionale karakter van het
gerandomiseerde experiment is de generaliseerbaarheid problematisch. De belang-
rijkste tekortkoming van de matching studies is dat er niet voldoende rekening
wordt gehouden met selectieprocessen. In dit artikel onderzoeken we de recidive
van werkgestraften ten opzichte van de recidive van gevangenisgestraften. We
pogen hierbij veel van de kenmerkende tekortkomingen van eerdere studies te ver-
mijden. We maken gebruik van een grootschalig databestand met gedetailleerde
gegevens van alle personen die in 1997 voor de eerste maal werden veroordeeld tot
een werk- dan wel gevangenisstraf (N=4.246). Om te controleren voor selectiepro-
cessen maken we gebruik van ‘matching’-methoden.
Werk- en gevangenisstraf in criminologisch perspectief
De penologische literatuur noemt retributie (vergelding/leedtoevoeging), generale

preventie (gericht op de totale bevolking) en speciale preventie (gericht op de dader)
als belangrijkste strafdoelen. De speciaal-preventieve werking van straf kan worden
bewerkstelligd door insluiting, waarbij het de dader tijdelijk fysiek onmogelijk wordt
gemaakt delicten te begaan, of door afschrikking en resocialisatie, waarvan wordt
verwacht dat zij na afloop van de straf invloed hebben op het gedrag van de gestrafte
dader. Dit artikel richt zich op speciale preventie met als centrale vraag in hoeverre
werkstraffen een goed alternatief zijn voor gevangenisstraffen in relatie tot de reci-
dive van de gestraften na afloop van hun straf. Een theoretisch antwoord op deze
onderzoeksvraag kan uit verschillende algemene criminologische theorieën worden
afgeleid, hetgeen leidt tot rivaliserende hypothesen.
Vanuit de afschrikkingstheorie kan worden verwacht dat recidive na werkstraffen
hoger is dan na gevangenisstraffen. De theorie gaat ervan uit dat sancties voor
gestraften een afschrikwekkende werking hebben. Afschrikking vindt plaats wan-
neer de gestraften zich geheel onthouden van toekomstig crimineel gedrag, minder
212 Tijdschrift voor Criminologie 2009 (51) 3
56
Recidive na werkstraffen en na gevangenisstraffen
delicten plegen of minder zware delicten plegen, uit angst in de toekomst opnieuw
gestraft te worden (Paternoster, 1991). Hoe zwaarder de straf, hoe meer de gestrafte
een nieuwe straf zal proberen te voorkomen, en dus hoe groter de afschrikwekkende
werking. Verwacht mag worden dat – hoewel vrijheids- en werkstraffen volgens de
wet inwisselbaar zijn – de ‘pains of imprisonment’ (Windzio, 2006) gemiddeld
genomen als zwaarder worden ervaren dan de ‘pains’ van werkstraffen. Hierdoor
zal de afschrikwekkende werking van werkstraffen minder sterk zijn dan die van
gevangenisstraffen.
Aan de andere kant zijn er diverse criminologische theorieën op basis waarvan wordt
verwacht dat werkstraffen juist leiden tot minder recidive dan gevangenisstraffen.
De differentiële-associatietheorie bijvoorbeeld veronderstelt dat het plegen van
delicten geleerd kan worden via interactie en communicatie met anderen die deviant
gedrag goedkeuren (Sutherland, 1947). Gevangenissen worden vanuit dit perspec-
tief gezien als ‘school voor criminaliteit’; een plek waar verschillende daders met
elkaar in aanraking komen en van elkaar deviante attitudes en crimineel gedrag
leren. De mogelijkheid om te leren van medegedetineerden is beduidend minder
voor daders die veroordeeld zijn tot een werkstraf. Daders die een werkstraf uit-
voeren, worden namelijk minder intensief en minder lang blootgesteld aan andere
daders.
Hirschi’s (1969) sociale-controletheorie en Sampson en Laubs (1993) leeftijdsge-
bonden theorie van informele sociale controle hebben beide als uitgangspunt dat
bepaalde controlemechanismen personen van delinquent gedrag kunnen weerhou-
den. Verschillen tussen personen in het plegen van crimineel gedrag zouden voort-
komen uit verschillen in de mate waarin geloofd wordt in de heersende normen in
de samenleving, de inzet in en tijd en energie besteed aan conventionele activiteiten,
en de mate waarin personen affectieve bindingen hebben met belangrijke anderen
die deviant gedrag afkeuren, zoals ouders, vrienden en collega’s (Hirschi, 1969).
Deze controlemechanismen worden aangetast wanneer daders voor een bepaalde
tijd uit hun sociale omgeving en conventionele routines worden gehaald, zoals bij
gevangenisstraffen. De negatieve gevolgen hiervan kunnen zich opstapelen over de
levensloop en zo de kans op criminaliteit verhogen (Sampson & Laub, 1993). Werk-
straffen, daarentegen, zijn gericht op (re-)integratie en beogen banden met de
omgeving in stand te houden of zelfs te versterken (Bazemore & Maloney, 1994).
Bovendien interfereren werkstraffen niet met conventionele bezigheden zoals
school of werk.
Ten slotte wordt binnen de labelingtheorie verondersteld dat officiële interventies
een bijdrage leveren aan een stabiele criminele carrière van daders (Becker, 1963).
Het labelen van daders, en in het bijzonder het officieel labelen van daders, kan
leiden tot economische en sociale stigmatisering (Pettit & Western, 2004), blok-
keert mogelijkheden voor een conventionele levensstijl (Becker, 1963; Farrington,
1977; Bernburg & Krohn, 2003) en verhoogt de sociale isolatie van de veroordeelde
ten opzichte van de bevolking die de wet niet overtreedt, waardoor het waarschijn-
lijker is dat de dader na bestraffing zal terugvallen in een criminele levensstijl (McA-
linden, 2005). Dit is in overeenstemming met de ‘disintegrative shaming’-theorie,
waarbij wordt gesteld dat stigmatisering in essentie een schaamtegevoel is, veroor-
zaakt doordat daders als ‘outcast’ worden gezien en uiteindelijk aangezet worden
57
tot het terugvallen in het plegen van criminele activiteiten (Braithwaite, 1989).
Werkstraffen worden geacht bij te dragen aan de humanisering van de strafrechts-
pleging, juist omdat wordt verwacht dat van werkstraffen een minder stigmatise-
rende werking uitgaat dan van gevangenisstraffen.
Samenvattend: verschillende criminologische theorieën leiden tot rivaliserende
voorspellingen. Vanuit deterrence-theorieën kan worden afgeleid dat werkstraffen
leiden tot meer recidive dan gevangenisstraffen, de overige theorieën voorspellen
het omgekeerde. Gedegen empirisch onderzoek om recidive van werk- ten opzichte
van gevangenisgestraften vast te stellen, is daarom geboden.
Eerder onderzoek
Twee recente overzichtsstudies (Villettaz e.a., 2006; Nagin e.a., 2008) analyseerden
een groot aantal studies waarin recidive na gevangenisstraf werd vergeleken met
recidive na andere straffen. Slechts vier studies met een experimenteel of matching
design vergeleken recidive na gevangenisstraffen met recidive na werkstraffen. Een
Nederlandse overzichtsstudie van Van Noije en Wittebrood (2008) noemt nog twee
Nederlandse matching-studies.
Elk van deze zes studies maakt gebruik van officieel geregistreerde data. In geen van
deze studies zijn de recidivecijfers na een werkstraf hoger dan die na een gevange-
nisstraf. Drie studies vinden, ook wanneer rekening gehouden wordt met bestaande
verschillen tussen de twee groepen gestraften, dat recidive na een werkstraf signi-
ficant lager is dan na een gevangenisstraf (Nirel e.a., 1997; Bol & Overwater, 1986;
Spaans, 1994). In de drie andere studies recidiveren werkgestraften ook minder dan
gevangenisgestraften, maar blijkt dat verschil niet significant (Killias e.a., 2000;
Gilliéron e.a., 2006; Muiluvuori, 2001). Nirel e.a. (1997), bijvoorbeeld, laten zien
dat gevangenisgestraften 1,7 keer vaker recidiveren dan werkgestraften. In het
enige gerandomiseerde experiment (Killias e.a., 2000) recidiveren werkgestraften
over een periode van twee jaar gemiddeld met 0,39 veroordelingen en gevangenis-
gestraften met 0,64 veroordelingen. Echter, in een follow-upstudie vinden Gilliéron
e.a. (2006) dat na een periode van elf jaar de verschillen in recidive tussen werken
gevangenisgestraften zijn verdwenen en dat het er zelfs op lijkt dat gevangenisge-
straften beter geïntegreerd zijn in de conventionele maatschappij dan werkgestraf-
ten.2
Ondanks het experimentele of matching-design kleven ook aan voornoemde studies
nog de nodige bezwaren. Zo maken niet alle studies gebruik van een voldoende
uitgebreide set controlevariabelen, waardoor de mogelijkheid blijft bestaan dat de
vergeleken groepen op voorhand van elkaar verschilden in belangrijke niet-geob-
serveerde kenmerken. Bovendien kijken eerdere studies niet in het bijzonder naar
de eerste veroordeling tot een werkstraf dan wel gevangenisstraf, waardoor de cau-
saliteit moeilijk is vast te stellen. Verder hanteren de meeste studies een korte fol-
low-upperiode, waardoor het niet mogelijk is om langetermijnverschillen in recidive
na verschillende sancties vast te stellen. Tot slot is de generaliseerbaarheid van
2 Gilliéron e.a. (2006) maakten gebruik van dezelfde data als Killias e.a. (2000).
58
sommige studies beperkt omdat alleen daders onderzocht werden die in een bepaald
arrondissement zijn veroordeeld, alleen zeer korte gevangenisstraffen met werk-
straffen worden vergeleken, of omdat slechts een zeer klein aantal daders vergeleken
kon worden.
Data en operationalisatie
Onze gegevens zijn afkomstig van de Justitiële Documentatiedienst van het Minis-
terie van Justitie. De gegevens zijn in bewerkte vorm beschikbaar gesteld door het
Wetenschappelijk Onderzoek- en Documentatiecentrum (WODC). De gegevens zijn
uit de OBJD (Onderzoeks- en Beleidsdatabase Justitiële Documentatie) gehaald en
in analyseerbare data omgezet volgens het procédé van de recidivemonitor van dit-
zelfde ministerie (zie voor uitgebreide informatie over het OBJD: Wartna e.a.,
2008). Het hier gebruikte databestand bevat de volledige strafbladen van alle per-
sonen die in Nederland in 1997 werden veroordeeld tot een werkstraf dan wel een
gevangenisstraf (N=28.072). Op deze strafbladen staan alle strafzaken vermeld die
door de politie bij het Openbaar Ministerie werden aangemeld. De gegevens bestrij-
ken voor elke persoon de volledige criminele carrière van vóór 1997 vanaf het 12e
levensjaar tot 2005.
Experimentele groep: eerste werkstraf

In dit onderzoek kiezen we als ‘experimentele’ groep alle volwassen daders van 18
tot 50 jaar die in 1997 veroordeeld zijn tot een werkstraf en die bovendien in de
jaren daaraan voorafgaand niet al eerder veroordeeld waren tot een werkstraf dan
wel een gevangenisstraf (N=7.806). De beslissing om recidive na werkstraffen alleen
te bepalen voor daders tussen de leeftijden 18 en 50 is gebaseerd op verschillende
overwegingen. Ten eerste valt het grootste deel van de daders, voor het eerst ver-
oordeeld voor een werkstraf, binnen deze leeftijdsgrenzen. Ten tweede worden voor
minderjarigen andere regelingen getroffen dan voor volwassenen. Om de groepen
zo vergelijkbaar mogelijk te houden, hebben we er daarom voor gekozen alleen vol-
wassen daders in onze analyses te betrekken. In 1997 kon werkstraf in theorie wor-
den opgelegd voor elk type delict, en alleen wanneer de veroordeelde bereid was
deze werkstraf uit te voeren (Tak, 2001).
De gemiddelde leeftijd van de 7.806 daders die in 1997 voor de eerste maal zijn
veroordeeld tot een werkstraf, is 28,4 jaar. 18,3 procent van hen is vrouw en 32,1
procent is in het buitenland geboren. Ten tijde van hun veroordeling in 1997 hadden
deze personen gemiddeld 0,81 feiten in hun criminele geschiedenis (gemeten tot
1987). De lengte van de in 1997 aan deze personen opgelegde werkstraffen varieerde
van 1 tot en met 240 uur, met een gemiddelde lengte van 106 uur.
Controlegroep: eerste gevangenisstraf

De controlegroep in dit onderzoek bestaat uit alle daders die in 1997 zijn veroor-
deeld tot een vrijheidsstraf van maximaal 6 maanden3 en die daaraan voorafgaand
3 In 1997 betreft dit 72,6 procent van het totale aantal personen dat een vrijheidsstraf kreeg opgelegd.
59
niet eerder veroordeeld waren tot een werkstraf dan wel gevangenisstraf (N=3.502).
We kiezen voor de controlegroep een vrijheidsstraf van maximaal 6 maanden, omdat
in 1997 de werkstraf wettelijk een alternatief werd geacht voor een gevangenisstraf
van ten hoogste die periode.
De gemiddelde leeftijd van de gevangenisgestraften in de controlegroep is 28,9
jaar. 12,2 procent van hen is vrouw en 68,6 procent is in het buitenland geboren.
Gemeten over de tien voorafgaande jaren telde het strafblad van deze personen in
1997 gemiddeld 0,9 eerdere veroordelingen. De lengte van de in de controlegroep
opgelegde gevangenisstraffen varieerde van 1 dag tot en met 6 maanden, met een
gemiddelde van 60,3 dagen. De verdeling van de gevangenisstraffen is zeer scheef:
bijna de helft van de controlepersonen heeft een gevangenisstraf opgelegd gekregen
tot 2 maanden. Bij langere straffen daalt het aantal personen snel.
Recidive
De uitkomstvariabele in deze studie is de recidive van daders na het uitgangsdelict
in 1997. Als uitgangsdelict kiezen we de eerste veroordeling in 1997 waarop een
werkstraf of gevangenisstraf volgde, alle veroordelingen daarna worden beschouwd
als recidive. Met ‘veroordeling’ doelen we op alle schuldigverklaringen en beleids-
sepots; vrijspraken en technische sepots vallen er niet onder. In de analyses verge-
lijken we het aantal veroordelingen van werkgestraften met dat van gevangenisge-
straften in de maximaal acht jaar na het uitgangsdelict, waarbij we controleren voor
eventuele perioden van detentie.
Methode: propensity score en variable matching
De beste methode om recidive van werkgestraften ten opzichte van gevangenisge-

straften vast te stellen, zou een experiment zijn waarbij daders willekeurig een
werkstraf of een gevangenisstraf wordt toebedeeld. Willekeurige of random toede-
ling van de verschillende strafsoorten zorgt ervoor dat daders in de werkgestrafte
groep op alle kenmerken, relevant voor recidive, vergelijkbaar zijn met de daders
uit de gevangenisgestrafte groep. In de praktijk zijn vergelijkingen gebaseerd op niet
experimenteel verkregen data vaak het maximaal haalbare. Selectieprocessen zor-
gen er dan voor dat het op voorhand al waarschijnlijker is dat gevangenisgestraften
vaker delicten plegen dan daders die veroordeeld zijn tot een werkstraf. Door
gebruik te maken van een uitgebreide set aan controlevariabelen en de methode
‘propensity score matching’ en ‘matching by variable’ wordt maximaal rekening
gehouden met deze selectieprocessen.4
Propensity score model

Om voor selectie-effecten te controleren maken we in dit artikel gebruik van ‘pro-
pensity score matching’. Deze methode is speciaal ontwikkeld om balans op cova-
riaten tussen de onderzoeksgroepen te verkrijgen (Rosenbaum & Rubin, 1983;
4 Modellen gebaseerd op propensity scores zijn robuuster wat betreft modelmisspecificatie in ver-
gelijking met modellen gebaseerd op regressietechnieken (Drake, 1993).
60
1984; zie ook Nieuwbeerta e.a., 2007). In deze context is de propensity score de
conditionele waarschijnlijkheid om op leeftijd t veroordeeld te worden tot een werk-
straf versus gevangenisstraf, gegeven de covariaten tot t. De propensity score wordt
geschat door gebruik te maken van een logistisch regressiemodel (Cox & Snell,
1989).
We maken gebruik van de ‘nearest neighbor matching’-strategie om daders op hun
propensity score te matchen. Na het schatten van de propensity worden daders uit
de experimentele groep een voor een gekoppeld, zonder teruglegging, aan daders
die een gevangenisstraf ondergingen (controlegroep), maar die op basis van hun
persoonlijke kenmerken een vergelijkbare voorspelde kans op werkstraf hadden.
Een persoon uit de controlegroep werd gekoppeld aan een individu uit de experi-
mentele groep, wanneer het verschil in de geschatte kans op werkstraf voor beide
personen niet meer bedroeg dan 0,05.
Na matching bestaan er tussen de beide groepen geen systematische verschillen
meer op geobserveerde variabelen, voorafgaand aan het opleggen van de straf. Door
te matchen wordt met andere woorden gecontroleerd voor selectie in de strafop-
legging op basis van de in het model opgenomen variabelen en daarmee voor de
mogelijke vertekening in recidivecijfers die voortkomt uit dit selectieproces. Het
aantal potentiële verstorende variabelen is groot en Nagin e.a. (2008) stellen dat in
ieder geval rekening gehouden moet worden met criminele geschiedenis, type delict,
leeftijd, geslacht en of de dader autochtoon is. Van deze kenmerken is ook bekend
dat zij van belang zijn voor rechters, wanneer zij beslissingen nemen ten aanzien
van het berechten van daders (Monahan, 2006). In ons model houden we uitgebreid
rekening met deze kenmerken (zie tabel 1).
Tabel 1 laat de resultaten zien van het propensity score model. De meeste kenmer-
ken hebben een significante invloed op de kans om veroordeeld te worden tot een
werkstraf. De resultaten laten zien dat de odds van vrouwen om veroordeeld te
worden tot een werkstraf 1,50 groter is dan die van mannen. In het buitenland
geboren daders hebben een vijf keer kleinere odds op een werkstraf dan autochto-
nen. Bij de meeste delicttypen is de kansverhouding kleiner om veroordeeld te wor-
den tot een werkstraf in plaats van een gevangenisstraf, in vergelijking met een
veroordeling voor een ‘overige wetten’-delict (de referentiecategorie). Zo zijn vooral
vermogensdelicten met geweld en brandstichting met levensgevaar voorspellers om
tot een gevangenisstraf veroordeeld te worden.
61
Tabel 1: Variabelen propensity score model (N=11.308, waarvan 7.806

personen veroordeeld tot werkstraf, 3.502 veroordeeld tot
gevangenisstraf)
B S.E. Sign. Exp(B)
Persoonskenmerken
vrouw 0,41 0,07 *** 1,50
leeftijd (/10) -1,29 0,22 *** 0,28
leeftijd kwadraat (/10) 0,21 0,03 *** 1,23
geboren in buitenland -1,61 0,05 *** 0,20
Uitgangszaak
aantal feiten in uitgangszaak 0,08 0,03 *** 1,08
ernst uitgangszaak1 0,11 0,02 *** 1,12
Type delict
overige wetten (ref.) (ref.) (ref.)
huis, lokaalvredebreuk -1,41 0,86 * 0,24
openlijk geweld -0,03 0,17 0,97
wederspannigheid -0,25 0,50 0,78
overig openbare orde -0,32 0,24 0,73
agressief ernstig letsel -0,44 0,18 *** 0,64
belediging -2,09 1,30 0,12
bedreiging, mishandeling -0,36 0,14 *** 0,69
mishandeling zwaar -0,60 0,24 *** 0,55
brandstichting levensgevaar -1,98 0,36 *** 0,14
overige zaken, dieren -0,81 0,22 *** 0,45
zedengeweld -1,00 0,32 *** 0,37
ontucht jonger 16 jaar -0,63 0,23 *** 0,53
bevoordeling -0,09 0,12 0,92
wegnemen -1,26 0,13 *** 0,28
gekwal. diefstal -1,27 0,13 *** 0,28
vermogensgeweld -1,61 0,20 *** 0,20
WvW 0,19 0,13 1,21
Opiumwet -0,59 0,11 *** 0,56
Vuurwapenwet -0,17 0,18 0,84
Criminele geschiedenis
aantal vermogensdelicten in afgelopen jaar -0,36 0,07 *** 0,70
aantal geweld in afgelopen jaar -0,28 0,14 ** 0,76
aantal overig in afgelopen jaar 0,03 0,08 1,03
aantal vermogen in laatste 10 jaar -0,03 0,03 0,97

aantal geweld in laatste 10 jaar 0,03 0,06 1,03
62
Tabel 1: (Vervolg)
B S.E. Sign. Exp(B)
aantal overig in laatste 10 jaar -0,09 0,03 *** 0,91
Constante 3,41 0,35 ***

* p<0,10; ** p<0,05; *** p<0,01
1 Op basis van maximale wettelijke strafdreiging.
Figuur 1 laat de verdeling van de propensity scores zien binnen de twee groepen.
Duidelijk is te zien dat personen die op basis van hun voorspelde kenmerken vooraf
een hoge kans hadden op een werkstraf, in de meeste gevallen ook daadwerkelijk
een werkstraf kregen opgelegd. Tegelijkertijd blijkt uit figuur 1 dat er voldoende
overlap – of common support – bestaat tussen de propensity score-verdeling van
de experimentele groep en die van de controlegroep. Dit is een aanwijzing dat pro-
pensity score matching voor deze data een geëigende methode vormt. Wanneer er
voldoende common support is, heeft de uiteindelijke matching niet enkel betrekking
op een select (niet-representatief) deel van de experimentele of controlegroep.
Figuur 1: Verdeling propensity score (kans op werkstraf, werkgestraften en

gevangenisgestraften, totaal bestand (N=11.308))
300
250
200
aantal personen
150
100
50
0
0,01 0,11 0,21 0,31 0,41 0,51 0,61 0,71 0,81 0,91
propensity score
10 per. Mov. Avg. (werkgestraften) 10 per. Mov. Avg. (gevangenisgestraften)
Figuur gebaseerd op een tienpunts zwevend gemiddelde.
Variabele matching
In deze studie kiezen we er aanvullend voor om – voordat we matchen op de pro-
pensity score – eerst daders uit de experimentele en de controlegroep specifiek te
koppelen op geslacht, leeftijdscategorie en straflengte. Bij het indelen van de werk-
en gevangenisstraffen zijn, conform artikel 22b Wetboek van Strafrecht (Sr), werk-
straffen tot 60 uur gelijkgesteld aan gevangenisstraffen tot een maand (Schuyt,
2008). Opvolgende werkstrafcategorieën zijn steeds minder breed gekozen, zodat
het wettelijk maximum van 240 uur niet wordt overschreden. Bij de straflengte
63
onderscheiden we zes categorieën. Doordat we daders in de experimentele groep

allereerst koppelen aan daders in de controlegroep op basis van leeftijd, geslacht en
straflengte, en vervolgens ook nog matchen op basis van propensity score, worden
de experimentele en de controlegroep optimaal vergelijkbaar gemaakt.
Tabel 2: Verschillen werkgestraften (experimentele groep) en

gevangenisgestraften (controlegroep) – gekoppeld databestand
(N=4.246)
Gemid- Gemid- Abso- t-stat D
delde delde luut
experi controle ver-
mentele groep schil
groep (N=2.123)
(N=2.123)
Vrouw 0,125 0,125 0,000 0,000 0,0
Leeftijd (/10) 2,723 2,723 0,000 0,000 0,0
Leeftijd kwadraat (/10) 9,003 9,011 -0,008 -0,052 -0,2
Geboren in buitenland 0,554 0,555 -0,001 -0,062 -0,2
Aantal feiten in uitgangszaak 1,581 1,574 0,007 0,253 0,8

Strafdreiging 4,315 4,181 0,134 1,547 4,7
Aantal feiten afgelopen jaar

vermogen 0,057 0,049 0,008 0,907 2,8
geweld 0,021 0,018 0,002 0,507 1,6
overig 0,060 0,069 -0,009 -0,965 -3,0
Aantal feiten afgelopen tien jaar

vermogen 0,314 0,329 -0,015 -0,545 -1,7
geweld 0,097 0,104 -0,007 -0,614 -1,9
overig 0,352 0,389 -0,036 -1,394 -4,3
* p<0,10; ** p<0,05; *** p<0,01
Het was mogelijk om aan ruim 27 procent van de werkgestrafte daders een gevan-
genisgestrafte dader te koppelen (zie tabel 2). Daarmee werd bijna 61 procent van
de controlegroep benut. Tabel 2 laat de resultaten van de matchingmethoden zien
in het creëren van balans op alle covariaten tussen de experimentele en de contro-
legroep. De daders uit beide groepen verschillen voor de koppeling statistisch sig-
nificant op veel verschillende kenmerken. Daders die een werkstraf kregen, zijn in
vergelijking met daders die een gevangenisstraf kregen, bijvoorbeeld vaker vrouw,
minder vaak van niet-Nederlandse herkomst en hebben minder vermogens- en
geweldsfeiten in hun criminele geschiedenis.5 Tabel 2 laat zien dat we succesvol zijn
in het creëren van balans: na de koppeling zijn er geen significante verschillen meer
tussen de daders veroordeeld tot een werkstraf en de daders veroordeeld tot een
5 Deze resultaten zijn in verband met de beschikbare ruimte niet getoond in tabel 2 en kunnen
desgewenst worden opgevraagd bij de eerste auteur.
64
gevangenisstraf.6 Dit houdt in dat we er zeker van kunnen zijn dat eventuele ver-
schillen in recidive tussen de groepen niet een reflectie zijn van gemeten verschillen
tussen de experimentele en controlegroep op deze kenmerken.
Resultaten: recidive van werken gevangenisgestraften
Tabel 3 laat de recidive na de eerste werkstraf zien ten opzichte van een eerste
gevangenisstraf. De recidive wordt weergegeven over verschillende follow-upperio-
den en zowel voor alle delicten samen als uitgesplitst naar het type delict waarmee
de gestrafte recidiveert. De recidive is voor elk van de verschillende follow-uppe-
rioden uitgedrukt in een jaarlijks gemiddeld aantal delicten.
Tabel 3: Gemiddelde jaarlijkse recidive na eerste werk-/gevangenisstraf,

verschillende followupperioden
Gemiddelde Gemiddelde Absoluut t-stat Sign. Relatief
experimentele controle- verschil verschil
groep groep
(N=2.123) (N=2.123)
1 jaar
totaal 0,273 0,683 -0,410 -3,229 *** -0,60
vermogen 0,132 0,398 -0,266 -2,404 ** -0,67
geweld 0,044 0,109 -0,065 -2,255 ** -0,60
overig 0,097 0,175 -0,079 -3,698 *** -0,45
3 jaar
totaal 0,292 0,575 -0,283 -5,401 *** -0,49
vermogen 0,129 0,292 -0,162 -5,418 *** -0,56
geweld 0,052 0,100 -0,049 -2,755 *** -0,48
overig 0,111 0,183 -0,072 -2,048 ** -0,40
5 jaar
totaal 0,276 0,518 -0,242 -5,181 *** -0,47
vermogen 0,115 0,250 -0,135 -5,339 *** -0,54
geweld 0,053 0,096 -0,043 -2,645 *** -0,45
overig 0,108 0,172 -0,064 -2,000 ** -0,37
8 jaar
totaal 0,249 0,444 -0,195 -4,943 *** -0,44
vermogen 0,094 0,203 -0,110 -5,422 *** -0,54
geweld 0,049 0,086 -0,037 -2,770 *** -0,42
overig 0,106 0,155 -0,049 -1,497 -0,31
* p<0,10; ** p<0,05; *** p<0,01
6 De significantie van deze verschillen werd vastgesteld met een conventionele t-test en een gestan-
daardiseerde verschiltoets (D). Er is ook balans verkregen op alle delicttypen en op de straf-
lengte – niet getoond in tabel 2.
65
In tabel 3 wordt verder onderscheid gemaakt tussen absolute en relatieve verschil-

len. Het absolute verschil van werkstraffen wordt bepaald door het verschil te
nemen tussen de recidive van de controlegroep – de gevangenisgestraften – en de
recidive van de experimentele groep – de werkgestraften. Bij het bepalen van het
relatieve verschil wordt ook nog rekening gehouden met het feit dat de recidive van
bijvoorbeeld mannen hoger is dan van vrouwen en dat die van jongeren hoger is dan
van ouderen, ongeacht de soort ondergane straf. Het relatieve verschil van werkstraf
wordt berekend door het quotiënt te nemen van de recidive na werkstraf en de
recidive na gevangenisstraf. Het relatieve verschil geeft daarmee de afwijking van
de base rate (de recidive van de controlegroep) weer.
Het absolute verschil van de eerste werkstraf ten opzichte van de eerste gevange-
nisstraf op de totale recidive is negatief en uiterst significant. De negatieve richting
impliceert dat, gemiddeld genomen, de ervaring van een eerste werkstraf tot minder
recidive leidt dan de ervaring van een eerste gevangenisstraf. Na vijf jaar bijvoor-
beeld is het verschil 1,21 veroordeling (5 * -0,242 = -1,21). Onze resultaten laten
verder zien dat, in relatieve termen, werkgestraften 47 procent minder recidiveren
dan gevangenisgestraften.
Wanneer we de recidive opsplitsen naar delicttypen ontstaat eenzelfde beeld. Reci-
dive na werkstraf is voor alle delicttypen lager dan na een gevangenisstraf. In het
jaar na een opgelegde werkstraf wordt 67 procent minder vermogensdelicten
gepleegd dan in het jaar na een gevangenisstraf. Voor geweldsdelicten is dit 60 pro-
cent. De gevonden verschillen blijven ook op lange termijn (acht jaar) bestaan (zie
figuur 2). Van de zestien uitkomsten van het absolute verschil van werkstraffen ten
opzichte van gevangenistraffen zijn vijftien negatief en statistisch significant.
Kortom: in vergelijking met gevangenisstraffen hebben werkstraffen een dempende
werking, zowel op de totale recidive als op de vermogens- en geweldsrecidive afzon-
derlijk, en zowel op de korte termijn als op de lange termijn.
Sensitiviteitsanalyse
Bij wijze van sensitiviteitsanalyse hebben we bij het bepalen van de verschillen in
recidive onderscheid gemaakt naar geslacht en naar verschillende leeftijdscatego-
rieën (tabel 4). Recidive na werkstraffen blijkt voor zowel mannen als vrouwen lager
dan na gevangenisstraffen. Gemeten over een follow-upperiode van acht jaar is het
absolute verschil van een werkstraf voor mannen -1,728 en statistisch significant;
dit is een gemiddelde relatieve daling van 45 procent ten opzichte van de recidive
van de gevangenisgestraften. Voor vrouwen is het absolute verschil ook negatief en
statistisch significant (-0,376) – een daling van 29 procent. Verder laat tabel 4 zien
dat recidive na werkstraffen lager is dan die na gevangenisstraf voor alle onder-
scheiden leeftijdscategorieën.
66
Figuur 2: Absoluut verschil in recidive na eerste werk-/gevangenisstraf,

verschillende follow-upperioden (N=4.246)
0 totaal
-0,05 vermogen
absoluut verschil in recidive
-0,1
geweld
-0,15
overig
-0,2
-0,25
-0,3
-0,35
-0,4
-0,45 1 jaar 3 jaar 5 jaar 8 jaar
follow-upperiode
Tabel 4: Gemiddelde jaarlijkse recidive na eerste werk-/gevangenisstraf, follow-

upperiode acht jaar, mannen en vrouwen, verschillende
leeftijdscategorieën
Gemid- Gemid- Absoluut t-stat Sign. Relatief
delde delde verschil verschil
experi- controle
mentele groep
groep
Geslacht
mannen (N=1.857/1.857) 0,269 0,485 -0,216 -4,640 *** -0,45
vrouwen (N=266/266) 0,114 0,161 -0,047 -1,865 * -0,29
Leeftijd
18-21 (N=431/431) 0,419 0,711 -0,292 -4,135 *** -0,41
22-25 (N=389/389) 0,279 0,631 -0,352 -3,902 *** -0,56
26-30 (N=504/504) 0,225 0,364 -0,139 -3,321 *** -0,38
31-35 (N=364/364) 0,198 0,299 -0,101 -2,015 ** -0,34
36-40 (N=222/222) 0,162 0,280 -0,117 -2,339 ** -0,42
41-45 (N=135/135) 0,108 0,215 -0,106 -1,867 * -0,49
46-50 (N=78/78) 0,054 0,112 -0,058 -1,763 * -0,52
* p<0,10; ** p<0,05; *** p<0,01
Conclusie
Doel van dit artikel was om voor de Nederlandse situatie de recidive van volwassen
werkgestraften te vergelijken met de recidive van volwassenen die een gevangenis-
straf van maximaal zes maanden ondergingen. Voor Nederland, maar ook interna-
tionaal, is dit de eerste grootschalige matching studie, gebaseerd op observationele
67
data over een langere periode, die de recidive na werkstraffen vergelijkt met die na
gevangenisstraffen (N=4.246). Gebaseerd op de officieel geregistreerde criminele
carrières van alle volwassen personen veroordeeld in 1997 laten onze bevindingen
zien dat daders significant minder recidiveren na een werkstraf dan na een gevan-
genisstraf. Dit gevonden resultaat blijft substantieel, ook wanneer door matching
gecontroleerd wordt voor mogelijke selectie-effecten. Na werkstraffen recidiveer-
den daders minder, zowel op de korte als op de lange termijn – gemiddeld 50 procent
minder veroordelingen over een periode van acht jaar. Dit wordt gevonden voor
zowel mannen als vrouwen en op verschillende leeftijden.
Onze resultaten zijn niet gebaseerd op een experiment met volledige random toe-
wijzing van werken gevangenisstraffen en daarom blijft voorzichtigheid geboden
bij het interpreteren van de gevonden verschillen in termen van causaliteit. Het
risico blijft immers bestaan dat ondanks de matching op geobserveerde variabelen
een of meer niet-gemeten variabelen verantwoordelijk zijn voor de gevonden ver-
schillen in recidive. Persoonlijke omstandigheden die niet worden meegenomen in
ons propensity score model, zoals werk of een woning hebben of verslaafd zijn aan
alcohol of drugs, zouden bijvoorbeeld zowel de kans om een bepaald straftype opge-
legd te krijgen, als de kans op recidive kunnen beïnvloeden. Echter, voor zover deze
ongeobserveerde variabelen ook van invloed zijn op het criminele gedrag vooraf-
gaand aan de werk- of gevangenisstraf, wordt voor hun invloed (deels) gecontroleerd
door het opnemen van kenmerken van de criminele geschiedenis in het propensity
score model. Verder bevat het OBJD informatie over vonnissen en niet over wer-
kelijk uitgevoerde straffen. Een onbekend aantal werkstraffen kan derhalve in een
later stadium alsnog zijn omgezet in een gevangenisstraf (dit is het geval als een
werkstraf wordt gestaakt of iemand helemaal niet komt opdagen) en dus in onze
analyses ten onrechte onder werkstraf zijn gerubriceerd. Dat zou – gezien de richting
van het gevonden resultaat –tot een conservatieve schatting leiden. Toekomstig
onderzoek zal, om de mogelijkheid van verborgen bias volledig uit te kunnen sluiten,
zich moeten richten op het bestuderen van de effecten van verschillende strafrech-
telijke sancties binnen een gerandomiseerd experimenteel design (zie ook: De Roos,
2007).
Vanuit een theoretisch perspectief zetten deze resultaten vraagtekens bij de theorie
van afschrikking – als onderdeel van het strafdoel speciale preventie. Terwijl ver-
wacht werd dat de gepercipieerde kosten van gevangenisstraf hoger en daarmee de
afschrikkende werking van gevangenisstraf groter zou zijn dan die van een werk-
straf, was er toch sprake van minder recidive na een werkstraf. Hoewel hier niet
direct getoetst, lijkt het, gegeven het karakter van de Nederlandse werkstraf, onaan-
nemelijk dat daders de ‘pains’ van een werkstraf hoger waarderen dan de ‘pains of
imprisonment’. Met de nodige voorzichtigheid kan daarom uit onze resultaten wor-
den geconcludeerd dat op het punt van speciale preventie afschrikking geen grote
rol speelt, of althans wordt overschaduwd door andere gevolgen van gevangenis-
straf. Vanuit de differentiële-associatie-, sociale-controle- en labelingtheorieën
werd wel voorspeld dat werkstraffen zouden leiden tot minder recidive dan gevan-
genisstraffen. Nadere studie is nodig om vast te stellen voor welke van de door deze
theorieën voorspelde processen de meeste empirische ondersteuning wordt gevon-
den.
68
Ook voor beleidsmakers zijn onze resultaten relevant. Met het detineren van daders
zijn hoge kosten gemoeid. Werkstraffen vormen een goedkoper alternatief.7 Boven-
dien blijkt uit onze studie dat daders na een werkstraf minder recidiveren dan na
een gevangenisstraf, wat een extra kostenbesparing oplevert in termen van het
voorkomen van delictschade. Daarnaast ligt bij werkstraffen sterker de nadruk op
resocialisatie en re-integratie, en minder op leedtoevoeging. Daardoor zijn naar
verwachting ook de humane kosten van werkstraf minder dan die van gevangenis-
straf. Aan de andere kant voorkomt gevangenisstraf, anders dan werkstraf, delicten
in de periode dat daders gedetineerd zijn. Ook zou de normbevestigende en alge-
meen preventieve werking van gevangenisstraf groter kunnen zijn dan die van
werkstraffen, wat juist weer voorkómt dat potentiële daders daadwerkelijk over de
schreef gaan. Tot slot is de vraag of werkstraffen wel tegemoetkomen aan de
behoefte tot genoegdoening die (potentiële) slachtoffers van misdrijven ervaren.
Juist om die reden wil het OM onderzoek gaan doen naar het vermeende vermin-
derde maatschappelijke draagvlak voor werkstraffen (Harm Brouwer in de Volks-
krant, 29 april 2009).
Of werkstraffen dus in alle relevante opzichten een goed alternatief zijn voor gevan-
genisstraffen, blijft daarom de vraag, hangende nader (experimenteel) onderzoek,
maar voor zover het recidive betreft, lijkt het antwoord vooralsnog ‘ja’ te zijn.
Literatuur
Bazemore, G. & Maloney, D. (1994). Rehabilitating community service: Toward restorative

service sanctions in a balanced justice system. Federal probation, 58(1), 24-35.
Becker, G.S. (1963). Outsiders: Studies in the sociology of deviance. New York: Free Press.
Bernburg, J.G. & Krohn, M.D. (2003). Labeling, life chances, and adult crime: The direct and
indirect effects of official intervention in adolescence on crime in early adulthood. Crim-
inology, 41(4), 1287-1318.
Bol, M.S. & Overwater, J.J. (1986). Recidive van dienstverleners: In het strafrecht voor volwas-
senen. Den Haag: Staatsuitgeverij.
Braithwaite, J. (1989). Crime, shame and reintegration. Sydney: Cambridge University Press.
CBS (2008). Misdrijven; Opgelegde straffen en maatregelen. Op 8 september 2008 geraadpleegd
op http://statline.cbs.nl/StatWeb/publication/?DM=SLNL&PA=03727&D1=0-25&D2
=0&D3=0&D4=0&D5=(l-11)-l&VW=T.
Cox, D.R. & Snell, E.J. (1989). Analysis of binary data. London: Chapman and Hall.
Drake, C. (1993). Effects of the propensity score on estimators of treatment effect. Biome-
trics, 49, 1231-1236.
Farrington, D.P. (1977). The effects of public labeling. British Journal of Criminology, 17(2),
112-125.
Gilliéron, G., Poglia, C., Villard, F., Vuille, J., Perisset, C. & Killias, M. (2006). Bessere Reso-
zialisierung nach gemeinnütziger Arbeit? Ein Vergleich mit kurzen Freiheitsstrafen – 11
Jahre später. Crimiscope, 34. ESC – UNIL – Lausanne. www.unil.ch/esc/page19064.html.
7 Het zes maanden detineren van een persoon kost de Nederlandse samenleving ongeveer € 34.200
(Moolenaar, 2005). Bij een werkstraf van 240 uur in plaats van zes maanden gevangenisstraf dalen
de totale kosten met zo’n € 30.000 naar ongeveer € 4.200 (Moolenaar, 2005; zie ook: Junger-Tas,
1994).
69
Heide, W. van der, Wartna, B.S.J. & Blom, M. (2004). Tenuitvoerlegging van sancties. In:
A.Th.J. Eggen & W. van der Heide (red.). Criminaliteit en rechtshandhaving. Den Haag:
Boom Juridische uitgevers, 173-210.
Hirschi, T. (1969). Causes of delinquency. Berkeley/Los Angeles: University of California Press.
Junger-Tas, J. (1994). Rapport van de werkgroep kosten-baten taakstraffen. Den Haag: Minis-
terie van Justitie.
Killias, M., Aebi, M. & Ribeaud, D. (2000). Does community service rehabilitate better than
short-term imprisonment? Results of a controlled experiment. The Howard Journal,
39(1), 40-57.
McAlinden, A. (2005). The use of ‘shame’ with sexual offenders. British Journal of Criminol-
ogy, 45, 373-394.
Monahan, J. (2006). Structured violence risk assessment. In: R. Simon & K. Tardiff (eds.).
American psychiatric publishing textbook on violence assessment and management. Wash-
ington, DC: American Psychiatric Publishing.
Moolenaar, D.E.G. (2005). Uitgaven aan criminaliteit. In: A.T.J. Eggen & W. van der Heide
(red.). Criminaliteit en rechtshandhaving 2004. Den Haag: WODC, 211-243.
Muiluvuori, M.L. (2001). Recidivism among people sentenced to community service in Fin-
land. Journal of Scandinavian Studies in Criminology and Crime Prevention, 2(1), 72-82.
Nagin, D.S., Cullen, F.T. & Jonson, C.L. (2008). Imprisonment on reoffending. Crime and
Justice, 38.
Nieuwbeerta, P., Nagin, D.S. & Blokland, A.A.J. (2007). Het meten van effecten van gevan-
genisstraf op crimineel gedrag in een niet-experimentele studie. Mens en Maatschappij,
82(3), 272-299.
Nirel, R., Landau, S.F., Sebba, L. & Sagiv, B. (1997). The effectiveness of service work: An
analysis of recidivism. Journal of Quantitative Criminology, 13(1), 73-92.
Noije, L. van & Wittebrood K. (2008). Sociale veiligheid ontsleuteld: Veronderstelde en werkelijke
effecten van veiligheidsbeleid. Den Haag: Sociaal en Cultureel Planbureau.
Paternoster, R. (1991). Capital punishment in America. New York: Lexington Books.
Pettit, B. & Western, B. (2004). Mass imprisonment and the life course: Race and class
inequality in U.S. incarceration. American Sociological Review, 69, 151-169.
Roos, T. de (2007). Experimenteren met sancties? Tijdschrift voor Criminologie, 49(1), 12-23.
Rosenbaum, P. & Rubin, D. (1983). The central role for the propensity score in observational
studies of causal effects. Biometrika, 70, 41-55.
Rosenbaum, P. & Rubin, D. (1984). Reducing bias in observational studies using subclassifi-
cation on the propensity score. Journal of the American Statistical Association, 94, 516-524.
Sampson, R.J. & Laub, J.H. (1993) Crime in the making; Pathways and turning points through
life. Cambridge, Mass: Harvard University Press.
Schuyt, P.M. (2008). Straffen. In: C.P.M. Cleiren & J.F. Nijboer (red.). Strafrecht. Deventer:
Kluwer, 55-150.
Sherman, L.W., Gottfredson, D.C. Mackenzie, D.L., Eck, J.E., Reuter, P. & Bushway, S. (1997).
Preventing crime: What works, what doesn’t, what’s promising. Washington, DC: Office of
Justice Programs.
Spaans, E.C. (1994). Appels en peren: Een onderzoek naar de recidive van dienstverleners en kort-
gestraften. Arnhem: Gouda Quint.
Sutherland, E.D. (1947). Principles of criminology. Philadelphia: Lippincott.
Tak, P.J. (2001). Sentencing and punishment in the Netherlands. In: M. Tonry & R.S. Frase
(eds.). Sentencing and sanctions in western countries. New York: Oxford University Press.
Villettaz, P., Killias, M. & Zoder, I. (2006). The effects of custodial vs. non-custodial sentences
on re-offending. A systematic review of the state of knowledge. www.campbellcollabora-
tion.org/doc-pdf/Campbell-report-30.09.06.pdf.
70
Wartna, B.S.J., Blom, M. & Tollenaar, N. (2008). De WODC-Recidivemonitor. Den Haag:

WODC.
Wartna, B.S.J., Kalidien, S.N., Tollenaar, N. & Essers, A.A.M. (2005a). Strafrechtelijke recidive
van ex-gedetineerden: Uitstroomperiode 1996-2003. Den Haag: WODC.
Wartna, B.S.J., Tollenaar, N. & Blom, M. (2005b). Recidive 1997. Een cijfermatig overzicht van
de strafrechtelijke recidive van volwassen en jeugdige daders. Den Haag: WODC.
Windzio, M. (2006). Is there a deterrent effect of pains of imprisonment? The impact of
‘social costs’ of first incarceration on the hazard rate of recidivism. Punishment Society,
8, 341-364.
71
72
14679930, 2005, 2, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1467-9930.2005.00200.x by Uva Universiteitsbibliotheek, Wiley Online Library on [28/11/2022]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
General Deterrence and Corporate
O
Thornton,
Oxford,
Law
LAPO
0265-8240
April
2
27
© riginal
Blackwell
LAW Article
Blackwell
&2005 Gunningham
&Policy
UKPublishing,
Publishing
POLICY & Kagan
AprilLtd.
Ltd.
2005 2005 DETERRENCE
Environmental Behavior*
DOROTHY THORNTON, NEIL A. GUNNINGHAM, and
ROBERT A. KAGAN
This research addresses the assumption that “general deterrence” is an important

key to enhanced compliance with regulatory laws. Through a survey of 233 firms in
several industries in the United States, we sought to answer the following questions:
(1) When severe legal penalties are imposed against a violator of environmental
laws, do other companies in the same industry actually learn about such “signal
cases”? (2) Does knowing about “signal cases” change firms’ compliance-
related behavior? It was found that only 42 percent of respondents could identify
the “signal case,” but 89 percent could identify some enforcement actions against
other firms, and 63 percent of firms reported having taken some compliance-
related actions in response to learning about such cases. Overall, it is concluded
that because most firms are in compliance already (for a variety of other reasons),
this form of “explicit general deterrence” knowledge usually serves not to enhance
the perceived threat of legal punishment, but as reassurance that compliance is not
foolish and as a reminder to check on the reliability of existing compliance routines.
In most regulatory programs, officials formally prosecute and obtain legal

sanctions against violators in only a small percentage of infractions. They deal
with most detected violations at the bottom of the “pyramid of sanctions”
(Ayres & Braithwaite 1992)—that is, by means of warnings, demands for
remedial action, repeated re-inspection, and other informal pressures. At
the same time, most regulatory officials and regulatory scholars believe that
governmental capacity to impose severe legal penalties, together with relatively
frequent use of that capacity, is crucial to the implementation of regulatory
norms. Pro-regulation advocacy groups often complain that regulatory
effectiveness suffers because violations are prosecuted too infrequently and
penalized too lightly.
* This work was supported by the U.S. Environmental Protection Agency’s STAR Grant Pro-
gram and the Center for the Study of Law and Society at University of California, Berkeley.
The authors are also grateful to Manuel Vallee and Peter Younkin, who conducted the bulk of
the survey interviews and provided other equally excellent research support.
Address correspondence to Neil Gunningham, REGNET, University House, Australian
National University, Canberra ACT 0200, Australia; e-mail: neil.gunningham@anu.edu.au
LAW & POLICY, Vol. 27, No. 2, April 2005 ISSN 0265 – 8240
© 2005 UB Foundation Activities Inc., for and on behalf of the Baldy Center for Law and
Social Policy and Blackwell Publishing Ltd.
73
Thornton, Gunningham & Kagan DETERRENCE 263
There is surprisingly little research, however, that tells us precisely how

important the threat of large legal sanctions really is in motivating regulated
business firms to comply with the law. This article, together with a companion
article (Gunningham, Thornton & Kagan 2005), reports the results of a research
project designed to explore that question.
I. EXPLAINING REGULATORY COMPLIANCE: GENERAL DETERRENCE

AND ALTERNATIVE HYPOTHESES
The basic theory of general deterrence rests on the notion that regulated business
entities are profit-driven “amoral calculators” (Kagan & Scholz 1984). Thus, only
fear of imminent legal penalties that exceed the cost of compliance can induce
profit-seeking firms to invest in compliance with regulatory demands. Each tough
legal penalty, it is assumed, sends a “threat message” that reverberates through the
community of regulated businesses. As the perceived risk and cost of violations
thereby increases, business executives increase their investments in compliance.
On the other hand, a considerable body of socio-legal scholarship suggests
rather different hypotheses about compliance-related behavior. Although the
research reported in this article was designed primarily to probe the standard
general deterrence theory, it will be useful first to review some of the alternative
theories that question it.
A. OTHER FEAR-BASED THEORIES
Some alternative theories retain the assumption that fear of legal sanctions
is the primary driver of compliance-related behavior, but question the potency
of the standard general deterrence model. For example, amidst the cacophony
of news, information, and demands of contemporary society, one might wonder
whether business firms actually learn about and attend to legal penalties imposed
on other firms in other places. Even if they do, business executives may not
think that their firm—which may differ in many ways from the sanctioned
firm—faces an enhanced risk of being found in violation and punished (see
Braithwaite & Makkai 1991). Thus, against the general deterrence thesis,
which assumes widespread dissemination and attention to clear deterrence
messages, one might counterpose a “weak signal, weak threat” hypothesis—
that is, the message often does not get through or send a meaningful threat.
Second, some research indicates that the chief driver of enhanced compliance
efforts by regulated firms is not general deterrence (hearing about legal
sanctions against others) but “specific deterrence”—the fear engendered by
the prior experience of being inspected, warned, or penalized themselves—
that is the chief driver of enhanced compliance efforts (Gray & Scholz 1991;
Gray & Shadbegian 2004; Mendeloff & Gray 2004).1
Third, some research indicates that many corporate officials regard the
risk of informal social and economic sanctions as far more salient and
© 2005 Baldy Center for Law and Social Policy and Blackwell Publishing Ltd.
74
264 LAW & POLICY April 2005
threatening than the risk of legal penalties. Negative publicity concerning

environmental harm caused by business facilities can alienate the firm’s
host community, result in loss of market share, and stimulate closer, more
suspicious scrutiny by regulators. Hence many corporate officials speak of
complying not only with regulations but with their “social license”—public
expectations concerning decent environmental performance, “enforced” in a
variety of ways by environmental activists, journalists, local politicians, and
sometimes by customers (Gunningham, Kagan & Thornton 2003). Corporate
concern for maintaining a reputation as a good environmental citizen helps
explain why many firms nowadays regard “overcompliance” with regulatory
obligations as a good business strategy (ibid. 2003; Mehta & Hawkins
1998; Prakash 2000).2 As a result of these concerns, general deterrence mes-
sages often may be redundant, exerting little impact on corporate compliance
behavior.
B. DUTY
Evidence abounds that regulatory violations by business firms are far from
infrequent.3 At the same time, studies indicate compliance with regulations
is much more common than deterrence theory would lead one to expect;
relatively high levels of regulatory compliance exist even when the threat of
legal enforcement appears to be remote.4 One possible explanation, referred
to above, is that many firms fear the negative publicity and the social, polit-
ical, and economic sanctions that can flow from serious violations. Another
explanation, suggested by a variety of scholars, is that for most firms, com-
pliance stems not from fear of legal sanctions, but from a sense of social or
legal obligation. Socio-legal research indicates that in democratic societies
with a strong rule of law tradition, most business managers have “internalized”
(or agree with) the social norms that under-gird many regulatory rules. Such
social norms, as articulated by Vandenbergh (2003: 88, 95) include “An
individual should not cause harm to human health,” and “An individual
should not harm the environment.”5 Moreover, most business executives in
the United States, it is not unreasonable to believe, are generally committed,
as a matter of socialization and citizenship, to complying with duly enacted
laws and regulations (Malloy 2003: 464–75). Furthermore, officials in charge of
corporate compliance efforts often are professionals—environmental engineers,
safety experts, and so on—who, like the chief nurses in the Australian nursing
homes studied by Braithwaite and Makkai (1991), have a strong sense of duty
about compliance with regulatory norms.6 May (2004) found that residential
construction company officials, in describing their motive to comply with
building code provisions, ranked “duty to comply with building requirements”
(as well as maintaining a reputation for quality) as much more important
than fear of regulatory fines. For firms sensitive to their normative obligations,
too, one might hypothesize that general deterrence signals are redundant,
adding little, if anything, to compliance efforts.
75
Figure 1. Fear, Duty, and the Role of General Deterrence.
C. MIXED MOTIVES
While some alternative theories of corporate compliance focus on fear of

legal sanctions (general and specific deterrence) and social sanctions, and
other theories focus on the potency of felt normative obligations, another
approach would emphasize the interaction of “fear” and “duty” and of variation
in motives across firms. Consider the schematic diagram in Figure 1, which
posits that, in principle, firms might display one of four combinations of duty
and fear of sanctions. This suggests that deterrence models would be meaningful
for those firms that score high on fear, but ineffective for those that are low
on fear, and they would be redundant for those that are high on duty.7
1. Implicit General Deterrence

If many firms are motivated by both fear and duty, general deterrence mes-
sages may matter but in a much more diffuse, non-specific manner than that
suggested by classic deterrence theory. Imagine that many businesses are
motivated by a combination of fear of sanctions (legal and social) and of felt
obligations to comply with most regulations.8 In polities like the U.S. with
a rich tradition of the rule of law and a history of significant regulatory
enforcement, such business firms might simply assume that (1) governmental
regulations (or the statutes that give rise to them) prescribe legal sanctions
for violations, (2) significant violations of those regulations entail a fairly
substantial risk of detection and punishment (and of related reputational
costs), and (3) it is both prudent and right to commit to a policy of full
compliance. For officials at Firm A, then, simply learning about an applicable
regulatory requirement evokes some level of perceived threat (plus a felt legal
obligation), inducing it to increase its compliance-related efforts—much as
a motorist in a strange city or country responds to posted speed limits and
other traffic signs. Firm A, therefore, would respond to the regulation
regardless of whether it hears about specific punishments of Firms B and C
for violating that rule. Such an “implicit general deterrence” mechanism
76
can be contrasted with the “explicit general deterrence theory” mentioned

at the outset, which assumes that learning about punishments against Firms
B and C will increase A’s fear that violations will be detected and punished,
motivating A to comply. In the “implicit general deterrence” theory, however,
variation in the frequency and severity of sanctions may have little effect, as
firms pay more attention in formulating their compliance strategies to
social, economic, and normative pressures—rather than carefully matching
compliance expenditures to their calculations of the precise level of legal risk.
2. The “Reminder” Function of General Deterrence Messages

Assume that because of the combination of legal threat, normative obliga-
tion, and concern for preserving their reputation as good corporate citizens,
most enterprises in the contemporary United States are committed to comply
with most governmental laws and regulations. That is not inconsistent with
the repeated finding, noted earlier, that violations of environmental regulations
are common. A great many violations occur, a number of studies indicate,
because corporate managers’ commitments to comply with the law, even if
embodied in corporate policies or prescribed routines, are neglected by
individual employees or sub-units that are subject to conflicting pressures,
or who misunderstand what precisely is required by a complex and changing
array of federal, state, and local regulatory demands (Malloy 2003; Spence
2001: 972–73).9 When that is the case, explicit general deterrence messages may
not motivate firms to comply, but they may serve as a reminder of preexist-
ing commitments to comply.10 Thus learning about legal penalties against
Companies B and C may lead managers in Company A to check whether
their compliance routines are being followed.
3. The “Reasssurance Function” of General Deterrence Messages

Lastly, one might imagine that explicit general deterrence messages matter
not because of the threat they signal but because of the symbolic reassurance
they provide to companies that make costly compliance-related investments.
Learning of penalties against Firm B reassures A that it will not be at a
competitive disadvantage vis-à-vis firms who cut costs by violating the law.
Chester Bowles (1971: 25), reflecting on his job as head of the U.S. Office
of Price Administration during World War II, said that 20 percent of the
population would comply with any regulation, 5 percent would attempt to
evade it, and the remaining 75 percent would go along with it as long as the
5 percent were caught and punished.11 Officials on other regulatory agencies
often echo that theory, arguing that penalizing the “bad apples” helps keep the
“contingently good apples” good (Bardach & Kagan 2002). Thus, the reminder
and reassurance functions interact to support and reinforce the assumptions
of implicit general deterrence: that rules have associated sanctions, sanctions
are enforced, and that compliance is both prudent and right.
77
4. Examining Explicit General Deterrence

The general theoretical issue, then, is how potent are “explicit general deterrence”
mechanisms as compared to the alternatives mentioned. The research
reported in this article is not structured as a rigorous test of the various
hypotheses, or even of the explicit general deterrence hypothesis. It does
represent, however, some first steps toward exploring the effects of explicit
general deterrence messages. To that end, we chose eight significant enforce-
ment actions and penalties (“signal cases”) announced by the U.S. Environ-
mental Protection Agency (EPA) between June and December of 2000. We
then conducted a telephone survey of officials in 233 business firms in the same
industries as those that had been penalized in the signal cases. In addition
to this survey, we conducted in-depth interviews with officials at eighteen
chemical manufacturing facilities and seventeen electroplating facilities in
the states of Washington and Ohio, gathering more detailed information
about their motivations and responses.12 The results of these more detailed
interviews are discussed in Gunningham, Thornton, and Kagan (2005). We
draw on that article however, in the discussion and conclusion of this article.
In the eight-industry survey discussed in this article, we focused on the
following general issues:
1. When regulatory penalties are imposed against a particular violator, to

what extent do other companies in the same industry actually learn about it?
2. Do respondents who know of regulatory enforcement actions against
other firms have, on average, higher perceptions of the risk of detection
and legal sanction than those who know of fewer enforcement actions?
3. Are respondents who know of regulatory enforcement actions against other
firms more likely, on average, to have taken environmental compliance
actions in response to hearing of such enforcement actions?
II. METHODOLOGY
A. THE SIGNAL CASES
To identify the “signal cases,” we examined all U.S. EPA headquarters press
releases from January 2000 through June 2001 that announced final enforcement
actions (n = 112).13 From these, we selected eight, issued between June 2000
and December of 2000.14 The eight were chosen to include a variety of industries,
localities, and penalties. We particularly wished to include some prosecutions
for behavior that would not be seen as obviously criminal. Table 1 describes
the infractions and the penalties assessed in these “signal cases.”
For each signal case, we identified all facilities in the same state and industry,
relying on a variety of sources.15 A random sample of these facilities was
contacted by telephone, seeking the “person responsible for environmental
compliance” at each facility. Officials in 233 facilities agreed to be interviewed,
78
Table 1. Signal Cases
268
Industry Infraction Company Fine Penalty Jail Sentence Individual Fine
Electroplating, The VP of a Denver plater, who, despite 56 $250,000 12 months +100hrs

Colorado warnings over ten years allowed Zn, Cd, Cu, Cr, community service
and Ni to be continually discharged into the Denver
municipal sewers.
Waste Water The district manager of a Rodeo, California, 5 months prison + 5 $3,000
Treatment, treatment plant who admitted to allowing months home
LAW & POLICY
California wastewater to bypass a chlorine contact chamber confinement + 1 year

and to tampering with monitoring methods on 473 probation
days between 1995 and 1997.
Chemical In 1995, a plant in KY stored fuming sulfuric acid $850,000 penalty +
Manufacturing in a tank that had cast iron piping instead of steel $650,000 on an
April 2005
or Blending, piping. The iron corroded, and the company did not emergency notification
Kentucky inspect the piping. This resulted in about 24,000 system
79
gallons of sulfuric acid solution being released into
the air in a four-hour period, creating a chemical
cloud. One thousand nearby residents had to be
evacuated and several were treated for burns of their
eyes, nasal passages and lungs.
Aluminum An aluminum fabricator in Port Allen, LA, who $1.1 million 5 years 100 hours of $2,000 to $5,000
Fabrication— discharged wastewater contaminated with hexanol probation community service
Southern States and with a COD of 1,737 ppm (13 × their permit
limit) into an inter-coastal waterway
Waste Water South Bay Utilities of Sarasota county, who $1.3 million $445,000
Treatment, discharged an estimated 290 gallons of inadequately (president of
Florida treated wastewater, along with additional periodic the company)
discharges amounting to 1.5 tons of nitrogen in a
two-year period, into Dryman Bay.
Table 1. Continued
Industry Infraction Company Fine Penalty Jail Sentence Individual Fine
Steel A corporation that settled allegations that it failed Civil penalty of $9

Fabrication, to control the pollution at eight steel mini-mills, million; $4 million on
Indiana resulting in thousands of tons of illegal air emissions environmental projects
of NOx, and mismanaged discharges of K061 dust $85 million on new
in the soil and groundwater. The company contends control tech.
that it had not violated any environmental law.
Asbestos While carrying out an asbestos abatement project, 41 months $59,700
Abatement between December 1997 and March 1998, the restitution
80
Services, New company failed to notify the EPA; knowingly sent
York workers into an asbestos “hot zone” for more than
twelve weeks, without providing them with
protective gear, or even informing them of the
presence of asbestos; failed to have a certified
contractor perform the work, to properly wet and
bag the asbestos, to properly label the containers
Thornton, Gunningham & Kagan
filled with asbestos, and to dispose of the asbestos
at a landfill approved for that purpose.
Chemical A chemical company in Westlake, LA, was charged $4.5 million penalty
Manufacturing, with releasing CFCs into the air in excess of the 35% and Fund an
Louisiana limit and then repeatedly failing to locate and repair “environmental
leaks. justice” project in
Westlake, LA
DETERRENCE 269
a response rate of 80 percent.16 Approximately 70 percent of the facilities whose

officials we interviewed had fewer than 100 employees; and only those in the
chemical industry had a significant (25 percent) proportion with more than
1,000 employees.
B. MEASURING KNOWLEDGE OF ENFORCEMENT ACTIONS
To assess respondents’ knowledge of regulatory enforcement action against

other firms, we employed three related but distinct measures:
1. Quantum of Knowledge
Respondents were asked: “In the last year or two, about how many instances
can you think of where a company, anywhere in the U.S., was fined for an
environmental problem.” The question was repeated with respect to recollec-
tion of individuals who had been fined in their personal capacity or imprisoned.
2. Knowledge of Particular Examples

Respondents were asked if they could recall a particular example of a person
or plant being penalized for an environmental crime. Beyond that, they were
not prompted. If they could recall an example, they were then asked what
the company had done that led to the penalty, what the penalty was, and
when and where the event had occurred.
3. Signal Case Knowledge

Respondents were given a description of the infraction that had led to the
EPA press release case for their industry. Respondents were then asked if
they recognized the signal case.
C. DEMOGRAPHIC VARIABLES
With respect to their own companies, respondents were asked:
1. The number of people employed by the company as a whole (company

size). Companies were then divided into two categories: small—less than
100 employees, or large—100 or more employees.
2. What percentage of their time they spent of environmental work (degree
of environmental professionalism).
D. RISK PERCEPTION VARIABLES
A number of questions sought to assess respondents’ perception of various

legal risks associated with regulatory enforcement:
81
1. Perceived Risk of Facility Closure. Respondents were asked: “In practice,

on a scale of 0 to 100, how often do you think that legal environmental
penalties lead to a plant being shut down?”
2. Perceived Risk of Detection. Later, respondents were asked to consider a
hypothetical situation modeled on the signal case. For example, chemical
manufacturers and blenders in Louisiana were asked: “Assume for a moment
that there was a chemical manufacturing plant that released CFCs into the
air, 35% in excess of their permit limits, and then repeatedly failed to locate
or repair the leaks that led to this excess. On a scale of 0 to 100, what do you
think the chances are that the plant would be found out by law enforcement?”
3. Perceived Risk of Company Fine (for the same scenario). “If they were
found out, on a scale of 0 to 100, what do you think the chances are that the
plant would be fined? “Can you give me a ballpark estimate of how much
they might be fined—hundreds, thousands, tens of thousands, millions?”
4. Perceived Risk of Incarceration. “If they were found out, on a scale of
0 to 100, what do you think the chances are that the plant operator/owner,
would be sent to jail or prison? Can you give me a ballpark estimate of
how long he might serve—weeks, months, years?”
5. Perceived Risk of Individual Fine. “If they were found out, on a scale of
0 to 100, what do you think the chances are that the plant operator/owner,
would be fined? Can you give me a ballpark estimate of how much he
might be fined—hundreds, thousands, tens of thousands, millions?”
E. ENVIRONMENTAL ACTIONS IN RESPONSE TO DETERRENCE SIGNALS
At the end of our interviews, after asking respondents if they recalled a signal
case, respondents were asked if hearing about a fine or prison sentence at another
company in their industry ever made them respond by: (1) reviewing their
environmental programs; (2) changing their management plans; (3) changing
how they kept track of or monitored things; (4) changing their employee
training; (5) changing their equipment; or (6) changing their physical plant
in some other way. We regarded a company as having “taken an environmental
action” if they reported having taken any of the actions listed above.
Figure 2 summarizes how we operationalized the classic deterrence theory
model, using the variables and measures discussed above.
III. FINDINGS
A. KNOWLEDGE OF ENFORCEMENT ACTIONS
Almost all (89 percent) of our 233 respondents remembered at least one instance
of a fine against some other company, 64 percent recalled at least one fine imposed
on an individual company official, and 31 percent remembered a prison sentence.
On the other hand, firms’ quantum knowledge of fines against other companies
82
Figure 2. Classic Deterrence Model Operationalized
did not appear to be very accurate in terms of the frequency of fines or their
magnitude. Respondents report having heard of far fewer fines than actually
occur. For example, the median number of fines against other companies (any-
where in the United States, in the last year or two) that respondents could recall
was only eight. Yet in Louisiana alone, in a one-year period (July 2001 through
June 2002), thirty-one companies were fined for environmental infractions.17
The majority of respondents (71 per cent) could describe at least one par-
ticular example of a person or business being penalized for an environmental
offense. However, if we examine the particular cases respondents described,
it becomes clear that, on the whole, they tended to remember only those
with unusually large financial penalties and/or cases where someone was
sentenced to jail. Of the 107 respondents who gave a magnitude estimate,
43 percent cited fines of $1 million or more, 67 percent cited fines of $100,000
or more, while 26 percent of respondents who could describe a specific
enforcement action noted that someone at the other company had been
incarcerated. At the same time, respondents overwhelmingly underestimated
the actual penalties when the signal cases were presented as hypotheticals.
Clearly, then, respondents have not been particularly attentive to penalty
information. Nor have they made special efforts to obtain timely and accurate
information. Thus consciousness of the possibility of a significant penalty
was high, but remarkably inaccurate, in the industries sampled.18
When respondents cited the examples of penalty-inducing non-compliance
by other firms, their accounts were often judgmental in tone, suggesting
support for underlying social norms condemning harm to the environment
and complying with law. For example:
Bottom line was they didn’t care and they did something that they thought they
would not get caught for.
Illegal removal of asbestos, dry removal (which is a violation), improper disposal,
improper personnel. [Interviewer: is this all in the same case?] No, different
cases, but you’d be surprised at what the EPA and DEP agents will find on a site.
It’s unbelievable.
83
Dumping in the creek. It killed bunches of fish.

They had an accident due to mismanagement, let hundreds of gallons of sewage
go into the ocean. It was on the news.
Not sure [what he did] but I think he screwed up and then tried to cover it up.
They polluted the local rivers with chemicals.
Burying toxic waste in ground and it leaked into the water.
These comments also seem to support the notion that explicit general deterrence
messages serve a “reassurance function,” informing contingent “good apples”
(firms committed to compliance for a combination of normative, reputational,
and “implicit general deterrence” reasons) that they are not foolish for doing
so, since their competitors who “cheat” are getting caught and punished.
Strikingly, diffusion and memory of the signal case was far from ubiquitous.
On average, only 42 percent of respondents recognized the specific EPA
“signal case” (see Table 1, above) in their industry.19 One reason may be that
the signal cases, despite their seriousness, generally did not get widespread
publicity in the news media.20
Of the respondents who were able to describe at least one particular example
of an enforcement action against another firm, 97 percent remembered the
infraction that led to the enforcement action, while somewhat fewer—83 per-
cent—recalled the penalty given as a result of the enforcement action. Of
the eighty-seven respondents who said they had heard of the signal case, only
61 percent believed they could remember the penalty in the case.
In sum, although 58 percent of respondents did not recognize a vignette
based on the signal case, previous penalties against other firms have a
cumulative effect: most firms are quite aware that environmental penalties
have been imposed on violators. At the same time, recollection of sanctions
against other firms tend to be general: respondents remember the infractions
more fully than the precise penalty, and they are more likely to think of
instances of fines against companies than fines against individuals or incar-
cerations. Indeed, a significant minority of respondents could not recall any
particular instance of a penalty against an individual.21 Thus the general
deterrence message as received is somewhat weak and diffuse, but loud
enough to create a noticeable background noise, so that most firms are
aware of its existence. Our in-depth interviews in two industries revealed a
similar effect (Gunningham, Kagan & Thornton 2005).
B. RISK PERCEPTION
Most respondents did not think that environmental penalties would result
in the closure of an offending facility.22 On the whole, however, respondents
perceived the probability of detection for serious infractions, such as those
described in the signal case, to be high; the median perception of detection
risk was 70 percent. However, respondents’ risk-of-detection perceptions were
highly variable, ranging from close to 0 to 100 percent in most industries.
84
Respondents generally felt that if a serious infraction resembling the signal

case were to be detected, the offending company would be penalized: 92
percent of respondents felt the odds of a company fine were greater than
50:50. But respondents were far less certain that an individual company official
or owner would be fined: 7 percent of respondents believed there was no
possibility of an owner or operator being fined in their personal capacity,
while 11 percent of respondents believed he would certainly be fined. The
median risk-of-individual-fine perception was 40 percent. Respondents were
even less certain that an individual would be incarcerated: 53 percent of
respondents believed that the chance that an owner or operator would be
incarcerated for a serious environmental infraction was 10 percent or less.23
Respondents’ expectations of the magnitude of company fines covered an
enormous range, from $0 to $20 million, and similarly, their estimates for
owner/operator fines varied from $0 to $2 million dollars. Individual fines
were always seen as lower than company fines, often by one or two orders
of magnitude. Fifty percent of respondents believed that if a company official
were incarcerated, the length of the sentence served would be six months or
less. The longest period of incarceration envisaged was ten years.24
C. THE EFFECT OF KNOWLEDGE ON RISK PERCEPTION
There was no clear association between knowledge of enforcement actions against

other firms and our measures of respondents’ perceptions of the risk of detection
and punishment. Five linear regression analyses were performed, each modeling
a risk perception variable (likelihood of facility closure, detection, company
fine, jail, individual fine) as a function of company size,25 degree of profession-
alization,26 knowledge (general deterrence),27 and industry. All models were
significant, but the overall goodness of fit for the model of perceived risk of
company fine was particularly weak.28 In general, knowledge variables were not
significantly associated with risk perception—nor was the direction of statistically
insignificant associations consistent for all models. However, recall of particular
examples was statistically significant in the “risk of company fine” model.
After we gave respondents a short description of the signal case as a
hypothetical case, we asked what they thought the possible size of the asso-
ciated company fine might be (assuming the infraction was detected and a
fine was levied). A large number of respondents (68) could offer no estimate.
For those that did, the majority (68 percent) of respondents underestimated
the fine actually imposed by an order of magnitude, 28 percent gave an esti-
mate of the same order of magnitude, and 4 percent overestimated the fines
by an order of magnitude. Those respondents that had heard of the signal
case also tended to underestimate the fine, but less often (59 percent) than
those who had not heard of the signal case (74 percent).
On the other hand, after being told the actual penalty in the signal case,
85 percent of respondents felt that the punishment in the case was reasonable.
There was no difference in that regard between those who remembered the
85
signal case and those who did not. Of the respondents who felt that the
penalty had been unreasonable, slightly fewer than half (40 percent) felt that
the punishments given were unreasonably stringent, while the remainder (60
percent) felt that the punishments were too lenient. These findings suggest
considerable support among respondents for tougher legal sanctions against
firms that had committed serious violations. And this is consistent with the
notion that publicized penalties against other firms serves a “reassurance
function” for firms that regard themselves as compliant “good apples.”
D. COMPLIANCE-RELATED BEHAVIOR
As noted earlier, at the end of our interviews, after asking respondents if they
recalled a signal case, respondents were asked if hearing about a fine or prison
sentence at another company in their industry ever made them respond by
reviewing their environmental programs or changing aspects of their operations
or compliance program. We regarded a company as having “taken an environ-
mental action” if they reported having taken any of the actions listed in Table 2.
Overall, 65 percent of facilities had taken an environmental action.29
In response to general deterrence signals, as shown in Table 2, facilities
were most likely to review their programs (57 percent did) than to change
any aspect of their behavior, and least likely to change their employee training
(only 23 percent of facilities did). However, 32 percent of facilities reported
having changed equipment. This suggests that a substantial fraction of
facilities respond proactively to environmental enforcement actions taken
against other facilities in their industry, and that the response is strong
enough in some cases to induce costly equipment changes.
E. KNOWLEDGE, RISK PERCEPTION, AND BEHAVIOR
What distinguishes firms that reported environmental actions in response to

deterrence messages from firms that did not? A logistic regression model of
company environmental action as a function of demographic, knowledge,
and risk perception variables30 was developed. Table 3 presents descriptive
statistics for the variables employed in the model. Table 4 presents the results
of the logistic regression.
Company size was significantly and positively associated with the likelihood
of taking environmental action. The degree of professionalization variable
was not significantly associated with taking environmental action.
With respect to knowledge variables, the number of particular examples of
enforcement actions that respondents could describe (0, 1 or 2) was signi-
ficantly and positively associated with whether a respondent reported having
taken an environmental action in response to deterrence signals. On the other
hand, remembering the signal case, or remembering a larger number of
instances of enforcement action in the last year or two, were not significantly
associated with taking environmental actions.
86
276
Table 2. Components of Corporate Environmental Action % of respondents reporting action in response to deterrence signals
LAW & POLICY
Asbestos Electroplating Sanitary Chem Al Manu Sanitary Steel Fab Chem All
– NY – CO Tx – CA Manu /Blend – KY – South – FL – IN Manu-LA
Took any environmental action 74 71 51 93 42 68 62 65 65

April 2005
Reviewed existing programs 54 53 42 85 38 66 59 64 57

Changed management plans 21 23 38 54 15 28 41 57 35
87
Changed how kept track of or 8 36 41 75 23 28 28 35 35
monitored things
Put in new equipment 0 36 36 30 31 25 45 52 32
Changed employee training 29 29 21 64 15 10 17 43 27
Changed physical plant in some way 0 27 21 30 27 13 28 39 23
Table 3. Responses to General Deterrence Messages: Descriptive Statistics
% Valid Missing
Took environmental action in response to deterrence signal 63 227 6

Company size 224 9
Large (>100 employees) 27
Percentage time spent on environmental work 228 5
0–25% 33
26–75% 33
Greater than 75% 33
No of instances of company fines recalled 228 5
0 11
1 6
2–5 25
6–10 18
>10 39
Maximum 2,000
Remember a particular example 232 1
0 29
1 45
2 26
Heard of the signal case 42 229 4
Probability of Detection* 228 5
0–25% 23
26–75% 41
76–100% 36
Probability of Company Fine 226 7
0–25% 4
26–75% 12
76–100% 84
Risk** 225 8
0–2500 28
2501–7500 42
7501–10000 30
Magnitude of Company Fine (dollars) 196 37
0 1
Thousands 9
Tens of thousands 38
Hundreds of thousands 18
Millions or more 34
Probability of Facility Closure 219 14
0 50
1–10 35
11–25 11
26–75 3
76–100 1
Notes: *Probability of Detection = Response to the question: “on a scale of 0 to 100, what do
you think the chances are that the plant (in hypothetical based on signal case) would be found
out by law enforcement?” Estimated Probability of Company Fine, Magnitude of Company
Fine, and Probability of Facility Closure measures based on similar question about fate of
company in hypothetical based on signal case.
**Risk = probability of detection × probability of company fine.
88
Table 4. Logistic Regression Model of Corporate Environmental Action31

Dependent Variable: Taking environmental action in response to deterrence signals
(binary)
B S.E. Wald df Sig. Exp(B)
Demographic Variables
Company size (large/small) 1.254 0.491 6.529 1 0.011 3.504
Degree of Professionalization 0.008 0.006 1.838 1 0.175 1.008
Knowledge Variables
Number of instances of company fines 0.002 0.002 0.840 1 0.359 1.002
Recall particular examples (0, 1, or 2) 0.980 0.289 11.516 1 0.001 2.665
Recognize signal case 0.386 0.405 0.908 1 0.341 1.470
Risk Perception Variables
Risk = prob of detection × prob co. fine 0.000 0.000 0.887 1 0.346 1.000
Magnitude of company fine (0, 1, 2, 3, 4) − 0.013 0.184 0.005 1 0.944 0.987
Risk that penalties will lead to closure 0.072 0.029 6.227 1 0.013 1.074
Constant − 1.775 0.691 6.608 1 0.010 0.169
Notes: Shaded and italicized results show variables significant at or below a p = 0.05 level.
Figure 3. Summary of Results
Of the risk perception variables, only the perception that penalties might lead
to facility closure was significantly associated with taking an environmental
action.32 However, this result appears to be driven by the electroplating
facilities in the sample, and is no longer significant (p = 0.095) if electro-
plating cases are excluded from the dataset.33 The findings regarding the
associations between variables are summarized in Figure 3 above.
IV. DISCUSSION
Classic deterrence theory predicts clear relationships between knowledge of

“high profile” enforcement actions (fines and incarcerations) and improved
89
compliance-related behavior. In most descriptions of the theory, (1) regulated

entities are presumed to monitor their environment for information about
enforcement activity and to have heard about high profile prosecutions and
penalties; (2) knowledge of high profile cases is presumed to increase perceived
risk of non-compliance; and (3) higher perceived risk of legal sanctions is pre-
sumed to improve overall compliance-related behavior.
Our results provide some, but very limited, support for this theory. The
majority of firms (63 percent) report having, at some point in the past, taken
an environmental action in response to hearing about an enforcement action
at another company. Our questions did not distinguish whether or not it was
knowledge of the signal case (as opposed to other penalty cases) that triggered
responsive environmental action. But employing a series of assumptions, we
can estimate that 10 –20 percent did respond to the signal case.34
On the other hand, we find only a weak association between increased
information about other penalty cases and increased perception of legal
risk. All other things being equal, respondents who recalled more particular
enforcement actions against others did report significantly higher perceived
risk of being fined (for violation like that in the signal case). However, no
other knowledge variable was significantly associated with increased risk per-
ception, and firms with higher risk perceptions of detection or fine were not sig-
nificantly more likely to have taken an environmental action than those with
lower risk perceptions.
In other ways, too, much of our data does not support the explicit general
deterrence theory. Firms’ quantum knowledge of fines against other companies
did not appear to be very accurate in terms of the frequency of fines or
their magnitude.
Respondents have not been particularly attentive to penalty information,
nor have they made special efforts to obtain timely and accurate information,
even though the classical deterrence model assumes that companies would
study and quantify the legal risks associated with noncompliance. In addition,
the lack of significant association between risk and the magnitude of fines,
and taking environmental action does not support the traditional model.
We are left with a puzzle: Why do we find a direct association between
recall of particular examples of prosecutions and improved environmental
behavior (see Figure 2) and yet find no chain of significant associations
between recall of particular examples, increased risk perception, and improved
environmental behavior? Our data provides no direct answers, but it does pro-
vide evidence relevant to several of the alternative hypotheses set forth earlier.
One possibility we mentioned earlier is that it is not fear of formal legal
sanctions that drives most firms to take environmental compliance actions.
Rather, environmental behavior stems more from fear of informal sanctions,
such as damage to a company’s reputation or to an environmental manager’s
job or professional standing. Thus, knowledge of enforcement actions might
lead to corporate environmental measures, not by increasing fear of legal
penalties, but by increasing fear of informal sanctions for violations.
90
We find some support for this theory in our data, but some counter-evidence
as well. First, some respondents, when describing the particular examples of
enforcement actions they remember, refer to cases in which individuals lost
their license to practice, or their job. Our lengthy qualitative interviews with
environmental managers in the chemical and electroplating industries
(Gunningham, Thornton & Kagan 2005) also suggest that, at least for some of
them, fear of losing a job or being sanctioned by an employer are more salient,
as possible consequences of serious violations, than are legal penalties. And
returning to the larger sample discussed in this article, those respondents with
greater fear of facility closure did not have, on average, higher perceptions
of the size of possible fines or the probability of being fined.35 On the other
hand, when describing the particular examples of enforcement actions they
recalled, respondents were far more likely to focus on formal sanction (“the
guy went to jail,” “they got a huge fine”) than they were to focus on informal
consequences (“it was all over the newspapers”). Thus it appears that more
dramatic legal sanctions are more likely to be remembered as salient than
are informal sanctions.36
The idea that general deterrence serves primarily as a “reminder mechanism”
provides a more plausible possible explanation for our findings. In this view,
deterrence signals remind the “contingent good apples”—firms already
committed to compliance as a general business strategy—that non-compliance
can occur due to slippage in their company’s own self-regulatory systems.
For good apples, compliance with environmental regulations is a key to
social and political legitimacy. For environmental managers in such firms,
the hypothesis continues, social and self-definitions of “goodness” require
continued compliance. Good apples do not calculate and calibrate the costs of
non-compliance; they assume that those costs are potentially disastrous. On
occasion, a deterrence signal will inform a good apple of non-compliance in their
own facility stemming from employee error or deviance, or of non-compliance
with a regulation they were unaware of or had interpreted incorrectly; hence
the signal will spur them into more than simple confirmation routines. In this
way, information could affect behavior without changing risk perceptions.
We find some support for this theory. First, the reminder function of
deterrence comes through quite dramatically in our in-depth interviews
(see Gunningham, Thornton & Kagan 2005). Enforcement actions are described
as “head turners” that draw the attention of environmental managers.
Second, in our eight-industry survey, the particular examples described by
respondents focus on large and dramatic formal sanctions with disastrous
consequences. Third, our surveys suggest that actions that threaten environ-
mental quality are seen as intrinsically “bad”—and hence that managers
actually valued environmental ends. For example, of the respondents who
thought that the environmental performance of the industry had improved
over the last fifteen years, 36 percent believed that the reason for this
improvement was (at least in part) attributable to greater awareness about
the environmental impacts of industrial operations and changes in their
91
industry’s attitude toward the environment. Essentially, this explanation is

a definition of the value of environmental ends and a definition of actions that
threaten environmental quality as unfavorable. In addition, the overwhelming
majority of facilities felt that the punishment in the signal case was reasonable
(86 percent) or unreasonably lenient (8 percent). And lastly, the examples of
non-compliance cited by respondents were often clearly judgmental in tone;
this suggests support for the notion that explicit general deterrence messages
serve a “reassurance function,” informing contingent “good apples” (firms
committed to compliance for a combination of normative, reputational, and
“implicit general deterrence” reasons) that they are not foolish for doing so,
since their competitors who “cheat’ are getting caught and punished. Thus
the legitimacy-reputation mechanisms are consistent with our findings of
the pattern of association between variables.
On the other hand, the weak support for the traditional “explicit” general
deterrence theory in our findings may reflect the possibility that the measures
we constructed do not accurately reflect the underlying constructs we hoped
they might. For example, to obtain our risk perception measures, we asked: if
a company is violating in this manner, what do you perceive the risk of detection
or punishment for that company to be? However, we could not ask respondents
how likely they were to commit the same violations. Thus, for “good apples”
we might have elicited an estimate of risk of detection for “bad apples” (not
their own firm). Nor did we directly ask: Did hearing about an enforcement
action at another company ever change your perception of risk?
In addition, we looked for an association between knowledge of enforcement
actions and risk perception in a cross-sectional study. Thus causality cannot be
inferred from our findings. Higher risk perceptions could cause greater knowledge
(rather than the other way around) because people with higher risk perceptions
look harder for enforcement information and remember it better. However, in
either case, one would expect a significant association, which we did not find.
Furthermore, we obtained only a snap shot of current risk perceptions,
but asked for an aggregate measure of behavior change, asking if companies
had “ever” taken environmental actions in response to deterrence signals.
Our measure thus does not rule out the possibility that firms that acted in
response to deterrence signals had higher risk perceptions at that prior time.
Lastly, it must be remembered that this research was conducted in the
United States in the early twenty-first century, more than a quarter century
after American states and the federal government started serious enforce-
ment of environmental laws. Hence the “implicit general deterrence” mech-
anism has matured, so that the enforcement and normative legitimacy of
environmental regulations is taken for granted by many firms. And social
and political support for environmental norms has given many companies a sub-
stantial economic stake in avoiding a reputation for being bad environmental
citizens. Thus, our research has little to say about the importance of explicit
general deterrence messages at earlier stages in regulatory programs, when their
value added may well be greater.
92
CONCLUSION
Our research provides only weak support, at best, for the classical “general
deterrence” hypothesis (which we would now label “explicit general deter-
rence”). Fewer than half (42 percent) of 229 respondents in regulated busi-
nesses recognized and remembered the specific signal case, suggesting at
least partial support for the “weak signal” hypothesis. On the other hand,
general deterrence seems to have a cumulative effect on the consciousness
of regulated companies: 89 percent of our respondents remembered at least
one instance of some company having been penalized for an environmental
violation in the past year or two. And some 63 percent of the companies we
surveyed reported having taken some environmental protection measures
after learning about penalties against other companies. Most often, the
reported reaction was to review their own compliance programs, but many
also changed equipment, monitoring practices, and employee training.
Yet many relationships predicted by the classical deterrence model did
not show up in our data. For example, respondents who recognized the signal
case or referred to a larger quantum of other cases were not more likely to
report having taken environmental action in response. Moreover, those
officials who saw the risk of formal detection and punishment as relatively
high were not, on average, more likely to report taking environmental measures
in response to general deterrence messages. Company managers were not
closely attentive to or knowledgeable about the penalties assessed against
violators, generally underestimating them. This suggests to us that penalties
against other firms—at least in the United States near the beginning of the
twenty-first century—play a somewhat different role from the one embedded
in the classical general deterrence theory, which assumes that the imminent
threat of legal punishment is the primary driver of compliance efforts.
Our survey and in-depth interview (Gunningham, Thornton & Kagan 2004)
evidence, rather, suggests that for most firms, general deterrence primarily
serves a reminder and a reassurance function. Most of the companies (some
63 percent of our sample) that respond to deterrence signals by taking some
environmental action probably are predominantly stimulated to check
whether they are in compliance; they are not firms (as assumed by deterrence
theory) who know that they are not in compliance and are stimulated by legal
threat to change their ways. We speculate that the mechanism by which
deterrence affects the behavior of these two groups is different. For those that
know they are not in compliance, deterrence cases that are exactly relevant
to their particular circumstances may well increase their risk perceptions
and change behavior (the classic deterrence model). But for the “good apples”
—firms that are generally committed to compliance for a variety of normative
and reputational reasons—each deterrent signal reinforces their perception
of the need to continue compliance activities and of the potential disastrousness
of non-compliance. Sometimes, a deterrence signal prods “good apples” to
check and learn that they are no longer in compliance and need to take further
93
action. Deterrence signals both reassure “good apples” that free-riders will
be punished and remind them to make sure that they are responsible corporate
citizens with no need to fear the social and economic costs that can be triggered
by serious violations.
dorothy thornton holds a Ph.D. in Health Services and Policy Analysis from the
University of California, Berkeley and currently is a research associate in Berkeley’s
School of Public Health.
neil gunningham holds professorial research appointments in the Regulatory Institu-

tions Network, Research School of the Social Sciences, and in the School of Resources,
Environment and Society, both at the Australian National University.
robert a. kagan is Professor of Political Science and Law at the University of California,
Berkeley.
NOTES
1. See also studies summarized in Vandenbergh (2003: 119–22).

