You are on page 1of 16

Self-Regulation Failure: An Overview

Author(s): Roy F. Baumeister and Todd F. Heatherton


Source: Psychological Inquiry, Vol. 7, No. 1 (1996), pp. 1-15
Published by: Taylor & Francis, Ltd.
Stable URL: http://www.jstor.org/stable/1449145
Accessed: 11/08/2010 13:39

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=taylorfrancis.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to Psychological
Inquiry.

http://www.jstor.org
Psychological
Inquiry Copyright1996by
1996, Vol. 7, No. 1, 1-15 LawrenceErlbaumAssociates, Inc.

__~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

TARGET ARTICLE
~~~ ~
I

Self-Regulation Failure: An Overview


Roy F. Baumeister
Departmentof Psychology
Case WesternReserve University

Todd F. Heatherton
Departmentof Psychology
DartmouthCollege

The majorpatterns of self-regulatoryfailure are reviewed. Underregulationoccurs


because of deficient standards, inadequate monitoring, or inadequate strength.
Misregulationoccurs because offalse assumptionsor misdirectedefforts,especially
an unwarrantedemphasis on emotion. The evidence supports a strength (limited
resource)model of self-regulationand suggests thatpeople often acquiesce in losing
control. Loss of control of attention,failure of transcendence,and various lapse-ac-
tivatedcauses all contributeto regulatoryfailure.

ModernAmericansociety suffersfroma broadrange their own behavior.Pioneers such as Mischel (1974)


of problemsthathave self-regulationfailureas a com- and Bandura(1977) proposed and demonstratedthat
mon core. Crime, teen pregnancy, alcoholism, drug humanbeings do seem to have the unique capacity to
addiction,venerealdisease, educationalunderachieve- alter their own responses. Over the past 2 decades,
ment, gambling, and domestic violence are among the theory and researchhave advancedthe understanding
social problemsthatrevolve aroundthe apparentinabil- of self-regulation considerably (Carver & Scheier,
ity of many individualsto disciplineand controlthem- 1981; Kanfer& Karoly, 1972) and models of self-reg-
selves. Although economic, political, and sociological ulationhave been appliedin diverse areas(e.g., educa-
causes may be relevant to such issues, the proximal tion, drug treatment, emotional control, and task
importanceof self-regulationfailure to many cases is performance). Despite the substantial progress in
undeniable.Moreover,therearemanyadditionalprob- studying how self-regulationcan function, however,
lems with self-regulationthat cause considerablesuf- relativelylittle effort has been devoted to directexam-
feringto individualseven if they do not menacesociety inationof failuresat self-regulation(cf. Kirschenbaum,
at large (e.g., eatingbinges, spendingsprees,procrasti- 1987).
nation,and inappropriategoal setting). The purpose of this article is to offer a theoretical
Researchersin the psychology of the self have re- treatmentof self-regulationfailure. We have recently
cently begun to recognize thatone of the most elusive, reviewedthe multipleliteraturesdealing with the many
important,anddistinctivelyhumantraitsis the capacity specific spheresof self-regulationfailure (Baumeister,
of humanbeings to altertheir own responsesand thus Heatherton,& Tice, 1994), and in this articlewe artic-
remove them from the direct effects of immediate, ulate some of our main conclusions. Because the em-
situationalstimuli.An understandingof self-regulation pirical literatureon these topics is extensive, we cite
failure would therefore have considerable value not evidence here only to illustrate key points. A com-
only for its applicationsto widespreadsocial and per- prehensive review of currentresearch knowledge is
sonal problems, but also to basic research and the beyond the scope of this article,and interestedreaders
construction of an adequate theoretical account of arereferredto the book.
humanselfhood. Self-regulationis a complex, multifacetedprocess,
Although conceptions of volition and self-control and so it can break down in several different ways.
have long been of philosophical, religious, and legal Therefore,it is notpossible to identify a single cause or
interest, only recently have psychologists focused on causal sequence that will explain all instances of self-
the extentto whichpeople influence,modify,or control regulationfailure.Instead,there are several main pat-
BAUMEISTER& HEATHERTON

terns, any one of which can produce self-regulation For instance,people who underestimatetheir abilities
failureindependently. may fail to initiateattemptsto achieve theirgoals.
