Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Standard view
Full view
of .
Save to My Library
Look up keyword
Like this
0 of .
Results for:
No results containing your search query
P. 1
A Comparison of Punishment Rules in Repeated Public Good Games an Experimental Study

A Comparison of Punishment Rules in Repeated Public Good Games an Experimental Study

Ratings: (0)|Views: 0 |Likes:
Published by Akbar Prasetya
Experimental Psychology
Experimental Psychology

More info:

Published by: Akbar Prasetya on Jan 18, 2014
Copyright:Attribution Non-commercial


Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less





A Comparison of Punishment Rulesin Repeated Public Good Games
 Institute for Operations Research Humboldt University Berlin
Strategic Interaction Group Max Planck Institute for Research into Economic Systems
 Department of Economics and International Institute of Infonomics Maastricht University, The Netherlands
Oneindividualandthreecollectivepunishmentrulesinapublicgoodsettingareanalyzed.Evidenceandexplanations for differences between the rules concerningpunishment intensity, contribution, profit levels,andjusticearepresented.Influencescrucialtoparticipants’supportforacollectiverulewhentheindividualruleisthestatusquoarealsoinvestigated.Resultsshowthatbesidesprofitdifferences,thedegreeofconsentrequired by the collective rule is essential for the degree of support by the participants.
 public good; free riding; cooperation; punishment; experiment 
distinctive characteristic of public goods is free access to the common good irre-spectiveofeachpersonscontributiontotheprovisionofthegood. Thischaracteristiccreates the incentive to free ride (i.e., enjoy the benefits while staying away from theprovision). Although free riding isrationalattheindividual level,itproduces sociallyundesirable outcomes at the aggregate level. Economists and scholars in other socialsciencesdirectedalotofresearchtosolvethisdilemma.Onewaytodosoisbyimple-menting additional institutions that create incentives to participate in the provision(e.g., a sanctioning system). Unfortunately, in a framework of rationality and selfish-ness, a new dilemmaarises:no individual iswilling to bear the costs of implementing
AUTHORSNOTE:WethankJörgBreitung,SimonGächter,WernerGüth,andMichaelKvasnickaforhelpful comments as well as Urs Fischbacher for z-Tree. Financial support by Deutsche Handelsbank AGBerlin and Deutsche Forschungsgemeinschaft (through SFB 373) is gratefully acknowledged. Additionalmaterialforthis article(i.e.,completedataset andexperimentinstructions)canbefoundontheWebsite of the journal (http://www.yale.edu/unsy/jcr/jcrdata.htm).
JOURNAL OF CONFLICT RESOLUTION, Vol. 47 No. 6, December 2003751-772DOI: 10.1177/0022002703258795© 2003 Sage Publications
or supporting the institution. This famous problem is known as the second-orderdilemma, or second-order public good (Oliver 1980; Bates 1988).From experiments and casual observations in reality, however, we have evidencethatpunishmentsystemsareindeedeffectivetoolstopromotecooperation.Asaresultof field studies, Elinor Ostrom (1990) worked out seven design principles, which arecrucial to the success or failure of a group or society facing a social dilemma. One of themissanctioning.
FehrandGächter(2000)showedexperimentallythatcostlypun-ishment opportunities, despite their dilemma characteristics, are used by the partici-pants and are able to raise and stabilize cooperation in a public good environment. Intheir experiment, they introduced a sanctioning system in which each participant hadto decide individually whether and to what extent he or she wants to punish anotherperson. The leverage, however, was rather high. The cost for imposing a fine onsomebody else was only one-third of the fine.Ifthecostsforpunishmentarehigh,however,membersofagroupprobablywishtodecidetogetheronthepotentialpunishmentoffreeriderssharingtheassociatedcosts.Ostrom(1990) reportscasesinwhichthedecisiononpunishmentwasfound inajointandorganizedway(e.g.,inavote).Anotherexample,theso-calledGrowthandStabil-ity Pact, contains a sanctioning system in which the members of the European Eco-nomic and Monetary Union (EMU) decide together on the punishment of countriesendangeringthestabilityoftheEuro.Acollectivesanctioningsystemmaybeusefulif the very structure of individual punishment bears strong incentives to abstain frompunishment (Weesie and Franzen 1998) or is likely to escalate and cause heavydamages.Two crucial questions may arise when designs of collective rules are considered:
1. Isitfeasiblethatpeopleacceptthecollectivedecisionwithallitsimplications,althoughthey favor a lower or higher punishment?2. Givenapositiveanswertoquestion1,whyisitnotfeasibletodirectlyenforceafullcon-tribution at the public good stage?
With respect to the first question, we argue that an institution, such as a collectivepunishment rule, issetup to continually gain benefitsfrom it.Opposing itin one casecouldputthewholeinstitutionatrisk.So,oncepeoplehaveerectedaninstitution,theywill probably continue to obey it for their long-term benefit. Moreover, it is conceiv-able that the enforcement of the institution is backed by an authority, which waserected by the people in advance and is out of control in the current situation. Ostrom(1990) reports on successful cases where people designed sophisticated systems toensurethatpunishmentcostsareshared.
Whetherpeopleareindeedwillingtosubmitthemselves to an institution, such as a collective punishment rule, is one of thequestions we seek to answer in this study.
1. Although Ostrom’s main field of study is common pool resources (CPR), she argues that “giventhesimilaritybetweenmanyCPR problemsandtheproblemsofprovidingsmallscalecollectivegoods,thefindingsof this volumeshouldcontribute an understandingof . . . the capabilities of individualsto organizecollective action related to providing local public goods” (Ostrom 1990, 27).2. Shereports,forexample,onirrigationsystemsinSpainwhereguardsandso-called“ditchriders”werepaidformonitoring,reportingviolations,andbringingchargesagainstfarmers(Ostrom1990,69-78).
The answer to the second question makes it important to distinguish between theinitial public good in the first stage and the public good “punishment” in the secondstage. A certain feature of collective punishment regulation is that punishmentbecomes a central issue and is, therefore, easy to enforce as discussed above. In con-trast, the contribution decision remains in private responsibility and cannot be trans-ferred to a central or an outside institution. In some cases, the contribution is even notoronlyexpostobservable.Thus,aparticipantdecidingbyhimselforherselfonhisorher contribution still faces incentives to free ride. Even a contract among the partici-pants that fixes contributions is no credible commitment and, therefore, is likely to beviolated.Forthereasonsgivenabove,weassumeinthisstudythatcontributionstothepublicgoodinthefirststagearenotenforceable,whereasthedecidedpunishmentinthesec-ond stage is. Our main goal is to investigate collective punishment rules and confrontthem with an individual rule, which we consider an initial position. The followingquestions are central to our analysis:
 Are collective punishment rules able to bring about stronger cooperation and/or higherprofits in a public good setting than an individual rule?
 To what extent do different collective rules perform differently from each other withrespect to contribution, efficiency, and justice?
 Willparticipantsagreetosubmitthemselvestoacollectiverule,evenifthismeanstogiveup some individual freedom? Which rule is preferred?
To find answers to the above questions, we designed an experiment in which partici-pantsrepeatedlyplayapublicgoodgame.Betweentherounds,theyhadthepossibilityto punish each other according to different rules. Every participant experienced a col-lective punishment rule and an individual punishment rule. After that, participantscould bid for the right to choose the institution for the last five rounds.The results of our experiment suggest basically that the more severe an institution,the higher the contribution to the public good but the lower the willingness of partici-pantstoacceptthisinstitution.Inthesecondsectionofthestudy,wewillfirstdescribetheexperimentaldesign.Then,inthethirdsection,weestablishsomelinkstotheliter-aturethatguided our expectations and report detailedresultsinthefourth section.Weconclude in the final section with some general remarks.
Inourexperiment,participantsplayedatwo-stagegamefor20periodsingroupsof four. The first stage consisted of a standard linear public good game (see Ledyard1995). Each group member
 received an initialendowment of 20 ECU (experimentalcurrency units) and had to decide on which amount
 to contribute to a public good(“pot”). The pot was multiplied by 1.6 and then equally distributed among the fourgroup members regardless of their contribution. Thus, ECU 1 contributed to the pothadamarginalreturnofECU0.40toeachparticipant,whereasECU1keptforonesel

You're Reading a Free Preview

/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->