2. In a recent survey of members of the American Bar Association Section of
Environment, Energy and Resources (practitioners of environmental law in both
the public and private sectors), Ruhl et al. (2002: 26) found that “Respondents
agreed or strongly agreed that noncompliance hurts the corporate public image
(85 percent), creates friction between business and government (81 percent),
increases administrative costs (82 percent), and demoralizes company personnel
(74 percent).”
3. A 2003 EPA survey found that approximately 25 percent of the 6600 facilities
with the largest discharges had engaged in “significant noncompliance” (U.S.
EPA 2003; see also Rechtschaffen 2004: 776).
4. Summarizing a number of studies, Vandenbergh (2003: 127) concludes, “Despite
the small risks of inspections and the small size of sanctions, compliance rates
[for environmental regulatory requirements] are widely regarded to be higher
than predicted by the standard deterrence model.”
5. Vandenbergh (2003: 90) cites a study of environmental managers in the metal
finishing industry (Flannery & May 2000), which concluded that the magnitude
of the human health consequences of environmental decisions is strongly correlated
with expressed intentions to comply with regulations, and “when the magnitude
of the consequences was high, financial cost did not affect the intended decision.”
6. Cordano and Frieze (2000: 635) found that corporate environmental managers
generally express attitudes strongly supporting pollution prevention.
7. Note that in many economic models of legal behavior (and the general deter-
rence theory), the presumption is that firms are low on duty (or that normative
obligation simply is irrelevant), while firms are quite sensitive to the magnitude
of legal threat.
8. Research on individual taxpayers has indicated that “fear” and “duty” tend to
interact in producing compliance (or non-compliance) with income tax law. That
is, taxpayers who are reminded of the threat of legal penalty for failing to report
income tend to display heightened sensitivity to the normative obligation to pay
taxes (Schwartz & Orleans 1967). Conversely, taxpayers to express a stronger
sense of duty to report all income also have a higher subjective estimate of the
94
risk of being caught for cheating (Scholz & Pinney 1995). However, another study
found that among taxpayers with a similar sense of duty, those who had lower fear
of being caught (greater opportunity to cheat) had lower levels of self-reported
compliance—indicating that “fear” has independent effects (Scholz & Lubell 1998).
9. The survey of environmental lawyers mentioned above (Ruhl et al. 2002) discov-
ered strong agreement that the sheer number, complexity, and changeability of
environmental regulations is the chief cause of non-compliance (far outranking
“costs of compliance”) and indeed made it virtually impossible to achieve full
compliance 100 percent of the time, even for committed firms (see also Aoki,
Axelrad & Kagan 2000).
10. Vandenbergh (2003: 73–74), referring to social psychologist Shalom Schwartz’s
theory of norm activation, writes “an enforcement intervention may activate [in
the minds of firm managers] if it provides information about the [adverse social]
consequences of a noncompliant act and the individual’s responsibility for or
ability to prevent those consequences.”
11. In terms of Figure 1, Bowles’s 20 percent of unconditional compliers are “high
on duty”, 5 percent are low on both fear and duty, and the other 75 percent are
contingently high on duty. Some legal theorists, too, have referred to the reas-
surance function of law enforcement. Kahan (1996: 604), for example, writes of
the prevalence of “conditional cooperation,” whereby, a commitment to obey
the law can be undercut by a sense that violators often are not punished, because
of the widespread “desire not to be suckered” (see also Kahan 1997).
12. In the smaller, in-depth study, the facilities were chosen in order to ensure that
the sample included respondents from urban areas (Seattle and Spokane in
Washington, Cleveland and Cincinnati in Ohio as well as rural areas; companies
that operated a number of facilities in a number of states as well as those that
operated only a single facility; and companies that ranged significantly in size
from mom-and-pop operations to multinationals. Response rates were
36 percent (8 out of 22) for WA electroplaters, 45 percent (9 of 20) and for OH
electroplaters. The most common reason given for non-response was lack of time
to participate in a one-hour interview. Response rates were 38 percent (8 of 21)
for WA chemical companies and 56 percent for OH chemical companies (10 of 18).
The non-respondents do not appear to have been disproportionately “bad apples;”
nor were the respondents disproportionately “good apples.” Using the EPA’s
“ECHO” on-line data set (http://www.epa.gov/echo/compliance_report.html), we
found that in 2002–03, the average “quarters in noncompliance” (according to
government inspectors) for electroplaters in our Washington sample was 1.38;
for Washington electroplaters who declined to participate, the figure was 1.25,
slightly less. We also compared electroplaters in our Ohio sample with all Ohio
electroplaters in the EPA database, and the average quarters in noncompliance
for both groups was virtually equal. Respondents in Washington more often
were larger firms than were non-respondents (which nevertheless were slightly
larger, on average, than the industry norm). But in Ohio respondents were about
the same size, on average, as the industry norm, according to the EPA data set.
13. We included only press releases of completed enforcement actions (e.g., we did
not include those simply announcing a prosecution) and excluded those involv-
ing “wholly illegal enterprises,” such as firms that operated entirely outside the
law (midnight dumpers, unlicensed businesses).
14. In truncating the period, we sought to concentrate on actions that were relatively
more recent, so that respondents might have a better chance of remembering
them, but not so recent that news of them might not have had time to circulate
in the industry.
15. We compiled a list of facilities by searching EPA’s Envirofacts database for
facilities in the same state and SIC code as the signal facility. In addition,
95
Switchboard.com and Yellowpages.com were searched for additional facilities in

the same industrial categories as the signal facility, as was Hoovers.com. Where
available, state databases of the relevant facilities were obtained.
16. Response rates were 100 percent for sanitary treatment facilities (n = 40 in
Florida, 39 in California), 76% for aluminum fabricators (26/34), 75% for steel
fabricators (30/40); 73% for chemical manufacturers and blenders (29/40), and
69% for Colorado Electroplaters (22/32), 75% for asbestos abatement companies
in New York (24/32) and 70% for chemical manufacturers in Louisiana (23/33).
Based on our examination of relative noncompliance rates (according to EPA
data) for electroplating firms in Washington and Ohio, see note 13 above, it appears
likely that respondent firms in the eight-industry survey, too, were not dispro-
portionately those with above-average compliance records, as compared to either
non-respondents or the industry norm.
17. The median penalty in this Louisiana example was $16,750 with a range of $2,300
to $1.6 million. Five of the thirty-one penalties (16 percent) exceeded $100,000.
18. All other things being equal, aluminum fabricators recalled significantly fewer
particular examples of enforcement actions than facilities in any other industry.
(Logistic Regression: Recall spc. ex.= f(#employees, professionalization, industry).
No other variables significant.
19. All other things being equal, electroplaters (71 percent) were significantly more
likely to recognize the signal case than were respondents from any other industry.
The more professionalized the environmental staff person, the more likely they were
to remember the signal case (Logistic Regression: Recall spc. ex.= f(#employees,
professionalization, industry). No other variables were significant.
20. We chose a random sample of forty EPA press releases announcing legal sanc-
tions against violators, and searched for media coverage via Lexis-Nexis, major
newspapers, local newspapers, radio and television news transcripts, industry news
outlets, newswires and regional newspaper files. Based on coverage, we found
that only ten of the press releases received wide media coverage (16 to 145 stories);
fourteen cases received “low” media coverage (0 – 6 stories); and sixteen received
“medium coverage (7–15 stories). The apparent threshold for obtaining wide media
attention was an unusually large fine (in excess of $4 million) or an unusually
long jail sentence (greater than 41 months).
21. The more of her time a respondent spent on environmental work the more likely
she was to recognize the signal case. However, less professionalized environmental
staff were just as likely to be able to describe particular examples of environmental
penalties. In general, knowledge did not vary significantly by industry. Only
aluminum fabricators were significantly less likely than facilities in other indus-
tries to be able to describe a particular example of an environmental penalty.
22. For 50 percent of respondents, there was no chance that environmental penalties
would eventually lead to facility closure, and 85 percent of facilities believed
the probability of such a closure was 10 percent or less—but for the remaining
15 percent of respondents, the risk of forced closure was real, and in a very few
cases, substantial. Electroplaters and asbestos abatement companies were likely
to think that fines might lead to facility closure. In fact, none of these respondents
felt that the probability of facility closure was zero. Conversely, the vast majority
of sanitary treatment facility respondents in both California and Florida deemed
closure impossible, which seems a reasonable assessment given the indispensability
of their function. Some chemical manufacturing facilities viewed the probability
of facility closure as reasonably high, while most aluminum fabricators and
steel fabricators viewed it as highly unlikely.
23. Electroplaters had a much higher risk perception than all other industries
(median probability is above 50 percent, while for all other industries it is at or
below 20 percent).
96
24. A large number of respondents did not give an estimate of the size of the penalty.
Many felt that the penalty would depend too much on the specifics of the case and
felt uncomfortable giving a single number, or even a range of possible penalties.
Others (5 percent of respondents) said they had no idea of the size of a possible
company fine, 7 percent had no idea of the size of a possible individual fine, and
13 percent had no idea of the length of a possible jail or prison sentence.
25. Company size is divided into “large” (100 or more employees) and “small” (less
than 100 employees).
26. Measured as a percentage of the time the respondent spent on environmental work.
27. Three different measures used. First, the quantum of fines recalled, categorized
as: none, 1 or 2, 3 to 9, 10 to 15, 16 to 30, more than 30. Second, the number of
particular cases recalled and described (none, one or two), and third, whether or
not the signal case was recalled.
28. Reference industry = sanitary treatment facilities in Florida.
Facility Closure: df = 190, F = 10.168, p < 0.000, Adj R2 = 0.367, Sig Vars: Asbes,
Elec, Steel, Chem-KY (all +ve)
Detection: df = 195, F = 3.679, p < 0.000, Adj R2 = 0.142, Sig Vars: Chem-KY (+ve)
Company Fine: df = 193, F = 1.851, p = 0.043, Adj R2 = 0.050, Sig Vars: Particular
Exs (+ve)
Individual Fine: df = 185, F = 2.390, p = 0.007, Adj R2 = 0.083, Sig Vars: Steel (−ve)
Jail: df = 186, F = 3.532, p < 0.000, Adj R2 = 0.140, Sig Vars: Elec (+ve) SanTx-
CA (+ve)
29. One reviewer speculated that that respondents may have over-reported having
taken an environmental action in response to learning about sanctions against
other firms, because they may have thought that would portray their firm in a more
positive light. But as will be shown in the next section, the most common environ-
mental response reported by respondents was merely to check their compliance
systems, and nothing more, which indicates that they were not inclined to exag-
gerate about how responsive they were to deterrence messages. Conversely, one
could also speculate that firms would be likely to under-report environmental
action in response to news of sanctions, because that might suggest that they had
not done enough in the past. Yet we think that unlikely, too, since the most common
reported response was only to review current corporate compliance systems.
30. Demographic variables included: (1) company size, and (2) degree of profession-
alization, as measured by the percent of time spent on environmental work.
Knowledge variables have been described earlier. The number of instances of
individual fines and the number of instances of jail /prison sentences were not
included in the model because of the high degree of correlation between vari-
ables. The company fine variable was chosen because company fines are the most
common enforcement tool in use.
Risk perception variables included the probability of detection multiplied by the
probability of a company fine, and the magnitude of the company fine. The
magnitude of the company fine was given as $0 = 0, $1,000s = 1, $10,000s = 2,
$100,000s = 3 and $1,000,000s or more = 4. These risk perceptions were for
serious infractions for which federal EPA headquarters had written press
releases, and varied from one industry to another. The probability of individual
fines and jail sentences were omitted because the number of missing values would
seriously bias the data.
The probability that enforcement would lead to facility closure was also
included in the model, this variable was directly comparable across all industries.
31. Number of cases included in the analysis = 176 (= 75.5% of all cases). The model
chi-square is 50.706 which is significant at p < 0.000. The −2 Log likelihood
value is 175.150 and the Cox and Snell R2 is 0.250. A second model was also run
including dummy variables for each industry. The addition of this block of
97
variables was not significant at a 0.05 level and so these variables were not
included in the model (chi-square = 9.812, df = 7, p = 0.199). A correlation matrix
was calculated. No bivariate correlations exceeded 0.30.
32. These results remained essentially unchanged if perceived risk of detection and
perceived risk of company fine were entered in the model separately, instead of
their product. Neither risk, the perceived risk of a company fine, nor perceived
risk of detection were significant, even when the recall of particular examples was
dropped from the model (this was done because recall of particular examples
was significantly associated with perceived risk of company fines).
33. The regression model action = f (size, professionalization, instances, particular
examples, signal case, risk, magnitude of fine, facility closure) was run eight times,
each time excluding those records from a single industry. Recall of particular
examples was found to be significant for all of these models. Size was found to
be significant for all but one model (the model excluding chemical companies
in Kentucky). And the perceived risk of facility closure was found to be sig-
nificant for all but one model (the model that excluded Colorado electroplating
facilities).
34. About 60 percent of facilities reported they had taken an action in response to
hearing about some legal penalty against some other company. About 40 percent
had heard of the signal case. If we apply the 60 percent response figure to the 40
percent, then perhaps 24 percent of facilities took environmental action in
response to the signal case. Since that may overestimate the signal case response,
our guess is that 10–20 percent would be more realistic
35. They did, however, have higher average perceptions of the risk of an infraction
being detected.
36. However, our study of media coverage of enforcement activities suggests that
more dramatic sanctions receive more press coverage, so there might well be an
association between publicity and other informal sanctions and more dramatic
legal sanctions.
REFERENCES
Aoki, Kazumasu, Lee Axelrad, and Robert A. Kagan (2000) “Industrial Effluent
Control in the United States and Japan.” In Regulatory Encounters: Multinational
Corporations and American Adversarial Legalism, edited by R. A. Kagan &
L. Axelrad. Berkeley: Univ. of California Press.
Ayres, Ian, and John Braithwaite (1992) Responsive Regulation. New York: Oxford
Univ. Press
Bardach, Eugene, and Robert A. Kagan (2002) Going by the Book: The Problem of
Regulatory Unreasonableness. Rev. ed. Philadelphia: Transaction Publishers.
Bowles, Chester (1971) Promises to Keep. New York: Harper & Row.
Braithwaite, John, and Toni Makkai (1991) “Testing an Expected Utility Model of
Corporate Deterrence,” Law & Society Review 25: 7– 40.
Cordano, Mark, and Irene Hanson Frieze (2000) “Pollution Reduction Preferences
of U.S. Environmental Managers,” Academy of Management Journal 43: 627.
Flannery, Brenda, and Douglas May (2000) “Environmental Ethical Decision
Making in the U.S. Metal-Finishing Industry,” Academy of Management Journal
43: 642.
Gray, Wayne, and John T. Scholz (1991) “Analyzing the Equity and Efficiency of
OSHA Enforcement,” Law & Policy 13: 185–214.
Gray, Wayne B., and Ronald J. Shadbegian (2005) “When and Why Do Plants
Comply? Paper Mills in the 1980s,” Law & Policy 27: 238–61.
98
Gunningham, Neil A., Dorothy Thornton, and Robert A. Kagan (2005) “Motivating
Management: Corporate Compliance in Environmental Protection,” Law & Policy
27: 289–316.
Gunningham, Neil, Robert A. Kagan, and Dorothy Thornton (2003) Shades of
Green: Business, Regulation and Environment. Berkeley, Calif.: Stanford Univ. Press.
Kagan, Robert A., Neil Gunningham, and Dorothy Thornton (2003) “Explaining
Corporate Environmental Performance: How Does Regulation Matter?” Law &
Society Review 37: 51–90.
Kagan, Robert A., and John T. Sholz (1984) “The Criminology of the Corporation
and Regulatory Enforcement Styles.” In Enforcing Regulation, edited by K. Hawkins
& J. M. Thomas. Boston: Kluwer-Nijhoff.
Kahan, Dan M. (1996) “What Do Alternative Sanctions Mean?” University of Chicago
Law Review 63: 591–653.
Kahan, Dan M. (1997) “Social Influence, Social Meaning, and Deterrence,” Virginia
Law Review 83: 349–95.
Malloy, Timothy (2003) “Regulation, Compliance and the Firm,” Temple Law Review
76: 451–531.
May, Peter J. (2004) “Compliance Motivations: Affirmative and Negative Bases,”
Law & Society Review 38: 41–68
Mehta, Alex, and Keith Hawkins (1998) “Integrated Pollution Control and Its
Impact: Perspectives From Industry,” Journal of Environmental Law 10: 61–77.
Mendeloff, John, and Wayne B. Gray (2005) “Inside the Black Box: How Do OSHA
Inspections Lead to Reductions in Workplace Injuries?” Law & Policy 27: 219–37.
Prakash, Asseem (2000) Greening the Firm: The Politics of Corporate Environmentalism.
Cambridge: Cambridge Univ. Press.
Rechtschaffen, Clifford (2004) “Enforcing the Clean Water Act in the Twenty-First
Century: Harnessing the Power of the Public Spotlight,” Alabama Law. Review 55:
775–814.
Ruhl, J. B., James Salzman, Kai-Sheng Song, and Han Yu (2002) “Environmental
Compliance: Another Integrity Crisis or Too Many Rules?’ Natural Resources &
Environment 17: 24–29.
Scholz, John T., and Neil Pinney (1995) “Duty, Fear and Tax Compliance: The
Heuristic Basis of Citizenship Behavior,” American Journal of Political Science 39:
490.
Scholz, John T., and Mark Lubell (1998) “Trust and Taxpaying: Testing the Heuristic
Hypothesis,” American Journal of Political Science 42: 398 – 417.
Schwartz, Richard D., and Sonya Orleans (1967) “On Legal Sanctions,” University
of Chicago Law Review 34: 274–300.
Spence, David (2001) “The Shadow of the Rational Polluter: Rethinking the Role
of Rational Actor Models in Environmental Law,” California Law Review 89:
917–98.
U.S. Environmental Protection Agency (EPA) (2003) A Pilot for Performance Analysis
of Selected Components of the National Enforcement and Compliance Assurance
Program. Washington, DC: U.S. EPA Office of Enforcement and Compliance
Assurance.
Vandenbergh, Michael (2003) “Beyond Elegance: A Testable Typology of Social
Norms in Corporate Environmental Compliance,” Stanford Environmental Law
Journal 22: 55–144.
99
100
Week 3
101
102
103
104
105
106
107
108
35 66
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
Social Influence, 2015
Vol. 10, No. 2, 109–118, http://dx.doi.org/10.1080/15534510.2015.1007082
Highlighting the injunctive norm to reduce phone-related

distracted driving
Natalie Kerr Lawrence*
Department of Psychology, James Madison University, 91 E. Grace Street, MSC 7704,

Harrisonburg, VA 22807, USA
(Received 15 September 2014; accepted 24 December 2014)
The purpose of this research was to measure the effectiveness of a social norm
intervention to reduce phone-related distracted driving on a university campus.
The intervention involved posting two different signs that highlighted the injunctive
norm (people’s disapproval of the behavior). One of the signs had a strong focus on the
injunctive norm, whereas the second sign merely implied that phone-related distracted
driving was disapproved. Research assistants recorded the phone-related distracted
driving behaviors of people as they drove away from the signs. The results suggest that
injunctive norm messages can be effective in reducing phone-related distracted driving,
but only when they focus people’s attention on the social disapproval of that behavior.
Keywords: social influence; injunctive norms; distracted driving
Phone-related distracted driving has become a dangerous epidemic in our society.

An estimated 26% of all motor vehicle accidents are caused by drivers using a cell phone
(National Safety Council, 2014). Researchers have found that driving while dialing a cell
phone makes the risk of having a crash or near-crash almost three times more likely than
non-distracted driving. Texting makes the risk of a crash or near-crash over 20 times more
likely (Dingus et al., 2006). Despite the danger, an alarming number of drivers continue to
use their cell phones while driving. In a recent survey, 69% of U.S. drivers of ages 18– 64
reported talking on their cell phone while driving and 31% reported sending or reading text
messages or emails. Nearly 50% of drivers of ages 18 – 24 reported texting while driving
(Center for Disease Control and Prevention, 2013). Other research confirms that teenagers
and young adults are much more likely to text behind the wheel than older drivers
(e.g., National Highway Traffic Safety Administration, 2014; Schroeder, Meyers, &
Kostyniuk, 2013).
There are numerous public service campaigns designed to reduce phone-related
distracted driving (hereafter shortened to distracted driving), but the behavior has
remained unchanged (Pickrell, 2014). It is vitally important for applied researchers and
other professionals to develop and assess new interventions to reduce distracted driving.
Social psychologists, because of their expertise in persuasion and social influence, are in a
unique position to combat this problem. Within the social psychological literature, it is
widely known that social norms have a powerful influence on behavior. According to the
focus theory of normative conduct (Cialdini, Reno, & Kallgren, 1990), there are two types
of social norms. Injunctive norms refer to what is commonly approved or disapproved of.
Descriptive norms refer to what is commonly done. For example, most people think that
*Email: lawrennk@jmu.edu
q 2015 Taylor & Francis
133
110 N.K. Lawrence
texting while driving is a bad idea (the injunctive norm) but many drivers engage in this
behavior anyway (the descriptive norm). Injunctive and descriptive norms can be aligned
(e.g., when people do what they approve of) or they can be in opposition (e.g., when
people do what they disapprove of). Survey research indicates that the norms related to
distracted driving are in stark opposition, especially among young drivers (e.g., Lawrence,
2013; Schroeder et al., 2013).
The focus theory of normative conduct states that norms will only affect behavior
when they are focal (i.e., salient) at the time of behavior. An extensive body of research
has demonstrated that both descriptive and injunctive norms can influence behavior when
people are thinking about them (e.g., Cialdini et al., 1990, 2006; Goldstein, Cialdini, &
Griskevicius, 2008; Nolan, Schultz, Cialdini, Goldstein, & Griskevicius, 2008; Schultz,
Khazian, & Zaleski, 2008; Schultz, Nolan, Cialdini, Goldstein, & Griskevicius, 2007).
Normative social influence often occurs through social interaction but it can also happen in
more indirect ways, such as through electronic communications or printed messages.
Using signs and other printed messages is an efficient way to communicate normative
information to a large number of people. Many researchers have used this strategy to
encourage pro-environmental and health behaviors (e.g., Bator, Tabanico, Walton, &
Schultz, 2014; Burger & Shelton, 2011; Cialdini et al., 2006; Nolan et al., 2008; Schultz
et al., 2007; Schultz et al., 2008). Several studies have found that drawing peoples’
attention to descriptive norms via electronic and printed messages can produce desirable
changes in behavior. For example, Burger and Shelton (2011) posted a sign near an
elevator that read, “Did you know? More than 90 percent of the time, people in this
building use the stairs instead of the elevator. Why not you?” The rate of elevator use
declined among people exposed to this descriptive norm sign, but not among people
exposed to a sign that mentioned the health benefits of taking the stairs.
Social norms marketing campaigns, which publicize descriptive normative
information, have become a popular way to discourage unhealthy behaviors such as
collegiate binge drinking (e.g., Neighbors, Larimer, & Lewis, 2004). These campaigns are
guided by the assumption that people sometimes overestimate the number of others who
engage in unhealthy behaviors and correcting their misperceptions can motivate a change
in behavior. Unfortunately, some of these interventions have backfired. In an attempt to
reduce heavy alcohol consumption among college students, researchers distributed
materials that highlighted the descriptive norm: “Seventy-five percent of (school name)
students drink 0, 1, 2, 3, or 4 drinks when they party.” This intervention corrected students’
misperceptions of the norm but it also increased the frequency of drinking among light
drinkers (Clapp, Lange, Russel, Shillington, & Voas, 2003). Researchers believe that
messages like this one can create a boomerang effect by normalizing undesirable behavior
(Cialdini, 2003; Schultz et al., 2007). Because people are motivated to align their behavior
with group standards, the descriptive norm acts like a “magnet for behavior for individuals
both above and below the average” (Schultz et al., 2007, p. 430). Social norms marketing
campaigns may decrease the undesirable behavior among those in the target audience, but
these campaigns may also increase the undesirable behavior among people who are doing
the right thing to begin with.
According to the focus theory of normative conduct, it is unwise to focus people’s
attention on the descriptive norm when the undesired behavior occurs too frequently.
Cialdini et al. (2006) demonstrated this in an applied setting. They posted signs in
Arizona’s Petrified Forest National Park to decrease the theft of petrified wood, a behavior
that was endangering the park. One of the signs induced focus on the descriptive norm:
“Many past visitors have removed the petrified wood from the park, changing the state of
134
Social Influence 111
the Petrified Forest.” Another sign created focus on the injunctive norm: “Please don’t
remove the petrified wood from the park.” The results showed that focusing park visitors’
attention on the descriptive norm increased theft, but focusing their attention on the
injunctive norm decreased it. This study confirms that it is better to highlight the injunctive
norm when trying to change socially disapproved behaviors that are “regrettably frequent”
(Cialdini et al., 2006, p. 4). This study also demonstrates that using descriptive normative
information can backfire even when the undesirable behavior is not performed by the
majority of people. Only a small percentage of visitors to the Petrified Forest National
Park steal petrified wood, but the descriptive norm signage implied that the behavior
was common.
Many of the public campaigns designed to reduce distracted driving focus on
the descriptive norm. These campaigns often feature statistics about the number of
distracted drivers on the road. For example, one prominent campaign features statistics
such as “78% of teens and young adults say they have read an SMS message while
driving” and “71% of teens and young people say they have composed/sent SMS
messages while driving” (www.stoptextsstopwrecks.org). This strategy is problematic
for a couple of reasons. First, the statistics may not be accurate (see Schroeder et al.,
2013). Second, publicizing this descriptive normative information may inadvertently
create a boomerang effect. Discovering that 3 out of 4 teenagers and young adults text
while driving may lead more young drivers to engage in this behavior (or do it more
often). Regardless of the percentages used, if the underlying message is that a lot of
people are driving distracted, the message can backfire (Cialdini et al., 2006). Research
stemming from the focus theory of normative conduct (Cialdini et al., 1990) suggests
that campaigns might be more successful if they focus drivers’ attention on the
injunctive norm rather than the descriptive norm. There are many public campaigns that
involve some type of plea for people to stop texting while driving (which may induce
focus on the injunctive norm), but it is not clear whether these messages are effective.
The purpose of this paper is to report the success of an injunctive norm intervention that
reduced distracted driving.
It is important to note that messages such as “please don’t steal the wood” or “please
don’t text while driving” contain implicit injunctive normative information. These
messages imply that the behavior is socially disapproved, but the injunctive norm is not
made explicit. People may read—and respond to—these messages without thinking
about others’ disapproval of the behavior. For example, some of Cialdini et al.’s
(2006) participants may have interpreted the sign, “Please don’t steal the wood,” as a
command from authority figures (e.g., park rangers) and complied with the sign to
avoid being punished. Indeed, a sign with an explicit injunctive normative message (i.e.,
one that explicitly states that others disapprove of the behavior) may be more likely to
focus people’s attention on the injunctive norm and, thus, more likely to change
behavior.
This research was guided by two primary questions: Are injunctive normative
messages an effective way to reduce distracted driving? Are these messages effective only
when the injunctive norm is made explicit? To answer these questions, I designed a social
norm intervention to reduce the rate of distracted driving on a large college campus. The
intervention involved posting two signs on campus that highlighted the injunctive norm
against distracted driving. Although researchers have used a number of different
techniques to create normative focus (e.g. Kallgren, Reno, & Cialdini, 2000), I believed a
sign would be the most practical way to express the injunctive norm against distracted
driving. One of the signs in this study had a weak focus on the injunctive norm (i.e., the
135
112 N.K. Lawrence
sign’s message implied that others disapprove of distracted driving, but it did not make the
norm explicit). The other sign had a strong focus on the injunctive norm (i.e., the message
clearly stated that other people disapprove of distracted driving). I predicted that the strong
focus sign would lead to a decrease in distracted driving, but the weak focus sign would
not. In recent years, drivers have been inundated with messages designed to persuade them
to stop driving distracted, but the behavior has not decreased (Pickrell, 2014). For this
reason, I believed the injunctive norm message would need to be explicit in order to
motivate behavior change.
Method
Participants
Participants were drivers exiting the parking decks on a university campus during a three-
week observation period. The vast majority of the participants in this study appeared to be
traditionally aged college students. A total of 8869 observations were made, but it is
impossible to ascertain how many different people were observed.
Procedure
The intervention involved posting signs near the exits of two of the three main parking
decks on campus. Only two of the three decks had a stop sign at the exit. The decks with
stop signs were chosen to be the experimental locations. The deck without a stop sign was
chosen to be the control location. Recent research suggests that injunctive messages are
more successful when people have the cognitive resources to process them (Kredentser,
Fabrigar, Smith, & Fulton, 2012). Thus, I reasoned that drivers who stop at a stop sign
would be more likely to notice and process the message than those who were not required
to stop.
Each sign was directly across the street from the stop sign. The signs were quite large
(i.e., 3 by 10 feet) and placed in a location where signs are normally prohibited, so they
were highly salient. Each sign had a black background with yellow lettering and a
photograph of a person texting while driving with a “no” symbol (i.e., a red circle with a
diagonal line through it) superimposed over the person’s hand. The sign with the weak
focus on the injunctive norm read, “Please don’t text and drive.” The sign with the strong
focus on the injunctive norm read, “97% of Dukes disapprove when you text and drive.”
Dukes refers to the faculty, staff, and students that make up the university community.
This statistic came from a university-wide survey I administered prior to the intervention.
I deliberately chose to associate the message with the in-group rather than the general
public because I wanted the sign to be as impactful as possible. Research has shown that
normative information is most effective when it comes from an in-group that individuals
strongly identify with (Terry & Hogg, 1996). Data collected by my university suggest that
most of our students strongly identify with the university community (Schreiner, 2014).
Even though this intervention was designed to reduce all phone-related distracted
driving (talking and texting), the signs focused on texting while driving. I chose to
highlight the norm against texting while driving because it is more socially disapproved
than talking on a cell phone while driving. To determine the effects of the signs, I chose to
measure all types of phone-related distracted driving. I did this for three reasons. First,
measuring only texting while driving might produce a floor effect. The rates of texting
while driving at any given moment are only 1– 2% in the general population (Pickrell,
2014). Second, past research has shown that activating one norm affects behavior related
136
to other, similar norms (Cialdini et al., 1990). It seems reasonable that activating the norm
against texting while driving would make drivers think about people’s disapproval of all
types of phone-related distracted driving. The third reason involves a measurement issue.
It can be difficult to detect the difference between talking on the phone and texting when
observing moving traffic. A person who holds the phone in front of them could be talking
on the phone, sending a text message, or reading a text message.
Trained research assistants recorded the rate of distracted driving before, during, and
after the intervention at each of the three parking decks. The research assistants stood near
the exit of each parking deck and recorded the behavior of each driver as he or she
accelerated away from the deck. Phone-related distracted driving was operationally
defined as (a) drivers holding phones to their ears, (b) drivers visibly manipulating hand-
held devices, and (c) drivers tilting their heads down. This definition was similar to the one
used by the NHTSA’s National Occupant Protection Use Survey (NOPUS; Pickrell & Ye,
2010). The NOPUS definition also included drivers speaking with visible headsets on, but
informal observations revealed that very few drivers used a headset in this geographic
area. Talking on a hand-held phone is not illegal in this area, so drivers are not required to
talk with a headset. Because texting while driving is illegal in this area, it is likely that
drivers try to hide the behavior by keeping their phone out of others’ sight. This makes it
difficult for stationary observers to ascertain whether drivers are texting. Consequently,
I chose to measure behaviors that co-vary with texting while driving: manipulating a hand-
held device and tilting one’s head. Of course, it is possible that people who have their head
titled down are not distracted by their phone. It is also possible that people holding phones
are not being distracted by them. For example, they may be holding their phone so they can
send a text as soon as they stop at a traffic light. The use of indirect measures in this study
may lead to an overestimate of the rate of distracted driving on campus, but more direct
measures were not feasible.
Data were collected over a three-week period. Each week, observations were made
daily between 3:30 pm and 4:30 pm, and on Thursdays between 2:00 pm and 3:00 pm.
I chose these hours because there was a heavy outflow of traffic during these times. There
were no signs posted during Weeks 1 and 3. During Week 2, the signs were in place from
8:00 am on Monday until 4:30pm on Friday. There were two research assistants present for
every observation hour at each location. One of the observers was designated as the
primary observer; the data collected from this individual were used in the analysis. The job
of the second observer was to collect data so that I could calculate inter-rater reliability.
The average inter-rater reliability across the three weeks of data collection was 97%.
Results
I performed a series of Pearson’s Chi-squared tests to determine whether the percentage of
distracted driving changed across each week in each of the three experimental conditions.1
There was no change in the rate of distracted driving in the control condition, x 2 (2,
N ¼ 2202) ¼ 1.46, p ¼ .48, f ¼ .03 or weak focus condition, x 2 (2, N ¼ 2448) ¼ .55,
p ¼ .76, f ¼ .02. However, there was a significant change in the strong focus condition, x 2
(2, N ¼ 4219) ¼ 10.66, p ¼ .005, f ¼ .05. Cell comparisons within the strong focus
condition revealed that the percentage of distracted driving dropped from Week 1 to Week
2, x 2 (1, N ¼ 2853) ¼ 9.36, p ¼ .002, f ¼ .06, and increased from Week 2 to Week 3, x 2
(1, N ¼ 2853) ¼ 9.36, p ¼ .002, f ¼ .06. There was no significant difference between
Weeks 1 and 3, x 2 (1, N ¼ 3112) ¼ .11, p ¼ .74, f ¼ .01. Figure 1 shows the percentage of
distracted driving across each week in each of the experimental conditions.
137
114 N.K. Lawrence
14
13
12
11 Week 1
10 Week 2
Week 3
9
6
Control Weak Focus Strong Focus
Figure 1. Percentage of phone-related distracted driving across each week in each of the
experimental conditions.
Discussion
The purpose of this study was to determine whether activating an injunctive norm related
to distracted driving would lead to a decrease in that behavior. The results showed that the
rate of distracted driving went down when drivers were exposed to a sign that clearly
expressed social disapproval of the behavior, but not when they saw a sign that merely
implied social disapproval. The findings suggest that injunctive norm messages can be
effective in reducing distracted driving, but only when there is a strong focus on the
injunctive norm. This study contributes to the literature in two important ways. First, it
provides evidence that injunctive normative messages can reduce risky driving behaviors,
at least in the short term. Second, it suggests that all injunctive normative messages are not
created equal. Messages that explicitly state the fact that people disapprove of the behavior
may be more effective than messages that merely imply it. These findings are consistent
with the focus theory of normative conduct (Cialdini et al., 1990), which states that norms
are only likely to influence behavior when they are focal in attention. Messages such as
“Please don’t drive distracted” or “Please don’t texting while driving” may not make
people think about others’ disapproval of their behavior. As a result, these types of
messages may not be maximally effective. Public service campaigns and other
interventions may be more successful if they make people cognizant of the fact that
distracted driving is socially disapproved.
For better and for worse, this study was conducted in a natural environment. It was
advantageous to record people’s actual behavior instead of measuring their self-reported
behavior or behavioral intentions, especially because the intervention targeted illegal
behavior. In fact, this is the first known study to empirically assess the effectiveness of a
social nom intervention for distracted driving. One of the problems of this study was that I
was unable to randomly assign participants to conditions. It may have been better to put
flyers on individuals’ vehicles, but campus policy prevented posting flyers in the parking
decks. It also would have been difficult to keep track of who received which experimental
treatment. Instead of using traditional random assignment, I assigned locations (i.e.,
parking decks) to conditions. It is possible that there were important differences among the
drivers at the different parking decks, but this seems unlikely given the similar baselines
rates of distracted driving among the decks. Another limitation of this study is that there
138
may have been a diffusion of treatment. That is, drivers could have been exposed to more
than one experimental condition. During the intervention week, drivers could have parked
at different decks on different days (and could have seen both signs). There is a tendency
for faculty, staff, and students to park in the same parking deck every day, but I cannot rule
out the possibility that drivers saw more than one sign.
The two signs used in this study were designed to vary in their strength of normative
focus. The sign that read, “Please don’t text and drive” had a weak focus on the
injunctive norm whereas the sign that read “97% of Dukes disapprove when you text and
drive” had a strong focus on the injunctive norm. It is clear, though, that the signs differ in
more than one way. In addition to varying in their strength of normative focus, they also
differ in terms of the presence of a statistic and the presence of an in-group referent (i.e.,
Dukes). Some studies indicate that messages containing statistical evidence are more
persuasive than messages (e.g., Kazoleas, 1993), but it seems unlikely that the mere
presence of a statistic is responsible for the decrease in distracted driving observed in this
study.
It is possible that the in-group referent in the strong-focus message increased the
persuasiveness of that message. Studies on normative social influence and message-based
persuasion suggest that in-groups wield more influence than out-groups, particularly
among people who strongly identify with the in-group (e.g., Fleming & Petty, 2000;
Mackie, Worth, & Asuncion, 1990; Terry & Hogg, 1996). I reasoned that the chosen
message would be more impactful than the one that did not include a specific referent (e.g.,
“97% of people disapprove when you text and drive”), and I wanted to create the most
impactful message possible. This study was not designed to determine whether the
presence of a referent (or the type of referent) makes a difference; it was designed to
examine whether an explicitly stated injunctive norm message would motivate drivers to
put down their phones. The results of this study suggest that it did. Future research should
examine whether in-group referents are an essential component of the message. Another
possibility for future research is to explore the relative impact of in-group versus out-group
referents.
It is important for applied researchers to develop interventions that are feasible for
others to implement. This intervention is certainly one that can be used on college
campuses and in other locations. Posting signs is a relatively easy and inexpensive strategy
that may have a large impact. Future researchers should investigate additional factors that
maximize the effectiveness of the signs. It would also be important to know how to prevent
drivers from becoming desensitized to the signs and their messages. One possibility for
college campuses is to have a variety of injunctive normative messages presented in a
variety of mediums (signs, posters, flyers, bumper stickers, pens, etc.). Injunctive
normative messages are already a part of national public service campaigns. Many of these
campaigns ask drivers to and “Stop the texts, stop the wrecks” (www.stoptextsstopwrecks.
org). These campaigns would likely be more successful if they drew more attention to the
fact that risky driving behaviors are socially disapproved. Consider a prototypical public
service announcement (PSA): a young driver is engaging in a text conversation and
crashes her car, killing the passengers. The PSA ends with voice-over urging viewers to
“Please don’t text and drive.” This PSA could more effective if one of the passengers
expressed disapproval and said, “What are you doing? Put down your phone and drive!”
Another strategy to induce focus on the injunctive norm is to feature testimonials of people
advocating against distracted driving. Importantly, public officials and researchers need to
empirically assess the success of these interventions so that institutions and agencies can
spend their dollars wisely.
139
116 N.K. Lawrence
As noted earlier, the effects of this injunctive norm intervention were short-lived.
The rate of distracted driving in the strong focus condition returned to baseline levels once
the intervention was over. Because this happened, one can have greater confidence that
the intervention—and not some extraneous variable—was responsible for the change.
Of course, it would have been nice for the intervention to promote long-lasting behavior
change, but this was not expected. Cell phone use has become habitual for many people
(Oulasvirta, Rattenbury, Ma, & Raita, 2012). Even texting while driving can occur
automatically, without awareness, attention, control, or intention (Bayer & Campbell,
2012). Getting people to ignore their phones in the car may be as hard as helping people
overcome an addiction such as smoking. For this reason, a solution to the problem will likely
require a combination of legal, technological, psychological, and social components. Still, it
is possible that repeated exposure to injunctive normative information will help make
distracted driving as socially disapproved as smoking is. Although the intervention used in
this study had only short-term effects, it is a significant first step toward creating widespread
campaigns that harness the power of normative social influence.
Acknowledgements
I would like to thank the following undergraduate research assistants for their help with data
collection: Timothy Adams, Rachel Browne, Sarah Janser, Hunter Kelly, Kathryn Vila, and Laura
Wiechecki. I am also grateful to Laura Williams for her advice regarding data analysis.
Disclosure statement
No potential conflict of interest was reported by the author.
Funding
James Madison University’s College of Health and Behavioral Studies provided grant support for
this research.
Note
1. Although the observations in this study are not independent, the Pearson’s Chi-square test is the
most appropriate test to use in this type of study. Burger and Shelton (2011) had a similar design
and used the same analysis.
References
Bator, R. J., Tabanico, J. J., Walton, M. L., & Schultz, P. W. (2014). Promoting energy conservation
with implied norms and explicit messages. Social Influence, 9, 69 – 82. doi:10.1080/15534510.
2013.778213
Bayer, J. B., & Campbell, S. W. (2012). Texting while driving on automatic: Considering the
frequency-independent side of habit. Computers in Human Behavior, 28, 2083– 2090. doi:10.
1016/j.chb.2012.06.012
Burger, J. M., & Shelton, M. (2011). Changing everyday health behaviors through descriptive norm
manipulations. Social Influence, 6, 69 – 77. doi:10.1080/15534510.2010.542305
Center for Disease Control and Prevention. (2013). Mobile device use while driving: United States
and seven European countries. Retrieved from http://www.cdc.gov/mmwr/preview/mmwrhtml/
mm6210a1.htm
Cialdini, R. B. (2003). Crafting normative messages to protect the environment. Current Directions
in Psychological Science, 12, 105– 109. doi:10.1111/1467-8721.01242
140
Cialdini, R. B., Demaine, L. J., Sagarin, B. J., Barrett, D. W., Rhoads, K., & Winter, P. L. (2006).
Managing social norms for persuasive impact. Social Influence, 1, 3 – 15. doi:10.1080/
15534510500181459
Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative conduct:
Recycling the concept of norms to reduce littering in public places. Journal of Personality and
Social Psychology, 58, 1015– 1026. doi:10.1037/0022-3514.58.6.1015
Clapp, J. D., Lange, J. E., Russel, C., Shillington, A., & Voas, R. (2003). A failed norms social
marketing campaign. Journal of Studies on Alcohol, 64, 409– 414.
Dingus, T. A., Klauer, S. G., Neale, V. L., Petersen, A., Lee, S. E., Sudweeks, J., . . . Knipling, R. R
(2006). The 100-car naturalistic driving study, Phase II: Results of the 100-car field experiment .
(Report No. DOT HS 810 593). Washington, DC: National Highway Traffic Safety
Administration.
Fleming, M. A., & Petty, R. E. (2000). Identity and persuasion: An elaboration likelihood approach.
In D. J. Terry & M. A. Hogg (Eds.), Attitudes, behaviors, and social context: The role of norms
and group membership (pp. 171– 199). Mahwah, NJ: Erlbaum.
Goldstein, N. J., Cialdini, R. B., & Griskevicius, V. (2008). A room with a viewpoint: Using social
norms to motivate environmental conservation in hotels. Journal of Consumer Research, 35,
472– 482. doi:10.1086/586910
Kallgren, C. A., Reno, R. R., & Cialdini, R. B. (2000). A focus theory of normative conduct: When
norms do and do not affect behavior. Personality and Social Psychology Bulletin, 26,
1002– 1012. doi:10.1177/01461672002610009
Kazoleas, D. C. (1993). A comparison of the persuasive effectiveness of qualitative versus
quantitative evidence: A test of explanatory hypotheses. Communication Quarterly, 41, 40 – 50.
10.1080/01463379309369866
Kredentser, M. S., Fabrigar, L. R., Smith, S. M., & Fulton, K. (2012). Following what people think
we should do versus what people actually do: Elaboration as a moderator of the impact of
descriptive and injunctive norms. Social Psychology and Personality Science, 3, 341–347.
doi:10.1177/1948550611420481
Lawrence, N. K. (2013). [University survey]. Unpublished raw data.
Mackie, D. M., Worth, L. T., & Asuncion, A. G. (1990). Processing of persuasive in-group messages.
Journal of Personality and Social Psychology, 58, 812– 822. doi:10.1037/0022-3514.58.5.812
National Highway Traffic Safety Administration. (2014). Driver electronic device use in 2012.
(Report No. 811884). Retrieved from http://www-nrd.nhtsa.dot.gov/Pubs/811884.pdf
National Safety Council. (2014). NSC releases latest injury and fatality statistics and trends [Press
release]. Retrieved from http://www.nsc.org/NewsDocuments/2014-Press-Release-Archive/3-
25-2014-Injury-Facts-release.pdf
Neighbors, C., Larimer, M., & Lewis, M. (2004). Targeting misperceptions of descriptive drinking
norms: Efficacy of a computer-delivered personalized normative feedback intervention. Journal
of Consulting and Clinical Psychology, 72, 434– 447. doi:10.1037/0022-006X.72.3.434
Nolan, J. M., Schultz, P. W., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V. (2008). Normative
social influence is underdetected. Personality and Social Psychology Bulletin, 34, 913–923.
doi:10.1177/0146167208316691
Oulasvirta, A., Rattenbury, T., Ma, L., & Raita, E (2012). Habits make smartphone use more
pervasive. Personal Ubiquitous Computing, 16, 105– 114. doi:10.1007/s00779-011-0412-2
Pickrell, T. M. (2014). Driver electronic device use in 2012 (Traffic Safety Facts Research Note:
Report No. DOT HS 811 884). Washington, DC: National Highway Traffic Safety
Administration. Retrieved from http://www-nrd.nhtsa.dot.gov/Pubs/811884.pdf
Pickrell, T. M., & Ye, T. J. (2010). Driver electronic device use observation protocol (Traffic Safety
Facts Research Note: Report No. DOT HS 811 361). Washington, DC: National Highway Traffic
Safety Administration. Retrieved from http://www-nrd.nhtsa.dot.gov/Pubs/811361.pdf
Schreiner, L. A. (2014). [University survey]. Unpublished raw data.
Schroeder, P., Meyers, M., & Kostyniuk, L. (2013). National survey on distracted driving attitudes
and behaviors – 2012. (Report No. DOT HS 811729). Washington, DC: National Highway
Traffic Safety Administration.
Schultz, P. W., Khazian, A. M., & Zaleski, A. C. (2008). Using normative social influence to
promote conservation among hotel guests. Social Influence, 3, 4 – 23. doi:10.1080/
15534510701755614
141
118 N.K. Lawrence
Schultz, P. W., Nolan, J. M., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V (2007). The
constructive, destructive, and reconstructive power of social norms. Psychological Science, 18,
429– 434. doi:10.1111/j.1467-9280.2007.01917.x
Terry, D. J., & Hogg, M. A. (1996). Group norms and the attitude-behavior relationship: A role for
group identification. Personality and Social Psychology Bulletin, 22, 776– 793. doi:10.1177/
0146167296228002
142
Week 4
143
144
5
Handhaving van de bijstandsregels

Uitvoerders zoeken ruimte voor eigen keuzes
Paulien de Winter
Paulien de Winter is rechtssocioloog en werkt als universitair docent aan de faculteit

Rechtsgeleerdheid van de Rijksuniversiteit Groningen
‘Iedereen gelijk behandelen dat zullen we moeten doen, omdat er regels zijn. (…)
en dat ik iemand dan zielig of leuk of aardig vind, dat doet er niet toe. Ik kan er
niets mee.’
‘Bijstand begint bij de mens, niet bij de regel. (…) Je kunt dit werk niet vanuit pa-
pier doen. Misschien gaat het dan feilloos en dan klopt het wetstechnisch, maar
het morele en het menselijke, ik zie niet hoe je daar dan inhoud en achting aan
kan geven.’
Dit zijn citaten van twee medewerkers bij Nederlandse gemeenten die da-
gelijks contact hebben met uitkeringsgerechtigden. Ze zijn verantwoorde-
lijk voor de handhaving van de verplichtingen van de Participatiewet. De
uitspraken laten zien dat de medewerkers verschillende opvattingen heb-
ben over wat belangrijk is in hun werk. De ene is voor uniformiteit, ter-
wijl de andere juist een voorstander van maatwerk lijkt te zijn. Deze ver-
schillende opvattingen kunnen leiden tot een verschil in uitvoering van
de Participatiewet. Het kan voor de uitkeringsgerechtigden mogelijk dus
uitmaken welke medewerker zij treffen. In dit hoofdstuk kijk ik naar dit
vraagstuk vanuit het perspectief van de uitvoerders zelf. De vraag die hier
centraal staat, is: Hoe denken uitvoerende medewerkers over handhaving
van de verplichtingen van de Participatiewet? Op basis van empirisch on-
derzoek bij drie gemeenten bespreek ik de wijze waarop medewerkers de
regels toepassen en de wijze waarop medewerkers reageren op overtredin-
gen (zie ook De Winter 2019). Deze bijdrage geeft daarmee een kijk in de
115
145
PAULIEN DE WINTER
wereld van de handhaving op uitvoeringsniveau. Daarvoor zal ik eerst een

beeld schetsen van de verplichtingen van de Participatiewet zoals die op
papier staan. Maar vooraf een korte toelichting op mijn onderzoek.
Het onderzoek: de praktijk in drie gemeenten

Handhaving betreft zowel preventieve als repressieve activiteiten van me-
dewerkers van gemeentelijke uitvoeringsorganisaties ter bevordering van
naleving van de Participatiewet door uitkeringsgerechtigden. Het onder-
zoek is gericht op uitvoerende medewerkers met handhavingstaken. Om
de handhaving van de verplichtingen van de Participatiewet in de prak-
tijk te bestuderen, zijn drie gemeenten onderzocht. Bij elke gemeente zijn
beleidsstukken geanalyseerd, participerende observaties uitgevoerd en
verdiepende interviews met medewerkers gehouden. De analyse van de
beleidsstukken geeft inzicht in het handhavingsbeleid en de handhavings-
keuzes van gemeenten. Participerende observaties geven inzicht in het ge-
drag van medewerkers. De interviews geven een beeld van de houding van
uitvoerders, de keuzes die zij maken en hun gedachten over handhaving.
Bij de keuze voor de gemeenten is gekeken naar locatie, aantal inwo-
ners, aantal uitkeringsgerechtigden en handhavingsreputatie (dit laatste
gebaseerd op gesprekken met mensen uit de praktijk en met andere onder-
zoekers). Gemeente I is een relatief kleine gemeente (minder dan 50.000
inwoners, 50 personen met een bijstandsuitkering per 1000 inwoners van
18 jaar en ouder, 7,7 procent werkloosheid) in het noorden van Neder-
land, met een zachte tot gemiddelde handhavingsreputatie. Gemeente II
is een grote gemeente (100.000 tot 500.000 inwoners, 74 personen met een
bijstandsuitkering per 1000 inwoners van 18 jaar en ouder, 8,6 procent
werkloosheid) in het oosten van Nederland, met een harde handhavings-
reputatie. Gemeente III is een grote gemeente (100.000 tot 500.000 inwo-
ners, 39 personen met een bijstandsuitkering per 1000 inwoners van 18
jaar en ouder, 4,6 procent werkloosheid) in het midden van Nederland,
met een gemiddelde handhavingsreputatie. In totaal zijn er 41 medewer-
kers geïnterviewd: 13 medewerkers reïntegratie, 8 medewerkers inkomen,
11 medewerkers handhaving en 9 managers.
Ik richt mij op twee aspecten van handhaving: de toepassing van regels
en de reactie op overtredingen. De wijze van regeltoepassing kan in meer
of mindere mate flexibel of strikt zijn. Als uitvoerders de regels flexibel
116
146
HANDHAVING VAN DE BIJSTANDSREGEL S
toepassen, dan staan ze vaker uitzonderingen toe, hebben ze een eigen de-
finitie van een overtreding, kiezen ze vaker voor de uitkeringsgerechtigde
(en niet voor de regel) en kiezen ze voor maatwerk (en niet voor iedereen
gelijk behandelen). Als uitvoerders de regels strikt toepassen, dan staan ze
uitzonderingen slechts zeer beperkt toe, interpreteren ze de wet letterlijk
als het gaat om overtredingen, kiezen ze voor regels (en niet voor uitke-
ringsgerechtigden) en kiezen ze voor uniformiteit (en niet voor maatwerk).
De reactie op een overtreding kan in meer of mindere mate overredend
of bestraffend zijn. Uitvoerders die overredend reageren op overtredingen,
geven voornamelijk voorlichting en advies aan uitkeringsgerechtigden,
geven vrijstelling van de sollicitatieplicht en geven waarschuwingen en
dreigen af en toe met sancties. Uitvoerders die bestraffend reageren op
overtredingen, sturen vaak waarschuwingsbrieven, leggen maatregelen of
boetes op en schorsen of beëindigen uitkeringen. Ik heb uitvoerders op
beide aspecten ingedeeld, gebaseerd op hun gedrag (participerende obser-
vaties) en hun uitspraken (interviews).1
Handhaving op papier: de wet en de verordeningen

Voordat ik de praktijk van handhaving beschrijf, ga ik eerst in op de regels
zoals die op papier staan en waartoe de uitvoerders zich moeten verhou-
den. In de Participatiewet zijn de standaarden voor de bestuurlijke boete
voor een overtreding van de inlichtingenplicht vastgelegd – alle gemeen-
ten moeten die toepassen. De maatregelen voor een overtreding van de
arbeidsverplichtingen verschillen per onderzochte gemeente.
De inlichtingenplicht is de verplichting tot het (uit eigen beweging) ver-
strekken van inlichtingen. Inlichtingen zijn feiten die van invloed kunnen
zijn op het recht op of de hoogte van de uitkering (Participatiewet art.
17). Bij een overtreding van de inlichtingenplicht móét de gemeente een
boete opleggen. De berekening van deze (bestuurlijke) boete is opgeno-
men in het Boetebesluit socialezekerheidswetten. Als er sprake is van te
veel ontvangen uitkeringsgeld, dan gelden de volgende standaarden: bij
‘opzet’ is de boete 100 procent van het te veel ontvangen uitkeringsgeld,
bij ‘grove schuld’ is de boete 75 procent, bij ‘normale verwijtbaarheid’ is de
boete 50 procent en bij ‘verminderde verwijtbaarheid’ bedraagt de boete
1 Het gaat om 29 uitvoerders; van 3 uitvoerders was er te weinig informatie om de handha-
vingsstijl vast te stellen.
117
147
PAULIEN DE WINTER
25 procent. Bij recidive geldt een boete van 150 procent. Deze standaarden
gelden voor alle gemeenten.
De arbeidsverplichtingen zijn verplichtingen om zo snel mogelijk betaal-
de arbeid te vinden, bijvoorbeeld de sollicitatieplicht. Deze verplichtingen
zijn verdeeld in geüniformeerde en niet-geüniformeerde verplichtingen.
Geüniformeerde verplichtingen zijn verplichtingen waarbij de wetgever
de maatregel vastgelegd heeft. Bij de niet-geüniformeerde verplichtingen
laat de wetgever de gemeenten vrij om in hun verordening zelf vast te stel-
len wat de hoogte en de duur van de maatregel is bij niet-naleving.
Gemeenten moeten dus een maatregel opleggen bij overtredingen van
de geüniformeerde verplichtingen. Het gaat om een verlaging van de bij-
stand met 100 procent. Gemeenten kunnen in hun verordening de duur
van de maatregel en de eventuele spreiding (de ‘verrekening’) van de
maatregel over drie maanden (bijvoorbeeld in plaats van één maand een
verlaging van 100 procent, drie maanden een verlaging van 33 procent)
vaststellen. De drie onderzochte gemeenten hebben gekozen voor een of
twee maanden. Een van de onderzochte gemeenten staat verrekening toe
bij alle overtredingen, de andere twee gemeenten alleen in ‘(zeer) bijzon-
dere omstandigheden’.
Bij een herhaalde overtreding (recidive) van de geüniformeerde ver-
plichtingen hebben de drie onderzochte gemeenten een maatregel van 100
procent van de bijstandsnorm gedurende twee maanden vastgesteld in de
verordeningen. Gemeenten hanteren vervolgens verschillende definities
van recidive. Bij twee van de drie onderzochte gemeenten wordt een on-
derscheid gemaakt in een eerste recidive (twee maanden) en herhaalde
recidive (drie maanden). Bij één gemeente maakt het niet uit welke ver-
plichting er wordt overtreden. Verder nemen twee van de drie gemeenten
in de verordening de ruimte om de maatregel aan te passen of op 0 procent
vast te stellen op grond van bijzondere omstandigheden.
Naast de wettelijk gespecificeerde verplichtingen die de Participatiewet

voorschrijft, kan een gemeente niet-geüniformeerde verplichtingen vast-
leggen in een verordening. Dit gaat om verplichtingen als het tijdig regis-
treren als werkzoekende, het opstellen, uitvoeren en evalueren van een
plan van aanpak of het voldoen aan een oproep om in verband met de ar-
beidsinschakeling op een gegeven plaats en tijd te verschijnen. De gemeen-
118
148
ten kiezen voor uiteenlopende maatregelenregimes: twee categorieën (10

procent en 50 procent verlaging van de bijstandsnorm), drie categorieën (5
procent, 10 procent en 20 procent verlaging van de bijstandsnorm) of vijf
categorieën (10 procent, 20 procent, 40 procent, 50 procent en 100 procent
verlaging van de bijstandsnorm).
Kortom, gemeenten verschillen in de hoogte en de duur van de maatre-
gelen, en gaan verschillend om met recidive en de spreiding van de maat-
regel over meerdere maanden.
Handhaving in de praktijk: de toepassing van de regels

Tot zover de handhaving op papier. Hoe ziet handhaven van de Participa-
tiewet er in de praktijk uit? Ik ga allereerst in op de toepassing van de re-
gels. De meeste onderzochte uitvoerders (59 procent) passen de regels min
of meer flexibel toe. Een kleiner gedeelte (28 procent) past de regels min
of meer strikt toe. Eén medewerker (3 procent) zit precies op de scheids-
lijn tussen flexibel en strikt (‘gemiddeld’ genoemd). Zie tabel 1 voor een
overzicht. Medewerkers van dezelfde gemeente verschillen onderling in
de wijze van regeltoepassing. Bijna alle onderzochte uitvoerders van ge-
meente I passen de regels flexibel toe. Bij gemeente II past het grootste
gedeelte van de onderzochte medewerkers de regels strikt toe. De overige
onderzochte medewerkers passen de regels flexibel of ‘gemiddeld’ toe. Bij
gemeente III passen de onderzochte uitvoerders de regels zowel flexibel als
strikt toe. Deze indeling van uitvoerders naar wijze van regeltoepassing
komt overeen met de handhavingsreputatie van de gemeenten (zie de eer-
dere passage over ‘Het onderzoek’).
Tabel 1 Indeling uitvoerders naar wijze van regeltoepassing

flexibel gemiddeld strikt totaal
totaal 17 (59%) 1 (3%) 11 (38%) 29
gemeente I 9 (82%) 0 (0%) 2 (18%) 11
gemeente II 3 (33%) 1 (11%) 5 (56%) 9
gemeente III 5 (56%) 0 (0%) 4 (44%) 9
Er blijkt dus veel variatie te zitten in de wijze waarop medewerkers de

regels toepassen, zowel binnen gemeenten als tussen gemeenten. Om de
verschillen in houding van de uitvoerders te illustreren, volgen enige cita-
ten over hoe zij denken over regels en gelijke behandeling.
119
149
PAULIEN DE WINTER
‘Als je de wet gaat bestuderen, dan zie je dat binnen elke wet een kader is [respon-
dent bedoelt: de wet geeft een kader waarbinnen ruimte is; PW]. En die vrijheid
heb je. Je ziet dat mensen [medewerkers] angstvallig gewoon die vrijheid niet dur-
ven te pakken, wat ingesloten zit binnen een wet. Als ik de mogelijkheid krijg, als
ik vind dat ik gebruik moet maken van de kaders, dan doe ik dat. Alles ter wille
van mijn klant.’ (respondent 4)
‘Soms moet je over een regel heen durven stappen.’ (respondent 6)
‘Als ik daar met een juridisch oogpunt naar kijk, dan zeg ik, is iedereen gelijk.
Want wat je niet wilt, is rechtsongelijkheid. Maar ik ben me er ook terdege van
bewust, dat maatwerk steeds meer moet. Puur juridisch gezien zeg ik, is iedereen
voor de wet gelijk. (…) Want wat je niet wilt, is dat je op een gegeven moment voor
de ene klant maatwerk levert terwijl de buurman die in dezelfde situatie zit, je het
maatwerk niet levert.’ (respondent 42)
‘Iedereen gelijk behandelen dat zullen we moeten doen, omdat er regels zijn. (…)
en dat ik iemand dan zielig of leuk of aardig vind, dat doet er niet toe. Ik kan er
niets mee.’ (respondent 43)
Respondenten 4 en 6 zijn ingedeeld als uitvoerders die regels flexibel

toepassen en respondenten 42 en 43 als uitvoerders die de regels strikt
toepassen. Dat uitvoerders verschillend denken over de toepassing van
regels, kan samenhangen met dat ze verschillende definities hanteren van
een (verwijtbare) overtreding. Het is mogelijk dat de ene medewerker in
een bepaalde situatie van mening is dat er een overtreding is begaan, ter-
wijl een andere medewerker anders denkt over dezelfde situatie.
Handhaving in de praktijk: reactie op overtreding

Bijna de helft van de onderzochte medewerkers (45 procent) reageert in
meer of mindere mate bestraffend op overtredingen. Een kleiner gedeelte
(34 procent) reageert ‘gemiddeld’ op overtredingen (de scheidslijn tussen
bestraffend en overredend). Een iets kleiner gedeelte (32 procent) reageert
in meer of mindere mate overredend op overtredingen. Zie tabel 2 voor
een overzicht. Medewerkers van dezelfde gemeente verschillen in hun
reactie op overtredingen. De meeste onderzochte medewerkers van ge-
120
150
meente I reageren ‘gemiddeld’ op overtredingen. De overige onderzochte

medewerkers reageren overredend of bestraffend. Bij gemeente II reageren
de meeste onderzochte medewerkers bestraffend op overtredingen. Bij ge-
meente III reageren de onderzochte medewerkers zowel bestraffend als
overredend op overtredingen, met de nadruk op bestraffend. Deze inde-
ling van uitvoerders naar wijze van regeltoepassing komt overeen met de
handhavingsreputatie van de gemeenten (zie de eerdere passage over ‘Het
onderzoek’).
Tabel 2 Indeling uitvoerders naar reactie op overtreding

overredend gemiddeld bestraffend totaal
totaal 6 (21%) 10 (34%) 13 (45%) 29
gemeente I 3 (27%) 6 (55%) 2 (18%) 11
gemeente II 0 (0%) 3 (33%) 6 (67%) 9
gemeente III 3 (33%) 1 (11%) 5 (56%) 9
Ik laat aan de hand van een aantal citaten van medewerkers de verschil-
len zien. Een voorbeeld van een medewerker die overredend reageert op
overtredingen:
‘Als het niet kan, als het niet hoeft, dan leg ik mensen geen boete op. Mensen heb-
ben al genoeg zorgen aan hun hoofd en dan moet ik nog een boete gaan opleggen.
Nou sorry, ik zit er niet op te wachten. Als ik eronderuit kan komen dan doe ik
dat ook wel. Dat probeer ik zoveel mogelijk, maar soms moet het gewoon. Dan ga
ik met de klant in gesprek. Geregeld is het mij al wel gelukt dat klanten van ons
zeggen “ik snap het wel, je doet ook je werk en je kunt ook niet anders, ja ik ben
dom geweest, blablabla”.’ (respondent 4)
Deze uitvoerder geeft aan dat hij liever geen sancties oplegt en dat hij in
plaats van een boete op te leggen een goed gesprek voert met de uitke-
ringsgerechtigde. Hij probeert het bestraffen van uitkeringsgerechtigden
zoveel mogelijk te voorkomen en kiest voor overreden.
Een voorbeeld van een uitvoerder die bestraffend reageert op overtre-
dingen, is respondent 65. Hij vindt dat zijn collega’s te weinig maatregelen
opleggen:
121
151
PAULIEN DE WINTER
‘Ik heb de cijfers gezien van vorig jaar en daar ben ik wel van geschrokken. (…) Dat
vind ik toch wel heel weinig. (…) Ik denk dat ze het meestal wel opmerken, maar er
zijn twee redenen waarom ze het niet doen. En één is dat het veel te veel tijd kost.
Het is wel een heel rapport hè, je moet toch wel even een stukje geschiedenis erin
zetten en dat kost veel tijd. En de andere kant, ze hebben een band met die klant en
ze zijn ook wel bang dat als ze een maatregel gaan opleggen dat de klant daar niet
blij mee gaat zijn en dan verstoor je die relatie en krijg je hem helemáál niet meer
aan het werk.’ (respondent 65)
Hij geeft verder aan dat hij geen moeite heeft met zijn controlerende rol:
‘Die vind ik heel belangrijk. Het is wel gemeenschapsgeld. Het is ook mijn belas-
tinggeld. En uiteindelijk help je de klant er ook niet mee als die onterecht een uit-
kering krijgt. Want ja, als het aan het licht komt, hij moet alles terugbetalen. Een
boete eroverheen. Ze komen alleen maar verder in de problemen. Dus dan heb ik
het liever maar gewoon duidelijk.’ (respondent 65)
Ook twee andere uitvoerders geven aan hun controlerende rol belangrijk
te vinden:
‘Ik ben wel heel fraude-alert. Dat is m’n werk. Ik vind ook dat dat moet.’ (respondent 6)
‘Ze ontvangen natuurlijk wel een uitkering en zij moeten er wel alles aan doen
om zo snel mogelijk weer uit de uitkering te geraken. Het is natuurlijk het laatste
sociale vangnet van alles in Nederland. Dus ik heb er geen problemen mee om ook
controlerend te zijn naar klanten.’ (respondent 7)
Tussen de regels ruimte vinden voor de beste werkwijze

Medewerkers reageren in de praktijk dus verschillend op overtredingen
en passen de regels op verschillende wijze toe. Voor uitkeringsgerechtig-
den houdt dit in dat het kan uitmaken welke medewerker zij treffen. Me-
dewerkers zijn zich bewust van de verschillen in handhaving:
‘Iedereen heeft zijn eigen werkwijze. Ik houd het bij mijn eigen werkwijze. Ik heb
mijn eigen manier van werken.’ (respondent 4)
122
152
‘We hebben beleid en we praten over uniform werken, maar iedereen neemt na-
tuurlijk z’n eigen pakketje met vaardigheden et cetera mee. (…) Iedereen is hier
ook heel eigenwijs, dus iedereen blijft het ook zeker op z’n eigen manier doen. (…)
Maar ik denk dat je, dat uiteindelijk elke klant of werkzoekende die hier binnen
komt lopen, die loopt wél met hetzelfde pakketje de deur uit.’ (respondent 11)
‘Ik ben daarin best een stukje eigenwijs, want ik denk dat mijn werkwijze goed is.
(…) Maar ze zijn altijd nog heel positief over mijn resultaten geweest, dus ik houd
het gewoon lekker bij mijn werkwijze.’ (respondent 54)
Medewerkers vinden tussen de regels manieren om te handhaven zoals

zij dat passend vinden. Dit is niet alleen het geval bij de drie onderzoch-
te gemeenten; dit is een bredere trend. Uit een rapport van de Inspectie
SZW (2017, p. 10) blijkt dat medewerkers van de sociale dienst verschillen-
de keuzes maken in de handhaving van de arbeidsverplichtingen. In dat
onderzoek blijkt dat bij een aantal onderzochte gemeenten medewerkers
terughoudend zijn met het opleggen van maatregelen omdat dit volgens
de gemeenten de relatie met de cliënt verstoort en contraproductief werkt
(p. 35). Medewerkers vinden sancties van 100 procent van de bijstand
voor een groot deel van de arbeidsverplichtingen te zwaar (p. 3, p. 12).
De uitvoerders vinden dat zij ‘ongeacht de wet, moeten bepalen wat de
beste interventie is bij nalatig gedrag’ (p. 27). De Inspectie constateert dat
gemeenten variëren in het opleggen van maatregelen terwijl de wet juist
uniformiteit nastreeft (p. 3). De Inspectie is van oordeel ‘dat er verschillen
bestaan bij uitvoering van de arbeidsverplichtingen in de Participatiewet,
waardoor de met de wet nagestreefde rechtsgelijkheid bij het handhaven
van die arbeidsverplichtingen in het gedrang komt (…).’ Ik ga dieper in op
een van de mogelijke redenen waarom medewerkers tussen de regels door
op zoek gaan naar een (in hun ogen) passende handhavingsaanpak.
De rol van leidinggevenden

De mate waarin managers ruimte en vrijheid aan medewerkers geven en
de mate waarin medewerkers ruimte nemen, verschilt. Sommige mana-
gers geven aan dat er wel ruimte is, maar dat niet alle medewerkers deze
ruimte (willen) zien. Er zijn ook managers die medewerkers juist stimu-
leren om tot creatieve oplossingen te komen en hen soms zelfs letterlijk
123
153
PAULIEN DE WINTER
aansporen om buiten de lijntjes te kleuren als dit in het voordeel is van de

klant. Een manager die aanstuurt op creativiteit en een medewerker die
aangeeft te mogen afwijken van werkprocessen:
‘Binnen de kaders hebben ze dus heel veel ruimte, en dat moet ook want ik heb ook
slimme mensen aangenomen. En die behandel ik ook als slimme mensen, en niet
als kleuters. (…) Die ruimte is er gewoon. Zelf nadenken, creatief zijn en dan is er
niemand die zegt: ‘’Het mag niet’’.’ (manager, respondent 13)
‘Ik mag zelfs beslissingen nemen die, zeg maar, buiten de geëigende paden liggen,
die buiten de afbakeningen liggen.’ (respondent 8)
Uit de interviews blijkt dat ook managers het belangrijk vinden dat uit-
voerders rekening houden met de uitkeringsgerechtigden:
‘Die discretionaire ruimte vind ik belangrijk, dat mensen [medewerkers] ook

vooral gaan nadenken. Daar heb ik ook wel een beetje op geselecteerd ook, dat
mensen net even iets verder gaan dan de bevelen, dus het domweg uitvoeren van.’
(manager, respondent 58)
‘Dat is altijd maatwerk! Ja, dat is geen standaard, er is geen standaard in mensen.
(…) De regels zijn de gelijke behandeling. De wijze waarop is het maatwerk.’ (ma-
nager, respondent 66)
Eigen opvattingen over wat de beste werkwijze is

Een bekend perspectief op ambtelijke uitvoeringsorganisaties is dat me-
dewerkers neutrale uitvoerders zijn. Vanuit dit oogpunt zijn verschillen in
toepassing van de regels en verschillen in reacties op overtredingen onwen-
selijk en onrechtvaardig voor uitkeringsgerechtigden. Uitvoerende mede-
werkers zijn echter geen mechanische uitvoerders van de regels, ze hebben
eigen opvattingen over hun werk (De Winter 2019). Het belang van op-
vattingen van medewerkers op de manier waarop zij hun werk uitvoeren,
staat in recent onderzoek steeds meer centraal. Zacka (2017, p. 87) stelt dat
de rolopvattingen de beslissingen van uitvoerders beïnvloeden. Uitvoe-
rende medewerkers moeten hun werk zo efficiënt mogelijk uitvoeren, hun
discretionaire ruimte gebruiken op een manier die eerlijk, respectvol en
124
154
zo responsief mogelijk is naar de bijzonderheden van de individuele zaak

(Zacka 2017, p. 242). Zacka beschrijft uitvoerende medewerkers als moral
agents die na verloop van tijd hun eigen stijl (professionele identiteit) ont-
wikkelen (p. 90), door hem moral dispositions genoemd.2 De verschillen in
handhaving ontstaan mogelijk doordat medewerkers anders denken over
wat juist en rechtvaardig en effectief is voor uitkeringsgerechtigden.
Rekening houden met de situatie van de uitkeringsgerechtigde

Het onderzoek laat zien dat veel uitvoerende medewerkers in hun werk-
wijze rekening houden met de situatie van de uitkeringsgerechtigde:
‘Het mooiste blijft natuurlijk het opheffen van de situatie waarin de klant verkeert.
Dat kan je door negatieve prikkels doen, maar dat kan je ook door positieve prik-
kels doen. Plus dat je de gelegenheid hebt het continu af te stemmen (…) Neem
je verantwoordelijkheid als werkcoach en als klantmanager. Jij bepaalt dat, jij zit
daar die wet uit te voeren, jij kent die klant.’ (respondent 6)
‘Niet elke situatie is dezelfde maar dat neemt niet weg dat elke partij gewoon zijn
eigen verhaal heeft en zijn eigen achtergrond. Dat is geen standaard iets wat je kunt
zeggen dat voor iedereen zou moeten gelden dus.’ (respondent 41)
Uitvoerders zien het belang in van gelijke behandeling en tegelijkertijd

vinden zij maatwerk belangrijk:
‘Gelijk behandelen, maar maatwerk daar waar nodig.’ (respondent 51)
‘Ik denk dat je in deze tijd niet zonder maatwerk kan. (…) Aan de andere kant
denk ik gelijke monniken, gelijke kappen.’ (respondent 57)
Sommige medewerkers geven nadrukkelijk aan dat gelijkheid in ongelijke

situaties om maatwerk vraagt:
‘Natuurlijk wil je alle klanten gelijk…, dat vind ik een lastige joh. Ik denk dat je
2 Morele disposities gaan over de interpretatieve lens van medewerkers, over hun normatieve
en affectieve gevoeligheden (sensibilities) en over de m anier waarop zij hun rol en verantwoor-
delijkheid interpreteren (Zacka 2017, p. 15).
125
155
PAULIEN DE WINTER
er niet aan ontkomt om maatwerk te leveren. Als je kijkt hoe verschillend klanten
zijn en wat klanten nodig hebben dan moet je, dan moet je eigenlijk wel maatwerk
leveren.’ (respondent 62)
Uitvoerders leggen bij hun afwegingen de nadruk op gelijkheid in de uit-

komst en niet op gelijkheid in de behandeling. Op basis van hun eigen
overtuiging en soms in overleg met collega’s, bepalen de medewerkers
welke afweging zij in welke situaties rechtvaardig vinden.
‘Als het goed is zijn ze streng maar rechtvaardig. Het zijn ook geen watjes, maar
we hebben hier ook wel zo’n gevleugeld woord: we zijn er wel om mensen op de
rails te zetten en niet om ze in de afgrond te duwen want dat is een klein kunstje.’
(manager, respondent 14)
‘Streng maar rechtvaardig. Nou nee, want kijk het is niet zo dat wij met de botte
bijl overal maar op inslaan of als een olifant door de porseleinkast heen rennen.
Helemaal niet, ik denk juist dat we hele goede afwegingen maken. Natuurlijk zijn
we wel streng, we zijn gewoon strikt. We weten wat de procedures zijn, we weten
hoe weten regelgeving in elkaar zit, maar we kunnen ook best buiten de kaders
denken. In die zin dat we ons ook wel inleven in een klant en in een situatie.’ (res-
pondent 57)
Medewerkers denken en handelen dus verschillend bij hun handhavings-

beslissingen. Welke overwegingen liggen hieraan ten grondslag? Veel me-
dewerkers zeggen dat ze het doel voor ogen proberen te houden. Wat ze
specifiek onder ‘het doel’ verstaan, laten ze vaak in het midden. Meestal
lijkt het over reïntegratie te gaan, dus dat uitkeringsgerechtigden weer
werk vinden en geen uitkering meer nodig hebben.
‘De uitkomst zal bij beiden hetzelfde zijn, absoluut, daar ben ik van overtuigd.
Alleen de wat meer zakelijke consulent, die laat ook wat meer bij de klant zelf
liggen, van nou, als jij iets zelf wilt, kom maar. Terwijl de meer sociale consulent
handvatten biedt, van als je dit nou even doet of dat nou even doet, als je dat dan
inlevert, hebben we de zaak rond. Dus in dat opzicht maakt het denk ik wel heel
veel verschil voor een klant uit wie jouw contactpersoon is.’ (respondent 42)
126
156
‘We zijn allemaal verschillende individuen, we hebben hetzelfde doel, maar de weg
daarnaartoe verschilt nog weleens.’ (respondent 51)
‘Ik weet wel dat iedereen heel verschillend op zijn eigen manier werkt. En ook dat
dat ook wel prima is omdat leidinggevenden zeggen: “Het maakt niet uit hoe je het
doet, als je het doel maar bereikt”.’ (respondent 54)
Medewerkers zijn op zoek naar ruimte tussen de regels omdat zij het idee
hebben dat wanneer zij rekening houden met de situatie van de uitke-
ringsgerechtigden, dit effectiever is en zij hiermee hun doel meer bereiken.
Meer ruimte voor uitvoerders, maar niet zonder voorwaarden

Aan de hand van een analyse van de verordeningen en citaten uit inter-
views met uitvoerende medewerkers (medewerkers inkomen, medewer-
kers reïntegratie en medewerkers handhaving) en hun managers, is in dit
hoofdstuk een beeld geschetst van de (uitvoerings)praktijk van handha-
ving van de verplichtingen in de Participatiewet. De eerste bevinding is
dat de gemeenten op papier een verschillende invulling geven aan de ar-
beidsverplichtingen. Deze verschillen gaan over de hoogte en de duur van
de maatregelen, de definitie van recidive en de eventuele spreiding van
maatregelen over meerdere maanden.
De tweede bevinding is dat uitvoerders verschillen in hoe ze regels toe-
passen en reageren op overtredingen. Bij beide aspecten is variatie aange-
troffen. Medewerkers maken eigen afwegingen en zeggen die af te stem-
men op de situatie van de uitkeringsgerechtigde, maar ook op hun eigen
werkwijze en de eigen opvatting hoe ze denken het uiteindelijke doel te
bereiken. Vanuit landelijk beleid draait handhaving om uniformiteit en ge-
lijke behandeling. Dit onderzoek laat zien dat medewerkers daarover een
andere opvatting hebben. Dat kan een verklaring zijn waarom medewer-
kers tussen de regels op zoek gaan naar een passende handhavingsaanpak.
Het is wenselijk dat bij het vormgeven van beleid en het opstellen van
wetten in de toekomst meer rekening wordt gehouden met de weerbarsti-
ge praktijk. In het rapport Weten is nog geen doen van de Wetenschappe-
lijke Raad voor het Regeringsbeleid (WRR 2017) wordt een lans gebroken
voor een realistische overheid die in haar regelgeving rekening houdt met
de gewone burger in al zijn variaties. Aansluitend pleit de Algemene Be-
127
157
PAULIEN DE WINTER
stuursdienst BZK in het rapport Regels en ruimte (2019) voor een vergro-
ting van de ruimte voor maatwerk door deregulering, professionalisering
van de uitvoering en versterking van de verantwoording voor de door de-
regulering ontstane discretionaire ruimten.
Bij meer ruimte voor uitvoerders past wel een kanttekening. De uit-
voeringsorganisatie moet rekening houden met de verschillende rolop-
vattingen en perspectieven op rechtvaardigheid van uitvoerders. Door-
dat medewerkers gefocust zijn op hun eigen aanpak en werkwijze, is er
kans op tunnelvisie waarbij medewerkers onterecht vasthouden aan hun
eigen opvattingen. Medewerkers moeten door de uitvoeringsinstantie en
door de managers gestimuleerd worden om gebalanceerde afwegingen te
blijven maken. Het is voor uitvoeringsinstanties daarom ook belangrijk
dat zij medewerkers met verschillende houdingen (dispositions) in dienst
hebben, zodat medewerkers in contact komen met collega’s met andere
opvattingen (gebaseerd op Zacka, p. 196-197).
Literatuur
Algemene Bestuursdienst BZK, 2019. Regels en ruimte. Verkenning maatwerk in dienst-
verlening en discretionaire ruimte. Den Haag.
Inspectie SZW, 2017. Handhaving arbeidsverplichtingen. R17/02. Den Haag: Ministerie
van Sociale Zaken en Werkgelegenheid.
May, P.J. & Burby, R., 1998. Making sense out of regulatory enforcement. Law & Policy,
20(2), 157-182.
May, P.J. & Winter, S., 2000. Reconsidering styles of regulatory enforcement: Patterns in
Danish agro-environmental inspection. Law & Policy, 22(2), 143-173.
May, P.J. & Winter, S.., 2011. Regulatory enforcement styles and compliance. In: Parker,
C. & Nielsen, V.L. (eds.), Explaining regulatory compliance. Business responses to
regulation, 222-244. Cheltenham: Edward Elgar Publishing.
Winter, P. de, 2019. Tussen de regels. Een rechtssociologische studie naar handhaving in
de sociale zekerheid. Groningen: Boom juridisch.
WRR, 2017. Weten is nog geen doen. Een realistisch perspectief op redzaamheid. Den
Haag: Wetenschappelijke Raad voor het Regeringsbeleid.
Zacka, B., 2017. When the state meets the street. Public service and moral agency. Cam-
bridge: The Belknap Press of Harvard University.
128
158
‘Vis ruikt nou eenmaal zo’
Responsive regulation door de Voedsel en Waren Autoriteit*
Peter Mascini en Eelco van Wijk
‘Responsive regulation’ is een theorie die sinds enige jaren nationaal en inter
nationaal furore maakt onder zowel beleidsmakers als wetenschappers (zie Van
damentele veronderstellingen te toetsen die aan de gangbare interpretatie van

deze theorie ten grondslag liggen. Dit doen wij aan de hand van een empirisch
onderzoek bij de Voedsel en Waren Autoriteit (VWA), omdat die ook responsive
regulation tot uitgangspunt van haar optreden heeft verheven (Voedsel en Waren
Autoriteit, 2006). We zetten eerst uiteen wat deze theorie inhoudt en welke ver
bespreken wij de bevindingen uit ons veldwerk en gaan wij in op de generaliseer

baarheid van onze bevindingen.
Wat houdt responsive regulation in?
Responsive regulation is vooral populair geworden door het gelijknamige boek
recht die zou kunnen volgen op de fasen van het ‘repressieve’ respectievelijk ‘au
bij toepassing van het recht rekening dient te worden gehouden met de gevolgen
die daarvan te voorzien zijn in de praktijk. In dit opzicht staan ze tegenover be
naderingen waarin het belang van de naleving van de letter van de wet centraal
staat.1
* We willen de reviewers en de redactie bedanken voor hun zorgvuldige, bruikbare commentaren
bedanken we de Voedsel en Waren Autoriteit en de controleurs voor hun ruimhartige medewer

king aan dit onderzoek. Dit onderzoek is mogelijk gemaakt door een subsidie van de interdepar
te (2002) responsive regulation als tegenpool opvat van ‘regulatory formalism’.
159
afgestemd op de specifieke context waarin ze worden toegepast, impliceert dat
Braithwaite, 1992, 5).2

Doorgaans wordt responsive regulation gebruikt ter verklaring van de meest ge
schikte handhavingsstijl in een specifieke context. In dit geval wordt het gezien
als mogelijkheid om de tegenstelling tussen een bestraffende – ‘command and
– ‘compliance’, ‘enabling’, ‘restorative justice’, ‘representative’ – gerichte hand

havingsstijl te overbruggen. Bij een uitsluitend overredende aanpak dreigt het
gevaar van inkapseling en vrijblijvendheid, en krijgen kwaadwillende normadres
santen de ruimte om regels te overtreden. Een louter bestraffende aanpak leidt
daarentegen tot afnemende betrokkenheid bij de doelen die aan overheidswetge
ving ten grondslag liggen, inefficiëntie wegens slepende rechtszaken, verstoorde
verhoudingen tussen controleurs en bedrijven, individueel en collectief verzet
van bedrijven en het achterhouden van informatie voor controleurs (Robens,
Voor deze specifieke invulling van responsive regulation is de centrale vraag hoe
kan worden bepaald wanneer voor een bestraffende aanpak moet worden gekozen
controlerende instanties hun wijze van optreden moeten toesnijden op de motie

ven en mogelijkheden tot regelnaleving van normadressanten (Braithwaite e.a.,
adaptability
controlerende overheidsinstanties zou een overredingsstijl dienen te zijn en pas
nadat bedrijven consequent weigeren mee te werken, zou op het toekennen van
1992). Op deze manier zouden controlerende instanties een stok achter de deur
hebben voor situaties van tegenwerking en kunnen zij volstaan met advisering of
lende manieren invulling worden gegeven aan responsive regulation en dat is dan ook gebeurd.
Zelf gebruiken Ayres en Braithwaite het om met voorstellen te komen om de tegenstelling te
overbruggen tussen overheidsregulering en zelfregulering. Dit kan volgens hen bijvoorbeeld
door middel van gedwongen zelfregulering en de versterking van de positie van derde partijen
als consumentenbewegingen.
complexere vorm van afschrikking waarin pure sancties worden gecombineerd met morele
boodschappen over de onwenselijkheid van regelovertredingen. Deze morele boodschappen
kunnen worden gebracht door middel van informele sociale controle, negatieve publiciteit en
stigmatisering. Doordat de reguleringscontext bij Ayers en Braithwaite geen vaste betekenis
heeft, onderkennen zij dat responsive regulation strikt genomen geen theorie is waarbij con
crete hypothesen worden afgeleid uit algemene vooronderstellingen. In feite komt responsive
regulation neer op een verzameling ideeën over de manier waarop regulering kan worden toege
sneden op uiteenlopende contexten.
160
tandaanpak zou als voordeel hebben dat de negatieve, onbedoelde gevolgen van
een bestraffende handhavingsstijl tot een minimum worden beperkt, doordat ze
wijze van optreden te differentiëren tussen de mogelijkheden en motieven van
bedrijfssectoren, tussen bedrijven binnen sectoren en binnen bedrijven tussen

verschillende regels; immers ‘a single company may be a ‘‘good apple’’ on some
Braithwaite, 1994a, 366).
Veronderstellingen van responsive regulation
Volgens de gangbare interpretatie van responsive regulation dient de keuze tus

sen handhavingsstijlen dus nauwkeurig te worden afgestemd op de motieven en
mogelijkheden van normadressanten om regels na te leven.
geschreven, staat op gespannen voet met de idee dat meerdere wegen naar Rome
sen aan de motieven en mogelijkheden tot regelnaleving van normadressanten

veronderstelt dat het mogelijk is om deze objectief vast te stellen. Dit gaat voorbij
hoe geringer de overeenstemming over de motieven en mogelijkheden tot regel

naleving van bedrijven, hoe groter de kans dat controleurs in vergelijkbare geval
len een verschillende handhavingsstijl voorstaan en hoe lastiger de juiste hand
havingstijl te bepalen is. Dit wil zeggen dat de toepasbaarheid van de gangbare
interpretatie van responsive regulation problematischer is naarmate de begrip
pen motivatie en mogelijkheden om regels na te leven rekbaarder zijn, aangezien
hierdoor minder makkelijk valt te bepalen welke handhavingsstijl het beste past
bij de overtreder en overtreding.
De gangbare interpretatie van responsive regulation veronderstelt ten tweede dat
het mogelijk is om de meest geschikte handhavingsstijl in de praktijk toe te pas
belemmeringen stuiten om met succes de handhavingsstijl toe te passen die aan

sluit bij hun beeld van de normadressanten, is de toepasbaarheid van responsive
regulation geringer.
Ze veronderstelt ten derde dat controleurs de perverse effecten van hun optreden
heid, juridisering, verstoorde verhoudingen, verzet en achterhouden van infor

matie wel optreden als controleurs een bestraffende stijl aanwenden, maar niet
als zij een overredingsstijl gebruiken. Ook achter de hand houden van een stok
om mee te slaan zou geen perverse gevolgen hebben; dit zou alleen gelden voor
recalcitrante bedrijven die er daadwerkelijk door worden geraakt, terwijl het be
sef dat controleurs ‘carry big sticks’ juist zou helpen om welwillende bedrijven in
het gareel te houden. Wederom is het een empirische vraag of deze veronderstel
wellicht voor dat controleurs er niet in slagen om de bedoelingen die aan hun
161
manier van optreden ten grondslag liggen over te brengen op de gecontroleerde

bedrijven? En indien zij hier niet in slagen, wat zijn daarvan de gevolgen? Zijn er
situaties waarin de negatieve onbedoelde gevolgen toch optreden, hoewel het niet
de bedoeling is geweest van de controleur om bestraffend op te treden?
De doelstelling van dit onderzoek is nagaan of er tussen controleurs van de VWA
overeenstemming bestaat over de motieven en mogelijkheden van bedrijven om
regels na te leven, of zij geen praktische belemmeringen tegenkomen om de meest
geschikte handhavingsstijl toe te passen en of zij de perverse effecten van hun
wijze van optreden in eigen hand hebben.
Methodologie
eerste helft van het onderzoek gerapporteerd.
Oost en Zuidwest. Elke regio kent twee teams die ieder een eigen gebied binnen de
regio bestrijken. Ieder team heeft zijn eigen teamleider en steunpunt met onder
andere werkplekken voor de controleurs. De dataverzameling vond plaats door
observaties, interviews en een vragenlijst. De observaties werden gedaan tijdens
het dagelijkse werk van de controleurs, het beoordelen van een vast aantal pun
ren en opslag van levensmiddelen en de beheersing van de bereidingsprocessen

van levensmiddelen door de ondernemer. De tweede auteur liep mee tijdens de
werkdag van de controleurs en interviewde ze vooraf over de situatie die ze ver
wachtten aan te treffen en hoe ze hun gedrag hierop dachten te gaan aanpassen.
Achteraf vroeg hij of zowel de situatie in het bedrijf als hun eigen handelen aan de
verwachtingen voldeden. Daarnaast observeerde hij de interactie tussen contro
leur en gecontroleerde. Bij de bezochte bedrijven heeft hij bovendien een vragen
lijst achtergelaten, over wat ze van de opstelling van de controleur vonden, wat ze
voor consequenties aan het bezoek van de controleur verbonden en wat ze in het
algemeen vonden van de regels en wetten waaraan ze zich moeten houden.
De resultaten van dit artikel zijn uitsluitend gebaseerd op de kwalitatieve gege
vens van het onderzoek. Ze zijn gebaseerd op observaties van zeventien contro
Alleen als dat expliciet in de tekst wordt vermeld, heeft de analyse betrekking op
het bovenindividuele niveau van controleteams.
Gebrek aan eenduidigheid
Taakopvattingen
Slagen controleurs van de VWA erin om eenduidig vast te stellen wat de juiste
aanpak is van bedrijven? Dit is niet het geval, omdat er onder controleurs zeker
geen eenduidige opvatting over het veld bestaat. Acht van de zeventien geobser
veerde controleurs vinden dat er meer foute dan goede bedrijven zijn, drie zijn
162
van mening dat de goede en foute bedrijven elkaar in evenwicht houden, terwijl
er vier denken dat er meer goede dan foute bedrijven zijn. Verder zijn er grote
verschillen tussen de controleurs als het gaat om waar ze op letten. Voor de één
is het papierwerk dat een ondernemer hoort bij te houden over allerlei tempera
tuurmetingen een belangrijk aspect van de inspectie, terwijl de ander deze zelfde
lijsten als een papieren tijger beschouwt, die niets zegt over de werkelijke bedrijfs
voering. Verder lopen de meningen uiteen over wat de meest effectieve manier
van handhaven is. Sommigen menen dat het schrijven van boetes de beste manier
is om ondernemers in het gareel te krijgen en te houden, anderen zijn juist van
mening dat het opbouwen van een vertrouwensband tussen controleur en gecon
over effectief handhaven variëren.
Handhavingsstijlen
Zijn er ook aanwijzingen dat controleurs in vergelijkbare gevallen verschillend op
west als oefening een aantal cases aan de controleurs voorgelegd met de vraag wat
voor maatregelen zij al dan niet op basis van de verschafte informatie zouden op
leggen, uitgaande van het maatregelbeleid dat op dat moment van kracht was. De
verschillen waren groot. De 23 controleurs die de cases beoordeelden, kwamen bij
één case tot niet minder dan elf verschillende maatregelen, met alle gevolgen van
dien voor het betreffende fictieve bedrijf. Bij één controleur zou de ondernemer
met twee waarschuwingen wegkomen, terwijl een ander voor meer dan tweedui
zend euro aan boetes zou opleggen. Opvallend was de duidelijke scheiding tussen
wijl het andere team dit maar in 26 procent van de gevallen deed.
Een tweede voorbeeld betreft de verschillen in de manier waarop teams controles
verrichtten tijdens een multicultureel zomerfestival waar onder andere bij ver
boete, terwijl het andere ‘compliance assistance’ als motto van de dag had uitge
roepen en vooral adviserend optrad. Standhouders aan de ene kant van het ter
rein werden dus wezenlijk anders benaderd en behandeld dan aan de andere kant.
De motivatie van beide groepen om anders op te treden is terug te voeren op een
duidelijk verschil in visie op de standhouders. Zo was de ene groep van mening
dat de standhouders – vaak particulieren – niet net zo benaderd konden worden
als de reguliere bedrijven die de controleurs bezoeken. Bovendien hielden zij re
kening met het feit dat de faciliteiten op het festival ook niet optimaal waren. De
andere groep benaderde de standhouders juist wel als reguliere bedrijven, en eiste
dan ook van de standhouders dat ze zich op een zelfde manier aan de dezelfde
regels hielden.
Een laatste voorbeeld betreft het omgaan met handhaving rondom Pekingeenden.
Er bestaat veel onduidelijkheid en onenigheid over hoe hiermee om te gaan. Er
als een bederfelijk product en eist van ondernemers dat de opslag aan bepaalde
voorwaarden voldoet. Volgens het andere kamp valt het met die bederfelijkheid
wel mee. Zij weigeren dan ook op te treden tegen het bewaren van Pekingeenden
163
buiten de koelkast. Veelzeggend was de vraag van een Chinese restauranthouder

die bij binnenkomst van de controleur vroeg of de controleur uit het ene of het
beboet zou worden voor zijn Pekingeenden.