The most basic distinctionis between underregula- The thirdingredientof self-regulationis contained
tion and misregulation(e.g., Carver& Scheier, 1981). in the operate phase of the feedback loop. The idea is
Underregulationentails a failure to exert self-control; that when the test phase reveals that the currentstate
often, the persondoes not botheror does not manageto falls short of the standards,some process is set in
controlthe self. In contrast,misregulationinvolves the motion to change the currentstate. Past theories have
exertionof controlover oneself, butthis controlis done not devoted a great deal of attention to how these
in a misguidedor counterproductivefashion,andso the processes actually function to bring about change,
desiredresultis not achieved.At present,thereis more partly because they may have seemed complex and
researchavailableon underregulation thanon misregula- heterogeneous.Still, it is clear thatself-regulationfail-
tion, andit also appearsthatunderregulationis the more ure can occur despite clear standardsand effective
commonsortof problem.Aftera briefdiscussionof the monitoring, simply because the person is unable to
natureof self-regulation,we examine underregulation bringaboutthe desiredchange.
first and then proceed to misregulation. We have found it useful to conceptualizesuch oper-
ate changes in termsof one internalprocess overriding
another.Certainresponses are set in motion, either by
Three Ingredients of Self-Regulation innate programming,learning,habit, or motivation-
and self-regulationinvolves overridingthem. In other
Feedback-loopmodels of self-regulation,suchas the words, a great many instances of self-regulation in-
one elaboratedby CarverandScheier(1981, 1982;also volve a response that is initiatedby a combinationof
Carver, 1979), indicate three main ingredientsof self- latent motivationsand activating stimuli; self-regula-
regulation,and these suggest threemainpossible path- tion is a matterof interruptingthat response and pre-
ways for self-regulationfailure.The first ingredientis ventingit from runningto its normal,typical outcome.
standards,which areideals, goals, or otherconceptions For example, a beer commercial(an activatingstimu-
of possible states. Without clear and consistent stan- lus) may bring to the fore one's liking for alcohol (a
dards,self-regulationwill be hampered.Therefore,ei- latent motivation) and create an impulse to consume
ther a lack of standardsor a dilemma of conflicting, alcohol; however, the person who is trying to reduce
incompatiblestandardscan preventeffective self-reg- his or her drinkingwill seek to override the response
ulation. There is indeed evidence that such inner con- sequence and preventit from leading to the consump-
flicts can impair action and undercut efforts at tion of such a beverage.1
self-regulation(e.g., Emmons& King, 1988;Van Hook In many cases, impulses are automaticin the sense
& Higgins, 1988). Moreover, inappropriatestandards of being beyonda person's volitionalcontrol.Thus,the
(i.e., those thataretoo high or too low) can also hamper termimpulsecontrolis misleading.Self-regulationis a
and thwart self-regulation (Heatherton& Ambady, controlled process that overrides the usual conse-
1993). quences of an impulse ratherthan preventingthe im-
The second ingredient is monitoring. The "test" pulse from occurring.The problem is not that people
phaseof feedback-loopmodels involves comparingthe have impulses;rather,it is thatthey act on them.
actual state of the self to the standards,and to do that
the personmust monitorhim- or herself. Keepingclose Self-Regulatory Strength: A Limited
track of one's actions and states is often vital to suc- Resource
cessful self-regulation, and so when people cease to
monitor themselves they tend to lose control. Eating We turnnow to the issue of whatenables a personto
binges, for example, seem to occur when the person override a habitual or motivated response sequence.
ceases to keep trackof what he or she is eating (for a How does thepacifistturnthe othercheek andhow does
review, see Heatherton& Baumeister, 1991; Polivy, the dieterrefrainfrom eating his or her fill? It is clear
1976). A particularlyimportantfactor is alcohol con-
sumption, which reduces self-attentionand therefore
makespeople less able or less willing to monitorthem- 'We use the term impulseto refer to an inclinationto performa
selves (Hull, 1981). Alcohol consumption has been particularaction on a particularoccasion. Thus, impulses are highly
foundto promoteself-regulatoryfailurein manydiffer- specific in contrastto motivations,which maybe generalor abstract.
Impulsesarisewhen motivationsencounterspecific, activatingstim-
ent spheres (Baumeister et al., 1994; Steele & uli in a particularsituation. For example, hunger is a motivation,
Southwick, 1985). The failure to judge one's abilities whereas the wish to devour one of those fragrant,sizzling cheese-
accuratelymay also impede successful self-regulation. burgerson the grill is an impulse.

2
SELF-REGULATIONFAILURE

that impulses and motivations vary according to ular,manypatternsof self-regulationbreakdown when


strength,and the weakerones are those that are easier people are understress, presumablybecause the stress
to controland stifle. If the impulseshave strength,then depletes their self-regulatorycapacities. People be-
whateverstifles themmustpresumablyconsist of some come moreemotionalandirritable,they aremorelikely
greaterstrength.Our own research(Baumeisteret al., to increasesmoking,breakdiets or overeat,abusealco-
1994) led us to concur with other scholars such as hol or other drugs, and so forth when under stress.
Mischel (in press) who have suggested that strength Glass, Singer, and Friedman(1969) found that coping
models are apt and useful for self-regulationtheory. with stress seemed to have a "psychic cost" that took
Underregulationis thus often a matter of the inade- the form of lowered self-regulatorycapacity, as mea-
quacy of one's strength to override the unwanted suredby subsequentcapacitiesto make oneself persist
thought,feeling, or impulse. More precisely, our over- in the face of frustrationandto concentrateon a difficult
view of the self-regulationliteraturesuggests thateach task.
person's capacity for self-regulationappearsto be a Likewise, if we assume that people are generally
limitedresource,which is renewableover time andcan fatigued late in the evening, then self-regulation
be increasedor decreasedas a resultof gradualdevel- should breakdown more at such times than at others.
opmentsor practice.One cannotregulateeverythingat Evidence about the timing of such self-regulatory
once. failures is consistent with the fatigue hypothesis (al-
Adopting a strength model of self-regulation has though some of these effects are confounded by the
several importantcorollaries for understandingself- fact that people are more likely to have consumed
regulationfailure. There will be importantindividual alcohol late in the day and alcohol impairs monitor-
differencesin self-regulatorystrength,which shouldbe ing, thereby also weakening self-regulation). Diets
consistent across a variety of spheres. There is some are most often brokenlate in the evening; sexual acts
evidence to supportthis view. Thus, individualdiffer- that one will later regret are likewise most common
ences in the capacity to delay gratificationpredict a then;people smoke and drinkmost heavily late in the
varietyof interpersonaltraitsandbehaviorsthatreflect day; most violent and impulsive crimes are commit-
self-control (Funder,Block, & Block, 1983) and can ted between 1:00 and 2:00 a.m.
even predictacademicperformanceover a decadelater These first two implicationsof the strengthmodel
(Mischel, Shoda, & Peake, 1988; Shoda, Mischel, & furnish a basis for predicting the intercorrelations
Peake, 1990). Also, the sameindividualsshow self-reg- among indicationsof self-controlin multiple spheres.
ulatorydeficits across a broadspectrumof both legal If there are individual differences in self-regulatory
and illegal behavior.A typical criminal,for example, strength,then over the long run there will be positive
will not specialize in one particularkind of illegal correlationsbecause strong people will tend to have
activity,butratherwill commita varietyof crimes,and relativelyhigh levels of self-controlin all spheres. On
he or she will also be proneto smoke cigarettes,spend the otherhand,in the shortrunthe correlationswill be
impulsively (thereby dissipating any financial gains negativebecausedevotingone's self-regulatoryefforts
from crime), become involved in unwantedpregnan- to one sphere will take away what is available for
cies, fail at marriage,abuse alcohol and drugs, have controllingoneself in otherspheres.Researchersinter-
high absenteeismat workor school, andengagein other ested in overlapsbetween self-regulatoryeffectiveness
behaviorsindicative of poor self-regulation(Gottfred- in differentspheresmay need to be alertto these oppos-
son & Hirschi, 1990). ing empiricaltendencies.