Opvallend bij al deze gevonden verschillen is dat deze steeds tussen dezelfde
op tijdens het multiculturele festival en maakte tijdens de behandeling van de ca
wordt gedacht over het aandeel amoreel calculerende bedrijven. Deze verschil
lende visies zorgen er dus voor dat het ene team bij uiteenlopende situaties naar
de ‘strenge’ kant neigt, terwijl het andere in vergelijkbare gevallen juist veel meer
de meest geschikte aanpak, omdat individuele controleurs of teams fundamen

teel anders tegen een zelfde bedrijf of regel aankijken. Daar waar de één een over
redende aanpak voorstaat omdat hij een positief beeld heeft over een bedrijf, kiest
de ander voor een bestraffende aanpak omdat hij juist een negatief beeld heeft
van een vergelijkbaar bedrijf.
Geldt deze conclusie alleen voor de VWA? Zeker niet. Tal van onderzoeken heb
ben uitgewezen dat er grote verschillen bestaan tussen publieke dienstverleners
mening verschillen over wat ‘in redelijkheid’ van cliënten mag worden verwacht,
staan typen medewerkers tegenover elkaar die bijvoorbeeld worden aangeduid
als hardliners versus softliners (Brants, 2003), bureaucraten versus professionals
en bikkels versus watjes (De Groot, 1999). En hoewel enkele onderzoekers betwij
felen of de rolopvatting daadwerkelijk doorwerkt in beslissingen (bijv. Wilson,
den zijn de verlening van arbeidsongeschiktheidsuitkeringen (Sainsbury, 2001),
bureaucrat’ voor een bestraffende aanpak kiest en de andere voor een overredende
aanpak, doordat zij verschillende opvattingen hebben over identieke gevallen. 3
De bevinding dat verschillen in taakopvattingen leiden tot ongelijkheid en wil
komt dit zelfs vaker voor bij controlerende overheidsinstanties dan bij vele an
dere uitvoeringsinstanties. Controleurs beschikken namelijk in de praktijk over
3 Dat deze meningsverschillen zullen verdwijnen naarmate controleurs meer kennis en ervaring
de verklaring van ongelijkheid in de uitvoering van het asielbeleid heeft namelijk uitgewezen
dat aanzienlijke verschillen in voorgestelde beslissingen over een identieke asielaanvraag even
groot zijn onder ervaren en onervaren beslismedewerkers.
164
veel beslissingsruimte. Ze hebben vaak te maken met algemene richtlijnen die
een breed terrein dekken. Dit dwingt hen tot het interpreteren van de regels, het
stellen van prioriteiten en het maken van keuzes. Deze noodzaak wordt verder
versterkt doordat zij het belang van naleving van de regels dienen af te wegen
zij dit niet, dan lopen zij het risico aan legitimiteit in te boeten bij bedrijven en
het publiek. Bovendien onttrekt het werk van controleurs zich in grote mate aan
het blikveld van superieuren doordat zij naar ondernemingen toe gaan om con
mensen en daarbij beschikken over een grote mate van eigen beslissingsruimte
heid over de juiste aanpak, en de achtergronden van het eventuele gebrek hieraan,
doen sterk afbreuk aan de praktische toepasbaarheid van responsive regulation.
bepalen wat als de juiste aanpak moet worden beschouwd.
Belemmeringen
maar dat ze dit niet mogelijk achtten. Veel controleurs voelen zich bijvoorbeeld
verplicht om een boete op te leggen, terwijl ze van mening zijn dat dit averechts
nodig om een boete op te leggen, maar deden dit niet omdat ze voorzagen dat
deze de toets der kritiek van het bureau bestuurlijke boetes niet kon doorstaan.
de maatregel toch niet geëffectueerd zou worden. Ook de onduidelijkheid over de

situaties waarin regels moeten worden toegepast, is voor controleurs een sta in de
weg om te handelen zoals hen dat goeddunkt. Zo gaat een strenge controleur die
vaak maatregelen oplegt, bewust niet naar binnen bij horecazaken die Pekingeen
den verkopen, omdat volgens hem te veel onduidelijkheid bestaat over de vraag of
dit als bederfelijke waar moet worden beschouwd.
De grootste belemmering waar de controleurs tegenaan lopen, zijn taalbarrières.
die kleine supermarkten, shoarmazaken en pizzeria’s exploiteren, en de Chinese

restauranteigenaars bevolken deze categorie. In veel gevallen was bij het bezoe
ken van dergelijke ondernemers van wederzijdse communicatie eigenlijk geen
sprake, terwijl controleurs dat wel nodig achtten.
Zo had een Chinese ondernemer een boete gekregen voor te veel schadelijke stof
fen in zijn bakolie. De ondernemer had het boeterapport niet goed begrepen en
dacht dat hij de boete had gekregen omdat de temperatuur van de olie te hoog
was. De controleur heeft een half uur lang geprobeerd uit te leggen dat de boete
niet voor de te hoge temperatuur was, maar dat die temperatuur een oorzaak van
165
de hoeveelheid schadelijke stoffen kon zijn. Zelfs na een half uur uitleggen had de
ondernemer het nog steeds niet begrepen.
Een tweede voorbeeld betreft de eigenaar van een kleine Turkse supermarkt. Op
de opslag van bederfelijke producten had de controleur behoorlijk wat kritiek.
man het beste kon doen om de situatie te verbeteren. De man knikte bij alles en
het leek of hij alles had begrepen. De controleur vertrok tevreden. Toen de tweede
auteur de volgende dag de achtergelaten vragenlijst kwam ophalen, bleek dat de
sprak. Aan de situatie was dan ook niets veranderd, laat staan dat de man begreep
wat er niet goed was aan zijn manier van opslaan.
De taalbarrière weerhoudt controleurs er dus niet van om de stijl te gebruiken
die zij het meest geschikt achten, maar verkleint wel aanzienlijk de kans dat deze
ving zo’n belangrijke positie innemen, is het niet voldoende kunnen communice
ren desastreus voor de effectiviteit van de handhaving.
niet de stijl toepassen die ze het meest effectief achten, of dat ze zich niet reali
seren dat de stijl die zij toepassen ineffectief is. Behalve communicatieproblemen
vormen ook regels een barrière om de meest geschikte stijl tot uitvoering te bren
gen. Zowel een gebrek aan ruimte om van de regels af te wijken als onduidelijk
heid over de manier waarop ze moeten worden geïnterpreteerd, kunnen hiervoor
verantwoordelijk zijn. Ten slotte gebeurt het ook dat controleurs niet de meest
geschikte stijl aanwenden, doordat zij anticiperen op ongewenste reacties van
andere partijen verderop in de beslisketen.
onderkennen zij dat gebrek aan trainingsmiddelen om controleurs een flexibele

houding aan te leren, en politieke onwil om controleurs voldoende beslissings
ruimte te geven om regels naar eigen inzicht toe te laten passen, serieuze belem
hun beleid hebben opgenomen géén rekening houden met de attitude en beweeg
redenen van regelovertreders. Zij laten de boetebepaling eerst en vooral afhangen
krijgt daarmee in de praktijk toch voorrang boven het bevorderen van regelnale
ving, ook al schrijft het beleid formeel een responsieve aanpak voor.
Dat controleurs niet de stijl toepassen die zij het meest geschikt achten, omdat
die op gespannen voet staat met andere doelstellingen dan het tot stand brengen
van regelnaleving, is een regelmatig terugkerende bevinding. Zo heeft Wilthagen
(1993) laten zien dat controleurs niet altijd overschakelen op een bestraffende
aanpak bij in hun ogen weigerachtige bedrijven, omdat zij opzien tegen de gespan
arbeidsinspecteurs veel minder vaak processen aanspannen over overtredingen
166
met betrekking tot algemene gezondheidsrisico’s die zich eventueel pas op lange
re termijn manifesteren dan over concrete, specifieke overtredingen die hebben
geresulteerd in aanwijsbare ongevallen. De eerste zijn namelijk veel moeilijker te
bewijzen en laten rechters meer ruimte voor verschillende, en dus ook van die van
de inspecteur afwijkende, interpretaties, terwijl ze volgens de arbeidsinspecteurs
veel schadelijker zijn voor werknemers dan de makkelijker te bewijzen overtre
dingen van de arbowetgeving. Daarmee laten de arbeidsinspecteurs het risico van
het verlies van een rechtszaak zwaarder wegen dan de ernst van de overtreding.
Controleurs wenden dus ervaringen met eerdere beslissingen aan om andere
tutionele benadering, dat individuele medewerkers proberen hun positie in de

organisatie zo sterk mogelijk te maken door terug te vallen op instituties die hun
ke dienstverleners in de gelegenheid vanwege het repetitieve karakter van hun

werk. Zij nemen bij herhaling beslissingen over gevallen die in vele opzichten op
elkaar lijken. Dit geeft hen de mogelijkheid om de effecten van eerdere beslis
reputatieschade te voorkomen voor hun eigen positie of de organisatie waarin zij

werkzaam zijn.
Dit leidt tot de conclusie dat het zelfs bij overeenstemming over de meest effec
tieve manier van handhaven toch allerminst zeker is dat die succesvol wordt toe
gepast. Dit problematiseert de veronderstelling dat controleurs in staat zijn om
de meest geschikte stijl toe te passen.
‘Handhavingsstijl-als-bedoeld’ versus ‘handhavingsstijl-als-ervaren’
De VWA heeft het omschakelen tussen ‘command and control’ en ‘compliance as

sistance’ uitgeroepen tot een van de uitgangspunten van het huidige, vernieuwde
en zacht waar het kan’ zou niet alleen zijn bedoeld om handhavingsstijlen tussen
bedrijven te variëren, maar ook om dat binnen bedrijven te doen (gesprek beleids
medewerker en teamleider VWA, 19 september 2006). In de praktijk blijft hand
having dan ook meestal niet tot één instrument beperkt, maar worden dwang,
overtuiging, advisering en stimulering gecombineerd.
Treden negatieve, onbedoelde gevolgen alleen op indien controleurs bestraffend
optreden? Dit is niet zo, omdat controleurs er niet altijd in slagen om de bedoe
lingen van hun stijl van optreden over te brengen op de gecontroleerde bedrijven.
Sommige ondernemers die vrijwel gedurende de hele inspectie coöperatief wor
den benaderd, maar op specifieke punten worden gewaarschuwd of beboet, leg
gen dit naast zich neer, accepteren dit, of geven aan het met strafoplegging eens
te zijn. Velen hebben echter de neiging zich geheel te concentreren op de waar
schuwing of boete. In deze gevallen treden toch onbedoeld negatieve gevolgen
op over inspectieonderdelen waar controleurs niet bestraffend wilden optreden.
Drie willekeurige voorbeelden – er zijn er veel meer – kunnen dit illustreren. Zo
verliep de inspectie bij een snackbarhouder rooskleuring en bijna gezellig, totdat
167
de eigenaar een boete kreeg voor het overschrijden van een temperatuurgrens.
De man ontstak vervolgens in een tirade over de VWA en de rechtsstaat. In een
ander geval kreeg de eigenaresse van een pannenkoekenrestaurant hoofdzakelijk
uitleg over de manier waarop zij aan de regels zou kunnen voldoen en hoe de
regels precies in elkaar zaten, maar ook een waarschuwing. De eigenaresse ging
uitgebreid in discussie met de controleur over de zwaarte van de overtreding en
maakte later in de vragenlijst over de controle duidelijk dat haar relatie met de
zelfs nog een klacht in bij de VWA over het naar hun mening pietluttige optre
den van de controleur. In een visrestaurant knoopte de controleur voortdurend
gezelligheidspraatjes aan en liet hij duidelijk blijken rekening te houden met de
omstandigheden van het bedrijf. Toen hij echter aan het einde van inspectie zei
dat er nog wat kleine verbeterpunten waren, begon de vrouw in kwestie meteen
te klagen dat controleurs altijd wat te zeuren hebben. Ze was ook verbolgen over
de constatering dat de koelcel stonk – ‘vis ruikt nu eenmaal zo’ – en bleef daar
tegenin gaan.
Deze drie voorbeelden maken duidelijk dat de onbedoeld negatieve gevolgen niet
beperkt blijven tot de inspectieonderdelen waarover een controleur waarschuwt
ondernemer het optreden van de controleur als geheel percipieert en leidt precies
tot de negatieve gevolgen die de controleurs met hun overwegend overredende
aanpak probeerden te voorkomen. Dit betekent dat controleurs die binnen bedrij
ven een bestraffende en een overredende stijl afwisselen de perverse effecten
van hun handelen niet in eigen hand hebben, doordat ondernemers het optreden
negatiever opvatten dan de controleur het heeft bedoeld. Ze verliezen uit het oog
dat de controleur overwegend overredend is opgetreden en concentreren zich op
het negatieve relationele signaal dat van een waarschuwing of boete uitgaat.4
Ook onze conclusie over de derde veronderstelling die aan responsive regulation
ten grondslag ligt, kan worden gestaafd door ander onderzoek. Zo is uit verschil
lende onderzoeken gebleken dat er een verschil kan bestaan tussen de intenties
van controleurs en de manier waarop normadressanten ze ervaren (Etzioni, 1961,
1980; Braithwaite e.a., 1994). Een bedrijf kan het optreden van een controleur als
coöperatief ervaren, terwijl het de bedoeling van de controleur was om streng op
sitely sensitive to criticism and are capable of imputing stigmatizing intent to

feedback from inspectors which is intended to be ‘‘factual’’ rather than to ‘‘put
van een overtreding in combinatie met het accepteren van een dader – reintegra
4 De conclusie dat controleurs de onbedoeld negatieve gevolgen van hun handelen binnen bedrij
ven niet in eigen hand hebben, betekent niet dat dit ook niet zo is wanneer ze hun aanpak dif
ferentiëren tussen bedrijven of sectoren. Wij hebben daar geen directe aanwijzingen voor gevon
den.
168
tot gevolg heeft dan het afkeuren van zowel de overtreding als de dader – desin
tegrative shaming.
Botchkovar en Tittle (2005) beargumenteren echter niet alleen dat de toetsing
zelf aan de hand van een verbeterde toetsing aan dat geen van beide vormen van
lens. De reden hiervan zou volgens hen kunnen zijn dat delinquenten niet langer
openstaan voor een positieve benadering nadat eerst hun daden zijn veroordeeld
santen geneigd zijn hun reacties primair te laten afhangen van het meest nega
een sanctieapparaat leidt ertoe dat actoren calculerender worden en het algemeen
belang uit het oog verliezen. Dit plaatst het centrale uitgangspunt van de hand
havingspiramide, dat een overredende stijl slechts kan gedijen indien controleurs
‘carry big sticks’, in een geheel ander daglicht (Braithwaite, 2002, 34; zie ook
signaal dat uitgaat van het achter de hand houden van ‘big sticks’ óók negatieve
onbedoelde gevolgen teweegbrengt bij de normadressanten die daar nooit door
way of ‘‘voluntary’’ compliance at the bottom of the pyramid’.

Deze bevindingen leiden tot de conclusie dat controleurs de perverse effecten van
hun handelen alléén in eigen hand hebben als ze erin slagen hun bedoelingen dui
tieve gevolgen en een overredingsstijl niet, indien bedrijven het optreden van de
controleur als zodanig ervaren. Als controleurs er níet in slagen om hun bedoelin
gen over te brengen, treden onbedoeld negatieve gevolgen tóch op als controleurs
denken dat zij zich overredend opstellen terwijl bedrijven vijandigheid ervaren.
Ze blijven juist uit in het omgekeerde geval. Controleurs slagen er echter niet al
tijd in om hun bedoelingen over te brengen en hebben dus de perverse effecten
van hun handelen niet volledig in eigen hand. Daarmee is ook de toepasbaarheid
problematisch gebleken van de veronderstelling dat controleurs de perverse ef
fecten van hun optreden in eigen hand hebben.
Conclusie en discussie: bedoelingen versus receptie
ten overschakelen op een bestraffende aanpak als is gebleken dat overreding

niet werkt, omdat op deze manier welwillende bedrijven niet onnodig op de kast
worden gejaagd, terwijl onwelwillende bedrijven zo uiteindelijk tot aanpassing
kunnen worden gedwongen. De toepasbaarheid van dit uitgangspunt blijkt in
de uitvoeringspraktijk van de VWA op duidelijke grenzen te stuiten. In de eer
ste plaats is niet altijd duidelijk wanneer op een van beide stijlen moet worden
overgeschakeld, omdat controleurs verschillend denken over de manier waarop
bedrijven kunnen worden bewogen om regels na te leven. Onder controleurs be
169
staat geen overeenstemming over de mate waarin bedrijven bereid en in staat

zijn tot naleving van de regels. Bovendien zijn er verschillende redenen waardoor
controleurs er niet in slagen de meest geschikte stijl toe te passen. Er kan sprake
zijn van taalproblemen, onduidelijke of te rigide regels, of de toepassing van de
meest geschikte handhavingsstijl kan op gespannen voet staan met doelen waar
aan controleurs een hogere prioriteit geven dan het bewerkstelligen van regelna
leving. Tenslotte blijken onbedoeld negatieve gevolgen ook niet beperkt te blijven
tot bedrijven die controleurs hard hebben willen aanpakken. Daarvan is ook vaak
sprake als ze de intentie hadden om overwegend overredend op te treden, door
dat normadressanten geneigd zijn het optreden als bestraffender te ervaren dan
controleurs het hebben bedoeld.
centrale uitgangspunt van responsive regulation, uniek zijn voor de uitvoerings
nen worden geproblematiseerd op grond van wetenschappelijke inzichten die al

bestonden toen deze theorie werd geïntroduceerd. Dit roept de vraag op hoe de
dere factoren dan kennisaccumulatie een rol. Wellicht is deze theorie zo algemeen
geformuleerd dat een aantal fundamentele dilemma’s en problemen als te gemak
kelijk oplosbaar worden voorgesteld.
Volgens responsive regulation moet handhaving bijvoorbeeld enerzijds worden
afgestemd op de specifieke context, terwijl het anderzijds belangrijk is om de ‘rule
sive regulation niet ten koste hoeft te gaan van rechtszekerheid en rechtsgelijk
heid, terwijl dit juist als een belangrijk bezwaar tegen deze theorie is ingebracht
conflictsituaties net zo lang dienen te laten escaleren totdat normadressanten

zich gedwongen zien om zich neer te leggen bij de overmacht van de controleur,
terwijl anderzijds controleurs geduldig en vergevingsgezind zouden moeten zijn
Braithwaite, 1992). In hoeverre geduld en vergevingsgezindheid te verenigen zijn

met het vermogen en de bereidheid om conflicten te laten escaleren, blijft echter
in het midden.
In het verlengde hiervan kan men zich op basis van dit onderzoek afvragen in
hoeverre responsive regulation negatieve vicieuze cirkels van antagonisme kan
voorkomen. Onder andere op basis van inzichten uit de speltheorie beveelt deze
theorie aan om het gedrag van de normadressanten te imiteren; toont een bedrijf
goede wil dan dient een overredende aanpak te worden gebruikt, terwijl een
wen dat in het dreigen met of toepassen van sancties ligt besloten, roept even
wel weerstand op bij recalcitrante bedrijven en ondermijnt hun toch al geringe
betrokkenheid bij gezamenlijke belangen. Daardoor neemt de kans dat zij regels
overtreden verder toe en daarmee de kans op een nog bestraffender aanpak van
de controleur. Deze polarisatie wordt nog eens extra in de hand gewerkt door de
170
neiging van normadressanten om de handhavingsstijl als bestraffender te erva
ren dan controleurs het hebben bedoeld. Bovendien is het eenvoudiger gebleken
building trust with orga
nizations who have never experienced legal threat is one thing, rebuilding trust
lation ten grondslag ligt, lijkt het doorbreken van negatieve vicieuze cirkels te
bemoeilijken.
Dit roept de vraag op of het in bepaalde gevallen voor een controleur wellicht
verstandiger is om een negatieve opstelling van normadressanten bewust niet
ven kan ertoe leiden dat de normadressant de controleur gaat nadoen (vgl. Vonk,
traden zich vervolgens gaan gedragen als ‘good citizens’ wanneer ze overredend
tegemoet worden getreden door een controleur, terwijl ze een bestraffende aan
pak verwachtten. Uiteraard heeft ook deze strategie een keerzijde, aangezien juist
onwillige bedrijven de mogelijkheid krijgen om misbruik te maken van de terug
houdende opstelling van de controleur. Dit leidt tot de enigszins pessimistische
conclusie dat het schipperen blijft tussen de Scylla van een bestraffende en de
Charibda van een overredende aanpak.
Toch kan uit ons onderzoek ten minste een nieuwe richting worden afgeleid voor
de optimalisering van handhaving. We hebben laten zien dat de bedoelingen van
controleurs regelmatig anders worden ervaren door normadressanten. Over de
vragen hoe dit komt en wat hieraan te doen valt, is vooralsnog weinig bekend. Wel
waren om te bluffen over de macht van hun inspectiedienst om regelovertredingen

te vervolgen. Wellicht zijn er ook strategieën denkbaar waarmee controleurs zich
bedoelingen over te brengen op normadressanten, opent de zoektocht naar de

manier waarop normadressanten de bedoelingen van controleurs interpreteren
en naar tactieken waarmee controleurs deze interpretaties kunnen beïnvloeden.
Dit leidt tot de aanbeveling om de aandacht te verleggen van het feitelijk optreden
van controleurs en hun bedoelingen naar de manier waarop normadressanten dit
meer aandacht in onderzoek en beleid voor het sociaal geconstrueerde karakter

van interacties tussen controleurs en normadressanten.
Literatuur
Responsive Regulation: Transcending the Deregulation
Debate
The Modern Law Review,
71
171
Going by the Book: The Problem of Regulatory Unreason-

ableness
Poortwachters van de Nederlandse arbeidsmarkt:
arbeidsvoorziening en de verlening van tewerkstellingsvergunningen
An Investigation Using Russian Respondents. Criminology, 43
Case of Immigration in Canada. Canadian Public Administration, 45

Restorative Justice and Responsive Regulation
University Press.
Law & Policy,
29
Law & Policy, 16
Postures, and Responsive Regulation. Law & Policy, 29
street-level bureaucrat. Amsterdams Sociologisch Tijdschrift, 30
Braithwaite? Tijdschrift voor Criminologie, 49(4), 386-399.
asielzaken. Migrantenstudies, 14
Hulpverlening aan arme gezinnen: naar een typologie van
hulpverleners.
Journal of Social Welfare Law, 2
The Uses of Discretion

Etzioni, A. (1961). A Comparative Analysis of Complex Organizations: On Power, Involvement,
and Their Correlates
A
Sociological Reader on Complex Organizations
Immigration Inspections. Law & Society Review, 26

Groot, C. (1999). De poortwachters van Nederland: over ambtenaren en de asielprocedure.
Universiteit van Amsterdam.
Corporate Regulation. Beyond ‘Punish or Persuade’

Press.
Journal of Law and Society, 27

Human Service Organizations
Environment and Enforcement: Regulation and the Social Definition of
Pollution
on Pearce and Tombs. British Journal of Criminology, 30
172
Organizations,
Uncertainties and Risk
Invloeden op regelnaleving door bedrijven. Inzichten uit
wetenschappelijk onderzoek
Regulatory Justice: Implementing a Wage-Price Freeze
Sage Foundation.
Handbook of Regulation and Administrative Law
Enforcing Regulation.
Assistance Agency. Social Problems, 23

Organisaties vergelijkerwijs. Ontwikkeling en relevantie van het
sociologisch denken over organisaties
Street-level Bureaucracy: Dilemmas of the Individual in Public Services
Journal of
Research in Crime and Delinquency, 31
Regulatory Standards. Criminology, 32

Risico’s in bedrijf: omgaan met gezondheidsrisico’s op de werkvloer.
van regels tot minder ongelukken? Tijdschrift voor gezondheidswetenschappen, 79(4),
het asielbeleid verklaard. Amsterdams Sociologisch Tijdschrift, 31

Law & Policy, 3
Law and Society in Transition. Toward Responsive Law
Enforcement. Law & Society Review, 40
Environmental Governance. Public Administration Review, 64

Safety and Health at Work: Report of the Committee
Rooij, B. van (2006). Regulating Land and Pollution in China. Lawmaking, Compliance, and
Enforcement; Theory and Cases
The Uses of Discretion
Journal of Public Administration Research and Theory,

7
Stokkom, B. van (2004). Handhaven: Eerst kiezen dan doen. Sociaal-wetenschappelijke
mogelijkheden en beperkingen
173
Cooperation. Administrative Science Quarterly, 44
Shuttle Challenger. Administrative Science Quarterly, 35

Veen, R. van der (1990). De sociale grenzen van beleid: een onderzoek naar de uitvoering en
effecten van het stelsel van sociale zekerheid
Voedsel en Waren Autoriteit (2006). Handhaven met verstand en gevoel. Vernieuwing van het
handhavingsbeleid van de VWA op basis van risico’s en proportionaliteit
Intermediair, 43, 41.
The Howard Journal, 42

Bureaucracy: What Government Agencies Do and Why they Do It
Sociaal Recht,
Wilthagen, T. (1993). Het overheidstoezicht op de arbeidsomstandigheden: Een onderzoek naar
het functioneren van de Arbeidsinspectie
174
Week 5
175
176
Bottom-Up Workplace Law Enforcement:
An Empirical Analysis
CHARLOTTE S. ALEXANDER* & ARTHI PRASAD**
This Article presents an original analysis of newly available data from a

landmark survey of 4387 low-wage, front-line workers in the three largest U.S.
cities. We analyze data on worker claims, retaliation, and legal knowledge to
investigate what we call “bottom-up” workplace law enforcement, or the reliance
of many labor and employment statutes on workers themselves to enforce their
rights. We conclude that bottom-up workplace law enforcement may fail to protect
the workers who are most vulnerable to workplace rights violations, as they often
lack the legal knowledge and incentives to complain that are prerequisites for
enforcement activity.
INTRODUCTION .................................................................................................... 1070

I. THEORETICAL FRAMEWORK ............................................................................. 1074
II. METHODOLOGY ............................................................................................... 1077
A. 2008 UNREGULATED WORK SURVEY .................................................... 1077
B. SUMMARY STATISTICS ........................................................................... 1080
C. OUR ANALYSIS ...................................................................................... 1082
III. RESULTS ........................................................................................................ 1085
A. WORKPLACE PROBLEMS ........................................................................ 1085
B. CLAIMS .................................................................................................. 1089
C. RETALIATION ......................................................................................... 1091
D. SUBSTANTIVE LEGAL KNOWLEDGE ....................................................... 1093
E. PROCEDURAL LEGAL KNOWLEDGE ........................................................ 1095
IV. IMPLICATIONS FOR BOTTOM-UP WORKPLACE LAW ENFORCEMENT.............. 1097
A. THE FAILURE OF OPERATIONAL RIGHTS ................................................ 1102
B. SELF-PERPETUATING ENFORCEMENT GAP ............................................. 1107
V. ALTERNATIVES ............................................................................................... 1108
A. DISSEMINATING LEGAL KNOWLEDGE .................................................... 1110
B. MINIMIZING THE COSTS ......................................................................... 1113
C. MAXIMIZING THE BENEFITS ................................................................... 1115
* Assistant Professor of Legal Studies, Department of Risk Management and

Insurance, J. Mack Robinson College of Business, Georgia State University; Secondary
Appointment, Georgia State University College of Law. We would like to thank Annette
Bernhardt and Trey Spiller for providing the data analyzed in this Article as well as
extensive and indispensable guidance on its use. Nicole Sykes, Teodor Markov, and Pam
Brannon provided excellent research assistance. We have also been fortunate to present early
versions of this paper at the Labor and Employment Law Colloquium, People and
Organizations Conference, Employment and Labor Law Scholars’ Forum, Emory Law
School Faculty Colloquium, and Cornell University School of Industrial and Labor Relations
Faculty Seminar. Thanks to David Lopez for his insight on the EEOC’s work on behalf of
vulnerable workers and to Shannon Gleeson, Steve Willborn, and David Weil for their useful
feedback.
** M.A. Economics, Andrew Young School of Policy Studies, Georgia State
University.
177
1070 INDIANA LAW JOURNAL [Vol. 89:1069
D. INCREASING GOVERNMENT ENFORCEMENT........................................... 1117

APPENDIX A: UNREGULATED WORK SURVEY QUESTIONS .................................. 1120
SURVEY QUESTIONS RELATED TO WORKPLACE PROBLEMS AND CLAIMS .. 1120
SURVEY QUESTIONS RELATED TO RETALIATION ........................................ 1121
SURVEY QUESTIONS RELATED TO SUBSTANTIVE AND PROCEDURAL LEGAL
KNOWLEDGE .............................................................................................. 1122
SURVEY QUESTIONS RELATED TO SOCIAL NETWORKS .............................. 1122
APPENDIX B ......................................................................................................... 1125
APPENDIX C ......................................................................................................... 1127
INTRODUCTION
Workplace rights in the United States are generally enforced from the bottom
up. With few exceptions, labor and employment laws contain private rights of
action that enable workers themselves to bring lawsuits when their rights are
violated.1 These private lawsuits vastly outnumber government enforcement actions
against law-breaking employers.2 Even what seems to be top-down government
enforcement is often bottom-up enforcement in disguise, as government agencies
depend in large part on worker complaints to direct their enforcement activity.3
1. The Occupational Safety and Health Act and, in large part, the National Labor
Relations Act are significant exceptions, as workers themselves may not sue to enforce
statutory health and safety protections or prohibitions on unfair labor practices. Michael J.
Wishnie, Immigrants and the Right to Petition, 78 N.Y.U. L. REV. 667, 731–32 (2003)
(discussing the lack of a private right of action under the OSHA and NLRA).
2. Between fiscal years 1997 and 2012, for example, the number of cases filed by
private plaintiffs in federal court to enforce civil rights employment laws was forty-eight
times the number of cases filed by the U.S. Equal Employment Opportunity Commission
(EEOC). Compare EEOC Litigation Statistics, FY 1997 through FY 2012, U.S. EQUAL EMP.
OPPORTUNITY COMM’N, http://www.eeoc.gov/eeoc/statistics/enforcement/litigation.cfm
(listing 5705 EEOC enforcement suits filed in the federal district courts in fiscal years 1997
through 2012), with Supplements Table C-2A: Cases Commenced by Nature of Suit,
Judicial Business Archive, U.S. COURTS, http://www.uscourts.gov/Statistics/JudicialBusiness
/archive.aspx (listing annual reports identifying 274,422 “Civil Rights Employment” and
“ADA—Employment” cases filed by private plaintiffs in federal district court in fiscal years
1997 through 2012). Likewise, between fiscal years 2000 and 2011, private plaintiffs filed
thirty-eight times more Federal Fair Labor Standards Act cases than did the U.S. Department
of Labor. Author-compiled data, Public Access to Court Electronic Records (PACER)
(finding 1382 cases with the FLSA Nature of Suit code filed in federal court by the U.S.
Secretary of Labor versus 52,865 cases filed by private plaintiffs) (on file with authors).
3. For example, almost all lawsuits brought by the EEOC, which enforces federal
antidiscrimination statutes, begin as worker-filed charges. Though the EEOC may
affirmatively initiate an investigation using a mechanism known as a commissioner charge,
this charge is rarely used, with only twelve charges filed in fiscal year 2012. Performance
and Accountability Report: FY2012, U.S. EQUAL EMP. OPPORTUNITY COMM’N,
http://www.eeoc.gov/eeoc/plan/2012par_performance.cfm (“[I]n FY 2012, 12 new
Commissioner charges were filed.”). The U.S. Department of Labor may also initiate its own
affirmative “directed investigations,” but those are vastly outnumbered by complaint-driven
investigations. DAVID WEIL, IMPROVING WORKPLACE CONDITIONS THROUGH STRATEGIC
ENFORCEMENT 8 (2010), available at http://www.dol.gov/whd/resources
178
2014] BOTTOM-UP WORKPLACE LAW ENFORCEMENT 1071
Workplace law enforcement therefore depends significantly on worker “voice,”

with workers themselves identifying violations of their rights and making claims to
enforce them.4
In theory, a bottom-up system should produce an accessible, responsive, and
efficient workplace law enforcement regime. The parties with the most information
about violations and the greatest incentive to correct them—the workers—drive the
enforcement process.5 Workers need not wait on cumbersome, budget-strapped, or
politically hamstrung government agencies, but can take enforcement duties into
their own hands.6 And when government agencies do act, the bottom-up system
should allow them to allocate resources efficiently: the “market” in complaints
should signal to agencies which employers are bad actors in need of reform. 7
However elegantly designed, there is good reason to believe that the system fails
in practice, and that it fails particularly badly in the case of workers who are most
vulnerable to workplace rights violations.8 These workers include women, those
/strategicEnforcement.pdf (reporting that over 75% of U.S. Department of Labor Wage and
Hour Division investigations in 2008 arose from worker complaints, as opposed to
agency-initiated complaints in accordance with the Department’s strategic enforcement
priorities); Janice Fine & Jennifer Gordon, Strengthening Labor Standards Enforcement
Through Partnerships with Workers’ Organizations, 38 POL. & SOC’Y 552, 556 (2010) (“The
[U.S. Department of Labor] is thus expending most of its resources on inspectors responding
to complaints filed by individual workers, without an underlying justification that this
approach is likely to be effective in detecting violators.”).
4. ALBERT O. HIRSCHMAN, EXIT, VOICE, AND LOYALTY: RESPONSES TO DECLINE IN
FIRMS, ORGANIZATIONS, AND STATES 30 (1970) (defining a worker’s choices when faced
with a problem within her organization as exit, voice, and loyalty: “Voice is here defined as
any attempt at all to change, rather than to escape from, an objectionable state of
affairs . . . .”); see also Mitchell v. Robert De Mario Jewelry, Inc., 361 U.S. 288, 292 (1960)
(“For weighty practical and other reasons, Congress did not seek to secure compliance with
prescribed standards through continuing detailed federal supervision or inspection of
payrolls. Rather it chose to rely on information and complaints received from employees
seeking to vindicate rights claimed to have been denied.”). But see HIRSCHMAN, supra at 21–
25 (discussing the option to “exit,” or take one’s labor or business elsewhere); id. at 76–105
(discussing a “theory of loyalty”).
5. See Gideon Yaniv, Complaining About Noncompliance with the Minimum Wage
Law, 14 INT’L REV. L. & ECON. 351, 351–52 (1994) (“Being the direct (and sole) victims of
noncompliance, underpaid workers are naturally perceived as the faithful guardians of the
minimum wage law.”).
6. Cynthia Estlund, Rebuilding the Law of the Workplace in an Era of Self-Regulation,
105 COLUM. L. REV. 319, 376–77 (2005) (“Private lawsuits can potentially help to fill the
enforcement gap left by the undercommitment of public resources; indeed, they can
sometimes supply a big gun where public enforcement has none to wield.”).
7. See David Weil & Amanda Pyles, Why Complain? Complaints, Compliance, and
the Problem of Enforcement in the U.S. Workplace, 27 COMP. LAB. L. & POL’Y. J. 59, 70
(2005) (“Ideally, regulators would like to assume two things: (1) that the workers who are
complaining are voicing legitimate grievances and representing them accurately (in other
words, that employees working under lawful conditions are not complaining); and (2) that
workers who are experiencing violations will complain.” (emphasis in original)).
8. See, e.g., id. at 73 (finding “only modest overlap between [worker] complaints and
[employer] compliance” with the Fair Labor Standards and Occupational Safety and Health
Acts).
179
with less education, nonunionized workers, and undocumented workers, all of

whom hold relatively disempowered positions both in the workplace and in society
as a whole.9 The system fails these workers because it is built on two foundational,
misplaced assumptions: (1) that workers have the substantive and procedural legal
knowledge to identify violations of their rights and access the proper enforcement
procedures, and (2) that workers have sufficient incentives to file suit or make
agency complaints.10 Using a rich source of newly available survey data, this
Article demonstrates that for many low-wage, front-line workers—those who earn
below the median wage for the city where they live and hold nonmanagerial,
nonsupervisory, or nontechnical jobs—neither assumption applies.
This Article analyzes data that were originally collected as part of the 2008
Unregulated Work Survey, a landmark study of 4387 low-wage, front-line workers
in the three largest U.S. cities (New York, Chicago, and Los Angeles).11 These data
are valuable, as they include “hidden” groups such as undocumented people or day
laborers who are often excluded from research due to their marginalized social,
political, and economic statuses. The data are also important because they enable
empirical analyses of the determinants of claiming behavior and the transformation
of problems into legal claims, subjects of much study within the sociolegal,
economic, and political science literatures.12 Finally, analyses such as these are
important as a policy matter. As Steven Willborn puts it, as “policymakers sort
through the many choices available when they are deciding how to enforce a labor
statute,” they should be guided by both theory and data on the flaws and benefits of
our current system of workplace law enforcement.13
Our analysis reveals gaps in workers’ legal knowledge and powerful incentives
to stay silent in the face of workplace problems. First, the data show that more than
three-quarters of surveyed workers did not know where to file a government
complaint about a problem on the job, and almost 60% of these workers
misunderstood their minimum wage and overtime rights.14 The data also suggest
that female and undocumented workers had less accurate procedural legal
knowledge than their male and documented counterparts.15 Legal knowledge
9. See generally infra Part III (reporting results pertaining to women, nonunionized,
less educated, and undocumented workers).
10. Steven Willborn describes workers’ lack of legal knowledge and lack of incentives
slightly differently, as problems with detection and discovery of labor regulation violations
and problems with prosecution. Steven L. Willborn, Labor Enforcement Theory: The Case of
Public vs. Private Enforcement, in REGULATING DECENT WORK (Colin Fenwick & John
Howe eds., forthcoming) (manuscript at 3).
11. See infra Part II.
12. See infra Part I.
13. Willborn, supra note 10, at 1 (“The consequences of [the differences in types of]
labor enforcement are under-studied empirically. Labor enforcement theory, if anything, is
even less well-studied. This is unfortunate, in part, because only a well-developed theory can
help policymakers sort through the many choices available when they are deciding how to
enforce a labor statute. Moreover, a well-developed theory is necessary to guide researchers
in their efforts to understand both general issues of labor enforcement and particular issues
relating to specific legislation.”).
14. See infra Part III.D–E.
15. See infra Part III.D–E.
180
therefore appears to decrease with a worker’s relative power and stability, and
many workers simply may not have the information necessary to become
workplace law enforcers.
Second, even informed workers may lack an incentive to exercise “voice” on the
job. The data show that about 43% of workers who had experienced a workplace
problem within the twelve months prior to the survey decided not to make a
claim.16 The most common reason for workers’ silence was their fear of employer
retaliation; the second was their belief that their claim would have no effect.17 Their
fear of retaliation was well founded: about 43% of workers reported experiencing
employer retaliation as a result of their most recent claim about a justiciable
workplace problem in the twelve months before the survey.18 Likewise, workers’
lack of faith in the efficacy of claiming was borne out by the data, as only about
15% of employers addressed the claim or promised to address it.19
Though workplace laws offer a set of protections and inducements to entice
workers to become law enforcers—what we call “operational rights”20—these
incentives are miscalibrated in the case of many low-wage, front-line workers,
whose fear of retaliation or doubt in the efficacy of complaining outweigh the
benefits that would accrue from workplace law enforcement.21
The data suggest, then, that low-wage, front-line workers such as the cashiers,
parking lot attendants, and dishwashers captured by the Unregulated Work Survey
may be particularly unsuited as bottom-up rights enforcers. At the same time,
numerous studies have shown that workplace rights violations are extraordinarily
prevalent in the very industries that employ these workers.22 Thus, the same
low-wage, front-line workers who are the most likely to experience workplace
rights violations may be the least likely to become rights enforcers. Workers are
overdeterred from claiming, and employers may be underdeterred from complying,
creating a self-perpetuating enforcement gap in labor and employment law.23
16. See infra Part III.A–B.

17. See infra Part III.B.
18. See infra Part III.C.
19. See infra Part III.C.
20. See infra Part IV.A; see also Charlotte S. Alexander, Explaining Peripheral Labor:
A Poultry Industry Case Study, 33 BERKELEY J. EMP. & LAB. L. 353, 386 (2012) (using the
term “operational rights” to refer to the set of incentives and protections designed to
operationalize or effectuate substantive rights, to entice workers to become bottom-up law
enforcers).
21. See Yuval Feldman & Orly Lobel, The Incentives Matrix: The Comparative
Effectiveness of Rewards, Liabilities, Duties, and Protections for Reporting Illegality, 88
TEX. L. REV. 1151, 1155 (2010) (“Legislators, as well as adjudicators, must consider
tailoring the incentives embedded in the law to the misconduct and the individual that it
targets as an enforcer.”).
22. See infra note 138 and accompanying text.
23. Estlund, supra note 6, at 330 (“Traditional enforcement mechanisms have often
failed to raise the cost of noncompliance high enough to outweigh the immediate gains from
noncompliance. Most enforcement actions secure only the back wages owed to employees
(if that); that means opportunistic employers risk very little by underpaying employees and
hoping—quite realistically—to avoid enforcement, either by inspection or complaint . . . .
Simply ignoring the law is an especially tempting strategy for marginal producers at the
181
The Article proceeds as follows: Part I reviews previous work in the sociolegal,
economic, and political science literatures that provides a theoretical framework for
the present analysis. Part II describes the methodology of the 2008 Unregulated
Work Survey, the data collected, and our own analysis. Part III reports our findings
on the frequency of worker claims, the prevalence of retaliation against workers
who did make claims, the level of workers’ substantive and procedural legal
knowledge, and workers’ incentives to engage in enforcement activity. Part IV
discusses the implications of a rights enforcement regime that leaves out the most
vulnerable workers. Part V concludes by examining possible alternative structures
for workplace law enforcement.
I. THEORETICAL FRAMEWORK
This Article fits into a tradition within the sociolegal, economic, and political
science literatures of studying the determinants of claiming behavior and provides
an opportunity to expand on existing models. The classic framework for
understanding the transformation of problems into legal claims comes from
William Felstiner and his coauthors.24 In this view, problems move through three
phases: naming, blaming, and claiming.25 First, a person must “say[] to [her]self
that a particular experience has been injurious.”26 This is naming. Second, she must
“attribute[] [the] injury to the fault of another individual or social entity”—
blaming.27 Third, in the process of claiming, she voices her grievance “to the
person or entity believed to be responsible and asks for some remedy.”28
Richard Miller, Austin Sarat, and Marc Galanter, among others, propose a more
complex dispute generation and resolution pyramid, with additional layers beyond
claiming.29 The pyramid accounts for claims that are disputed by the opposing
bottom of the production chain, who have little fixed capital or stake in their reputation, who
tend to operate under the regulatory radar, and who often rely heavily on undocumented
immigrant workers who are too fearful or desperate to complain.” (footnotes omitted)); see
also David Weil, “Broken Windows,” Vulnerable Workers, and the Future of Worker
Representation, FORUM, Jan. 2012, at 1, 1 (“In the presence of persistent violations, keeping
one’s head down, ‘staying out of other people’s business,’ and turning a blind eye to unfair
treatment of others is a survival strategy.”).
24. William L.F. Felstiner, Richard L. Abel & Austin Sarat, The Emergence and
Transformation of Disputes: Naming, Blaming, Claiming . . ., 15 LAW & SOC’Y REV. 631,
635–36 (1980–81).
25. Id.
26. Id. at 635.
27. Id.
28. Id. at 635–36.
29. Marc Galanter, Reading the Landscape of Disputes: What We Know and Don’t
Know (and Think We Know) About Our Allegedly Contentious and Litigious Society, 31
UCLA L. REV. 4, 11 (1983); Richard E. Miller & Austin Sarat, Grievances, Claims, and
Disputes: Assessing the Adversary Culture, 15 LAW & SOC’Y REV. 525, 545 (1980–81); see
also Laura Beth Nielsen & Robert L. Nelson, Rights Realized? An Empirical Analysis of
Employment Discrimination Litigation as a Claiming System, 2005 WIS. L. REV. 663, 681
(“At the base of the dispute pyramid are perceived injurious experiences—the broad mass of
injuries that people recognize. Some proportion of these perceived injurious experiences are
grievances: injuries that involve a violation of right or entitlement. . . . Only some grievances
182
party and the resulting suit filing and disposition in court, and the pyramid’s shape
reflects the fact that many “named” problems—the wide bottom layers—never
progress all the way through to litigation at the pyramid’s peak.30
In a separate strand of economics literature, Albert Hirschman’s seminal work
examines problems that arise within organizations such as the workplace.31
Hirschman posits that people have three options in such circumstances: exit, voice,
and loyalty.32 Exit is the option to take one’s labor elsewhere; voice is a worker’s
attempt “to change, rather than to escape from, an objectionable state of affairs”;
and loyalty can result in a decision to remain within the organization and silently
withstand recognized problems.33 In many ways, Hirschman’s “voice” can be
thought of as Felstiner’s “claiming.”
The present Article takes this prior work as the starting point for investigating
the empirics of workplace dispute generation and resolution and for constructing a
more complex model that incorporates workers’ legal knowledge and incentive
structures. Figure 1 illustrates how such a model might look, presenting the
panoply of options that workers face when identifying problems on the job and
deciding whether and how to address them. The darkened arrows represent a direct
path from problem identification to claim resolution using formal legal means such
as a lawsuit or government agency complaint. This is the path that our current
system of bottom-up workplace law enforcement envisions: workers have the legal
knowledge to engage in “naming”; they then “blame” the appropriate party and are
incentivized and knowledgeable enough to exercise “voice,” to make a “claim” in
the form of a lawsuit or agency complaint and pursue it through resolution.
However, as the complexity of the diagram illustrates, there are many alternative
paths. And as the diminishing size of each pyramid layer indicates, many problems
“escape” along the way. Workers may not have the substantive legal knowledge to
recognize justiciable problems (violations of their legal rights at work) as such.34
Workers who “name” the problems they face may nevertheless choose exit or
loyalty over voice, or may not know who to “blame,” as a legal matter, for the
issues they face. Workers who do make claims may be faced with inaction or
become claims: when an individual contacts the party allegedly responsible for the
grievance. Fewer still are disputes: when the party allegedly responsible for an individual’s
claim initially denies their responsibility. Some number of disputes results in filings: a
formal complaint (in a litigation model, a court filing). The smallest category of all is made
up of trials: cases that are adjudicated.” (emphases omitted) (footnotes omitted)).
30. Galanter, supra note 29, at 11–26 (describing the pyramid).
31. HIRSCHMAN, supra note 4.
32. Id. at 21–35 (discussing exit); id. at 30 (discussing voice); id. at 76–105 (discussing
loyalty).
33. See id. at 21–25 (discussing exit), id. at 30 (discussing voice); id. at 76–105
(discussing loyalty, which, in the alternative, might result in a decision to stay within the
organization and exercise voice).
34. Justiciable problems are those that “raise issues in civil law, have civil legal aspects,
and have consequences shaped by civil law.” Rebecca L. Sandefur, The Fulcrum Point of
Equal Access to Justice: Legal and Nonlegal Institutions of Remedy, 42 LOY. L.A. L. REV.
949, 968 (2009) (footnote omitted). See generally HAZEL GENN WITH SARAH BEINART,
STEVEN FINCH, CHRISTOS KOROVESSIS & PATTEN SMITH, PATHS TO JUSTICE: WHAT PEOPLE
DO AND THINK ABOUT GOING TO LAW (1999) (setting out theory of justiciable problems,
unmet legal needs, and access to justice).
183
retaliation by their employers, at which point they may choose exit or loyalty, or
may start again at the bottom of the pyramid and progress through the steps of
“naming,” “blaming,” and choosing whether or not to make a retaliation “claim.”
Moreover, if a claim is escalated and becomes a lawsuit or agency complaint, the
worker’s incentive structure must push her to pursue the claim to resolution in
whatever form it takes (e.g., win or loss at trial, case dismissal, settlement). The
costs of formal legal action in time and treasure may build over the duration of the
process, and the worker may be retaliated against at any point, which then triggers
its own decision tree. Throughout the process, the worker’s incentive structure must
continue to encourage pursuit of her claim; for her to persevere through the claim’s
resolution, the benefits of claiming must consistently outweigh its costs.
Figure 1. Workplace dispute pyramid.
Thus, this more complex workplace dispute generation and resolution pyramid
synthesizes Hirschman’s theories with those of Felstiner and his progeny. It also
adds to the model the concepts of substantive and procedural legal knowledge and
incentives as prerequisites for problem identification, the transformation of
problems into claims, and then those claims’ resolution.
The very complexity of the model suggests that the simplistic assumptions
behind our bottom-up workplace law enforcement regime (indicated by the
darkened arrows in Figure 1) may be misplaced. Workers do not simply progress
through a process of naming, blaming, claiming, and claim resolution; they must at
every step choose among alternatives and have the legal knowledge and appropriate
incentive structure to encourage progression.
The pyramid also raises many empirical questions. Which workers have the
substantive legal knowledge to recognize certain workplace problems as justiciable
rights violations? Which workers have the procedural legal knowledge required to
transform their claims into lawsuits or agency complaints? Which workers choose
exit or silent loyalty over voice, at which stage in the process, and why? What
forms of extralegal dispute resolution might workers be accessing as alternatives to
184
court, and how do those mechanisms work? How are workers’ incentive
structures—their assessments of claiming’s costs and benefits—functioning to
encourage or discourage claiming? How do those incentives shift and change over
the course of the process, and do they work differently for claims of different type
and magnitude? Finally, as a policy matter, how do we want these incentive
structures to function, and what are the implications for our bottom-up workplace
law enforcement regime?
The data presented in this Article provide an opportunity to examine some of
these questions, focusing on low-wage, front-line workers as potential workplace
rights enforcers. In particular, this Article investigates which of these workers make
claims on the job, the prevalence of retaliation against workers who do engage in
claiming behavior, the level of workers’ substantive and procedural legal
knowledge, and workers’ incentives to engage in claiming activity. However, these
data also leave many questions unanswered, paving the way for additional future
research.
II. METHODOLOGY
A. 2008 Unregulated Work Survey
The data analyzed in this Article were originally collected as part of the 2008
Unregulated Work Survey, a large-scale study of 4387 low-wage, front-line
workers in New York, Chicago, and Los Angeles. With foundation funding, the
study was conducted by researchers from the National Employment Law Project,
the University of Illinois at Chicago’s Center for Urban Economic Development,
and the UCLA Institute for Research on Labor and Employment, in partnership
with researchers at Cornell University, the City University of New York, Rutgers
University, and the UCLA Downtown Labor Center. The survey team originally
published some of its findings in a 2009 report, Broken Laws, Unprotected
Workers: Violations of Employment and Labor Laws in America’s Cities,35 and
released its full data set to outside researchers, including the present authors, for the
first time in January 2012.
The survey targeted workers in each city who were at least eighteen years old
and who worked in industries where the median wage was less than 85% of the city
median wage.36 Workers also had to hold “front-line,” or nonmanagerial,
nonsupervisory, and nontechnical jobs.37 Workers surveyed included cooks and
dishwashers, car-wash workers, groundskeepers, maids and housekeepers, parking
lot attendants, cashiers, cafeteria workers, tellers, garment workers, teacher’s
assistants, and security guards.38 Their median hourly wage was $8.02. Each
35. ANNETTE BERNHARDT, RUTH MILKMAN, NIK THEODORE, DOUGLAS HECKATHORN,

MIRABAI AUER, JAMES DEFILIPPIS, ANA LUZ GONZÁLEZ, VICTOR NARRO, JASON
PERELSHTEYN, DIANA POLSON & MICHAEL SPILLER, BROKEN LAWS, UNPROTECTED
WORKERS: VIOLATIONS OF EMPLOYMENT AND LABOR LAWS IN AMERICA’S CITIES (2009),
available at http://www.nelp.org/page/-/brokenlaws/BrokenLawsReport2009.pdf?nocdn=1.
36. Id. at 56.
37. Id.
38. Id. at 16.
185
worker participated in a face-to-face interview in his or her native language,

including English, Spanish, Polish, Russian, Chinese, French, Urdu, Hindi, and
Korean.39
Many of these workers would not be reached via traditional survey methods, as
they might not appear on voter registration rolls or other official lists from which
survey respondents are typically drawn or, due to their immigration status, might be
unwilling to give personal information to a stranger survey-taker. Thus, a typically
drawn random sample could exclude or undersample these workers, making it
extremely difficult to infer anything about their characteristics or presence in the
general population. As a result, the survey team employed a sampling methodology
known as respondent-driven sampling (RDS).40 First pioneered in public health
research, RDS uses survey respondents’ own social networks to allow researchers
to gain access to “hidden” or hard-to-reach groups.41 RDS methodology solves the
problems of lack of access and mistrust by relying on each early respondent to
recruit additional later respondents. In an RDS survey, each initial “seed” receives a
limited number of coded coupons to distribute to other people in her social
39. Id. at 13–14.

40. For more on RDS methodology, see Douglas D. Heckathorn, Extensions of
Respondent-Driven Sampling: Analyzing Continuous Variables and Controlling for
Differential Recruitment, 37 SOC. METHODOLOGY 151 (2007); Douglas D. Heckathorn,
Respondent-Driven Sampling: A New Approach to the Study of Hidden Populations, 44 SOC.
PROBS. 174 (1997) [hereinafter New Approach]; Douglas D. Heckathorn, Respondent-Driven
Sampling II: Deriving Valid Population Estimates from Chain-Referral Samples of Hidden
Populations, 49 SOC. PROBS. 11 (2002) [hereinafter Chain-Referral]; Matthew J. Salganik &
Douglas D. Heckathorn, Sampling and Estimation in Hidden Populations Using Respondent-
Driven Sampling, 34 SOC. METHODOLOGY 193 (2004); Matthew J. Salganik, Variance
Estimation, Design Effects, and Sample Size Calculations for Respondent-Driven Sampling,
83 J. URB. HEALTH 98 (2006); Erik Volz & Douglas D. Heckathorn, Probability Based
Estimation Theory for Respondent Driven Sampling, 24 J. OFFICIAL STAT. 79 (2008). The
above sources were all cited with approval in Michael W. Spiller, Annette Bernhardt, Jason
Perelshteyn & Douglas Heckathorn, Technical Report: Sampling, Fielding, and Estimation
in the 2008 Unregulated Work Survey (Ctr. for the Study of Econ. & Soc’y Working Paper
Series, Paper No. 62, 2010), available at http://www.economyandsociety.org/wp
-content/uploads/2013/08/wp62_SpillerHeckathornetal_Sampling.pdf.
41. More specifically, some populations may be too socially stigmatized to self-identify
as members of the relevant group, some may live in socially insular communities that are not
readily accessible to researchers, and some may be members of groups that are simply too
small to be captured in a typical randomly drawn sample from the U.S. population as a
whole. See, e.g., Douglas D. Heckathorn & Joan Jeffri, Finding the Beat: Using Respondent-
Driven Sampling to Study Jazz Musicians, 28 POETICS 307, 308 (2001) (defining jazz
musicians as a “hidden population” in that “(1) no sampling frame exists, so the size and
boundaries of the population are unknown, (2) there are strong privacy concerns, not because
of illegal or stigmatized behavior, but because of the tight but informal networks which
outsiders find hard to penetrate, and (3) the population constitutes a small proportion of the
general population”); Heckathorn, New Approach, supra note 40, at 174 (“‘Hidden
populations’ have two characteristics: first, no sampling frame exists, so the size and
boundaries of the population are unknown; and second, there exist strong privacy concerns,
because membership involves stigmatized or illegal behavior, leading individuals to refuse to
cooperate, or give unreliable answers to protect their privacy.”).
186
network.42 “Seeds” are compensated for their participation in the survey, as well as
for each additional respondent they successfully recruit. They vouch for the
integrity of the survey process, solving the trust problem, and they provide
researchers with an entrée to hidden communities of respondents, solving the
access problem.
Though RDS allows researchers to reach respondents who would likely be
missed by traditional survey methods, it creates problems of its own by generating
a nonrandom sample of the population. Each respondent’s participation is not
determined by chance, but instead by whether she is connected socially to a “seed.”
Because randomness is the foundational assumption upon which all of statistical
inference is built, RDS must compensate for the nonrandom nature of the samples
that its network-recruitment method generates. RDS does so by weighting each
response by the size and homophily, or sameness, of that respondent’s social
network.43 (Researchers determine size and homophily by asking a series of
carefully designed questions of each respondent.)44 Weighting produces an estimate
of the prevalence of the particular variable of interest in the general population,
known as the “population estimate,” within an RDS-specified confidence interval.45
In other words, it allows the characteristics and experiences of the respondents who
were surveyed to be generalized, roughly, to the three cities’ entire population of
low-wage, front-line workers—here, approximately 1.64 million total workers.46
Though RDS-like “snowball” or chain-referral survey methodologies have
existed for years, this weighting procedure represents an innovation, allowing
researchers to reliably account for the nonrandomness of respondents’ inclusion in
the sample. RDS methods remain an active area of research in statistics and survey
design, but the methodology is generally accepted as valid and is in wide use,
particularly within the fields of sociology and public health.47
42. Spiller et al., supra note 40, at 4 (“Our RDS survey began with an initial set of
population members to be surveyed, which we located through our contacts in each city.
These ‘seeds’ were then given a fixed number of uniquely numbered dollar-bill sized
coupons to pass on to other eligible population members.”).
43. See Chain-Referral, supra note 40, at 13, 20 (discussing the nonrandom nature of
RDS referrals and the resulting introduction of “homophily bias” and overrepresentation of
respondents with larger social networks).
44. A list of survey questions regarding respondents’ social networks appears in
Appendix A. Researchers also map respondents’ social networks by tracking coupon codes
and charting the links between seeds and recruited respondents. Researchers are then able to
gauge the level of homophily of each social network along a variety of axes to determine
whether, for example, female seeds recruited exclusively female respondents and Latino/a or
Hispanic seeds recruited exclusively Latino/a or Hispanic respondents.
45. See Chain-Referral, supra note 40, at 14 (discussing computation of population
estimates and standard errors).
46. BERNHARDT ET AL., supra note 35, at 2.
47. See, e.g., Matthew J. Salganik, Respondent-Driven Sampling in the Real World, 23
EPIDEMIOLOGY 148 (2012) (discussing widespread use of RDS and ongoing refinements of
RDS analytical techniques).
187
B. Summary Statistics
Table 1 reports the number and percentage of survey respondents by a variety of

demographic and job characteristic variables. The term “high road,” which appears
in Table 1 and throughout this Article, is used by labor economists to designate
employers that provide some fringe benefits and pay their employees more than the
required minimum:
It is commonly posited that employers face the choice of either

competing on the basis of cost or competing on the basis of quality,
variety, and service. In popular discussion, the former is referred to as
the “low road” and the latter as the “high road,” on the assumption that
the latter implies more generous employment conditions (such as
wages) and new work systems.48
The original survey team designated a “high-road” employer as one that provided
two or more of the following: health insurance, paid vacation days, paid sick days,
and pay raises to employees. In this data set, the “high-road” designation was not
particularly correlated with employer size. Employers with fewer than one hundred
employees were roughly as likely as employers with more than one hundred
employees to offer “high-road” benefits.
Table 1. Characteristics of Survey Respondents
Variable Frequency (Percent)
City
Chicago 1097 (26%)
Los Angeles 1772 (42%)
New York 1391 (33%)
Gender
Female 2250 (53%)
Male 2009 (47%)
Race/Ethnicity
Latino/a or Hispanic 2506 (59%)
Black or African American 919 (22%)
Asian or other race 614 (14%)
White 214 (5%)
Age
46+ years 1258 (30%)
36–45 years 1071 (26%)
26–35 years 1036 (25%)
18–25 years 785 (19%)
48. Paul Osterman, How Common Is Workplace Transformation and Who Adopts It?, 47
INDUS. & LAB. REL. REV. 173, 179 (1994) (citations omitted); see generally ARNE L.
KALLEBERG, GOOD JOBS, BAD JOBS: THE RISE OF POLARIZED AND PRECARIOUS EMPLOYMENT
SYSTEMS IN THE UNITED STATES, 1970S TO 2000S (2013) (describing the rise of precarious
labor and the growing gap in job quality between high-wage and low-wage work).
188
Table 1 (continued)

Education
Less than high school degree, no GED 1715 (40%)
High school degree or GED 1385 (33%)
Some college or associate’s degree 819 (19%)
College or higher degree 326 (8%)
Nativity and immigration status
U.S.-born citizen 1506 (36%)
Undocumented resident 1445 (34%)
Foreign-born citizen 518 (12%)
Documented resident 757 (18%)
Union membership
Not a union member 3911 (92%)
Union member 330 (8%)
Occupation during previous workweek49
Service occupations 2156 (51%)
Production, transportation, and material 944 (22%)
Sales and office occupations 760 (18%)
Construction, extraction, and maintenance 333 (8%)
Management, professional, and related 66 (2%)
Industry during previous workweek50
Arts, entertainment, recreation, accommodations, and food
services 833 (20%)
Other services (except public administration) 746 (18%)
Retail trade 662 (16%)
Nondurable goods manufacturing 605 (14%)
Educational, health, and social services 431 (10%)
Professional, scientific, and management 411 (10%)
Construction 267 (6%)
Transportation and warehousing 201 (5%)
Finance, insurance, real estate, and rental and leasing 58 (1%)
Durable goods manufacturing 32 (1%)
Wholesale trade 6 (0%)
Public administration 4 (0%)
Information and communications 3 (0%)
49. Occupation categories are those used by the U.S. Department of Labor’s Bureau of
Labor Statistics. See BUREAU OF LABOR STATISTICS, DOL, 2010 CENSUS OCCUPATIONAL
CLASSIFICATION (2011), available at http://www.bls.gov/cps/cenocc.pdf.
50. Industry categories are those used by the U.S. Department of Labor’s Bureau of
Labor Statistics. See BUREAU OF LABOR STATISTICS, DOL, 2007 CENSUS INDUSTRIAL
CLASSIFICATION (2009), available at http://www.bls.gov/cps/cenind.pdf.
189
Table 1 (continued)

Job tenure
0–5 years 3675 (87%)
6–10 years 372 (9%)
11–15 years 130 (3%)
16–20 years 29 (1%)
Employer size
Under 100 employees 2371 (62%)
100 employees or more 1484 (38%)
“High-road” employer
0–1 “high-road” indicators 2967 (69%)
2+ “high-road” indicators 1233 (31%)
(N = 4387)
C. Our Analysis
The original survey team’s 2009 Broken Laws report focused primarily on
unpaid wages and health and safety violations experienced by low-wage, front-line
workers.51 Though the report also presented figures on the prevalence of retaliation
and workers who chose not to make claims about workplace problems (variables of
interest here), it did so only descriptively, without the regressions included in our
analysis that allow study of the characteristics of workers who made different
claiming choices.52 Nor did the report investigate our other variables of interest: the
level of workers’ substantive and procedural legal knowledge and the factors
associated with workers’ being more or less informed about their rights.
Drawing on the raw survey data, our analysis focuses on five main variables: (1)
workplace problems, or the frequency with which workers reported experiencing
problems on the job (the “naming” of problems); (2) claims, or the frequency with
which workers engaged in claiming behavior after identifying a workplace
problem; (3) retaliation, or the prevalence of retaliation against workers who made
claims about justiciable problems; (4) substantive legal knowledge, or the extent to
which workers knew their minimum wage and overtime rights;53 and (5) procedural
legal knowledge, or the extent to which workers knew how to make a government
complaint about a workplace problem.54
51. BERNHARDT ET AL., supra note 35, at 2 (“This report exposes a world of work in
which the core protections that many Americans take for granted—the right to be paid at
least the minimum wage, the right to be paid for overtime hours, the right to take meal
breaks, access to workers’ compensation when injured, and the right to advocate for better
working conditions—are failing significant numbers of workers.”).
52. Id. at 24–25.
53. The original survey asked only about workers’ knowledge of minimum wage and
overtime law and not about their knowledge of other types of workplace rights. Relevant
survey questions are reproduced in Appendix A.
54. The original survey asked only whether workers knew where to file a government
complaint and not how to engage in other methods of formal claiming such as contacting an
190
Using the RDS network-size weighting process, we first generated population

point estimates and confidence intervals for the five main variables.55 This allowed
us to calculate the percentage (within a given margin of error) of the three cities’
low-wage, front-line workforce that fell into each of the five main variable
categories, that is, the approximate percentage of workers who had identified
workplace problems, made claims, experienced retaliation, and had accurate
substantive and procedural legal knowledge. We also generated RDS estimates for
two additional variables: (1) the reasons workers gave for their decisions not to
make claims; and (2) for those who did make claims, the method(s) and subject(s)
of those claims.56
Next, we explored which groups of workers were more or less likely to identify
problems, make claims, experience retaliation, and have accurate substantive and
procedural legal knowledge. Using logistic regression, we modeled the associations
between these dependent variables and a number of independent variables,
including worker demographics and job characteristics.57 Though logistic
regression does not allow us to identify causal relationships, it does enable us to
identify whether workers with certain characteristics—for example, women or
undocumented workers—were more or less likely, holding all else constant, to
attorney, union representative, or other third party. Relevant survey questions are reproduced
in Appendix A.
55. We used the Respondent Driven Sampling Analysis Tool software (RDSAT),
available at http://www.respondentdrivensampling.org, to produce RDS population point
estimates and confidence intervals.
56. We report these results in Part III and Table 7 in Appendix B.
57. We used Stata to run all regressions. We chose to use a logit model because our
dependent variables are binary, that is, they have two values (0, 1). To test model fit, we
used a nonlinear version of the Hausman test (a combination of the suest and testn1
commands in Stata) to compare the logit and ordinary least squares (OLS) coefficients. The
null hypothesis of the test is that the coefficients from both estimators are not significantly
different from each other so the OLS estimator is consistent and efficient, versus an
alternative hypothesis that the coefficients from both estimators are significantly different
from each other so the OLS estimator would be inconsistent. The test results for our
regressions led us to reject the null hypothesis, indicating that logit is the better model. In
addition, we used probability weights provided by the original survey team in estimating the
logistic regression equations to correct for the sample bias introduced by the RDS sampling
methodology. These weights also took into account the undersampling of certain
races/ethnicities among survey respondents. For a further explanation of the weighting
process, see Spiller et al., supra note 40, at 13. Finally, following the protocol recommended
by the original survey team, we dropped from our analysis all respondents identified as home
health workers, due to changing coverage of this occupational category under federal and
state wage and hour law. For other uses and explanations of logistic regression methods and
interpretation, see James J. Brudney, Sara Schiavoni & Deborah J. Merritt, Judicial Hostility
Toward Labor Unions? Applying the Social Background Model to a Celebrated Concern, 60
OHIO ST. L.J. 1675, 1694–1713 (1999) (using logistic regression to model the effect of a
variety of independent variables on judicial voting in cases involving unions); Cass R.
Sunstein, David Schkade & Lisa Michelle Ellman, Ideological Voting on Federal Courts of
Appeals: A Preliminary Investigation, 90 VA. L. REV. 301, 316 (2004) (using logistic
regression to investigate the relationship between a judge’s ideology and voting patterns).
191
identify problems, make claims, experience retaliation, or have accurate substantive

and procedural legal knowledge.58
A preliminary note is warranted on possible limitations of the survey data. First,
the survey was not designed primarily to investigate questions of problem
identification, claiming behavior, retaliation, and legal knowledge. Though the
particular survey questions analyzed in this Article do address those topics, they
contain some gaps. For example, the questions concerning workers’ substantive
legal knowledge asked only about knowledge of minimum wage and overtime
rights, and not about other areas of employment and labor law, and the questions
concerning procedural legal knowledge inquired only about claims made to a
government agency, and not about claims made via litigation. As a result, the
conclusions drawn in this Article are necessarily limited.
Second, the data are now relatively old, as they were collected in 2008, five
years before the time of this writing. However, 2008 came after the onset of the
Great Recession, the effects of which low-wage, front-line workers continue to
feel.59 To the extent that depressed economic conditions have an impact on
workers’ claiming behavior, conditions are still relatively depressed, and
readministering the survey today would likely produce substantially similar
results.60 In fact, some commentators have observed that conditions for low-wage,
front-line workers have worsened since 2008, as employers increasingly use
temporary, part-time, and other contingent work arrangements, and union density
continues to decline.61
Third, one could argue that selection effects may be present among the
respondents who chose to participate in the survey. Only the more assertive
workers would choose to be surveyed, the argument goes, and those more assertive
workers might also be more likely to have made claims on the job. However, as
noted in Part III, only about 57% of workers made any claim at all in the twelve
months before the survey. If this figure represents only assertive, claim-prone
workers, then the real incidence of claiming behavior was even lower, further
emphasizing the problems with a claims-driven, bottom-up workplace law
enforcement regime.
58. Part III and Tables 8–12 in Appendix C present the results of our logistic regression
models.
59. See, e.g., Long-Term Unemployment: Causes, Consequences, and Solutions:
Hearing Before the Joint Econ. Comm., 111th Cong. 35, 36 (2010) (statement of Lawrence
F. Katz, Professor, Harvard University) (testifying about lingering “labor market distress”
well into 2010).
60. See Willborn, supra note 10, at 8 (postulating that “[e]conomic downturns may . . .
affect discovery of violations through private enforcement”).
61. PAMELA LOPREST & AUSTIN NICHOLS, UNEMPLOYMENT & RECOVERY PROJECT,
URBAN INST., LESS-EDUCATED CONTINUE TO LOSE JOBS IN RECOVERY—EVEN IN LOW-WAGE
INDUSTRIES (2011), available at http://www.urban.org/url.cfm?id=412382&RSSFeed
=UI_PovertyandSafetyNet.xml (“[S]ince the end of the recession, low-skill workers continue
to be the biggest net job losers across industries.”); WEIL, supra note 3, at 9–11 (discussing
the breakup, or “fissuring,” of the traditional employment relationship, declining
unionization rates, and other factors contributing to the degradation of conditions for low-
wage, front-line workers).
192
III. RESULTS
This Part, along with the tables in Appendices B and C,62 reports the results of
our analysis of the survey data. The subsequent Part discusses possible
explanations for our findings and considers the implications for bottom-up
workplace law enforcement.
A. Workplace Problems
About one-third of low-wage, front-line workers identified a problem on the job

in the twelve months before the survey.63 In Felstiner’s terms, these are the workers
who perceived, or “named” problems. However, the occurrence of actual
workplace problems was probably much higher. The original survey team gathered
respondents’ raw wage and hour data and determined that 26% of workers had been
paid less than the minimum wage during the previous workweek and 76% who had
worked more than forty hours had not been paid proper overtime.64 Thirty percent
of tipped workers had not been paid lawfully, and 70% of workers who had worked
beyond their scheduled shift were not paid for this extra working time.65 These
calculations focus only on violations of wage and hour laws; occurrences of actual
workplace problems may have been higher if other rights violations were also
considered.66
62. Table 7 in Appendix B reports the RDS population point estimates and confidence
intervals for the five main variables (workplace problems, claims, retaliation, substantive
legal knowledge, and procedural legal knowledge), along with the reasons workers gave for
their decisions not to make claims, and, for those who did engage in claims making, the
method(s) and subject(s) of their claims. Tables 8–12 in Appendix C show the full results of
our logistic regression equations for the workplace problems, claims, retaliation, and
substantive and procedural legal knowledge variables.
63. The original survey did not ask respondents directly about workplace problem
identification. As a result, we calculated the rate of problem identification by adding together
the workers who had made a claim about a problem with the workers who had identified a
problem but did not make a claim. See infra Appendix A (reproducing two separate survey
questions).
64. BERNHARDT ET AL., supra note 35, at 2–3.
65. Id.
66. For example, surveys of almost five hundred low-wage workers in the San
Francisco, California, area conducted by sociologist Shannon Gleeson revealed that:
Almost all claimants had had a rest or meal break denied or shortened
throughout their working life, and half reported problems with timely payment.
Over a third had been denied time off for illness and half had been become
injured or ill because of their job. Verbal abuse or degrading treatment had at
some point been present for two-thirds of the sample, and over a quarter of all
women reported being sexually harassed at some point.
Shannon Gleeson, The Limits to Lawyering: Findings from a Survey of Low-Wage Workers
in the San Francisco Bay Area, Remarks at the Labor and Global Solidarity, American
Sociological Association Labor & Labor Movements Section Mini-Conference (Aug. 12,
2013) (on file with author).
193
The difference between the one-third of workers who identified workplace

problems and the higher rates of actual problems points to issues with “naming”:
workers were not “saying to [themselves] that a particular experience has been
injurious.”67 As explored further below, the discrepancy may also be explained by
workers’ lack of substantive legal knowledge, creating an inability to identify
justiciable problems and an underreporting of problems at work.68
Table 2 lists the characteristics of the workers who were more and less likely to
identify workplace problems, all else equal. This and subsequent tables in this Part
list only the regression results that are statistically significant and that likely can be
explained by an actual association between the independent variables (worker
characteristics) and dependent variable (workplace problems) rather than by
chance. These results are indicated by asterisks. The table also lists results that are
just outside the bounds of significance, but appear to be intuitively correct,
indicated by a dagger. Full regression results for workplace problems are reported
in Table 8 in Appendix C.69
The regression results summarized throughout this Part should be interpreted
differently depending on the type of independent variable at issue. For independent
variables such as gender that have a discrete set of values (female and male), each
value that was included in the regression model is compared to its counterpart that
was not in the regression. These comparison values are known as the “base
category,” and are listed for each regression in Tables 8–12 in Appendix C. For
example, in Table 2, the variable “female” is included in the model, while “male” is
in the base category. Table 2 shows that, holding all else constant, women workers
were less likely to have identified a workplace problem in the previous year than
male workers, the value in the base category. For continuous independent variables
such as age, the results reflect the impact on the dependent variable when the value
of the independent variable increases by one. Referring again to Table 2, for
example, each additional year of age reported made a respondent less likely to
identify a workplace problem.
67. Felstiner et al., supra note 24, at 635.

68. The discrepancy might also be explained by respondents’ reluctance to identify
workplace problems to survey takers, out of a fear that their complaints might somehow
reach their employers’ ears.
69. Throughout this Part, we have chosen only to examine the sign, or direction, of the
association between independent and dependent variables, rather than the relative strength of
that association. Though additional work could be done (using odds ratios or marginal
effects) to rank the relative impact of each independent variable on the dependent variables,
because many of our results lack statistical significance, we have chosen to restrict our
analysis to an examination of only the sign or direction of the coefficient and not its relative
magnitude.
194
Table 2. Workplace Problems
Less likely to identify workplace problem More likely to identify workplace problem
Female** Education***
Age*** Does not know where to file
government complaint (procedural
Not a union member† legal knowledge)**
Job tenure of fewer than 12 months* Employer is not “high road”***
Employer has fewer than 100
employees**
(N = 3620)
***p 0.01; **p 0.05; *p 0.10; †p 0.15
As Table 2 shows, women, nonunion members, workers who had job tenure of
less than one year, and workers at employers with fewer than 100 employees were
all less likely than their counterparts in the base category to have identified a
problem at work in the previous twelve months. In addition, for every year of age a
respondent reported, she was less likely to have identified a workplace problem.
The results with respect to gender, union membership, and increasing age appear
generally to follow from theory. Women, nonunion members, and older workers,
who tend to hold less powerful positions in the workplace and society as a whole,
may be less accustomed to taking the relatively transgressive step of “naming”
workplace problems, to “saying to [themselves] that a particular experience has
been injurious.”70 These groups of workers may also have less legal knowledge
than other worker groups to enable them to identify problems at work. For
example, Table 5, infra, reports that each increasing year of a respondent’s age was
associated with a lower chance that a worker knew her minimum wage and
overtime rights. In addition, labor unions may serve an educational or
consciousness-raising function, enabling their members to engage in more
“naming” of workplace problems than their nonunion counterparts.71 Social science
research has also shown that women in particular may be reluctant to “name”
workplace problems as such, “to perceive themselves as targets of discrimination
[including sexual harassment], notwithstanding evidence” that these workplace
problems have occurred.72 Researchers posit that this may be a self-esteem
preservation strategy: if women workers’ failings are the result of personal
deficiencies, rather than their gender, their self-esteem may suffer in the
“performance domain,” but may be intact with respect to their standing in society
as a whole.73 Such a dynamic may be at play in the results reported in Table 2,
where women workers were less likely than men to “name” workplace problems.

71. The results reported in Table 5, infra, complicate this narrative somewhat.
72. Deborah L. Brake, Retaliation, 90 MINN. L. REV. 18, 26 (2005) (collecting social
science research).
73. Id. at 26–27.
195
With respect to job tenure, this result may merely be a consequence of exposure:
workers who had been employed for less than a full year may have had fewer
opportunities to experience, and to identify, a problem on the job during that year.
Alternatively, workers with shorter job tenures may have had less experience in the
workplace generally and be less aware of and able to identify workplace problems
when they occurred.
Finally, the positive association between a lower likelihood of naming
workplace problems and a worker’s employment at a relatively small employer
(fewer than 100 employees) could be interpreted in multiple ways. On the one
hand, there may be fewer workplace problems to identify in the first place at
smaller employers, because those workforces may be closer-knit and better at
avoiding workplace conflict than larger ones.74 On the other hand, the rate of
underlying workplace problems may be the same at smaller and larger employers,
but workers at smaller employers may be less likely to identify those problems
because of a greater sense of loyalty generated by small workplace size, or because
smaller “mom and pop” employers may tend to hire workers who are a priori less
educated and assertive than workers employed at larger, more sophisticated
operations.75
Table 2 also shows that every year of education reported by a worker made it
more likely that she had identified a workplace problem in the previous twelve
months. Workers who did not know where to file a government complaint and
those employed by “low-road” employers were also more likely to have identified
problems on the job. The education result makes sense: as education levels
increase, one would expect workers to be aware of their legal rights and norms of
fair treatment on the job, and to “name” violations of those rights and norms as
they occur. The result with respect to workers employed by low-road employers is
also intuitive: employers that do not provide health insurance, paid sick and
vacation leave, and pay raises may generate more dissatisfaction in the workplace,
and their workers may be more likely to identify workplace problems.
The result with respect to lack of procedural legal knowledge is harder to
explain, as one might expect that workers who did not know where to file a
government complaint would be less likely to have identified a workplace problem
74. See, e.g., Harry Matlay, Employee Relations in Small Firms: A Micro-Business
Perspective, 21 EMP. REL. 285, 292 (1999) (describing the results of a study of
approximately 6000 small businesses in the United Kingdom; finding that “[t]he deliberate
informality inherent in the majority of [small business] owner/managers’ styles seems to
have facilitated a personal approach that, in most cases, resulted in amicable solutions to
complex and occasionally acrimonious work-related situations”).
75. See, e.g., Marilyn Carroll, Mick Marchington, Jill Earnshaw & Stephen Taylor,
Recruitment in Small Firms: Processes, Methods and Problems, 21 EMP. REL. 236, 238
(1999) (describing results of a study of forty small firms in the United Kingdom; reporting
complaints from small firms about “the quality of labour available, including a lack of basic
literacy skills, particularly among young people”); George Saridakis, Rebeca Muñoz Torres
& Stewart Johnstone., Do Human Resource Practices Enhance Organizational Commitment
in SMEs with Low Employee Satisfaction?, 24 BRIT. J. MGMT. 445, 452 (2013) (finding in
analysis of nationwide survey of British firms that “employees in small organizations are
more committed [to their organizations] than their counterparts in large firms”).
196
in the previous twelve months. However, this result may in fact signal the
importance of a worker’s having both substantive and procedural legal knowledge
in order to progress up the dispute identification and resolution pyramid. Without
knowledge of her substantive legal rights in the workplace, procedural legal
knowledge alone—knowing where or how to file a workplace complaint with the
government—may not help a worker take the initial step of “saying to [herself] that
a particular experience has been injurious.”76
B. Claims
Of the workers who had identified a workplace problem within the year prior to
the survey, about 57% decided to make a claim of some type.77 Almost all of these
workers (96%) chose to make a claim to their employer, while very few
complained directly to a government agency or to an intermediary such as an
attorney or union representative. Sixty-five percent of workers’ claims were about
justiciable workplace problems, that is, violations of workers’ legal rights around
pay, discrimination, harassment, abuse, or occupational safety. The remaining
claims were nonjusticiable, centering on quality of worklife (e.g. commute time,
workload, and work schedule) and benefits. Of all workers who had identified a
workplace problem, then, approximately 37% made a claim about a justiciable
problem, 20% made a nonjusticiable claim, and 43% made no claim at all.
Of the 43% of workers who decided not to make a claim about an identified
workplace problem, the top two reasons workers gave for their decision were a fear
of being fired and a belief that the claim would make no difference.78 The next two
reasons (apart from “other”) were also retaliation related: the worker’s fear of
having her wages or hours cut and the worker’s knowledge of retaliation against
others for claiming behavior. The next reason points to a lack of procedural legal
knowledge: uncertainty as to how to make a claim.
Table 3 lists the worker characteristics that were associated with a greater or
lesser likelihood of claims making.79

77. Appendix A reproduces the relevant survey questions. Here, workers were asked,
“During the last 12 months, did you make a complaint, either by yourself or with co-
workers, about your working conditions, by going to your employer, supervisor or going to a
government agency?”
78. Workers were asked, “During the past 12 months, were there times when you DID
NOT complain, even though you had a problem at your job with dangerous working
conditions, discrimination, not being paid the minimum wage or not being paid overtime?”
and “What stopped you from complaining?” The menu of choices is listed in Appendix A.
79. Full regression results for the claims model are reported in Table 9 in Appendix C.
197
Table 3. Claims
Less likely to make claim More likely to make claim
Does not know minimum wage and overtime rights No results

(substantive legal knowledge)*
Does not know where to file a government
complaint (procedural legal knowledge)**
Employer is not “high road”*
Not a union member†
Age†
Job tenure of fewer than 12 months***
Asian or other race†
(N = 1422)
***p 0.01; **p 0.05; *p 0.10; †p 0.15
The regression results summarized in Table 3 suggest that workers without

substantive and procedural legal knowledge, as well as workers who were not
union members or who worked for a low-road employer, were less likely to make
claims about a workplace problem than their counterparts in the base category. In
addition, for each additional year of age reported by a worker, the likelihood that
she had made a claim in the twelve months before the survey decreased. Finally,
workers who had been employed for fewer than twelve months were less likely to
have made a claim during that time, as were workers who identified as “Asian or
other race.”80
These results are, on the whole, as one might expect: less powerful and
economically stable workers appear less likely to engage in claiming behavior than
their more powerful and stable coworkers. For example, claims may be riskier for
older than younger workers, because if an older worker is fired in retaliation,
finding a replacement job may be relatively more difficult.81 Likewise, nonunion
members might be expected to make claims less frequently than their unionized
counterparts, as they have less support and fewer channels through which to air
their grievances.
In addition, workers who were employed by low-road employers were less
likely to make claims than their counterparts in the base category. At first glance,
80. “Other races” include any worker who did not identify as “White,” “Black or
African American,” or “Latino/a or Hispanic.”
81. See, e.g., Sewin Chan & Ann Huff Stevens, Job Loss and Employment Patterns of
Older Workers, 19 J. LAB. ECON. 484, 485 (2001) (finding that employment rates of workers
aged 50 and older “are substantially below the employment rates of comparably aged
nondisplaced workers” and that “a job loss at age 50 or above has substantial and long-
lasting employment effects”). Older workers may also have more dependents to support than
younger workers, increasing the harm caused by a retaliatory job loss and making those
workers more reluctant to risk retaliation by making claims.
198
this result may seem counterintuitive: workers at low-road companies would

presumably have more to complain about than their counterparts who receive
benefits and pay raises. However, low-road employers may hire a less empowered
workforce to begin with, or might send signals, either explicitly or implicitly by
providing substandard wages and working conditions, that worker claims would be
futile.
It is also unsurprising that workers who did not know their substantive legal
rights or how to make a government complaint were less likely to make claims than
workers who did have this knowledge. This result squares with the self-reported
reasons, discussed above, that workers gave for choosing not to make claims after
identifying workplace problems: a top explanation given was that workers did not
know how to make a claim.
Workers who had held their job for fewer than twelve months were also less
likely to have made a claim about a workplace problem during that time. This may
simply reflect the opportunity (or lack thereof) to experience a problem: the longer
a worker remains on the job, the more potential exposure she has to workplace
problems and the more chances she has to make claims. In addition, the more
experience a worker accrues at a job, the more valued she may feel at the company
and the more secure she may feel in engaging in claiming behavior.
Finally, it is difficult to explain why workers who identified as “Asian or other
race” were less likely to make claims than other workers who identified as “White,”
“Black or African American,” or “Latino/a or Hispanic.” As Tables 5 and 6 show,
workers who identified as “Asian or other race” were less likely than members of
other racial and ethnic groups to have accurate substantive and procedural legal
knowledge. This lack of legal knowledge—itself an interesting result—probably
contributed to these workers’ lower likelihood of engaging in claiming behavior.
C. Retaliation
Of the workers who had made claims about justiciable problems in the twelve
months before the survey, about 43% experienced some form of employer reprisal
in response to their most recent complaint.82 Of these reprisals, roughly 35%
constituted unlawful retaliation in violation of labor and employment laws,
including termination or suspension, calls to the police or immigration, decreases in
hours, and abuse and harassment.83 The remaining reprisals were acts such as
threats that likely did not rise to the level of an “adverse employment action,” a
requirement for retaliation to be unlawful,84 but nevertheless likely had a silencing
effect on workers.
82. Workers were asked, “Did your employer or supervisor know you made this
complaint?” and “Did your employer or supervisor do any of the following as a direct result
of this complaint?” The menu of choices is listed in Appendix A.
83. To identify unlawful retaliation, the original survey team analyzed respondents’
descriptions of the reprisals they experienced and counted only those acts that would be
actionable under the relevant labor and employment laws. Spiller et al., supra note 40, at 25–
26 (describing criteria used).
84. Jennifer Clemons, FLSA Retaliation: A Continuum of Employee Protection, 53
BAYLOR L. REV. 535 (2001) (discussing adverse employment action requirement for FLSA
199
Apart from those workers who experienced some form of retaliation,

approximately 15% of workers reported that their employers addressed or promised
to address the workers’ claims. The remaining workers’ employers took no action,
ignored the workers’ claims, or engaged in some other response.
Thus, in sum, of all workers who had made claims about justiciable workplace
problems, about 15% experienced unlawful retaliation, 28% experienced some
other form of reprisal, another 15% had their claims addressed or promised to be
addressed, and 42% were met with employer inaction or some other response.
Table 4 lists the variables that were associated with higher and lower risks of
employer retaliation of all types, not just reprisals that would constitute violations
of the law. Full regression results on retaliation are reported in Table 10 in
Appendix C.
Table 4. Retaliation
Less likely to experience retaliation More likely to experience retaliation
Made claim alone* No results
(N = 560)
***p 0.01; **p 0.05; *p 0.10; †p 0.15
As Table 4 shows, we achieved only one statistically significant result with

respect to retaliation. We suspect that a smaller number of observations (560 as
opposed to between 1422 and 3620 for the other dependent variables) may be
driving this outcome. It also may be explained by the very nature of retaliation.
Individual-level worker characteristics such as gender and race/ethnicity may not
be associated strongly with the experience of reprisals because an employer’s
retaliatory threats or acts may be directed toward the workforce generally, as
opposed to singling out a particular worker or workers.85 For example, an employer
may gather the entire low-wage, front-line workforce and threaten them about the
negative consequences of claiming behavior.86 Such threats may have a silencing
effect across the workforce, without regard to individual worker characteristics, and
may therefore not register in the regression analyses performed in this Part.
Turning to the one result in Table 4, workers who made claims alone were less
likely to experience retaliation of any type than those who complained in a group.
This may seem counterintuitive, as individual claimants would appear to be weaker
and more vulnerable to being “picked off” by employer retaliation. However,
employers may perceive claims brought by groups of workers as more threatening
than those brought by individuals, and therefore be more likely to engage in
reprisals against group rather than individual claimants.
retaliation cases); Rosalie Berger Levinson, Parsing the Meaning of “Adverse Employment
Action” in Title VII Disparate Treatment, Sexual Harassment, and Retaliation Claims: What
Should Be Actionable Wrongdoing?, 56 OKLA. L. REV. 623 (2003) (same for Title VII).
85. Thanks to Pauline Kim for this insight.
86. See, e.g., infra Part V.A (recounting a circumstance in which an entire workforce
was likely silenced by an employer’s threats).
200
D. Substantive Legal Knowledge
Approximately 41% of workers had accurate substantive legal knowledge in the

area of wage and hour law, meaning that 59% misunderstood their minimum wage
and overtime rights.87 With respect to the minimum wage, workers were asked to
name the current applicable minimum wage in their city, and the original survey
team checked the responses for accuracy based on the federal, state, and local laws
in place at the time. Workers over- and underestimated the required minimum wage
in about equal numbers: 25% gave a number below the required minimum, while
21% gave a number above.
Table 5 lists the variables that were associated with a greater or lesser likelihood
of a worker’s knowing her minimum wage and overtime rights. Full substantive
legal knowledge regression results are reported in Table 11 in Appendix C.
Table 5. Substantive Legal Knowledge
Less likely to know minimum wage and More likely to know minimum wage and
overtime rights overtime rights
Not a union member*

Employer has fewer than 100
employees*** No legal immigration status†
Age***
Black or African American***
White***
Asian or other race†
(N = 3620)
***p 0.01; **p 0.05; *p 0.10; †p 0.15
Here, workers who worked for relatively small employers (those with 100 or
fewer employees) were less likely than employees of larger companies to know
their minimum wage and overtime rights. As workers reported higher ages, the
likelihood that they had accurate substantive legal knowledge also fell. In addition,
workers who identified as “Black or African American,” “White,” or “Asian or
other race” were less likely to know their minimum wage and overtime rights than
those who identified as Latino/a or Hispanic. On the other hand, workers who were
not union members and who had no legal immigration status (i.e. were
undocumented workers) were more likely to report accurate knowledge of their
wage and hour rights.
The result with respect to employer size is relatively intuitive: smaller
companies that lack human resources departments or formal employee screening
87. Workers were asked, “As far as you know, what is the current minimum wage in
[CITY]?” and “As far as you know, do employers have to pay workers more than their usual
wage when they work more than 40 hours in a week?” See infra Appendix A.
201
procedures may hire a less sophisticated workforce with lower levels of legal
knowledge. Alternatively, smaller companies may do a worse job than larger
companies of educating their employees about workplace laws through training and
compliance programs.88
With respect to age, older workers may be at a disadvantage in knowing their
minimum wage rights because the minimum wage changes over time. For example,
the oldest respondent was eighty-one years old at the time of the survey. Since his
or her eighteenth birthday in 1945, federal minimum wage laws have changed
twenty-six times, California’s state minimum wage law has changed at least eleven
times, Illinois’ at least fourteen times, and New York’s at least sixteen times.89 A
worker may learn the minimum wage when she enters the workforce and may
update that knowledge periodically, but may not always have accurate, current
minimum wage information for her locality.
Turning to the race and ethnicity results, it is unclear why members of certain
races or ethnicities would be less likely to have substantive legal knowledge than
members of other races. Here, “Latino/a or Hispanic” is the variable in the base
category; workers who identified as every other race/ethnicity were less likely than
Latino/a or Hispanic workers to know their minimum wage and overtime rights.
This may be because Latino/a or Hispanic workers are generally overrepresented in
the type of low-wage, front-line jobs targeted by this survey.90 Latino/a workers
may therefore have stronger networks on the job than other workers and be more
likely to educate one another about their wage and hour rights. Precisely because of
that overrepresentation, these workers may also be the recipients of targeted
know-your-rights campaigns and outreach efforts by advocacy groups, designed to
inform Latino/a workers of their substantive legal rights.91
Finally, the two variables associated with a greater likelihood of a worker’s
knowing her substantive minimum wage and overtime rights, “not a union
member” and “no legal immigration status,” appear initially to be counterintuitive.
However, it may be that workers who are union members tend to earn relatively
88. For example, research conducted in the United Kingdom has shown that employer
size was correlated with legal knowledge: smaller firms were “less likely to be
knowledgeable about employment rights.” ROBERT BLACKBURN & MARK HART, SMALL
FIRMS’ AWARENESS AND KNOWLEDGE OF INDIVIDUAL EMPLOYMENT RIGHTS (Small Business
Research Centre, Kingston Univ., Employment Relations Research Series No. 14) 70–71
(2002).
89. See Wage & Hour Div., Changes in Basic Minimum Wages in Non-Farm
Employment Under State Law: Selected Years 1968 to 2013, U.S. DEP’T OF LAB. (Apr.
2013), http://www.dol.gov/whd/state/stateMinWageHis.htm#.UNHUt3ek04U; Wage & Hour
Div., History of Federal Minimum Wage Rates Under the Fair Labor Standards Act, 1938–
2009, U.S. DEP’T OF LAB., http://www.dol.gov/whd/minwage/chart.htm#.UM_ECqDyXRQ.
90. See, e.g., Janelle Jones & John Schmitt, Low-Wage Latino Workers, CENTER FOR
ECON. POL’Y & RES. (Apr. 4, 2012, 11:15 AM), http://www.cepr.net/index.php/blogs/cepr
-blog/low-wage-latino-workers (concluding that, given their overall presence in the
population, “Latinos are indeed over-represented among low-wage workers,” defined as
those who earn $10 per hour or less).
91. For example, the U.S. Department of Labor’s Wage and Hour Division runs a
campaign that specifically targets undocumented workers with wage and hour claims. Wage
& Hour Div., We Can Help, U.S. DEP’T OF LAB., http://www.dol.gov/wecanhelp/.
202
high wages under their particular collective bargaining agreements and are
therefore less informed about universally applicable minimum wage requirements
under the law.92 In addition, undocumented workers, who, like Latino/a and
Hispanic workers, hold a disproportionate number of low-wage, front-line jobs,93
may be the recipients of the same sort of targeted “know-your-rights” outreach
efforts by advocacy groups discussed above. Undocumented workers may also be
less complacent about knowing the law than their documented counterparts, as the
law shapes their existence to a great extent, with the risk of arrest, detention, and
deportation looming large in their working lives.94
E. Procedural Legal Knowledge
Approximately 23% of workers had accurate procedural legal knowledge; a

corresponding 77% did not know where to file a workplace complaint with the
government. Interestingly, even those workers who knew how to make a
government complaint were unlikely to act on that knowledge, as 96% of workers
who did make a claim on the job made it directly to their employers, rather than to
a government agency or other third party. Moreover, the 23% result may actually
be inflated, as a worker’s knowledge of government complaint procedures was
self-reported, with no way for researchers to test the accuracy of workers’
responses.95
Table 6 lists the variables that were associated with higher and lower likelihoods
of a worker’s knowing where to file a government complaint.96
92. Lawrence Mishel, Unions, Inequality, and Faltering Middle-Class Wages, 342
ECON. POL’Y INST. ISSUE BRIEFS 1 (2012) (reporting that workers covered by a union contract
earn 13.6% higher wages than those without a union contract).
93. RANDY CAPPS, MICHAEL FIX, JEFFREY S. PASSEL, JASON OST & DAN PEREZ-LOPEZ,
URBAN INST., BRIEF NO. 4, A PROFILE OF THE LOW-WAGE IMMIGRANT WORKFORCE 1 (2003)
(finding that “[i]mmigrants compose an increasingly large share of the U.S. labor force and a
growing share of low-wage workers” and “[t]wo of every five low-wage immigrant workers
are undocumented”).
94. Shannon Gleeson, Labor Rights for All? The Role of Undocumented Immigrant
Status for Worker Claims Making, 35 LAW & SOC. INQUIRY 561, 563 (2010) (contending
that, “like the holistic experiences of race, class, and gender,” “undocumented status is
similarly a master status that is constructed by the law and that in turn shapes an individual’s
relationship to the law” (emphasis omitted) (citations omitted)).
95. Workers were asked, “Do you know where to file a complaint with the government
if you are having a problem with an employer?” See infra Appendix A.
96. Full procedural legal knowledge regression results are reported in Table 12 in
Appendix C.
203
Table 6. Procedural Legal Knowledge
Less likely to know where to file More likely to know where to file
government complaint government complaint
Female*** Age***
No legal immigration status*** Education**
Employer is not “high road”*** Black or African American**
Asian or other race**
(N = 3606)
***p 0.01; **p 0.05; *p 0.10; †p 0.15
As Table 6 shows, women, workers without legal immigration status, workers

employed by low-road employers, and those who identified as “Asian or other
race” were all less likely than those in the base category to know where to file a
workplace complaint with a government agency. On the other hand, increasing age,
education, and identification as “Black or African American” were associated with
a greater likelihood of a worker’s having accurate procedural legal knowledge.
These results generally follow from theory. Low-wage, front-line women, who
have lower social standing than their male counterparts, may have had less
exposure than men to information about government complaint procedures. And
undocumented workers, despite having relatively greater substantive legal
knowledge, might not know how to make a government complaint because of an
overarching view that government contact of any sort would risk detection and
deportation.97
The result regarding low-road employers is also consistent with intuition:
low-road employers may be less likely to educate their workers about government
complaint procedures, or may hire workers who are a priori less knowledgeable
about their procedural legal rights. In addition, workers who identified as “Asian or
other race” were less likely to have procedural legal knowledge. As with the results
in Tables 3 and 5, some cultural norm against claiming may be in effect, resulting
in workers who identify as “Asian or other race” being less likely to have legal
knowledge or engage in claiming behavior.
On the flip side, age, education, and identification as “Black or African
American” were positively associated with accurate procedural legal knowledge.
With respect to age, unlike the substantive wage and hour laws, channels for
government complaint tend not to change over time. One would therefore expect
procedural legal knowledge to increase as workers grow older and have more
exposure to workplace problems and channels for complaint. In addition, workers
with higher levels of education overall may know more about how to lodge a
government complaint than those with fewer years of schooling. Finally, workers’
97. See, e.g., Stephen Lee, Monitoring Immigration Enforcement, 53 ARIZ. L. REV.
1089, 1101–03 (2011) (discussing unauthorized immigrants’ mistrust of even friendly or
status-neutral U.S. government institutions).
204
identification as “Black or African American” may be positively correlated with

procedural legal knowledge because African Americans may have a comparatively
longer history of seeking government redress for legal wrongs than other racial or
ethnic groups.
IV. IMPLICATIONS FOR BOTTOM-UP WORKPLACE LAW ENFORCEMENT
Taken together, the results reported in Part III suggest some conclusions about
the empirics of workplace dispute generation and resolution and the functioning of
our system of bottom-up workplace law enforcement. First, though our workplace
law enforcement regime is designed to identify, surface, and resolve justiciable
problems from the bottom up, some number of actual workplace rights violations
are never even identified, or “named,” by workers. This can be seen in Figure 2,
which synthesizes the findings reported in Part III and replicates a workplace
dispute pyramid, turned on its side.
Figure 2. Workplace problems, claims, and employer responses.