The second implicationof a strengthmodel is thata The third implication is just as it is possible to
personcan become exhaustedfrommanysimultaneous increasestrengthby regularexercise, so self-regulation
demandsandso will sometimesfail at self-controleven should become easier the more one does it. This has
regardingthings at which he or she would otherwise been assertedby James (1890/1950) and many other
succeed. As a limitedresource,self-regulatorystrength observersof human behavior but we do not know of
can be temporarilydepleted.At any given time, a given strongempiricaltests of the hypothesis.In this connec-
person will only be able to regulateso much of his or tion, it is of considerablerelevancethat new programs
her behavior, and so when strength is depleted by forprisoners(e.g., "bootcamps")involve military-style
demands in one sphere, self-regulatorybreakdowns training,in which an attemptis made to instill self-dis-
may occur in others.In particular,fatigue or overexer- cipline by means of enforcing externaldiscipline. Al-
tion will depletethe person's strengthandhenceunder- though the effectiveness of these programshas yet to
mine some patternsof self-control. be decided, we predictthattheirsuccess at rehabilitat-
The evidence regardingsuch short-termdepletions ing prisonerswill be in proportionto their success at
is not extensive but it is broadandconsistent.In partic- strengtheningself-regulatorycapacities.

3
BAUMEISTER& HEATHERTON

One implicationof the notion of increasingstrength that managing attention was the most common and
is thatpeople may become betterat practicingself-de- often the most effective formof self-regulationandthat
nial or impulse controlover time. Ironically,this could attentionalproblemspresageda greatmany varietiesof
meanthatpeople who repeatedlyquitsmokingor go on self-regulationfailure.Withcontrollingthoughts,emo-
diets may graduallybecome more effective and suc- tions and moods, task-performanceprocesses, and ap-
cessful. Schachter (1982) contended that people im- petites and impulses, the effective management of
prove at quittingsmokingwith practice.Prochaskaand attentionwas a powerful and decisive step, and self-
DiClemente (1984, 1986) arguedthat people become regulatoryfailure ensued when attentioncould not be
betterat quittinga varietyof addictionswhen they do managed(Baumeisteret al., 1994; see also Kirschen-
it multiple times. Of course, the fact that they are baum, 1987; Wegner, 1994).
quittingagainmeansthatthe prioreffortto quitwas not Forourpurposes,the key pointis thatthe importance
a permanentsuccess but it may be the case that one of attentionis at least partlyattributableto the inertia
learns to quit through successive approximations. principle.Effective managementof attentioncan pre-
There could be several reasons for progressive im- vent the unwantedresponse sequence from starting,
provementat impulse control but one of them clearly which makesit relativelyeasy to preventthe unaccept-
is the possibility of increasingstrength. able outcome. In contrast,if attentionescapes control
it can set the unwantedresponses in motion, and once
they acquireinertiathey are more difficult to control.
Inertia and Attention In simple terms,it is easier to avoid temptationthanto
overcome it.
A fair amountof evidence suggests thatpsycholog-
ical responsesaremarkedby somethingakinto inertia,
which makes them difficultto interrupt.The terminer- Transcendence
tia is borrowedfromphysics, in which it referredto the
(now discredited)theorythatbodies in motionacquired One particularlyimportantform of attentioncontrol
a force that sustainedthem in motion. is transcendence.Transcendenceis a matterof focusing
We proposethatpsychologicalprocessesdo acquire awareness beyond the immediate stimuli (i.e., trans-
a kindof inertia(unlikephysicalprocesses).Indeed,the cendingthe immediatesituation).This does not neces-
longer a responsehas gone on the moreinertiait seems sarilymeanignoringthe immediatepresentso much as
to have and hence the more difficult it is to override. seeing it in the context of more distal concerns (e.g.,
This theoretical principle is not new (indeed, the values, goals, and motivations). Phenomenologists
Zeigarnikeffect involvedthe principlethatinterrupting have emphasizedtranscendenceas a particularlyim-
an activity becomes more strenuous as it nears its portantcapabilityof humanconsciousness.