* Because of a small number of observations, RDSAT was not able to generate a population
point estimate for the 42% figure, which is therefore a very rough estimate.
As the left-most column shows, though the original survey team calculated rates
of actual wage and hour law violations of up to 76%, only about 33% of workers
“named” a workplace problem. This slippage between actual and perceived
workplace problems may be attributed to workers’ lack of substantive legal
knowledge: about 59% of workers did not know their minimum wage and overtime
rights. Moreover, as the workplace problem regression results in Table 2, supra,
reflect, less politically, socially, and economically powerful and secure workers—
women workers, older workers, nonunion workers—were less likely to engage in
the process of “naming” workplace problems. “Naming” therefore represents a
point where claims are escaping from the workplace dispute pyramid, confounding
the expectations of our bottom-up system of workplace law enforcement.
205
Second, claiming is another point of escape from the pyramid. As the second
column in Figure 2 illustrates, about 33% of workers had identified workplace
problems, but 43% of those workers chose not to make claims. It is unknown from
these data whether workers chose silent loyalty or exit instead of voice. However,
median job tenure among survey respondents was 2.25 years,98 about half the
national median,99 suggesting that many workers may have chosen to exit their jobs
after having identified workplace problems.
Third, nearly all of the claims that workers made—represented by the second
column in Figure 2—were internal, made directly to the employer, while only
about 4% of workers chose to make claims in the form of a lawsuit or government
complaint. This may be attributed to workers’ general lack of procedural legal
knowledge: only 22% knew where to file a government complaint. Yet the 4%
external complaint figure suggests that even procedurally knowledgeable workers
may have been unlikely to act on their knowledge, perhaps pointing to a problem
with workers’ incentives around claiming.100
Fourth, though we do not know from these data why some workers chose to
make claims, we do know that the workers who chose not to engage in claiming
behavior did so because they feared retaliation or doubted the efficacy of claims
making. The third column in Figure 2 shows that these beliefs were rational: about
43% of workers who had made a justiciable claim experienced some form of
retaliation for their most recent claim. In addition, when asked why they chose not
to make a claim, 14% of workers reported having witnessed a coworker being
retaliated against. Moreover, as Figure 2 shows, of the workers who did make
claims about justiciable workplace problems, only 15% of employers addressed
those claims or promised to do so, while roughly 42% simply ignored the claims or
gave some other response. These statistics again suggest failures in workers’
incentives to make claims.
Thus, the results reported in Part III support our contention that the bottom-up
workplace law enforcement regime may be failing the very workers who most need
protection. Workers do not simply progress up the workplace dispute pyramid
through consecutive phases of naming, blaming, claiming, and resolution. Instead,
our analysis of the Unregulated Work Survey data suggests that workers are
shunted off the pyramid, at minimum, at the naming and claiming stages. We
contend that this drop-off occurs because low-wage, front-line workers often lack
the legal knowledge and incentives needed for bottom-up workplace law
enforcement. Compounding this problem, claiming incentives and legal knowledge
are distributed unevenly among workers. As our analysis suggests, the least
politically, economically, and socially powerful and secure workers were the least
likely to make claims, the most likely to experience retaliation, and the least likely
98. Analysis available from authors.

99. BUREAU OF LABOR STATISTICS, U.S. DEP’T OF LABOR, USDL 08-1344, EMPLOYEE
TENURE IN 2008 1 (2008), available at http://www.bls.gov/news.release/archives
/tenure_09262008.pdf (“The median number of years that wage and salary workers had been
with their current employer was 4.1 years in January 2008.”).
100. We do not know whether some workers who initially made internal claims later
chose to pursue their claims as lawsuits or government complaints.
206
to have accurate substantive and procedural legal knowledge: women, workers

without legal immigration status, and workers with low education levels.
These findings are consistent with other research on worker claiming behavior,
retaliation patterns, and legal knowledge.101 For example, David Weil and Amanda
Pyles have characterized the frequency of worker complaints to the U.S.
Department of Labor’s Wage and Hour Division and Occupational Safety and
Health Administration as “exceedingly low,” concluding:
Under [the Fair Labor Standards Act], although an average of about

29,000 workers complained each year between 2001 and 2004, when
deflated by the total number of workers, this amounts to an average of
less than 25 complaint cases for every 100,000 workers. The rate was
even lower for [complaints under the Occupational Safety and Health
Act] over the same period, averaging a mere 17 complaints for every
100,000 workers.102
Weil and Pyles also echo our findings on the distribution of claiming behavior,
noting that “workers [who] feel vulnerable to exploitation are less likely to use their
rights—these include immigrant workers, those with less education or fewer skills,
and those in smaller workplaces or in sectors prone to a high degree of informal
work arrangements.”103 Other researchers have made similar observations in the
area of employment discrimination:
[I]t is not clear that litigation protects all kinds of employees equally
well. Most employment discrimination suits are brought by employees
who have already left the job where the discrimination took place.
Further, those ex-employees who bring suit tend to come from the
ranks of managers and professionals rather than from lower-level
workers.104
101. Research on crime-victim reporting patterns is also relevant here. For example,
“[r]esearch indicates that the utilization of formal support services by victims of crime is
relatively low” and that victim characteristics like gender and ethnicity influence the
likelihood of help seeking. Michael R. McCart, Daniel W. Smith & Genelle K. Sawyer, Help
Seeking Among Victims of Crime: A Review of the Empirical Literature, 23 J. TRAUMATIC
STRESS 198, 199–200 (2010). Thanks to Leah Daigle for identifying this literature.
102. Weil & Pyles, supra note 7, at 69; see also Michele Hoyman & Lamont Stallworth,
Suit Filing by Women: An Empirical Analysis, 62 NOTRE DAME L. REV. 61, 76 (1986)
(reporting in study of suit filing among public sector workers in Illinois that “filing is a rare
event,” and “[o]nly 63 of 876 people filed suits (or charges)” regarding workplace
problems).
103. Weil & Pyles, supra note 7, at 91; see also Mitchell Langbert, Voice Asymmetries in
ERISA Litigation, 16 J. LAB. RES. 455, 462–63 (1995) (noting underrepresentation of women
as plaintiffs in ERISA lawsuits); Willborn, supra note 10, at 7 (noting that workers who sue
tend to be “considerably wealthier than the average employee”).
104. U.S. COMM’N ON THE FUTURE OF WORKER-MGMT. RELATIONS, THE DUNLOP
COMMISSION ON THE FUTURE OF WORKER-MANAGEMENT RELATIONS—FINAL REPORT 49–50
(1994), available at http://digitalcommons.ilr.cornell.edu/cgi/viewcontent.cgi?article=1004
&context=key_workplace; see also Brake, supra note 72 at 36 (“[T]he social costs of
207
With respect to retaliation, a 2012 report by the Ethics Resource Center on the
incidence of retaliation against whistleblowers (workers who uncover and report
unlawful conduct within their companies) found that 45% of employees “observe
misconduct each year,” 65% of those workers blow the whistle, and about one in
five, or 22% of those who choose to become whistleblowers, “perceives retaliation
for doing so.”105 Other studies in a variety of employment settings have found that
fear of retaliation is often the main driver of workers’ decisions not to confront the
workplace problems they identify.106
Other researchers have also found that low-wage, front-line workers routinely
misunderstand their workplace rights. A survey of low-income Latino workers in
Southern states found that “[m]ost people surveyed (about eighty percent) had no
idea how to contact government enforcement agencies such as the Department of
Labor. Many respondents did not know such agencies even exist.”107 Similarly,
only about half of day laborers in an Arizona survey knew the minimum wage,
while a survey of over two thousand contingent workers in twenty states and the
District of Columbia found that 70% of respondents did not know where to report
claiming discrimination are primarily reserved for low-power or stigmatized social

groups.”); Lauren B. Edelman, Scott R. Eliason, Virginia Mellema, Linda H. Krieger &
Catherine R. Albiston, When Organizations Rule: Judicial Deference to Institutionalized
Employment Structures, 117 AM. J. SOC. 888, 904 (2011) (“Research shows that minorities
and the poor face substantial obstacles in mobilizing legal rights ranging from psychological
and structural barriers to filing claims, to weaker legal counsel, to structural aspects of
litigation that disadvantage parties that lack social clout.” (citations omitted)); Nielsen &
Nelson, supra note 29, at 685 (collecting studies of discrimination claims-making; noting
that relatively few employees who have identified a problem with workplace discrimination
made a claim).
105. ETHICS RES. CTR., RETALIATION: WHEN WHISTLEBLOWERS BECOME VICTIMS: A
SUPPLEMENTAL REPORT OF THE 2011 NATIONAL BUSINESS ETHICS SURVEY 2 (2012),
available at http://www.ethics.org/nbes/index.php.
106. Brake, supra note 72, at 28 (collecting citations to studies) (“Most of the research on
the low level of reporting discrimination has been done in the area of sexual harassment.
Social science literature on sexual harassment abounds with findings showing that sexually
harassed women most often choose coping strategies of avoidance or denial and that the least
likely response is to report the harassment to someone in a position of authority.”); Gleeson,
supra note 66, at 13 (“A significant portion of workers [in a study of low-wage workers in
the San Francisco, California, area] attempt to negotiate with their employer before seeking
legal counsel, yet those who lack English proficiency, and those in the restaurant industry
(which represents one of the largest and least unionized employers in the private sector) are
more likely to have their attempts met with threats. Fear of job loss remains a big hurdle for
workers, who are less likely to approach a government agency directly before coming to a
law clinic if they are still employed, despite clear formal protections against retaliation.”); cf.
EILEEN APPELBAUM & RUTH MILKMAN, CTR. FOR ECON. & POLICY RESEARCH, LEAVES THAT
PAY: EMPLOYER AND WORKER EXPERIENCES WITH PAID FAMILY LEAVE IN CALIFORNIA 4–5
(2011), available at http://www.cepr.net/publications/reports/leaves-that-pay (finding that
over one-third of workers “for whom data are available were worried that if they took [paid
family leave for which they were eligible under California law], their employer would be
unhappy, that their opportunities for advancement would be affected, or that they might
actually be fired”).
107. S. POVERTY LAW CTR., UNDER SIEGE: LIFE FOR LOW-INCOME LATINOS IN THE SOUTH 6
(2009), available at http://www.splcenter.org/sites/default/files/downloads/UnderSiege.pdf.
208
workplace abuses.108 Likewise, a study of the sexual harassment policy of a large

Midwestern university reported that many administrative and clerical women
employees—relatively more powerful, secure employees compared to the
low-wage, front-line workers surveyed here—were confused about sexual
harassment law as it applied to them.109
Stephen Lee has written about the conditions that give rise to inaccurate
substantive and procedural legal knowledge among low-wage, front-line immigrant
workers in particular. As he has observed, undocumented workers may perceive
government enforcers of labor rights as enforcers of immigration law as well.110 As
a result, they may choose not to report a workplace problem to a government
agency for fear that the report will trigger immigration consequences. Immigrant
workers may also import legal knowledge from their home countries that is
inapplicable in their U.S. workplaces and derive inaccurate beliefs about their
workplace rights from relatively insular information “islands” and ethnic
networks.111
108. Mary Nell Trautner, Erin Hatton & Kelly E. Smith, What Workers Want Depends:
Legal Knowledge and the Desire for Workplace Change Among Day Laborers, 35 LAW &
POL’Y 319 (2013); ABEL VALENZUELA JR., NIK THEODORE, EDWIN MELÉNDEZ & ANA LUZ
GONZALEZ, ON THE CORNER: DAY LABOR IN THE UNITED STATES 16 tbl.7 (2006), available at
http://www.urbaneconomy.org/sites/default/files/onthecorner_daylaborinUS_39p_2006.pdf
(cited in Gleeson, supra note 94 at 562).
109. Anna-Maria Marshall, Idle Rights: Employees’ Rights Consciousness and the
Construction of Sexual Harassment Policies, 39 LAW & SOC’Y REV. 83, 115 (2005); see also
Pauline T. Kim, Bargaining with Imperfect Information: A Study of Worker Perceptions of
Legal Protection in an At-Will World, 83 CORNELL L. REV. 105, 110, 133 (1998) (quizzing
over 330 unemployed workers on knowledge of at-will employment legal rules and finding
that workers gave correct answers only 51% of the time on average); Pauline T. Kim, Norms,
Learning, and Law: Exploring the Influences on Workers’ Legal Knowledge, 1999 U. ILL. L.
REV. 447, 458 (administering the same survey in multiple additional states; finding
corresponding correct answer rates ranging from 25.2% to 40%); Weil, supra note 23, at 5
n.6 (summarizing the “literature on the lack of knowledge of statutory rights under a variety
of laws”).
110. See Lee, supra note 97, at 1101–03 (discussing unauthorized immigrants’ mistrust
of even friendly or status-neutral U.S. government institutions). This mistrust is rational,
given immigration agents’ history of such tactics as disguising themselves as labor officials
in order to locate, detain, and deport undocumented workers. See U.S. Ends Job “Safety”
Immigration Raids, ASSOCIATED PRESS, Mar. 29, 2006, available at
http://www.apnewsarchive.com/2006/U-S-Ends-Job-Safety-Immigration-Raids/id-e232
4a8e97a434063791bb95b30ff419 (describing immigration agents masquerading as
occupational health and safety workers). The mistrust persists despite a Memorandum of
Understanding between the U.S. Departments of Labor and Homeland Security that requires
immigration authorities to proceed with caution in detaining and deporting workers involved
in a labor dispute. U.S. DEP’T OF LABOR & U.S. DEP’T OF HOMELAND SEC., REVISED
MEMORANDUM OF UNDERSTANDING BETWEEN THE DEPARTMENTS OF HOMELAND SECURITY
AND LABOR CONCERNING ENFORCEMENT ACTIVITIES AT WORKSITES (2011), available at
http://www.dol.gov/_sec/media/reports/HispanicLaborForce/DHS-DOL-MOU.pdf.
111. As Jennifer Gordon, R.A. Lenhardt, and Shannon Gleeson have observed,
undocumented workers may also lack the legal consciousness, or view of themselves in
relation to the law, to make claims. Gleeson, supra note 94, at 590 (“When asked why they
209
A. The Failure of Operational Rights
The drafters of the major federal labor and employment laws were not blind to
the possibility that workers would need incentives to become the drivers of
workplace law enforcement in a bottom-up system. 112 As a result, they wrote into
the laws a set of protections and inducements that were meant to tip workers’
cost-benefit scales in the direction of claims making. We call these “operational
rights” because they operationalize or effectuate substantive statutory
protections.113
Operational rights implicitly acknowledge a major contention of this Article,
that workers’ incentives are a key determinant of claiming behavior.114 More
precisely, workers assess the size of the possible benefit they might receive as a
result of a claim, the certainty of receiving that benefit, and how costly the benefit
will be to achieve. They weigh these factors against the benefits and costs of their
chose not to come forward about long days or dangerous working conditions, many
undocumented workers repeatedly explained that to do so would simply not be characteristic
of a good worker, championing their willingness to do work others would not.” (emphasis
omitted)); Jennifer Gordon & R.A. Lenhardt, Rethinking Work and Citizenship, 55 UCLA L.
REV. 1161, 1223 (2008) (labeling undocumented workers’ view of themselves as
hardworking and uncomplaining “identity work designed to respond in some way to the
negative stereotypes and stigma associated with their particular groups”); see also Catherine
R. Albiston, Legal Consciousness and Workplace Rights, in NEW CIVIL RIGHTS RESEARCH: A
CONSTITUTIVE APPROACH 55, 56 (Benjamin Fleury-Steiner & Laura Beth Nielsen eds., 2006)
(defining legal consciousness as “the dynamic process through which actors draw on legal
discourse to construct their understanding of and relation to the social world, but that process
takes place within a social context already structured in part by the law itself”).
112. See, e.g., Mitchell v. Robert DeMario Jewelry, Inc., 361 U.S. 288, 292 (1960)
(“Congress did not seek to secure compliance with [the FLSA’s] prescribed standards
through continuing detailed federal supervision or inspection of payrolls. Rather it chose to
rely on information and complaints received from employees seeking to vindicate rights
claimed to have been denied. Plainly, effective enforcement could thus only be expected if
employees felt free to approach officials with their grievances. This ends the prohibition of
[the statute] against discharges and other discriminatory practices was designed to serve. For
it needs no argument to show that fear of economic retaliation might often operate to induce
aggrieved employees quietly to accept substandard conditions. By the proscription of
retaliatory acts set forth in [the FLSA] . . . Congress sought to foster a climate in which
compliance with the substantive provisions of the Act would be enhanced.” (citation
omitted)).
113. Alexander, supra note 20, at 386; see also Margaret H. Lemos, Special Incentives to
Sue, 95 MINN. L. REV. 782, 783 (2011) (discussing “suit boosters” or incentives offered to
plaintiffs and their attorneys to bring private litigation).
114. See also Hoyman & Stallworth, supra note 102, at 63 (“The three factors which
determine whether women will file are: 1) expected benefits which are a function of future
positions (job property rights) available in jobs without discrimination; 2) expected costs of
relief (financial and other); 3) resources which are used to file suit rather than pursue other
jobs.”); id. at 62 (“An employee calculates the cost of filing a complaint based on an
estimate of the risks associated with any action to seek relief from discrimination.”); Weil &
Pyles, supra note 7, at 91 (“The likelihood that workers exercise their rights depends on both
the benefits and the risks of doing so. . . .”).
210
other options: exit or silent loyalty. Operational rights attempt to increase the size
and certainty of the benefit and reduce the costs of achieving it.115
In the context of private lawsuits, operational rights offer payment (sometimes
double116) for violations of a plaintiff’s rights and extra payments to class
representatives.117 These supplement the benefits that would accrue anyway to
plaintiffs who are successful in a workplace lawsuit: the opportunity to work for an
employer who now complies with the law and the chance to see justice done.
Operational rights also seek to minimize the costs of claiming. They offer the
chance for reimbursed attorneys’ fees for prevailing plaintiffs, and they outlaw
retaliation and provide a private right of action for workers who have been
retaliated against.118 They also allow workers to band together in class or collective
actions to bring joint claims against their employer.119
In the context of government complaints, operational rights also include
agencies’ ability to litigate claims on workers’ behalf at no financial cost to the
worker, their capacity to collect workers’ back pay and other damages owed to
them,120 and agencies’ promise to keep complainants’ identities confidential, at
least during the initial, prelitigation phases of an investigation.121
Yet our analysis of the Unregulated Work Survey data suggests that, despite
their promise, operational rights may be largely ineffective in tipping many
workers’ cost-benefit analyses in the direction of claims. Indeed, those workers
who did make claims nearly universally did so internally, and never filed a lawsuit
115. See Weil & Pyles, supra note 7, at 91–92.

116. Under the Fair Labor Standards Act, plaintiffs may collect double their lost wages as
liquidated damages, unless the defendant shows that its violation of the FLSA was in good
faith. 29 U.S.C. §§ 216, 260 (2006).
117. Cobell v. Salazar, 679 F.3d 909, 922 (D.C. Cir. 2012) (describing class
representative incentive payments).
118. See, e.g., 42 U.S.C. § 2000e-5 (2006) (Title VII of the Civil Rights Act of 1964’s fee
shifting provision); 29 U.S.C. § 216(b) (same for the FLSA); 42 U.S.C. § 2000e-3(a) (2006)
(Title VII’s antiretaliation provision); 29 U.S.C. § 215(a)(3) (2006) (same for the FLSA); see
also Mitchell v. Robert DeMario Jewelry, Inc., 361 U.S. 288, 292 (1960) (describing the
purpose of antiretaliation protections in encouraging lawsuits).
119. 29 U.S.C. § 216(b) (allowing FLSA plaintiffs to form collective actions by
individually opting into a lawsuit); FED. R. CIV. P. 23 (allowing plaintiff class actions).
120. For a description of the U.S. Department of Labor’s Wage and Hour Division’s
ability to litigate on behalf of and collect workers’ unpaid back wages, see WEIL, supra note
3, passim. For a description of the Equal Employment Opportunity Commission’s
capabilities in these areas, see Administrative Enforcement and Litigation, U.S. EQUAL EMP.
OPPORTUNITY COMMISSION, http://www.eeoc.gov/eeoc/enforcement_litigation.cfm.
121. For example, Title VII of the Civil Rights Act of 1964 requires that the Equal
Employment Opportunity Commission keep employee charges of discrimination confidential
from the public (though not from the defendant), under penalty of a $1,000 fine or
imprisonment of up to one year. 42 U.S.C. § 2000e-5(b). See also Solis v. Seafood Peddler
of San Rafael, No. C 12-0116 PJH, 2012 U.S. Dist. LEXIS 172137, at *3–4 (N.D. Cal. Dec.
4, 2012) (noting that the U.S. Department of Labor may “use the ‘confidential informant’s
privilege’ to shield the identity of the reporting employee(s),” but that “even plaintiff
concedes that the defendants will eventually need to learn the identities of the informants
‘for purposes of trial preparation and impeachment’”).
211
or made a government complaint. For low-wage, front-line workers, the costs of

this sort of external, formal claim may be too great, and on the other side of the
scale, the benefits too small.122
First, statutory antiretaliation provisions often do not work, and workers know
this. As Emily Spieler has observed in the context of workers’ compensation,
“[r]etaliatory discharge lawsuits are a useful tool primarily for professionals,
managerial, and other upper income workers,” but not for the low-wage, front-line
workers studied here.123 Likewise, Pauline Kim points out, quoting Cynthia
Estlund:
Anti-retaliation remedies can be invoked only after the employee has

suffered discharge or discipline, and offer, at best, the possibility of an
uncertain remedy after a long delay. . . . Given the difficulties of
pursuing a retaliation claim, [Estlund] argues, “all but the most intrepid
employees will be deterred, or ‘chilled’ from speaking out in ways that
might provoke the employer’s displeasure.”
To put the point more directly, speaking out at the workplace in the
ways that the law encourages is hard.124
Indeed, in our analysis, about 43% of workers who had made claims in the
twelve months prior to the survey had been retaliated against for their most recent
claim. The workers who chose not to make claims listed retaliation-related issues (a
fear of being fired, a fear of having wages or hours cut, and the worker’s
knowledge of retaliation against other workers who had made claims) as three of
the top four reasons they decided to stay silent.
Statutory antiretaliation provisions are simply too narrow and too back-loaded to
be effective in protecting low-wage, front-line workers. The “remedy” that labor
and employment law offers workers who have been retaliated against is a chance to
bring a retaliation claim in court, which may or may not be successful, and which
may be cold (and late) comfort to a worker who has already lost her job, been
deported, or suffered other adverse employment actions. And with the exception of
the FLSA, which offers double damages, a successful retaliation plaintiff is only
restored to her position ex ante, receiving back pay and possibly reinstatement, and
often is not compensated for the costs of the retaliation itself.
122. Weil, supra note 23, at 3 (“[T]he need to keep one’s job trumps other considerations
such as being denied overtime pay or potential exposures to health hazards, let alone issues
related to supervisory treatment that may have no direct legal consequence.”).
123. Emily A. Spieler, Perpetuating Risk? Workers’ Compensation and the Persistence
of Occupational Injuries, 31 HOUS. L. REV. 119, 230 (1994); see also Langbert, supra note
103, at 459 (“Like other individual rights regulation, ERISA’s emphasis on court remedies
may prove to be of mostly symbolic value for plaintiffs most in need of protection: those
who are low-wage.”).
124. Pauline T. Kim, Electronic Privacy and Employee Speech, 87 CHI.-KENT L. REV.
901, 928 (2012) (quoting Cynthia L. Estlund, Free Speech and Due Process in the
Workplace, 71 IND. L.J. 101, 135 (1995)); see also Brake, supra note 72, at 20 (“To a large
extent, the effectiveness and very legitimacy of discrimination law turns on people’s ability
to raise concerns about discrimination without fear of retaliation.”).
212
In addition, despite some recent broadening of retaliation protections under Title

VII of the Civil Rights Act of 1964 and the Fair Labor Standards Act by the
U.S. Supreme Court,125 retaliation’s definition in workplace statutes and its
interpretation by courts tends to be quite narrow. For example, an employer’s
threats to retaliate may not themselves be actionable, even though they may be just
as effective in silencing workers as when threats are carried out.126 Moreover,
employers may effectively retaliate against one complaining worker as an example
to the rest of the workforce: that one worker may have a cause of action for
retaliation, but her coworkers, who are now scared silent, have no remedy.
For plaintiffs who are undocumented or hold work-contingent visas, the
consequences of retaliation may be especially severe. These workers risk losing not
only their job as a result of a workplace claim, but also their home, as their
employers may report them to immigration authorities.127 Because the downsides
include not only job loss, but also loss of a visa and removal from the country,
workplace claiming becomes even more risky in the transnational labor market in
which immigrants work.128
Second, while complaining to a government agency offers some increased level
of protection—a guarantee of confidentiality, at least initially—a complainant’s
identity must eventually be revealed to an employer once a case moves into
litigation or another form of dispute resolution.129 In this way, a worker’s costs may
increase over the course of the claiming process, requiring the worker to repeatedly
reassess the costs and benefits of claims making. In addition, incompetence and
processing delays may render government agencies an unattractive forum for
workers seeking effective and timely redress for workplace problems. 130
125. See, e.g., Kasten v. Saint-Gobain Performance Plastics Corp., 131 S. Ct. 1325, 1335
(2011) (holding that an oral complaint under the FLSA counted as a protected activity for
purposes of the statute’s antiretaliation provisions); Thompson v. N. Am. Stainless, LP, 131
S. Ct. 863, 867–68 (2011) (holding that Title VII’s antiretaliation provisions covered
“third-party retaliation,” or reprisals against people associated with the worker who engaged
in protected activity).
126. See, e.g., Baloch v. Kempthorne, 550 F.3d 1191, 1199 (D.C. Cir. 2008) (finding in a
Title VII case that an employer’s threats to suspend an employee did not amount to “material
adversity” that was actionable under the statute’s antiretaliation provisions).
127. See, e.g., Sure-Tan, Inc. v. NLRB, 467 U.S. 883, 886–87 (1984) (describing
employer’s retaliatory reporting of undocumented workers who had engaged in union
organizing to immigration authorities, which either deported the workers or accepted their
voluntary departure from the United States).
128. Charlotte S. Alexander, Explaining Peripheral Labor: A Poultry Industry Case
Study, 33 BERKELEY J. EMP. & LAB. L. 353, 378–79 (2012) (discussing consequences of
retaliation for immigrant workers).
129. See, e.g., Confidentiality, U.S. EQUAL EMP. OPPORTUNITY COMMISSION,
http://www.eeoc.gov/employees/confidentiality.cfm (“Once a charge is filed, the individual's
name and basic information about the allegations of discrimination will be disclosed to the
employer.”).
130. See, e.g., Highlights to U.S. GOV’T ACCOUNTABILITY OFFICE, GAO-09-458T, WAGE
AND HOUR DIVISION’S COMPLAINT INTAKE AND INVESTIGATIVE PROCESSES LEAVE LOW WAGE
WORKERS VULNERABLE TO WAGE THEFT (2009) (“[The Government Accountability
Office’s] overall assessment of the [Wage and Hour Division] complaint intake, conciliation,
213
Against these costs, the benefits of claiming appear paltry. Though back pay
may be available to a plaintiff at the end of a lawsuit, if that amount is
insufficiently large—and for plaintiffs who sue because they were paid less than
the minimum wage, back pay awards will, by definition, be quite small—then
enduring the uncertain, stressful, drawn-out process of litigation may not be worth
it.131 Moreover, our analysis reveals that workers doubt the certainty of the benefit
they might receive from claims making: the second-most frequent reason for
workers’ choices not to make claims was that workers doubted their claims would
make any difference. (On top of this, some measure of substantive and procedural
legal knowledge may be necessary for workers even to know what benefits may be
available to them at the end of a lawsuit or government complaint process.)
Changing legal standards have also made it harder for groups of workers to
bring claims as class actions and access the benefit of joint representation.132 In
addition, the Fair Labor Standards Act, Equal Pay Act, and Age Discrimination in
Employment Act have never allowed traditional class actions, in which a few
named plaintiffs represent a largely anonymous class, but instead require each
individual additional plaintiff to affirmatively opt into a case.133 As a result,
workers are increasingly forced to bring cases alone or in small groups, with each
worker’s identity exposed, setting the stage for retaliation by unscrupulous
employers.
Thus, the failure of operational rights to create appropriate incentives in either
the context of private lawsuits or government complaints means that the risk
associated with enforcement is shifted to the parties who can least bear it. The
workers who are most vulnerable to workplace rights violations in the first place
are also especially unable to bear the costs associated with claiming. As a result,
the would-be law enforcers for whom costs are particularly heavy, the low-wage,
front-line workers studied here, may choose simply to stay silent or to exit and drop
out of the workplace dispute pyramid altogether.
and investigation processes found an ineffective system that discourages wage theft
complaints.”).
131. Willborn, supra note 10, at 7 (“[V]ictims with larger potential awards will be more
likely to enforce for a number of related reasons, such as their ability to obtain counsel, the
likelihood that the award will be larger than the pecuniary and non-pecuniary costs of suit,
and the likelihood that the expected recovery will still exceed expected costs even as the
probability of winning decreases.”).
132. Myriam Gilles & Gary Friedman, After Class: Aggregate Litigation in the Wake of
AT&T Mobility v. Concepcion, 79 U. CHI. L. REV. 623 (2012) (discussing Wal-Mart v.
Dukes and AT&T Mobility Limited Liability Company v. Concepcion, which narrowed the
availability of class actions to employment plaintiffs); see also Pamela S. Karlan, Disarming
the Private Attorney General, 2003 U. ILL. L. REV. 183, 187 (discussing multiple ways that
court decisions have narrowed the ability of private individuals to bring suit to seek redress
for rights violations).
133. 29 U.S.C. § 206(d) (2006) (incorporating the FLSA’s opt-in requirement into the
Equal Pay Act, which in any case is technically a part of the FLSA); 29 U.S.C. § 216(b)
(2006) (“No employee shall be a party plaintiff to any [FLSA] action unless he gives his
consent in writing to become such a party and such consent is filed in the court in which
such action is brought.”); 29 U.S.C. § 626(d) (2006) (incorporating the FLSA’s opt-in
requirement into the Age Discrimination in Employment Act).
214
B. Self-Perpetuating Enforcement Gap
Despite this gloomy picture of workplace law enforcement, it is important to

note that workplace claiming does occur. In fact, private claims under the Fair
Labor Standards Act have risen dramatically over the past decade.134 However,
research has shown that only about 15% of eligible workers opt into, or participate
in, each case.135 This suggests a claiming gap: 85% of workers who could become
bottom-up enforcers of the FLSA and demand their unpaid minimum wage or
overtime pay choose instead not to participate in litigation.136
Likewise, workers do complain to government agencies about workplace
problems, but those complaints may not be coming from the workplaces most in
need of agency enforcement. For example, an analysis of wage and hour complaints
to the U.S. Department of Labor found an “average of less than 25 complaint cases
for every 100,000 workers” and “only modest overlap” between complaints and
employers’ compliance with Fair Labor Standards Act requirements.137 In other
words, the “market” in complaints may be failing to identify the workplaces most
in need of agency intervention.138
This produces, in the end, a self-perpetuating enforcement gap in low-wage
workplaces, in which worker claims are the driver of workplace law enforcement
but only about half of workers are likely to make claims. Employers know this, and
the unscrupulous among them feel free to underpay workers’ wages, discriminate,
134. See Table C-2A: U.S. District Courts—Civil Cases Commenced, by Nature of Suit,
During the 12-Month Periods Ending September 30, 2008 Through 2012, U.S. CTS.,
http://www.uscourts.gov/uscourts/Statistics/JudicialBusiness/2012/appendices
/C02ASep12.pdf (listing 5393 Fair Labor Standards Act cases filed in federal court in fiscal
year 2008 compared to 8152 cases filed in fiscal year 2012).
135. Charlotte S. Alexander, Would an Opt In Requirement Fix the Class Action
Settlement? Evidence from the Fair Labor Standards Act, 80 MISS. L.J. 443, 466 (2010)
(reporting opt-in rate in FLSA cases of 15%); see also Rutti v. Lojack Corp., No. SACV 06-
350 DOC (JCx), 2012 U.S. Dist. LEXIS 107677, *14–20 (C.D. Cal. July 31, 2012) (noting
problems with the FLSA opt-in structure that tend to discourage opting in by fearful,
apathetic, inertia-bound workers, and those who lack legal knowledge).
136. Moreover, most of the FLSA suits that are filed are likely brought by managers
claiming to have been illegally misclassified as exempt from overtime, not by low-wage,
front-line workers who are making minimum wage claims.
137. Weil & Pyles, supra note 7, at 69, 73.
138. For statistics on the background level of workplace rights violations in low-wage
workplaces, see BERNHARDT ET AL., supra note 35, at 2 (“We found that many employment
and labor laws are regularly and systematically violated, impacting a significant part of the
low-wage labor force in the nation’s largest cities.”); Donald M. Kerwin with Kristin
McCabe, Labor Standards Enforcement and Low-Wage Immigrants: Creating an Effective
Enforcement System, MIGRATION POL’Y INST. 21 (2011) (claiming that “there is substantial
evidence that unauthorized immigrants work at high rates in industries with long track
records of labor standards violations” and listing agriculture, meat, poultry, and fish
processing, garment assembly, and sewing as industries in which wage and hour violations
are “endemic”); Siobhán McGrath, A Survey of Literature Estimating the Prevalence of
Employment and Labor Law Violations in the U.S., BRENNAN CTR. FOR JUST. (Apr. 15,
2005), http://brennan.3cdn.net/bdeabea099b7581a26_srm6br9zf.pdf.
215
maintain unsafe working conditions, engage in unfair labor practices, and otherwise
violate workers’ rights with relative impunity.139 Workers are overdeterred from
claiming, while employers are underdeterred from complying.
V. ALTERNATIVES
This Part examines alternatives to the current bottom-up workplace law

enforcement regime. Heeding John Coffee’s admonition that “the academic
reformer has a duty to specify not only what the optimal solution is, but what the
first tentative steps are that should be taken toward it,”140 we first propose reforms
that would address the problems of lack of legal knowledge and claiming
incentives among low-wage, front-line workers. These changes would preserve the
bottom-up structure of workplace law enforcement, but would attempt to do a
better job of increasing benefits and decreasing costs than the present system of
operational rights, thereby preventing claims from escaping the workplace dispute
pyramid.
Possible reforms include minimizing bottom-up enforcers’ costs by
strengthening retaliation protections, eliminating the opt-in requirement of some
employment statutes, and establishing an arbitration system that workers could
choose to access to achieve a quick resolution of their claims. On the benefits side,
we consider sweetening the proverbial pot by lifting damages caps, offering treble
damages, and making punitive damages more readily available.141
139. As Weil and Pyles put it, “an instrinsic [sic] problem arising from the statutory
structure of workplace rights is that if left to the decision of an individual worker, the
threshold for exercise of rights lies above the threshold optimal from the workplace—and
societal—level.” Weil & Pyles, supra note 7, at 86.
140. John C. Coffee, Jr., Rescuing the Private Attorney General: Why the Model of the
Lawyer as Bounty Hunter Is Not Working, 42 MD. L. REV. 215, 286 (1983).
141. Each of these solutions focuses specifically on reforming bottom-up workplace law
enforcement rather than empowering workers generally to change conditions at work. If this
were the broader goal, then union organizing and worker centers, among other solidarity-
building strategies, would certainly need to be added to the mix. See, e.g., Jennifer Gordon,
We Make the Road by Walking: Immigrant Workers, the Workplace Project, and the
Struggle for Social Change, 30 HARV. C.R.-C.L. L. REV. 407 (1995) (describing the
operation of a worker center in advocating for worker rights and empowerment). In fact,
Allison Morantz has demonstrated that the unionization of a workplace can not only
empower workers, but also increase legal compliance, at least in the area of mine health and
safety. Alison D. Morantz, Coal Mine Safety: Do Unions Make a Difference?, 66 INDUS. &
LAB. REL. REV. 88 (2013). In addition, other reform methods harnessing technology and
market discipline are possible. For example, the U.S. Department of Labor has recently
developed a smartphone app that allows consumers to research whether a particular business
owes its employees back wages or has been cited for health and safety violations. R. Moore,
Eat Shop Sleep, ITUNES, http://itunes.apple.com/us/app/eat-shop-sleep/id465262611?mt=8
(app updated Mar. 19, 2012). Similarly, the new initiative Coworker.org allows workers to
launch online petitions and organize themselves virtually across companies and industries to
call attention to problems on the job. COWORKER.ORG, http://www.coworker.org/. Finally,
including better measures of job quality in corporate social responsibility (CSR) indices
might recruit activist consumers to the project of improving low-wage work by directing
216
Second, we consider workplace law enforcement regimes that would rely less
heavily on workers themselves to initiate claims. Here, we propose increasing
top-down, government investigative, and prosecutorial activity, and setting
government enforcement priorities such that agencies take action in the workplaces
where workers are least likely to step forward themselves. This set of proposals is
far from comprehensive, and is meant to complement calls for reform previously
suggested by other scholars and advocates.142
Of course, proposals to increase the level of workplace law enforcement beg the
question of what the desirable level of enforcement is in the first place.143 One’s
answer might depend on whether one is an employer or an employee, or whether
one is more likely to become a claimant in a workplace dispute or the target of such
a claim. Nevertheless, increased enforcement of labor and employment standards
would likely benefit both employers and employees. Employees benefit from
lawful pay, safe working conditions, and nondiscriminatory policies; employers
benefit by operating on a level playing field, without being undercut by competitors
who save money by paying lower wages, skimping on costly safety measures, or
otherwise disregarding their legal obligations. Indeed, though one might argue that
increased workplace law enforcement could increase employers’ costs and thereby
eliminate jobs, legal compliance should not be an optional cost. At the very least,
any given employer’s legal compliance should not hinge on the characteristics of
its workforce, with employers of low-wage, front-line workers particularly able to
evade detection by silencing potential claims.144
their dollars toward companies with high CSR values in the labor and employment category.
See, e.g., Leonardo Becchetti & Rocco Ciciretti, Corporate Social Responsibility and Stock
Market Performance, 19 APPLIED FIN. ECON. 1283 (2009) (describing corporate social
responsibility rankings issued by RiskMetrics-KLD, which include items around employee
relations).
142. See, e.g., NAT’L EMP’T LAW PROJECT, JUST PAY: IMPROVING WAGE AND HOUR
ENFORCEMENT AT THE UNITED STATES DEPARTMENT OF LABOR (2010), available at http://www
.nelp.org/page/-/Justice/2010/JustPayReport2010.pdf?nocdn=1 (proposing improvements to the
U.S. Department of Labor’s operations); Estlund, supra note 6, at 319 (proposing monitored
self-regulation of company compliance with labor standards); Lora Jo Foo, The Vulnerable
and Exploitable Immigrant Workforce and the Need for Strengthening Worker Protective
Legislation, 103 YALE L.J. 2179 (1994) (making a series of proposals for FLSA reform);
Daniel C. Lopez, Note, Collective Confusion: FLSA Collective Actions, Rule 23 Class
Actions, and the Rules Enabling Act, 61 HASTINGS L.J. 275, 308 (2009) (same); WEIL, supra
note 3 (proposing improvements to the U.S. Department of Labor’s operations).
143. For an attempt to identify rational criteria for enforcement of laws, see, for example,
George J. Stigler, The Optimum Enforcement of Laws, in ESSAYS IN THE ECONOMICS OF
CRIME AND PUNISHMENT 55 (Gary S. Becker & William M. Landes eds., 1974); see also
Margaret H. Lemos & Alex Stein, Strategic Enforcement, 95 MINN. L. REV. 9 (2010)
(discussing methods of optimizing law enforcement).
144. The question of whether increased workplace law enforcement would in fact
decrease employment levels is an empirical one that merits study. Similar predictions have
been made with respect to the negative effects of a minimum wage increase on employment
levels, but repeated studies have called these theories into question. See, e.g., Liana Fox,
Minimum Wage Trends: Understanding Past and Contemporary Research at 1 (Econ. Policy
Inst., Briefing Paper No. 178, 2006) (“There is a growing view among economists that the
minimum wage offers substantial benefits to low-wage workers without negative effect.
Although there are still dissenters, the best recent research has shown that the job loss
217
A. Disseminating Legal Knowledge
At first glance, the goal of disseminating legal knowledge—a necessary

condition for workplace claims—would seem to be low hanging policy fruit.
However, educating workers effectively about their substantive and procedural
legal rights is a relatively complex and difficult endeavor. Many employers,
including those covered by the Unregulated Work Survey, are already required to
post information on a variety of workers’ rights. In fact, the U.S. Department of
Labor provides eleven different “know your rights” posters, in a variety of
languages, that employers may be required to display in the workplace.145 Yet our
findings that approximately 59% of low-wage, front-line workers did not know
their minimum wage and overtime rights and 78% did not know how to file a
government complaint (both subjects of Department of Labor posters) suggest that
this method of disseminating legal knowledge is highly ineffective. Moreover,
Pauline Kim’s theory that powerful social norms can override explicit employer-
provided messages about legal rights calls into question whether the workplace is
the most appropriate place for educating workers about their legal rights and
protections.146
Instead, advocacy groups, unions, worker centers, and government agencies
should distribute materials, hang posters, give presentations, perform skits, and
make announcements in the places where workers are otherwise accustomed to
receiving information: at places of worship, at community centers, at festivals and
fairs, on the radio, and on television. These outreach efforts could specifically
reported in earlier analyses does not, in fact, occur when the minimum wage is increased.”).
145. See Poster Page: Workplace Poster Requirements for Small Businesses and Other
Employers, U.S. DEP’T OF LABOR, http://www.dol.gov/oasam/boc/osdbu/sbrefa/poster
/matrix.htm. States have also experimented with notice and posting requirements of various
specificity. See, e.g., Division of Labor Standards Enforcement, Wage Theft Protection Act,
CA.GOV (Apr. 2013), http://www.dir.ca.gov/dlse/Governor_signs_Wage_Theft_Protection
_Act_of_2011.html (describing California’s notice template requiring specific information
on rates of pay, pay schedules, workers’ compensation, and overtime); Notice of Pay Rate,
N.Y. ST. DEP’T OF LABOR, http://www.labor.ny.gov/workerprotection/laborstandards
/employer/wage-theft-prevention-act.shtm (describing the notice requirements of New
York’s Wage Theft Prevention Act, which requires disclosure of detailed pay information in
English and in the worker’s native language; making notices available in Spanish, Chinese,
Haitian Creole, Korean, Polish, or Russian).
146. Kim, Bargaining with Imperfect Information, supra note 109, at 110
(“[R]espondents overwhelmingly misunderstand the background legal rules governing the
employment relationship. More specifically, they consistently overestimate the degree of job
protection afforded by law, believing that employees have far greater rights not to be fired
without good cause than they in fact have.” (emphasis omitted)); Kim, Norms, Learning, and
Law, supra note 109, at 447 (“Contrary to the assumption commonly made by defenders of
the at-will rule, [surveys of workers in Missouri, New York, and California] indicate that
workers do not understand the default presumption [of at-will employment], but erroneously
believe that the law affords them protection akin to a just cause contract, when, in fact, they
can be dismissed at will.”); see also Gleeson, supra note 94, at 562 (“[I]t is clear that
immigrant workers, like the average low-wage worker, often lack sufficient knowledge
about the laws governing work in America. Language barriers and lack of culturally
appropriate information intensify this barrier.”).
218
target the groups of workers shown by our analysis to be less likely to have
accurate substantive and procedural legal knowledge: undocumented workers,
those employed by relatively small or low-road employers, and women workers.
The U.S. Department of Labor’s “We Can Help” initiative is an example of this
sort of campaign, incorporating language-appropriate public service
announcements and materials designed specifically to assure undocumented
workers that they have substantive workplace rights and to educate them on
enforcement procedures.147 Perhaps in this way accurate substantive and procedural
legal knowledge will find its way into the web of information that workers carry
with them into the workplace and that influences their ability and willingness to
make claims on the job.
Because lawyers themselves can also be sources of information about the law,
access to legal advice and representation may be a key component to increasing
workers’ legal knowledge. Though the phenomenon of lawyers ginning up cases by
recruiting potential plaintiffs is often portrayed negatively by the media and
academic commentators, little attention is paid to the potentially salutary function
of so-called ambulance chasers and bounty hunters.148 Lawyers who take an active
role in outreach efforts, whether to recruit clients or merely to perform a public
service, serve a valuable education and empowerment function. Increasing access
to legal services may then help address the legal knowledge deficit suggested by
the data analyzed in this Article. (Though the survey data do not answer the
question of whether lawyers were available to the respondents, lawyer usage was
likely very low, as only 4% of workers chose to take their claims to any third party
or intermediary outside the company.)149
However, legal knowledge alone, even when paired with attorney access, is not
enough to set the wheels of bottom-up workplace law enforcement into motion and
send claims up the dispute resolution pyramid. Even the most informed worker may
still choose not to make a claim, because the costs of claiming outweigh the
benefits. A brief story, recounted in a Human Rights Watch report, illustrates this
point.150 In 1999, 500 men came to North Carolina from Mexico with temporary,
H-2A “guestworker” visas to work as farm laborers. Before they arrived, staff from
a legal services organization had given them information booklets written in
Spanish that described the U.S. laws that governed their wages and working
147. See Wage & Hour Div., supra note 91.

148. See, e.g., Myriam Gilles & Gary B. Friedman, Exploding the Class Action Agency
Costs Myth: The Social Utility of Entrepreneurial Lawyers, 155 U. PA. L. REV. 103 (2006)
(describing and countering criticisms of plaintiffs’ class action lawyers).
149. Study after study has documented the gulf between civil legal needs and lawyer
availability, particularly for ex ante legal advice. See, e.g., Gillian K. Hadfield, The Cost of
Law: Promoting Access to Justice Through the Corporate Practice of Law 1 (Ctr. in Law,
Econ. & Org., Research Papers Series No. C12-16, 2012) (gathering studies and noting that
“ordinary people are largely shut out from legal assistance; the great majority of legal work
is done for corporations, organizations and governments”).
150. HUMAN RIGHTS WATCH, UNFAIR ADVANTAGE: WORKERS’ FREEDOM OF ASSOCIATION
IN THE UNITED STATES UNDER INTERNATIONAL HUMAN RIGHTS STANDARDS 218–25 (2000),
available at http://www.hrw.org/reports/pdfs/u/us/uslbr008.pdf.
219
conditions and explained the process of obtaining free legal services if their rights
were violated.
When the workers arrived in North Carolina, they attended an orientation
session. Their employer warned them that the legal services organization was their
enemy, to be avoided.151 He then told them to throw away their booklets and
distributed an employee handbook, which warned:
FLS [Farmworker Legal Services] has a hidden motive when they

approach you. They say that they are your friends and they are
concerned about your rights and well being [sic], but in reality their
motive is to destroy the program which brings you to North Carolina
legally . . . FLS discourage [sic] the growers with excessive suits which
are for the most part without merit. The history of FLS shows that the
workers who have talked with them have harmed themselves. Don’t be
fooled and allow them to take away your jobs.152
On the wall at the orientation site also hung a banner which read, in Spanish,
“Legal Services Want to Destroy the H-2A Program” and “Don’t be a puppet of
Legal Services.”153 Workers reported feeling that “if they keep [the] booklets or if
they are ever seen with one of [the] booklets, they will be fired or have serious
problems” with their employer.154
In the eyes of worker advocates, the story of the legal services booklet describes
a nearly perfectly-executed “know-your-rights” campaign.155 Workers received
language-appropriate materials that fully explained their substantive and procedural
legal rights. They even knew whom to call to get free legal representation if they
ran into a problem. Yet even full substantive and procedural legal knowledge was
likely neutralized by employer threats. Though we do not know from the report, it
would be surprising if any of the North Carolina H-2A workers actually made
claims about workplace problems that harvest season. Instead, in Hirschman’s
construction, they likely chose exit or silent loyalty over voice.
151. Id. at 218 (“[The workers’ employer] spoke at length about the Farmworker Unit of
Legal Services of North Carolina . . . He told the workers that Legal Services was their
‘enemy.’ He told the workers they should avoid Legal Services.”).
152. Id. at 218–19 (“‘Mr. Hill then held up a copy of the ‘Know Your Rights’ booklet
produced by Legal Services. He ordered workers to toss the ‘Know Your Rights’ booklet
into the trash can. . . .’ Following Hill’s admonition, workers discarded their booklets en
masse.” (emphasis in original)).
153. Id. at 221.
154. Id. at 220.
155. See, e.g., Margaret Martin Barry, A. Rachel Camp, Margaret E. Johnson, Catherine
F. Klein & Lisa V. Martin, Teaching Social Justice Lawyering: Systematically Including
Community Legal Education in Law School Clinics, 18 CLINICAL L. REV. 401, 404 (2012)
(characterizing effective community legal education, including know-your-rights campaigns,
as that which “‘encourag[es] planning on the basis of legal rights and obligations’;
‘mobiliz[es] individuals and groups to pursue their rights’; ‘facilitat[es] and strengthen[s]
community organizations’; ‘foster[s] self-help activities for which lawyers will not be
necessary’; and ‘demystif[ies] the law’”).
220
B. Minimizing the Costs
How, then, can workers be convinced to make claims? Part III has illustrated
that roughly 48% of workers in the Unregulated Work Survey who had identified
workplace problems chose not to make claims; Part IV has detailed the failure of
the current system of operational rights to maximize workers’ benefits and
minimize the costs of claiming. We suggest the following strategies to further
minimize the risks faced by would-be bottom-up workplace law enforcers.
First, Congress and the courts should strengthen existing retaliation protections
within labor and employment laws by broadening the definition of “adverse
employment action” to unambiguously include threats.156 This would acknowledge
the chilling effect that such threats, even if never executed, have on workplace
claiming behavior. It would also likely address the prevalence of non-actionable
reprisals in the workplace, as illustrated by the 28% of workers in this analysis who
made claims about justiciable workplace problems and then faced forms of
retaliation that fell outside the coverage of labor and employment law.
However, it is important to note that even legal regimes with extremely robust
anti-retaliation provisions can fail to protect claimants and encourage claiming
behavior. For example, Richard Moberly has reviewed the success of the
Sarbanes-Oxley Public Company Accounting Reform and Corporate Responsibility
Act, which concerns corporate accounting standards and is considered the “gold
standard” among retaliation protections in encouraging whistleblower claims.157 He
reports that “the percentage of employee tips actually decreased after
Sarbanes-Oxley” and that “the Act often failed to protect [whistleblowers] from
reprisals and failed to compensate them consistently for the retaliation they
suffered.”158
To make it harder for employers to identify the workers to retaliate against, then,
our second proposal is that Congress should eliminate the opt-in requirement of the
Fair Labor Standards Act, the Equal Pay Act, and the Age Discrimination in
Employment Act. Particularly for low-wage, front-line workers, the opt-in
requirement acts as a barrier to participation in litigation because it requires
workers to take the very public step of joining a lawsuit, rather than being part of
an anonymous class represented by a few named plaintiffs.159 Anonymous class
156. Charlotte S. Alexander, Anticipatory Retaliation, Threats, and the Silencing of the
Brown Collar Workforce, 50 AM. BUS. L.J. 779 (2013) (arguing that courts should recognize
a cause of action for “per se” or “facial” violations of statutory anti-retaliation provisions to
cover threats by employers designed to silence workers’ future potential protected activity).
See generally Levinson, supra note 84; Clemons, supra note 84.
157. Public Company Accounting Reform and Corporate Responsibility (Sarbanes-
Oxley) Act, 15 U.S.C. § 7201 (2012); Estlund, supra note 6, at 376 (“Sarbanes-Oxley
represents the gold standard in protection of employee whistleblowers, with both criminal
sanctions and fully compensatory private civil remedies against reprisals.”).
158. Richard Moberly, Sarbanes-Oxley’s Whistleblower Provisions: Ten Years Later, 64
S.C. L. REV. 1, 27 29 (2012) (emphasis in original) (noting that only 1.8% of employees
who made claims of retaliation won their case before an administrative body, and no
employee won for three straight years).
159. Alexander, supra note 135, at 484 (“In many ways, the opt in requirement of the
221
membership protects workers from retaliation, because employers may not be able
to identify with certainty which particular workers are participating in a lawsuit.
Moreover, workers have some measure of deniability, because they are included in
the plaintiff class not because they chose to join, but rather because they fall into
the class definition. Of course, this feature of the opt-out class action available
under Federal Rule of Civil Procedure 23 has been criticized for being
undemocratic and discouraging plaintiff autonomy.160 However, this criticism fails
to recognize the benefits of the anonymous, automatic nature of the traditional class
action for low-wage, front-line plaintiffs who are extremely susceptible to
retaliation.161 Thus, all areas of employment law should be made consistent with the
great majority of federal statutes, in which plaintiffs are permitted to use the class
action device of Federal Rule of Civil Procedure 23 and receive the protection of
relative anonymity within the class.162
Third, we propose an alternative method of claim resolution. For justiciable
workplace claims that are not already covered by an administrative exhaustion
requirement, the U.S. Department of Labor should establish an optional, one-way
binding arbitration system, modeled on a program in place at The Coca-Cola
Company, that would give workers a forum for speedy claim resolution.163 This
FLSA seems to eliminate any possibility of collective action, making the FLSA enforceable
only by individuals or very small groups. The opt in requirement may, therefore, undermine
the project of law enforcement generally, dissuading plaintiffs from acting collectively as
private attorneys general to enforce the FLSA’s minimum wage and overtime mandates.”).
At the very least, courts should permit language in the opt-in notice that advises
undocumented workers that their immigration status is not relevant to the case. See, e.g.,
Enriquez v. Cherry Hill Market Corp., Case No. 10-CV-5616-FB-MDG, (E.D.N.Y. Feb. 10,
2012), available at http://docs.justia.com/cases/federal/district-courts/new-york/nyedce
/1:2010cv05616/312066/36/0.pdf?ts=1328966650 (order approving FLSA opt-in notice
language informing potential opt-in plaintiffs that their immigration status is irrelevant).
160. See, e.g., Alexander, supra note 135, at 484–86 (summarizing criticisms of Fed. R.
Civ. P. 23 class actions).
161. Research is being conducted on the benefits of anonymity in the analogous
whistleblower context and one study has found that “anonymity, trust, and risk are highly
salient in the [whistle-blowing reporting system] context.” Paul Benjamin Lowry, Gregory
D. Moody, Dennis F. Galletta & Anthony Vance, The Drivers in the Use of Online
Whistle-Blowing Reporting Systems, 30 J. MGMT. INFO. SYS. 153, 154 (2013); see also Weil,
supra note 23, at 13 (noting “over two decades of evidence that shows that workers are more
likely to exercise rights given the presence of a collective workplace actor”).
162. We acknowledge that class members’ identities are often exposed at the end of a
lawsuit when they must collect their portion of settlement or judgment proceeds. However,
the risk of retaliation at this point is relatively low, as the litigation has already concluded.
We also acknowledge that courts have restricted the Rule 23 class action as a vehicle for
collective complaint, but we nevertheless think it preferable to the current opt-in regime in
place in the Fair Labor Standards Act, Equal Pay Act, and Age Discrimination in
Employment Act.
163. Suzette M. Malveaux, Is It the “Real Thing”? How Coke’s One-Way Binding
Arbitration May Bridge the Divide Between Litigation and Arbitration, 2009 J. DISP. RESOL.
77, 78 (2009) (“This unilateral arrangement—voluntarily adopted by the Coca-Cola
Company following its historic employment discrimination class action settlement—is a
groundbreaking and novel approach to promoting arbitration, while also protecting employee
choice and access to the court system. One-way binding arbitration also offers employers an
222
system would not increase the size or certainty of the benefits of claiming, but
would reduce the costs associated with trying to achieve those benefits. The
program could be funded by civil monetary penalties collected by the Department
from law-breaking employers, and arbitration would be free for workers who could
demonstrate financial need, much as in forma pauperis determinations are made.
The parties could choose to use this system at their discretion, and, as in Coca-
Cola’s system, the arbitrator’s decision would be binding on the employer only.164
Particularly for workers who have been fired in retaliation for claims making, this
system would allow them to collect back pay and/or return to their jobs more
quickly than in a typical lawsuit, without requiring them to relinquish their right to
file their claim in court. Employers, too, would benefit from the speed and
efficiency of arbitration.165 (Employees with group claims, however, would still
have to pursue their claims in court, as arbitration would not be well-suited for
resolutions of collective or class actions.166)
Each of these proposals seeks to lower the costs of claims making by increasing
retaliation protections, facilitating group action by plaintiffs, and speeding up the
process of workplace dispute resolution. However, as Steven Willborn has
observed, minimizing worker costs may not be effective in and of itself in
encouraging claims; increasing the benefits may also be necessary.167
C. Maximizing the Benefits
Workers might also be incentivized to act if there were greater benefits

associated with making claims. Here, Congress could lift the cap on Title VII
opportunity to forge new ground. Companies can enjoy all of the benefits of arbitration—
such as efficiency, privacy, cost savings and litigation avoidance—while bolstering
workplace relations that may enhance profitability.”); see also David Sherwyn, J. Bruce
Tracey & Zev J. Eigen, In Defense of Mandatory Arbitration of Employment Disputes:
Saving the Baby, Tossing Out the Bath Water, and Constructing a New Sink in the Process, 2
U. PA. J. LAB. & EMP. L. 73 (1999) (arguing that employees should reconsider arbitration as a
tool for speedy resolution of disputes).
164. Malveaux, supra note 163, at 78 79 (describing Coca-Cola’s “one-way binding
arbitration” program).
165. Id. at 119–20 (describing benefits to employers of one-way binding arbitration
system including speed, efficiency, improved employee perception, and advantage in future
litigation).
166. Id. at 134–35 (“Finally, while [Coca-Cola’s] Solutions program may be appropriate
for addressing individual employee work-related matters, it is not appropriate for resolving
class actions or other actions alleging systemic corporate misbehavior. Because of the
important role the courts play in resolving cases with larger societal implications, employees
must be permitted to challenge an employer’s pattern or practice of conduct without having
to go through an internal company resolution process. The judiciary’s unique public function
and procedural attributes make it especially well-suited for addressing large-scale
employment issues.”) (emphasis in original).
167. Willborn, supra note 10, at 12–13 (“[E]ven with retaliation protection, individuals
are more likely to come forward in situations in which there is little risk of retaliation than in
situations where they have to rely on retaliation protections. In turn, this implies that positive
inducements, rather than protection against negative consequences, may be necessary to
produce adequate private enforcement in some circumstances.”).
223
damages,168 make punitive damages more readily available to plaintiffs, especially

in cases of retaliation,169 and allow plaintiffs to collect treble damages, perhaps
upon a showing of outrageous conduct and/or recidivism by the defendant.170
Federal statutes in areas other than labor and employment provide models for
offering increased incentives to expose wrongdoing. For example, the Internal
Revenue Service entices people to report tax fraud by offering between 15% and
30% of the back taxes and other funds ultimately collected by the Service;171 the
Dodd-Frank Wall Street Reform and Consumer Protection Act offers
whistleblowers between 10% and 30% of monies recovered as a result of tips about
insider trading and other violations of securities law;172 the qui tam provision of the
False Claims Act allows people who report fraud against the federal government to
collect up to 30% of recovered damages.173
However, in practice, such incentives may not be very effective at encouraging
claims making. A study of federal statutory whistleblower incentives reveals that,
“although [individual] rewards under existing whistleblower programs may be
substantial, general use of the programs is not high.”174 This, again, points to the
168. 42 U.S.C. § 1981a(b)(3) (2006) (imposing caps on Title VII damages of between
$50,000 and $300,000 depending on employer size).
169. At present, punitive damages are generally not available for retaliation claims under
the FLSA, the National Labor Relations Act, the Family and Medical Leave Act, and the
Americans with Disabilities Act, and are available, but with statutory caps, under Title VII. 42
U.S.C. § 1981a(b)(1) (“A complaining party [under Title VII] may recover punitive damages
under this section against a respondent . . . if the complaining party demonstrates that the
respondent engaged in a discriminatory practice or discriminatory practices with malice or with
reckless indifference to the federally protected rights of an aggrieved individual.”); see also 42
U.S.C. § 1981a(b)(3) (capping Title VII damages at between $50,000 and $300,000 depending
on employer size); Jackson v. Estelle’s Place, LLC, 391 Fed. Appx. 239, 247 (4th Cir. 2010)
(noting that punitive damages “are not available for claims under the FLSA”); Alvarado v.
Cajun Operating Co., 588 F.3d 1261, 1265, 1270 (9th Cir. 2009) (noting courts’ “divergent
approaches” to availability of punitive damages for ADA retaliation claims and holding that
punitive damages are not available); Farrell v. Tri-Cnty. Metro. Transp. Dist., 530 F.3d 1023,
1025 (9th Cir. 2008) (“It is well-settled that the FMLA, by its terms, ‘only provides for
compensatory damages and not punitive damages.’” (quoting Liu v. Amway Corp., 347 F.3d
1125, 1133 n.6 (9th Cir. 2003))); NLRB v. Pepsi Cola Bottling Co. of Fayetteville, Inc., 258
F.3d 305, 314 (4th Cir. 2001) (holding that “an explicitly punitive method of calculation [of
backpay] is contrary to the purposes of the [National Labor Relations] Act”).
170. The Fair Labor Standards Act already allows plaintiffs to collect double or liquidated
damages. 29 U.S.C. § 260 (2006). The Sherman Anti-Trust Act and Racketeer Influenced and
Corrupt Organizations Act both allow collection of treble damages. Trusts in Restraint of Trade
Illegal, 15 U.S.C. § 1 (2006); Racketeer Influenced and Corrupt Organizations Act, 18 U.S.C. §
1961 (2006).
171. Karie Davis-Nozemack & Sarah Webber, Paying the IRS Whistleblower: A Critical
Analysis of Collected Proceeds, 32 VA. TAX REV. 77, 86 (2012).
172. Consumer Financial Protection Act of 2010, Pub. L. No. 111-203, 124 Stat. 1376,
1957 (2010).
173. Civil Actions for False Claims, 31 U.S.C. § 3730(c)–(d) (2006).
174. Jonathan L. Awner & Denise Dickins, Will There Be Whistleblowers?, 34 REG. 36, 39
(2011). At the time of writing, the Dodd-Frank Act was still fairly new, and so the success of its
whistleblower incentive provisions are still relatively untested. In addition, in 2012, an IRS
224
tricky problem of cost and benefit: if existing reward programs targeting

sophisticated potential whistleblowers in the financial sector have been relatively
unsuccessful in incentivizing claims,175 it is likely that the analog in the labor and
employment law context—offering much higher damages awards as an increased
potential benefit for plaintiffs—may also fail to encourage low-wage, front-line
workers to exercise voice.176
D. Increasing Government Enforcement
Each of the foregoing reforms would preserve the fundamental bottom-up nature
of workplace law enforcement, while attempting to improve workers’ legal
knowledge and recalibrate their incentive structures such that low-wage, front-line
workers might be enabled to act as enforcers of their workplace rights. However,
given the fundamental problems with bottom-up enforcement suggested by the
Unregulated Work Survey data, such tinkering around the margins might be
insufficient, akin to rearranging the deck chairs on the Titanic.
A more fundamental reform would advocate shifting more enforcement from the
bottom to the top, increasing government investigation and oversight rather than
relying so heavily on workers themselves to be the drivers of enforcement activity.
After all, government enforcement agents are insulated from many of the risks
faced by worker-claimants: their salaries are paid regardless of the outcome of a
lawsuit, and they are unaffected by defendant retaliation.177
Michael Waterstone has made just such a proposal in the disability law context.
He argues that certain case types—failure to hire and public accessibility—are
particularly well suited to public enforcement.178 In the labor and employment
context, Steven Willborn has suggested case types that would be appropriate for
increased government enforcement, in areas where private litigation leaves gaps:
“low-damages cases, large cases which involve heavy litigation-related
investments, and violations not producing wage losses.”179 Likewise, David Weil
has advocated that the U.S. Department of Labor’s Wage and Hour Division focus
its investigative resources on “fissured” industries, in which businesses employ
tipster received an unprecedented award of $104 million. Paul Sullivan, The Price
Whistle-Blowers Pay for Secrets, N.Y. TIMES, Sept. 22, 2012, at B1. It remains to be seen
whether this highly-publicized award will increase the level of IRS whistleblower activity.
175. See, e.g., Sullivan, supra note 174 (describing the costs that whistleblowers pay, “If
you look at the field of whistle-blowers, you see a high degree of bankruptcies. You may
find yourself unemployable. Home foreclosures, divorce, suicide and depression all go with
this territory.”).
176. In addition, research has shown that, among the options of rewards, liabilities,
duties, and protections, the availability of rewards for reporting illegal conduct may trigger
“less reporting than merely offering protection or establishing a duty.” Feldman & Lobel,
supra note 21, at 1155.
177. See Yaniv, supra note 5, at 352 (“While the expected gain from complaining is
obvious, there is a serious risk involved: losing one’s job. However, when a worker’s
complaint is placed anonymously (or when the enforcement agency is bound not to disclose
his identity), there is no reason to expect a personal reaction on the part of his employer.”).
178. Michael Waterstone, A New Vision of Public Enforcement, 92 MINN. L. REV. 434,
437 (2007).
179. Willborn, supra note 10, at 13.
225
high numbers of subcontractors and other contingent workers and disclaim any
legal responsibility for wages and working conditions.180
To these lists we add cases in which workers would be particularly unlikely to
complain on their own, where workers have low levels of substantive and
procedural legal knowledge, and where workers are highly susceptible to
retaliation. Indeed, the U.S. Equal Employment Opportunity Commission’s most
recent Strategic Enforcement Plan (SEP) names as an enforcement priority
“[p]rotecting [i]mmigrant, [m]igrant, and [o]ther [v]ulnerable [w]orkers,” and calls
for a focus on “[i]ssues affecting workers who may lack an awareness of their legal
protections, or who may be reluctant or unable to exercise their rights.”181
In implementing goals such as these, government enforcers could develop a sort
of “vulnerability index,” taking into account workers’ wages, immigration status,
unionization levels, and education levels within a particular industry.182 Industries
with high vulnerability index scores could be prioritized in agencies’ strategic
enforcement plans.183 Likewise, agencies might prioritize retaliation cases and send
a message to unscrupulous employers that they will not be permitted to silence
worker claims through threatened and actual reprisals.
180. WEIL, supra note 3, at 18–26.

181. U.S. EQUAL EMP’T OPPORTUNITY COMM’N, STRATEGIC ENFORCEMENT PLAN: FY
2013–2016, 1, 8 (2012), available at http://www.eeoc.gov/eeoc/plan/sep.cfm. The EEOC has
also explored using the commissioner’s charge mechanism strategically, considering “factors
such as current economic conditions, demographic data, relevant labor markets, industry
data, underserved areas or populations and charge data.” LESLIE E. SILVERMAN ET AL.,
SYSTEMIC TASK FORCE REPORT: TO THE CHAIR OF THE EQUAL EMPLOYMENT OPPORTUNITY
COMMISSION 15 (2006), available at http://www.eeoc.gov/eeoc/task_reports/systemic.cfm.
182. Martha Albertson Fineman’s extensive work on vulnerability provides a theoretical
backdrop for creation of a vulnerability index; Kerry Rittich’s work on employee
vulnerability is informative as well. See, e.g., Martha Albertson Fineman, The Vulnerable
Subject: Anchoring Equality in the Human Condition, 20 YALE J.L. & FEMINISM 1, 20–21
(2008) (“The vulnerability inquiry examines the ways in which societal resources are
channeled to see if the result is to privilege and protect some while tolerating the
disadvantage and vulnerability of others. This focus on the structuring of societal institutions
reflects the fact that the state has an affirmative obligation not to privilege any group of
citizens over others and to actively structure conditions for equality.”); KERRY RITTICH, LAW
COMM’N OF CAN., VULNERABILITY AT WORK: LEGAL AND POLICY ISSUES IN THE NEW
ECONOMY 3 (2004) (“Vulnerability and insecurity at work can arise from: 1) the distribution
of risks, costs, benefits and powers among workers and employers; 2) the (in)capacity of
workers to conform to or perform according to workplace rules and norms; 3) the allocation
of work among workers, including unpaid work; 4) (in)access to resources; 5)
discrimination, either directly on the basis of a particular characteristics or grounds or
through their intersectional operation or indirectly, because of the connection between these
grounds and the factors listed above.”).
183. The U.S. Department of Labor’s public communications speak of advocacy on
behalf of “vulnerable workers” such as homeless people, but do not appear to define that
term. See, e.g., Vulnerable Workers, U.S DEP’T LAB. NEWS BRIEF, Oct. 18, 2012, available
at http://www.dol.gov/_sec/newsletter/2012/20121018.pdf (reporting on “roundtable
discussions” between U.S. Secretary of Labor Hilda Solis and “representatives of vulnerable
communities in Seattle, New Orleans and San Francisco”).
226
NAMING, BLAMING, AND JUST PLAIN GIVING UP
By Matthew Fritz-Mauer
Abstract
Millions of Americans experience civil justice problems each year, but most never assert
their rights in the justice system. This “access to justice” crisis has long received attention from
the legal profession, commentators, and bar associations. This analysis largely emphasizes unmet
legal needs, and pushes to expand the availability of legal assistance. This narrow approach does
not meaningfully examine understandings of “access” and “justice,” interrogate the problem
within particular socio-legal contexts, or consider more radical policy solutions.
This Article does. It analyzes access to justice issues among low-wage workers suffering
wage theft in Washington, D.C., in the wake of that city enacting one of the strongest workers’
rights regimes in the country. Through original qualitative and quantitative data, this Article
evaluates whether these legal changes have improved access to justice for the working poor.
While there has been some success on the margins, there has not been a meaningful shift,
because the government is largely inactive and low-income people face many non-legal barriers
to bringing claims. Using Denver, Colorado as a case study, this Article concludes that rights
enforcers must fundamentally change their approach to tangibly improve access to justice.
Electronic copy available at: https://ssrn.com/abstract=4029827
227
Introduction
In 2012, “Lisa” started working as a counselor at The Giving House, a non-profit in the
District of Columbia that provides housing and support for at-risk teenagers.1 Lisa was supposed
to work forty hours per week for a set salary, and earn overtime beyond that. She also got to live
in an on-site apartment.
Problem was, Lisa’s employer did not pay her overtime. Lisa brought this issue up with
her supervisors, but they “disagreed” and she “didn’t push it because I was glad to have a job.”
Things reached a tipping point in early 2016, however. A blizzard struck, and most
employees could not make it in to work. “[T]he management team made a decision,” Lisa
explains, “instead of bringing in other people to relieve the [counselors], they just wanted [us] to
work eight hours on, eight hours off, come back and work eight hours on, eight hours off.” This
went on for three days. It was exhausting, especially with Lisa’s regular schedule the rest of the
week. Although The Giving House promised extra compensation, none arrived.
Lisa finally reached her limit, and filed a wage claim with the D.C. Department of
Employment Services (DOES). “I knew it was a risk,” she says, “but I just thought that that was
the right thing to do, that they can’t get away with that, they can’t treat people like that and think
it’s okay.”
Eight months later, DOES determined Lisa’s employer had underpaid her by about $800.
This seemed low, though, and a review of the agency’s math revealed significant errors. Lisa
appealed, and the agency set up a mediation a few months later.
1
Throughout this Article, I have given pseudonyms to low-wage workers and their employers. The other
participants in this research typically chose to identify themselves. Where I use a person’s first name only, it is a
pseudonym; where I use a first and last name, it is not.
228
But in the meantime, Lisa lost the will to fight. During this long, slow process, Lisa had
left her job to work for the Clinton presidential campaign, but the non-profit was still her
landlord. While out of town, The Giving House suddenly and illegally evicted Lisa. Her former
employer moved her things out, and when Clinton lost the 2016 presidential race, Lisa became
unemployed. She returned to D.C. without a job, without a home, and without many of her
belongings. Despite her misgivings, Lisa gave up and accepted DOES’ flawed determination.
She was living in her car. She needed the money.
At first, Lisa tried to get DOES to take action over her eviction, which she thought was
retaliatory. “The hard part, the trauma,” she explains, “is the retaliation, is violating my rights.”
But almost two years later, the agency had done nothing. “I would push it,” Lisa says, “then I
would stand back because I didn’t want to go through the pain.” Ultimately, she gave up.
“There’s no driving force from their end,” she says, speaking through tears. “It’s always had to
be me pushing, and that’s why . . . I’m where I am now.”
* * *
What happened to Lisa was wage theft, which occurs whenever an employer denies a
worker the wages or benefits to which they are legally entitled (Bobo 2014; Fritz-Mauer 2021).
Wage theft comes in many forms, and includes minimum wage and overtime violations,
misclassifying employees as independent contractors, refusing to pay for all hours worked, and
denying workers paid sick days or breaks (Bobo 2014; Fritz-Mauer 2021). In broad strokes,
employers steal tens of billions of dollars from millions of low-wage workers every year
(Bernhardt et al. 2009; Eastern Research Group 2014; Galvin 2016; Cooper & Kroeger 2017).
Lisa’s story is upsetting and brazen, but not unique.
229
But despite wage theft’s prevalence and the fact that it is often plainly illegal, the
overwhelming majority goes unpunished. This is not because wage theft is a minor crime with
little significance to its victims, though. Minimum wage violations alone are estimated to cost
more than two million individuals more than $15 billion per year, outstripping the costs of all
other property crimes combined (Cooper & Kroeger 2017). Independent contractor
misclassification suppresses wages and increases individual taxes, and denies federal, state, and
local governments billions per year in revenue (Carré 2015; Ruckelshaus & Gao, 2017).
Individuals lose thousands per year; affected communities, especially immigrant communities
and communities of color, are deprived of billions (Cooper & Kroeger 2017). “In many low-
wage industries, wage theft . . . is pervasive and endemic” (Galvin 2016), a massive, widespread
social problem (Fritz-Mauer 2021).
That most wage theft goes unreported and unaddressed should not be surprising. Most
potential legal disputes are never resolved through the civil justice system (Albiston et al. 2014).
For decades, scholars have discussed the crisis of “access to justice,” which is typically
understood to be a “crisis of unmet legal need” (Sandefur 2019; Albiston & Sandefur 2013).
Because the problem is predominately defined as a lack of legal representation, the typical
solution is to expand legal services (Sandefur 2019).
In the past decade, wage theft has received a significant amount of scholarly and legal
attention. In the face of an inactive federal government (see Hacker & Pierson 2010), local
policymakers have passed a slate of laws geared towards finding and remedying wage theft
(Doussard & Gamal 2016; Galvin 2016). These reforms are largely designed to empower
individuals to bring claims (Fritz-Mauer 2021). They create new causes of action and
230
administrative processes; enhance civil and criminal sanctions; and increase the value of cases
through expanded damages and mandatory attorneys’ fees (Galvin 2016; Fritz-Mauer 2021).
This backdrop raises a number of social, legal, and empirical questions. Why is it that so
many poor people do not bring wage claims? How well do different reforms work at sweeping
more cases into the civil justice system? How effective are various interventions at helping
victims find substantive justice?
This Article examines access to justice through the lens and within the context of the
problem of wage theft among the working poor. Using original research to build on existing data,
I explore whether, when, why, and how people are likely to dispute their wage theft, and the
efficacy of the District of Columbia’s anti-wage theft reforms.
In doing so, I take on three tasks. First, I join a growing body of scholars in arguing that
the issue of access to justice has been superficially analyzed, examined in ways that are largely
divorced from the lives and experiences of the people who have the least ability to assert their
rights. Researchers must consider access to justice issues within particular social, political, and
legal frameworks, and account for the perspectives of those most affected by this crisis.
Second, I challenge the overriding understanding of access to justice. If there is a crisis,
the thinking goes, then the solution is to make it easier for people to go out and find legal
assistance. But this mindset conflates “justice” with legal representation or processes, and it still
places the burden on individuals who often lack the time, energy, knowledge, and wherewithal to
do that. When it comes to wage theft, the phrase “access to justice” should refer to something
more.
Third, in analyzing the practical barriers that low-wage workers face in claims-making, I
also evaluate whether and to what extent existing, popular reforms succeed at helping poor
231
people. Using D.C. and Denver as case studies, I conclude that the access to justice crisis, as it
applies here, will never be meaningfully addressed unless and until governments embrace
strategies that bring access and justice to low-income people.
Part I elaborates on access to justice in more detail. Part II lays out the District’s anti-
wage theft enforcement regime, and Part III discusses my study methodology. Part IV details my
findings regarding the scope and impact of wage theft among the low-wage workers who
experience it, and explains why so many people fail to take any kind of formal legal action to
enforce their rights. Part V ties this research back to access to justice, and pushes for a more
contextualized, broader understanding of that term. Finally, Part VI proposes new, more radical,
and more effective reforms.
I. Access to Justice (or if that isn’t available, some kind of legal process)
Every year, there are more than one hundred million events that could be addressed
through the civil justice system. These problems affect between one-half and two-thirds of
Americans (Sandefur 2019) and run the gamut of issues, including bankruptcy, eviction, torts,
domestic violence, and wage theft. Most are never addressed in any formal way, however, in
what many call the “access to justice crisis.”
Concerns over access to justice flow, in part, from social science research on disputing.
Scholars have paid a great deal of attention to the process by which people recognize an injury
(name), assign responsibility (blame), and take action to obtain a remedy (claim) (e.g., Albiston
et al. 2014; Calavita & Jenness 2015). In the 1970s, research revealed that there are many more
potential legal disputes than previously believed, but the vast majority are never acted upon
(Albiston et al. 2014).
232
These findings remain true today. Most legal problems never receive any attention from a
lawyer (Sandefur 2019) or even a quasi-legal professional. Traditionally, lawyers, bar
associations, and other legal actors have called this a “crisis of unmet legal need.” This
understanding frames the problem as being about the provision of legal services, implicitly
defining “access” as a person’s ability to obtain legal representation, and “justice” as a lawful
outcome that results from formal processes (Sandefur 2019).
According to this logic, people with disputes have trouble finding justice because they
cannot afford a lawyer or legal advice. Solutions, then, have largely focused on increasing the
ability of people to obtain representation and guidance (Pruitt & Showman 2014; Sandefur
2019). To that end, the legal profession encourages pro bono work; jurisdictions create legal
advice clinics; and reforms attempt to the smooth the path to claiming by making cases more
attractive to attorneys and making it easier to file claims in court or with administrative agencies
(see Sandefur 2019; Fritz-Mauer 2021).
This framing is simple, sounds reasonable, and fits neatly within the idealization of the
adversarial system. But these solutions prioritize access to legal processes, not justice, and
emphasize individualized, case-based interventions to civil problems (see MacDowell 2014).
This approach has been criticized as a “thin” conception of access to justice that does not
adequately grapple with or respond to what it means to be a person with a legal problem (e.g.,
Albiston & Sandefur 2013; Pruitt & Showman, 2014; Sandefur 2019).
There is much more to the story of why people do not try to formally enforce their rights,
and lately the topic of access to justice is experiencing something of a renaissance (see Albiston
& Sandefur 2013). In particular, social scientists challenge the conflation of “justice” with “legal
services,” arguing that the problem is multifaceted and complex, and its solutions must be as
233
well (Pruitt & Showman 2014; Sandefur 2019). This is particularly true when it comes to
marginalized people, who face “daunting barriers” to identifying and taking action over legal
problems (Calavita & Jenness 2013, p. 51; Sandefur 2008).
The real problem is that the thin conception overlooks key hurdles to both seeking and
providing legal services. It inherently treats civil justice problems as personal ones. But many
civil justice problems are complex, pervasive, and far-reaching enough to impact entire
communities. Beyond that, many problems are caused by larger social and economic forces.
One-off claims cannot shift such tides (see Fritz-Mauer 2021).
The thin understanding also relies on individuals to identify their own legal needs and
take action, even though people frequently do not interpret their experiences through a legal lens
(Pruitt & Showman 2014). Some, for example, think of their problems as flowing from
impersonal or abstract forces, like bad luck or fate (e.g., Levitsky 2008: 568-69). Such people are
likely to internalize their problems, rather than assign blame elsewhere after a consideration of
legal rights.
Finally, the thin approach does not recognize or attempt to redress the underlying causes
of a problem. For example, somebody experiencing eviction clearly has a need for legal
assistance. But there are many interconnecting issues in that situation that exacerbate each other:
the minimum wage is too low, America’s social safety net is inadequate, there are only weak
tenant protections, or the person grew up in poverty and is disadvantaged. None of these
problems are solved through individual representation, although that representation may be
crucial in the moment.
Wage theft illustrates these critiques. It is endemic among low-wage workers,
disproportionately affecting immigrants, women, and non-whites (Bernhardt et al. 2009; Cooper
234
& Kroeger 2017). It is both an economic crime and a way to reinforce ancient social hierarchies,
and each year, it artificially forces millions of people into greater economic precarity (Cooper &
Kroeger 2017). It is best understood as a social problem that cannot be adequately resolved
through one-off acts of legal advocacy. In a best-case scenario, representation of low-wage
victims of wage theft earns them unpaid wages and damages, while delivering an unpleasant
experience to their employers. But while that work might—might—also benefit their coworkers,
it does nothing for those who do not go to a lawyer for help. Nor does it offer a solution to the
other evils of poverty.
In light of these deficiencies, some have argued for a “thicker” approach to access to
justice that recognizes that “[n]ot all barriers [to justice] are in the judicial system.” (Rhode
2009: 873). Some of these barriers reflect economic disadvantage more broadly, or race- and
sex-based hierarchies that discourage certain kinds of people from asserting their rights.
Social science research bears out these arguments. Members of disadvantaged groups,
including lower-income people, those lacking formal education, and racial minorities perceive
fewer legal problems and make fewer claims (Sandefur 2008; see Best & Andreasen 1976).
“[C]laiming behavior varies inversely with socioeconomic status,” (Albiston et al. 2014: 114),
with higher-income people being more likely to take legal action (Sandefur 2008). Although
wealth is “clearly part of the story . . . an explanation based on cost, resources, and stakes is
insufficient to explain the full pattern of class differences” (Sandefur 2008: 347). Other factors
related to social rank are also key to understanding why lower-class people are less likely to
dispute their problems, including “a sense of entitlement or feelings of powerlessness” (Sandefur
2008: 347).
235
The thicker approach analyzes legal problems within the context of the practical realities
of peoples’ lives; works with communities to understand what problems exist and what “justice”
looks like; and considers alternative policies that could create longer-lasting systemic change
(Pruitt & Showman 2014; Albiston & Sandefur 2013). Leading socio-legal scholars have urged
researchers to deemphasize lawyers and legal services, and instead closely analyze how effective
existing policies are, and “how current definitions and understandings of access to justice may
blind policy makers to more radical, but potentially more effective, solutions.” (Albiston &
Sandefur 2013, p. 103).
I adopt the thicker conception here. The debate on access to justice has been too myopic,
too focused on the need for representation and what that means for outcomes in cases involving
poverty and inequality. To really address the “crisis in access to justice, we need a broader
understanding of what access to justice means and what the current lack of access entails,”
(Sandefur 2019) and that requires deeper analyses of problems as they exist in specific social,
legal, and regulatory environments.
To that end, this Article examines access to justice and disputing among the working
poor in the District of Columbia who experience wage theft. The District recently and
dramatically overhauled its anti-wage theft laws and policies. The overall effect of these reforms,
on paper, was to create multiple points of entry to the justice system, encourage and empower
more people to bring more claims, entice lawyers to take on more cases, and make it easier to
collect wages and damages.
II. Best (Written) Practices in the District of Columbia
10
236
In 2018, OxFam analyzed all fifty states and D.C., ranking them based on how well their
laws protect workers’ rights (Rose et al. 2018). The District came out on top overall, scoring
second-highest on wage laws and worker protection policies, and highest on union organizing
rights. D.C., declared OxFam’s researchers, is “the national model for worker rights and
protections” (Rose et al. 2018).
This was no accident, but followed years of conscious organizing and reform. In writing,
the city provides workers’ rights that are significantly stronger than those secured by federal law
or many other jurisdictions. For example, it has a $15 minimum wage that is tied to inflation
(D.C. Code § 32-1003(a)), paid family leave (D.C. Code. § 32-541.01 et seq.), and paid sick days
(D.C. Code 32-531.02).
Most significantly for this research, though, is D.C.’s Wage Theft Prevention Act
(WTPA). In 2014, the City Council unanimously approved this sweeping legislation, which
followed a two-year organizing campaign by community advocates, plaintiffs’ lawyers, and low-
wage workers.
On paper, the WTPA is one of the strongest anti-wage theft laws in the country. As one
advocate explains, in drafting the legislation they “wanted to be as robust as possible . . . .
[O]ften, what we [did] is grab what looked like the best language, mixing and matching from
other jurisdictions,” and then trying to “go stronger than any other example.”
The end result was a strong and thorough omnibus bill that was designed, from the
perspective of the workers’ rights community, to be the best response possible to wage theft. To
that end, the WTPA and its amendments changed D.C.’s enforcement regime in three primary
ways.
11
237
A. Penalty Enhancement
First, the Act dramatically increased civil and criminal penalties for wage theft and
retaliation. Even a negligent violation of wage laws is now a misdemeanor (D.C. Code § 32-
1307). On the civil side, failing to pay a worker all earned wages carries the threat of significant
fines and empowers workers to recover treble damages (D.C. Code §§ 32-1308, 1308.01). While
a number of jurisdictions have adopted treble damages, which have been shown to reduce
minimum wage violation rates (Galvin 2016), the District is unique in imposing treble damages
in addition to unpaid wages.
B. Liability and Collections
Second, the WTPA made it easier for workers to recover wages and penalties. Even when
workers win, it can be very difficult to collect on judgments (see Cho et al., 2013; Weil, 2014a).
One analysis, for example, found that more than eighty percent of successful claimants never
collected even a single dollar (Cho et al. 2013). Some employers are “judgment-proof” and have
no real assets; others are simply resistant, and use a variety of strategies to avoid paying (Weil
2014a; Cho et al. 2013).
The WTPA attempts to address these problems. It expands what’s commonly called
“joint and several” or “up the chain” liability in the construction and temporary employment
industries (D.C. Code §§ 32-1012(c), (f)(1)). This change is key to addressing the problem of
“fissuring,” which refers to the fact that many companies have reduced the scope of their
operations, electing to streamline their business by focusing on a relatively small number of core
tasks to cut costs (Weil 2014a). Organizations outsource many jobs, and use subcontracting to
obtain workers (Weil 2014b). Frequently, such workers are not considered employees of the
12
238
organization that is higher up the employment food chain. The WTPA simply states that general
contractors and subcontractors, and temporary staffing firms and their clients, share civil
responsibility for acts of wage theft (D.C. Code § 32-1303(5), (6)).
C. Expanded Claims-Making
Third, the WTPA attempted to expand access to justice by creating multiple points of
entry into the justice system, empowering three distinct groups: the private bar; the D.C. Office
of the Attorney General (OAG); and the D.C. Department of Employment Services (DOES).
The private bar is incentivized to take on more cases, both because they are now worth
more and because the WTPA provides for relatively high attorneys’ fees. While many laws
entitle successful plaintiffs to recover fees and costs, D.C. now applies the “Salazar Matrix”
(Salazar v. District of Columbia, 123 F. Supp. 2d 8 (D.D.C. 2008)) to wage cases. This matrix—
normally reserved for complex litigation—sets fees that better reflect the prevailing market rate
for litigating complex federal cases in the District. As of June 1, 2021, an attorney with between
one and three years of experience has a statutory hourly rate of $381, while lawyers with twenty
or more years of experience are entitled to $919 (Laffey Matrix, 2022). According to plaintiffs’
attorneys, this has expanded the range of cases they may take.
Second, the D.C. OAG now has independent enforcement authority (D.C. Code § 32-
1306(b)(2)). It has taken a strong, public stance against wage theft, hiring dedicated attorneys to
prosecute civil wage theft and issuing frequent press releases (e.g., OAG 2017). In addition, it
13
239
has the exclusive power to bring criminal prosecutions against employers (D.C. Code. § 32-
1306(a)(1)).
Finally, the WTPA created sweeping changes at DOES, which enforces the District’s
wage and hour laws. Generally, it enhanced the agency’s authority by ratcheting up penalties and
empowering DOES to order the exact same remedies as a court, including treble damages,
attorneys’ fees, and reinstatement. It also established that wage claims may be finally adjudicated
through the District’s existing administrative process (see D.C. Code § 32-1308.01).
The goal of such processes is to simplify dispute resolution and improve access to justice.
Administrative hearings are supposed to be a gentler, speedier alternative to the slow,
procedurally-laden court system. In theory, a worker should be able to navigate the agency’s
process, which takes no more than six months (see D.C. Code § 1308.01).
This is the legal, social, and regulatory framework in which the District’s working poor
experience and respond to wage theft. On paper, the end result of these reforms was a
comprehensive system of enforcement, designed to serve workers’ needs, respond quickly and
efficiently to complaints, proactively find wage theft, and create rippling effects throughout
industries.
In practice, the District has fallen far short of its promise.
III. Study Methodology
Original qualitative and quantitative research forms the backbone of this analysis. In
2018 and 2019, I conducted in-depth, semi-structured interviews with approximately sixty
people belonging to four groups: low-wage workers (n=33), workers’; rights activists (n=12),
employment lawyers (n=10), and employers (n<5).
14
240
I recruited participants in three ways. First, I partnered with a non-profit workers’ rights
organization called the Washington Lawyers’ Committee for Civil Rights Under Law (the
“WLC”). The WLC is the key anti-wage theft legal services organization in the District of
Columbia. It advocates for legislation, engages in impact litigation, and runs a workers’ rights
clinic that engages with more than 1200 people per year. I recruited participants from among the
clinic’s clients, and leveraged the WLC’s broader network to recruit employment lawyers and
workers’ rights activists. To expand beyond this network, I also used snowball sampling. Finally,
I sent cold e-mails to employment lawyers who practice in the District.
There is no agreed-upon scholarly definition of what constitutes a “low-wage worker.”
Some researchers include all workers in the bottom 20% of earners (Cooper & Kroeger 2017);
others focus on those making less than 1.5x the local minimum wage (Galvin 2016); still others
on those who earn no more than two-thirds the local median wage (Boushey et al. 2007). But
some of these approaches do not adequately account for the cost of living of the place where
people live and work. Especially in a place as expensive as the District of Columbia, this should
be accounted for. For this study, workers qualified if they made no more than 300% of the
federal poverty line, as shown in Table 1:
For this research, it is important to understand how low-wage workers think about their
income and earnings. The goal was to understand and analyze the perspectives of people who are
economically insecure. How a person experiences and responds to legal problems turns heavily
on their socioeconomic status (see Sandefur 2008; Albiston et al. 2014), and how concerned they
are about the consequences of taking action. This threshold allowed me to do that. While these
earnings would allow people in some areas to live decent, secure lives, they do not in the
15
241
District. This is highlighted by research from the D.C. City Council, which found that in 2018 “a
single adult would need to earn approximately $36,988 per year to afford their basic needs”
(D.C. Council 2018).
The workers I interviewed reflect the diversity of the District and the demographics of its
low-wage workforce, which is dominated by people of color. Of thirty-three worker-participants,
twenty-eight are racial minorities: eighteen identified as African American; five as Latinx; two
as Asian; and three as mixed. Four identified as white, and one participant declined to self-
identify.
Participants ranged in age from twenty-two to sixty-six, with an average age of forty-
seven. Many of these people have had long careers in low-wage industries, but not all. A few
were recent entrants into the world of financial struggle, reflecting the fact that poverty can arise
suddenly.
The average household size of participants was about two. Estimating an average twelve-
month household income for this sample is difficult. Rarely did somebody give a definite
number for annual earnings. Many were unable to, because they did not know. Instead, I often
heard ranges, or vague, generalized responses, like “Very little” or “I get unemployment.” Some
of these people had not kept track; others were ashamed.
Most, however, were nowhere close to meeting this project’s income thresholds. Out of
thirty-three people, only four thought they were within $5000 of the applicable income ceiling.
Most workers reported making close to minimum wage. Often, their low wages were
compounded by erratic work schedules, a scarcity of work, and wage theft. Others had the ability
to earn a decent hourly rate, but faced the same income-depressing issues.
16
242
I likely under-sampled Latinx people, who comprise approximately 11% of the District’s
population (Census 2021), but are more heavily concentrated in low-wage jobs (DOES 2017).
This was due to my own language limitations, although I hired a translator. The harm of this flaw
is blunted by the fact that a significant amount of research on the low-wage workforce has
focused heavily on Spanish-language workers and their experiences with wage theft (e.g.,
Valenzuela et al. 2006; Fussell 2011; Gleeson 2015, 2016; Galemba 2022).
I supplemented this qualitative data with existing quantitative research and some original
data about enforcement metrics and case outcomes from DOES.
IV. The Harms of Wage Theft and Barriers to Claiming
The majority of research on wage theft has emphasized its prevalence and economic
impact, both to individuals and in the aggregate. This research is important, but it is also
incomplete. What is missing is a strong representation of the viewpoints and experiences of low-
wage workers themselves, because by and large, workers’ experiences with and responses to
wage theft have not been closely examined. Consequently, researchers have not detailed the full
measure of wage theft’s harms. This research fills this gap.
A. Cascading economic harms, emotional damage, and health consequences.
Wage theft causes much more damage, both to individuals and to communities, than
research has yet reported. Low-wage workers report significant harm extending far beyond the
immediate consequences of being underpaid or unpaid. Wage theft frequently causes cascading
economic harms among those with little money to spare, as well as emotional and mental
distress.
17
243
Cascading Economic Consequences: The most frequent victims of wage theft are the
working poor, both because they are frequently deemed ripe for exploitation (e.g., Fussell 2011;
Fritz-Mauer 2021) and because they tend to work in industries that skirt workplace laws and
regulations (Weil 2014(a); Bernhardt et al. 2009). By definition, poor people do not have a lot of
money. They live paycheck to paycheck, and have little or nothing in the way of savings (Federal
Reserve 2019). As a result, delayed, reduced, or just non-existent paychecks inflame economic
insecurity. Low-wage workers struggle to pay for their basic needs, and frequently report sinking
into debt as they fall behind on bills or seek loans.
Frequently, when they are short of cash, workers must pick and choose among their needs
and obligations. The problem is, not one basic life commitment is easily cast aside. Marion, who
worked approximately twenty hours of unpaid overtime a week, chose to go hungry rather than
speak up because she desperately needed to be on good terms with her employer. Miranda
prioritized rent and food over utilities. “My budget is week by week,” says Kira, who has
frequently had trouble getting her supervisor to input her paid sick leave on time. “[I]f you don’t
pay me for a day, and I have to get back to work, that means maybe the babysitter doesn’t get
paid. Maybe we don’t eat the way we should that week.”
Others choose to fall behind on rent, favoring more immediate needs. Cora owed her
landlord approximately $10,000. Carol was in debt much, much more than that, while Lydia was
a month behind and scrambling to catch up. Ameen and Naomi were being evicted. Electricity,
food, rent, transportation, medicine, water—these things are all necessary to a person’s
existence, they all cost money, and those who control access demand regular payments. If a
person falls too far behind, or takes too long to pay, there are often further consequences. Credit
agencies become notified of delinquencies, which makes it harder to find an apartment, open a
18
244
bank account, or take out a loan. Some employers even pull credit reports, further locking people
into cycles of poverty.
Emotional Distress: When we first speak, James cracks a joke about the problems he has
been having at Market, a large, unionized grocery store chain. Some are the kinds of
degradations that tend to accompany low-wage work, like verbal abuse. But his real issues stem
from the fact that his employer does not pay him for all of his hours. Several times, his
supervisors have corrected the “error” and paid him after he complained, but his problems also
escalated over time. His managers bullied him, denied him bereavement leave, transferred him to
another store and, at one point, illegally suspended him.2 “I got to laugh to keep from crying,” he
says. As he becomes more comfortable, he goes into detail about what impact these experiences
have on him:
I get at the point where I want to kill myself sometimes, you know? So I know some
of the trigger things, I take medicine with depression now, and it, it just . . . . It’s
rough, you know what I’m saying? It’s really rough. It’s already hard enough, and
to have to fight just to get the money that . . . it’s rough.
Although the stress, anger, and sadness that people experience is related to the strain of
economic insecurity, there is more to it than that. Many low-wage workers do not experience
wage theft as just a financial crime. For most, wage theft represents an unjust and deeply unfair
exercise of power. It signals to a person that their rights do not matter, that work is not worth
what it should be, and that they are not, either.
Sixteen of the thirty-three workers interviewed for this project expressed feeling
depressed or ashamed, both over their wage theft and their perceived inability to stop it. Harriet,
2
During our interview, James showed me a decision from the National Labor Relations Board finding his
suspension unlawful.
19
245
who was repeatedly paid late despite her protests, felt “useless.” Cora, who attempted suicide,
felt “worthless” because she earned less than minimum wage, while Carl—who also
contemplates suicide—described his treatment as “cruel.” Other researchers have reported that
low-wage victims of wage theft describe feeling “very depressed,” “upset and powerless,” and
“really sad” (Cho et al. 2013, pp. 4, 8).
Not every victim expressed strong negative feelings, but those who did not typically were
unaware of their rights violations. Most workers, however, generally knew of their own
mistreatment, and had strong reactions.
Wage Theft as a Public Health Problem: Finally, wage theft also has negative physical
health consequences. Researchers estimate that minimum wage violations alone artificially place
hundreds of thousands of working people below the poverty line (Eastern Research Group 2014;
Cooper & Kroeger 2017), while millions more have less money than they otherwise would and
should.
“The relationship between poverty and health has long been documented in the literature
of medicine” (Mansfield & Novick 2012, p. 1). Poor people are more likely to experience a
variety of chronic health conditions, including diabetes, heart disease, obesity, stress, headaches,
and ear infections (Mansfield & Novick 2012; Conway 2016). They are also significantly less
likely to treat these conditions, because the United States does not have universal healthcare.
Most of the uninsured are poor, and they “tend to forego preventative care and to wait until an
illness is severe before seeking medical care” (Simon 2012: 1).
The relationship between poverty and health is cyclical: Poverty leads to health problems,
which contribute to poverty (Conway 2016). Relatively speaking, poor people experience bad
housing (Desmond 2016), suffer more pollution, lack ready access to healthy food, and live more
20
246
sedentary lifestyles (Mansfield & Novick 2012). Sick people cannot work, and must also fit
treatment costs into a shoestring budget. These negative effects persist across generations and
radiate into the larger community (Conway 2016).
This story about poverty’s negative effects on health is illustrated by this research. Cora
described feeling lethargic, “tight, and pain-ridden,” as though “rigamortis had set in,” while
Olivia lost sleep and struggled to concentrate. Maylin’s doctors placed her on a home-monitoring
system for her blood pressure, but she laughs while explaining that “[s]ince I’ve been terminated
my blood pressure’s been good.” Ameen eventually became sick and had to go to the hospital,
where he received a $1500 bill. For a man struggling to avoid homelessness, this was an
impossible burden.
Recognizing these problems, some public health experts have drawn attention to “a small
but growing body of literature suggesting that wage theft is a serious and largely overlooked
problem with important public health implications,” because “by causing or exacerbating low
income or income insecurity, wage theft is . . . [a problem with] which health departments and
other public health organizations must increasingly be concerned” (Minkler et al. 2014).
To summarize, wage theft causes significant harms that go beyond the dollars and cents
of it all. Given these facts, and the District’s slate of favorable laws, one might expect that low-
wage workers are empowered and practically capable of asserting their rights. This is far from
the case, however.
B. Naming, Blaming, and Barriers to Claiming
Each year, tens of millions of Americans experience problems that could, potentially,
become legal disputes. The vast majority do not, however. In studying disputing, scholars focus
21
247
on a three-step process known as naming, blaming, and claiming, which refers to individuals 1)
identifying a problem, 2) assigning responsibility, and 3) seeking redress (Albiston et al. 2014).
An extensive body of research on disputing shows that the ability and willingness to
name a problem, blame somebody for it, and make a claim are socially and culturally patterned
(Calavita & Jenness 2015). As Rebecca Sandefur (2008: 342) explains, “studies frequently find
that people often do not think of their justiciable problems as having any connection to law or
rights and also reveal powerful influences of local social context on how disputes are understood
and pursued.” The relevant “local social context” that affects how a person thinks about and
responds to their experiences is made up of many factors, including individual people,
organizations and institutions, widespread cultural beliefs, and a person’s own background and
upbringing. For example, scholars point to “agents of transformation”—including lawyers,
friends, and government institutions—as playing an important role in managing and shaping a
person’s feelings and expectations (Albiston 2005; Edelman et al. 1993). The messages that
these agents of transformation provide influence how people think about an issue, including
whether a person decides that a given event is a legal problem, and whether and how that
problem should be addressed (Albiston et al. 2014; Gleeson 2015).
Likewise, the type of potential dispute matters. For example, people more often name,
blame, and claim torts, while victims of discrimination rarely confront their offender, even
informally (Miller & Sarat 1980: 561, 563-64). Rights knowledge, in particular, is key (Sandefur
2008: 342; Trautner et al. 2013: 320; Albiston 2005). The more prevalent a particular area of law
and rights, the more likely people are to dispute (e.g., Albiston 2005; Levitsky 2008).
Knowledge alone, however, is not enough (Sandefur 2008).
22
248
In many ways, the District seems like an ideal place for low-wage workers to dispute
their wage theft. There are powerful laws, an active workers’ rights community which engages in
significant community outreach and education, and government actors broadcast that wage theft
is morally wrong and deserves a serious, comprehensive response. For example, when the OAG
sued an employer named Power Design for allegedly misclassifying hundreds of workers, the
Attorney General himself publicly stated that “Power Design cheated hundreds of District
workers out of their hard-earned wages and stripped them of their legal rights” (OAG 2018). City
Councilmember Elissa Silverman called for Power Design to be removed from a government
apprenticeship program, arguing that “[w]e shouldn’t be putting our tax dollars toward
companies that are robbing workers of their wages” (Thebault 2018). It is not that Power Design
was “skirting its responsibilities” or engaging in some other euphemistic act. In these leaders’
words, wage theft is starkly immoral, indistinguishable from traditional crimes of theft.
Nevertheless, workers overwhelmingly do not pursue formal legal claims. Naming and
blaming are not the largest points of failure here. With few exceptions, workers reported being
aware that their employers had stolen from them, and understood these acts to be both wrong and
illegal. They also universally assigned responsibility to their employers.
These findings may seem surprising in light of research establishing that many
Americans have a poor understanding of their workplace rights (see, e.g., Bernhardt et al. 2009;
Alexander & Prasad 2014: 1072; Trautner et al. 2013). For example, almost sixty-percent of the
people in a landmark 2009 study of wage theft “misunderstood their minimum wage and
overtime rights” (Alexander & Prasad 2014: 1072).
To be fair, workers frequently misstate the details of their rights. But, the majority are
largely aware of their wage theft, either when it occurs or shortly thereafter. This might sound
23
249
counterintuitive, but the explanation lies in the fact that wage theft is often nothing more than a
casual repudiation of a person’s most basic entitlements. “Wage theft” is an umbrella term, but
the most common forms violate our country’s oldest and most rudimentary workplace laws.
These laws are not only longstanding, they line up with our own cultural and moral
understandings of how things should be.
So even as the details of these laws escape many people, the working poor still
comprehend the outlines of their own mistreatment when it involves repudiations of their basic
rights. With this background in mind, why don’t workers pursue the legal claims that they are so
obviously entitled to make, over acts they describe as harmful and upsetting? The problems
overwhelmingly occur at the claiming stage, and are the result of larger social and economic
forces that prevent people from coming forward to assert their rights.
i. The Frequency and Futility of Informal Claims
Claiming can involve formal or informal processes, with everything ranging from talking
to a supervisor to filing a lawsuit. For a variety of reasons, low-wage workers who believe they
have valid legal disputes are extremely reluctant to take formal action. In a study of wage theft
among more than 4,300 low-wage workers, fewer than 3% of those who perceived a workplace
rights violation attempted formal legal action (Alexander & Prasad 2014: 1085-86, 1089).
Instead, the overwhelming majority of attempt to assert their rights do so informally by speaking
to employers, supervisors, or Human Resources departments (Alexander & Prasad 2014).
These approaches are often unsatisfactory. Workers find themselves at a significant
power disadvantage relative to their employers, who may themselves be ignorant of the law or
just unwilling to follow it. For instance, Myron’s supervisor told him point blank that he would
24
250
not pay him for a full shift because Myron had not been “authorized” to work that day, a
confusing claim without any legal justification. “I was just shocked, stung, and I ain’t know what
to say,” Myron says, “so I just left it alone.” Like Myron, workers who speak up and are denied
rarely escalate. And even where informal action resolves a particular issue, it may fail to address
underlying problems. Miranda, Kira, and Tonya complained repeatedly about their wage theft to
their supervisors, but found themselves locked into a cycle of rights violations and arguments.
As I explain, workers fail to aggressively pursue their claims for a variety of overlapping
reasons: they fear retaliation, especially the kind that leads to a loss of income; do not know how
to file a formal complaint; and express a significant lack of faith in both their ability to navigate
the system and its ability to protect them. What these barriers have in common is that they are
inextricably tied to workers’ low position on our society’s social and economic hierarchies. This
actual and perceived lack of power generates fear and insecurity, and marks silence as a safe, if
deeply aggravating path.
ii. Barrier to Claiming: Fear of Retaliation
Perhaps the biggest explanation for why low-wage workers do not attempt to assert their
workplace rights is because they are afraid of retaliation. For people who live in poverty or just
on the edge of it, loss of income is an overriding concern. Such retaliation can take many forms,
but workers are acutely aware of both the danger of losing their jobs and the risk that their
employers will contact immigration authorities (Gleeson 2015; see also Fussell 2011). Even a
bad employer can be better than the prospect of debt, hunger, and homelessness.
Earl, for example, is a journeyman plumber who has experienced regular wage theft, and
who believes that employers will not follow the law until somebody holds them accountable.
25
251
When asked whether he has ever tried to do that, he laughs. “[I]f you step forward, what are the
odds? You’re outnumbered with all the contractors here, and then you’ll be scapegoated. You
won’t be able to get no work. You won’t be able to find a job. You won’t be able to get nothing.
So how do you step forward?”
Low-wage jobs are ubiquitous. Many believe that if a worker loses their job, they can
find another. But even those who work “typical” low-wage jobs say that insecurity buys their
silence. Maria, who cleans office buildings, explains that “for the employers it’s easy, they can
just take the job away from you, but for you it’s really difficult to get a job.”
Retaliation need not involve job loss to deter action, though. Supervisors have a variety
of strategies. Harriet believes that her manager held up her paychecks to punish her for filing a
grievance, which had devastating effects. James’ managers unlawfully suspended him, Ruben’s
cut his hours, and Miranda told me that once, she got the distinct feeling that her manager falsely
told a potential employer that Miranda had been in a physical fight at work.
Fears of job loss are worsened by qualities that make people especially vulnerable to the
vagaries of an uncertain job market. Being older, having dependents, or having a criminal record
heightens insecurity and increases the need to put up with a bad situation. “There’s the fear,
especially when you’re a single parent,” says Agda, a bartender. “You’re responsible for
someone else. It’s like this huge weight.”
“It’s hard [to find a job] because I have a record now,” Marion explains, “and sometimes
I don’t get [interviews], sometimes I get interviews that go well, sometimes they be like ‘I can’t
hire you because of your background.’” Criminal records are major barriers to employment,
especially for African Americans, who have more difficulty finding work in the low-wage labor
market than equally-qualified whites (Pager 2003; Pager et al. 2009). James felt stuck at his job
26
252
because “I have a criminal record and. . . . I’m guessing that deters a lot of employers from
hiring me, especially with me being an older gentlemen.” Lydia echoes James’ thoughts about
age: “So that’s the fear now, with me being 50. Who wants to invest?”
Perhaps the biggest point of vulnerability is immigration status. Undocumented people
have a difficult time finding work due to laws requiring employers to verify citizenship status
(Galemba 2022). As a result, undocumented workers have limited options, and often must
choose to either 1) present false proof of citizenship, which carries its own risks, or 2) accept
low-grade, wage theft-riddled jobs in the underground economy. Immigration concerns are a
major reason why undocumented workers do not confront their employers over exploitation
(Gleeson 2015; Fussell 2011).
In short, many low-wage workers will never take formal legal action over their wage
theft because the potential consequences are too great. Wrapped up in this fear of retaliation is a
sense that nobody will protect them, at least not before they suffer significant harm. This is a
rational belief: In 2019, DOES took an average of 42 days to even contact employers accused of
unlawful retaliation, and an average of 117 days to “resolve” (that is, close out) those complaints.
iii. Barrier to Claiming: Concerns over navigating the system
Low-wage workers also report that they do not formally pursue their claims because they
feel unprepared to navigate the legal system. There are two related aspects to this: workers lack
procedural knowledge, and are intimidated by the prospect of legal action.
Many do not understand the basics of how to bring a lawsuit or file an administrative
claim (see Alexander & Prasad 2014). The process is confusing and nebulous, and many do not
know how much authority administrative agencies have. Even those who are familiar with DOES
27
253
express confusion. Cora and Earl, for example, have both participated in a variety of DOES-run
programs designed to help residents find work. In fact, Cora is one of the few people I
interviewed who spoke well of DOES. When I ask whether she might file a wage claim, she says
“I don’t know, and I wouldn’t know how to do it.”
Procedural hurdles may be worsened by agencies themselves. At the time of this
research, for example, DOES insisted wage claims be notarized. Nothing in the law required this,
but the agency nevertheless refused to process un-notarized complaints. Similarly, DOES would
also refuse workers who self-identified as independent contractors on the wage claim form,
notwithstanding the fact that misclassification is common (Carré 2015) and whether somebody is
an employee is a complex legal question.
Beyond the lack of procedural knowledge, some low-wage workers are reluctant to take
formal action because they do not believe they can successfully navigate the system. It is not a
lack of knowledge that holds them back, necessarily, but a desire to avoid a process that is
intimidating, confusing, time-consuming, and emotionally draining. “I just didn’t want any more
added stress,” says Nate, a caterer whose employer denied him overtime pay.
These facts are well-known among activists and employment lawyers, and are
particularly relevant in the District. “Just getting DOES to answer the phone is difficult,” says
Shlomo Katz, a management-side attorney who sometimes takes pro bono wage theft cases.
“[Y]ou call a switchboard, you get switched around, eventually you leave a message, and they
may or may not call you back.” Jason Rathod and Nick Migliaccio, two experienced lawyers,
explain that at first they were excited at the prospect of working with DOES, but quickly gave
up. “[T]he statute created this administrative body that was supposed to be able to hear claims,”
says Jason, “and we’ve tried to avail ourselves of it and just ran into a brick wall.”
28
254
iv. Barrier to Claiming: Lack of power, lack of faith
Finally, low-wage people do not make formal legal claims because they simply do not
think it will do any good. There are, again, two components: first, workers understand and are
intimidated by the stark power imbalance between them and their employers; and second, many
lack faith in the District government and do not believe it will serve them.
“Power” refers not only to the dominant ideas that permeate and structure society, but
also to the dynamics of interactions between people. It may be understood as both “the outcome
of social transactions” and the ability of a person to “achieve foreseen and intended effects” in a
dispute (Ewick & Silbey 2003: 1333). In D.C., workers and activists brought a significant
amount of power to bear to broadly define wage theft as a gross injustice that must be remedied,
but the effects of these efforts have been limited. Wage theft is common largely because
employers and workers alike understand that low-wage workers lack the power to really assert
themselves.
Jaime Cruz, who serves on the board of an immigrant and workers’ rights organization
called Trabajados Unidos de Washington D.C., knows well how these dynamics force silence,
especially among undocumented people. “[Employers] know the circumstances and they play
with it,” he says, “And then they don’t pay [workers] and they know that generally speaking
[those] laborers whose wages they stole is not going to raise any issue with it, is not going to go
to the [government].” Ultimately, this power imbalance is closely tied to the fear that speaking
up will only make things worse. “[Your rights are] in your handbook, but if you try to pursue it
then you’re retaliated on,” says Lydia.
29
255
The lack of power workers feel is closely related to the belief many of them have that the
government will not protect them, whether they speak up or not. There is an acceptance that, at
best, government processes are slow and inefficient; at worst, that agencies are designed to favor
the Haves over the Have-Nots (see Galanter 1974). This general belief is not limited to those
with personal experience. Even Susanna, who has not experienced wage theft, says that “[t]he
government typically takes too long in the process of addressing [people’s] concerns and issues
that they’re having, and it’s not taken care of in a timely manner.”
In her book Precarious Claims, Professor Shannon Gleeson illustrates how pursuing an
administrative claim can be a deeply upsetting and emotional experience, and that these
problems are not unique to the District:
While we tend to think of legal mobilization as an empowering process, I found

that many workers focused on what was missing or lost throughout the ordeal. A
lack of access to key brokers and experts, issues with language and communication
during appointments and proceedings, and the time, monetary, and emotional costs
of endless wrangling all weighed heavily on them . . . . For some, the lesson to be
learned was to always speak up and defend your rights; for many others it was to
learn to remain quiet, as there is much to be lost (Gleeson 2016: 127).
Low-wage workers are not alone in feeling cynical and skeptical. Activists and
employment lawyers also take a dim view of the District management of wage theft, with DOES
being a particular culprit. “A worker, honest[ly], if they have a wage claim they have a better
chance of waiting for the next semester at a law school to open up so [a law school clinic] can
potentially take their case than they do at having DOES review their claim,” says Allen
Cardenas, who coordinated a workers’ rights clinic, because “[t]he law school will be more
responsive.” Union organizer Ray Diaz is less polite: “[At DOES], they don’t give a shit about
investigating, really resolving issues for people.”
30
256
v. Overcoming barriers to claiming
These barriers to claiming prevent the vast majority of wage theft’s victims from taking
legal action. But, some still do. For most people to pursue action, though, they must reach a
breaking point where the offensiveness of their exploitation outweighs the possible costs of
asserting rights. When wage theft is small or infrequent it is easier to ignore; similarly, when a
given approach to claiming carries with it relatively few life or job-related consequences,
workers are more likely to use it. This is why many are willing to confront their supervisors, but
reluctant to escalate their dispute.
Low-wage workers think through their options and make rational decisions. Almost all
understand their wage theft to be morally and legally wrong; almost all believe that taking action
is not worth it. When weighing costs and benefits to formal claiming, the scale tilts towards
silence. It would be wrong to conclude from this discussion, though, that many low-wage
workers are willing to put up with some amount of exploitation. It is more accurate to say that
low-wage workers are required to tolerate mistreatment and abuse because they do not have
another choice.
For the overwhelming majority of low-wage workers, if they want to take formal action,
their only real option is to go to the government for help. Lawyers are prohibitively expensive
and take few cases. Self-help is often no help. And, because the D.C. Office of the Attorney
General has very limited resources, and the federal Department of Labor does not enforce D.C.’s
more protective laws, most people can really only file a complaint with DOES. Given wage
theft’s harms and the dynamics of how it plays out, what does this mean for the issue of access to
justice?
31
257
V. Access, Justice, and the Failures of Wage Theft Enforcement
Socio-legal scholar Rebecca Sandefur has argued that “America has a massive crisis,
with two parts.” (Sandefur 2019). First, access to justice is restricted: only some kinds of people
and some types of problems get addressed. Others do not, especially where there is a power
imbalance, as with many of the issues that affect poor people. Second, access to justice is
systematically unequal: some groups, especially whites and wealthier people, are more likely to
have it than others (Sandefur 2019; see also Albiston et al. 2014).
This Article highlights these two problems within the particular context of wage theft
among poor people. Notwithstanding D.C.’s progressive laws, very little wage theft gets
reported, discovered, and addressed.
Recent quantitative data from DOES underscores this fact. In 2017, researchers estimated
that approximately 39,500 workers are paid less than minimum wage each year (DOES 2017).
But DOES receives fewer than 100 minimum wage complaints per year (DOES 2019), and is not
proactive about finding violations (Fritz-Mauer 2021). The key agency simply does not learn
about more than 99.5% of minimum wage violations. If these numbers seem stark, they
shouldn’t. Complaint rates are, and always have been, extremely low (see Alexander & Prasad
2014: 1085-86, 1089; Fritz-Mauer 2021: 136). In fact, the industries with the highest violation
rates have low complaint rates (Weil & Pyles 2005).
Reliable data concerning other types of wage violations in D.C. are not available. But, it
is safe to conservatively assume that there are tens of thousands more victims annually (see
Bernhardt et al. 2009). Nevertheless, DOES receives only 700-800 wage theft-related complaints
per year, across all categories (DOES 2019).
32
258
The fundamental problem is that DOES utilizes a passive, complaint-based enforcement
strategy. In essence, it sits and waits for workers to self-report. As Janice Fine (2017) explains,
this has “been the dominant approach taken by the federal government for many years,” along
with most states and cities. This reactive practice misses large numbers of violations and fails to
create sustainable and lasting compliance (Weil & Pyles 2005; Weil et al. 2010).
These empirical realities require us to rethink what access to justice means for poor
people experiencing wage theft. Because despite the District’s comprehensive legal reforms, the
stunning majority of the city’s working poor are effectively denied access either to legal
processes and substantive justice.
In this context, “access” and “justice” must refer to something more than they
traditionally have. To respond to this crisis, we cannot understand “access” to mean the
technical, formal right to bring a claim. Formulating this problem as one of “unmet legal need”
(Sandefur 2019) leads to solutions that are too narrow, too focused on granting individuals the
written authority to stand up and assert their rights, especially by obtaining free or low-cost legal
advice. The District has attempted that approach, and in many ways its efforts outstrip those of
other progressive jurisdictions. But the same problems persist across the board: workers must be
willing and able to speak up and step out, and for all practical intents and purposes, most cannot.
The District’s passive approach to remedying wage theft is fundamentally failing the people who
are supposed to be the beneficiaries of its reforms, because these people are systematically
locked out of the process. “Access” will never be achieved if we wait for workers to come
forward. Too many cannot. Justice must be brought to them.
Worse, even when people do go to the government for help, they fail to find true justice.
The process is frequently slow, frustrating, and ineffective, defined by errors,
33
259
miscommunications, and weak monetary sanctions. But so much of the harm of wage theft
happens during and immediately after the fact. These delays, which often end with workers
receiving little or no money (see Fritz-Mauer 2021; Cho et al. 2013), mean that to the extent
“justice” is served, it is a day late and a dollar short.
These chronic problems breed deep resentment, and frequently came up in interviews
with those who did attempt formal enforcement of their rights. Ruben, for example, summed up
his experience with DOES like this: “Everybody on some slow motion time, you know?” The
agency did not seem to understand the urgency. Caleb agreed. So did Olivia, and Lisa, and every
single lawyer and activist who participated in this research.
Similarly, “justice” should not be understood merely as the technical provision of legal
services or a lawful outcome following a defined, regulated process. This is a narrow
interpretation that, at best, encompasses only the procedural side of the concept. The workers in
this study who went to DOES all received legal guidance and formal outcomes. But none of
these people found “justice” in any meaningful sense. Caleb’s claim languished before DOES for
more than a year; Olivia’s for more than 9 months. Ruben and Lisa received their unpaid wages
and some damages, but both experienced significant retaliation. Even absent that retaliation, the
agency’s delays and mistakes ensured a negative experience. And, of course, those who never
make a claim at all never receive anything approaching justice.
At a minimum, “justice” for wage theft requires three components:
1. Payment of unpaid wages and damages appropriate to compensate workers for both
the harms of wage theft and the time and energy they must spend asserting their
rights;
34
260
2. Sanctions sufficient to deter future bad acts by an employer, especially retaliatory
ones; and
3. Processes that are simple, accessible, and speedy.
But more fundamentally, every single person here who experienced wage theft also
suffered injustice, regardless of their ultimate outcome. The District’s weak enforcement scheme
has created an environment where the stakes are low and wage theft is rampant. “Justice” is not
only something that becomes possible following rights violations. When violations are
prevented, and workers are able to enjoy jobs free of pay-based degradations, that is a more
valuable form of justice than a winning lawsuit could ever be.
The final section outlines how we must think about wage theft reform if we want to make
access to justice something more than a catchphrase. In doing so, it proves that these demands
are not unrealistic pipe dreams by highlighting one public entity that has embraced these
principles and worked to adopt best practices: Denver Labor.
VI. Access, Justice, and the Promise of Legal Reform
A system of “bottom-up workplace law enforcement,” which places responsibility for
finding and reporting violations, fails the people who need protection the most (Alexander &
Prasad, 2014). For many in D.C., recent legal changes have not shifted the basic power dynamics
that extract silence from the very people the system depends upon. Access to justice requires
moving beyond traditional ideas of legal empowerment, whereby statutory reforms address a
problem through a surface-level expansion of individual legal authority. Rights enforcers must
account for the problems described here and proactively mitigate them.
35
261
While these findings may be depressing, these problems are not intractable. There are
solutions, opportunities for meaningful improvement, because how people think about and
respond to violations of their rights, and how they conceive of the justice system itself, may be
shaped. Rebecca Sandefur (2007) has explained that potential legal claimants’ past experiences
influence their responses to present problems, including what options they consider and pursue.
Rather than waiting for claimants to come forward, a better policy “would meet people where
they are, either by changing the way problems’ solutions are institutionalized or by aggressively
seeking out people with problems and marketing solutions to them” (2007, p. 127). This
approach will require both statutory legal changes designed to sweep more people and
institutions into the system and shift structural barriers, and for governments to adopt proactive,
creative strategies.
A. Statutory Reform and Transforming Public-Private Relationships
Formal legal reform is not enough, but it is far from meaningless. This is especially true
if reforms focus on empowering institutions—non-profits, labor unions, and worker centers—to
action.
Many jurisdictions have done this, in part, by allowing for class action lawsuits and third-
party wage and hour complaints. But best practices contemplate and require robust partnerships
that empower groups to address wage theft in close coordination with the government, and to
supplement the government’s resources through private action. To that end, policymakers should
pass laws granting private institutions the power to bring qui tam enforcement actions, and
enforcement agencies should actively partner with community organizations.
36
262
Qui tam laws allow private actors to file a lawsuit to enforce a public interest. These
private parties stand in the shoes of the government itself, able to sue not only to remedy
individual harms, but also injuries to society, like the payroll tax fraud that accompanies
misclassification (see Fritz-Mauer, 2021, pp. 788-89). These laws fill an enforcement gap, and
better allow institutions to advocate for those who cannot easily do it themselves.
Rights enforcers can also partner with community organizations in a variety of ways.
Recently, a new approach that formally empowers private citizens to act as limited government
agents to conduct workplace investigations has become more popular. Community members
(often union organizers) with knowledge of wage theft-riddled industries, like construction, are
“badged” or “deputized” to go to job sites, interview workers, and otherwise collect and remit
information to determine whether and to what extent employers are violating wage laws. This
strategy comes at a very low cost, but can dramatically improve the knowledge-gathering
capabilities of enforcement agencies. Since the mid-1990s, it has been successful in the Los
Angeles Unified School District (see Fine & Gordon, 2010). More recently, Multnomah County
in Oregon greenlit such a program for publicly-funded construction projects (McIntosh, 2020).
B. New Strategies
This problem will not be seriously addressed by formal legal reform alone, no matter how
comprehensive. Furthermore, enforcement agencies are dramatically underfunded (see, e.g.,
Cooper & Kroeger, 2017; Galvin, 2016). In an age characterized by the politics of austerity, that
fact is not likely to change anytime soon.
But there is much that can be done. Experts argue for “strategic enforcement,” an
approach that emphasizes long-term, widespread compliance through smart, punctuated efforts
37
263
(see, e.g., Weil, 2018; Fine, 2017; Su, 2016). By analyzing wage theft in context, identifying
high-risk industries, and then conducting targeted investigations, governments may do much
more to ensure that workers suffering rights violations will find justice without having to risk so
much.
At a minimum, policymakers must understand the perspectives of low-wage workers and
work to mitigate the hurdles that preclude claiming. Agencies must move beyond their passive,
complaint-based approach, which focuses on individual cases and individualized harm. Instead,
they must be proactive, and use all of their tools, labor, and authority to engage in
comprehensive, thoughtful investigations. This will require expanding the frequency of agency-
initiated investigations and changing how the agency interacts with and mobilizes the
community.
Enforcement agencies can do much more to work with motivated community members,
many of whom are ideologically or even financially invested in having a compliant labor market.
Soliciting third-party complaints and tips, attending community meetings, and forging closer
professional working relationships with workers’ rights advocates are all easy, cost-effective
strategies.
Such partnerships are also crucial for engagement. Many low-wage workers perceive a
wide gap between them and the government, which can appear monolithic, inaccessible, and
even threatening. But community organizations are often deeply embedded among the people
they serve. These groups can serve as effective intermediaries, bridging the gap between
enforcement agencies and people who may be skeptical of them, and drawing attention to bad
actors known to commit wage theft.
38
264
These strategies are not merely theoretical. Leaders of both the California Division of
Labor Standards Enforcement and the Department of Labor have reported that by partnering with
workers’ rights organizations and unions, they were able to learn about and act upon violations
that otherwise might have stayed under the radar (Weil 2018; Su 2016). And today, this approach
has found purchase at Denver Labor, a division of the Auditor’s Office.
C. Best Practices at Denver Labor
“I think one of the biggest changes we’ve had, that’s had the biggest impact, was that we
created communications,” says Jeff Garcia, the Executive Director of Denver Labor. Not just a
communications department, though. “Every investigator has periods of time where they’re in
the community, meeting with organizations, whether they’re labor organizations, public interest
organizations, or employers, everyone does outreach.” To Jeff, the benefit is clear: “Those types
of relationships have made it so we can find the wrongdoing or the mistakes without us actually
having to do any [additional] work.”
Denver Labor comprises one third of the Denver Auditor’s Office, and it enforces the
city’s minimum wage, overtime, and prevailing wage laws. For a long time, though, that branch
of the office was largely defunct. For about sixty years, the Auditor has been responsible for
enforcing the prevailing wage; for about fifty years, the law was “dramatically ignored.”
But when Denver elected a new Auditor—Timothy O’Brien—in 2015, things started to
change very quickly. In part, this is due to the fact that the city updated its existing laws and
policies, and passed new ones regarding wage theft. In the Auditor’s Office, Jeff spearheaded an
effort to modernize the city’s prevailing wage law, and then in 2019, the Auditor formally
39
265
created Denver Labor. That same year, the City Council granted the Auditor authority to enforce
the minimum wage for city contractors, and passed its own minimum wage ordinance, again
vesting power with the Auditor.
These laws are far from the gold standard of anti-wage theft legislation. Denver Labor
has limited authority to subpoena records, for instance. There is no “up the chain” liability to
ease collections. Perhaps most significantly, the division has no power at all to collect damages
on top of unpaid wages, and only obtained the authority to conduct proactive investigations on
January 1, 2022. But while Denver Labor lacks the same powerful tools of deterrence as DOES,
it has adopted a categorically different enforcement approach.
In short, there are three key components to what Denver Labor does, all of which fit
neatly within the umbrella of strategic enforcement.
First, Jeff and his people use cutting-edge technology to find wage theft and collect
industry-level data. Prior to Jeff’s tenure, the office only used its auditing software inefficiently
and sparingly. “We turned every bell and whistle on for a while,” says Jeff, which allowed them
to automate “overtime calculations, apprentice ratios, [and] classification checks,” which “were
never checked before.” This same technology also identifies noncompliant sectors, which
Denver Labor is now using to drive targeted investigations.
Second, the Auditor’s Office simply dedicated more resources to wage theft. It did so
both by shifting money and personnel internally, and seeking a larger budget from the City. In
2015, there were four investigators; today, Denver Labor has more than 30 employees.
Third, Denver Labor adopted the understanding that wage theft is a problem that can be
addressed, but only if government agencies leverage limited resources effectively. Some of these
enforcement changes simply draw on common sense. For example, prior to 2015, the office only
40
266
examined certified payroll records every fourth week. “One of the things that we found out is
that bad employers figured out the pattern and knew that every fourth payroll was the one we
checked,” says Jeff Garcia. “Once we went back and checked all of the other ones, they were out
of compliance three out of four times.”
But others required a much more thoughtful analysis of what meaningful regulation
requires. In the last six years Jeff has professionalized the office, developed institutional
expertise, and smoothed processes. Previously, the office hired former construction workers to
be investigators, but the modern world of wage theft requires more specialized expertise and
knowledge. “My investigators now are actuaries and attorneys, because I think those are the two
types of professional people who can uncover fraud . . . and [do the] forensic kind of work that
we do,” Jeff explains. Denver Labor has also worked to make the complaint process accessible,
streamlining enforcement in an effort to sweep more people into the system, generate cases, and
resolve them quickly.
Most importantly, the division set out to become a visible presence in the community,
and in doing so, to overcome the barriers that prevent people from seeking help. When asked
why this is important, Jeff explains he cannot expect most people to come forward to report their
wage theft, especially if they’re poor:
[P]eople don’t have the time to step aside and say “I’m going to find a lawyer, and
I’m going to put together the evidence I need to show to my lawyer….I have to feed
my kids, I have to find transportation, I have to find a new job, I can’t keep working
for this person!” There’s all these other things that are just higher on the list.
The point is not to find victims of wage theft at every event, but to spread awareness and for the
future.
There is reason to think this approach has rippling reputational effects, building trust and
confidence among stakeholders and in communities, and thereby improving enforcement over
41
267
time. At least among workers’ rights activists and attorneys in the District of Columbia, there is
both a desire and a hesitancy to work with DOES. People want to partner with the agency
because it occupies a unique and powerful position, but are reluctant to try because they have
been disappointed too many times. Likewise, although many low-wage workers express that the
government was too inactive and would not adequately enforce or protect their rights, they also
frequently reported wanting more help, support, and effort from the government. In other words,
many of the people represented here reported what social scientists have long argued: that their
views on the legal system were fluid and open to change (see, e.g., Ewick & Silbey 1998).
To some degree, Denver experiences these same problems. But, the key difference is that
the primary public body responsible for redressing wage theft has embarked on a years-long,
conscious effort to mitigate the practical barriers to claims-making.
This approach contrasts starkly with that of the District of Columbia. Empirically, the
results speak for themselves. In Fiscal Year 2019, D.C. DOES recovered $194,000 in unpaid
wages (Fritz-Mauer, 2021, p. 50). In the same timeframe, Denver Labor recovered nearly
$680,000 (Denver Auditor’s Office 2021), and did so with a weaker mandate, fewer employees,
and a smaller budget. In fact, DOES only even assessed about $483,000 in unpaid wages (Fritz-
Mauer, 2021: 50).
This is not an exhaustive discussion, but an illustration of what is needed to effectively
deal with the pernicious realities of wage theft. These strategies are directly responsive to the
access to justice problem as it exists for poor people who have their wages stolen. They are
geared towards overcoming barriers to bringing legal claims and creating industry-wide
compliance, and acknowledge that policymakers must move beyond the narrow strategies that
have failed to systemically improve access to justice. For many low-wage workers, proactive and
42
268
community-embedded interventions are not just the best source of aid, but the only plausible one.
Practically speaking, they cannot access justice; it must be brought to them.
Conclusion
In discussing the issue of access to justice, socio-legal scholars have urged researchers to
develop contextualized, detailed research plans that move beyond the traditional understanding
of the problem as one of unmet legal need (Sandefur 2019). To truly understand this “crisis,” we
must closely analyze how people think about and respond to their potential legal claims, and
“what access to justice means and what the current lack of access entails” (Albiston & Sandefur
2013, p. 105).
This Article has taken on this call, revealing the full scope of wage theft’s harms and why
so much of it goes unremedied. Although existing enforcement strategies are largely failing to
respond to the problem, there is much that can be done to improve access to justice for the
working poor. Doing so will require enforcers to reconsider what they know about the dynamics
of wage theft; empathize with the viewpoints and experiences of low-income people; and
embrace new strategies designed to bring justice to the poor by meeting them where they are.
43
269
References
Albiston, Catherine R. "Bargaining in the shadow of social institutions: Competing discourses
and social change in workplace mobilization of civil rights." Law & Society Review 39, no. 1
(2005): 11-50.
Albiston, Catherine R., and Rebecca L. Sandefur. "Expanding the empirical study of access to
justice." Wis. L. Rev. (2013): 101.
Albiston, Catherine R., Lauren B. Edelman, and Joy Milligan. "The dispute tree and the legal
forest." Annual Review of Law and Social Science 10 (2014): 105-131.
Alexander, Charlotte S., and Arthi Prasad. "Bottom-up workplace law enforcement: An
empirical analysis." Ind. LJ 89 (2014): 1069
Bernhardt, Annette, Ruth Milkman, Nik Theodore, Douglas D. Heckathorn, Mirabai Auer, James
DeFilippis, Ana Luz González, Victor Narro, and Jason Perelshteyn. "Broken laws, unprotected
workers: Violations of employment and labor laws in America's cities." (2009).
Best, Arthur, and Alan R. Andreasen. "Consumer response to unsatisfactory purchases: A survey
of perceiving defects, voicing complaints, and obtaining redress." Law & Soc'y Rev. 11 (1976):
701.
Bobo, Kim. Wage Theft in America: Why Millions of Americans Are Not Getting Paid—And
What We Can Do about It. The New Press, 2014.
Boushey, Heather, Shawn Fremstad, Rachel Gragg, and Margy Waller. Understanding low-wage
work in the United States. Washington, DC: Center for Economic and Policy Research, 2007.
Calavita, Kitty, and Valerie Jenness. "Inside the pyramid of disputes: Naming problems and
filing grievances in California prisons." Social problems 60, no. 1 (2013): 50-80.
Carré, Francoise. "Dependent contractor misclassification." Economic Policy Institute 8 (2015).
Cho, Eunice Hyunhye, Tia Koonse, and Anthony Mischel. "Hollow Victories: The Crisis in
Collecting Unpaid Wages for California's Workers." (2013).
Claire Conway. “Poor Health: When poverty becomes disease.” UCSF Magazine, Fall 2015,
https://magazine.ucsf.edu/poor-health.
Cooper, David, and Teresa Kroeger. "Employers steal billions from workers’ paychecks each
year: Survey data show millions of workers are paid less than the minimum wage, at significant
cost to taxpayers and state economies." Economic Policy Institute. May 10 (2017).
Council of the District of Columbia, “Minimum Income Study Released by Council Budget
Office, April 5, 2018. https://dccouncil.us/minimum-income-study-released-council-budget-
office/ [DC Council 2018].
44
270
D.C. Office of the Attorney General, “Attorney General Racine to Enforce Workers’ Rights
Laws Against Abusive Employers,” accessed March 21, 2022.
https://oag.dc.gov/release/attorney-general-racine-enforce-workers-rights [OAG 2017].
D.C. Office of the Attorney General, “AG Racine Sues Power Design for Cheating Hundreds of
Electrical Workers Out of Wages and Benefits,” accessed March 21, 2022.
https://oag.dc.gov/release/ag-racine-sues-power-design-cheating-hundreds [OAG 2018].
D.C. Department of Employment Services. “Minimum Wage Impact Study.” Nov. 1, 2017,
https://does.dc.gov/sites/default/files/dc/sites/does/page_content/attachments/Minimum%20Wag
e%20Impact%20Study%20Report_r1.pdf (Rep.) [DOES 2017].
D.C. Department of Employment Services, Response to 2020 Request Under the Freedom of
Information Act (on file with author) [DOES 2019].
Denver Auditor’s Office, “2021 Annual Report.” Denver Auditor’s Office, March 30, 2022.
https://www.denvergov.org/Government/Agencies-Departments-Offices/Agencies-Departments-
Offices-Directory/Auditors-Office/Audit-Services/2021-Annual-Report
Desmond, Matthew. Evicted: Poverty and profit in the American city. Crown, 2016.
Doussard, Marc, and Ahmad Gamal. "The rise of wage theft laws: can community–labor
coalitions win victories in state houses?." Urban Affairs Review 52, no. 5 (2016): 780-807.
Eastern Research Group. "The social and economic effects of wage violations: Estimates for
California and New York." Final Report prepared for the US Department of Labor (Lexington,
MA: Eastern Research Group (2014).
Ewick, Patricia, and Susan S. Silbey. The common place of law: Stories from everyday life.
University of Chicago Press, 1998.
Ewick, Patricia, and Susan Silbey. "Narrating social structure: Stories of resistance to legal
authority." American journal of sociology 108, no. 6 (2003): 1328-1372.
Fine, Janice, and Jennifer Gordon. “Strengthening labor standards enforcement through
partnerships with workers’ organizations.” Politics & Society 38, no. 4 (2010): 552-585.
Fine, Janice. “Enforcing labor standards in partnership with civil society: Can co-enforcement
succeed where the state alone has failed?” Politics & Society 45, no. 3 (2017): 359-388.
Fritz-Mauer, Matthew. “The ragged edge of rugged individualism: wage theft and the
personalization of social harm.” U. Mich. JL Reform 54 (2020): 735.
Fussell, Elizabeth. “The deportation threat dynamic and victimization of Latino migrants: Wage
theft and robbery.” The Sociological Quarterly 52, no. 4 (2011): 593-615.
45
271
Galanter, Marc. “Why the haves come out ahead: Speculations on the limits of legal change.”
Law & Soc'y Rev. 9 (1974): 95.
Galemba, Rebecca Berke. n.d. Unjust Wages: Day Laborers and Advocates Fighting Wage
Theft. Under contract with Stanford University Press.
Galvin, Daniel J. "Deterring wage theft: Alt-labor, state politics, and the policy determinants of
minimum wage compliance." Perspectives on Politics 14, no. 2 (2016): 324-350.
Gleeson, Shannon. “Brokered pathways to justice and cracks in the law: A closer look at the
claims-making experiences of low-wage workers.” WorkingUSA 18, no. 1 (2015): 77-102.
Gleeson, Shannon. Precarious claims: The promise and failure of workplace protections in the
United States. Oakland: University of California Press, 2016.
Hacker, Jacob S., and Paul Pierson. Winner-take-all politics: How Washington made the rich
richer--and turned its back on the middle class. Simon and Schuster, 2010.
“matrix,” Laffey Matrix, March 20, 2022, http://www.laffeymatrix.com/see.html.
Levitsky, Sandra R. “‘What rights?’ The construction of political claims to American health care
entitlements.” Law & Society Review 42, no. 3 (2008): 551-590.
MacDowell, Elizabeth L. “Reimagining Access to justice in the Poor People's Courts.” Geo. J.
on Poverty L. & Pol'y 22 (2014): 473.
Mansfield, Christopher, and Lloyd F. Novick. “Poverty and health: focus on North Carolina.”
North Carolina Medical Journal 73, no. 5 (2012): 366-373.
McIntosh, Don. "Union volunteers help enforce laws against construction wage theft on
Multnomah County projects." (2020).
Miller, Richard E., and Austin Sarat. "Grievances, claims, and disputes: Assessing the adversary
culture." Law and Society Review (1980): 525-566.
Minkler, Meredith, Alicia L. Salvatore, Charlotte Chang, Megan Gaydos, Shaw San Liu, Pam
Tau Lee, Alex Tom, Rajiv Bhatia, and Niklas Krause. “Wage theft as a neglected public health
problem: An overview and case study from San Francisco’s Chinatown district.” American
journal of public health 104, no. 6 (2014): 1010-1020.
Pager, Devah. “The mark of a criminal record.” American journal of sociology 108, no. 5 (2003):
937-975.
Pager, Devah, Bart Bonikowski, and Bruce Western. “Discrimination in a low-wage labor
market: A field experiment.” American sociological review 74, no. 5 (2009): 777-799.
46
272
Pruitt, Lisa R., and Bradley E. Showman. “Law stretched thin: Access to justice in rural
America.” SDL Rev. 59 (2014): 466.
Reis Thebault. “DC board won't penalize company at center of massive wage-theft case.” The
Washington Post, September 03, 2018. https://www.washingtonpost.com/local/dc-politics/dc-
agency-wont-penalize-company-at-center-of-massive-wage-theft-case/2018/09/03/d8deb71c-
ac83-11e8-b1da-ff7faa680710_story.html?utm_term=.5a650ac3a5df.
Rhode, Deborah L. “Whatever happened to access to justice.” Loy. LAL Rev. 42 (2008): 869.
Rose, Jonathan, et al. The Best States to Work Index: A Guide to Labor Policy in US States
(OxFam, 2018).
Ruckelshaus, Catherine, and Ceilidh Gao. “Independent Contractor Misclassification Imposes

Huge Costs on Workers and Federal and State Treasuries.” New York: National Employment
Law Project (2017).
Sandefur, Rebecca L. “Access to what?” Daedalus 148, no. 1 (2019): 49-55.
Sandefur, Rebecca L. “Access to civil justice and race, class, and gender inequality.” Annu. Rev.
Sociol 34 (2008): 339-358.
Simon, Dan. “Poverty fact sheet: Poor and in poor health.” UW Institute for Research on Poverty
(2012).
Su, Julie A. “Enforcing Labor Laws: Wage Theft, the Myth of Neutrality, and Agency
Transformation.” Berkeley Journal of Employment and Labor Law 37, no. 1 (2016): 143-156.
Trautner, Mary Nell, Erin Hatton, and Kelly E. Smith. “What workers want depends: Legal
knowledge and the desire for workplace change among day laborers.” Law & Policy 35, no. 4
(2013): 319-340.
U.S. Census Bureau, “QuickFacts: District of Columbia,” July 1, 2021,
https://www.census.gov/quickfacts/fact/table/dc/PST045217
Valenzuela, Abel. On the corner: Day labor in the United States. Latino Union of Chicago,
2006.
Weil, David. The fissured workplace. Harvard University Press, 2014a.
Weil, David. “Mending the fissured workplace.” What Works for Workers?: Public Policies and
Innovative Strategies for Low-Wage Workers 108 (2014).
Weil, David. “Creating a strategic enforcement approach to address wage theft: One academic’s
journey in organizational change.” Journal of Industrial Relations 60, no. 3 (2018): 437-460.
Weil, David, and Amanda Pyles. “Why complain-complaints, compliance, and the problem of
enforcement in the us workplace.” Comp. Lab. L. & Pol'y. J. 27 (2005): 59.
47
273
Cases Cited
Salazar v. District of Columbia, 123 F. Supp. 2d 8 (D.D.C. 2008)
Statutes Cited
D.C. Accrued Sick and Safe Leave Act, D.C. Code § 32-531.01 et seq. (2013).
Fair Shot Minimum Wage Amendment Act, D.C. Code § 32-1001 et seq. (2016)
Wage Theft Prevention Amendment Act of 2014, D.C. Code § 32-1301 et seq. (2014).
48
274
Tables and Figures
Table 1
Persons in Maximum income for inclusion

family/household
1 $36,180
2 $48,720
3 $61,260
4 $73,800
49
275
276
Week 6
277
278
This content downloaded from
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
All use subject to https://about.jstor.org/terms
279
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
280
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
281
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
282
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
283
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
284
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
285
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
286
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
287
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
288
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
289
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
290
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
291
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
292
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
293
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
294
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
295
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
296
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
297
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
298
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
299
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
300
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
301
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
302
146.50.98.82 on Thu, 01 Dec 2022 17:31:07 UTC
303
304
647 Wetenschap......
Verschillen in
straftoemeting in
soortgelijke zaken
Een kwantitatief onderzoek naar de rol van specifieke
kenmerken van de dader
Hilde Wermink, Jan de Keijser en Pauline Schuyt1
Spelen kenmerken van een verdachte als een al dan niet Nederlands uiterlijk en het al dan niet spreken van
de Nederlandse taal een rol bij het bepalen van de straf? In het hier beschreven onderzoek, dat door externe
peer reviewers is beoordeeld, wordt dit, voor zover bekend, voor het eerst onderzocht op basis van directe
observaties tijdens strafzittingen van de politierechter. Daders met een Nederlands uiterlijk die ook de Neder-
landse taal spreken blijken de minste odds te hebben om veroordeeld te worden tot een onvoorwaardelijke
gevangenisstraf. In vergelijking met die groep zijn de odds om veroordeeld te worden tot een onvoorwaardelij-
ke gevangenisstraf voor daders met een buitenlands uiterlijk die de Nederlandse taal spreken ruim vijf keer
hoger. Voor daders met een buitenlands uiterlijk die ook nog eens de Nederlandse taal niet machtig zijn, zijn
de odds om achter de tralies te verdwijnen maar liefst twintig keer hoger.
D
e relatief grote vrijheid van de rechters bij het bepa- macht de afgelopen twintig jaar veel aandacht besteed
len van de straf is van belang voor de wijze waarop aan het ontwikkelen van instrumenten ten behoeve van
zij tot hun straf komen en de factoren die daarbij een consistentere straftoemeting. Na de invoering van die
een rol spelen. In eerder onderzoek is aangetoond dat de instrumenten is onderzoek naar verschil in straftoeme-
straf die wordt opgelegd door de rechter voor het grootste ting in Nederland, op enkele uitzonderingen na6, van de
gedeelte wordt bepaald door het delict dat is gepleegd en de empirische radar verdwenen. Tevens ontbreekt in veel
criminele geschiedenis van de dader.2 Echter, op basis van voorgaand onderzoek aandacht voor hetgeen zich op de
de ‘focal concerns’ theorie wordt verondersteld dat de rech- zitting afspeelt, terwijl men doorgaans meent dat specifie-
ter bij het bepalen van de straf onder andere rekening ke omstandigheden tijdens de strafzitting, zoals het optre-
houdt met de geschatte gevaarlijkheid van de dader, waarden van de raadsman of -vrouw van de verdachte, de
door kenmerken van de persoon van de verdachte ook een opstelling van de verdachte en de slachtofferverklaring de
rol spelen bij de straftoemeting. In het huidige onderzoek straftoemeting mede beïnvloeden.7
richten wij ons op de rol van het al dan niet betuigen van In eerder straftoemetingsonderzoek is vooral
spijt, het geslacht en het gecombineerde kenmerk van een gebruikgemaakt van officieel geregistreerde gegevens
al dan niet Nederlands uiterlijk met het al dan niet spreken (rechtbankstatistieken) en van quasi-experimentele
van de Nederlandse taal bij het bepalen van de straf. Bij het methoden aan de hand van schriftelijke casussen. In bei-
vaststellen van die relaties zal rekening worden gehouden de gevallen ontbreekt een realistisch beeld van wat zich
met andere kenmerken waarvan bekend is dat zij een rol op de zitting afspeelt. Het is daarom goed nader onder-
kunnen spelen bij de straftoemeting, zoals delictskenmer- zoek te doen naar deze factoren. Observaties tijdens het
ken en kenmerken van criminele geschiedenis. onderzoek ter terechtzitting vormen daarvoor een geëi-
Eerder onderzoek naar straftoemeting heeft zich met gende methode. In dit onderzoek richten wij ons op één
name gericht op het verklaren van verschillen in straftoe- aspect van straftoemetingsbeslissingen, namelijk de
meting op basis van demografische kenmerken van beslissing om al dan niet een onvoorwaardelijke gevange-
daders, zoals etniciteit3 en verschillen in straftoemeting nisstraf op te leggen. De gevangenisstraf is de zwaarste
tussen rechters en rechtbanken4. Naar aanleiding van straf die in Nederland opgelegd kan worden en heeft gro-
onder meer die onderzoeken5 is binnen de rechtsprekende te gevolgen, zowel bedoelde als onbedoelde.
726
305
In Nederland heeft onderzoek zich overigens al eer- Naast het uiterlijk van de verdachte
der gericht op gegevens afkomstig van observaties tijdens
zittingen.8 In die studies is onderzocht hoe het communi- wordt ook verwacht dat de taal die
catieproces verloopt tijdens de zitting, hoe verdachten
zich tijdens de zitting gedragen en hoe de sanctie door de de verdachte tijdens de zitting
rechter wordt gemotiveerd. De rol van hetgeen zich ter zit-
ting afspeelt, zoals het al dan niet betuigen van spijt en spreekt de straftoemeting beïnvloedt
het spreken van de Nederlandse taal, bij straftoemeting is
echter niet eerder onderzocht op basis van directe obser-
vaties.
baarheid van de dader. Daders met een criminele geschie-
Theoretische achtergrond en bevindingen denis zouden zwaarder gestraft kunnen worden, omdat
van eerder onderzoek een dergelijke geschiedenis een grotere mate van verwijt-
Een gezaghebbende sociaalwetenschappelijke benadering baarheid suggereert voor het huidige gepleegde delict.10
van factoren die een rol spelen bij de straftoemeting is de Daarnaast zouden daders met een strafblad ook als
‘focal concerns’ theorie. Volgens deze theorie grijpt de gevaarlijker kunnen worden gepercipieerd door het ver-
rechter bij het nemen van de straftoemetingsbeslissing hoogde risico van recidive. Eerder onderzoek heeft aange-
terug op de volgende drie criteria: toond dat deze delictskenmerken en kenmerken van cri-
1) de mate van verwijtbaarheid van de dader; minele geschiedenis het grootste deel van de
2) de gevaarlijkheid van de dader / bescherming van de straftoemetingsbeslissing verklaren.11 Desalniettemin is
maatschappij; er bewijs dat de straftoemetingsbeslissing zelfs na contro-
3) de praktische gevolgen van de beslissing voor organisa- le op delictskenmerken en kenmerken van criminele
ties en individuen.9 geschiedenis wordt beïnvloed door socio-demografische
Verondersteld wordt dat de straftoemetingsbeslis- kenmerken, zoals het geslacht en de etniciteit van de
sing het resultaat is van een complexe besluitvorming dader.12
waarin alle relevante kenmerken en verscheidene ‘focal Gezien de beperkte tijd en informatie die voor de
concerns’ door de rechter in overweging worden genomen. rechter voor handen is over de gevaarlijkheid van de
Het criterium van de mate van verwijtbaarheid is dader stelt de ‘focal concerns’ theorie dat de rechter terug-
gerelateerd aan de ‘just deserts’ filosofie van straffen, valt op eigen ervaringen met daders uit bepaalde sociale
waarin gesteld wordt dat de straf moet passen bij de ernst groepen en criminele stereotypen.13 Zelfs wanneer uitge-
van het delict dat is gepleegd en de schuld van de dader. breide informatie beschikbaar is over de verdachte is reci-
Tevens wordt verondersteld dat een criminele geschiede- dive nooit volledig voorspelbaar en daarnaast is het
nis van invloed zou kunnen zijn op de mate van verwijt- karakter van de verdachte nooit volledig te doorgronden.14
Auteurs University 1999. Zie ook D. Steffensmeier 7. S.G.C. van Wingerden en P. Nieuwbeer- applied perspectives, Oxford and Portland,
1. H.T. Wermink MSc, dr. J.W. de Keijser en en S. DeMuth, ‘Ethnicity and sentencing ta, De vervolging en berechting van moord Oregon: HART publishing 2010.
mr. dr. P.M. Schuyt zijn resp. als promoven- outcomes in U.S. federal courts: Who is en doodslag in Nederland, 1993- 2004. 11. Zie bijv. B. Rovers, Klassenjustitie.
dus, universitair hoofddocent Criminologie punished more harshly’, American Sociolo- Rapport NSCR-2006-2. Overzicht van onderzoek naar selectiviteit
en universitair hoofddocent Straf- en straf- gical Review, 2000-65, p. 705-729 en S.D. 8. Zie bijv. G.P. Hoefnagels, Een eenvoudi- in de Nederlandse Strafrechtketen, Rotter-
procesrecht verbonden aan het Instituut Bushway en A.M. Piehl, ‘Judging judicial ge zitting, Alphen aan den Rijn: Samsom dam: Erasmus University 1999 en J.T.
voor Strafrecht & Criminologie van de Uni- discretion: Legal factors and racial discrimi- 1980; P. Bal, Dwangkommunikatie in de Ulmer, Social Worlds of Sentencing: Court
versiteit Leiden. De auteurs willen Johan nation in sentencing’, Law & Society rechtszaal, Arnhem: Gouda Quint BV 1988; Communities under Sentencing Guidelines,
van Wilsem en Anke Ramakers bedanken Review, 2001-35(4), p. 733-764. G. Kannegieter, Ongelijkheid in de straftoe- Albany, NY: State University of New York
voor het leveren van constructief commen- 4. J.P.S. Fiselier, ‘Regionale verscheidenheid meting. De invloed van de sociale positie Press 1997 en J.D. Wooldredge, ‘Analytical
taar op een eerdere versie van dit artikel. in strafrechtspleging’, DD 1985, p. 204- van de verdachte op strafrechtelijke beslis- rigor in studies of disparities in criminal case
211. Zie ook J.W. de Keijser, Punishment singen, Groningen: Wolters Noordhoff B.V. processing’, Journal of Quantitative Crimi-
Noten and Purpose: From moral theory to punish- 1994; D. Ecoma-Verstege en S. van Win- nology 1998-14, p. 155-179.
2. Zie bijvoorbeeld B. Rovers, Klassenjusti- ment in action, Amsterdam: Thela Thesis gerden, ‘Ook de mondelinge motivering 12. K. Daly en R.L. Bordt, ‘Sex effects and
tie. Overzicht van onderzoek naar selectivi- 2000. A.C. Berghuis, ‘De harde en de zach- van de straftoemetingsbeslissing laat te sentencing: An analysis of the statistical
teit in de Nederlandse Strafrechtketen, te hand; een statistische analyse van ver- wensen over’, Proces 2009-88(1), p. 36-46; literature’, Justice Quarterly 1995-12, p.
Rotterdam: Erasmus University 1999 en J.T. schillen in sanctiebeleid’, Trema, 1992-3, p. W. van Rossum, Verschijnen voor de rech- 141-176 en M. Zatz, ‘The convergence of
Ulmer, Social Worlds of Sentencing: Court 84-93. ter. Hoe het hoort en het ritueel van Turkse race, ethnicity, gender, and class on court
Communities under Sentencing Guidelines, 5. Zie ook R. Jongman en T. Schilt, ‘Gelet verdachten in de rechtszaal, Amsterdam: decisionmaking: Looking toward the 21st
Albany, NY: State University of New York op de persoon van de verdachte…’, TvCr Duizend en Een 1998. century’, Criminal Justice: The national
Press 1997 en J.D. Wooldredge. ‘Analytical 1976-6, p. 273-287; A.C. Berghuis, ‘Gelijk- 9. D. Steffensmeier, J.T. Ulmer en J.H. Kra- institute of justice journal 2000-503-552.
rigor in studies of disparities in criminal case heid van straffen’, Trema, 1992-3, p. mer, ‘The interaction of race, gender and 13. C. Albonetti, ‘An integration of theories
processing’, Journal of Quantitative Crimi- 77-84. age in criminal sentencing: The punishment to explain judicial discretion’, Social Pro-
nology 1998-14, p. 155-179. 6. Zie bijv. F. van Tulder en B. Diephuis, cost of being young, Black and male’, Cri- blems 1991-38, p. 247-266.
3. B. Rovers, Klassenjustitie. Overzicht van ‘Afgewogen straffen. Analyse en verbetering minology 1998-36, p. 763-797. 14. M. Bagaric, Punishment and Senten-
onderzoek naar selectiviteit in de Neder- van de databank consistente straftoemeting’, 10. J.V. Roberts en A. von Hirsch, Previous cing: A rational approach, London: Caven-
landse Strafrechtketen, Rotterdam: Erasmus research memoranda, 2007-3(4), p. 1-68. convictions at sentencing. Theoretical and dish 2001.
727
306
Wetenschap......
Bij Nederlands onderzoek naar beslissingen over schuld

en straf onder niet-juristen is een aanwijzing gevonden
dat negatieve stereotypering van de verdachte bij beslis-
singen onder tijdsdruk leidt tot een grotere kans op een
schuldigverklaring en tot een zwaardere straf.15
Met betrekking tot het geslacht van de verdachte is in
eerder onderzoek aangetoond dat vrouwen door rechters
ingeschat worden als minder gevaarlijk, minder verwijtbaar
en meer berouwvol in vergelijking met mannen.16 Allochto-
nen of verdachten met een buitenlands uiterlijk zouden
vanwege stereotypen juist als gevaarlijker kunnen worden
ingeschat door rechters, bijvoorbeeld door hun buitenpro-
portionele deelname aan crimineel gedrag en oververtegen-
woordiging in de gevangenispopulatie in vergelijking met
Nederlandse daders.17 Wanneer de rechter de samenleving
wil beschermen tegen recidive zouden zij deze daders moge-
lijk eerder veroordelen tot een onvoorwaardelijke gevange-
nisstraf, zodat zij in ieder geval tijdelijk geen delicten in de
samenleving kunnen plegen. In een overzichtsstudie laat
Rovers zien dat verdachten met geringe sociaaleconomische
hulpbronnen (inclusief allochtonen) naar verhouding vaker
een onvoorwaardelijke straf krijgen opgelegd (in plaats van
een voorwaardelijke straf) en ook vaker een vrijheidsstraf in
plaats van een geldboete.18 Op basis van de ‘focal concerns’
theorie en de bevindingen van eerder onderzoek verwach-
ten wij dat daders met een buitenlands uiterlijk zwaarder
gestraft zullen worden dan daders met een Nederlands
uiterlijk en dat vrouwen lichter zullen worden gestraft dan
mannen. Wij verwachten dat dit zelfstandige effecten zijn,
dus onafhankelijk van andere kenmerken zoals het delict en
het al dan niet hebben van een strafblad.
Naast het uiterlijk van de verdachte wordt ook verwacht
dat de taal die de verdachte tijdens de zitting spreekt, de
straftoemeting beïnvloedt. Het spreken van een andere taal
door de verdachte kan namelijk leiden tot taalbarrières tij-
dens het onderzoek ter terechtzitting. Voor verdachten is het
moeilijker om emoties en gevoelens over te brengen op de
rechter en door de aanwezigheid van een tolk verloopt het
communicatieproces niet direct, wat tevens kan leiden tot
misvattingen. Ook zouden daders met een buitenlands uiter-
Rechters kunnen onbewust

(negatieve) stereotyperingen
gebruiken om de gevaarlijkheid
van de dader te bepalen
lijk die ook niet de Nederlandse taal spreken door negatieve

stereotypering als gevaarlijker ingeschat kunnen worden dan
daders met een buitenlands uiterlijk die Nederlands spreken,
bijvoorbeeld door een verwachte mindere binding met de
samenleving. Daders met een buitenlands uiterlijk die ook
niet de Nederlandse taal spreken, lopen daardoor het risico
zwaarder gestraft te worden dan zowel daders die er Neder-
lands uitzien en de Nederlandse taal spreken als daders die
er buitenlands uitzien en de Nederlandse taal spreken.19 Bij
728
307
het bepalen van een straf bij daders die de Nederlandse taal vaties werden geregistreerd aan de hand van gestandaar-
niet beheersen, kunnen daarnaast praktische bezwaren een diseerde checklists.24 In totaal zijn in het voorjaar van
rol spelen. Indien verdachten de Nederlandse taal niet 2010 tien verschillende rechtbanken bezocht: Alkmaar,
beheersen, zal een rechter wellicht geen taakstraf willen Amsterdam, Arnhem, Breda, Den Bosch, Den Haag, Dord-
opleggen vanwege de problemen die bij de tenuitvoerlegging recht, Haarlem, Utrecht en Zutphen. Van de geobserveerde
daarvan kunnen ontstaan door het niet beheersen van het zaken werd het aantal en het type delict(en) opgenomen
Nederlands. Bij de tenuitvoerlegging van taakstraffen is waarvoor is veroordeeld, alsmede gegevens over eventuele
namelijk niet voorzien in tolken. Bij het grotendeels wegval- eerdere veroordelingen, of de verdachte in voorlopige
len van deze strafoptie zou de rechter eerder geneigd kunnen hechtenis heeft gezeten, of het zwaarste delict alleen of
zijn om een gevangenisstraf op te leggen. samen is gepleegd en de leeftijd van de verdachte. Er zal
Naast het al dan niet spreken van de Nederlandse taal rekening worden gehouden met deze kenmerken als we
door een verdachte is er nog een ander kenmerk dat zich bij de rol van kenmerken van de persoon van de verdachte bij
uitstek op de zitting openbaart en dat de rechter als hand- het bepalen van de straf onderzoeken.
vat voor stereotypering wellicht onbewust gebruikt bij het Op de verzamelde gegevens is de volgende selectie
inschatten van gevaarlijkheid. Dat betreft de mate van het toegepast. Ten eerste hebben de analyses alleen betrek-
betuigen van spijt door verdachte. Volgens Drass en Spencer king op verdachten die zijn veroordeeld. Daarom zijn
is het betuigen van spijt een indicatie dat het minder waar- zaken verwijderd waarin sprake was van vrijspraak, ont-
schijnlijk is dat de dader in de toekomst hetzelfde of soort- slag van alle rechtsvervolging (OVAR) of aanhouding van
gelijk gedrag opnieuw zal vertonen.20 De studie van Drass en de zitting (90 zaken). Daarnaast zijn zaken verwijderd
Spencer toont, met behulp van data afkomstig uit de Ver- waarin sprake was van andersoortige uitspraken, zoals
enigde Staten, aan dat het betuigen van spijt inderdaad van vorderingen tot verlenging van de proeftijd of vorderin-
invloed is op de gepercipieerde gevaarlijkheid van de dader. gen tot de tenuitvoerlegging van een eerder opgelegde
In een studie van Pipes en Allessi wordt aangetoond dat het (deels) voorwaardelijke straf (7 zaken). Vervolgens zijn de
betuigen van spijt in zaken waarin de dader verdacht wordt zaken verwijderd waarin de straf van de rechter onbekend
van serieuze aanranding leidt tot een lagere straf. 21 Naast of onduidelijk was (3 zaken). Aangezien verdachten aan-
de rol van spijt bij stereotypering heeft het betuigen van wezig moeten zijn op de strafzitting om te kunnen spre-
spijt ook een zelfstandig en meer direct effect op straftoe- ken of spijt te betuigen, zijn alle zaken verwijderd waarin
meting. Het is namelijk van invloed op de morele afkeuring de verdachte afwezig was tijdens het onderzoek ter
van daad en dader door de rechter.22 Door dit proces van terechtzitting (76 zaken). De voorafgaande restricties heb-
gepercipieerde gevaarlijkheid en morele afkeuring verwach- ben geleid tot een bestand van 365 veroordeelden (68%
ten we dat verdachten die tijdens de strafzitting spijt betui- van het originele aantal geobserveerde zaken).25
gen minder kans hebben op een onvoorwaardelijke gevan-
genisstraf dan verdachten die geen spijt betuigen. Straftoemeting
We toetsen de hypothesen aan de hand van gegevens De strafsoort die de politierechter oplegt aan de verdachte
uit gestructureerde en gestandaardiseerde observatielijsten staat centraal in dit onderzoek. Meer specifiek is onderzocht
die verzameld zijn bij zittingen van de politierechter in het of een onvoorwaardelijke gevangenisstraf dan wel een ande-
voorjaar van 2010. Van alle zittingen in Nederland vindt de re straf is opgelegd (Wel = onvoorwaardelijke gevangenis-
overgrote meerderheid, ruim 80%, plaats bij de politierech- straf; Niet = geldboete, taakstraf, voorwaardelijke gevange-
ter, slechts een kleine 10% vindt plaats in de meervoudige nisstraf, schuldigverklaring zonder straf). Wanneer een
kamer en de overige 10% bij de kinderrechter.23 onvoorwaardelijke gevangenisstraf in combinatie met een
andere straf werd opgelegd, is deze ingedeeld als onvoor-
Gegevens waardelijke gevangenisstraf. In totaal is bij 81 verdachten
Onze gegevens zijn afkomstig van observaties die zijn ver- een onvoorwaardelijke gevangenisstraf opgelegd (22%) en
richt tijdens politierechterzittingen (541 zaken). De obser- bij de overige 284 verdachten (78%) een andere straf.
15. A. van Knippenberg, A. Dijksterhuis en 18. Zie B. Rovers, Klassenjustitie. Overzicht of punishment and estimated likelihood of a gewoond om de gestandaardiseerde check-
D. Vermeulen, ‘Judgment and memory of a van onderzoek naar selectiviteit in de repeated offense’, Psychological Reports list in te vullen. Voordat zittingen werden
criminal act: The effects of stereotypes and Nederlandse Strafrechtketen, Rotterdam: 1999-85,p. 246-248. bijgewoond is een uitgebreide uitleg bij de
cognitive load’, European Journal of Social Erasmus University 1999. 22. Zie bijv. B.C. Everett en R.S. Nienstedt , checklist gegeven. Van elke volledige zitting
Psychology 1999-29, p. 191-202. 19. D. Steffensmeier en S. Demuth, ‘Ethni- ‘Race, remorse, and sentence reduction: Is is een checklist ingevuld.
16. D. Steffensmeier, J.T. Ulmer en J.H. city and sentencing outcomes in U.S. fede- saying your sorry enough?’, Justice Quar- 25. Daarnaast komt het voor dat er een
Kramer, ‘The interaction of race, gender ral courts: Who is punished more harshly?’, terly 1999-16(1), p. 99-122. ontbrekende waarde is op een van de varia-
and age in criminal sentencing: The punish- American Sociological Review 2000-65(5), 23. Centraal Bureau voor de Statistiek. belen in ons model. Zo zijn er bijvoorbeeld
ment cost of being young, Black and male’, p. 705-729. Geraadpleegd op 15-04-2011 via http:// 10 zaken waarin het onbekend was of de
Criminology 1998-36, p. 763-797 en D. 20. K.A. Drass en J.W. Spencer, ‘Accounting statline.cbs.nl/StatWeb/publication/?DM=S verdachte een Nederlands of buitenlands
Steffensmeier, J. Kramer en C. Streifel, for pre-sentencing recommendations: Typo- LNL&PA=37961&D1=1- uiterlijk had. Indien er sprake is van een
‘Gender and imprisonment decisions’, Cri- logies and probation officers’ theory of 7&D2=0&D3=0&D4=0&D5=5- ontbrekende waarde wordt die verdachte
minology 1993-31, p. 411-446. office’, Social Problems 1987-34(3), p. 15&HDR=T&STB=G1,G2,G3,G4&VW=T. verwijderd uit de logistische regressieanalyse.
17. P. Linckens en J. de Looff, Gevangenis- 277-293. 24. Verschillende studenten van de oplei-
wezen in getal, Den Haag: Dienst Justitiële 21. R.B. Pipes en M. Allessi, ‘Remorse and a ding Criminologie (Universiteit Leiden)
inrichtingen 2010. previously punished offense in assignment hebben zittingen van de politierechter bij-
729
308
Wetenschap......
De persoon van de verdachte 2) verdachte heeft een buitenlands uiterlijk en spreekt

Tijdens de strafzittingen is geobserveerd of de verdachte Nederlands;
Nederlands sprak of niet. Van alle verdachten spraken 315 3) verdachte heeft een buitenlands uiterlijk en spreekt
verdachten de Nederlandse taal en spraken 47 een andere niet de Nederlandse taal.28
taal.26 Uit de gegevens blijkt dat niet-Nederlands sprekende Bij het observeren van gedrag van de verdachten
daders vergelijkbare delicten plegen als Nederlands spreken- tijdens de zittingen is onderscheid gemaakt tussen
de daders; overwegend (winkel)diefstal. De niet-Nederlands geen spijt betuigen, een beetje spijt betuigen, erg spijt
sprekende daders vormen dus geen selecte groep op basis betuigen, en heel erg spijt betuigen. Dit betrof (natuur-
van het type delict waarvoor zij zijn veroordeeld. Verder zijn lijk) de eigen, subjectieve interpretatie van de observan-
de niet-Nederlands sprekende daders ook naar verhouding in ten.29 Om de betrouwbaarheid van dit kenmerk te ver-
dezelfde rechtbanken berecht als de Nederlands sprekende groten is ervoor gekozen slechts te onderzoeken of het
daders; overwegend Den Haag en Amsterdam. Deze daders al dan niet betuigen van spijt de kans op een onvoor-
vormen dus ook geen selecte groep op basis van de recht- waardelijke gevangenisstraf beïnvloedt. In totaal waren
bank waarin die strafzittingen hebben plaatsgevonden.27 er 149 verdachten die geen spijt hadden betuigd tijdens
In de beoordeelde zaken gaat het om 185 verdach- de strafzitting en 212 verdachten die wel spijt hadden
ten met een Nederlands uiterlijk en 170 met een buiten- betuigd. Verder waren van de 365 verdachten slechts 45
lands uiterlijk. Om te onderzoeken of taal een effect heeft vrouw.
naast het hebben van een buitenlands uiterlijk zijn deze
twee kenmerken als volgt samengevoegd: Resultaten
1) verdachte heeft een Nederlands uiterlijk en spreekt Tabel 1 laat zien dat een typische dader die een gevan-
Nederlands; genisstraf krijgt opgelegd in voorlopige hechtenis heeft
gezeten, een strafblad heeft, in de leeftijdscategorie 31
Andersge- Gevangenis- tot 40 jaar valt, is veroordeeld voor gekwalificeerde dief-
straften gestraften
stal en een zitting heeft gehad die in Amsterdam
Percentage Percentage
plaatsvond.30 Met name daders die een delict hadden
Persoon van de verdachte
gepleegd dat is ingedeeld in de categorie ‘overige delic-
Vrouw (N = 45) 91,1 8,9
ten’, zijn veroordeeld tot een andersoortige straf dan
Man (N = 320) 75,7 24,3
onvoorwaardelijke gevangenisstraf. De categorie ‘overige
Geen spijt betuigd (N = 149) 76,5 23,5
Wel spijt betuigd (N = 212) 78,8 21,2
delicten’ bestaat voor het merendeel uit wegenverkeers-
Ned. uiterlijk en spreekt Ned. (N = 182) 89,0 11,0
wetdelicten. Daarnaast is ook te zien dat met name ver-
Buitenlands uiterlijk en spreekt Ned. (N = 72,0 28,0 dachten die ouder zijn dan 50 jaar en van wie de zitting
125) plaatsvond in Haarlem, minder vaak worden veroor-
Buitenlands uiterlijk en spreekt geen Ned. 44,4 55,6 deeld tot een gevangenisstraf.
(N = 45) Tabel 1 laat verder zien dat mannen vaker wor-
Controlevariabelen den veroordeeld tot een onvoorwaardelijke gevange-
Rechtbank: nisstraf dan vrouwen. Daarnaast is het aandeel van
Amsterdam (N = 87) 64,4 35,6 de verdachten die geen spijt betuigen enigszins gro-
Den Haag (N = 142) 80,3 19,7 ter in de tot gevangenisstraf veroordeelden in verge-
Haarlem (N = 43) 95,3 4,7 lijking met verdachten die wel spijt betuigen. Wan-
Utrecht (N = 57) 73,7 26,3
Overige rechtbanken (N = 36) 86,1 13,9
Type delict:
Diefstal (N = 70)
Licht geweldsdelict (N = 49)
57,1
87,8
42,9
12,2
Het al dan niet betuigen van spijt
Bedreiging (N = 15)
Gekwalificeerde diefstal (N = 18)
73,3
27,8
26,7
72,2
beïnvloedt de kans op een
Opium (N = 23)
Overige delicten (N = 148)
78,3
92,6
21,7
7,4
onvoorwaardelijke gevangenisstraf
Zwaar geweldsdelict (N = 19) 78,9 21,1
Overig vermogensdelict (N = 23) 65,2 34,8
niet significant
Overige controlevariabelen:
Niet in voorlopige hechtenis (N = 270) 94,8 5,2
Wel in voorlopige hechtenis (N = 93) 28,0 72,0
Verdachte heeft geen strafblad (N = 94) 86,2 13,8 neer we de resultaten van het uiterlijk en de taal van
Verdachte heeft strafblad (N = 255) 73,3 26,7 de verdachte bestuderen, vallen twee bevindingen op.
Delict alleen gepleegd (N = 299) 80,3 19,7 Met name de groep die er Nederlands uitziet en ook
Delict samen gepleegd (N = 66) 66,7 33,3 Nederlands spreekt, is ondervertegenwoordigd bij de
Verdachte is 18-30 jaar (N = 152) 78,9 21,1 tot gevangenisstraf veroordeelden en juist die daders
Verdachte is 31-40 jaar (N = 93) 69,9 30,1 die er buitenlands uitzien en ook niet de Nederlandse
Verdachte is 41-50 jaar (N = 77) 77,9 22,1 taal spreken, zijn oververtegenwoordigd bij de popu-
Verdachte is ouder dan 50 jaar (N = 37) 89,2 10,8 latie van gevangenisgestraften. Deze bevindingen
Tabel 1: Beschrijvende resultaten van andersgestraften (N = 284) en sluiten aan bij de verwachtingen die eerder waren
gevangenisgestraften (N = 81) geformuleerd.
730
309
Echter, bij interpretatie van de gegevens uit tabel 1 Exp (B) Sign S.E.
is geen rekening gehouden met verschillende kenmer- Onafhankelijke variabelen
ken waarvan bekend is dat zij een rol spelen bij de straf- Vrouw ,151 ** ,832
toemetingsbeslissing, namelijk het type delict dat is Wel spijt betuigd 1,527 ,497
gepleegd, het al dan niet aanwezig zijn van justitiële Ned. uiterlijk en spreekt Ned. ref. ref. ref.
documentatie, of de verdachte in voorlopige hechtenis Buitenlands uiterlijk en spreekt Ned. 5,251 *** ,593
heeft gezeten, de rechtbank waar de zitting plaatsvond, Buitenlands uiterlijk en spreekt geen Ned. 20,654 *** ,797
het aantal feiten waarvan de persoon wordt verdacht en Controlevariabelen
of het delict alleen is gepleegd of samen. In een aparte Rechtbank:
analyse (een zogeheten logistische regressieanalyse) die Amsterdam ref. ref. ref.
wij uitvoerden, is dat wel gedaan (zie tabel 2). Op die Den Haag 3,018 * ,666
wijze zijn uitspraken te doen over de zelfstandige Haarlem ,792 1,156
invloed van kenmerken van de persoon van de verdach- Utrecht 9,215 *** ,804
te op de straftoemetingsbeslissing.31 Overige rechtbanken 1,944 ,849
De resultaten in tabel 2 laten zien dat de kans op Type delict:
een onvoorwaardelijke gevangenisstraf groter is voor Diefstal ref. ref. ref.
mensen die in voorlopige hechtenis hebben gezeten, Licht geweldsdelict ,205 ** ,772
voor daders tussen de 31 tot 40 jaar oud in vergelijking Bedreiging 2,007 1,093
met de jongste en oudste verdachten, als de veroorde- Gekwalificeerde diefstal 16,587 *** 1,014
ling plaatsvond in de Rechtbank Utrecht of Den Haag Opium ,953 ,913
en als de verdachte wordt veroordeeld voor gekwalifi- Overige delicten ,149 *** ,683
Zwaar geweldsdelict ,395 1,115
ceerde diefstal. Met name de lichte geweldsdelicten en
Overig vermogensdelict 1,382 ,867
categorie overige delicten, die voor een groot deel uit
Overige controlevariabelen:
wegenverkeerswetdelicten bestaat, leiden tot een lagere
Wel in voorlopige hechtenis 64,909 *** ,577
kans op een onvoorwaardelijke gevangenisstraf.
Verdachte heeft strafblad 2,125 ,659
Verder blijkt uit de resultaten van tabel 2 dat ken-
Delict samen gepleegd ,757 ,613
merken van de persoon van de verdachte ook zelfstan-
Aantal feiten waarvoor veroordeeld ,719 ,305
dig van invloed zijn op de kans op een onvoorwaarde-
Verdachte is 18-30 jaar ref. ref. ref.
lijke straf. Vrouwen worden minder snel veroordeeld
Verdachte is 31-40 jaar 2,646 * ,574
tot een onvoorwaardelijke gevangenisstraf. De odds
Verdachte is 41-50 jaar 1,438 ,616
voor vrouwen om veroordeeld te worden tot een Verdachte is ouder dan 50 jaar ,660 1,059
onvoorwaardelijke gevangenisstraf blijken ruim 6,5 Pseudo R2 (Nagelkerke) 72,1%
keer lager in vergelijking met mannen. Daarnaast blij- * p < .10; **p < .05; ***p<.01
ken daders met een Nederlands uiterlijk die ook de
Tabel 2: Logistische regressieanalyse onvoorwaardelijke gevangenis-
Nederlandse taal spreken de minste kans te hebben straf (N = 333)
om veroordeeld te worden tot een gevangenisstraf. In
vergelijking met die groep zijn de odds om veroor- uit dat het al dan niet betuigen van spijt de kans op
deeld te worden tot een onvoorwaardelijke gevangenis- een onvoorwaardelijke gevangenisstraf niet significant
straf voor daders met een buitenlands uiterlijk die de beïnvloedt. De kans op een onvoorwaardelijke gevan-
Nederlandse taal spreken ruim vijf keer hoger. De genisstraf is dus niet anders voor daders die wel spijt
groep die echter nog zwaarder wordt gestraft, is de betuigen in vergelijking met daders die geen spijt
groep daders met een buitenlands uiterlijk die niet de betuigen.
Nederlandse taal spreekt. De odds voor hen om veroor-
deeld te worden tot een onvoorwaardelijke gevangenis- Conclusie en discussie
straf zijn ruim twintig keer hoger in vergelijking met Het doel van deze studie was om voor de Nederlandse
mensen met een Nederlands uiterlijk. Alle hiervoor situatie te onderzoeken in hoeverre bepaalde kenmerken
beschreven resultaten zijn statistisch significant, zoals van de persoon van de verdachte een eigenstandige
ook in tabel 2 is weergegeven. Verder wijst de analyse invloed hebben op straftoemeting. Met name hebben wij
26. Van zes personen die Nederlands spre- 27. Deze resultaten zijn niet gepresenteerd observaties zijn gedaan, zijn samengevoegd den zijn lager dan 1,5. De laagste en hoog-
ken en twee personen die niet de Neder- in dit artikel. Geïnteresseerde lezers kunnen in de categorie ‘overig’. Voor de overzichte- ste waarde van de bivariate correlaties zijn
landse taal spreken, is bekend dat zij deze gegevens opvragen bij de auteurs. lijkheid is ervoor gekozen het aantal feiten respectievelijk 0,0 en 0,7. De correlatie
gebruik hebben gemaakt van het zwijg- 28. Er waren slechts twee personen met een waar de dader voor is veroordeeld niet op tussen voorlopige hechtenis en een onvoor-
recht. Van de overige verdachten die niet Nederlands uiterlijk die niet de Nederlandse te nemen in deze tabel. Geïnteresseerden waardelijke gevangenisstraf is 0,7. De resul-
Nederlands spreken, is niet bekend of er taal spraken. Deze personen zijn verwijderd kunnen deze resultaten opvragen bij de taten van tabel 2 veranderen niet van rich-
een tolk aanwezig was of welke taal ze dan voor vervolganalyses aangezien de groep te auteurs. ting en significantie wanneer voorlopige
wel spraken. De focus in dit onderzoek is klein is. 31. In een aparte analyse is onderzocht of hechtenis niet wordt opgenomen in het
daarmee beperkt tot de invloed van het al 29. Observanten hebben op basis van eigen de verschillende kenmerken in het model model.
dan niet spreken van de Nederlandse taal interpretatie aangegeven in hoeverre de problematisch onderling met elkaar samen-
op de kans op een onvoorwaardelijke verdachte spijt betuigd heeft. hangen. De resultaten van die analyse laten
gevangenisstraf. 30. De rechtbanken waar minder dan tien zien dat dit niet het geval is. Alle VIF-waar-
731
310
Wetenschap......
ons gericht op de rol van het al dan niet betuigen van te invloed hebben op de kans op een onvoorwaardelijke
spijt en het gecombineerde kenmerk van een al dan niet gevangenisstraf.
Nederlands uiterlijk met het al dan niet spreken van de De bevinding dat vrouwen minder kans hebben op
Nederlandse taal bij het bepalen van de straf. Bij het een onvoorwaardelijke gevangenisstraf sluit aan bij
vaststellen van die relaties werd rekening gehouden met bevindingen in eerder onderzoek33 en kan verklaard wor-
andere kenmerken waarvan bekend is dat zij een rol den aan de hand van de ‘focal concerns’ theorie die ver-
kunnen spelen bij de straftoemeting, zoals delictsken- onderstelt dat rechters vrouwen als minder gevaarlijk
merken en kenmerken van criminele geschiedenis. Voor inschatten. Het effect van het al dan niet spreken van de
Nederlandse taal is multi-interpretabel. Ten eerste kan
het effect van het al dan niet spreken van de Nederland-
se taal door de verdachte worden verklaard aan de hand
Het zwaarst gestraft werden de van bewust of onbewust gebruik van (negatieve) stereo-
typeringen door rechters om de gevaarlijkheid van de
daders met een buitenlands uiterlijk dader te bepalen. Ten tweede kan het niet spreken van
de Nederlandse taal het direct uitdrukken van emoties
die ook niet de Nederlandse taal en gevoelens belemmeren. Ten derde kan een andere,
meer praktische, reden worden aangedragen voor het
spraken tijdens de strafzitting substantiële effect van het al dan niet spreken van de
Nederlandse taal door de verdachte. Indien verdachten
de Nederlandse taal niet beheersen, zal een rechter wel-
licht geen taakstraf willen opleggen vanwege de proble-
Nederland, maar ook internationaal is dit voor zover ons men die bij de tenuitvoerlegging daarvan kunnen ont-
bekend de eerste studie die dat doet op basis van directe staan door het niet beheersen van het Nederlands. Bij
observaties tijdens strafzittingen en kwantitatieve analy- het wegvallen van deze strafoptie zou de rechter dan eer-
ses. der geneigd zijn een gevangenisstraf op te leggen. Indien
De resultaten van deze studie zijn gebaseerd op de taakstraf die anders opgelegd zou worden, relatief
observaties tijdens strafzittingen bij de politierechter in gezien kort is, zou het echter ook mogelijk zijn om in
Nederland in het voorjaar van 2010 (N = 333). Voor deze een dergelijk geval eerder een geldboete op te leggen.
studie zijn alleen de zittingen geselecteerd waar de ver- Het is dus onduidelijk hoe dit de kans op een onvoor-
dachte bij aanwezig was. Onze bevindingen lieten zien dat waardelijke gevangenisstraf precies beïnvloedt.
van de controlevariabelen het type delict en het al dan Het betuigen van spijt tijdens de strafzitting heeft
niet in voorlopige hechtenis hebben gezeten de kans op geen effect op de straftoemeting in ons onderzoek, ter-
een onvoorwaardelijke gevangenisstraf beïnvloeden. Daar- wijl dit op basis van de theorie wel de verwachting was.
naast vonden we ook verschillen in de kans op een Ook in eerder onderzoek met gegevens uit de Verenigde
onvoorwaardelijke gevangenisstraf tussen rechtbanken; Staten is aangetoond dat het betuigen van spijt van
wat duidt op regionale verscheidenheid.32 Deze gevonden invloed kan zijn op de gepercipieerde gevaarlijkheid van
effecten sluiten aan bij bevindingen in eerder onderzoek de dader34 en dat het betuigen van spijt in zaken waarin
en de ‘focal concerns’ theorie. de dader verdacht wordt van serieuze aanranding leidt
De hoofdbevindingen van het onderzoek lieten zien tot een lagere strafeis.35 Een mogelijke verklaring waar-
dat daders met een Nederlands uiterlijk die ook de Neder- om in het huidige onderzoek geen effect werd gevonden,
landse taal spraken tijdens de zitting de minste kans had- is onze beperking tot zittingen van de politierechter
den op een gevangenisstraf. Het zwaarst gestraft werden waar relatief minder ernstige delicten worden behan-
de daders met een buitenlands uiterlijk die ook niet de deld. Het zou kunnen zijn dat het betuigen van spijt bij
Nederlandse taal spraken tijdens de strafzitting. De odds bepaalde delicten wel effect heeft op de straftoemeting,
op een onvoorwaardelijke gevangenisstraf voor deze laat- bijvoorbeeld in ernstigere zaken als zware mishandeling,
ste groep zijn ruim twintig keer hoger in vergelijking met verkrachting, doodslag en moord. Om meer inzicht te
mensen met een Nederlands uiterlijk en ongeveer twee geven in het effect van spijt op straftoemeting zou toe-
keer hoger in vergelijking met mensen met een buiten- komstig onderzoek zich kunnen richten op zaken waarin
lands uiterlijk die de Nederlandse taal spreken. De resulta- ook deze ernstigere delicten behandeld worden. Daar-
ten lieten verder zien dat de odds op een onvoorwaardelij- naast hebben wij spijt alleen gemeten door te onderzoe-
ke gevangenisstraf voor vrouwen ruim 6,5 keer lager is in ken in hoeverre het expliciet betuigen van spijt tijdens
vergelijking met mannen. Dit zijn zelfstandige effecten, de strafzitting invloed heeft op straftoemeting. Voor ver-
dus onafhankelijk van andere kenmerken van de zaak volgonderzoek zou het goed zijn om dit aan te vullen
waarvoor gecontroleerd is in onze analyses, zoals het aan- met een maat voor geloofwaardigheid van de spijtbetui-
tal en het type delict(en) en het al dan niet hebben van ging en verduidelijking waarvan de verdachte precies
een strafblad. Het al dan niet betuigen van spijt door de aangeeft spijt te hebben.
verdachte tijdens de strafzitting heeft geen invloed op de Met betrekking tot taal heeft dit onderzoek zich
kans op een onvoorwaardelijke gevangenisstraf. Daarnaast slechts gericht op het al dan niet spreken van de Neder-
laten de resultaten zien dat van de controlevariabelen het landse taal. Voor vervolgonderzoek zou taal uitgebreider
al dan niet hebben van een strafblad, het aantal feiten onderzocht kunnen worden waarbij ook aandacht wordt
waar de verdachte voor is veroordeeld en het uitgangsde- geschonken aan, bijvoorbeeld, verdachten die gebrekkig
lict al dan niet samen te hebben gepleegd geen significan- Nederlands spreken of een sterk buitenlands accent heb-
732
311
Stereotyperingen spelen juist een rol bij beslissingen
die onder tijdsdruk genomen moeten worden
ben. Verder zouden we op basis van de theorie verwach- Toekomstig onderzoek zal moeten aantonen of ook
ten dat een negatieve stereotypering meer geldt voor tijdens zittingen van de meervoudige kamer deze effecten
mensen die bijvoorbeeld een niet-westerse taal spreken gevonden worden. In dergelijke zittingen hoeft de rechter
in vergelijking met een verdachte die bijvoorbeeld niet direct een beslissing te nemen over de straftoemeting
Engels spreekt. Bij vervolgonderzoek zou daarom ook zoals wel het geval is bij de politierechter. Dat heeft wel-
onderscheid gemaakt kunnen worden naar welke taal de licht consequenties voor de beslissingen die worden geno-
verdachte spreekt tijdens de rechtszitting. Dit laatste men en de mogelijke rol van stereotypering daarbij.
geldt overigens ook voor het uiterlijk van de dader. Voor Immers, zoals hiervoor besproken, spelen stereotyperingen
vervolgonderzoek zou het aan te bevelen zijn om meer juist een rol bij beslissingen die onder tijdsdruk genomen
onderscheid te maken tussen verschillende etniciteiten moeten worden. Bij straftoemeting door een meervoudige
en tevens gebruik te maken van kenmerken zoals het kamer kan daar veel minder sprake van zijn dan bij de
geboorteland van de dader en het geboorteland van de beslissing direct na het onderzoek ter terechtzitting door
ouders van de dader. een politierechter. Bovendien zijn er drie rechters bij de
Dat mannen en daders met een buitenlands uiterlijk straftoemetingsbeslissing betrokken in de meervoudige
die niet de Nederlandse taal spreken tijdens de strafzit- kamer; wat een mediërende werking kan hebben. Drie rech-
ting een substantieel verhoogde kans hebben op een ters zouden verschillende meningen kunnen hebben, zodat
onvoorwaardelijke gevangenisstraf zou op gespannen voet de uiteindelijke straf een compromis is van drie mogelijk
kunnen staan met het doel van consistentie in straftoe- verschillende ideeën over de straftoemeting. Voor toekom-
meting in Nederland. Ten behoeve van een consistentere stig onderzoek is het tevens belangrijk om invloeden van
straftoemeting zijn in Nederland sinds de jaren negentig rechterkenmerken op straftoemeting te onderzoeken.
verschillende richtlijnen geïntroduceerd.36 Daarnaast In eerder empirisch onderzoek naar straftoemeting
behoeft dit extra aandacht, omdat het gaat over de zwaar- zijn de mogelijke effecten van specifieke omstandigheden
ste straf die in Nederland opgelegd kan worden aan tijdens de zitting nog weinig onderzocht. De aandacht in
daders. Uit onderzoek is bekend dat gevangenisstraffen de onderzoekstraditie naar het verklaren van verschillen
(onbedoelde) effecten kunnen hebben, zoals een grotere in straftoemeting gaat vooral uit naar demografische ken-
kans om te scheiden37 en verminderde kansen op de merken van de dader en verschillen in straftoemeting tus-
arbeidsmarkt38. Deze relaties kunnen vervolgens leiden tot sen rechters en rechtbanken. Deze bijdrage toont het
een langdurige criminele carrière.39 De selectie van welke belang aan van wat zich op de zitting afspeelt in het straf-
daders een onvoorwaardelijke gevangenisstraf opgelegd toemetingsonderzoek. Toekomstig onderzoek zou zich
krijgen, heeft dus ook consequenties voor de verdere verder moeten richten op het effect van zittingskenmer-
levens van deze individuen. ken op straftoemeting.
32. Zie voor meer informatie over regionale 34. K.A. Drass en J.W. Spencer, ‘Accounting 1999-85, p. 246-248. Criminology 2010-26, p. 269-300.
verscheidenheid ook J.T. Ulmer, Social for pre-sentencing recommendations: Typo- 36. G.K. Schoep en P. M. Schuyt, Instru- 38. D. Pager, ‘The mark of a criminal
Worlds of Sentencing: Court Communities logies and probation officers’ theory of menten ter ondersteuning van de rechter record’, American Journal of Sociology
under Sentencing Guidelines, Albany, NY: office’, Social Problems 1987-34(3), p. bij de straftoemeting, Leiden, E.M. Meijers 2003-108, p. 937–975.
State University of New York Press 1997. 277-293. Instituut 2005. 39. D.S. Nagin, F.T. Cullen en C.L. Jonson.
33. K. Daly en R.L. Bordt, ‘Sex effects and 35. R.B. Pipes en M. Allessi, ‘Remorse and a 37. R. Apel, A.A.J. Blokland, P. Nieuwbeerta ‘Imprisonment and re-offending’, in: M.
sentencing: An analyses of the statistical previously punished offense in assignment en M. van Schellen, ‘The impact of imprison- Tonry (Ed.), Crime and Justice: A Review of
literature’, Justice Quarterly 1995-12(1), p. of punishment and estimated likelihood of a ment on marriage and divorce: A risk set Research, Volume 38, Chicago: University
141-175. repeated offense’, Psychological Reports matching approach’, Journal of Quantitative of Chicago Press 2009, p. 115-200.
733
312
Week 7
313
314
“IT’S NOT ABOUT THE MONEY!”: A THEORY ON
MISCONCEPTIONS OF PLAINTIFFS’ LITIGATION AIMS
Tamara Relis*
This Article examines from a new angle a long-standing debate on two

central questions of the legal system: why do plaintiffs sue, and what do they
seek from litigation? Legal research has documented various extra-legal aims
or non-economic agendas of plaintiffs who commence legal proceedings for
various case-types.1 However, current debates have failed to address this issue
in depth from the perspectives of plaintiffs themselves, subsequent to lawyers
conditioning them on “legal system realities” and translating their disputes
into legally cognizable compartments. Nor have understandings of plaintiffs’
aims been examined from the perspectives of defense lawyers. These are
significant gaps in the knowledge, as plaintiffs’ objectives directly impact
their experiences in litigation and litigation-linked processes such as
mediation. Likewise, attorneys’ approaches and conduct throughout litigation
and mediation processes are premised upon their basic understandings of what
those who commenced these suits want.
In providing new empirical research, I offer disconcerting evidence of the
surprising degree to which disparate perceptions of plaintiffs’ litigation aims
exist as between plaintiffs and attorneys—at times even between attorneys and
* Dr. Tamara Relis, Fellow, Columbia University Law School and London School of Economics,
Department of Law. This work has been presented at the American Law & Society Association Annual
Meeting (2004), the London School of Economics and the Kyoto Law School Symposium on “The
Participation of Non-legal Professionals and Citizens in Judiciary” in Japan (2004). I would like to thank
Simon Roberts, Richard Abel, Michael Palmer, Kendall Thomas, Howard Erlanger, and Craig McEwen for
their insightful comments and suggestions. I am indebted to the lawyers, disputants and mediators involved
in this project. Although their names must remain confidential, without them this research would not have
been possible. I would also like to thank Columbia/LSE and the Economic & Social Research Council for
their financial support: award PTA-026-27-0979. This article supersedes any other prior published version.
1. I have termed non-monetary objectives “extra-legal” as monetary damages are discussed within
legal rules and the traditional world of legal practice. Some of the literature on legal policy, victims’ rights,
mediation and legal consciousness refers to a number of the non-monetary litigation aims I have termed
“extra-legal” as “legal,” e.g. plaintiffs’ desires to obtain acknowledgment of error. See generally PATR ICIA
EWICK & SUSAN SILBEY, THE COMMON PLACE OF LAW (1998); David Engel, The Oven Bird’s Song:
Insiders, Outsiders and Personal Injuries in an American Community, 18 LAW & SOC ’Y REV . 551 (1984).
However, I have used this term to highlight an important recurrent theme within this work and to reflect
the views and perceptions of the lawyers within my sample. In their opinion, these aims are not within the
province of the law or legal practice, and thus are “extra-legal.”
701
315
702 UNIVERSITY OF PITTSBURGH LAW REVIEW [Vol. 68:701
their own clients. As such, this article presents a picture of discontinuity

where there appears to be little understanding of plaintiffs’ motivations
amongst the lawyers involved in their cases. Employing a novel angle of
juxtaposing the views of all sides within the same or similar fatality and
medical injury cases, I show that irrespective of allegiances the bulk of
attorneys understand that plaintiffs sue solely or predominantly for money.
Even many plaintiff lawyers, who are more aware of their clients’ extra-legal
aspirations, swiftly translate these objectives into finance alone, as “that is all
the legal system can provide.” Yet simultaneously, virtually all plaintiffs
vehemently insist, “it is not about the money!” with only a minority saying
financial compensation was even a secondary aim. Instead, plaintiffs’
articulations of why they sued and what they sought from the legal
system—irrespective of whether their cases had been litigating for months or
years—were thickly composed of extra-legal aims of principle. Yet,
notwithstanding any needs or desires for monetary compensation, plaintiffs’
objectives of obtaining admissions of fault, prevention of recurrences,
retribution for defendant conduct, answers, apologies and acknowledgments
of harm remained invisible to virtually all lawyers throughout the duration of
the processing of their cases.
To explain this phenomenon, I present a theory that argues that these
“parallel worlds of understanding” occur largely due to the institutional
framework of the civil justice system coupled with the practical and economic
realities of legal practice that result in dispute transformation. In offering two
proposals to increase attorneys’ understanding of plaintiffs’ extra-legal
agendas and needs within litigation processes, this Article implicitly argues
that the operating economic premises of the civil justice system require urgent
and serious reevaluation.
316
2007] “IT’S NOT ABOUT THE MONEY!” 703
Table of Contents
I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704
II. Opposing Partisans’ Similar Comprehensions of What
Plaintiffs Want . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 710
A. Physician Lawyers: It’s Only About Money . . . . . . . . . . . . . 713
B. Hospital Lawyers: It’s Mostly About Money . . . . . . . . . . . . . 715
C. Plaintiff Lawyers: It’s Mostly About Money, But Also
Other Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 718
III. Plaintiffs: It’s about Principles, Not Money! . . . . . . . . . . . . . . . . 720
A. Plaintiffs’ Articulations of Their Motives and Litigation
Aims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 721
B. Case Studies Revealing Parallel Worlds of Understanding
Plaintiffs’ Aims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 728
IV. “System Conditioning” by Lawyers . . . . . . . . . . . . . . . . . . . . . . . . 733
V. An Institutional Theory on Lawyers’ Understandings of
Plaintiffs’ Litigation Aims . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 740
VI. Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 742
317
I. INTRODUCTION
The phenomenon of dispute transformation, formulated now long ago,2

is generally accepted and remains largely unchallenged within legal
scholarship.3 I present it here as a twofold process. First, notwithstanding
plaintiffs’ dispute descriptions and initial expressed desires, lawyers condition
clients on “legal system realities” and persuade them to aim for what they
view as legally realistic.4 Attorneys then reframe litigants’ dispute
2. See William Felstiner et al., The Emergence and Transformation of Disputes: Naming,
Blaming, Claiming, 15 LAW & SOC ’Y REV . 631, 631-32, 650-51 (1981).
3. Although dispute transformation in medical malpractice cases has not been examined in depth,
it has been found in various other case types. These include divorce, poverty law, consumer, general injury,
small claims and harassment. See, e.g., JAMES WHITE, JUSTICE AS TRANSLATION : AN ESSAY IN CULTURAL
AND LEGAL CRITICISM 20-21, 260-61, 269 (1990); Carl Hosticka, We Don’t Care About What Happened,
We Only Care About What Is Going To Happen: Lawyer-Client Negotiations Of Reality, 26 SOC . PROBS.
600, 600-04 (1979) (utilizing solely observation data); Clark Cunningham, The Lawyer as Translator,
Representation as Text: Towards an Ethnography of Legal Discourse, 77 CORNE LL L. REV . 1298,
1339-57, 1367-85 (1992); Tamara Relis, Civil Litigation from Litigants’ Perspectives: What We Know
and What We Don’t Know About the Litigation Experience of Individual Litigants, in STUD . L., POL. &
SOC ’Y 162-63, 167-68, 193 (Austin Sarat & Patricia Ewick eds., 2002). See generally Anthony Alfieri,
Reconstructive Poverty Law Practice: Learning Lessons of Client Narrative, 100 YALE L.J. 2111 (1991);
Christopher Gilkerson, Poverty Law Narratives: The Critical Practice and Theory of Receiving and
Translating Client Stories, 43 HASTINGS L.J. 883 (1992); Austin Sarat & William Felstiner, Law and
Strategy in the Divorce Lawyer’s Office, 20 LAW & SOC ’Y REV . 93, 116 (1986); Jerome Bruner, A
Psychologist and the Law, 37 N.Y.L. SCH . L. REV . 173, 182 (1992); JOHN CONLEY & WILLIAM O’BARR,
RULES VERSUS RELATIONSHIPS: THE ETHNOGRAPHY OF LEGAL DISCOURSE (1990); MAUREEN CAIN , THE
GENERAL PRACTICE LAWYER AND THE CLIENT: TOWARDS A RADICAL CONCEPTION (1983). Legal
transformation has also been found in various jurisdictions. For instance, in discussing the Kenyan legal
system it has been noted that lawyers have “elaborated highly sophisticated techniques for explaining
disputes in terms of the norms involved . . . diverging from that employed by society at large.” See Richard
Abel, A Comparative Theory of Dispute Institutions in Society, 8 LAW & SOC ’Y REV . 234, 271 (1974).
Likewise, utilizing case studies from both western and non-western settings (e.g. Tanzania), it has been
found that “disputes altered as they were processed in response to the interests of various participants . . .”
with “narrowing being the most common process of dispute transformation.” See Lyn Mather & Barbra
Yngvesson, Language, Audience, and the Transformation of Disputes, 15 LAW & SOC ’Y REV . 775, 778,
780, 789, 796 (1981). The legal narrowing of disputes “determined by the needs and purposes of the legal
process” was also noted in Pasargada. See Boaventura de Sousa Santos, The Law of the Oppressed: The
Construction and Reproduction of Legality in Pasargada, 12 LAW & SOC ’Y REV . 5, 18 (1977).
4. Divorce clients have been found to repeatedly attempt to broaden the dialogue with their lawyers.
However, lawyers’ dispute interpretations reflecting their own sense of legal reality generally prevailed.
See AUSTIN SARAT & WILLIAM FELSTINER , DIVORCE LAWYERS AND THEIR CLIENTS 85, 106-07 (1995)
[hereinafter SARAT & FELSTINER , DIVORCE LAWYERS] (relying upon observations of conversations between
divorce lawyers and clients in two U.S. states); Austin Sarat & William Felstiner, Law and Social
Relations: Vocabularies of Motive in Lawyer/Client Interaction, 22 LAW & SOC ’Y REV . 737, 741-42,
766-67 (1988) [hereinafter Sarat & Felstiner, Law and Social Relations]; Sarat & Felstiner, supra note 3,
318
experiences, feelings and extra-legal aims to fit into legally cognizable

compartments suitable for processing within the legal system.
I argue here, however, that current debates have failed to address two
important issues within this phenomenon. First, as with many aspects of
litigants’ perspectives, there has been little investigation from plaintiffs’
viewpoints during case processing how, if at all, dispute transformation affects
their understandings of their cases and what they seek from the justice system,
how they perceive what they hear from their lawyers, and whether their
objectives change throughout the litigation process.5 Second, there is little
knowledge of defense lawyers’ understandings of plaintiffs’ aims, and of
whether dispute transformation plays any part in their comprehensions.6 This
is significant, as attorneys’ approaches to their cases and their conduct during
litigation and litigation-linked processes such as mediation are premised upon
their basic understandings of what plaintiffs want. Likewise, for plaintiffs,
their motivations to sue and consequent litigation aims have a marked impact
upon their experiences during litigation and litigation-track mediations.7
Thus, understanding plaintiffs’ objectives is particularly important if civil
justice evaluations or reforms are to be responsive to their needs.
In offering findings from new empirical research into lawyers’ and
plaintiffs’ understandings within the same or similar medical malpractice
cases, this article argues that dispute transformation does affect defense
lawyers’ understandings of opposing plaintiffs’ objectives, yet it does little to
at 116-17.
5. Little is known about what litigants really want from the civil justice system, what they aim to
achieve and how their objectives may change. Consequently, we have little knowledge of whether litigants’
real aims are met by the realities of civil litigation. See generally HAZEL GENN , PATHS TO JUSTICE : WHAT
PEOPLE DO AND THINK ABOUT GOING TO LAW (1999) [hereinafter GENN , PATHS TO JUSTICE ]; Hazel Genn,
Access to Just Settlements: The Case of Medical Negligence, in REFORM OF CIVIL PROCEDURE: ESSAYS
ON “ACCESS TO JUSTICE ” (A. Zuckerman & R. Cranston eds., 1995) [hereinafter Genn, Access to Just
Settlements]; JOHN BALD WIN , SMALL CLAIMS IN THE COUNTY COURTS IN ENGLAND AND WALES (1997).
Even Sarat’s & Felstiner’s research into divorce lawyers and clients lacks much input from clients on their
understanding of what was occurring during their attorney-client relationships. Nor does it tell us whether
clients absorb and internalize what they hear from their lawyers and if they do, how they feel about this.
See Cunningham, supra note 3, at 1339-57, 1367-85.
6. Research has failed to examine in depth the practical effects of dispute transformation. See Relis,
7. Procedural justice studies and legal research suggest that the manner in which litigation
processes are subjectively perceived is affected by litigants’ expectations, which are linked to their aims.
See generally CONLEY & O’BARR, supra note 3; Tom Tyler et al., Maintaining Allegiance to Political
Authorities: The Role of Prior Attitudes and the Use of Fair Procedures, 33 AM . J. POL. SCI. 629 (1989);
William O’Barr & John Conley, Lay Expectations of the Civil Justice System, 22 LAW & SOC ’Y REV . 137
(1988).
319
extinguish plaintiffs’ own extra-legal litigation aspirations. I show how

plaintiffs interpret their lawyers’ explanations of the “monetary realities of the
legal system.” I also provide evidence of plaintiffs intermeshing “principles”
with pecuniary recompense. Thus, there may be some transformation in
plaintiffs’ goals, particularly on the surface. However, challenging a central
tenet of transformation theory, that a transformed dispute can actually become
the dispute,8 this article demonstrates that plaintiffs’ extra-legal aims of
principle do not dissipate after dispute transformation by their lawyers, nor
with the passage of time.9 In terms of the effects upon lawyers, the findings
here suggest practical consequences of dispute transformation, something
questioned in previous literature.10 Indeed, it has been argued that
transformed legal descriptions of conflicts commonly alter the nature of
disputes and may hold little meaning for litigants, resulting in remedies that
do not deal with their needs as they perceive them. 11 The findings here shed
further light on this issue, which has generally been neglected.12
In comparing attorneys’ and plaintiffs’ comprehension of plaintiffs’ aims,
I provide disconcerting evidence of the degree to which perceptions are
largely diverse. I demonstrate that, notwithstanding their different allegiances,
the bulk of lawyers for both plaintiffs and defendants have similar conceptions
of why plaintiffs have sued in particular cases—the reason being solely or
predominantly for money. Concurrently, I show how, notwithstanding any
needs or desires for pecuniary recompense,13 plaintiffs’ own articulations of
their litigation objectives are thickly composed of extra-legal aims of
principle, regardless of whether their cases have been in litigation for three
months or five years. However, claimants’ desires for acknowledgments of
harm, retribution for defendant conduct, admissions of fault, prevention of

9. It has been argued that litigants’ goals may change over time. See GENN , PATHS TO JUSTICE ,
supra note 5, at 11. Others have noted that litigants’ perceptions of their cases and needs may change
during litigation. See Austin Sarat, Alternatives in Dispute Processing: Litigation in a Small Claims
Court, 10 LAW & SOC ’Y REV . 341, 342, 342 n.7 (1976); SARAT & FELSTINER , DIVORCE LAWYERS, supra
note 4, at 53. As a separate issue, litigants’ perceptions of case facts have been found to change over time.
Felstiner et al., supra note 2, at 631. See generally CONLEY & O’BARR, supra note 3.
10. Relis, supra note 3, at 193.
11. See generally TAMARA RELIS , PARALLEL WORLDS OF DISPUTES AND MEDIATION: LITIGANTS,
LAWYERS AND GENDERED ACTORS’ PERSPECTIVES (forthcoming 2007) (which illustrates how this, in fact,
is the case); CONLEY & O’BARR, supra note 3 (examining small claims litigants); Relis, supra note 3, at
163; Hosticka, supra note 3 (research into poverty law cases).
12. See Cunningham, supra note 3, at 1300.
13. Which, of course, could include not only claims of compensation for pain and suffering and loss
of amenity, but also for reimbursement of expenditures incurred as a result of the injuries.
320
reoccurrences, answers and apologies remain invisible to most lawyers

throughout the litigation and mediation of these cases.14
As such, this article presents a picture of discontinuity in terms of basic
understandings of what these cases were about. This discontinuity appears to
be due, in part, to failed communication. Yet, instead of viewing this
discontinuity in terms of conflicting perceptions at a single moment, I argue
that it is part of a system that acts on those perceptions and channels behavior
in ways that reinforce those perceptions. So the structures of roles and
14. The full dataset derives from 131 in-depth interviews, questionnaires and observation files of
actors (plaintiffs, defendants, lawyers and mediators) involved in 64 medical malpractice cases that
underwent voluntary and court-linked mandatory mediations in three sites in Toronto (Ontario Court
Mandatory Mediation Program, ADR Chambers, College of Physicians and Surgeons). Specifically, there
are case interviews and questionnaire files from 17 plaintiffs (co-plaintiffs in any cases were interviewed
separately), 13 defendant physicians, 27 plaintiff lawyers (comprising views from seven general
practitioners & 11 specialist medical lawyers, being those who worked on 10 or more medical cases per
annum and amongst the top specialist lawyers in the region—nine discussed more than one case), 17
physician lawyers (with views from eight physician lawyers—four discussed more than one case), 23
hospital lawyers (comprising views of seven hospital lawyers—five discussed more than one case), 29
mediators (with views from 17 mediators, 10 being lawyers and seven being non-lawyers—seven discussed
more than one case) and two hospital representative/insurer interviews. All physician lawyers and hospital
lawyers were medical malpractice specialists. Of all the plaintiffs involved in the study, perhaps only one
or two, if any, might be viewed as “wealthy” in terms of being non-representative of plaintiffs generally.
This conclusion is based upon information obtained during the research process, e.g. from what plaintiffs
said during interviews and observed mediations including private caucuses, respondents’ questionnaire
responses about their homes, and their job descriptions as well as those of the principal financial providers
in each household. See generally Relis, supra note 11.
Interviews were conducted with 10 plaintiffs subsequent to voluntary mediations and seven plaintiffs
subsequent to mandatory mediations on the litigation-track. Mandatory court-linked mediations (similar
to many U.S. court-connected programs, although generally occurring prior to discovery) generally had to
take place within 90 days after the first defense had been filed, unless the court ordered otherwise. (ONT .
R. CIV . P. 24.1.09). Voluntary mediations at ADR Chambers generally occurred far later (often years later)
and close to trial (all cases were already on trial lists). See Julie Macfarlane, Culture Change? A Tale of
Two Cities and Mandatory Court-Connected Mediation, 2002 J. DISP. RESOL. 241, 244-45. Disputes that
did not settle at mediation continued to follow normal litigation routes. See Ontario Ministry of the
Attorney General, Ontario Mandatory Mediation Program at http://www.attorneygeneral.jus.gov
.on.ca/english/courts/manmed/rule24.asp (last visited Nov. 6, 2006). Plaintiffs would have commenced
litigation within one year of the disputed incidents as in Ontario the limitation period for medical
malpractice cases is only one year. [Limitation of Actions Act, R.S.A., ch. L-15, §§ 55(a), 56 (1980);
Health Disciplines Act, R.S.O., ch. H4, § 17 (1980); Hospitals Act, R.S.P.E.I., ch. H-10, § 13 (1988); The
Medical Profession Act 1981, 1980-81 S.S., ch. M-10.1, § 72]. College mediations, which were pre-
litigation, were not included in the analysis of plaintiffs’ litigation aims.
A number of the lawyers and mediators provided data on more than one case as these individuals were
repeatedly listed on the court files as acting in the various medical cases. This appeared to be due to the
relatively small size of the specialist medical-legal community in the region. Nonetheless, I obtained a wide
range of exposure to different mediators and lawyers. Accordingly, I am not aware of any bias resulting
from the cases included in the study.
321
expectations outlined here might be seen as creating some of the problems and
the perceptions. More specifically, I explain this discontinuity as a product
of a system that forces legal action for dollars partly because it is structured
that way, and partly because lawyers and insurance representatives understand
disputes in those narrow terms. These understandings are reinforced by the
behavior of plaintiffs, who have had their disputes transformed by both their
own lawyers and the non-responsiveness that a defense posture focused on
money creates.
In proffering detailed empirical data from actual plaintiffs on their
motivations and objectives, this article offers new insight into the critical area
of litigants’ non-economic goals for litigation.15 The findings also inform the
lawyering theory literature, answering questions such as “how do attorneys
understand what a particular client’s ‘problem’ is?” and “what ‘facts’ do
attorneys perceive versus those that escape them?”16 As such, this article
highlights the role of plaintiffs’ extra-legal agendas within the discontinuity
between claimants’ aims, lawyers’ understandings and realities, and the
pervasive economic assumptions of the civil justice system.17 In so doing, the
article underscores the significance of fully comprehending litigants’ extra-
legal agendas as an integral facet of the civil justice system.18
15. See Chris Guthrie, The Lawyer’s Philosophical Map and the Disputant’s Perceptual Map:
Impediments to Facilitative Mediation and Lawyering 6 HARV. NEGOT. L. REV . 145, 165-66 (2001); Genn,
Access to Just Settlements, supra note 5, at 11; John Conley & William O’Barr, Hearing the Hidden
Agenda: The Ethnographic Investigation of Procedure, 51 LAW & CONTEMP. PROBS., 181, 182-84, 196-97
(1988) [hereinafter Conley & O’Barr, Hearing the Hidden Agenda]; Neil Vidmar, The Small Claims Court:
A Reconceptualization of Disputes and an Empirical Investigation, 18 LAW & SOC ’Y REV . 515, 515
(1984); Charles Vincent et al., Why Do People Sue Doctors? A Study of Patients and Relatives Taking
Legal Action, 343 THE LANCET 1609, 1609-13 (1994). The data here shed light upon how lawyers and
disputants think and speak about the meaning of cases. They also represent talk about motives, which I
hope will be regarded as revealing and useful. I do not claim that this answers the ultimate psychological
question of why people do things. This is one of several windows on motivations, adding to the scant depth
of research on the needs of plaintiffs and why they sue.
16. Richard Sherwin, Lawyering Theory: An Overview What We Talk About When We Talk About
Law, 37 N.Y.L. SCH . L. REV . 9, 42-43, 48, 51 (1992).
17. See Conley & O’Barr, Hearing the Hidden Agenda, supra note 15, at 181, 182-84, 196-97.
Conley and O’Barr’s ethnographic research into small claims litigants in three U.S. states provides
extensive linguistic evidence of the significance of litigants’ hidden extra-legal, non-economic agendas.
Stressing the “importance, if not the pre-eminence of non-economic factors for litigants,” they conclude,
“the discontinuity between litigant agendas and the operating assumptions of the system may be a
fundamental source of dissatisfaction with the law.”
18. This research applies interpretive theory, drawing upon the agency-structure paradigm that
emerges in the works of Anthony Giddens and Pierre Bourdieu. As such, the focus is upon recovering
actors’ understandings and meanings whose actions are informed by structure—here, being litigation and
mediation processes. See PIERRE BOURDIEU , THE LOGIC OF PRACTICE 18, 53 (1990); PIERRE BOURDIEU ,
322
In an attempt to remedy the specific issues discussed here, I present two

proposals. First, direct dialogue early on in litigation between defense lawyers
and plaintiffs concerning plaintiffs’ litigation aims is necessary. Early
mediations present opportunities for litigants to articulate their extra-legal
realities and needs. This awareness should affect the way attorneys
understand and subsequently deal with these cases if they are not resolved
through mediation. Second, core law school curriculums for law students and
continuing legal education and ethical rules for attorneys must provide greater
emphasis upon litigants’ extra-legal dispute realities, needs and objectives
during case processing.
This article proceeds as follows. Part II examines similarities and
differences in the manner in which plaintiffs’ aims are understood across three
separate attorney groups: physicians’ attorneys; hospital attorneys; and
plaintiffs’ attorneys. Part III presents data on the dramatically disparate
motivations and litigation objectives articulated by plaintiffs within many of
the same disputes, most of which were rarely mentioned by the lawyers in
their cases. To illustrate the marked discontinuity of understandings of
plaintiffs’ aims, Section B of Part III presents three case studies juxtaposing
all actors’ discourse concerning their comprehensions of what plaintiffs
wanted. Part IV provides evidence of plaintiffs’ lawyers conditioning their
clients on the monetary realities of the legal system through the phenomenon
of system conditioning. It then examines plaintiffs’ understandings of this
phenomenon, as well as the fusion of principles and money within some
plaintiffs’ explanations of their aims. Part V offers an institutional theory to
explicate the discontinuity between lawyers’ and plaintiffs’ understandings.
OUTLINE OF A THEORY OF PRACTICE 3, 21, 72 (Richard Nice trans., Cambridge Univ. Press 1977);
ANTHONY GIDDENS, THE CONSTITUTION OF SOCIETY: OUTLINE OF THE THEORY OF STRUCTURATION 5, 25,
30 (1984); George Walsh, Introduction to ALFRED SCHUTZ , THE PHENOMENOLOGY OF THE SOCIAL WORLD ,
at xxi, xxiii, xxviii (George Walsh & Frederick Lehnert trans., Northwestern Univ. Press 1967); MAX
WEBER , WIRTSCHAFT UND GESELLSCHAFT 1 (A.M. Henderson & Talcott Parsons trans., J.C.B. Mohr 1957)
(1922). In contrast to positivism, for Giddens, social reality is the “ongoing construction of and by
knowledgeable actors whose recurrent interactions create, reproduce and transform the social world.” See
NICOS P. MOUZELIS, BACK TO SOCIOLOG ICAL THEORY: THE CONSTRUCTION OF SOCIAL ORDERS 19 (1991).
Yet, the constitution of agents and structures represents a duality because structure’s role in organizing
human conduct is not external to individuals. See id. at 26. Structure consists of recursively organized sets
of rules and resources comprising the “situated activities of human agents, reproduced across time and
space.” See GIDDENS, supra note 18, at 25. Not far from Giddens, Bourdieu’s theory of practice stipulates
that “each agent . . . is a producer and reproducer of objective meaning.” See BOURDIEU , OUTLINE OF A
THEORY OF PRACTICE , supra note 18, at 21. Thus, social practices are fundamental in understanding how
social structures are produced and reproduced, as practices constitute structures while also being determined
by structures. See id.
323
Finally, Part VI summarizes the findings and proposes two ways to address the
problems discussed.
II. OPPOSING PARTISANS’ SIMILAR COMPREHEN SIONS OF WHAT

PLAINTIFFS WANT
This Part examines the understandings of lawyers on all sides of medical

malpractice cases concerning the elementary issue of why plaintiffs
commenced these litigations, their goals, and their views of what these cases
were about.19 In terms of the legal and practical context of these cases, it is
generally accepted that despite marked increases in medical malpractice
litigation in the United States, Canada, Britain and elsewhere, very few
victims of medical accidents sue.20 Of those who do, only a minority receive
compensation through settlements, with even fewer obtaining court awards in
19. The research methodology was based predominantly on a qualitative paradigm, using a multiple
case study design. Although the nature of the sample excluded the possibility of meaningful regression or
multivariate analysis, a small amount of quantitative analysis was also undertaken, utilizing descriptive
statistics. This was done to assess percentages and proportions of particular responses and to evaluate, as
far as the data permitted, the strength of any associations between certain variables. Variables studied
include individuals’ views and perceptions, and their gender and actor positions in the litigation process.
Analysis of participants’ discourse was facilitated by the ATLAS.ti qualitative analysis computer program,
enabling grounded theory. The dataset was segmented into 5508 coded units and grouped into 263 code
families representing the main areas of focus. The integrity of the data was carefully preserved, and
triangulation of methods was utilized to enhance internal validity. As to external validity, natural groupings
representing social segmentation were used, enabling generalizations to wider populations, as is common
in qualitative research and similar projects. Additionally, by approaching all litigated medical malpractice
cases that mediated during the Fieldwork Period (May 2000 to October 2001) in the three sites, I conducted
a form of census. The number of individuals approached who consented to partake in the research was over
70%, a response rate considered reasonable in terms of generalizing to outside populations.
Of course, as in all such qualitative legal research a number of methodological weaknesses exist. The
fact that a random probability sample was impossible to obtain limits the generalizations that can be made
from the findings across populations. There are also inherent risks in extrapolating findings without
qualification from one location to another. However, it is acceptable to generalize thematic conclusions.
Additionally, the small number of directly observed mediations (seven), due to permission difficulties, may
be considered a weakness. To cure the effect of this limitation, observation data was used solely to support
or contradict interview and questionnaire data. Due to the small numbers within each actor group, the
findings must be regarded as tentative. However, in view of the consistency of responses and in support
of what the findings reveal regarding general legal practices, although I looked at particular respondents
involved in one case-type in particular institutions, there is no a priori reason to think that these findings
would be different from other cases in other culturally similar jurisdictions.
20. There has been substantial growth in the frequency and severity of medical litigation in Canada.
See Donald N. Dewees et al., The Medical Malpractice Crisis: A Comparative Empirical Perspective,
LAW & CONTEMP. PROBS., Winter 1991, at 220-21, 237, 250; Joan M. Gilmour, Overview of Medical
Malpractice Law in Canada, 3 ANNALS HEALTH L. 179, 183 (1994).
324
their favor.21 This is a result of various factors. First, there are difficulties in
successfully proving the elements of negligence by the appropriate evidentiary
standard, irrespective of whether claims are in fact legally meritorious.22
Plaintiffs must prove on the balance of probabilities that a duty of care
existed, breach of the standard of care by the medical professional, causation,
and consequent injury or loss.23 However, only a proportion of medical errors
21. See Gerald B. Robertson, The Efficacy of the Medical Malpractice System: A Canadian
Perspective, 3 ANNALS HEALTH L. 157, 170 (1994). In 1991, only 23% of plaintiffs in Canada succeeded
at trials, by 1992 that statistic had increased to a mere 34% success rate. Id. at 173. This is similar to
success rates in the United States and Britain. See P. Fenn & C. Whelan, Medical Litigation: Trends,
Causes, Consequences, in SOC IO -LEGAL ASPECTS OF MED . PRAC ., 8 (Robert Dingwall ed., 1989); Walter
Orlando Simmons, An Economic Analysis of Mandatory Mediation and the Disposition of Medical
Malpractice Claims, J. LEGAL ECON ., Fall 1996, at 42. Canadian plaintiffs who do succeed at trials
generally receive far less than successful plaintiffs in the United States, as non-pecuniary damages awards,
which include loss for pain and suffering, loss of expectation of life, and loss of amenities, were capped at
$100,000 CAD in 1978 (though adjusted for inflation) pursuant to a trilogy of decisions by the Supreme
Court of Canada. See Gilmour, supra note 20, at 185-86, 188 n.39 (citing Andrews v. Grand & Toy Alta.
Ltd., [1978] 2 S.C.R. 229, 263-65; Thornton v. Bd. of Sch. Tr. of Sch. Dist. No. 57 (Prince George), [1978]
2 S.C.R. 267, 284). Additionally, although compensatory aggravated damages may be awarded, Canadian
courts very rarely award punitive or exemplary damages. See Gilmour, supra note 20, at 186-87.
22. See Gilmour, supra note 20, at 197-98; Vincent et al., supra note 15, at 1609, 1612; Robertson,
23. See Gilmour, supra note 20, at 189-90. The legal position as to medical liability in Ontario is
virtually the same as that in most of common law Canada. See id. at 187; Robertson, supra note 21, at 168.
Apart from Quebec, which is a civil law jurisdiction, Canada’s remaining provinces and territories are
common law jurisdictions where there is a notable coherence in case law, partially because the Supreme
Court of Canada has acted to create doctrinal uniformity among all provinces as to medical liability. See
Gilmour, supra note 20, at 189-91; Robertson, supra note 21, at 173-74, 173 n.35. Medical accidents in
Ontario are compensated according to general civil liability principles and negligence tort law in particular,
which aims to compensate those injured negligently as well as to prevent substandard practice. See
Gilmour, supra note 20, at 189. This is the case notwithstanding the fact that other civil wrongs, such as
breach of contract, breach of fiduciary duty and battery, could also be relevant. All physicians are
consequently under a duty to exercise reasonable care when treating patients. See id. at 190. For medical
negligence to be substantiated, plaintiffs must, inter alia, prove that the doctor conducted herself in a
manner which was below “that degree of care and skill which could reasonably be expected of a normal,
prudent practitioner of the same experience and standing.” See id.; Robertson, supra note 21, at 173, 173
n.31 (citing Crits v. Sylvester, [1956] O.R. 132, 143 (C.A.), aff’d, [1956] S.C.R. 991). Additionally, apart
from emergency situations, patients in Canada who are able to comprehend the nature and consequences
of medical treatment must explicitly or impliedly provide informed consent to receive treatment, or
physicians may be sued for assault or battery. See Ciarlariello v. Schacter, [1993] 2 S.C.R. 119; Reibl v.
Hughes, [1980] 2 S.C.R. 880, 888-89; Gilmour, supra note 20, at 191-93; Robertson, supra note 21, at 174,
174 nn.41-42. As for liability of hospitals in medical negligence cases, they may be in breach of their direct
duty of care owed to patients. See Gilmour, supra note 20, at 190-91. This includes liability for failures
in their system to assure proper and punctual patient care. See id. at 191. However, only interns, residents
and nurses are considered hospital employees by the Canadian courts. See id. Consequently, hospitals are
not vicariously liable for the negligence of doctors who are viewed as independent contractors, and are sued
as separate entities. See id.
325
will be caused by legal negligence.24 If an incident could be classified as an

excusable error of clinical judgment, plaintiffs will fail in their claims.25
Second, plaintiffs face many practical hurdles, as even cases with real merit
demand enormous amounts of evidence to succeed. Yet, when commencing
legal proceedings, plaintiffs are normally poorly informed about the level and
quality of care they received, as compared with hospitals and physicians who
have access to more detailed information.26 Third, these cases are often
viewed as being “battles of the experts” where the defendant has far more
economic power than the plaintiff to obtain expert opinions and expert
counsel. Therefore, even in legitimate cases, plaintiffs could be shut out.
Fourth, notwithstanding the possibility of contingent fee agreements between
plaintiffs and their lawyers, only a small proportion of medical negligence
claims will be economically feasible to pursue due to high legal expenses in
conducting these cases.27 Thus, lawyers generally will not be interested in
taking cases in which plaintiffs have relatively modest claims.
Looking at these cases from a different perspective, medical malpractice
disputes often comprise issues of immeasurable, life-altering significance,
such as death or serious injury, affecting individuals’ lives as well as the lives
of their family members. These cases are multi-faceted, involving health care
professionals, whose life purpose, aspirations, training and work, is focused
upon care and healing; individuals in whom patients put their trust during
difficult life periods. The emotional difficulties of litigating for injured
24. See HARVEY TEFF, REASONABLE CARE: LEGAL PERSPECTIVES ON THE DOCTOR-PATIENT
RELA TIO NSH IP 58 (1994); Clifford Hawkins & Ian Paterson, Medicolegal Audit in the West Midlands
Region: Analysis of 100 Cases, 295 BRIT . MED . J. 1533, 1535 (1987); Sheila M. Johnson, A Medical
Malpractice Litigator Proposes Mediation, DISP. RESOL. J., Spring 1997, at 44.
25. See, e.g., Edward A. Dauer & Leonard J. Marcus, Adapting Mediation to Link Resolution of
Medical Malpractice Disputes with Health Care Quality Improvement, LAW & CONTEMP. PROBS., Winter
1997, at 199-200. See Gerald Robertson, Reform of the Law of Medical Liability: The Position in the
Common Law Jurisdictions of Canada, in CONTEM PORARY LAW : CANADIAN REPORTS TO THE 1990
INTER NATION AL CONGRESS OF COMPARATIVE LAW 175 (H. Patrick Glenn ed., 1992).
26. Henry S. Farber & Michelle J. White, A Comparison of Formal and Informal Dispute Resolution
in Medical Malpractice, 23 J. LEGAL STUD . 777, 781 (1994).
27. Even with contingency fees, plaintiffs still risk having to pay a proportion of defendants’ legal
costs if they lose at court, since Canada has implemented the British rule that costs follow the event in all
civil litigation. Thus, plaintiffs whose cases are not sufficiently strong to result in early settlements must
be ready to risk large legal costs. See Gilmour, supra note 20, at 184; Dewees et al., supra note 20, at 227;
HERBERT KRITZER, RISKS, REPUTATIONS AND REWARDS: CONTINGENCY FEE LEGAL PRACTICE IN THE
UNITED STATES 258 (2004) (discussing contingency fee arrangements in Ontario). It must also be noted
that, although technically available in Ontario, legal aid is virtually defunct. See Dewees et al., supra note
20, at 228, 249. This situation deters the commencement of actions and encourages litigants to reach
settlement. See Gilmour, supra note 20, at 184-85.
326
plaintiffs or their surviving families are patent. For defendants, the non-
economic costs of defending litigation arising from medical negligence claims
are also high, entailing much anxiety and emotional pain, as medical
professionals often view such disputes as attacks on their integrity and
competence.28 Yet, a different picture emerges through examining how
lawyers involved on all sides of these disputes understand what these cases
were about in terms of plaintiffs’ litigation motivations. Despite some
differences among the groups, attorneys representing physicians, hospitals and
plaintiffs either immediately or ultimately described the issue as one of
money—solely or predominantly.
A. Physicians’ Lawyers: It’s Only About Money
Virtually all physicians’ lawyers in the study were of the strong belief that
plaintiffs had sued for financial compensation alone. Even the two who
mentioned that non-fiscal objectives might also have been involved put much
emphasis on claimants’ primary monetary aims.
The following interview excerpts are typical of physicians’ attorneys’
responses in answering the global question: “What, in your view, were the
plaintiff’s aims in litigating?”
“My view is [that] the issue was money, to compensate for the pain associated with the
deterioration, and to compensate for lost income associated with the surgery that was
necessary. So it was money alone? I believe so.” Male attorney-50’s-prescription
alleged to have destroyed plaintiff’s bone tissue, resulting in the 40 year old plaintiff
undergoing hip replacement surgery-litigating several months.
“To settle it. Their assumption was that this would never go to trial; that they would get
money out of this beforehand. So you feel it is solely an issue of obtaining financial
compensation? Yes, but I also think that they are of the view that if they obtain financial
compensation it will make . . . them feel better. I think they’re misguided on that.”
Female attorney-30’s-plaintiff’s abdomen was not left intact after surgery-litigating
several months.
“I think in virtually all cases, it’s directly driven by financial compensation, by their
desire for compensation . . . . The sole aim, you know, in most of the cases it is to be
financially compensated for the wrong. And I would say that’s in 99% of the cases I do,
that’s what plaintiffs want.” Male attorney-30’s-child fatality case-litigating 4 years.
28. See TEFF, supra note 24, at 13, 57; Chris Currie, Mediation and Medical Practice Disputes, 15
MEDIATION Q. 215, 217 (1998); Dauer & Marcus, supra note 25, at 190-91, 191 n.29; Johnson, supra note
24, at 45; Janice E. Linn, Alternative Dispute Resolution for Health Care Providers: An Introduction, 22
J. HEALTH & HOSP . L. 336, 336 (1989); Scott Forehand, Note, Helping the Medicine Go Down: How a
Spoonful of Mediation Can Alleviate the Problems of Medical Malpractice Litigation, 14 OHIO ST . J. ON
DISP. RESOL. 907, 910 n.14 (1999); Thomas Hobbs & Gregory Gable, Coping With Litigation Stress,
PHYSICIAN’S NEWS DIGEST , Jan. 1998, http://physiciansnews.com/law/198.html.
327
The following chart summarizes the responses of physicians’ lawyers.
Figure 1—Physicians’ lawyers’ views on plaintiffs’ litigation aims

The vast majority of defense physicians’ attorneys felt that litigating plaintiffs
sued for monetary compensation alone.
To understand the stance of the bulk of physicians’ attorneys, it may be
relevant to bear in mind that the individuals studied were part of a group of
defense firms working on generally high paying cases that generate large legal
fees. Like most attorneys, they were judged by their peers and superiors, at
least to a material degree, by the amount of money they brought to their firms
in terms of billable hours, and particularly the amount of money they were
able to save their paying client, the Canadian Medical Protective Association
(“CMPA”), which underwrites the vast majority of trial judgments and
settlements in medical negligence cases in Canada.29 Defense lawyers for
29. The CMPA, which employs virtually all medical defense lawyers in Canada, is a non-profit
medical mutual defense organization incorporated by an Act of Parliament and owned and run by its
physician members. See Dewees et al., supra note 20, at 221-22; Gilmour, supra note 20, at 182. Though
it is within the organization’s discretion to refuse claims altogether and it does not sell insurance per se, the
CMPA acts partially as an insurer. Gilmour, supra note 20, at 181-82. It provides its members with legal
advice and counsel in legal actions based on allegations of malpractice, payment of legal expenses, and
provides no upper limit on the court awards, costs and settlements that it will pay out, with no deductibles.
See id.; Canadian Medical Protective Association (CMPA), About Us, http://www.cmpa-
328
doctors and hospitals in these cases each represented two clients: the
individual physician or the hospital (and/or its employees), and the insurer or
insurer-type body, which is the attorney’s paying client. Thus, the attorneys’
own personal achievement criteria and career trajectories most likely had
much to do with finances, as well as getting rid of cases and settling them.
Yet, as one prominent female physicians’ lawyer put it,
What I can tell you is what I see and what I’ve heard. I’ve heard many people say, “I
never wanted to go to a lawyer. I only went because the hospital or the doctor wouldn’t
answer my questions.” They say, “I only commenced an action because I wanted to be
very sure this never happens to anybody else,” or they say, “I wanted to hear the doctor
say ‘I messed up’ or ‘I’m sorry.’ ” And they always say “It isn’t about the money.” I’m
not so sure I always buy that. But I guess it’s possible that it’s also about something else.
This “something else” was somewhat more prevalent within the views of
hospital lawyers, who very often represented co-defendants in the same
actions.
B. Hospital Lawyers: It’s Mostly About Money
Similarly to physicians’ lawyers, the widespread view of hospital lawyers

in the cases studied was that plaintiffs sued primarily, though not solely, to
obtain financial compensation. However, contrasting somewhat with
physicians’ attorneys, only two individuals, representing 17% of those
interviewed, felt that the issue was money alone. Thus, hospital attorneys’
discourse about claimants’ primary fiscal aims was regularly coupled with talk
of other non-pecuniary or non-legal motivations, albeit on a secondary level.
The disparity between the views of hospital attorneys and physicians’
attorneys may have been due to the identity of their clients. For physicians’
attorneys, claims may have been viewed more as personal attacks against their
individual clients. Admissions of error can be personally devastating for
doctors. Hospital attorneys represented large entities where claims may have
been perceived as less personal, given that the client’s responsibility is
collective. This, in turn, may have created a less defensive environment for
acpm.ca/cmpapd02/pub_index.cfm?FILE=HOME_MAIN&LANG=E (follow “About the CMPA”

hyperlink; then follow “About Us” hyperlink). Hospitals are normally self-insured by various companies,
and nurses have a protection organization not unlike the CMPA. See Gilmour, supra note 20, at 182; J.
ROBERT S. PRICHARD , LIABILITY AND COMPENSATION IN HEALTH CARE: A REPORT TO THE CONFERENCE
OF DEPUTY MINISTERS OF HEALTH OF THE FEDERAL /PROVINCIAL /TERRITO RIAL REVIEW ON LIABILITY AND
COMPENSATION ISSUES IN HEALTH CARE, App. A at 58-59, 61, 63-64 (1990).
329
hospital lawyers, allowing for greater openness to plaintiffs’ extra-legal aims.

Moreover, if it was clear that there was going to be financial liability, there
would likely be little personal cost to admitting error.30 The chart below
highlights the views of hospital lawyers.
Figure 2—Hospital lawyers’ views on plaintiffs’ litigation aims

As with physicians’ lawyers, most hospital lawyers viewed monetary
recompense as the primary driver for plaintiffs. However, other secondary
objectives were recognized by some.
The following interview excerpts are illustrative of the responses of
hospital attorneys to the question: “What do you think the plaintiffs’ aims
were in suing?”
“To recover money. I think that was one of the big things. But, [also] to gain a sense
of control over what had happened. I think there’s just a lot of bitterness. And I think
it’s finding someone to blame for a lot of things that happened in [the plaintiff’s] life.”
Female attorney-30’s-plaintiff’s abdomen was not left intact after surgery-litigating
approximately one year.
“Oh, well [the plaintiffs are] definitely very angry. And they also want a lot of money
. . . . I think there’s [sic] a lot of other issues going on. But I also have the sense that
money was a primary driver.” Female attorney-20’s-dispute over failed vasectomy and
30. Comments on manuscript, from Richard Abel, UCLA Law School, to author (Sept. 5, 2005)
(unpublished, on file with author).
330
consequent abortion, where plaintiffs were practicing Catholics unable to support

a further child-litigating just under one year.
“Partially financial compensation. I think clearly they’re a struggling family . . . Even
though it could be said that these cases aren’t worth very much, they’re probably worth
a great deal to them. And I think also there’s the . . . element of . . . getting at the truth
and the matter of principle.” Female attorney-40’s-death of an infant subsequent to
labor and delivery-litigating five months.
A possible cause for the disparity in responses of counsel for hospitals

and those of counsel for physicians could be attributed to the gender
composition of the groups studied. The hospital lawyers’ group was
predominantly female (67%) (eight females, four males), whereas 80% of the
physicians’ lawyers group were male.31 Although these gender findings must
be regarded as tentative due to the small number of individuals studied, the
data suggests that defense lawyers’ awareness or understanding of plaintiffs’
extra-legal litigation aims was affected by gender. Only one of four male
respondents discussed plaintiffs’ extra-legal aims, whereas seven of the eight
female respondents did. Moreover, when aggregating the views of all 22
defense hospital and physician lawyers (10 female, 12 male), 80% of females
mentioned plaintiffs’ extra-legal objectives when discussing litigation aims as
compared with only 25% of males. This finding was surprising amongst the
professionals, as attorneys are trained to view disputes objectively, according
to the law. Thus, one would think that the gender of a lawyer would not affect
the way that s/he understands a case. However, this finding supports the view
that women have different interpretations of events than men, lending support
to writings in feminist legal theory claiming that even at its most neutral, there
are gender relations that continuously affect the way law works.32 Thus, it
may be that female professionals have the potential to offer greater assistance
in addressing disputants’ extra-legal needs during litigation-linked processes.33
31. Plaintiffs’ lawyers were omitted from the analysis, as there were only two females amongst that
group.
32. See CAROL GILLIGAN , IN A DIFFERENT VOICE: PSYCHO LOGICAL THEORY AND WOMEN ’S
DEVELOPMENT 25-29, 62-63, 173 (1982).
33. See Jeffrey W. Stempel, Forgetfulness, Fuzziness, Functionality, Fairness, and Freedom in
Dispute Resolution: Serving Dispute Resolution Through Adjudication, 3 NEV . L.J. 305, 311-12 (2003);
Stephen N. Subrin, A Traditionalist Looks at Mediation: It’s Here to Stay and Much Better Than I
Thought, 3 NEV . L.J. 196, 207-08 (2003).
331
C. Plaintiffs’ Lawyers: It’s Mostly About Money, But Also Other Issues
Plaintiffs’ lawyers were clearly in a different position than defense

lawyers in terms of knowing plaintiffs’ aims. Yet, in exploring some of the
questions from the lawyering theory literature—such as “how do attorneys
understand what a particular client’s ‘problem’ is?” and “what ‘facts’ do
attorneys perceive and what ‘facts’ escape them?”34 —it was interesting to note
that, as with defense lawyers, most plaintiffs’ lawyers felt that money was
plaintiffs’ primary litigation aim. Still, the bulk of plaintiffs’ lawyers viewed
non-pecuniary litigation objectives as being more important to plaintiffs than
their defense lawyer counterparts. These non-pecuniary factors, illustrated in
the chart below, included obtaining answers about what happened;
acknowledgements of harm; apologies; defendants’ acceptance of
responsibility, and retribution for insulting physician conduct.
Figure 3—Plaintiff lawyers’ views on plaintiffs’ litigation aims

The majority of plaintiffs’ lawyers viewed monetary recompense as their
clients’ primary aim. Yet, they were more aware of extra-legal litigation
objectives as compared with defense lawyers.
Nonetheless, claimants’ monetary aims were prominent in most plaintiffs’
lawyers’ parlance. As one female generalist lawyer noted in a case that had
been litigating for several months:
34. Sherwin, supra note 16, at 42-43, 48, 51.
332
I think always, when a client proceeds with a lawsuit, including a medmal lawsuit, money
is one factor. But I think that’s because that’s the only thing our courts can compensate
an individual with. But often the individual wants to understand what happened to them
and why it happened. So I think that for the individual is very important . . . as compared
with the money involved . . . So the plaintiff’s aim would be . . . information and
understanding . . . . And I guess there’s always the monetary consideration. I don’t think
people ever launch a lawsuit unless at some level there’s a monetary consideration.
Similarly, an experienced male specialist lawyer in his fifties stressed,

“My aims are to restore to my clients some kind of dignity, as their lives have
been shattered. I feel a responsibility to financially restore to them some kind
of security.”
In fact, a significant minority of both specialist and generalist plaintiffs’
lawyers viewed financial compensation as their clients’ only aim. Yet, in
comparison to most defense lawyers, plaintiffs’ lawyers generally explained
this on a less legal, more “human” level. As one male specialist lawyer
explained:
My clients are devastated psychologically, emotionally and financially . . . . They feel

they have been betrayed by their doctors . . . . My clients see the first tyranny as that of
the “cover ups” . . . i.e., the tyranny of medicine. They perceive the second tyranny as
being that of the litigation process, the length of the process, the exhaustion of the
process, the feeling that they are nothing but a number, nothing but a debt . . . . They rely
on their lawyers implicitly. More often than not, their lawyers let them down.
Plaintiffs’ lawyers would generally be most likely to acknowledge other,

non-pecuniary motives, as plaintiffs express their emotions directly to their
legal counsel. Perhaps some plaintiffs’ lawyers may also want to see
themselves as champions, not just hired guns. Perhaps they want to see their
clients as representing the interests of all potential victims. At the same time,
the only remedy plaintiffs’ lawyers can get for their clients through the formal
civil justice system is money. Therefore, plaintiffs’ attorneys also seek to
monetize other client motives.35
As with defense counsel, plaintiffs’ lawyers, other than sole practitioners,
are also judged by their peers and superiors based on fiscal considerations
such as how much money they bring into the firm in legal fees and the
amounts of settlements they are able to achieve. Judgments based on these
considerations have a significant impact on lawyers’ own career trajectories,
including election to partnership within their firms. A further practical reality
in these cases relates to contingency fee arrangements between plaintiffs’
35. See Abel, supra note 30.
333
lawyers and their clients, which existed de facto for nearly all of the cases
studied. Contingency fee lawyers only accept cases if they see the opportunity
for an economic return; if the cases look like sound investments. As such,
regardless of clients’ motives, contingency fee systems drive lawyers to
evaluate their cases initially as investments and to constantly re-evaluate them
in the same way. Thus, economic realities of legal practice may have played
a role in how the various legal actors in the study talked about and thought
about plaintiffs’ motivations and litigation aims.
Further interview excerpts from plaintiffs’ attorneys are included in the
case studies below. Notwithstanding the economic realities of legal practice,
it was startling that not a single plaintiffs’ lawyer suggested that money had
little to do with what their clients wanted from litigation. Yet, this was
something that most plaintiffs vehemently stressed.
III. PLAINTIFFS: IT ’S ABOUT PRINCIPLES, NOT MONEY !
Of all respondents, plaintiffs’ views are arguably the most important in

answering the question of what motivated them to sue and what precisely they
wanted. Although some empirical research has been done on litigants’
motivations, much of it is dated or relies upon survey evidence. Thus,
numerous lacunas still exist. Some sources argue that the vast majority of
suits are about money. 36 However, studies in tort, divorce, injury and small
claims cases have shown that plaintiffs frequently have material hidden aims
or agendas that involve more than just money, and which may, in fact, be their
cardinal goals.37 Moreover, the fact that the formal justice system does not
36. See HERBERT KRITZER, THE JUSTICE BROKER : LAWYERS AND ORDINARY LITIGATION 28-34
(1990) (identifying monetary stakes in the vast majority of cases, although Kritzer also identifies
nonfinancial stakes). Kritzer also makes the point that a distinction should be made between plaintiffs’
goals in litigation and their goals in the underlying dispute, as few litigants would choose to invest their
own money in litigation if there were not a prospect of a significant recovery. E-mail from Herbert Kritzer,
Professor of Law, William Mitchell College of Law (July 11, 2006) (on file with author). This is a valid
point; however, in the present study, some of the claims were for relatively low damages (e.g. infant death
case, child operation case, vasectomy case). Moreover, some litigants were paying their lawyers on an
hourly-rate basis (i.e. they did not have contingency fee agreements with their attorneys) with no guarantees
of financial recovery. Additionally, plaintiffs in the present study were asked specifically about their why
they instituted legal proceedings and their goals for litigation itself, not simply about their underlying
dispute objectives. Although, arguably these would be intertwined once litigation commenced. Finally,
I do not make the claim that plaintiffs’ aims for litigation never include monetary objectives, but simply
that plaintiffs at the same time have a whole host of non-fiscal objectives for litigation that are not
recognized by many legal actors.
37. See Conley & O’Barr, Hearing the Hidden Agenda, supra note 15, at 181; Marc Galanter,
Adjudication, Litigation, and Related Phenomena, in LAW AND THE SOCIAL SCIENCES 151, 191 (Leon
334
deal with litigants’ perceived requirements and real objectives, including their
non-monetary agendas, has been found to be a major reason for litigants’
dissatisfaction in small claims and injury cases.38 Thus, in various contexts
scholars have posited that plaintiffs may sue for psychological reasons, to
ascertain the truth, to express anger, or to achieve moral vindication or
“justice” by finding opponents in the wrong.39
A. Plaintiffs’ Articulations of Their Motives and Litigation Aims
Plaintiffs’ articulations of their litigation objectives rarely correlated with

what legal actors perceived as their prime litigation aims. In fact, a regular
and conspicuous occurrence was the failure of plaintiffs to mention financial
compensation as an objective at all (occurring in 65% of interviews), unless
probed. Instead, what plaintiffs recurrently repeated was a lexicon of non-
fiscal, extra-legal objectives for their litigation. “It’s not about the money!”
was a common theme throughout virtually all plaintiffs’ discourse. The issue
of “principle” was prominent for plaintiffs as revealed in the various
objectives they passionately spoke about. Many of their comments concerned
dignity and respect after the injury, inability to be heard, refusal to listen,
dismissal and victim blaming. All of this added insult to injury.40
Overall, the plaintiffs’ group was uniform in terms of the litigation aims
they discussed. There were no material differences in plaintiffs’ articulated
Lipson & Stanton Wheeler eds., 1986) [hereinafter Galanter, Adjudication, Litigation, and Related
Phenomena] (illustrating that litigants desire moral vindication albeit to varying degrees); Marc Galanter,
Reading The Landscape of Disputes: What We Know and Don’t Know (and Think We Know) About Our
Allegedly Contentious and Litigious Society, 31 UCLA L. REV . 4, 30-31 (1983) [hereinafter Galanter,
Landscape of Disputes]; Sally Engle Merry & Susan S. Silbey, What Do Plaintiffs Want? Re-Examining
the Concept of Dispute, 9 JUST . SYS. J. 151, 157 (1984); SARAT & FELSTINER , supra note 4, at 93 (finding
that litigants often seek acknowledgement that they have been treated unjustly as evidenced by the fact that
they commonly seek apologies; and that some litigants seek solely emotional vindication). See generally
SALLY ENGLE MERRY, GETTING JUSTICE AND GETTING EVEN : LEGAL CONSCIOU SNESS AMONG WORKING-
CLASS AMERICANS (1990) (illustrating that litigants frequently have important hidden agendas and
unrevealed aims which may be their cardinal goals and providing an example of a professional female
making a $30 claim against a repairman for allegedly causing damage in her house); Engel, supra note 1
(litigants want what they perceive as “justice” by the system ascertaining the truth and finding opponents
in the wrong).
38. See CONLEY & O’BARR, supra note 3, at 126-27, 141-49; Conley & O’Barr, Hearing the Hidden
Agenda, supra note 15, at 182-84, 196-97; MERRY, supra note 37, at 134-37, 142-46, 170-71; O’Barr &
Conley, supra note 7, at 159-60. See generally T. MATR UG LIO , PLAINTIFFS AND THE PROCESS OF
LITIGATION (1994).
39. See Engel, supra note 1, at 558-62; Merry & Silbey, supra note 37, at 153; Sarat, supra note
9, at 346.
335
objectives between claimants of different genders or education levels.41

Moreover, plaintiffs’ extra-legal objectives did not appear to be affected by
the passage of time, as plaintiffs’ views derived from cases that had been
litigating between three months and five years. Nor did opinions differ based
on whether claimants had complained to the College of Physicians and
Surgeons as well as having sued.42 The following chart illustrates the
spectrum of objectives iterated by claimants.
41. Although no substantive differences were found within the plaintiffs’ group in terms of gender,
a further tentative finding relates to evidence of a degree of unease amongst female plaintiffs in discussing
the financial objectives inherent in their claims. This was noted in 73% of female plaintiff interviews (11
individuals), yet in 0% of male plaintiff interviews (6 individuals). Even in cases where a mistake was
admitted and/or an apology given and a settlement agreed upon, female plaintiffs often appeared less
comfortable in speaking about pecuniary issues. This was evident even in cases comprising the most
serious harm. Seeming more concerned than males about how others perceived their litigation aims, only
female plaintiffs stressed that they were not the sorts of people who sued for money. The following are
excerpts from interviews with female plaintiffs whose cases had been ongoing for three to five years.
“From day one I said, ‘I’m all about truth here. I have to feel good within me.’ And that’s the
reason why I said, ‘I’m not the type who would get a slight hit on the bumper and, in all this neck
brace and what have you.’ I am not that.” female-50’s
“As . . . it got progressively worse, we weren’t able to cope with it financially . . . . Also starting a
law suit, because you know, we don’t sue people. We never have. . . . My parents never sued
anybody. We don’t do that kind of thing. And that car accident . . . . We don’t do that . . . .”
female-40’s
“To me, suing, it was almost a sense of like, the right thing to do . . . . That’s your only recourse
. . . . She would have needed money for these things. So we ‘had’ no choice. But, you know, it was
that sense too, thinking ‘Well what do they think? That we’re after money?’ Do they think that
money is really going to settle this type of thing, and is going to make it okay?” female-40’s
“We’re not like that. We don’t want money” female-40’s
42. Only nine plaintiffs (53% of those interviewed) sued as well as filed College complaints,
whereas all but one plaintiff discussed extra-legal objectives in litigating.
336
Figure 4—Claimants’ descriptions of their litigation aims

Plaintiffs’ articulated litigation aims were largely composed of extra-legal
objectives of principle, with 41% not mentioning monetary compensation at
all, 35% saying it was of secondary importance, 18% describing money as
their primary objective in suing, and only one person (6%) saying it was
money alone.
One might argue that plaintiffs may have felt that money is a culturally
inappropriate goal, and thus offered other motives as rationalizations to the
interviewer. However, findings in several studies on medical dispute plaintiffs
in various jurisdictions (predominantly using survey data) lend credence to
those in the present research. In these studies, claimants’ desires for monetary
compensation were never found to be a predominant litigation motivation or
of primary importance to most plaintiffs. Instead, they found extra-legal
litigation aims similar to those noted by plaintiffs in the present research.43
43. Survey findings on medical injury plaintiffs’ motivations and litigation aims, coupled with the
uniformity of responses proffered by plaintiffs within the present research on their aims of principle leave
little doubt as to their veracity. This is also supported by procedural justice findings that litigants can view
negotiations, which concentrate on money alone, as trivializing issues they view as important. See Lind
et al., In the Eye of the Beholder: Tort Litigants’ Evaluations of their Experiences in the Civil Justice
System, 24 LAW & SOC ’Y REV . 953, 965-67 (1990).
337
For instance, a Florida survey study of 127 perinatal injury plaintiffs

found that litigation motivations included not only requiring money (24% of
cases studied), but also wanting to reveal cover-ups and to protect others in
future.44 In Britain, questionnaire data indicated that less than half of medical
malpractice plaintiffs felt that obtaining financial compensation was a “very
important” objective. Prevention of similar occurrences, physician
acknowledgement, admissions that something had gone wrong, and answers
were objectives most often rated as “very important.”45 Likewise, other
British findings deriving from 227 medical negligence plaintiffs’ self-
administered questionnaires (via 5 attorneys’ firms) suggest that legal action
was taken to prevent re-occurrences, for explanations and apologies,
accountability, and admissions of negligence. Only 17% described
compensation as a reason for suing. 46 Other research in the U.S. and Australia
notes similar findings.47 However, this knowledge appears not to have filtered
through to legal practitioners in many cases.
It might also be argued that plaintiffs in medical malpractice cases may
be more invested in acceptance of responsibility than other tort plaintiffs
because doctors are so invested in seeing themselves as motivated only by
patient well-being. In comparison, the stranger defendant in automobile
accidents or products liability cases rarely pretends to be interested in victims’
well-being.48 Yet, non-monetary litigation objectives have also been found for
other case-types. For instance, research into community disputes, entailing
court and mediation observations and disputant interviews, found that
plaintiffs pursue cases not only for material interests, but also for integrity and
44. G. Hickson et al., Factors that Prompted Families to File Medical Malpractice Claims
Following Perinatal Injuries, 267 JAMA 1359, 1359, 1361 (1992).
45. Genn, Access to Just Settlements, supra note 5, at 393-412. In fact, the British National
Consumer Council survey, discussing diverse dispute types, stated that only one third of respondents said
monetary compensation was the most important thing they wanted to achieve, with apologies and
prevention of similar occurrences also being frequently noted primary goals. See NAT ’L CONSUMER
COU NC IL, SEEKING CIVIL JUSTICE : A SURVEY OF PEOPLE ’S NEEDS AND EXPERIENCES (1995).
46. Vincent et al., supra note 15, at 1609-13.
47. A.E. Daniel et al., Patients’ Complaints about Medical Practice, 170 MED . J. AUST . 598,
598-601 (1999); Dauer & Marcus, supra note 25, at 185-86, 218; Catherine Meschievitz, Mediation and
Medical Malpractice: Problems with Definition and Implementation, 54 LAW & CONTEMP. PROBS. 195,
200-01 (1991); Paul Nisselle, Editorial: Angered Patients and the Medical Profession, 170 MED . J. AUST .
576, 576-77 (1999). In view of these multi-national findings, I would argue that the issue of money being
of secondary importance to many plaintiffs should not be regarded as an instance of North American
exceptionalism. However, such findings might differ for plaintiffs in non-western contexts, where money
might more often represent core needs for survival.
338
self-respect.49 More recently, survey data deriving from 30 claimants in

sexual assault and violence cases who sought legal compensation similarly
indicated that most plaintiffs who sued sought more than money from civil
litigation.50 In fact, most claimants discussed therapeutic as opposed to
monetary goals. They frequently wanted validations of their experiences,
public declarations of the wrong, closure, retribution, and apology, as well as
to be heard and to deter similar events.51
In the present research, in comparing legal actors’ understandings with
those of plaintiffs in the same or similar cases, it was particularly remarkable
that very few lawyers noted either of plaintiffs’ most recurrently stated
objectives. Out of 40 lawyers interviewed, not a single physician lawyer or
hospital lawyer, and only three plaintiff lawyers (comprising 17% of those
interviewed) mentioned that plaintiffs wanted defendants to admit fault or to
accept responsibility. Likewise, not a single physician lawyer, only one
hospital lawyer (8% of that group) and one plaintiff lawyer (6% of that group)
noted that plaintiffs wanted to “ensure it never happened again.” Similarly,
“obtaining answers or explanations” about what had happened was the second
most repeated litigation objective of plaintiffs. All plaintiffs had stated that
they either had not received any explanation or had received an unsatisfactory
one after the incident, regardless of the severity of their injuries. Though
discussed by 50% of plaintiff lawyers, a mere 25% of hospital lawyers and
only a single physician lawyer mentioned that plaintiffs wanted answers or
explanations. Likewise, the third-most repeated plaintiff objective was to
obtain apologies and retribution for insulting physician conduct. Not a single
physician or hospital lawyer ever mentioned the issue of apology, and only
three plaintiff lawyers (17% of that group) did. Neither did any physician or
hospital lawyer bring up issues of physician conduct. Thus, it was surprising
that out of the numerous recurrent litigation aims vehemently stressed by
plaintiffs, so few were mentioned by lawyers who generally worked on these
same cases on a regular basis for several years.
This discontinuity of understandings may have related to the fact that
these aims were not within the norms of traditional legal practice. Still, apart
from suggesting that academic knowledge of plaintiffs’ litigation objectives
is not filtering through into the field of legal practice, it appeared that critical
needs of claimants who had generally suffered traumatic injuries or family
49. Merry & Silbey, supra note 37, at 153-54, 160.

50. Nathalie Des Rosiers et al., Legal Compensation for Sexual Violence: Therapeutic
Consequences and Consequences for the Judicial System, 4 PSYCHOL. PUB . POL’Y & L. 433 (1998).
51. Id. at 422.
339
fatalities were not being recognized. Although outside the scope of the
present article, it became clear within the larger research project that lawyers’
misconceptions or incomplete understandings of plaintiffs’ aims had a marked
affect upon how they approached these cases and how they behaved towards
plaintiffs during litigation-track processes, such as mediation, attempting to
resolve these disputes. For instance, a recurrent theme for plaintiffs during
litigation-track mediations was their astonishment, anger and incomprehension
of why defendants or their lawyers generally did not provide any explanations,
acknowledgements of their harm or apologies, even in a qualified manner (so
as not to increase defendants’ vulnerability during lawsuits).52
A case involving a bladder operation, which had been in litigation for
several months, provides one example of this discontinuity between legal and
lay understandings of plaintiffs’ aims. The case involved a man who had a
severe bladder problem, resulting in urination 8-10 times daily, headaches,
vomiting and strong stomach pain. The defendant-urologist performed an
operation to attempt to alleviate the problem. Subsequent to the operation, the
plaintiff felt that his problem had materially worsened, having to urinate
approximately 20-25 times a day or more and still suffering persistent
headaches, stomach pain and backaches. As a result, the plaintiff could not
keep his job as a carpenter. The plaintiff’s lawyer suggested that the initial
diagnosis and operation may not have been correct, and may have actually
worsened the problem. However, other issues were involved as well. The
following excerpts from separate discussions with both the plaintiff and his
attorney illustrate the discontinuity in their respective understandings of the
case.
“Why did he sue? Money. Solely? Yeah. I think the guy is simply looking for
compensation for what he feels has been done to him and worsened his quality of life.
There isn’t a lot of anger, there isn’t a desire to make the doctor accountable, as far as I
can tell . . . . He did start a complaints process . . . . So maybe he’s got a desire to see the
doctor brought to account.” Male general practice plaintiffs’ attorney-40’s
“Why did you sue? After the surgery, I said to [the] doctor . . . ‘Everything’s getting
worse,’ and he replied, ‘Oh, you complain too much. I have no idea what the problem
is,’ . . . took some other guy’s file . . . . He ignored me . . . . He didn’t have time for me.
He just ignored me like that . . . . He disappointed me . . . . He must see what he did; and
he must pay for what he did . . . . I believe he’s just not sorry . . . . I want the truth . . . .
52. See generally RELIS , supra note 11 (containing chapters on various aspects of case processing
for cases that ultimately underwent mediation, and examining the diverse understandings and goals of the
lay, professional and gendered actors involved in these cases; chapters five and six include examinations
and analyses of defendants’ and defense lawyers’ explanations relating to why they acted as they did during
mediations).
340
I don’t know what he did. Is that why you sued? Yeah. I want to get to the bottom of
this, what he did . . . . He didn’t care about his patients. He cares for his professional
career. Was your main aim not financial compensation? Financial compensation is not
so important . . . . Compensation is not really the problem or the solution . . . . After they
didn’t cooperate with me after complaining at the hospital and the College, then I went
to a lawyer . . . . My situation is very bad . . . . I am in a lot of pain . . . . If he tried to fix
the problem, or the hospital tried to help me, I would not go to my lawyer. But they tried
to brush me off.” Male plaintiff-40’s
In addition to the discontinuity of understandings between lawyers and

plaintiffs, the data here suggest that the extra-legal litigation motivations and
aims of plaintiffs do not change with the passage of time.53 Nor do they
appear to be dramatically affected by lawyers conditioning plaintiffs on legal
system “realities” as has been suggested in other contexts such as divorce
cases.54 Indeed, there were no marked disparities in the way plaintiffs spoke
of why they sued and what they wanted from the civil justice system as
between plaintiffs who had commenced litigation three to four months earlier
(interviewed subsequent to court-mandatory mediations) and claimants who
had been litigating for several years (interviewed after voluntary mediations
of cases already on trial lists). Moreover, in examining the discourse of all
actors on the separate issue of what they wanted to do during mediation
processes,55 plaintiffs’ articulated mediation objectives were not only
consistent with their accounts of their litigation aims (further emphasizing
their extra-legal objectives), but they were uniform for plaintiffs at all stages
of litigation. Yet, as with plaintiffs’ expressed litigation aims, there appeared
to be little awareness by defense lawyers and even some plaintiff lawyers of
what plaintiffs sought from mediations on the litigation-track. 56
53. But see Felstiner et al., supra note 2, at 638 (arguing that disputants’ feelings and objectives
change over time and complicate the dispute).
54. See Austin Sarat & William Felstiner, Legal Realism in Lawyer-Client Communication, in
LANGUAGE IN THE JUD . PROCESS 139-48 (Judith Levi & Anne Walker eds., 1990); William Felstiner &
Austin Sarat, Enactments of Power: Negotiating Reality and Responsibility in Lawyer-Client Interactions,
77 CORNE LL L. REV . 1447, 1451-59 (1992); John Griffiths, What Do Dutch Lawyers Actually Do in
Divorce Cases?, 20 LAW & SOC ’Y REV . 135, 135, 146-49 (1986); Stuart Macaulay, Lawyers and Consumer
Protection Laws, 14 LAW & SOC ’Y REV . 115, 115, 136-40 (1979); Mather & Yngvesson, supra note 3, at
775, 778, 780, 789, 796; SARAT & FELSTINER , DIVORCE LAWYERS, supra note 4, at 106-07; Austin Sarat
& William Felstiner, Law Talk in the Divorce Lawyer’s Office, 98 YALE L.J. 1663, 1664, 1671-76 (1989)
[hereinafter Sarat & Felstiner, Law Talk]. See generally WHITE, supra note 3; Felstiner et al., supra note
2.
55. Litigation objectives was the first topic of discussion during interviews, whereas mediation aims
was discussed later on during interviews.
56. See generally RELIS , supra note 11, at chapter 5 (discussing extensively defense attorneys’ lack
of awareness regarding plaintiffs’ aims during litigation-track mediation).
341
These findings serve to negate arguments that lawyers’ conditioning of

plaintiffs on legal system “realities” results in claimants materially revising
the aims of principle they seek from the justice system. Notwithstanding any
delineation laid down by lawyers to clients regarding what the litigation
system can provide or attorneys’ legal translations of cases, plaintiffs’
“original” meaning of disputes as well as their extra-legal resolution
objectives were not dislodged, even at later stages of their cases’ processing.
This “reality” is highlighted in the case studies below.
B. Case Studies Revealing Parallel Worlds of Understanding Plaintiffs’

Aims
The following three case studies examine perceptions of plaintiffs’

litigation objectives within particular disputes. The existence of parallel
worlds of understanding and perception of plaintiffs’ aims quickly becomes
apparent.
Incomplete Child Operation Case
This case, which had been litigating for less than one year, involved an
operation on a four-year-old child in which only one of the two parts
necessary was completed. The child had undergone other medical procedures
and had developed a material fear of going to doctors. As a result of the
incomplete operation he had to undergo three operations in total, having to be
sedated three times. As far as the plaintiff parents were concerned, the dispute
involved much more than simply undergoing another operation to complete
what had not been done. As the father noted during mediation, “I said
everything would be okay after the first operation and then it wasn’t. I lost
credibility with my four year old.” Yet, as the following excerpts illustrate,
understandings of the case were disparate as amongst all the actors involved.
“What do you think the plaintiffs’ aims were when they sued? Money. Do you think
that’s the sole aim? Yes. Any underlying issues? No.” Male attorney for defendant
surgeon-50’s
“Why do you think [plaintiff] decided to pursue legal proceedings? Well I, I have no
idea. But I suspect because there was a very identifiable error. They may have spoken
to a lawyer . . . . So let’s see if we can collect some money out of this. Do you see any
other reasons? I don’t . . . . I don’t think they purely were angry at me. I don’t think it
was a ‘get even’ type thing.” Male defendant surgeon-60’s
“They sued obviously to recover some money for their . . . pain and suffering . . . . I think
that in any lawsuit there’s an aim to get some money for pain and suffering. Any other
kind of underlying objectives? I don’t think so.” Female hospital attorney-20’s
“I think there was as much hurt on an emotional level, as there was, you know, a desire
for compensation . . . . First of all he didn’t apologize for what he did . . . . He offered no
words of condolence . . . no words of encouragement . . . . So I think there was an
342
element of personal scorn there . . . and I think also recognition that there was a wrong
that there was going to be compensation for. So their goals were twofold I think: to get
him to acknowledge the error of his ways, and also to ensure some compensation for their
son.” Male plaintiffs’ attorney-30’s
“Why did you sue? I wanted them to apologize for what happened to my son and to make
sure that stuff like this . . . shouldn’t end up in such a huge ordeal like it was for us . . . .
I can’t believe, like he did not want to admit fault . . . he wasn’t apologetic about it. And
the bottom line in this whole thing was he still is a doctor, . . . He told me that it was a
mistake, . . . . So that’s where I felt betrayed. Because, you know, if he would have come
out and said he was sorry and he would try to fix it right away. But he wasn’t apologetic
at the hospital. You know, ‘it was an error; it’s not my fault.’” Co-plaintiff, mother of
victim-30’s
“What was your reason . . . for starting a legal action? I think mostly because of the
lack of concern. The surgeon didn’t seem to be concerned about the mistake that he
made. If we really got a sincere apology and they said, ‘Oh, we’ll do anything we can
to make you and your family, your son, more comfortable’ . . . . But there was no
apology. It was like they tried to smooth us over with some free parking passes. It was
just the principle of it. They really affected somebody’s life. So, I wanted them to know
that . . . . Because we really didn’t want, like I don’t need the money. My son doesn’t
need the money. He’s well off. His mother’s well off. It was the fact that we wanted
somebody to notice. We didn’t want it to happen to anybody else . . . .
It’s about somebody being accountable for their actions and what they’ve done . . . . Let’s
make sure they don’t screw up and then we don’t have to go to court . . . . I mentioned
in the mediation, that’s why we get fined when we speed down the highway because it
deters us from speeding again. If you dip into somebody’s pocket . . . then obviously
somebody is held responsible for that . . . . Then maybe next time they’ll be a second
look at . . . and maybe the surgeon will double check and triple check, and make sure that
the procedure he’s doing is right.” Co-plaintiff, father of victim-30’s
Fatality Case Relating To Disputed Do Not Resuscitate (“DNR”) Order

The next case had been litigating for approximately one year at the time
of the study. It involved a dispute over a “do not resuscitate” order and the
consequent death of the plaintiffs’ husband and father. Notwithstanding his
age, the mediator noted, “it came out in mediation that this man was looking
forward to living and that this man had a lot to live for . . .”
“I think it was motivated by emotional factors, and also money.” Male physician
attorney-50’s
“I think their aim was that they just wanted to be heard. I think . . . they wanted to be
compensated financially for what they viewed as a wrong.” Female hospital attorney-
30’s
“They sought compensation for the loss of the family member who made a contribution
to a . . . business, and to the spouse who . . . depended upon him much more than might
be expected . . . and she felt that loss very acutely. So would you say then that it’s a mix
of financial compensation as well as other elements? Um, by definition, a civil action
is there for financial compensation. Whether that also represented a symbolic suing for
hurt feelings is hard to say.” Male specialist plaintiffs’ attorney-40’s
“Why did you sue? My husband was neglected in hospital the last six days until he died;
and I blame the doctor . . . You know the doctor told my son, ‘You dropped your father
343
in the hospital because you wanted to get rid of him’ . . . Those were hard words. So you
felt angry about how the doctor spoke to you? I know that it is such negligence . . . But
our main point, we think that many old people, sick people are neglected, and we say we
have to teach them. We suffered, but so many others are suffering from this. So that was
your goal when you decided to sue? Yes . . . we wanted to stop this . . . We want to open
the whole system, and what’s going on in the hospitals, particularly for old people. So
the reason that you went to a lawyer was not only because you wanted to have financial
compensation? No, no! Absolutely not. Court, you know, it’s lost of headaches, lots
of worries and lots of financial. But we wanted, you know, we feel very hurt . . . I told
the doctor, ‘why did you tell my son that we dropped my husband in the hospital because
we want to get rid of him?’ This was ‘very’ hard for me and for him. Big insult. I think
it pushed us to go to legal proceedings . . . All this situation pushed us to go ask for legal
help.” Co-plaintiff, widow of deceased-70’s
“What was your goal in suing? Primarily I wanted to . . . bring to light . . . any other
malpractices . . . until he admits he made an awful mistake . . . . I’m not fighting for my
father, but for whoever else he’s erred as well . . . The physician was totally arrogant. He
didn’t say ‘Listen, I’m really sorry about the loss of your father. I made a mistake’ . . .
He could have . . . and I think I would have felt good about that. We’re all human. We
make mistakes. He could have said that . . . . We’re not seeking anything. We wanted
answers . . . . It was his arrogance that pushed us to sue him . . . When we spoke initially
. . . he made implications we’re only here to hunt him for money. So I said, ‘I don’t want
nothing. I don’t need money. Just give me your diplomas. I lost a father. I think you
should lose something too . . . What if it was your daughter or son? This is not about
money. This is about the respect and dignity of my father . . . We didn’t get an apology
. . . This could have been done so civilly, before legal proceedings. You push us into a
corner. So the reason you went to a lawyer was not only because you wanted financial
compensation? Well you can tell by where we live and everything, the money is not part
of it. And financial loss? Yeah, well, okay. They could have said . . . ‘Let’s pick up the
tab for the funeral expenses . . . .’ Or even, you know, do something. It’s not a matter
of the money now. It’s just, ‘here’s a gesture,’ you know. You made a mistake. And I
think we would have probably gone away a lot quieter . . . We even waited right to the
last month until we said we’re going to commence a claim. This is what we want to do
because, in our hearts, Dad unjustly died.” Co-plaintiff, son of deceased-40’s
In the two cases above where finances were not required in order to
maintain life, it appeared that the sums sought represented the degree to which
the plaintiffs perceived defendants to be valuing the wrong, their harm or their
suffering. Just as salary levels represent not only what individuals can buy,
but also how much people are valued (influencing feelings of self-worth), a
higher degree of acknowledgement or importance ascribed to the harm would
be represented by higher amounts of compensation, and vice versa.
Loss of Sight Case
This case, which had been litigating for three years, involved a woman
who had glaucoma. She subsequently lost all vision in her right eye and sixty
percent of her vision in her left eye, which was expected to diminish. The
dispute related to the issue of failure to diagnose the disease in its early stages,
344
which was argued could have prevented or arrested the loss of sight. She had
been going to the same optometrist (the defendant in the suit) for 16 years.
“Apart from financial compensation, any underlying reasons for why they sued? Purely
financial.” Male attorney for optometrist-50’s
“Compensation. Financial compensation? Yeah. Any underlying issues? From when
my involvement started later in the case, what I gathered was mostly a motivation for
compensation.” Male general practice plaintiffs’ attorney-40’s
The plaintiff explained that she had asked the optometrist to do a

glaucoma test because her father had been inflicted with the disease, and she
felt film on her eye. She said, “He told me he’d already done the test and I
was negative . . . I then went to another clinic and tested positive for glaucoma
and was told that the right eye was at an advanced stage and could not be
repaired—and that it should have been detected many years ago.”
“When you instituted legal proceedings, what did you want to achieve? I hate this cliché,
but there is no amount of money that can compensate for an eye. My objective was more
or less to see that this man is either punished for what he did, or he’s kicked out . . .
stopped from practicing. I don’t know how many people he has overlooked. If he
continues to practice, there’s going to be a lot of people like me who will lose their sight.
So that was your main aim? Oh, yes. I take this very seriously. This punishment is a
very serious thing . . . We also lodged a complaint at the College of Optometrists. Do
you feel you should be compensated also? Oh, definitely. Presently I am taking drops
every day . . . eighty bucks every three weeks. This is something that’s coming out of our
pockets . . . and this will be for life. Now, I have to have the money at least for the drops.
So it is a mixture of you needing financial compensation but there are a lot of other
reasons; the main reason being that you feel that it shouldn’t happen to somebody else?
That’s right. And that was the truth. That was very important to me; and my lawyer
knew that.” Female plaintiff-40’s
“Wrong diagnosis . . . no explanation . . . She lost the right eye’s vision and sixty percent
of the left eye . . . . I think one of the reasons for the dispute is to ‘cover.’ It’s not
accepting blame. It’s obvious that the doctor is wrong. So it’s like ‘let us not admit
fault, sweep it under the carpet’ that sort of attitude. The reasons you sued, was it for
financial compensation? The principle. Also the principle. I don’t want that my wife
loses her eye, and nothing is done. Somebody has to take responsibility for it. And I
believe that the doctor and the insurance company need to be responsible for it . . .
although you can never compensate . . . .” Co-plaintiff, victim’s husband-50’s
In sum, it appeared that the vast majority of plaintiffs wanted other things,
or at least more than simply financial compensation, for their harm. Yet, as
seen above, what plaintiffs often explained to be of greatest importance to
them in terms of their aims of principle was regularly not even mentioned by
the legal actors in their cases, sometimes not even by their own lawyers.
These few examples highlight the recurrent finding within the research of the
discontinuity of understandings of plaintiffs’ aims within particular cases.
345
Although understanding plaintiffs’ goals would seem to be a basic

requirement for litigating attorneys,57 some argue that lawyers frequently fail
to even attempt to comprehend plaintiffs’ desires or how they feel about their
disputes.58 Nevertheless, these findings correlate with procedural justice
studies, which argue that lay disputants want disparate things from the justice
system than what is routinely delivered to them.59 What they desire reflects
a psychological paradigm. Yet this is inherently different from the paradigm
of legal discourse and decision-making that attorneys and judges are
socialized into in law school, and which dominates discussions by legal
actors.60
In attempting to understand this phenomenon from the viewpoint of
attorneys, one may argue that disputants’ extra-legal objectives are not within
the legitimate fields of lawyers’ operations. Some posit that when we deal
with institutions, we want a certain distance, an impersonal contact with
people acting in an official, professional capacity. 61 Yet, as lawyers have
achieved over generations a near monopoly over dispute management,62 one
might argue that if lawyers do not work to assist disputants with these issues,
who else is in the position to? Who else has the power to? It has been argued
that “lawyers create a particular epistemic understanding of what they can
know and do out of the adversary and advocate’s conception of their work that
frequently shuts them off from other ways of knowing, being, and doing.”63
In this context, the present findings empirically “highlight the need for
lawyers to be reoriented to a different set of assumptions about legal
problems, and advocate a broader, social welfare approach for lawyers to
determine what underlying conflicts may be at issue in disputes.”64
57. See generally ANTHONY KRONMAN , THE LOST LAWYER : FAILING IDEALS OF THE LEGAL
PROFESSION (1993).
58. See generally Tamara Goriely, Quality of Legal Services: The Need for Consumer Research,
3 CONSUMER POL’Y REV . 112 (1993); MICHAEL ZANDER , LEGAL SERVICES FOR THE COMMUNITY (1978).
59. Tom Tyler, Citizen Discontent with Legal Procedures: A Social Science Perspective on Civil
Procedure Reform, 45 AM . J. COMP. L. 871, 894 (1997). Id. at 872-84.
60. Id. at 872-84.
61. Interview with Joseph Raz, Professor of Law, Columbia Univ., in N.Y., N.Y. (Sept. 29, 2004).
62. MICHAEL PALMER & SIMON ROBERTS, DISPUTE PROCESSES: ADR AND THE PRIMARY FORMS
OF DECISION MAKING 25 (1998).
63. Carrie Menkel-Meadow, The Lawyer As Problem Solver And Third-Party Neutral: Creativity
And Non-Partisanship In Lawyering, 72 TEMP. L. REV . 785, 787 (1999).
64. Carrie Menkel-Meadow, From Legal Disputes to Conflict Resolution and Human Problem
Solving: Legal Dispute Resolution in a Multidisciplinary Context, 54 J. LEGAL EDUC . 7, 20 (2004). The
findings also support those who have previously urged lawyers to integrate clients’ legal and non-legal
interests. See DAVID A. BINDER & SUSAN C. PRICE, LEGAL INTERVIEWING AND COUNSELING: A CLIENT-
CENTERED APPROACH 22, 185-86 (1977); ROBERT H. MNO OK IN ET AL ., BEYOND WINNING : NEGOTIATING
346
IV. “SY ST EM CONDITIONING ” BY LAWYERS
This Part examines evidence of system conditioning and its effects upon
plaintiffs’ understandings and aims for litigation—both as a phenomenon in
itself as well as a means of comprehending how claimants’ stated aims of
principle were overlooked by many attorneys in their cases.
It has been argued that extra-legal concerns are often the most important
issues for plaintiffs,65 far outweighing legal worries.66 Thus, litigants
frequently pursue social and emotional benefits from litigation in contrast to
lawyers, who are predominantly interested in clients’ financial positions.67
Indeed, research into divorce, injury, welfare, consumer and small claims
cases has highlighted sharp distinctions between litigants’ expectations, goals
and views of justice and lawyers’ far narrower objectives and explanations of
law’s capabilities. Consequently, attorneys in various case types have been
found not to respond to issues relating to litigants’ feelings or other matters
of importance to clients that would be viewed as legally irrelevant. Instead,
they focus on cases’ legalities.68 Moreover, lawyers often treat what plaintiffs
describe as their aims as something ephemeral, and regularly urge clients not
TO C REATE VALUE IN DEALS AND DISPUTES 169 (2000); LEONARD L. RISKIN & JAMES E. WESTBROOK ,
DISPUTE RESOLUTION AND LAWYERS 86-95 (2d ed. 1997); Warren Lehman, The Pursuit of a Client’s
Interests, 77 MICH. L. REV . 1078, 1079-80 (1979); Peter Margulies, “Who Are You to Tell Me That?”:
Attorney-Client Deliberation Regarding Nonlegal Issues and the Interests of Nonclients, 68 N.C. L. REV .
213, 213 (1990); Jacqueline M. Nolan-Haley, Lawyers, Clients, and Mediation, 73 NOTRE DAME L. REV .
1369, 1386 (1998); Stephen L. Pepper, Counseling at the Limits of the Law: An Exercise in the
Jurisprudence and Ethics of Lawyering, 104 YALE L.J. 1545, 1602-04 (1995).
65. See Ann Shalleck, Constructions of the Client Within Legal Education, 45 STAN . L. REV . 1731,
1743 (1993). See generally BINDER & PRICE, supra note 64.
66. See DAVID BINDER ET AL ., LAWYER S AS COUNSELORS: A CLIENT-CENTERED APPROACH 33
(1991).
67. See Marygold S. Melli et al., The Process of Negotiation: An Exploratory Investigation in the
Context of No-Fault Divorce 40 RUTGERS L. REV . 1133, 1157 (1988). See generally CONLEY & O’BARR,
supra note 3; Sarat & Felstiner, supra note 3; Sarat & Felstiner, Law Talk, supra note 54; Felstiner & Sarat,
supra note 54.
68. See Anthony V. Alfieri, Speaking Out of Turn: The Story of Josephine V., 4 GEO . J. LEGAL
ETHICS 619, 620-628 (1991); Relis, supra note 3, at 166, 168. See generally Hosticka, supra note 3
(discussing poverty lawyers); Griffiths, supra note 54 (discussing divorce); Macaulay, supra note 54
(discussing consumer cases); SARAT & FELSTINER , DIVORCE LAWYERS, supra note 4 (discussing divorce).
Other observational data on divorce lawyers and clients has also found many lawyers reluctant to deal with
“non-legal” issues or to engage these more personal concerns, which were often the most important ones
to clients. LYNN MATHER ET AL ., DIVORCE LAWYERS AT WORK : VARIETIES OF PROFESSION ALISM IN
PRACTICE 68-69, 91-92 (2001).
347
to pursue goals such as emotional or moral vindication.69 By relying upon

their knowledge, experience and views on trial results, lawyers then condition
plaintiffs on civil justice system realities and direct clients on how to view
their disputes.70 In so doing, they attempt to persuade clients to aim for what
they view as legally realistic as opposed to plaintiffs’ initial desires.71
Consequently, injurious incidents, which are not experienced as legal events,
are translated, reconstituted and coerced by lawyers to fit into legal
compartments. This ignores aspects of disputes deemed irrelevant in law and
generally excludes many matters of greatest importance to plaintiffs.72
After some negotiation between lawyers and clients, system conditioning
has been found to result in claimants generally not demanding from lawyers
emotional or moral vindication, social justice or anything different from what
69. It has been argued that most litigants’ expectations are consequently deflated or shaped, and they
are persuaded not to expect too much from the legal process and not to demand things such as emotional
or moral vindication. See, e.g., DOUG LAS E. ROSENTHAL , LAWYER AND CLIENT: WHO ’S IN CHARGE? 76-77
(1974) (discussing personal injury cases); Herbert M. Kritzer, Contingent-Fee Lawyers and Their Clients:
Settlement Expectations, Settlement Realities, and Issues of Control in the Lawyer-Client Relationship,
23 LAW & SOC . INQUIRY 795, 803-05 (1998) (discussing personal injury cases); Richard J. Maiman et al.,
Gender and Specialization in the Practice of Divorce Law, 44 ME . L. REV . 39, 50 (1992); SARAT &
FELSTINER , DIVORCE LAWYERS, supra note 4, at 85, 106-07. Although some of these studies include
litigant interviews, much research relies upon observational data and interviews with lawyers alone. The
data from litigants in the present research suggests that simply because plaintiffs may cease to demand
things overtly from their lawyers does not mean that they do not still have extra-legal aspirations for
litigation. See generally W.A. Bogart & Neil Vidmar, Problems and Experience with the Ontario Civil
Justice System: An Empirical Assessment, in ACCESS TO CIVIL JUSTICE : BRIDGES AND BARRIERS (Allan
C. Hutchinson ed., 1990) (discussing diverse case-types). But see Felstiner et al., supra note 2 (arguing that
in reaction to various subsequent experiences and to what others say, do or expect, litigants redefine their
complaints as well as the way they perceive injurious experiences).
70. See Anthony V. Alfieri, Reconstructive Poverty Law Practice: Learning Lessons of Client
Narrative, 100 YALE L.J. 2107, 2118-25 (1991); Sarat & Felstiner, supra note 3, at 116; SARAT &
FELSTINER , DIVORCE LAWYERS, supra note 4, at 85, 106-07; Sarat & Felstiner, Law Talk, supra note 54,
at 1663, 1664, 1671-76. Sarat & Felstiner’s research was based upon observations and interviews with
matrimonial clients in 40 cases in two American towns. They found that clients were given “law talk” by
their lawyers in response to their questions. Lawyers described legal processes and actors in critical, cynical
and negative ways, stressing limitations whilst at the same time attempting to be sympathetic to clients’
situations and emphasizing their insider knowledge, connections, positions and/or reputations.
71. See, e.g., ROSENTHAL , supra note 69, at 63, 65 (discussing injury cases); Felstiner & Sarat,
supra note 54, at 1672-76 (discussing matrimonial cases); Kritzer, supra note 69, at 803-05; Sarat &
Felstiner, supra note 3, at 116; SARAT & FELSTINER , DIVORCE LAWYERS, supra note 4, at 85, 106-07. See
generally CONLEY & O’BARR, supra note 3 (discussing small claims cases).
72. See C. Campbell, Lawyers and their Public, in LAWYERS IN THEIR SOCIAL SETTING 195, 208-10
(D. MacCormick ed., 1976); Clark D. Cunningham, Legal Storytelling: A Tale of Two Clients: Thinking
About Law as Language, 87 MICH. L. REV . 2459, 2474 (1989); Galanter, Adjudication, Litigation, and
Related Phenomena, supra note 37, at 3. See generally CAIN , supra note 3.
348
they are told courts would provide.73 Indeed, having experienced system
conditioning and hearing about litigation’s “monetary realities” during
contacts with their attorneys, plaintiffs in the present study often appeared to
have simply ceased to discuss their extra-legal objectives with their lawyers.
Yet, as illustrated above, this did not divert plaintiffs’ focus from their aims
of principle throughout their litigation experiences.
The way in which a number of plaintiffs’ lawyers described plaintiffs’
aims highlights the phenomenon of “system conditioning” by lawyers in
present study’s cases. This was done in order to steer clients away from extra-
legal objectives and to train them to view their cases as relating to money
alone, as “that is all the system can provide.” Of course, the ability to frame
the terms of a dispute demonstrates the significant power that lawyers have in
setting out the nature of a problem and what should be done about it, shaping
its outcome.74 The following excerpts from interviews with plaintiffs’ lawyers
provide evidence of system conditioning in medical injury cases.
“Why do you think the plaintiff sued? That’s a little hard to say. I think ultimately . . .
they do get the message that . . . the object of the exercise . . . is trying to resolve their
case based on a reasonable compromise of the prospects of the case . . . I do believe, by
and large, that they understand when it’s explained to them that this is about
compensation. So therefore it is about money. And therefore it’s about trying to achieve
an outcome where you have received, you know, a check that represents your fair
compensation.” Male specialist attorney-40’s.
“I’ve worked on about 150 medical malpractice cases. Their aims once it gets to
litigation are primarily financial, because they listen to the lawyer and lawyers can only
quantify these things in dollar terms . . . . Yes, they would have to say that it’s finances;
but the undercurrent for them is an explanation and an apology . . . which they never got
. . . the doctor never showed up to even explain anything to them . . . .” Male specialist
attorney in infant death case-40’s.
“Day in and day out people say to me ‘We’ve got to go forward, even if it’s only to make
sure nobody else has to go through what I’ve gone through.’ So there’s an altruistic
aspect to people’s motives . . . or money doesn’t matter, you know ‘I just need enough
to do this . . .’ or getting back at the doctor. Although once they begin to understand the
financial implications of getting into litigation I think they find that they can’t afford
litigation on principle. There has to be the prospect of an economic recovery to make it
worthwhile. But I see that from the start in eighty percent of cases . . . . Then the focus
later on does become money. Perhaps because I continually remind people that’s all a
lawsuit is about. It doesn’t make the world a better place. It doesn’t heal their injury.
73. See, e.g., Cunningham, supra note 3, at 1339-57, 1367-85; Felstiner et al., supra note 2, at
634-37, 645; Sarat & Felstiner, supra note 3, at 116-17, 125-28; SARAT & FELSTINER , DIVORCE LAWYERS,
74. MERRY, supra note 37, at 2-3, 4-5, 34; Sally Engle Merry, Culture, Power, and The Discourse
of Law, 37 N.Y.L. SCH . L. REV . 209, 215, 218 (1992); Susan S. Silbey & Sally E. Merry, Mediator
Settlement Strategies, 8 LAW & POL’Y 7, 25-27 (1986).
349
It provides compensation, and as such it has to be looked at in a business-like way. As

you get into litigation, suddenly money does become the focus. Much more so. And you
find people that profess to have very modest expectations at the outset, suddenly when
you tell them they have a pretty good case, start becoming somewhat more demanding
in their expectations of monetary gain, more bullish . . . . Whether it’s a change that
would happen regardless of what I say to my clients and what I try and train them to do,
or whether I have some input into that, I don’t know. But I tell clients that investing in
a malpractice case is no different from investing in the stock market. It’s all about
money, and you should look at it that way.” Male specialist attorney-50’s.
Thus, it seemed that lawyers’ efforts to condition plaintiffs on the

monetary realities of the civil justice system played a role in causing some
plaintiffs to intertwine their aims of principle with talk of financial
recompense, or to express these extra-legal aims through monetary
compensation. A good example of claimants’ perceptions of “system
conditioning” and its effects was provided by a female plaintiff in her fifties
who described the inherent financial part of her legal action that had been
ongoing for three years. However, as the following interview excerpt
illustrates, financial compensation was not driving her claim following the
death of her son.
“Was any part of the reason to sue to get compensation? I wanted the truth about this.
And they said they couldn’t promise me the truth; they could just promise me dollars.
Dollars weren’t going to bring my son back . . . . The dollars feel really bad when you
compare them to a life. No matter how big they are, they feel bad. They hurt . . . . So
then you move that you’d like X number of dollars to do something. Like I’d like to give
a million dollars to this hospital. That was my big dream . . . . So then it didn’t make me
angry that this is about dollars. And you were able to do that? Early up [sic]. I went
into the whole civil case saying ‘I want the truth. I don’t want dollars, I want the truth’
. . . . And then I was told ‘well, the truth isn’t going to be; that isn’t part of what a civil
case is about’ . . . My ultimate goal was ‘never again’ . . . to get that particular drug that
killed my son off the market. . . . All along you’re told ‘civil equals dollars’ even though
dollars mean nothing. You want the truth; you want other people to know about this; and
you want doctors not to lie and cover up . . . . But you don’t get any of that from a civil
case. It’s just dollars, I learned that early up . . . . So, I’m not going in with a price tag.
But unfortunately you come out with a price tag, even though you didn’t go in with that
mindset.”
It was interesting to note that while this legal claim was not about money,
after system conditioning this plaintiff in some way accepted that her case had
been transformed on some level into being about dollars. Indeed, system
conditioning did appear to have an effect on some plaintiffs’ understandings
of their litigation aims. The explanation of the following male plaintiff in his
forties, who sued subsequent to the death of his father, further illustrates how
350
clients’ perceptions and aims are affected by system conditioning. The case
had been litigating for nearly one year.
“Would you say your primary aim when deciding to start legal proceedings was to get
financial compensation for your loss? Unfortunately, through lawsuits it’s only financial
compensation you can achieve . . . . In the mediation . . . their lawyer made an offer
saying ‘We’ll pay $5,000 or $7,000.’ But it’s not a matter of the price . . . So in that
case, I’m going to ask you again, when you sued, what was your aim? Well, when we
went see the first lawyer, he said, ‘Look, litigation has nothing to do with what you want.
This is only about money.’ . . . That’s . . . the way we were explained it . . . . The legal
system, all it can give you is only financial . . . not answers or an apology. The legal
proceedings can’t do that for you, unfortunately . . . . The doctor’s lawyer said at the
mediation, ‘Well, if you’re struggling and hurting for money . . . .’ He doesn’t know
where we live, doesn’t know about our financial situation. Meanwhile there’s no
mortgage on our house. He’s thinking maybe we’re trying only to get money out of the
doctor for financial benefit. But no . . . this is our family, and there’s not a dollar value
to it . . . . I’m doing it for my father; it’s my passion, my heart . . . And I want everyone
to know . . . . People should have dignity in their lives . . . . I’m selling my factory, I’m
devoting my life to whatever I can . . . .”
For those plaintiffs who said that obtaining financial compensation for
their harm was one objective, a duality of principles and money was present
throughout their articulations on what they sought. For example, a dispute
over a failed vasectomy and a consequent pregnancy and abortion provided an
example of plaintiffs intertwining their aims of principle with financial
compensation, or simply expressing these objectives through monetary
recompense. The plaintiffs were practicing Catholics who were unable to
financially support an additional child. The case had been litigating for just
under one year.
“What was the reason you sued? Really, when it comes down to it, it was just the hurt
that I put my wife through due to the unsuccessful operation . . . because it wasn’t my
fault; but we had a lot of trouble at that time . . . . So I more or less wanted to sue because
of her; because of what she went through . . . . I just wanted the doctor to pay for what
he has done to me and my family. And the question isn’t merely about money. I just
wanted answers of why, you know? I just wanted him to tell me why did this happen to
me . . . . I wasn’t given no answers, nothing.” Male plaintiff-30’s.
A hospital accident case that had been litigating for five years provides
a further example of financial compensation entangled with the non-monetary
objectives plaintiffs sought. The case involved the behavior of a nurse and her
failure to assist a hospitalized patient who repeatedly asked for her help but
did not receive it. The patient subsequently fell while in the hospital. The
damage resulting from the fall forced the patient to wear large casts on her leg
continuously for three years. These casts rendered her immobile for what
351
were to be the last three years of her life. Although subsequent care for the
victim of the accident necessitated money, as with virtually all plaintiffs in the
study, the expressed motivations and objectives of her surviving children in
bringing the suit clearly showed they did not sue for money. Instead, they
intensely spoke of how it was principles that were of greatest importance to
them in terms of what they sought from litigation. As the following interview
excerpts illustrate, they sought acknowledgement, an apology, the prevention
of recurrences and acceptance of responsibility on the part of the defendant.
“That’s what this dispute has always been about . . . . They not acknowledging that this
had transpired . . . the fact that she fell, nothing . . . I’m sure that nurse is still doing it to
others . . . she was so rude to my mother . . . And then . . . trying to get them to say,
‘We’re sorry your mother fell.’ You know, nothing . . . . The whole reason for all of this
is, as we said to our lawyer was ‘My mother needed an apology.’ ‘That’ is what this is
about: the mistakes and no apology, nothing. That’s what this is about . . . apologizing
. . . What’s so difficult about acknowledging the fact that this was a mistake, and saying
we’re sorry? That’s all we wanted. So that was the reason you decided to sue? Yeah.
And my mother’s care. What was going to happen to her? And all these casts she had
to wear on her leg for the years until she passed away . . . . It wasn’t as if they had all this
money stored away for a time like this, because they didn’t have it. So, what are you
supposed to do? You know, somebody had to be responsible for this. Because if this
happened to my mother, how many other people was it going to happen to? . . . It wasn’t
a money issue. It became a money issue in the end only because of care for my mother
. . . . Something had to be done for this error.” Co-plaintiff, daughter of victim-50’s
“Why did you decide to take legal proceedings? We had become so frustrated. The
hospital wasn’t accepting accountability. They weren’t even acknowledging it to start
with. It was unbelievable . . . just to have somebody say that this happened. We’re sorry
. . . But nothing was done . . . . They just kept sweeping it under the rug, that this
accident had even happened . . . . It’s important they should admit to such things, and
take care of it properly; and promptly. Otherwise, you know what, there wouldn’t have
been any kind of lawsuit . . . Part of it was that I felt the hospital should be accountable
. . . . We were looking at my mother losing her leg. That’s what we had basically been
told . . . as a result of the fall . . . How was she going to be able to cope with all of this
because of the accident in the hospital? We told our lawyer, ‘we want an apology . . . and
some acknowledgement of this’ . . . . It wasn’t about a dollar amount to be compensated,
it was about so they could know what damage was done to the family.” Co-plaintiff,
daughter of victim-40’s
Some plaintiffs may have sought compensation as a signal of being heard

and acknowledged or as a deterrent to future conduct. This seemed to be the
case in the incomplete child operation case noted above. Indeed, “money is
often a symbol of other things, such as finding a fair and honorable result,
validation of injury, vindication of justice, ‘winning’ . . . attribution of fault,
352
and perception of (not) being ‘taken.’ ”75 One plaintiffs’ lawyer provided
insight into the phenomenon of the concretization of principles expressed as
monetary aims, as illustrated in the following interview excerpt.
“What in your view were the plaintiffs’ litigation aims? Money was not the issue. I
think there was a desire for some revenge . . . along with a desire to see the wrongdoer
punished . . . and a desire to find the truth . . . and ultimately making the defense pay an
amount . . . . But for non-financial reasons? Yeah, that’s right.” Male specialist
plaintiffs’ attorney in a fatality case-50’s
The issue was expounded from a slightly different perspective by an

experienced female non-lawyer mediator, as evinced in the following excerpt.
“Do these cases relate to monetary compensation alone, or do you see other issues
involved? No, not even usually relating to money. There are many other issues. I mean
money is the way it often gets described as a problem . . . but financial remuneration, in
my cases anyway, has not been a focus . . . I’m not saying there aren’t legitimate
monetary issues, but there are many other factors at play . . . . If it ‘is’ just money, we’re
talking about a different animal . . . . Sometimes I think lawyers in particular make it a
substantive issue when it’s really more than that . . . . In the mediations I mediated at the
College, in every single case they withdrew their complaint . . . even fatality cases . . . .
People only know that the way to get some recognition is to do legal proceedings.”
Plaintiffs’ lengthy accounts of their motives and objectives for litigation

showed that their extra-legal aims of principle remained important objects of
plaintiffs’ desires throughout the processing of their cases. This was
notwithstanding their lawyers’ delineations of what the litigation system could
provide, the legal reformulation of plaintiffs’ cases, or the passage of time.
Thus, although the intermeshing of principles and money added complexity
to the issue, it was clear that, notwithstanding any needs or desires for
pecuniary recompense, plaintiffs were intensely seeking extra-legal principles
from the civil justice system. Plaintiffs sought “truth” as they had suffered
harm and deserved answers. They wanted perceived wrongdoers to “do right”
in recognizing their wrongs, admit fault, accept responsibility and apologize.
This appeared to be the case regardless of the degree to which claimants
required funds to continue life in their newfound situations. Yet, through civil
litigation, these principles were ultimately manifested in money alone, in a
system that was repeatedly described to them as one which could only provide
financial compensation for their harm. Consequently, plaintiffs who also
75. John Lande, Toward More Sophisticated Mediation Theory, 2000 J. DISP. RESOL. 321, 328
(2000).
353
mentioned pecuniary recompense when speaking of their aims for litigation,

at the same time vehemently stressed, “It’s not about the money!”
Some plaintiffs may have initially pursued solely extra-legal aims of
principle through litigation, with lawyers subsequently persuading them to
substitute money. Others may have had multiple objectives from the outset,
seeking from litigation both extra-legal issues of principle as well as financial
compensation for their harm. In these cases, plaintiffs’ lawyers may have then
simply translated all of plaintiffs’ aims into monetary ones. Regardless of
which of these versions was most common, the fact remains that many of the
things that plaintiffs uniformly indicated they sought from the civil justice
system were not mentioned by most defense lawyers, or by many plaintiff
lawyers.
V. AN INSTITUTIONAL THEORY ON LAWYERS ’ UNDERSTANDINGS OF

PLAINTIFFS’ LITIGATION AI MS
This Article has provided a picture of discontinuity in terms of the

comprehensions of legal actors regarding what plaintiffs want from the civil
justice system versus plaintiffs’ own articulations of their motivations and
aims for litigation. This discontinuity was found to continue throughout
litigation processes until settlement or trial. I present here an institutional
explanation involving a fourfold process as to why this may have occurred.
This should arguably describe both cases where plaintiffs approach lawyers
seeking solely extra-legal aims of principle through litigation, as well as those
where plaintiffs seek principles as well as financial compensation for their
harm.
First, consistent with legal system norms, lawyers are trained to operate
according to rights and rules, applying law to facts and placing people and
occurrences into legal categories. Thus, it is more difficult for
attorneys—who tend to be rational, logical and analytical—to be sensitive to
parties’ interpersonal, emotional or extra-legal needs.76 Consequently,
76. See, e.g., Susan J. Bell & Lawrence R. Richard, Anatomy of a Lawyer: Personality and Long-
Term Career Satisfaction, in FULL DISCLOSURE : DO YOU REALLY WANT TO BE A LAWYER ? 152-53 (1992);
Susan Daicoff, Lawyer, Know Thyself: A Review of Empirical Research on Attorney Attributes Bearing
on Professionalism, 46 AM . U. L. REV . 1337, 1405 (1997); Guthrie, supra note 15, at 155-56, 158-59;
James L. Hafner & M. Ebrahim Fakouri, Early Recollections of Individuals Preparing for Careers in
Clinical Psychology, Dentistry, and Law, 24 J. VOCATION AL BEHAV . 236, 236-41 (1984); Sandra Janoff,
The Influence of Legal Education on Moral Reasoning, 76 MINN. L. REV . 193, 228-29, 234 (1991); Russell
Korobkin & Chris Guthrie, Psychology, Economics and Settlement: A New Look at the Role of the Lawyer
76 TEX . L. REV . 77, 87 (1997); Leonard L. Riskin, Mediation and Lawyers, 43 OHIO ST . L.J. 29, 43-46
354
lawyers often focus exclusively on or emphasize money issues77 in contrast to

disputants’ tendencies to be more emotional about their cases.78 Indeed,
litigants in diverse case-types have been known to largely talk past their
lawyers who interpret their descriptions without there ever being a shared
comprehension of events.79
Second, as a result of the civil justice system’s denial of certain human
needs deemed legally irrelevant, lawyers condition plaintiffs on legal system
realities, preparing them to expect only money as “that is all the system can
provide.” Lawyers then inevitably re-define plaintiffs’ perceptions and
transform the “reality” that litigants describe.80 This involves narrowing the
scope of the issues and reformulating all of plaintiffs’ stories and objectives
(both extra-legal and monetary) to fit into legally cognizable
categories—ultimately relating to monetary compensation alone. This
commonly results in litigants being effectively silenced while lawyers purport
to recount their stories.81 The effects of dispute translation should not be
underestimated, as the linguistic definitions of conflicts are critical aspects of
disputes.82 Indeed, psychological research has demonstrated how language
and words influence the perceptions and views of listeners and readers.83
Thus, I argue that system conditioning and dispute translation not only alter
the nature of disputes for lawyers and observers,84 but also contribute to some
(1982); Marjorie A. Silver, Emotional Intelligence and Legal Education, 5 PSYCHOL. PUB . POL ’Y & L.
1173, 1198-1200 (1999). Research focusing upon attorneys’ psyches also suggests that lawyers are
predominantly left-brain dominant, indicating analytical orientation. See Guthrie, supra note 15, at 157;
Graham B. Strong, The Lawyer’s Left Hand: Nonanalytical Thought in the Practice of Law, 69 U. COLO .
L. REV . 759, 761-62 (1998).
77. Jean R. Sternlight, Lawyers’ Representation Of Clients In Mediation: Using Economics and
Psychology To Structure Advocacy in a Nonadversarial Setting, 14 OHIO ST . J. ON DISP. RESOL. 269, 323,
326, 342 (1999).
78. See Felstiner & Sarat, supra note 54, at 1456; Sternlight, supra note 77, at 324.
79. See Austin Sarat & William Felstiner, Law and Social Relations: Vocabularies of Motive in
Lawyer/Client Interaction, in THE LAW & SOCIETY READER 406 (R. Abel ed., 1995) (discussing divorce
cases). See generally Alfieri, supra note 68 (discussing poverty lawyers); Hosticka, supra note 3
(discussing poverty lawyers); Macaulay, supra note 54 (discussing consumer cases).
80. See Cunningham, supra note 3, at 1298, 1339-57, 1367-85.
81. Alfieri, supra note 68, at 2119.
82. See Mather & Yngvesson, supra note 3, at 780-81. It has been argued that the manner in which
lawyers define litigants’ problems leads to the unhappiness and frustration of litigants in the long term. See
MATR UG LIO , supra note 38 (discussing injury cases); Jeffrey M. Fitzgerald, The Contract Buyers League
and the Courts: A Case Study of Poverty in Litigation, 9 LAW & SOC ’Y REV . 165, 177 (1975) (discussing
poverty litigation).
83. Paul Kay & Willet Kempton, What is the Sapir-Whorf Hypothesis?, 86 AM . ANTHR OPOLO GIST
65 (1984).
84. See, e.g., Alfieri, supra note 68, at 2111; Sara Cobb, The Domestication of Violence in
355
plaintiffs eventually articulating that very same interest in economic

recompense despite having other strong extra-legal objectives for litigation.
Third, plaintiff lawyers’ dispute translations must also serve to cement
their own convictions that financial compensation is either entirely or
primarily what their clients seek, as plaintiffs were found to cease or reduce
the extent to which they discussed their aims of principle with their counsel
over the years in which litigation ensued. This was similarly found with U.S.
divorce lawyers and clients.85
Fourth, the translated fiscal manifestations of plaintiffs’ extra-legal
objectives are then transmitted by plaintiffs’ lawyers to defense lawyers in
their subsequent, often regular, bilateral communications and monetary
negotiations that continue throughout the months or years of case processing.
The defense attorneys interviewed had little contact with plaintiffs themselves.
Yet, they regularly mentioned their communications with plaintiff lawyers.
This, coupled with the economic realities of legal practice for plaintiff
lawyers, such as contingency fee arrangements, and for defense
attorneys—who generally act for insurance-type entities that view cases
monetarily—must further reinforce lawyers’ misunderstandings or incomplete
understandings of what plaintiffs want. In fact, it may be that these
misunderstandings of what plaintiffs seek are subsequently passed down from
defense lawyers to their defendant clients. So the cycle of at least partial
misconception of claimants’ aims continues throughout litigation processes,
including litigation-track mediations.86 All of this occurs whilst plaintiffs
almost uniformly insist, “it’s not about the money!”
VI. CONCLUSION
Through an examination of the understandings of both lay and

professional actors within the civil justice system, this article has reconsidered
a core element of the legal system, legal practice, and the resolution of
disputes: why plaintiffs sue and what they seek from the civil justice system.
In comparing plaintiffs’ articulations and lawyers’ comprehensions of
plaintiffs’ aims within the structure of litigation processes, I have argued that
a picture of near pervasive misunderstanding or incomplete understanding
exists between lawyers and plaintiffs on the elemental issue of what these
Mediation, 31 LAW & SOC ’Y REV . 397, 436-37 (1997); Hosticka, supra note 3, at 600-04; Sarat &
Felstiner, supra note 3, at 116-17.
85. Felstiner & Sarat, supra note 54, at 1460, 1463-64.
86. See RELIS , supra note 11, at chapter 9.
356
cases are about. By looking beyond the perceptions captured, I have

embarked upon a theory as to why these disparities exist.
I have shown that the disparate lawyer groups, regardless of allegiances,
tended overall to hold the same view: that plaintiffs sued mostly or solely for
money. Physicians’ lawyers saw claimants as solely questing for funds,
whereas hospital and plaintiff lawyers generally viewed financial
compensation as litigants’ primary aim to compensate for harm. Yet, even for
those plaintiffs’ lawyers who mentioned clients’ extra-legal objectives, these
were swiftly translated into monetary compensation alone—as “that was all
the legal system could offer.” At the same time, virtually all plaintiffs
vehemently stressed that they sued not for money, but rather for principles.
Although financial recompense was involved at some level within some
claimants’ descriptions of their objectives,87 plaintiffs’ articulations of what
they sought from the civil justice system were thickly composed of extra-legal
aims. Yet, remarkably, the things that plaintiffs repeatedly described as being
of prime importance to them were generally omitted by virtually all
physicians’ lawyers, most hospital lawyers, and sometimes even by plaintiffs’
own lawyers.88 Thus, plaintiffs’ expressed motivations and litigation aims
rarely correlated with legal actors’ understandings of this basic premise.
I have further shown that plaintiffs’ aims of principle do not appear to
dissipate over time, as their descriptions of what they wanted were the same
in cases that had been litigating for three months or for several years.
Likewise, I have demonstrated that plaintiffs’ extra-legal objectives do not
change subsequent to lawyers’ conditioning them on civil justice system
“realities” and legal transformations of their disputes. System conditioning
resulted in plaintiffs’ increased perceptions of the litigation game being about
money. Claimants’ expressed aims of principle were also at times intertwined
with pecuniary recompense or described as their concretization. These issues
were interwoven because, as some claimants’ explicated, the system, via
lawyers, forced them to be. Plaintiffs approached the legal system wholly,
predominantly, or at least in part seeking principles, yet they were conditioned
by their lawyers to understand that litigation can only be about money—even
for cases that later underwent mediations. Nevertheless, the non-economic
principles that plaintiffs sought from litigation remained driving ambitions for
them throughout.
87. Clearly, the more critically funds were required to maintain lives subsequent to medical
accidents, the greater importance would be ascribed to them in terms of litigation objectives.
88. As previously pointed out, gender may play a role in lawyers’ sensitivity to plaintiffs’ extra-legal
needs. See supra Part II.B.
357
The findings here illustrate in the context of medical injury cases the
ideological dissonance that exists between how plaintiffs and the justice
system, through its actors, view disputes.89 The data further highlight the
indivisibility of plaintiffs’ legal and extra-legal needs in case processing. Yet,
“the operations of the legal system hinge on law’s ability to sharply
distinguish between the legal and the extra-legal.”90 The findings here,
however, show this is ultimately misleading, as human disputes have not only
legal implications, but also often a host of other concerns, such as emotional,
interpersonal and moral issues.91 Some have argued that disputants expect that
the legal system can deliver more than it is actually able to, and thus their
desires may be unrealistic.92 However, based upon the premise that the law
and its workings must exist for the benefit of laymen rather than for that of
legal or institutional actors, this article implicitly argues that conceptions of
the meaning of litigation processes within the civil justice system must be
broadened to include litigants’ extra-legal needs. If lawyers view litigation
processes solely as legal and financial mechanisms, many fundamental issues
within disputes will not be addressed or resolved.93 Thus, on a policy level,
a rethinking of civil justice assumptions and objectives is urgently necessary.94
I offer two proposals to address the problems discussed in this article.
First, as these findings derive in part from issues of failed communication,
direct dialogue early on in litigation between defense lawyers and plaintiffs
on plaintiffs’ litigation aims is necessary. This could occur during early
litigation-track mediations, which present opportunities for litigants to
articulate their extra-legal objectives in a confidential environment. Assuming
that plaintiffs partake in a meaningful way, this would serve to sensitize
89. This has similarly been found for other case-types. See generally BALD WIN , supra note 5; JOHN
BALD WIN , MONITORING THE RISE OF THE SMALL CLAIMS LIMIT : LITIGANTS’ EXPERIENCES OF DIFFERENT
FORMS OF ADJUDICATION (1997) (discussing small claims); Felstiner & Sarat, supra note 54 (discussing
divorce); MERRY, supra note 37 (discussing family and neighbourhood disputes); ROGER BRYANT
HUNTING & GLORIA S. NEUWIR TH, WHO SUES IN NEW YORK CITY? (1962) (discussing automobile accident
cases).
90. REZA BANAKAR , MERGING LAW AND SOCIOLOGY 7-8 (2003).
91. See Carrie Menkel-Meadow, The Trouble with the Adversary System in a Postmodern,
Multicultural World, 38 WM . & MARY L. REV . 5, 6-7, 14-15, 19, 26-27 (1996).
92. See, e.g., Felstiner & Sarat, supra note 54, at 1459-60; MERRY, supra note 37, at 179
(commenting on expectations of the justice system).
93. Vincent et al., supra note 15, at 1612; RELIS , supra note 11 (providing evidence of various
adverse consequences resulting from lawyers and litigants’ disparate understandings and aims for litigation
and mediation processes).
94. See Des Rosiers et al., supra note 50, at 433, 442. It has been similarly argued that the attributes
of the adversary system as the “ideal type” of a legal system must be re-examined. See Menkel-Meadow,
supra note 91, at 5, 7, 42.
358
defense attorneys and defendants to case realities from plaintiffs’ perspectives

and may affect the way lawyers subsequently approach these cases. Second,
core law school curriculums for law students and continuing legal education
and ethical rules for attorneys must provide greater emphasis upon litigants’
extra-legal dispute realities, needs, and objectives during case processing.95
This should result in a broadening of the meaning of litigated disputes from
those dictated by narrow, adversarial approaches to conflict resolution. This,
in turn, should assist in transforming attorneys’ orientations to focus upon
problem solving, as opposed to money alone, so that money need not be
“regarded as a proxy for everything.”96 Various changes have occurred in
American law schools to include new approaches to lawyering, such as
therapeutic jurisprudence,97 holistic lawyering and collaborative lawyering.
Yet, although these approaches have been taught for a number of years,98 they
still remain on the periphery of legal education and practice.99 Thus, much
remains to be done to bring these perspectives into the mainstream.
It is my hope that this article was successful in its primary objective of
generating further recognition of the need to look beyond the monetary
elements of claims to understanding plaintiffs’ motivations and their aims
95. See, e.g., Kimberlee K. Kovach, Lawyer Ethics Must Keep Pace with Practice: Plurality in
Lawyering Roles Demands Diverse and Innovative Ethical Standards, 39 IDAHO L. REV . 399, 429 (2003);
Kimberlee K. Kovach, Good Faith in Mediation: Requested, Recommended, or Required? A New Ethic,
38 S. TEX . L. REV . 575, 619 (1997); Menkel-Meadow, supra note 63, at 9; Carrie Menkel-Meadow, Ethics
in Alternative Dispute Resolution: New Issues, No Answers from the Adversary Conception of Lawyers’
Responsibilities, 38 S. TEX . L. REV . 407, 409-10, 428 (1997); Nolan-Haley, supra note 64, at 1372-73.
Others too have discussed behavioral transformation of attorneys, focusing upon relationship and
communication skills as well as listening, empathizing and problem-solving skills. See Guthrie, supra note
15, at 180-82; Joshua D. Rosenberg, Interpersonal Dynamics: Helping Lawyers Learn the Skills, and the
Importance, of Human Relationships in the Practice of Law, 58 U. MIAMI L. REV . 1225, 1228, 1234
(2004).
96. Carrie Menkel-Meadow, Mothers and Fathers of Invention: The Intellectual Founders of ADR,
16 OHIO ST . J. ON DISP. RESOL. 1, 36 (2000).
97. Therapeutic jurisprudence deals with emotional and psychological implications of the legal
system and attempts to apply law in a more therapeutic way. See Loraine E. Ferris, Using Therapeutic
Jurisprudence and Preventive Law to Examine Disputants’ Best Interests in Mediating Cases About
Physicians’ Practices: A Guide for Medical Regulators, 23 MED . & LAW 183, 185-86 (2004).
98. Kimberlee Kovach, New Wine Requires New Wineskins: Transforming Lawyer Ethics for
Effective Representation in a Non-Adversarial Approach to Problem Solving: Mediation, 28 FORDHAM
URB . L.J. 935, 972-76 (2001); Menkel-Meadow, supra note 96, at 427.
99. See, e.g., Paul Brest, Skeptical Thoughts: Integrating Problem Solving into Legal Curriculum
Faces Uphill Climb, A.B.A. DISP. RESOL. MAG . Summer 2000, at 20; Leonard L. Riskin, The
Contemplative Lawyer: On the Potential Contributions of Mindfulness Meditation to Law Students,
Lawyers, and their Clients, 7 HARV. NEGOT. L. REV . 1, 8, 10, 14, 16-21, 23 (2002); Frank E.A. Sander &
Robert H. Mnookin, A Worthy Challenge: The Teaching of Problem Solving in Law Schools, A.B.A. DISP.
RESOL. MAG . Summer 2000, at 21.
359
during litigation processes. More attention needs to be paid by attorneys to

the possibility that plaintiffs want things other than money from litigation.
For this to happen, the law and legal actors must first recognize a broader
definition of legal disputes. It is my hope that this article will assist in
bringing about such recognition.
360
APOLOGIES AND LEGAL SETTLEMENT:
AN EMPIRICAL EXAMINATION
Jennifer K. Robbennolt*
TABLE OF CONTENTS
INTRODUCTION ........................................................................................ 461

I. LEGAL DEBATE OVER APOLOGIES ........................................... 463
A. Apologies and Admissibility ................................................ 465
B. Expressions of Sympathy...................................................... 468
C. Statutorily Protected Apologies ............................................ 470
D. Debate over Perceptions of Safe Apologies ........................ 473
II. PRIOR PSYCHOLOGICAL RESEARCH ON APOLOGIES ............. 475
A. Psychological Responses to Apologies ................................ 475
B. The Nature of the Apology ................................................... 479
C. Apologies and Legal Settlement ........................................... 480
III. THE PRESENT STUDIES ................................................................ 482
A. Effects of Apologies on Settlement Decisionmaking ......... 484
l. Effects of Apology on Settlement................................... 485
2. Effects of Evidentiary Rules on Settlement ................... 490
3. Summary .......................................................................... 491
IV. IMPLICATIONS OF RESULTS......................................................... 501

A. Evidentiary Protection for Apologies .................................. 502
B. Defendants ............................................................................. 505
C. Plaintiffs ................................................................................. 509
* Associate Professor of Law & Senior Fellow at the Center for the Study of Dispute
Resolution, University of Missouri-Columbia School of Law. B.S. 1991, Willamette
University; J.D. 1996, Ph.D. 1998 (Psychology), University of Nebraska-Lincoln. I am
indebted to Chris Guthrie, Monica Johnson, Phil Peters, Len Riskin, Grant Robbennolt,
Jean Sternlight, and Christina Studebaker for their very helpful advice on earlier versions of
this paper. Portions of these data were presented at the biannual meeting of the American
Psychology-Law Society in March 2002 and at the annual meeting of the Law and Society
Association in May 2002; I thank the participants at those conferences for their comments.
Valuable research assistance was provided by Patrick Pinkston, Mitch Berry, Jay Hastings,
Cecily Helms, Elizabeth McNichols, Jason Moore, Josie Pottebaum, and Tammy Stienle.
Finally, my thanks go to the University of Missouri Law School Foundation and the Hewlett
Foundation for financial support.
460

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
361
December 2003] Apologies and Legal Settlement 461
D. The Role of the Lawyer ........................................................ 511

CONCLUSION ............................................................................................ 515
INTRODUCTION
It is often said that U.S. legal culture discourages apologies. 1

Defendants, defense counsel, and insurers worry that statements of
apology will be admissible at trial and will be interpreted by jurors and
judges as admissions of responsibility.2 In recent years, however,
several legal commentators have suggested that disputants in civil
lawsuits should be encouraged to apologize to opposing parties.3 They
claim that apologies will avert lawsuits and promote settlement.4
1. See, e.g., Hiroshi Wagatsuma & Arthur Rosett, The Implications of Apology: Law and
Culture in Japan and the United States, 20 LAW & SOC'Y REV. 461 (1986).
2. See infra Part I.A.
3. See, e.g., Jonathan R. Cohen, Advising Clients to Apologize, 72 S. CAL. L. REV. 1009
(1999) (hereinafter Cohen, Advising Clients to Apologize]; Steven Keeva, Does Law Mean
Never Having to Say You're Sorry? AB.A. J., Dec. 1999, at 64; Aviva Orenstein, Apology
Excepted: Incorporating a Feminist Analysis into Evidence Policy Where You Would Least
Expect It, 28 Sw. U. L. REV. 221 (1999); Daniel Shuman, The Role of Apology in Tort Law,
83 JUDICATURE 180 (2000) (hereinafter Shuman, The Role of Apology in Tort Law];
Deborah L. Levi, Note, The Role of Apology in Mediation, 72 N.Y.U. L. REV. 1165 (1997).
4. Approximately fifteen million civil cases and hundreds of thousands of tort claims are
filed in state courts each year. See COURT STATISTICS PROJECT STAFF, NAT'L CTR. FOR
STATE COURTS, EXAMINING THE WORK OF STATE COURTS, 2001 (2001), available at
http://www.ncsconline.org/D_Research/esp/CSP_Main_Page.html; COURT STATISTICS
PROJECT STAFF, NAT'L CTR. FOR STATE COURTS, STATE COURT CASELOAD STATISTICS,
2001 194 tbl.16 (2001), available at http://www.ncsconline.org/D_Research/csp/2001_Files/
200l_Tables_l0-16.pdf. An untold number of additional disputes occur, but are resolved in
some way prior to a formal legal filing. See William L.F. Felstiner et al., The Emergence and
Transformation of Disputes: Naming, Blaming, and Claiming, 15 LAW & Soc'Y REV. 631
(1980-81). The majority of these cases are ultimately resolved through settlement in some
way. See id.; Marc Galanter & Mia Cahill, "Most Cases Settle": Judicial Promotion and
Regulation of Settlements, 46 STAN. L. REV. 1339 (1994); Chris Guthrie, Procedural Justice
Research and the Paucity of Trials, 2002 J. DISP. RESOL. 127; Michael J. Saks, Do We Really
Know Anything About the Behavior of the Tort Litigation System - and Why Not? 140 U.
PA. L. REV. 1147 (1992). In fact, the percentage of cases that are resolved by trial is
declining. See Hope Viner Samborn, The Vanishing Trial, A.B.A. J., Oct. 2002, at 24. These
settlements, however, are often not reached before the parties have been subjected to
considerable expense - in terms of time, finances, and emotion. As Judge Learned Hand
has said, "I must say that as a litigant I should dread a lawsuit beyond almost anything else
short of sickness and death." Learned Hand, The Deficiencies of Trials to Reach the Heart of
the Matter, 3 LECTURES ON LEGAL TOPICS 89, 105 (1921); see also Thomas B. Metzloff,
Resolving Malpractice Disputes: Imaging the Jury's Shadow, 54 LA w & CONTEMP. PROBS.
43, 59 & n.54 (1991) (finding that many cases settle right before trial). At the same time
there is continuing debate about the merits of settling civil cases. See, e.g., Owen Fiss,
Against Settlement, 93 YALE L.J. 1073 (1984); Galanter & Cahill, supra; Andrew McThenia
& Thomas L. Shaffer, For Reconciliation, 94 YALE L.J. 1660 (1985); Carrie Menkel
Meadow, For and Against Settlement: Uses and Abuses of the Mandatory Settlement
Conference, 33 UCLA L. REY. 485 (1985) (hereinafter Menkel-Meadow, For and Against
Settlement]; Carrie Menkel-Meadow, Whose Dispute is it Anyway? A Philosophical and
Democratic Defense of Settlement (In Some Cases), 83 GEO. L.J. 2663 (1995) (hereinafter,
Menkel-Meadow, Whose Dispute].

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
362
462 Michigan Law Review [Vol. 102:460
Consistent with this view, legislatures in several states have

enacted statutes that are intended to encourage and protect apologies
by making them inadmissible. 5 In addition, some commentators argue
that defendants might offer expressions of sympathy, rather than
apologies that explicitly accept responsibility for having caused injury,
in order to reap the benefits of apologizing while minimizing the risks. 6
Critics of these so-called "safe" apologies argue, however, that
apologies that avoid the legal consequences of apologizing - whether
because the apology is merely an expression of sympathy or because it
is protected by statute and is inadmissible - are devoid of moral
content and likely ineffectual. 7
Despite the recent surge of interest in, and debate over, the
potential benefits of apologizing in legal cases, there has been very
little empirical exploration8 of the ways in which apologies actually
affect settlement decisionmaking. 9 This Article seeks to fill the gap by
providing much-needed data. The studies described here explore the
proposition that apologies facilitate the settlement of civil disputes
either by increasing potential plaintiffs' inclination to accept a
particular settlement offer or by altering parties' perceptions and
attributions in ways that might smooth the progress toward reaching a
mutually satisfactory settlement agreement. More specifically, these
studies explore the differing ways in which apologies are perceived
and responded to when crafted to better insulate the offeror from
legal liability (e.g., expressions of sympathy and statutorily protected
apologies). This research suggests that an apology may favorably
impact the prospects for settlement but that attention must be paid to
both the nature of the apologetic expression and the circumstances of
the individual case.
Part I describes the legal debate over the role of apologies in
settlement decisionmaking. In particular, this Part describes the legal
admissibility of apologies and the debate surrounding the offering of
legally safe apologies. Part II details the previous empirical research
that has examined the effects of apologies on perception and
decisionmaking. Part III describes the empirical studies and presents
5. See infra Part I.C.

6. See infra Part I.B.
7. See, e.g., Lee Taft, Apology Subverted: The Commodification of Apology, 109 YALE
L.J. 1135 (2000). For further discussion, see infra notes 56-59 and accompanying text.
8. For discussion of the role of empiricism in legal scholarship, see Lee Epstein & Gary
King, The Rules of Inference, 69 U. CHI. L. REV. 1 (2002); Michael Heise, The Importance of
Being Empirical, 26 PEPP. L. REV. 807 (1999).
9. Apologizing may influence a variety of legal decisions in civil cases, including effects
on settlement decisions, decisions about liability, and decisions about compensatory or
punitive damages. All of these decision points are in need of empirical examination. The
focus here is limited to one of these decision points - decisions about settlement.

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
363
the results. Part IV examines the implications of these results for the
debate over evidentiary protection for apologies, for parties to
litigation, and for attorneys representing clients in such litigation.
I. LEGAL DEBATE OVER APOLOGIES

Recently, legal scholars have argued that apologizing has
important benefits for both parties to a lawsuit, including increasing
the possibilities for reaching settlements. 10 Accordingly, these scholars
have suggested that lawyers should discuss apologies with their clients
more often than they now do. They suggest that apologizing may
avoid litigation altogether, and even where it does not it may reduce
tension, antagonism, and anger so as to allow less protracted, more
productive, more creative, and more satisfying negotiation.11 Survey
research suggests that claimants desire apologies and that some would
not have filed suit had an apology been offered. 12 In addition, there is
10. See Cohen, Advising Clients to Apologize, supra note 3; Keeva, supra note 3; Levi,
supra note 3; Orenstein, supra note 3; Peter H. Rehm & Denise R. Beatty, Legal
Consequences of Apologizing, 1996 J. DISP. RESOL. 115; Carl D. Schneider, What It Means
To Be Sorry: The Power of Apology in Mediation, 17 MEDIATION Q. 265 (2000); Shuman,
The Role of Apology in Tort Law, supra note 3; see also Erin Ann O'Hara & Douglas Yarn,
On Apology and Consilience, 77 WASH. L. REV. 1121 (2002).
11. STEPHEN B. GOLDBERG ET AL., DISPUTE RESOLUTION 138 (2d ed., 1992); Cohen,
Advising Clients to Apologize, supra note 3; Levi, supra note 3; Orenstein, supra note 3;
Shuman, The Role of Apology in Tort Law, supra note 3. In addition to the strategic benefits
of apologies for settlement, which are the focus here, a number of nonstrategic benefits of
apologies in civil cases are also posited. Apologies may reduce negative emotions, repair
relationships, fulfill a need to make reparations and to restore equity, make forgiveness
possible, and facilitate psychological growth. GOLDBERG ET AL., supra, at 138; see also
Cohen, Advising Clients to Apologize, supra note 3; Michael E. McCullogh et al.,
Interpersonal Forgiving in Close Relationships, 73 J. PERSONALITY & SOC. PSYCHOL. 321
(1997); Orenstein, supra note 3, at 243-44; Elaine Waister et al., New Directions in Equity
Research, 25 J. PERSONALITY & Soc. PSYCHOL. 1 (1973); Gerald R. Williams, Negotiation as
a Healing Process, 1996 J. DISP. RESOL. 1, 53; Charlotte vanOyen Witvliet et al., Please
Forgive Me: Transgressors' Emotions and Physiology During Imagery of Seeking Forgiveness
and Victim Responses, 21 J. PSYCHOL. & CHRISTIANITY 219 (2002); Charlotte Witvliet et al.,
Victims' Heart Rate and Facial EMG Responses to Receiving an Apology and Restitution,
PSYCHOPHYSIOLOGY 588 (2002).
12. See Thomas H. Gallagher et al., Patients' and Physicians' Attitudes Regarding the
Disclosure of Medical Errors, 289 JAMA 1001 (2003) (finding that patients emphasized a
desire to receive an apology following a medical error); Gerald B. Hickson et al., Factors
That Prompted Families to File Medical Malpractice Claims Following Prenatal Injuries, 267
JAMA 1359, 1361 (1992) (noting that 24% filed claims "when they realized that physicians
had failed to be completely honest with them about what happened, allowed them to believe
things that were not true, or intentionally misled them"); Charles Vincent et al., Why Do
People Sue Doctors? A Study of Patients and Relatives Taking Legal Action, 343 LANCET
1609, 1612 (1994) (finding that 37% of respondents said that they would not have sued had
there been a full explanation and an apology and 14% indicated that they would not have
sued had there been an admission of negligence); Amy B. Witman et al., How Do Patients
Want Physicians to Handle Mistakes? A Survey of Internal Medicine Patients in an Academic
Setting, 156 ARCHIVES OF INTERNAL MED. 2565, 2566 (1996) (finding that 98% of
respondents "desired or expected the physician's active acknowledgement of an error. This

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
364
anecdotal evidence of injured parties who would not have filed

lawsuits had apologies been proffered, 1 3 of settlement negotiations
coming to a standstill over the issue of apology even after agreement
on an appropriate damage amount has been reached, 14 of plaintiffs
who would have preferred an apology as part of a settlement, 15 and of
occasions on which a failure to apologize promoted litigation by
adding insult to injury. 16 In those few cases that go to trial, it is
suggested that a defendant who has apologized will look better in
front of a jury who is asked to award damages, particularly punitive
damages.17 Decisions by corporations such as Wal-Mart and Ford to
ranged form a simple acknowledgement of the error to various forms of apology" and that
" [p]atients were significantly more likely to either report or sue the physician when he or she
failed to acknowledge the mistake.").
13. Bruce W. Neckers gives an example:
In a case in which I represented the plaintiff, the wrongdoer himself tearfully acknowledged
his role in the tragic accidental death of my client's son. It had a huge impact on the
settlement of the case. There never would have been a lawsuit if the same person had made
the same comments to the mother during the 30-day period in which her son lay dying in the
hospital, or during the three days his young body was at the funeral home. The sad part in
that case is that the defendant and his company wanted to express the same thought near the
time of the accident, but claimed to have been prohibited from doing so by their insurance
carrier.
The Art of the Apology, MICH. B.J., June 2002, at 10, 1 1 .
14. See Schneider, supra note 1 0 , at 274 (describing negotiations stalling "over the
plaintiff's demand for an apology, even after the sides had agreed on the damages to be
paid") (emphasis omitted).
15. See, e. g. , Piper Fogg, Minnesota System Agrees to Pay $500,000 to Settle Pay-Bias
Dispute, CHRON. HIGHER EDUC., Feb. 14, 2003, at A12 (describing class-action plaintiff's
disappointed reaction to the settlement: "I want an apology," she said, "and I am never
going to get it") (internal quotes omitted); Editorial, The Paula Jones Settlement, WASH.
POST, Nov. 15, 1998, at C6; Nathalie Des Rosiers et al., Legal Compensation for Sexual
Violence: Therapeutic Consequences and Consequences for the Judicial System, 4 PSYCHOL.
PUB. POL'Y & L. 433, 442 ( 1 998).
16. Cohen, Advising Clients to Apologize, supra note 3; Orenstein, supra note 3, at 243.
17. Cohen, Advising Clients to Apologize, supra note 3; Williams, supra note 1 1 , at 52-53
n.147. A number of states specifically provide that an apology can be considered in
mitigation of damages in defamation cases. See, e. g. , FLA. STAT. § 770.02 (2003); MISS CODE
ANN. § 95-1-5 (1999); TENN. CODE ANN. § 29-24-103 (2000); TEX. CIV. PRAC. & REM. CODE
ANN. § 73.003 (1997); VA. CODE ANN. § 8.01-48, 8.01-46 (2000); W. VA. CODE § 57-2-4
(1997). Some, however, worry that apologizing, "which implies knowledge of wrongdoing,
will exacerbate damage awards by showing a level of intent beyond mere negligence." Levi,
supra note 3, at 1 1 87. It is likely that the effects of apologizing on damage-award
decisionmaking will be complex. Bornstein conducted a set of empirical studies to examine
the effects of remorse on damage-award decisionmaking. Brian H. Bornstein et al., The
Effects of Defendant Remorse on Mock Juror Decisions in a Malpractice Case, 20 BEHAV.
SCI. & L. 393 (2002). In the first study, male participants awarded marginally Jess in damages
against the physician who expressed remorse at the time of trial or who did nothing to
indicate remorse or a lack thereof than they did against physicians who were remorseless or
who expressed remorse early (at the time of the incident) . Id. at 399-400. For female
participants there was no effect of remorse. Id. In a second study, participants awarded more
in compensatory damages against the physician who displayed remorse at the time of the
event and then again at the time of trial than they did in the other three conditions. Id. at
403. Given potential spillover between liability and damages decisions, it is unclear how

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
365
issue apologies in recent cases presumably reflect this intuition about

apologies. 1 8
A. Apologies and Admissibility

Many are concerned, however, that fears of apologies being used
to establish legal liability chill the offering of apologies. In the context
of civil disputes, the conventional wisdom among legal actors has been
that an apology will be viewed as an admission of responsibility and
will lead to increased legal liability - and accordingly, that apologies
ought to be avoided. 19 Consistent with this view, Hiroshi Wagatsuma
these results might have been affected by telling jurors to assume that the defendant was
liable. It is possible that jurors used the available decision to achieve goals that might
otherwise have been achieved through the rendering of a liability verdict. See generally
Jennifer K. Robbennolt et al., Symbolism and Incommensurability in Civil Sanctioning:
Decision-Makers as Goal Managers, BROOK. L. REV. (forthcoming).
18. Patti Waldmeir, Ford Goes Into the Business of Saying Sorry, FIN. TIMES, Jan. 31,
2001, at 8; Wendy Zellner, Wal-Mart: Why an Apology Made Sense, Bus. WK., July 3, 2000,
at 65. There are numerous recent examples of apologies in cases with possible legal
implications. See, e.g., WILLIAM J. CLINTON, Remarks Prior to the House Judiciary
Committee Vote on the First Article of Impeachment, in PUBLIC PAPERS OF THE PRESIDENTS
OF THE UNITED STATES 2158 (1 999); Pam Belluck with Frank Bruni, Law, Citing Abuse
Scandal, Quits as Boston Archbishop and Asks for Forgiveness, N .Y. TIMES, Dec. 14, 2002 at
Al; Cathy Burke, N.J. 's Top Cop Apologizes for Race Profiling, N.Y. POST, Jan. 1, 2001, at
24; Richard Huff, ABC Apology Snuffs Out Tobacco Firms' Lawsuits, DAILY NEWS, Aug.
22, 1995, at 4; John McCormick, At MIT, the Party's Over: A Tragic Alcohol Case Concludes
With a Costly Apology, NEWSWEEK, Sept. 25, 2000, at 45 (describing MIT's apology to the
family of a student who was killed in a fraternity drinking incident); John Schmeltzer, We 're
Sorry, United's CEO Says in TV Ad, CHI . TRIB., Aug. 25, 2000, at 1; Man Pleads Guilty to
Diluting Drugs, N.Y. TIMES, Feb. 27, 2002, at A16 (describing apology by pharmacist
accused of diluting chemotherapy drugs). See generally Jonathan R. Cohen, Apology and
Organizations: Exploring an Example from Medical Practice, 27 FORDHAM URB. L.J. 1447
(2000) [hereinafter Cohen, Apology and Organizations]; Laurent Belsie, The Rise of the
Corporate Apology, CHRISTIAN SCI. MONITOR, Sept. 1 3, 2000, at 1; Mike France, The Mea
Culpa Defense, Bus. WK., Aug. 26, 2002, at 76. Recently, even a group of jurors offered an
apology to the defendant they convicted. Dean E. Murphy, Jurors Who Convicted Marijuana
Grower Seek New Trial, N.Y. TIMES, Feb. 5, 2003, at A14 (describing public apologies by the
jurors).
19. See generally Cohen, Advising Clients to Apologize, supra note 3; Levi, supra note 3.
Empirically, it is not clear whether, under what circumstances, or to what degree an apology
might alter the risk of an adverse liability determination. I am aware of no empirical studies
that address whether apologies influence liability decisionmaking in civil cases. The single
published study of the effects of apologies on juror decisionmaking in civil cases examined
only the effect of apologizing on damage awards. Bornstein et al., supra note 17. Studies
examining attributions of responsibility in nonlegal contexts have found that offenders who
apologize are seen as having acted less intentionally and are blamed less. See Bruce W.
Darby & Barry R. Schlenker, Children's Reactions to Apologies, 43 J. PERSONALITY & Soc.
PSYCHOL. 742, 746, 749 (1982) [hereinafter Darby & Schlenker, Children's Reactions to
Apologies]; Bruce W. Darby & B arry R. Schlenker, Childrens ' Reactions to Transgressions:
Effects of the Actor's Apology, Reputation, and Remorse, 28 BRIT. J. Soc. PSYCHOL. 353,
358-59 (1989) [hereinafter Darby & Schlenker, Children's Reactions to Transgressions]; Ken
ichi Ohbuchi & Kobun Sato, Children's Reactions to Mitigating Accounts: Apologies,
Excuses, and Intentionality of Harm, 134 J. Soc. PSYCHOL. 5, 1 1 (1994); Steven J. Scher &
John M. Darley, How Effective Are the Things People Say to Apologize? Effects of the
Realization of the Apology Speech Act, 26 J. PSYCHOLINGUISTIC RES. 127, 1 34-36 (1997).

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
366
and Arthur Rosett note that "[a] crucial inhibition to a person making
an apology in an American legal proceeding is the possibility that a
sincere apology will be taken as an admission: evidence of the
occurrence of the event and of the defendant's liability for it. "20
Weiner and his colleagues found that offering a "confession" (that included an apology and
acceptance of responsibility) reduced attributions of responsibility to internal causes and
increased attributions to external causes. Bernard Weiner et al., Public Confession and
Forgiveness, 59 J. PERSONALITY 281 , 291 ( 1 991) ("This is in accord with a conceptualization
linking confession to perceived lack of responsibility (in spite of an admission of
responsibility! ) ."). This was true in particular when the cause or causes of the incident were
less clear. Id. at 295 . Some studies, however, have found that an apology increases
attributions of responsibility to the offender. See, e.g., Ken-ichi Ohbuchi et al., Apology as
Aggression Control: Its Role in Mediating Appraisal of and Response to Harm, 56 J.
PERSONALITY & SOC. PSYCHOL. 219, 221 ( 1 989).
20. Wagatsuma & Rosett, supra note 1, at 483. For comparative work on the role of
apologies in different cultures, see, for example, Dean C. B arnlund & Miho Yoshioka,
Apologies: Japanese and American Style, 14 lNT'L J. lNTERCULTURAL REL. 193 (1990); John
0. Haley, Apology and Pardon: Learning From Japan, 41 AM. BEHA V. SCIENTIST 842
(1998) ; V. Lee Hamilton & Shigeru Hagiwara, Roles, Responsibility, and Accounts Across
Cultures, 27 INT'L J. PSYCHOL. 157 (1992); Letitia Hickson, The Social Contexts of Apology
in Dispute Settlement: A Cross-Cultural Study, 25 ETHNOLOGY 229 (1986); Janet Holmes,
Apologies in New Zealand English, 19 LANGUAGE SOC'Y 155 ( 1 990); Ritsu Itoi et al., A
Cross-Cultural Study of Preference of Accounts: Relationship Closeness, Harm Severity, and
Motives of Account Making, 26 J. APPLIED Soc. PSYCHOL. 913 ( 1996) ; Naomi Sugimoto, A
Japan- U. S. Comparison of Apology Styles, 24 COMM. RES. 349 ( 1 997); Wagatsuma & Rosett,
supra note 1 .
367
Thus, attorneys and others fear that any apology will be admitted
into evidence as an admission of fault.27 Consequently, some clients
are hesitant to apologize. Likewise, lawyers and insurance companies
may be unlikely to advise their clients to apologize or to make any
statement that could be construed as an apology. In fact, they may
actively discourage such statements.28
27. Cohen, Advising Clients to Apologize, supra note 3, at 1010 ("If a lawyer
contemplates an apology, it may well be with a skeptical eye: Don't risk apology, it will just
create liability."); Gallagher et al., supra note 12 (describing physicians' concern that
offering an apology will lead to legal liability).
28. See GOLDBERG ET AL., supra note 11, at 138; Gallagher et al., supra note 12; Levi,
supra note 3, at 1 1 86.
368
B. Expressions of Sympathy
In response to these concerns, several commentators have
suggested that attorneys and clients consider crafting safe apologies
that minimize the risk that they will be used as evidence of liability. 29
Specifically, defendants might offer partial apologies that merely
express sympathy, but do not admit responsibility.
As a general matter, in accounting for wrongful behavior, an
offender can assert his or her innocence (e.g., by declaring he or she
was not involved), offer an excuse that minimizes responsibility, or
offer a justification for the act that legitimizes it. Alternatively, the
offender can offer an apology that admits blame and expresses
regret. 30 In defining apologies as remedial work, Erving Goffman
writes:
In its fullest form, the apology has several elements: expression of
embarrassment and chagrin; clarification that one knows what conduct
had been expected and sympathizes with the application of negative
sanction; verbal rejection, repudiation, and disavowal of the wrong way
of behaving along with vilification of the self that so behaved; espousal of
the right way and an avowal henceforth to pursue that course;
performance of penance and the volunteering of restitution. 3 1
Similarly, Nicholas Tavruchis suggests that, at a minimum, an apology
must incorporate "acknowledgment of the legitimacy of the violated
rule, admission of fault and responsibility for its violation, and the
29. See, e.g., Cohen, Advising Clients to Apologize, supra note 3.

30. Sociologists and psychologists distinguish apologies from other types of accounts or
remedial exchanges, such as denial, excuse, or justification. See ERVING GOFFMAN,
RELATIONS IN PUBLIC: MICROSTUDIES OF THE PUBLIC ORDER (1971); NICHOLAS
TAVUCHIS, MEA CULPA: A SOCIOLOGY OF APOLOGY AND RECONCILIATION (1991); Barry
R. Schlenker & Michael F. Weigold, Interpersonal Processes Involving Impression
Regulation and Management, 43 ANN. REV. PSYCHOL. 133 ( 1 992); Marvin B. Scott &
Stanford M. Lyman, Accounts, 33 AM. Soc. REV. 46 ( 1 968); see also MARTHA MINOW,
B ETWEEN VENGEANCE AND FORGIVENESS 1 15 (1998) ("Full acceptance of responsibility
by the wrongdoer is the hallmark of an apology.").
3 1 . GOFFMAN, supra note 30, at 1 13. Other authors have defined the elements of a full
apology in various ways. For example, Orenstein argues that:
At their fullest, apologies should: (1) acknowledge the legitimacy of the grievance and
express respect for the violated rule or moral norm; (2) indicate with specificity the nature of
the violation; (3) demonstrate understanding of the harm done; (4) admit fault and
responsibility for the violation; (5) express genuine regret and remorse for the injury; (6)
express concern for future good relations; (7) give appropriate assurance that the act will not
happen again; and, if possible, (8) compensate the injured party.
Orenstein, supra note 3, at 239. Wagatsuma and Rosett distinguish five components of a
"meaningful" apology: the offender must acknowledge that " l . the hurtful act happened,
caused injury, and was wrongful; 2. the apologizer was at fault and regrets participating in
the act; 3. the apologizer will compensate the injured party; 4. the act will not happen again;
and 5. the apologizer intends to work for good relations in the future." Wagatsuma &
Rosett, supra note 1 , at 469-70. Cohen identifies three elements: "(i) admitting one's fault,
(ii) expressing regret for the injurious action, and (iii) expressing sympathy for the other's
injury." Cohen, Advising Clients to A pologize, supra note 3, at 1014-15.

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
369
expression of genuine regret and remorse for the harm done." 32 In

some sense, then, it is the admission of responsibility that distinguishes
an apology from other forms of accounting for one's behavior.
Nonetheless, some legal commentators suggest that defendants
might "safely" offer apologies that are incomplete according to the
above definitions - specifically, apologies that express sympathy, but
do not admit fault or responsibility for the incident. For example,
Deborah Levi suggests that "lawyers protective of their clients'
interests might serve those interests by encouraging clients to
apologize short of admitting liability." 33 Similarly, Jonathan Cohen
suggests that " [w] hile a full apology will usually be most powerful,
'merely' expressing sympathy can often be a large step." 34 Defendants,
therefore, might offer an apology that conveys sympathy - "I am
sorry that you have been injured" - but does not acknowledge
responsibility for the injuries inflicted. 35
Relying primarily on anecdotes, commentators assert that even
these partial apologies36 can avoid lawsuits and facilitate settlement. 37
While the intuition is that such apologies will not be as effective as
more complete apologies that acknowledge responsibility, partial
apologies are thought by some to be better than failing to apologize at
all. 38 Some attorneys offer such expressions of sympathy, without
admitting fault, in the belief that it is the right thing to do and that it
helps to settle cases more quickly and favorably. 39
32. TA VU CHIS, supra note 30, at 3. Tavuchis argues that other features of an apology are
implicit in "two fundamental requirements: the offender has to be sorry and has to say so."
Id. at 36.
33. Levi, supra note 3, at 1 1 88. See also Gallagher et al., supra note 12, at 1004
(describing physicians' tendencies to " 'choos[e] their words carefully' when talking with
patients about errors.").
34. Cohen, Advising Clients to Apologize, supra note 3, at 1048.
35. This distinction was central to the disagreement over the U.S.-China surveillance
plane incident as the United States offered only an expression of sympathy for the death of a
Chinese pilot, while China demanded a full apology that took responsibility for the incident.
Careful Language Breaks Washington-Beijing Impasse, at http://www.cnn.com/2001/
WORLD/asiapcf/east/04/1 1/collision.letter.analysis.02/ (April 1 1 , 2001). While this incident
differs from civil litigation because it involved an apology in a public, political context, it
illustrates the general tension at issue.
36. In the present Article, "partial apology" will be used to refer to a statement that
expresses sympathy, but does not admit responsibility. These will be contrasted with "full
apologies," in which the offender both expresses sympathy and accepts responsibility. These
have been variously referred to as "effective" apologies, "authentic" apologies, "genuine"
apologies, and "happy-ending" apologies.
37. See Cohen, Advising Clients to Apologize, supra note 3 ; Orenstein, supra note 3;
Elizabeth Latif, Note, Apologetic Justice: Evaluating Apologies Tailored Toward Legal
Solutions, 81 B.U. L. REV. 289 (2001).
38. See Orenstein, supra note 3.
39. See Cohen, Apology and Organizations, supra note 18.
370
Moreover, many supporters of partial apologies believe that judges

and jurors will be able to distinguish between expressions of sympathy
for a plaintiff's injuries and apologies that accept responsibility for
having caused those injuries. Accordingly, they argue that the risk that
a partial apology will be misinterpreted as evidence of liability is
minimal. Peter Rehm and Denise Beatty review the few published
appellate opinions involving apologies and conclude that "Li]udges and
juries understand that expression of sympathy, regret, remorse and
apology are not necessarily admissions of responsibility or liability. "40
Similarly, Cohen argues that "most judges and juries are good at
distinguishing between expressions of sympathy and expressions of
fault or regret, and may even be favorably disposed toward defendants
who express sympathy. " 41 Others, however, "fear that a statement of
sympathy will be mistaken for an assumption of culpability [ and] do
not issue" even these partial apologies.42
C. Statutorily Protected Apologies

Such concerns have led some to argue that apologies ought to be
rendered safe by preventing them, through the rules of evidence, from
being admissible in court as evidence of liability. 43 In recent years,
several states have adopted rules of evidence that offer such
protection to certain apologies44 and a number of additional states
have considered or are considering similar legislation. 45 Such safe
40. Rehm & Beatty, supra note 10. See also William K. Bartels, The Stormy Seas of
Apologies: California Evidence Code Section 1160 Provides a Safe Harbor for Apologies
Made After Accidents, 28 W. ST. U. L. REV. 141 (2000-01).
42. See Orenstein, supra note 3, at 240. This is, of course, an empirical question that
awaits further research. For cases in which the admissibility or sufficiency of particular
apologetic expressions was at issue, see, for example, Denton v. Park Hotel, 180 N.E.2d 70
(Mass. 1962); Giangrasso v. Schimmel, 207 N.W.2d 517 (Neb. 1973); Phinney v. Vinson, 605
A.2d 849 (Vt. 1992); and Senesac v. Assocs. in Obstetrics & Gynecology, 449 A.2d 900 (Vt.
1982).
43. See, e.g., Cohen, Advising Clients to Apologize, supra note 3, at 1062-63; Jonathan R.
Cohen, Legislating Apology: The Pros and Cons, 70 U. CIN. L. REV. 819 (2002) [hereinafter
Cohen, Legislating Apology]; Orenstein, supra note 3, at 247-48 (proposing to "except
apologies and admissions of fault in civil cases."). See generally O'Hara & Yarn, supra note
10.
44. See CAL. Evrn. CODE § 1160 (West Supp. 1995); COLO. REV. STAT. § 13-25-135
(2003); FLA. STAT. ch. 90.4026 (2003); MASS. GEN. LAWS ch. 233, § 23D (2002); TEX. CIV.
PRAC. & REM. CODE ANN. § 18.061 (Vernon 2004); WASH. REV. CODE § 5.66.010(1) (2002).
45. See e.g., S. 577, 2001 Gen. Assem., Reg. Sess. (Conn. 2001); S. 1477, 21st Leg., Reg.
Sess. (Haw. 2001); S. 439, 92d Gen. Assem., Reg. Sess. (Ill. 2001); H.R. 1002, 112th Gen.
Assem., 1st Reg. Sess. (Ind. 2001); S. 1071, 79th Gen. Assem., 1st Sess. (Iowa 2001); S. 467,
2001 Gen. Assem., 415th Sess. (Md. 2001); S. 528, 92d Leg., Reg. Sess. (Mich. 2003); S. 280,
92d Gen. Assem., 1st Reg. Sess. (Mo. 2003); Draft Bill 1000, 2003 Leg., 58th Sess. (Mont.
2003) (draft bill never formally introduced) ; S. 1262, 48th Leg., 2d Sess. (Okla. 2002); S. 339,
2003 Gen. Assem., Jan. Sess. (R.I. 2003); H.R. 6905, 2002 Gen. Assem., Jan. Sess. (R.I.

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
371
harbors have been said to encourage parties to apologize and,

consequently, to promote the settlement of civil cases.46
In 1986, Massachusetts became the first state to adopt a rule of
evidence designed to preclude admission of apologies. The
Massachusetts statute provides that:
Statements, writings or benevolent gestures expressing sympathy or a
general sense of benevolence relating to the pain, suffering or death of a
person involved in an accident and made to such person or to the family
of such person shall be inadmissible as evidence of an admission of
liability in a civil action. 47
This provision renders inadmissible apologies that are expressions of
sympathy (e.g., "I am sorry that you are hurt"), but does not clearly
address the admissibility of more fully apologetic expressions that also
take responsibility for the injury-causing incident ( e.g., "I am sorry
that I hurt you").
Subsequently, several other states, including Texas (1999),
California (2000), Florida (2001), and Washington (2002), have
enacted rules of evidence protecting apologies that are expressions of
sympathy from being admissible at trial. 48 The Florida statute, for
example, protects statements expressing sympathy from admissibility
but makes it clear that statements acknowledging fault remain
admissible:
The portion of statements, writings, or benevolent gestures expressing
sympathy or a general sense of benevolence relating to the pain,
suffering, or death of a person involved in an accident and made to that
person or to the family of that person shall be inadmissible as evidence in
a civil action. A statement of fault, however, which is part of, or in
2002); H.R. 5628, 2001 Gen. Assem., Jan. Sess. (RI. 2001); H.R. 67, 103d Gen. Assem., 1st
Reg. Sess. (Tenn. 2003); H.R. 2185, 102d Gen. Assem., 2d Reg. Sess. (Tenn. 2002); S. 587,
75th Leg., 1st Sess. (W. Va. 2001). A handful of cases also provide some implicit protection
for certain apologetic expressions. See, e.g. , Denton, 180 N.E.2d 70 (Mass. 1962) (finding that
defendant's apology was an expression of sympathy and should not have been admitted);
Giangrasso, 207 N.W.2d at 518 ("The word 'sorry' in conjunction with other language or
circumstances may constitute an admission, denoting apology. Standing alone, it is not an
admission of negligence; it may mean regret, not apology. The statement was not an
admission of negligence, and the evidence would have been insufficient to support a verdict
for plaintiff."); Phinney, 605 A.2d 849 (Vt. 1992) (finding that a defendant's apology was
insufficient to show a breach of the standard of care); Senesac, 449 A.2d 900 (Vt. 1982)
(same).
46. See Cohen, Advising Clients to Apologize, supra note 3; Cohen, Legislating Apology ,
supra note 43; Latif, supra note 37; Orenstein, supra note 3; Shuman, The Role of Apology in
Tort Law, supra note 3.
47. MASS. GEN. LAWS ch. 233, § 23D (2002).
48. CAL. EVID. CODE § 1160 (West Supp. 1995); FLA. STAT. ch. 90.4026 (Supp. 2004);
MASS. GEN. LAWS ch. 233, § 23D (2002); TEX. Crv. PRAC. & REM. CODE ANN. § 18.061
(Vernon 2004); WASH. REV. CODE § 5.66.010(1) (1995).
372
addition to, any of the above shall be admissible pursuant to this

section. 49
While most of this legislation specifically protects expressions of
sympathy and does not protect statements that admit fault,50 some
have argued that such protection ought to also be afforded to full
apologies that accept responsibility. 5 1 Most recently, in 2003 Colorado
adopted a rule that appears to take a step in this direction. The
Colorado rule provides:
In any civil action brought by an alleged victim of an unanticipated
outcome of medical care, or in any arbitration proceeding related to such
civil action, any and all statements, affirmations, gestures, or conduct
expressing apology, fault, sympathy, commiseration, condolence,
49. FLA. STAT. ch. 90.4026(2) (Supp. 2004). The California, Washington, and Texas
statutes are substantially similar to the Florida statute. The Washington statute closely tracks
the Florida statute in providing:
The portion of statements, writings, or benevolent gestures expressing sympathy or a general
sense of benevolence relating to the pain, suffering, or death of a person involved in an
accident, and made to that person or to the family of that person, shall be inadmissible as
evidence in a civil action. A statement of fault, however, which is part of, or in addition to,
any of the above shall not be made inadmissible by this section.
WASH. REV. CODE § 5 .66.010(1) (2002). Similarly, the California statute provides:
The portion of statements, writings, or benevolent gestures expressing sympathy or a general
sense of benevolence relating to the pain, suffering, or death of a person involved in an
accident and made to that person or to the family of that person shall be inadmissible as
evidence of an admission of liability in a civil action. A statement of fault, however, which is
part of. or in addition to, any of the above shall not be inadmissible pursuant to this section.
CAL. Evrn. CODE § 1 1 60(a) (West Supp. 1995). The Historical Notes to the California
statute include hypothetical examples that make clear the intentions of the drafters that
statements admitting fault will still be admissible:
Hypothetical #1: An automobile accident occurs and one driver says to the other: "I'm
sorry you were hurt." -or-"I'm sorry that your car was damaged.'' Under the bill, these
statements would not be admissible in court.
Hypothetical #2: The same accident occurs and one driver says to the other: "I'm sorry you
were hurt, the accident was all my fault.''-or-"I'm sorry you were hurt, I was using my cell
phone and just didn't see you coming." Under the bill, only the portions of the statements
containing the apology would be inadmissible; any other expression acknowledging or
implying fault would continue to be admissible, consistent with present evidentiary
standards.
CAL. ASSEMBLY COMM. ON JUDICIARY, HISTORICAL NOTES TO CAL. EVID . CODE § 1 160
(West Supp. 1995) . The Texas statute provides:
(a) A court in a civil action may not admit a communication that: (1) expresses sympathy or
a general sense of benevolence relating to the pain, suffering, or death of an individual
involved in an accident; (2) is made to the individual or a person related to the individual . . .
and (3) is offered to prove liability of the communicator in relation to the individual. . . . (c)
Notwithstanding the provisions of Subsections (a) and (b), a communication . . . which also
includes a statement or statements concerning negligence or culpable conduct pertaining to
an accident or event, is admissible to prove liability of the communicator.
TEX. CIV. PRAC. & REM. CODE ANN. § 18.061 (Vernon 2004).
50. See CAL. Evrn. CODE § 1 1 60 (West 1 995) ; FLA. STAT. ch. 90.4026 (2002); TEX. Civ.
PRAC. & REM. CODE ANN. § 18.061 (Vernon 2004); WASH. REV. CODE § 5.66.010(1 ) (1995).
5 1 . Orenstein, supra note 3; see also Cohen, Legislating Apology, supra note 43.

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
373
compassion, or a general sense of benevolence which are made by a

health care provider or an employee of a health care provider to the
alleged victim, a relative of the alleged victim, or a representative of the
alleged victim and which related to the discomfort, pain, suffering, injury,
or death of the alleged victim as the result of the unanticipated outcome
of medical care shall be inadmissible as evidence of an admission of
liability or as evidence of an admission against interest. 52
While the Colorado rule is limited in scope to expressions made by
health-care providers, the provision is notable in that it appears to be
the first statute that explicitly protects statements expressing fault. 53
D. Debate over Perceptions of Safe Apologies

How these safe apologies - either those made safe through
careful wording or those that are statutorily protected - will be
perceived has been a subject of some controversy but little empirical
examination. First, it is the intuition of many that a partial, or
sympathy-only, apology is less satisfying and less effective at
promoting settlement than a more complete apology in which the
other party not only expresses sympathy, but also accepts
responsibility for having caused injury. 54 Something seems to be lost
when a wrongdoer fails to acknowledge wrongdoing. Thus, Steven
Scher and John Darley comment that "[t]he admission of
responsibility for the transgression is a necessary gesture of an apology
because it conveys to the listener that the speaker is aware of the
social norms that have been violated . . . , and therefore conveys that
the speaker will be able to avoid the offense in future interactions." 55
Similarly, while recognizing that allowing apologies to be
introduced against the apologizer in a subsequent legal proceeding
52. COLO. REV. STAT. § 13-25-135 (2003).

53. The Colorado legislation, widely viewed as being part of a "concerted effort" of tort
reform in Colorado, was billed by its sponsors as the "I'm sorry legislation," while criticized
by opponents as being anti-patient rights by insulating doctors from liability. Arthur Kane,
GOP Pushes Tort Reform, DENVER POST, Apr. 6, 2003, at B4 (characterizing the bill as part
of a "flurry of bills to limit lawsuits and damage awards"); Al Lewis, Malpractice Measure is
"Sorry " Protection, DENVER POST, Apr. 13, 2003, at Kl (describing the bill as "among a
dozen tort reform proposals"); Peggy Lowe, "Sorry " Bill Advances, ROCKY MOUNTAIN
NEWS, Apr. 2, 2003, at 22A (describing the different characterizations of the bill by its
sponsors and opponents). Also in 2003, the Texas legislature considered, as part of a larger
tort reform package, a measure that would have repealed the provision in its statute that
makes clear that statements that acknowledge fault remain admissible. H.R. 4, 78th Leg.,
Reg. Sess. (Tex. 2003) (March 28, 2003 version). This provision, however, did not make it
into the final version of the bill that was signed by the governor. H.R. 4, 78th Leg., Reg. Sess.
(Tex. 2003) (June 1 1, 2003, enacted version).
54. See Cohen, Advising Clients to Apologize, supra note 3; Levi, supra note 3.
55. Scher & Darley, supra note 19, at 130. See also O'Hara & Yarn, supra note 10, at
1137 ("The more unambiguously and emphatically a transgressor accepts blameworthiness,
the more likely the apology will induce meaningful reconciliation.").
374
may have a chilling effect on such expressions of remorse, some

commentators argue that avoiding the legal consequences of
apologizing, either by offering a partial apology or because the
apology is legally protected from admissibility, diminishes the
apology's moral content. 56 Thus, Lee Taft argues that:
The law recognizes that an apology, when authentically and freely made,
is an admission; it is an unequivocal statement of wrongdoing. The law
permits such an acknowledgement to enter the legal process as a way to
allow the performer of apology to experience the full consequences of
the wrongful act. An apology made in this context, with full knowledge
of the legal ramifications, is much more freighted than an apology made
in a purely social context. 57
Accordingly, Taft argues that when an apology is inadmissible to show
liability, "the moral dimension of apology is entirely eclipsed. "58 In
other words, to protect an apology or to make it safe is to diminish it
and to change its meaning. Moreover, Cohen argues that " [t ] o many
the message, 'I'm sorry, but I don't want anything I'm saying to be
used against me to make me pay for it,' rings empty. Many may think
that if you are unwilling to 'put your money where your mouth is' you
are insincere. " 59 Such interpretation could lead recipients to discount
safe apologies, rendering them ineffective.
Others, however, make the case for carefully crafted apologies and
increased evidentiary protection for apologies, arguing that even
legally safe apologies are socially useful and can promote settlement,
particularly where more complete apologies are unlikely to be
forthcoming. 60 "While most people would prefer receiving an apology
that involves financial exposure to one that does not, I think most
people would prefer receiving an apology that does not involve
financial exposure to receiving no apology at all. "61 Thus, proponents
of protecting apologies argue that " [b] ecause apologies are so crucial
to social interaction and personal peace, it is desirable that law
facilitate or, at least, not hinder the possibility of this healing ritual. "62
56. See Taft, supra note 7. See also arguments articulated in Cohen, Legislating
Apology, supra note 43.
57. Taft, supra note 7, at 1 157.
58. Id. at 1 150.
60. Id.; Cohen, Legislating Apology, supra note 43; Latif, supra note 37; Orenstein,
supra note 3.
61. Cohen, Legislating Apology, supra note 43, at 858.
62. Orenstein, supra note 3, at 247.
375
II. PRIOR PSYCHOLOGICAL RESEARCH ON APOLOGIES

Given recent calls for apologies in civil litigation and efforts to
reform the law to protect them, it is important to know whether and in
what ways apologies really do influence settlement decisionmaking.
Such an understanding will better equip courts and legislatures to
draft sensible and effective rules. In addition, such information will
help parties to a dispute more clearly understand the utility of
apologies for resolving disputes or, perhaps more importantly, for
resolving disputes more quickly63 or more satisfactorily. 64
A. Psychological Responses to Apologies

A variety of perceptions and attributions are thought to underlie
negotiation behavior and to influence the outcomes of settlement
negotiations. 65 Experimental studies in primarily nonlegal contexts
have demonstrated that an apology or an expression of remorse may
influence a number of these perceptual and attributional judgments.
Apologies or expressions of remorse have been found to influence
attributions of responsibility or blame for the incident,66 beliefs about
the stability of the behavior (i.e., its likelihood of recurrence),67
expectations about the future relationship between the parties,68
perceptions of the character of the wrongdoer,69 affective reactions
63. Heise found that the median disposition time for a civil case from the time of filing
to a j ury verdict was over two years. Michael Heise, Justice Delayed?: An Empirical Analysis
of Civil Case Disposition Time, 50 CASE W. RES. L. REV. 813 (2000). Given the tendency of
many filed cases to settle late in the process, see Metzloff, supra note 4, at 59 & n.54, the
settlement process may be protracted even for cases that will ultimately be resolved by
settlement.
64. See Chris Guthrie & James Levin, A "Party Satisfaction " Perspective on a
Comprehensive Mediation Statute, 13 OHIO ST. J. ON DISP. RESOL. 885 (1998); infra notes
237-242 and accompanying text. But cf Galanter & Cahill, supra note 4; David Luban, The
Quality of Justice, 66 DENY. U. L. REV. 381 (1989).
65. See generally Leigh Thompson, Negotiation Behavior and Outcomes: Empirical
Evidence and Theoretical Issues, 108 PSYCHOL. BULL. 515, 527 (1990) .
66. See, e. g. , Darby & Schlenker, Children's Reactions to Apologies, supra note 19;
Darby & Schlenker, Children 's Reactions to Transgressions, supra note 19; Ohbuchi et al.,
supra note 19; Ohbuchi & Sato, supra note 19; Scher & Darley, supra note 19; Weiner et al.,
supra note 19.
67. See, e.g. , Gregg J. Gold & Bernard Weiner, Remorse, Confession, Group Identity,
and Expectancies About Repeating a Transgression, 22 BASIC & APPLIED Soc. PSYCHOL.
291 (2000); Ohbuchi et al., supra note 19; Jennifer R. Orleans & Michael B. Gurtman,
Effects of Physical Attractiveness and Remorse on Evaluations of Transgressions, 6 ACAD.
PSYCHOL. BULL. 49 (1984) ; Weiner et al., supra note 19.
68. Holley S. Hodgins & Elizabeth Liebeskind, Apology Versus Defense: Antecedents
and Consequences, 39 J. EXPERIMENTAL SOC. PSYCHOL. 297 (2003).
69. See, e.g., Darby & Schlenker, Children's Reactions to Apologies, supra note 19;
Darby & Schlenker, Children's Reactions to Transgressions, supra note 1 9; Gold & Weiner,
supra note 67; Marti Hope Gonzales et al., Victims as "Narrative Critics ": Factors Influencing
Rejoinders and Evaluative Responses to Offenders' Accounts, 20 PERSONALITY & Soc.

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
376
such as anger and sympathy,70 and behaviors such as forgiveness,7 1

aggression,72 and recommendations for punishment. 73
Experimental studies of reactions to criminal defendants have
generally shown that remorseful defendants are perceived more
positively and sentenced more leniently than are defendants who do
not show remorse. 74 Similarly, Brian Bornstein found that defendants
in civil trials who show remorse were perceived more positively by
mock-jurors than those who did not. 75 Each of these psychological
reactions may underlie any effect of apology on settlement
decisionmaking.
PSYCHOL. BULL. 691 ( 1994); Ohbuchi et al., supra note 19; Ohbuchi & Sato, supra note 19;
Orleans & Gurtman, supra note 67; Weiner et al., supra note 19.
70. See, e. g., Mark Bennett & Deborah Earwaker, Victims ' Response to Apologies: The
Effects of Offender Responsibility and Offense Severity, 134 J. Soc. PSYCHOL. 457 ( 1994);
Gold & Weiner, supra note 67; Ohbuchi et al., supra note 19; Weiner et al., supra note 19.
For a discussion of research in emotion in negotiation generally, see Max H. Bazerman et al.,
Negotiation, 51 ANN. REV. PSYCHOL. 279, 285-86 (2000).
71. See, e.g., Darby & Schlenker, Children's Reactions to Apologies, supra note 19; Gold
& Weiner, supra note 67; Ohbuchi & Sato, supra note 19; Weiner et al., supra note 19.
72. See, e.g., Ohbuchi et al., supra note 19.
73. See, e.g., Darby & Schlenker, Children's Reactions to Apologies, supra note 19;
Darby & Schlenker, Children's Reactions to Transgressions, supra note 19; Gold & Weiner,
supra note 67; Weiner et al., supra note 19.
74. See Chris L. Kleinke et al., Evaluation of a Rapist as a Function of Expressed Intent
and Remorse, 132 J. Soc. PSYCHOL. 525 ( 1992) (finding both that a convicted rapist was
judged to have acted less intentionally, to be of less negative character, and to have more
potential for rehabilitation if he demonstrated remorse than if he did not and that
recommended sentences for the convicted rapist were predicted by perceived remorse);
Randolph B. Pipes & Marci Alessi, Remorse and a Previously Punished Offense in
Assignment of Punishment and Estimated Likelihood of a Repeated Offense, 85 PSYCHOL.
REP. 246 ( 1999); Michael G. Rumsey, Effects of Defendant Background and Remorse on
Sentencing Judgments, 6 J. APPLIED Soc. PSYCHOL. 64 ( 1976) ( finding that participants gave
a defendant in a drunk-driving case who was described as "extremely remorseful" a shorter
sentence than they did a defendant who gave "no indication of remorse"); Christy Taylor &
Chris L. Kleinke, Effects of Severity of Accident, History of Drunk Driving, Intent, and
Remorse on Judgments of a Drunk Driver, 22 J. APPLIED Soc. PSYCHOL. 1641 ( 1992)
(firtding that a defendant who expressed remorse was rated as being a person of greater
responsibility and sensitivity than a defendant who did not express remorse, but not finding
significant differences in sentences). See the review of the role of apologies in criminal cases
in Carrie J. Petrucci, Apology in the Criminal Justice Setting: Evidence for Including Apology
as an Additional Component in the Legal System, 20 BEHA V. SCI. & L. 337 (2002). For some
boundary conditions on these types of effects, see Keith E. Neidermeier et al., Exceptions to
the Rule: The Effects of Remorse, Status, and Gender on Decision Making, 31 J. APPLIED
Soc. PSYCHOL. 604 (2001). Interviews with jurors in capital cases also provide evidence that
the degree to which jurors perceive defendants to be remorseful influences their choice
between a sentence of life in prison and death. See Theodore Eisenberg et al., But Was He
Sorry? The Role of Remorse in Capital Sentencing, 83 CORNELL L. REV. 1599 ( 1998).
75. Bornstein et al., supra note 17. In his first study, Bornstein found that remorse had a
significant positive effect on jurors' overall perceptions of the defendant. Id. at 400. In a
second study, Bornstein found that physicians who expressed remorse were perceived as
having suffered more than defendants who did not express remorse. Id. at 404. But see
Bornstein et al., supra note 17, for Bornstein's findings about damage awards.
377
More generally, research on the factors that lead injured parties to

instigate and pursue legal claims suggests several avenues by which an
apology might influence settlement behavior. Research investigating
the claiming process suggests that the course of a dispute is in part
determined by how the dispute is construed by the participants.
Moreover, the research identifies a number of variables that affect the
likelihood that injured persons will name their injury, blame a third
party, and pursue a claim. 76 Whether or not a legal claim is made is
influenced by factors such as whether the injured person identifies an
injury, attributes causation and fault to a third party, perceives that he
or she has been treated unfairly, and knows how to go about pursuing
a claim. 77 Apologies may influence perceptions of fair versus unfair
treatment, attributions of responsibility, and perceived dignity vis-a-vis
the wrongdoer. In these ways, apologies may lead to greater
willingness to settle claims and greater satisfaction with outcomes.
Equity theory also suggests that apologies may be useful in
facilitating settlement. Equity theory posits that a transgression by an
offender against an injured party results in an inequity in their
relationship; that is, the wrong creates a moral imbalance between the
parties. Moreover, "when individuals find themselves participating in
inequitable relationships, they become distressed. The more
inequitable the relationship, the more distress individuals feel. " 78
Upon discovering that a relationship is inequitable, individuals are
motivated to attempt to restore equity to the relationship. 79
Equity theorists have suggested a number of ways in which equity
might be restored to the relationship between the parties, including
the offender offering an apology. 80 An apology may persuade the
injured party that the relationship is, indeed, equitable,81 perhaps in
part because it demonstrates that the wrongdoer has suffered as a
result.82 To apologize is to engage in a social "ritual whereby the
wrongdoer can symbolically bring himself low ( or raise us up)." 83
76. See Felstiner et al., supra note 4; E. Allan Lind, Litigating and Claiming in
Organizations: Antisocial Behavior or Quest for Justice? in ANTISOCIAL BERA VIOR IN
ORGANIZATIONS 150 (R.A. Giacalone & J. Greenberg eds., 1997) (reviewing research).
77. See DEBORAH R. HENSLER ET AL., COMPENSATION FOR ACCIDENTAL INJURIES IN
THE UNITED STATES (1991); Dan Coates & Steven Penrod, Social Psychology and the
Emergence of Disputes, 15 LAW & Soc'Y REV. 655 (1980-81); Felstiner et al., supra note 4;
Richard E. Miller & Austin Sarat, Grievances, Claims, and Disputes: Assessing the Adversary
Culture, 15 LAW & SOC'Y REV. 525 (1980-81).
78. See Waister et al., supra note 11, at 153.
79. See id. at 154.
80. See id. at 163.
81. See id. at 163.
82. See William Austin et al., Equity and the Law: The Effect of a Harmdoer 's "Suffering
in the Act " on Liking and Assigned Punishment, 9 ADVANCES EXPERIMENTAL Soc.
PSYCHOL. 163 (1976); Dana Bramel et al., An Observer's Reaction to the Suffering of his
378
Accordingly, as sociologist Tavuchis recognizes:

Genuine apologies . . . may be taken as the symbolic foci of secular
remedial rituals that serve to recall and reaffirm allegiance to codes of
behavior and belief whose integrity has been tested and challenged by
transgression, whether knowingly or unwittingly. An apology thus speaks
to an act that cannot be undone but that cannot go unnoticed without
compromising the current and future relationship of the parties, the
legitimacy of the violated rule, and the wider social web in which the
participants are enmeshed. 84
Similarly, Jean Hampton argues that "by apologizing, we deny the
diminishment of the victim, and our relative elevation, expressed by
our wrongful action. " 85 In this way, an apology offered by the
transgressor to the victim may repair the breach created by the
wrongful conduct and affirm the relative value of the parties.
One empirically supported reason that an apology may be an
effective repair strategy is that observers may infer from an apology
that the cause of the incident is less stable and, therefore, less likely to
be repeated. 86 Gregg Gold and Bernard Weiner asked participants to
read a scenario describing an incident that was followed by either an
expression of remorse or no such expression. 87 They found that
wrongdoers who expressed remorse were judged to be of higher moral
character and as less likely to act similarly in the future. 88 Consistent
with this notion, Randolph Pipes and Marci Alessi found that a
criminal defendant who demonstrated remorse was perceived as less
likely to reoffend than was a defendant who did not express remorse. 89
In addition, Dawn Robinson, Lynn Smith-Lovin, and Olga Tsoudis
found that a criminal defendant who was perceived as more
remorseful from his visible emotional reaction was rated as having
engaged in the behavior at issue fewer times in the past and as less
Enemy, 8 J. PERSONALITY & Soc. PSYCHOL. 384 (1968); Jerry I. Shaw & James A.
McMartin, Perpetrator or Victim? Effects of Who Suffers in an Automobile Accident on
Judgmental Strictness, 3 SOC. BEHA V. & PERSONALITY 5 ( 1 975); Harry S. Upshaw & Daniel
Romer, Punishment For One's Misdeeds as a Function of Having Suffered From Them, 2
PERSONALITY & SOC. PSYCHOL. BULL. 162 ( 1976).
83. Jeffrie Murphy, Forgiveness and Resentment, in FORGIVENESS AND MERCY 14, 28
(Jean Hampton & Jeffrie G. Murphy eds., 1 988).
84. TAVUCHIS, supra note 30, at 13.
85. Jean Hampton, Correcting Harms Versus Righting Wrongs: The Goal of Retribution,
39 UCLA L. REv. 1659, 1698 (1992).
86. See, e.g., FRITZ HEIDER, THE PSYCHOLOGY OF INTERPERSONAL RELATIONS
(1958) .
87. Gold & Weiner, supra note 67.
88. Id. Participants also expressed more sympathy for remorseful wrongdoers, were
more likely to forgive them, and recommended less punishment. Id.
89. Pipes & Alessi, supra note 74.
379
likely to engage in the behavior in the future. 90 Gold and Weiner posit
that "[o]ne reason for the anticipation of positive future behavior may
be that when an individual confesses with remorse, the moral
character of the offender is recovered," and the wrongful behavior is
not seen as representative of the offender's true character. 9 1 Many
injured parties say that in pursuing lawsuits they are interested in
effecting a change in the wrongdoer's future behavior and desire
assurance that the wrongful behavior will not recur. 92 If these self
reports are accurate, then a recipient's interpretation of an apology as
an indication that the behavior will not be repeated may predict
willingness to settle.
B. The Nature of the Apology

Whether an apology is merely an expression of sympathy or also
includes an acknowledgment of responsibility may affect its impact on
its recipient. Several studies have demonstrated that more elaborate
apologies result in improved attributions. Bruce Darby and Barry
Schlenker evaluated children's responses to different apologies given
following an incident. 93 The children judged wrongdoers who offered
more elaborate apologies more favorably, as better persons who they
liked more, blamed less, were more willing to forgive, and thought
should be punished less.
More recently, Scher and Darley identified several potential
components of an apology: the admission of responsibility, an
expression of remorse, a promise of forbearance, and an offer of
repair. 94 They found that each of these components independently
contributed to the effectiveness of the apology. 95 Nevertheless, "the
greatest improvement in perceptions came from the addition of one
apology strategy - i.e., the offering of an apology, compared to no
90. Dawn T. Robinson et al., Heinous Crime or Unfortunate Accident? The Effects of
Remorse on Responses to Mock Criminal Confessions, 73 Soc. FORCES 175 (1994). These
judgments were found, in turn, to influence sentencing decisions. Id.
91. Gold & Weiner, supra note 67, at 292. Goffman suggests that the act of apologizing
causes a "splitting of the self into a blameworthy part and a part that stands back and
sympathizes with the blame giving, and, by implication, is worthy of being brought back into
the fold." GOFFMAN, supra note 30, at 1 13.
92. See Gallagher et al., supra note 12; Hickson et al., supra note 12; Vincent et al. , supra
note 12; Witman et al., supra note 12.
93. Darby & Schlenker, Children's Reactions to Apologies, supra note 1 9. The authors
compared reactions to no apology, a perfunctory "excuse me," an expression of remorse,
and an apology with an offer of repair. Id. at 744.
94. Scher & Darley, supra note 19, at 132.
95. Id. at 133. The effectiveness of the apology was measured by adults' judgments
about the appropriateness of the wrongdoer's response, how bad he felt, the degree to which
he was to blame and would be condemned, and how reliable and conscientious he was. Id.
380
apology." 96 This is consistent with the intuition that an apology that

affirmatively accepts responsibility is probably most effective, but any
apology, even one that merely expresses sympathy, may be better than
nothing.
C. Apologies and Legal Settlement

Thus, in addition to the anecdotal evidence of the importance of
apologies in legal-settlement negotiation, there is a growing body of
empirical evidence that apologies can have important effects on a
variety of judgments that may underlie legal-settlement decisions.
There has been very little systematic investigation, however, of the
particular effects of apologies on legal-settlement decisionmaking.
Moreover, apologies offered in the course of a legal dispute are likely
to involve a different dynamic than apologies offered in other
contexts, and the constraints and influences of the legal system may
influence how apologies are understood.
Russell Korobkin and Chris Guthrie have conducted the only
experimental study to examine the influence of apology on litigants'
settlement decisions. 97 They asked undergraduates to assume the role
of the tenant in a landlord-tenant dispute over a broken furnace and
to indicate their willingness to accept a settlement offer from the
landlord. Some participants were told that "the landlord apologized to
you for his behavior. 'I know this is not an acceptable excuse,' he told
you, 'but I have been under a great deal of pressure lately.' "
Participants who were told that they had received this apology from
the landlord were marginally more likely to accept the offer than were
participants who had not received the apology. 98
The Korobkin and Guthrie study provides an indication that
apologies may have a role to play in settling legal disputes. Numerous
questions, however, are left unanswered. First, as the authors noted,
participants may not have considered the "apology" described in the
study materials to be an actual apology, as it did not include the
prototypical phrase "I'm sorry" and it included an excuse for the
landlord's behavior. 99 Second, the effects of different types of
apologies and the effects of different evidentiary rules remain
96. Id. at 137; see also Donald E. Conlon & Noel M. Murray, Customer Perceptions of
Corporate Responses to Product Complaints: The Role of Explanations, 39 ACAD. MGMT. J.
1040, 1042 ( 1996) (discussing apologies in the consumer-complaint context).
97. Russell Korobkin & Chris Guthrie, Psychological Barriers to Litigation Settlement:
An Experimental Approach, 93 MICH. L. REV. 107 ( 1994) [hereinafter Korobkin & Guthrie,
Psychological Barriers to Litigation Settlement].
98. Id. at 148 (describing a marginally statistically significant result at the p < .I level).
99. Id. at 149 n.176. See also supra note 30 and accompanying text (distinguishing
excuses and apologies).
381
unexplored. Finally, in addition to exploring participants' willingness

to accept a particular offer of settlement, it would be useful to inquire
into a broader range of the parties' underlying attributions and
perceptions in order to more fully understand the role of apologies in
legal settlement.
Cognitive approaches to understanding negotiation posit that a
central aspect of negotiation is how the parties to the negotiation
perceive, interpret, and understand the context of the negotiation, the
other party, and themselves. 100 Importantly for present purposes,
negotiator perceptions are not necessarily static - participation in the
process and interaction between the parties to the negotiation can
shape perceptions. 101 As Felstiner, Abel, and Sarat have noted,
"disputes are not things: they are social constructs." 102 Thus, an
apology might operate to influence settlement decisionmaking by
altering the injured party's perceptions of the situation, the other
party, or the offer itself. 103
It is important to consider these reactions for two distinct reasons.
First, because it is thought that apologies may lead to dispute
resolution indirectly through the apologies' effects on the parties'
construal of the situation, the mechanisms by which an apology might
lead to greater willingness to settle ought to be explored. Thus,
100. See Bazerman et al., supra note 70, at 281; Max H. Bazerman & John S. Carroll,
Negotiator Cognition, 9 RES. ORGANIZATIONAL BEHAV. 247, 268 (1987); John S. Carroll &
John W. Payne, An Information Processing Approach to Two-Party Negotiations, 3 RES.
NEGOTIATION ORG. 3, 8 ( 1 991); Margaret A. Neale & Gregory B. Northcraft, Behavioral
Negotiation Theory: A Framework for Conceptualizing Dyadic Bargaining, 13 RES.
ORGANIZATIONAL BEHAV. 147, 165 (1991); Thompson, supra note 65, at 527; Leigh
Thompson & Reid Hastie, Social Perception in Negotiation, 47 ORGANIZATIONAL BEHAV.
& HUM. DECISION PROCESSES 98 (1990).
101. Carroll & Payne, supra note 100, at 8 ("During this process, preferences and
aspirations may change, facts may be revealed or reinterpreted, and new issues arise.");
Thompson, supra note 65, at 528.
102. Felstiner et al., supra note 4, at 631 .
103. For example, a n apology that changes a party's construal o f the situation o r the
other party may also change the party's reactions, posturing, or communication. See Michael
W. Morris et al., Misperceiving Negotiation Counterparts: When Situationally Determined
Bargaining Behaviors Are Attributed to Personality Traits, 77 J. PERSONALITY & Soc.
PSYCHOL. 52, 52 (1999) ("Conflict theorists have suggested that a negotiator's attributions
for a counterpart's behavior exert a pivotal influence on the negotiator's strategic decisions."
(citation omitted)); see also Donald G. Gifford, The Synthesis of Legal Counseling and
Negotiation Models: Preserving Client-Centered Advocacy in the Negotiation Context, 34
UCLA L. REV. 811, 830 (1987) (recognizing legal negotiation as a "cyclical process" during
which perceptions and goals may evolve). Neale and Northcraft write:
There is a more global sense in which negotiator's information processing strategies
influence the processes and outcomes of negotiation. Negotiator cognition is the lens
through which negotiation occurs. . . . Negotiators do more than just process (perhaps
incorrectly) information about the context in which negotiations occur. Negotiators also
perceive that context, and react to their perceptions in ways that validate or enact those
perceptions. Thus, it is negotiators' cognitions which contextualize negotiations.
Neale & Northcraft, supra note 100, at 169.

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
382
participants' perceptions and attributions, affective reactions, and

willingness to forgive are of interest as variables which may mediate
any effects of apologies on decisions to accept or reject a particular
settlement offer. Second, it is likely that apologies may not always ( or
even frequently) result in the settlement of a dispute through their
effect on the acceptance of a particular settlement offer. Rather, an
apology may favorably change the dynamics of the negotiation to
smooth the way toward an eventual settlement - a settlement that
may come more quickly or be more satisfactory as a result of the
apology - even if the apology does not convince the victim to accept
the immediate offer of settlement. 104
Ill. THE PRESENT STUDIES
While many have become convinced of the value of an apology in

the legal context, it can be seen from the above review that there is a
paucity of empirical studies exammmg the legal effects of
apologizing. 105 This Part reports several original experimental studies
that were designed to begin systematic study of these effects.
In a typical experiment, a large number of participants evaluate the
same simulated case. All aspects of the case ( e.g., the factual situation
and the settlement offer) are held constant, except the manipulated
variable that is of interest in the study ( e.g., the nature of the apology
offered)_1ll6 Then, all participants are asked the same set of questions
(e.g., a question about how likely they would be to accept the offer).
Experimental simulation methods afford several benefits for
studying the role of apologies in legal-settlement negotiation. First,
there is often no record of the millions of settlement negotiations that
occur in civil cases each year, some of which may involve apologies. 107
Using experimental simulations is a useful way to begin this systematic
study of the effects of apologies on settlement decisionmaking.
104. See Bazerman & Carroll, supra note 1 00, at 249 (describing "sequences of
interaction over time during negotiation"); Leonard L. Riskin, Understanding Mediators '
Orientations, Strategies, and Techniques: A Grid for the Perplexed, l HARV. NEGOT. L. REV.
7, 20 ( 1 996) (arguing that "participants often must address the relational and emotional
aspects of their interactions in order to pave the way for settlement of the narrower
economic issues").
105. On the dearth of empirical studies examining legal-settlement negotiation generally
or the legal effects of apologies in particular, see generally Cohen, Legislating Apology,
supra note 43; Korobkin & Guthrie, Psychological Barriers to Litigation Settlement, supra
note 97; Shuman, The Role of Apology in Tort Law, supra note 3; and John 0. Haley,
Comment: The Implications of Apology, 20 LAW & Soc'Y REV. 499, 506 (1 986).
106. Thus, for example, participants might all be exposed to identical descriptions of an
injury-causing incident and would receive the same offer of settlement. Different groups of
participants, however, would be provided with different information about their
communication with the offender.
107. See Waldmeir, supra note 18, at 8.
383
Second, experimental simulation allows the researcher to obtain the

reactions of a large number of people to the same factual scenario.
Third, because all factors except the variable of interest are controlled,
any observed differences in the responses of participants who
evaluated cases involving different levels of the manipulated variable
can be attributed to that variable, unconfounded by other
influences. 108
In each study, participants were asked to visit a website and read a
scenario describing an incident in which one party was injured. 1 09
Respondents were asked to take on the role of the injured person and
to evaluate a settlement offer (covering out-of-pocket costs) from the
other party. In addition to their willingness to accept the settlement
offer, respondents were asked to indicate their assessment of the
sufficiency of the apology; the motives they ascribed to the other
party; their assessment of the nature of the injuries, the other party's
conduct and character, and each party's responsibility for causing the
incident; their degree of anger at, sympathy for, and willingness to
forgive the other party; and their assessment of how the incident
would likely affect the relationship and the other party's conduct in
the future. Ho
108. Ideally, the results reported here will converge with results from research
examining reactions of litigants in actual cases if and when such research is conducted.
Research using actual litigants in actual cases has the benefit of realism, but the researcher is
unable to control the myriad factors that make cases differ from one another, making the
isolation of the variable of interest impracticable. See Coates & Penrod, supra note 77, at 667
("[T]hese studies [of actual cases] measure rather than manipulate variables, with the
consequence that clear causal inferences are impossible to make. "). For a discussion of
simulation research, see generally Brian H. Bornstein, The Ecological Validity of Jury
Simulations: ls the Jury Still Out?, 23 LAW & HUM. BEHAV. 75 (1999); Robert M. Bray &
Norbert L. Kerr, Use of the Simulation Method in the Study of Jury Behavior: Some
Methodological Considerations, 3 LAW & HUM. BEHAV. 107 (1979); Shari Seidman
Diamond, Illuminations and Shadows from Jury Simulations, 21 LAW & HUM. BEHAV. 561
(1997); Robert MacCoun, Inside the Black Box: What Empirical Research Tells Us About
Decisionmaking by Civil Juries, in VERDICT: ASSESSING THE CIVIL JURY SYSTEM 137
(Robert E. Litan ed., 1993); Jennifer K. Robbennolt, Evaluating Empirical Research
Methods: Using Empirical Research in Law and Policy, 81 NEB. L. REV. 777 (2002); Wayne
Weiten & Shari Seidman Diamond, A Critical Review of the Jury Simulation Paradigm: The
Case of Defendant Characteristics, 3 LAW & HUM. BEHAV. 71, 75-83 (1979). For examples of
simulation research in legal settlement negotiation, see, for example, Mark Kelman et al.,
Context-Dependence in Legal Decision Making, 25 J. LEGAL STUD. 287 (1996); Korobkin &
Guthrie, Psychological Barriers to Litigation Settlement, supra note 97; Thomas D. Rowe, Jr.
& Neil Vidmar, Empirical Research on Offers of Settlement: A Preliminary Report, L. &
CONTEMP. PROBS., Autumn 1988, at 13.
109. In each of the studies described here, university staff members were randomly
selected from a university directory and were recruited by direct e-mail solicitation to
participate in the study. Thus, the resulting samples, which did not include students or
faculty, included a wide range of participants who could potentially find themselves in the
position of a person injured in an accident - participants included, fo r example, secretarial
and administrative staff, maintenance and grounds staff, athletic staff, food service staff,
library staff, and staff employed by the university-affiliated radio and television stations.
110. After they submitted their responses, participants were asked to provide
identifying information so that they could receive payment; these responses were sent to a

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
384
A. Effects of Apologies on Settlement Decisionmaking

The first study examined recipients' interpretations of apologies
and the effects of those apologies on willingness to accept a settlement
offer in the dispute. The hypothetical scenario detailed a relatively
simple personal-injury dispute: a pedestrian-bicycle accident. 1 1 1 All
participants reviewed this scenario and evaluated the same settlement
offer.
Control participants evaluated a version of the scenario in which
no apology was offered. Additional participants evaluated versions of
the scenario in which two variables, the nature of the apology offered
and the nature of the applicable evidentiary rule, were varied. First,
the nature of the apology offered was varied to compare the effects of
a partial apology, in which the other party merely expressed sympathy
for the potential claimant's injuries, with the effects of a full apology,
in which he or she also took responsibility for causing the injuries. 112
Second, the evidentiary rule described to participants was varied to
examine how different evidentiary rules influence the interpretation
and effectiveness of an apology, comparing respondents' reactions to
an apology where the evidentiary rules protected the apology, where
the evidentiary rules did not protect the apology, and where no
evidentiary rule was described. 1 1 3 Thus, there were seven different
variations of the basic scenario. 1 1 4 One hundred forty-five people
separate data file and were not linked with participants' substantive responses. For the few
individuals who wanted to participate in the study but were limited in their web access, hard
copies of the study materials were sent and returned via campus mail.
1 1 1. The hypothetical included evidence that the offender, described as someone from
the injured party's neighborhood with whom the injured party was acquainted, had been
riding quite fast, had been reaching fo r a water bottle, and had almost hit another pedestrian
shortly before the incident. The resulting injuries included a broken arm that required
surgery and a pin to set the fracture, bruised ribs, and numerous bruises and scratches.
1 12. The offender who offered a partial apology stated, in relevant part: "I am so sorry
that you were hurt. I really hope that you feel better soon." The offender who offered a full
apology stated: "I am so sorry that you were hurt. The accident was all my fault. I was going
too fast and not watching where I was going until it was too late." In both cases, the apology
was offered shortly after the accident, prior to the filing of a lawsuit.
113. This variable was manipulated by describing an over-the-fence conversation with
another neighbor who was an attorney. In the condition in which the apology was protected,
the neighbor/attorney described "a new rule preventing apologies from being used against
defendants in court," that the offender's statement "will not be admissible in court to prove
that [the offender] was at fault for the accident," and that the attorney "had a conversation
with [the offender] immediately after the accident and had mentioned to [the offender] that
if she apologized it couldn't be used against her if there was a lawsuit. " In the condition in
which the apology was explicitly not protected, the neighbor/attorney explained that the
offender's statement "could be introduced in court to prove [the offender] was at fault" and
that the attorney "had a conversation with [the offender] immediately after the accident and
had mentioned to [the offender] that if she apologized it could be used against her if there
was a lawsuit."
1 14. Thus, the study used a 2 (apology) x 3 (evidentiary rule) factorial design with an
additional no-apology control condition.

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
385
participated in the study and were randomly assigned to the

experimental conditions. 1 1 5
l. Effects of Apology on Settlement

The first set of analyses compared the two different types of
apology (partial and full) to the no-apology control where no
evidentiary rule was specified. It was predicted that the nature of the
apology would influence participants' settlement decisions. Moreover,
based on the previous research on the psychological effects of
apologies, it was predicted that an apology would influence settlement
decisionmaking through its effects on participants' perceptions and
attributions. That is, participants' interpretations of the apology, their
affective reactions, and their evaluations of the situation and the
offender were expected to mediate the effects of an apology on their
settlement decisions: the nature of the apology would influence these
perceptions and attributions, which would in turn influence settlement
decisions. 1 16
a. Effect of Apology on Settlement Decisions. First, even though all
participants were told that they had suffered the same injuries and
received the same offer of settlement, the nature of the apology
offered influenced recipients' willingness to accept the offer. 117 These
results are presented in Figure 1 . When no apology was offered 52 %
1 15. Participants ranged in age from twenty-one to seventy (mean = 40) ; thirty percent
of the participants were male.
1 16. As Baron and Kenny describe, "In general, a given variable may be said to function
as a mediator to the extent that it accounts for the relation between the predictor and the
criterion. [M]ediators explain how external physical events take on internal psychological
significance. . . . Mediators speak to how or why such effects occur." Reuben M. B aron &
David A. Kenny, The Moderator-Mediator Variable Distinction in Social Psychological
Research: Conceptual, Strategic, and Statistical Considerations, 51 J. PERSONALITY & Soc.
PSYCHOL. 1 173, 1176 (1986). In order to establish a mediating relationship, four
requirements must be established. First, the initial variable (e.g., apology) must be
correlated with the outcome variable (e.g. , settlement decisions) . Second, the initial variable
must be correlated with the proposed mediator (e.g., perceptions and attributions). Third,
the proposed mediator must be correlated with the outcome variable. And, finally, the direct
effect of the initial variable on the outcome variable must be diminished upon controlling for
the proposed mediator. Id. at 1 176-77.
1 1 7. j (4) = 13.388, p = . OJ. Regression analysis measures the relationships between a
set of predictor variables and a dependent variable. Because the settlement rate variable
consisted of three categories (settle, not settle, and unsure), this variable was analyzed using
multinomial logistic regression. See VAN! K. BOROOAH, LOGIT AND PROBIT: ORDERED
AND MULTINOMIAL MODELS (2002); TIM FuTING LIAO, INTERPRETING PROBABILITY
MODELS: LOGIT, PROBIT, AND OTHER GENERALIZED LINEAR MODELS ( 1 994); SCOTT
MENARD, APPLIED LOGISTIC REGRESSION ANALYSIS ( 1 995). Throughout this Article
differences are considered to be "statistically significant" if the statistical test used indicates
that the likelihood that the difference would occur by chance is less than 5% (reported by
the p-value as p < . 05). Differences are reported as "marginally significant" if the likelihood
of such a difference occurring by chance is greater than 5% but less than 10%. See generally
BARBARA G. TABACHNICK & LINDA S. FIDELL, USING MULTIVARIATE STATISTICS (2d ed.
1 989).

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
386
of respondents indicated that they would definitely or probably accept

the offer, while 43% would definitely or probably reject the offer and
5 % were unsure. When a partial apology was offered, only 35 % of
respondents were inclined to accept the offer, 25 % were inclined to
reject it, and 40% indicated that they were unsure. In contrast, when a
full apology was offered, 73% of respondents were inclined to accept
the offer, with only 13-14 % each inclined to reject it or remaining
unsure.
F i g u r e 1 . R e s p o n s e s t o S e t t l e m e n t O ffe r
1 00 %
80%
60% ■ A ccept
40% Iii U n c e rt a i n
20% □ R ej e c t
0%
N o A p o lo g y P a rtia l F u ll A p o lo g y
A p o lo g y
Comparing each type of apology to the condition in which no

apology was received, receiving a partial apology increased the
likelihood that the respondent would be unsure about how to respond
to the settlement offer, 1 18 and receiving a full apology increased the
likelihood that the respondent would choose to accept the offer1 19 and
decreased the likelihood that the respondent would choose to reject
the offer. 1 20
b. The Role of Perceptions and Attributions. Next, the study
examined the effects of the nature of the apology on a number of
constructs thought to underlie the effect of apology on settlement
decisionmaking. Where there were differences in participants'
responses across conditions, the differences follow a strikingly similar
pattern: offering no apology or a partial apology elicited equivalent
responses that were both different from the responses elicited when a
1 18. As compared to no apology, a partial apology increased the likelihood that the
respondent would be unsure rather than either accept (p = .030) or reject (p = . 026) the
offer.
1 19. As compared to no apology, a full apology increased the likelihood that the
respondent would accept the offer rather than reject it (p = .057) or be unsure (p = .553).
120. As compared to no apology, a full apology marginally decreased the likelihood that
the respondent would reject the offer rather than accept it (p = . 057) or be unsure (p = .099) .

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
387
full apology was offered. 1 2 1 The average ratings for variables on which
an apology had a significant effect are displayed in Table 1.
TABLE l
PARTICIPANTS' PERCEPTIONS AND ATTRIBUTIONS (STUDY 1)
Partial Full
No Apology
Apology Apology
Sufficient apology 1.90• 2.30• 3.82 b
Regret 2.86. 2.63. 4. 14b
Moral character 2.65. 2.10· 3.86b
Careful in future 3_50• 3_30• 4. 18b
Belief that responsible 2.86. 3.oo· 4.68b
Bad Conduct 4_05• 3.85. 3.lQb
Sympathy 1.57' 1.65. 2.60 b
Anger 3.71· 3_74• 2.86b
Forgiveness 3.62· 3.85. 4.23•
Damage to Relationship 3.29. 3_50• 2.00b
Offer make up for injury 2.52• 2_35• 3.55 b
Note: All constructs were measured on 5 point scales; higher numbers represent "more" of
the construct. For each rating, means with different superscripts differ significantly (p < . 05) .
Thus, a full apology was viewed as more sufficient than either a

partial apology or no apology. 1 22 An offender who offered a full
apology was seen as experiencing more regret, 123 as more moral, 1 24 and
as more likely to be careful in the future1 25 than one offering a partial
or no apology. While an offender offering a full apology was seen as
believing that he or she was more responsible for the incident than
121. A multivariate analysis of variance ("MANOV A") conducted on participants'

assessments of the case and the other party revealed a significant multivariate effect of
apology ( F (28, 72) = 1.642, p = . 048) . MANOVA measures the likelihood that observed
differences between groups on a combination of measures are due to actual differences on
those measures. See generally TABACHNICK & FIDELL, supra note 117, at 371-72. Follow-up
analyses of variance ("ANOVA") were conducted for each variable separately; it is these
univariate effects that are presented in Table 1. ANOV A measures the likelihood that
observed differences among a set of means are due to actual differences among the groups
as opposed to chance. See TABACHNICK & FIDELL, supra note 117, at 37. Except where
otherwise noted, all post-hoc ps < . 05.
122. F(2,59) = 15. 790, p < . 001 .
123. F(2,59) = 9.578, p < . 001.
124. F(2,59) = 9.065, p < . 001 .
125. F(2,59) = 6.959, p = . 002.

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
388
one who offered a partial or no apology, 1 26 the conduct of the full

apologizer was judged more favorably than that of offenders who
offered either a partial or no apology. 1 27 Participants expressed greater
sympathy and less anger at the offender who offered a full apology
than they did at offenders who offered either a partial or no
apology. 128 Participants also indicated more willingness to forgive an
offender who gave a full apology than they did for offenders offering a
partial or no apology129 and expected that less damage to the parties'
relationship would result following a full apology than they did
following a partial or no apology.13° Finally, participants indicated that
the settlement offer would better make up for their injuries when they
had received a full apology than when they had received either a
partial or no apology . 131
To facilitate further analysis, two scales were constructed from
these measures1 32 that represented participants' overall assessments of
the apologies1 33 and their evaluations of the offer.1 34 Consistent with
126. F(2,59) = 15.578, p < . 001.

127. F(2,58) = 4.645, p = . 013.
128. F(2,58) = 7.317, p = .001 and F(2,59) = 3.858, p = .027, respectively.
129. F(2,60) = 4. 150, p = .021. Post-hoc tests for this variable showed only a marginally
significant difference (p = .086) between offenders who offered full and partial apologies.
130. F(2,58) = 9.840, p < .001.
131. F(2,60) = 5.865, p = .005.
132. The groups of variables to be included in each scale were determined by a principal
components factor analysis. Factor analysis is a procedure that is used to determine whether
a larger number of variables cluster together into a smaller number of constructs. The
procedure is "applied to a single set of variables where the researcher is interested in
discovering which variables in the set form coherent subsets that are relatively independent
of one another. Variables that are correlated with one another but largely independent of
other subsets of variables are combined into factors." TABACHNICK & FIDELL, supra note
117, at 597. The factor analysis was conducted on fourteen questions (each measured on a
five-point Likert scale); varimax rotation· was utilized to enhance the interpretability of the
factors. The analysis demonstrated that the items comprised three orthogonal factors
(labeled Assessment of Apology, Evaluation of Offer, and Attribution of Responsibility)
accounting for a total of sixty-one percent of the variance. Scales based on the first two
factors (Assessment of Apology and Evaluation of Offer) are described more fully infra
notes 133-134, and were used in subsequent analyses. Because a scale constructed with
variables based on the third factor (ratings of the offender's responsibility, and the
participant's own responsibility) (alpha = .59) was not influenced by whether an apology was
offered (F(2,59) = .052, p = .950), nor was it related to settlement decisions (x1 (2) = .870, p =
.647), this factor was not included in subsequent analyses.
133. The first scale, labeled "Assessment of Apology," included participants' ratings of
the sufficiency of the apology, the degree to which the offender thought he or she was
responsible, the offender's regret, the degree to which the offender would be careful in the
future, sympathy for the offender, and the offender's morality (alpha = .87).
134. The second scale, labeled "Evaluation of Offer," appeared to comprise an
evaluation of the offer relative to the offense and included participants' ratings of the degree
to which the offer would "make up for" their injuries (reverse coded), their willingness to
forgive the offender (reverse coded), the severity of their injuries, the damage done to the
relationship, their anger, and the conduct of the offender (alpha = .81).

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
389
December 2003) Apologies and Legal Settlement 489
the results reported above for the individual underlying questions, 1 35

the nature of the apology received influenced scores on both scales.
Specifically, participants who received a full apology made more
positive assessments of the apology and the information conveyed by
the apology than did participants who were told they had received
only a partial apology or no apology. 1 36 Moreover, participants who
had received a full apology made more positive evaluations of the
offer than did participants who were told that they had received only a
partial apology or no apology. 1 37
In addition, each of these scales influenced settlement
decisionmaking. Participants who made more favorable assessments of
the apology and the information that it communicated were more
likely to agree to settle the case than they were to be unsure or to
reject the offer. 1 38 Similarly, participants who made more favorable
evaluations of the offer were more likely to agree to settle the case
than they were to be unsure or to reject the offer. 139
FIGURE 1: THE EFFECTS OF APOLOGIES ON SETTLEMENT DECISIONS
--- -------------
Evaluation of
Nature of Assessment Settlement
Offer, Damage,
Apology of Apology Decision
Forgiveness
Note: Statistically significant effects are indicated by solid lines; marginally significant
effects are indicated by dashed lines.
Figure 1 depicts the relationships among the nature of the apology,

participants' attributions and perceptions, and settlement decisions. 140
135. See supra notes 122-131 and accompanying text.

136. No Apology mean = 2.56; Partial Apology mean = 2.59; Full Apology mean = 3.90.
F(2,60) = 19.531 , p < .001 .
137. No Apology mean = 2. 60; Partial Apology mean = 2.46; Full Apology mean = 3.37.
F(2,60) = 9.050, p < .001 .
138. j(2) = 8. 747, p = .013. More favorable assessments increased the likelihood that
the respondent would accept the offer rather than reject it (p = .016) or be unsure (p = .034).
139. j(2) = 20.150, p < .001 . More favorable evaluations increased the likelihood that
the respondent would accept the offer rather than reject it (p = .001 ) or be unsure (p = .004).
140. Recall that in order to show that an apology operates to influence settlement
decisionmaking through the ways in which it influences the injured party's perceptions and
attributions about the situation, it must be shown not only that the apology affects both
settlement decisions and individual perceptions and attributions, and that the perceptions
and attributions directly influence settlement decisionmaking, but also that the effect of the
apology on decisionmaking is lessened when perceptions and attributions are controlled. See
supra note 116. Given the content of the two scales, it was hypothesized that these two sets

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
390
490 Michigan Law Review (Vol. 102:460
First, as noted above, the nature of the apology was found to influence
assessments of the apology and of the information an apology is
thought to convey. 1 4 1 Second, there was a significant relationship
between the nature of the apology and participants' evaluations of the
offer that was mediated by the participants' assessments of the
apology. 1 42 Third, participants' assessments of the apology and their
evaluations of the offer together partially mediated the relationship
between the nature of the apology and participants' settlement
decisions; apology, however, retained a marginally significant direct
effect on settlement decisions. 1 43
2. Effects of Evidentiary Rules on Settlement

The second set of analyses explored the effects of the different
evidentiary rules on settlement decisionmaking. 144 Differences in
evidentiary rules did not produce significant differences in settlement
rates145 nor did they produce differences in participants' perceptions
and attributions. 1 46 Importantly, there were no effects of the
of perceptions would work in tandem to mediate the effect of an apology on settlement

decisionmaking. Thus, it was expected that the nature of the apology would influence
assessments of the apology and perceptions of the information conveyed by the apology. It
was expected that these assessments, in turn, would influence evaluations of the offer, and
that this evaluation would influence decisionmaking. A series of additional regression
analyses were conducted to test these propositions. See infra notes 141-143 and
accompanying text.
142. As demonstrated above, the nature of the apology was correlated with both
assessments of the apology and evaluations of the offer. See supra notes 1 36-137 and
accompanying text. In addition, assessments of the apology and evaluations of the offer were
significantly related (F(1,61) = 28. 789, p < . 001). Moreover, when evaluations of the offer
were regressed on both the nature of the apology and assessments of the apology (F(3,59) =
10.600, p < .001), only assessments of the apology had a statistically significant influence on
these evaluations (t = -3.280, p = .002). In this model, the nature of the apology did not
significantly influence evaluations of the offer (t = -1.055, p = .296 and t = -1. 639, p = .107).
143. The nature of the apology, participants' assessments of the apology, and
participants' evaluation of the offer were examined as predictors of participants' settlement
decisions. Analyses revealed that when settlement decisions were regressed on a set of
predictors that included the nature of the apology and both perception scales (x,2 (8) = 28. 967,
p < .001), participants' evaluation of the offer was the only statistically significant predictor
of settlement decisions (x,'(2) = 10.953, p = .004), the effect of the apology on settlement
decisionmaking became marginally statistically significant (x,2 (4) = 8.465, p = . 076), and the
effect of assessments of the apology was not statistically significant (x,'(2) = .547, p = . 761). In
other words, these mediators partially explain the effect of apology on settlement decisions.
144. These analyses were conducted on the 2 (partial or full apology) x 3 (evidentiary
rules) factorial.
145. j(6) = 3.404, p = . 757. Neither the evidentiary rule (x,'(4) = 1.551, p = .818) nor the
apology (x,1 (2) = 1.898, p = .387) had a statistically significant influence on settlement rates.
A model that included interaction effects was also not statistically significant (x,'(10) = 8. 786,
p = .553).
146. A 3 (evidentiary rule) x 2 (partial v. full apology) MANOVA conducted on
participants' assessments of the case and the other party revealed no significant differences

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
391
evidentiary rules on ratings of the sufficiency147 or sincerity 148 of the

apology given. Participants were, however, aware of the differences in
the rules as they assessed the scenario; analysis of participants' ratings
of the likely motives for the apology revealed that apologies that were
not protected by an evidentiary rule were seen as less likely to have
been motivated by desire to avoid a lawsuit. 149 Thus, participants were
aware of the content of the different evidentiary rules, but did not
adjust their assessments of the apologies received in response to those
rules.
3. Summary
Therefore, apologies influenced the inclination to accept or reject a
settlement offer. The effect of an apology on settlement decisions was
complex, however, and depended on the type of apology offered. Only
the full, responsibility-accepting apology increased the likelihood that
the offer would be accepted. The partial, sympathy-expressing,
apology, in contrast, increased participants' uncertainty about whether
or not to accept the offer.
Moreover, apologies were found to influence ratings of numerous
variables that are thought to underlie the settlement decision. The
effects of apologies on these underlying constructs, however, were
limited to the full apology in which the offender accepted
responsibility. Offering a partial apology was no different from
offering no apology at all. These underlying judgments are likely to
favorably change the dynamics of the negotiation and also provided
the mechanism by which apologies influenced settlement decisions.
Importantly for the debate over evidentiary protection for
apologies, while participants were aware of the different evidentiary
rules governing the admissibility of the apology, the nature of the
applicable rule did not influence the apologies' effect on settlement
decisions, nor did these rules influence participants' perceptions of the
situation or the offender.
among the evidentiary rules (F(28,184) = 1.057, p = .395) and no significant interaction
(F(28,184) = 0.906, p = . 605). This analysis did reveal a statistically significant multivariate
effect of the apology (F(14,92) = 6. 778, p < .001). Since the follow-up analyses revealed
differences between the partial and full apologies that were similar to those demonstrated by
the planned comparisons of these conditions with no apology control, see supra notes 122-
131 and accompanying text, details will not be presented here.
147. F(2, 118) = .646, p = .526.
148. F(2, 116) = .870, p = .422.
149. F(2, 116) = 5. 987, p = .003. This test was a follow-up to a 3 (evidentiary rule) x 2
(partial v. full apology) MANOVA conducted on participants' judgments about the
offender's motives for apologizing, which revealed statistically significant effects of both the
type of apology (F(5,110) = 4. 130, p = . 002) and the type of evidentiary rule (F(J0,220) =
2. 039, p = . 031). No significant interaction was detected (F(J0,220) = 0.899, p = .534).
392
IV. IMPLICATIONS OF RESULTS
Legal actors in the United States have traditionally assumed that

apologies would be viewed by judges and juries as admissions of
responsibility that will lead to increased legal liability. This perception
is thought to be largely responsible for a perceived reluctance on the
part of potential defendants in civil lawsuits to apologize to those that
they have, or may have, injured. 194
Recently, scholars have debated the utility of safe apologies in civil
litigation. Despite concern that apologies will increase the risk of
liability, many are now promoting apologies as beneficial in settling
lawsuits. There is increasing interest in the possibility that an apology
can serve to resolve a dispute so as to avoid or bring an end to
litigation or, where it cannot do so directly, the possibility that an
apology may set the stage for the parties to engage in more productive
settlement discussions. In particular, safe apologies have been
promoted in an attempt to balance aspirations for the beneficial
effects of apologies on settlement with the potentially increased risk of
adverse liability findings. 195 Legal reform efforts have begun to follow
this path- moving to provide legal protection for apologies. 196 At the
same time, however, other commentators have criticized safe
apologies as diminished in meaning, moral value, and effectiveness. 197
The data reported here comprise an initial empirical examination
of the effects of apologies on settlement decisionmaking. Importantly,
these studies provide evidence that offering an apology can facilitate
conflict resolution. These studies also make clear, however, that the
effects of apologies on settlement decisionmaking are influenced by a
variety of factors. These complex effects present policymakers and
litigants or potential litigants with some difficult decisions about the
appropriate evidentiary protection for apologies, whether to offer an
apology to an opposing party in civil litigation, and how to respond to
an apology so offered.
194. See supra Part I.A.

195. See supra Parts I.B-C.
196. See supra Part J.C.
197. See supra Part I.D.
393
A. Evidentiary Protection for Apologies

One mechanism by which several states have attempted to make
apologies safe is to provide them with statutory protection that
prevents the admissibility of some apologies as evidence of liability.
The results of the present study inform decisions to be made by
policymakers considering safe-harbor statutes to protect apologetic
expressions. The data reported here provide no support for the
intuition that evidentiary protection for apologies will affect the ways
in which apologies are perceived. Instead, these results are consistent
with the contrary intuition that even statutorily protected apologies
have a role to play in defusing conflict. Across two studies, the
different evidentiary rules did not produce significant differences in
settlement rates or in participants' perceptions of the negotiation
situation or the other party. 198
What might explain the apparent lack of effects of safe-harbor
statutes on assessments of apologies and on settlement behavior?
198. This was true even though it was made clear that both parties were aware of the
rules. In the real world, where parties to a dispute like this one are less likely to know about
the rules of evidence as they apply to apologies, this might be even more so.
199. RICHARD E. NISBETT & LEE Ross, HUMAN INFERENCE: STRATEGIES AND
SHORTCOMINGS OF SOCIAL JUDGMENT (1980); LEE Ross & RICHARD E. NISBETT, THE
PERSON AND THE SITUATION 4 (1991); Lee Ross, The Intuitive Psychologist and His
Shortcomings: Distortions in the Attribution Process, 10 ADVANCES IN EXPERIMENTAL Soc.
PSYCHOL. 173 (1977) ("People's inflated belief in the importance of personality traits and
dispositions, together with their failure to recognize the importance of situational factors in
affecting behavior, has been termed the 'fundamental attribution error.'").
394
Finally, participants may have discounted the influence of the

statutory protection given the specific facts of the situation with which
they were faced. In particular, in the studies described here, the
apology at issue was offered very soon after the accident (before a
lawsuit was filed) in the context of an interpersonal dispute between
neighbors, where the offender's responsibility was judged to be
relatively clear,209 and the offer was fairly reasonable.210 On these facts,
then, the evidentiary value of the apology may have been thought to
be minimal and, accordingly, recipients may not have attributed the
apology or the lack of an apology to the various legal rules. Thus, it is
possible that respondents would be more skeptical of protected
apologies offered in conjunction with less generous offers, apologies
offered outside the interpersonal context (e.g., by a corporation), or
apologies offered later in the process. Clearly, additional research is
needed to explore these boundaries.
There is, then, at present, no evidence to suggest that protected
apologies will be less effective or less valued by claimants than
unprotected apologies. Accordingly, providing evidentiary protection
for apologies may serve to encourage the offering of apologies,211 or at
least to signal that apologies are a desired response to an injury
producing event,212 without diminishing the value and effectiveness of
apologies so offered. 213
208.
209. See supra note 152. Thus, even in the conditions in which the offender's
responsibility was less clear, participants judged the offender to be highly responsible
overall, though less responsible than in the conditions in which the offender's responsibility
was more clear.
210. The offer in the present studies was designed to cover out-of-pocket costs.
211. Note that whether and the degree to which more apologies might be offered under
evidentiary protection is a separate empirical question. The data presented here address
only the responses of injured parties to apologies that are offered by offenders under various
legal protections. The data do rtot address whether the types of statutory protections that
have been enacted, or are being considered, will in fact encourage apologies.
212. Safe-harbor statutes might serve this expressive purpose even if there were no
instrumental consequences for the offering of apologies or for settlement. See generally, Cass
R. Sunstein, On the Expressive Function of Law, 144 U. PA. L. REV. 2021 (1996).
213. While the fact that participants did not devalue apologies that were protected by
statute bodes well for the ability of these statutes to promote the useful exchange of
apologies, it does raise the concern that plaintiffs may have difficulty distinguishing between
sincere and strategic apologies. See infra notes 231-234 and accompanying text.
395
To the extent that the goals of such provisions are to encourage

apologies in order to facilitate settlement, 214 however, the current
statutes may be protecting the wrong apologetic expressions. The
current and proposed statutes predominantly protect partial apologies
and those portions of full apologies that constitute expressions of
sympathy, not admissions of responsibility. 215 The results presented
here suggest that it is full, responsibility-accepting, apologies that have
a positive impact on settlement decisionmaking, rather than the partial
apologies that are typically protected by the statutes. Moreover, full
apologies, because they are admissions, are more likely to raise
defendants' concerns about adverse liability rulings and are more
likely deterred by potential admissibility. At the same time, however,
offering protection to full apologies may result in the exclusion of
probative evidence and may limit a plaintiff's ability to bring a
successful lawsuit. 216
Accordingly, these data suggest that policy discussion ought to
focus on the appropriateness of statutory protection for full apologies.
Such policy discussion must consider the present findings regarding
the beneficial effects of full apologies on settlement decisionmaking in
light of other relevant considerations such as how best to encourage
apologies, concerns about undue limits on the ability to bring lawsuits,
the probative value of full apologies, and so on. 217
B. Defendants
In the absence of statutory protection, defendants or potential
defendants who desire to apologize are faced with balancing the
effects of apologies on different types of judgments - settlement
decisionmaking, liability decisions, and decisions about appropriate
damage awards - as well as weighing the less strategic aspects of
apologizing. The results of the instant studies inform one important
component of these interrelated decisions- the impact of apology on
settlement decisions. An important lesson for defendants to draw from
these data is that apologies can have beneficial effects on settlement,
altering the injured parties' perceptions of the situation and the
214. See, e.g., CAL. ASSEMBLY COMM. ON JUDICIARY, COMMENT TO CAL. EVID. CODE
§ 1160 ("The author introduced this bill in an attempt to reduce lawsuits and encourage
settlements by fostering the use of apologies in connection with accident-related injuries or
death.").
215. See supra Part I.C.
216. See Cohen, Legislating Apology, supra note 43, at 856-63; see also supra note 53
(describing objections to the Colorado statute).
217. Evaluating the overall merits of evidentiary protection for full apologetic
expressions is beyond the scope of this Article. See Cohen, Legislating Apology, supra note
43, at 841-66 (describing the pros and cons of such protection); Orenstein, supra note 3
(proposing evidentiary protection for full apologies).
396
offender so as to make them more amenable to settlement discussions

and ultimately more likely to accept an offer.
As a general matter, the results of the present study provide
evidence that a full apology that both expresses sympathy for the
victim's injuries and accepts responsibility for those injuries influences
a variety of perceptions and attributions about the situation and the
other party that might lead to a settlement or allow the parties to
begin discussions. 21 8 Full apologies were seen as more sufficient
apologies, as evidencing more regret and a greater likelihood of care
in the future, and as offered by people of higher moral character. Full
apologies favorably altered assessments of the conduct leading to the
injuries and changed the emotions of the injured party so as to reduce
anger and increase sympathy for the offender. Full apologies were
seen as mitigating potential damage to the relationship, were more
likely to lead to forgiveness, and inclined injured parties to look more
favorably on the settlement offer. In addition, the results of the first
study demonstrated that full apologies, through these effects on
perceptions and attributions, increased the likelihood that the
settlement offer would be accepted. 21 9
Accordingly, full apologies that include accepting responsibility for
the incident may facilitate the settling of lawsuits. These
"responsibility-accepting" apologies, however, are precisely the type
of apologies that most clearly raise concerns about the effects of
apologizing on liability decisionmaking and are not likely to be
protected by evidentiary rules protecting apologies. Making a
statement that admits fault and that might be admissible at trial, while
improving the prospects for settling the case, is thought to increase the
risk that the offender will be found liable. 220
For this reason, there is growing interest in ways in which
offenders can apologize without exposing themselves to the same risks
attendant to a full, responsibility-accepting apology. The present
research suggests that the effects of such partial apologies are complex
and identifies several aspects of the case that defendants ought to take
into account when considering a partial apology. First, the effects of
partial apologies on settlement decisionmaking appear to be much
more complicated than the effects of full apologies. On the whole,
partial apologies did not appear to facilitate settlement in the ways
218. See supra Part III.Al.

219. See supra Part III.Al.
220. It is not clear what risk an apology entails for liability determinations. See supra
note 19. The present study did not address this issue directly, as the participants were not
asked to take on the role of third-party decision makers and were not asked to make liability
determinations. Participants, however, were asked to judge the responsibility of the alleged
offender. In the first study, apologies had no effect on judged responsibility for the incident.
In the second study, offenders offering a partial apology were judged to be more responsible
than those who offered no apology. See supra note 181 and accompanying text.
397
D ecember 2003] Apologies and Legal Settlement 507
hoped by proponents. The most consistent finding was that partial

apologies tended to be no better ( or worse) than not offering an
apology. Across both studies,
, there were no differences between those
receiving partial apologies and no apology in their evaluations of the
offender's conduct, the offender's regret, the offender's belief that he
or she was responsible, damage to the relationship, anger, the degree
to which the offer would make up for the injuries, or forgiveness. 221
221. See supra Parts III.A l , III.B.2.

222.
398
Overall, then, it appears that a partial apology, while perhaps

minimizing the risk that the apology will be considered an admission
and result in an adverse liability decision,226 may not be terribly
effective at improving the prospects for settlement. This may be
because partial apologies do not communicate the same messages to
recipients that full apologies do. In these studies, the partial apology,
unlike the full apology, did not consistently convey to the recipient
that the offender had accepted responsibility for his or her behavior,
that he or she regretted the behavior, and, accordingly, that the
offender would not repeat the conduct. Consistent with the
fundamental attribution error, 227 injured parties may attribute this
failure to completely apologize to the character of the person, rather
than to any constraints attendant to the legal system. Perhaps there is
also a sense that the offender who offers only a partial apology has not
sufficiently owned up to the consequences of his or her behavior. 228 It
is also possible that there are other sets of circumstances, not reflected
in the vignettes used here, in which partial apologies are either
particularly helpful or particularly detrimental. 229
Thus, a defendant or potential defendant who wants to fully
apologize may face the uncertain effects of the apology on liability and
damage-award decisions if settlement negotiations fail, but may
benefit from an improved settlement climate and an improved chance
of avoiding litigation altogether. Defendants contemplating a partial,
sympathy-expressing apology may not face the same liability risks.
The beneficial effects of such apologies are not clear, however, and
there is evidence of some risk that perceptions will be negatively
impacted. Thus, defendants must be carefully attuned to the types of
factors identified here, in particular the severity of the plaintiff's injury
and the degree to which the offender appears to be responsible, 230 in
evaluating these rules.
226. As noted above, this is still an open empirical question. See supra note 19.
228. Taft, supra note 7.
229. In particular, even under conditions in which the offender was less clearly
responsible, the offender was judged to be relatively responsible. See supra note 1 5 6 and
accompanying text. It is possible t hat when the level of the offender's responsibility is even
less clear, a partial apology would be more effective.
230. For example, a full apology might be considered (in particular) when the offender's
responsibility is relatively clear. Under such conditions, there is evidence that a partial
apology may impede settlement, t here is much to be gained from a full apology in terms of
the possibilities for settlement, and the ma rginal risk of increased liability du e to accepting
responsibility may be minimal given the independent evidence of fault.

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
399
D ecember 2003] Apologies and Legal Settlement 509
C. Plaintiffs
In contrast to defendants, plaintiffs or potential plaintiffs must
determine how to respond to an apology from someone they believe
has wronged them. Many observers have expressed concern that
plaintiffs may be induced by an apology to agree to a settlement that
does not provide them with the monetary recovery to which they are
entitled. 231 Importantly, this concern is sometimes dismissed by
assuming that plaintiffs are capable of evaluating an apology for its
sincerity and strategic motivation and can assess apologies and
settlement offers accordingly. 232
The results of the present studies suggest that plaintiffs may, in
fact, be able to critically evaluate the content of an apology and to
distinguish those that they find credible and that communicate the
necessary information from those that do not. Participants in the
studies evaluated partial apologies very differently from full apologies.
Participants did not interpret partial apologies as conveying the same
evidence of regret, acceptance of responsibility, or likelihood of
greater care in the future as full apologies. 233 On the other hand,
participants consistently did not distinguish among apologies that
potentially exposed the offender to an increased risk of legal liability
(no statutory protection) and those that did not (statutorily
protected). Across both studies, the different evidentiary rules had no
23 1. Cohen, Advising Clients to Apologize, supra note 3; Cohen, Legislating Apology,

supra note 43; Levi, supra note 3, at 1171 ("For instance, critics might ask, if a plaintiff settles
because she's emotionally fulfilled by an apology, i sn't she being duped out of her legal
entitlement - an entitlement that the apology itself makes concrete? "); O'Hara & Yarn,
supra note 10, at 1186 (" [A]pology can be used as a tool for organizations to strategically
take advantage of individual victims' instincts to forgive in the face of apology."). There is
also some empirical evidence that an apology " script" dictates that an apology by an
offender is to be followed by forgiveness by the recipient. See Mark Bennett & Christopher
D ewberry, "I've said I'm sorry, haven 't I? " A Study of the Identity Implications and
Constraints that Apologies Create for Their Recipients, 13 CURRENT PSYCHOL. 10 (1994).
Thi s concern is part of a larger debate concerning the appropriateness of private settlement
and i ssues of self-determination in settlement a nd mediation. See Fiss, supra note 4; Galanter
& Cahill, supra note 4; Gary LaFree & Christine Rack, The Effects of Participants ' Ethnicity
and Gender on Monetary Outcomes in Mediated and Adjudicated Civil Cases, 30 LAW &
Soc'Y REV. 767 ( 1996); M cThenia & Shaffer, supra note 4; Menk el-Meadow, For and
Against Settlement, supra note 4; Menk el-Meadow, Whose Dispute, supra note 4; Ellen
Waldman, Substituting Needs for Rights in Mediation: Therapeutic or Disabling?, 5
PSYCHOL. PUB. POL'Y & L. 1103 ( 1999); Nancy A. Welsh, The Thinning Vision of Self
Determination in Court- Connected Mediation: The Inevitable Price of Institutionalization?, 6
HARV. NEGOT. L. REV. 1 (2001).
232. See, e.g. , Cohen, Advising Clients to Apologize, supra note 3, at 1067 ("If, as the
recipi ent of an apology, you k now that th e apologizer has apologized i n such a way as to
ensure insulation from l egal liability, you may attach less worth to that apology than
otherwise, and a plaintiff' s lawy er sh ould be sure to point thi s out to her client."); see also
R eh m & B eatty, supra note 10 ( claiming that judges and jurors can distinguish apologies that
admit fault from those that do not).
233. See supra Parts III.A. 1, III.B.2.
400
influence on participants' perceptions and attributions nor on

settlement decisions. 234 To the extent that participants did not
distinguish these apologies, there is some risk that plaintiffs will be
convinced by insincere or purely strategic apologies to settle for less
than they might otherwise.
The results of the present studies, however, do not clearly indicate
whether or not an apology will induce a plaintiff to accept a smaller
financial settlement. 235 Specifically, these studies only assessed
participants' propensities to accept or reject a particular settlement
offer. In addition, while a full apology did result in an increase in the
settlement rate in the first study, the more consistent result was that
apologies influenced perceptions and attributions. Improvements in
perceptions may not directly induce plaintiffs to accept inadequate
offers that they might not otherwise be inclined to accept, but they
may make it possible to begin or continue settlement discussions.
Similar settlement terms may have ultimately been reached, but via a
more difficult and time-consuming process. While it might be
reasonable to infer that improved perceptions or a greater propensity
to accept a particular offer might result in acceptance of a smaller
amount than might otherwise have been agreed to, further empirical
study ought to explore the extent to which potential litigants are
willing to trade-off apologies and financial compensation. 236
With that caveat in mind, however, it is plausible that some
plaintiffs would be willing to accept a smaller financial settlement if
they receive an apology. If this is true, one explanation may be that
they have been unfairly induced to forego their rightful compensation.
An alternative explanation may be that they are more satisfied by a
combination of financial compensation and the apology than they
would have been with a larger monetary amount and no apology.
Plaintiffs may value the apology more than or differently from
financial compensation. Indeed, motivations other than legal
entitlement and monetary recovery may be important to plaintiffs. 237
234. See supra Parts III.A.2, 111.B.3.

235. Given that most cases settle, settlements reached when an apology has been
offered are most appropriately compared to what the plaintiffs would have gotten in
bilateral settlement absent the apology rather than to what the plaintiffs could have or would
have gotten at trial.
236. For a general discussion of incommensurability in law, see Margaret Jane Radin,
Compensation and Commensurability, 43 DUKE L.J. 56 (1993), and Cass R. Sunstein,
lncommensurability and Valuation in Law, 92 MICH. L. REV. 779 (1994).
237. Menkel-Meadow, Whose Dispute, supra note 4, at 2677 ("[P]eople and entities in
disputes may have a wide variety of interests (of which legal principles may be one class) and
may decide that, in any given case, social, psychological, economic, political, moral, or
religious principles should govern the resolution of their dispute. "); Jean R. Sternlight,
Lawyers' Representation of Clients in Mediation: Using Economics and Psychology to
Structure Advocacy in a Nonadversarial Setting, 14 OHIO ST. J. ON DISP. RESOL. 269, 302-06
(1999) (detailing a number of nonmonetary goals that parties may have - "may reflect a

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
401
December 2003] Apologies and Legal Settlement 51 1
While a monetary settlement may adequately restore financial losses

resulting from an injury, an apology may be a better mechanism for
restoring less tangible damage, 238 expressing the proper relative moral
positions of the parties,239 assuring the injured party that the offender
will not reoffend, 240 or achieving restorative justice. 241 As Cohen notes,
" [p]aying monetary damages may help take care of the financial
consequences of an injury, but it may take an apology to 'wipe the
moral ledger' clean and construct an understanding of the injury and
the relationship which both parties can accept . " 242 As described in the
next Part, appropriate legal counseling may help to ensure that these
preferences are respected while limiting the chances that an
inappropriate settlement is reached.
D. The Role of the Lawyer

Defendants, then, face difficult decisions about whether or not to
offer an apology and plaintiffs face difficult decisions about how to
respond to offers of settlement that include or are accompanied by an
rationality that is broader than the mere maximization of wealth"); Vincent et al., supra note
12 (identifying four primary reasons plaintiffs pursued litigation: "accountability - wish to
see staff disciplined and called to account; explanation - a combination of wanting an
explanation and feeling ignored or neglected after the incident; standards of care - wishing
to ensure that a similar incident did not happen again; and compensation - wanting
compensation and an admission of negligence"); see also Robbennolt et al., supra note 17
(arguing that legal decision makers simultaneously pursue numerous instrumental and
symbolic goals). The extent to which negotiation and settlement, more broadly, track legal
principles is unclear. See, e.g. , HERBERT KRITZER, LET'S MAKE A DEAL 30-56 (1991)
[hereinafter KRITZER, LET'S MAKE A DEAL]; Janet Cooper Alexander, Do the Merits
Matter? A Study of Settlements in Securities Class Actions, 43 STAN. L. REV. 497 (1991);
Robert J. Condlin, "Cases on Both Sides ": Patterns of Argument in Legal Dispute
Negotiation, 44 MD. L. REV. 65 (1985); John Lande, Getting the Faith: Why Business Lawyers
and Executives Believe in Mediation, 5 HARV. NEGOT. L. REV. 137 (2000); Robert H.
Mnookin & Lewis Kornhauser, Bargaining in the Shadow of the Law: The Case of Divorce,
88 YALE L.J. 950 (1979).
238. See Shuman, The Role of Apology in Tort Law, supra note 3, at 181 ("Practically,
tort damages for these intangible losses [pain and suffering, loss of consortium, indignity,
and grief] defy the formulation of an empirically grounded metric . . . "); see also Daniel W.
Shuman, The Psychology of Compensation in Tort Law, 43 U. KAN. L. REV. 39, 71 (1994).
Note also that apologies may mitigate damages in defamation cases. See supra note 17.
241. See generally JOHN BRAITHWAITE, RESTORATIVE JUSTICE AND RESPONSIVE
REGULATION (2001); G. Bazemore, Restorative Justice and Earned Redemption , 41 AM.
BEHAV. SCIENTIST 768 (1998); K. Daly & Russ lmmarigeon, The Past, Present, and Future of
Restorative Justice: Some Critical Reflections, 1 CONTEMP. JUST. REV. 21 (1998).
242. Cohen, Advising Clients to A pologize, supra note 3, at 1020 ; see also Lon L. Fuller,
Mediation - Its Forms and Functions, 44 S. CAL. L. REV. 305, 308 (1971) (describing
mediation as having the capacity to reorient the parties toward each other, not by imposing
rules on them, but by helping them to achieve a new and shared perception of their
relationship, a perception that will redirect their attitudes and dispositions toward one
another).

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
402
apology. 243 For those parties who are represented by counsel, their
attorneys may help them make these decisions. Indeed, there is
empirical evidence that lawyers can have a notable influence on their
clients' settlement decisions. 244
Through a process of counseling and deliberation, attorneys and
clients together should attempt to evaluate the strategic and
nonstrategic considerations surrounding apologies. David Binder, Paul
Bergman, and Susan Price suggest that the attorney should engage in a
process of counseling that assists the client in identifying the client's
objectives, the possible options, and the likely consequences of those
options, and in weighing the options given their probable
consequences. 245 Importantly, the attorney ought to facilitate the
client's consideration of both the legal and nonlegal (e.g., economic,
psychological, social, moral, political, and religious) consequences of
the decision246 and recognize that objectives may change during the
course of the representation.247
Attorneys, as a consequence, must pay attention to what Len
Riskin describes as the "problem definition" of the matter in
dispute. 248 Attorneys are trained to focus on settling a dispute in the
243. Ethical rules reserve these decisions for the clients. Model Rules of Professional
Conduct Rule 1.2(a) provides that a "lawyer shall abide by a client's decision whether to
accept an offer of settlement of a matter." MODEL RULES OF PROF'L CONDUCT R. l.2(a)
(2004).
244. Russell Korobkin & Chris Guthrie, Psychology, Economics, and Settlement: A New
Look at the Role of the Lawyer, 76 TEXAS L. REV. 77 (1997) [hereinafter Korobkin &
Guthrie, Psychology, Economics, and Settlement] (finding attorney influence on clients
through education or explicit recommendations). See also DAVID A. BINDER & SUSAN C.
PRICE, LEGAL INTERVIEWING AND COUNSELING 186-87 (1977) (suggesting ways to
minimize this influence); Sternlight, supra note 237, at 318-19 (describing dependence of
client on attorney). See generally William L.F. Felstiner & Ben Pettit, Paternalism, Power,
and Respect in Lawyer-Client Relations, in HANDBOOK OF JUSTICE RESEARCH IN LAW 135
(Joseph Sanders & V. Lee Hamilton eds., 2000). For an empirical description of lawyer
client relations see HERBERT M. KRITZER, THE JUSTICE BROKER 55-67 (1990).
245. DAVID A. BINDER ET AL., LAWYERS AS COUNSELORS 287-308 (1991); see also
Robert F. Cochran, Jr., Legal Representation and the Next Steps Toward Client Control:
Attorney Malpractice for the Failure to Allow the Client to Control Negotiation and Pursue
Alternatives to Litigation, 47 WASH. & LEE L. REV. 819 (1990); Gifford, supra note 103;
Jacqueline M. Nolan-Haley, Lawyers, Clients, and Mediation, 73 NOTRE DAME L. REV.
1369, 1381-82 (1998) ("Prudential discussions between lawyer and client about the relative
merits of particular courses of action help to achieve participatory and educated client
decisionmaking. ").
246. BINDER ET AL, supra note 245, at 2-15 (emphasizing the importance of nonlegal
aspects of clients' problems); see also Nolan-Haley, supra note 245, at 1384 (arguing that the
"content of attorney-client deliberation takes into account the totality of the clients'
circumstances and may include the economic, social, psychological, moral, political, and
religious consequences of actions.").
247. BINDER ET AL, supra note 245 , at 29; see also Gifford, supra note 103, at 838-39
(emphasizing that information gained during the course of the representation or negotiation
"often lead[s] the client to change his expectations concerning the negotiation or his
preferences as to concessions").
248. Riskin, supra note 104 (describing problem definition in the context of mediation).
403
December 2003 ] Apologies and Legal Settlement 513
shadow of the law, with attention to the rights and responsibilities of

the parties and to the likely outcome of a trial process. This focus
defines the problem in its most narrow, legal sense. 249 As a general
matter, attorneys are likely to be more attuned to the legal
implications of the apology than are the parties. It is important for
attorneys to educate their clients about these legal implications. As
Jacqueline Nolan-Haley argues, "clients must have a general
knowledge about the relevant law governing their case, so that during
deliberation they may meaningfully evaluate alternative courses of
actions." 250 In addition, attorneys may be more likely to evaluate the
situation analytically rather than emotionally,251 and to be less
influenced by concerns for equity in responding to settlement offers
than are litigants. 252 Accordingly, attorneys are likely to be well suited
to assist the client in identifying the available alternatives and to
identify and predict the probable legal consequences of the
contemplated options. 253
At the other end of the spectrum, a broader conception of the
problem would include interests that are less likely to be the focus of a
legal proceeding. 254 For example, a broad problem definition might
include considerations of repamng and maintaining ongoing
relationships, restoring equity, handling emotional reactions, achieving
forgiveness - or the need for an apology. Attorneys are
249. Id. at 19.

250. Nolan-Haley, supra note 245, at 1385.
25 1. Susan Daicoff, Lawyer, Know Thyself- A Review of Empirical Research on Attorney
Attributes Bearing on Professionalism, 46 AM. U. L. REV. 1337 (1997); Chris Guthrie, The
Lawyer's Philosophical Map and the Disputant's Perceptual Map: Impediments to Facilitative
Mediation and Lawyering, 6 HARV. NEGOT. L. REV. 145 (2001); Korobkin & Guthrie,
Psychology, Economics, and Settlement, supra note 244, at 87 (" [L] egal training teaches
lawyers to analyze legal conflicts carefully and unemotionally rather than react to them
viscerally. "); see also Carrie Menk el-Meadow, The Transformation of Disputes by Lawyers:
What the Dispute Paradigm Does and Does Not Tell Us, 1985 J. DISP. RESOL. 25 [hereinafter
Menkel-Meadow, Dispute Paradigm] . Riskin writes:
Lawyers are trained to put people and events into categories that are legally meaningful, to
think in terms of rights and duties established by rules, to focus on acts more than persons.
This view requires a strong development of cognitive capabilities, which is often attended by
the under-cultivation of emotional faculties.
Leonard L. Riskin, Mediation and Lawyers, 43 OHIO ST. L.J. 29, 45 (1982).
252. Korobkin & Guthrie, Psychological Barriers to Litigation Settlement, supra note 97
(finding that attorneys were marginally less influenced by a concern for equity than were
students in the role of l itigants). But cf Richard B irke & Craig R. Fox, Psychological
Principles in Negotiating Civil Settlements, 4 HARV. NEGOT. L. REV. 1 (1999) (detailing a
variety of ways that psychological heuristics may affect legal negotiations by attorneys);
Sternlight, supra note 237 (detailing the same); Chris Guthrie et al. , Inside the Judicial Mind,
86 CORNELL L. REV. 777 (200 1) (demonstrating a variety of ways that psychological
heuristics may affect the legal decisionmaking of judges).
253. See BINDER ET AL., supra note 245, at 273-74.
254. Riskin, supra note 104, at 20-2 1.
404
correspondingly likely to be less well suited to identify and weigh

these nonlegal psychological and social considerations of the parties255
- including the valuation of either giving or receiving an apology.
Therefore, attorneys should remain mindful of the nonlegal
significance of apologies while guiding clients through a process of
considering and weighing the options given both their legal and non
legal consequences. In the process, attorneys must provide
information to allow the client to understand the legal consequences
of the options, while taking care to avoid monetizing the dispute when
the client could gain more utility from a settlement that combines
monetary and nonmonetary elements. 256
In this way, defense attorneys have a key role to play in assisting
their clients to critically evaluate the legal risks and benefits of an
apology.257 The traditional view has been to conceive of the defense
attorney's role as "to protect their clients from having to admit
wrongdoing or having to make legal compensation for the harms they
cause."258 A more nuanced advisory role, however, may be
appropriate. Defense attorneys can help their clients identify the
potential legal and nonlegal consequences of offering an apology. The
attorney may be best positioned to evaluate the merits of the case, the
probable outcomes of negotiation or trial, and the legal implications of
the apology for those outcomes - legal implications for the
possibilities for settlement, for liability judgments, and for damage
awards. The defendant may be best positioned to evaluate the
importance to her of offering the apology and of the other non-
255. See BINDER ET AL., supra note 245, at 272-73. Williams studied a sample of cases
that were scheduled for trial. He found that in fifty-three percent of the cases that did
proceed to trial, "the reason they went to trial was not a failure of the two attorneys to work
out a framework for agreement. Rather, it was the failure of one lawyer or the other to
'bring his client along.' " Williams, supra note 11, at 24. Williams notes that one way to
characterize this finding is as "a failure of one lawyer or the other to adequately
communicate with his or her own client, by which I mean a failure to adequately understand
the underlying interests and needs of one's own client." Williams, supra note 1 1 , at 25 n.74.
256. See Korobkin & Guthrie, Psychology, Economics, and Settlement, supra note 244,
at 134 (observing that clients may act in ways that maximize their utility, though not their
wealth); Menkel-Meadow, Dispute Paradigm, supra note 251 ; Riskin, supra note 251, at 44
(noting that on the lawyer's "standard philosophical map" the "victory is reduced to a
money judgment"); Sternlight, supra note 237, at 321-22 ("[The attorney] may regard these
[nonmonetary] interests as having little or no value."); id. at 324 ("Lawyers will also tend to
dismiss, as fluff, offers or requests for apology."). Moreover, lawyers may have to address
their own disincentives to settle cases for less cash and an apology. See ROBERT H.
MNOOKIN ET AL., BEYOND WINNING 1 18 (2000) ("As one plaintiff's attorney told us, 'You
can't pay the rent with one-third of an apology.' "); see also Sternlight, supra note 237, at 322
("[T]hese nonmonetary goals likely have little appeal for the attorney, who, after all, cannot
take a one-third contingency of an apology."). See generally KRITZER, LET'S MAKE A DEAL,
supra note 237, at 100-10.
257. See generally Cohen, Advising Clients to Apologize, supra note 3 (describing the
role of the attorney in advising a client to apologize).
258. Williams, supra note 1 1 , at 53 (describing this role).
405
tangible benefits of offering an apology. 259 In addition, particularly if a

partial apology is contemplated, the defendant, assisted by counsel,
must also consider aspects of the case that are likely to make such an
apology more or less effective.260 The defendant, then, in consultation
with defense counsel, can determine whether or not to offer an
apology given the way in which the defendant weighs these
considerations.
Similarly, plaintiffs' attorneys may play a key role in assisting their
clients in evaluating the credibility of an apology and any
corresponding settlement offer. 26 1 Plaintiffs' attorneys can help their
clients identify the potential legal and nonlegal consequences of
accepting a settlement offer following or accompanied by an apology.
Plaintiffs' attorneys may be best positioned to evaluate the merits of
the case, the legal consequences for the plaintiff of the defendant's
apology, and the effects of any applicable evidentiary rules. Moreover,
plaintiffs' attorneys may be able to help their clients avoid an unjust
financial settlement by explaining the client's legal entitlements and
predicting the outcomes of accepting the offer, continuing to
negotiate, or proceeding to trial. 262 In contrast, the client can provide
an assessment of the importance and meaning that receiving the
apology has for her. The plaintiff, then, in consultation with her
attorney, can assess the degree to which she is satisfied with the
settlement offer.
CONCLUSION
The findings reported here support the conclusion that apologies

can be beneficial in facilitating settlement of disputes. The results also
suggest, however, that apologies influence settlement in relatively
complex ways. Several factors, such as the nature of the apology, the
severity of the injury, and the other evidence of responsibility, affect
the capacity of an apology to facilitate settlement. Full, responsibility
accepting apologies positively impacted participants' perceptions of
the situation and the prospects for settlement. Conversely, partial
sympathy-expressing apologies had fewer effects, had both positive
259. See supra note 1 1 (describing nonstrategic benefits of a pologizing).

260. For example, the nature of the plaintiff's inj uries , the degree to w hich the
defendant appears responsible, etc.
261. See, e.g. , Cohen, Advising Clients to Apologize, supra note 3, at 1 067 ("If, as the
recipient of an a pology, you k now that the a pologizer has apologized in such a way as to
ens ure ins ulation from legal liability, you may attach less worth to that apology than
otherwise, and a plaintiff's lawyer should be sure to point this out to her client. ").
Conversely, plaintiffs' attorneys can raise the possibility with their clients that the other
party's failure to apologize s tems from the situational constraints of the legal system.
262. See Sternlight, supra note 1 1 , at 274 (arguing that even in mediation " the attorney
must be active and assertive to ensure that her client is not coerced by the opposing party" ).

145.109.109.226 on Thu, 12 Jan 2023 08:17:04 UTC
406
and negative effects, and were more dependent on context.

Policymakers, litigants, and lawyers must take into account these
complex effects when making decisions about the appropriate role of
apologies in settling civil disputes.
407
408

Recht en Menselijk Gedrag - Reader 2022-2023 (Ov)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Recht en Menselijk Gedrag - Reader 2022-2023 (Ov)

Uploaded by

Copyright:

Available Formats

Faculteit der Rechtsgeleerdheid

Amsterdam College of Law

2022-2023 / Semester 2 blok 1

Pleasance, Balmer en Denvir 2017

Rijnhout, Giesen, van Aken en van Hoof 2016

WEEK 2 – Implementatie van het recht I: Afschrikkende werking van straffen

Wermink, Blokland, Nieuwbeerta en Tollenaar 2009

Thornton, Gunningham en Kagan 2005

WEEK 3 – Implementatie van het recht II: Naleving in de sociale context

Kerr Lawrence 2015

Mascini en Van Wijk 2008

WEEK 5 – Geschilbeslechting I: Mobilisatie van rechten

Alexander en Prasad 2014

WEEK 6 – Geschilbeslechting II: Presentatie voor de rechter

Conley, O’Barr en Lind 1979

Wermink, de Keijser en Schuyt 2012

WEEK 7 – Geschilbeslechting III: Beleving door rechtzoekenden

Drie typen van rechtswetenschappelijk onderzoek

Figuur 1: Drie typen van juridisch onderzoek

Er zijn dus drie typen van juridisch onderzoek:

Typen van Empirisch Onderzoek

Voor- en nadelen van methoden: validiteit, 1 generaliseerbaarheid, en causaliteit

Hoe bestudeer je empirisch onderzoek

E. Voor- en nadelen (inclusief beperkingen) van de gebruikte methoden

F. Implicaties van de studie

Type Voordelen Nadelen en Beperkingen

Validiteit Generaliseerbaarheid Causaliteit

We live in a ‘law-thick’ world1 in which our daily activities are played

(2017) 80(5) MLR 836–859 837

11 Answering 64% of questions correctly, as compared to 60%.

838 (2017) 80(5) MLR 836–859

in resolving ‘justiciable’21 problems; notably in relation to whether advice is

BELIEFS ABOUT LAW, ATTITUDES AND SOCIAL NORMS

(2017) 80(5) MLR 836–859 839

Similarly, Darley et al’s study of people’s understanding of aspects of criminal

BELIEFS ABOUT LAW AND REFERENCE TO TIME

840 (2017) 80(5) MLR 836–859

CHALLENGES TO DISLODGING ERRONEOUS BELIEFS

(2017) 80(5) MLR 836–859 841

AIMS AND HYPOTHESIS

44 Williams and Hall, n 6 above.

842 (2017) 80(5) MLR 836–859

(2017) 80(5) MLR 836–859 843

844 (2017) 80(5) MLR 836–859

(2017) 80(5) MLR 836–859 845

First we used simple descriptive statistics to look at how respondents answered

846 (2017) 80(5) MLR 836–859

General knowledge and chancing answers

Overall, respondents answered 59 per cent of the fact-based scenario questions

(2017) 80(5) MLR 836–859 847

employment scenarios and, in stark contrast, towards a lesser number in the

848 (2017) 80(5) MLR 836–859

Scenario Specific question (percentage correct)

Housing scenario Is the landlord entitled to enter the house? (77.4%)

(2017) 80(5) MLR 836–859 849

Knowledge where it is needed

850 (2017) 80(5) MLR 836–859

Factors associated with knowledge

71 This included self-employed respondents.

(2017) 80(5) MLR 836–859 851

Whether Statistically Significant in Model

Variable Housing Employment Consumer