completion) but its importancefor understandingself- Dieting offers a clear example of transcendence.
regulationhas been neglected. Humanbeings may be the only species on the planetin
Effective self-regulationoften seems to involve in- which hungry individuals will voluntarily refuse to
terveningas early as possible. For example, if the goal consume readily available, appealing food. Effective
of self-regulationis the preservationof chastity, it is dieting does, however, requirethe person to transcend
often more effective to interruptsexual activitiesat the the effects of the immediatestimuli. By contemplating
first kiss ratherthanafteran hour's worthof escalating long-rangegoals and concerns, such as how one will
physical contact. The effectiveness of early interven- look in a bathingsuit next summer,people areavailable
tion may well reflect the operationof inertia:To mini- to framethe attractivefood as a problematicor danger-
mize inertia, self-regulatory efforts may be most ous obstacle ratherthanas an appealingmorsel.
profitably focused on the very first stages of all re- Therefore, one proximal cause of self-regulation
sponse sequences. failure is the failureof transcendence.When attention
Most models of the cognitive control of behavior slips off of long-rangegoals andhigh ideals andinstead
begin with attentionbecause noticing somethingis by becomesimmersedin the immediatesituation,self-reg-
definitionthe first stage in informationprocessing.As ulation is in jeopardy. Whatever functions to direct
a result, one would expect that managing attention attentionto the here and now will tend to weaken the
would be importantin many or all spheresof self-reg- capacity for self-regulation. This may include both
ulation,and, as a corollary,the loss of attentionalcon- situationalanddispositionalfactors.Situationalfactors
trol will be a common firstharbingerof self-regulatory include those thatpromotedeindividuation.There are
failure. Our review of multiple, empirical literatures also individualdifferencesin theextentto whichpeople
confirmedthese hypotheses.Over and over, we found are influenced by environmental cues. Schachter's

4
SELF-REGULATIONFAILURE

(1971) influentialexternalitytheoryof obesity argued murdererscan hardly recall even the next day what
thatsome individuals(i.e., the overweight)were espe- made them so violent. However, in the heat of the
cially prone to be influenced by external cues about moment(i.e., the short-termattentionalfocus causedby
eating. high emotion), people fail to consider long-range im-
The capacity to delay gratification is one of the plications and act in response to short-termconcerns,
importantroots of self-regulationtheory.A successful which may include winning the dispute at all costs and
delay of gratification requires the person to forego by violent means.
immediately available rewardsin favor of larger but A second mechanism by which emotional distress
remoterones, andkeeping oneself fromthinkingabout may thwarttranscendenceand impairself-regulation
the immediate rewards is often a vital part of that occurs when the source of emotional distress is not
success (Mischel, 1974;Rodriguez,Mischel, & Shoda, present in the immediate situation but is highly avail-
1989). KarniolandMiller (1983) showed thatself-reg- able in memory (e.g., just after one has received a
ulatory failure (in this case, the failure to choose the major rejection or failure experience). Under such
delayed gratification) is often preceded by shifts in circumstances, people will seek to distract them-
attentionto the immediatereward.This shift in atten- selves to prevent themselves from thinking about the
tion to the immediate situationis a form of transcen- upsetting event; immersion in powerful, short-term
dence failure. stimuli may be an effective means. Unfortunately,
Transcendenceis often a vital aspect of emotion some of the most compelling short-term stimuli are
regulation.People overcome anger,frustration,or dis- precisely the things that the person is otherwise try-
appointmentby looking beyond the immediate situa- ing to control (e.g., alcohol, sweet foods, or drugs).
tion. They imagine how thingscould have been worse, A great deal of binge behavior, whether it be shop-
conjureup possible positive outcomes thatmay derive ping, gambling, eating, drinking, or having sex,
from the currentsetback, or speculate about possibly seems to result when people are seeking to keep their
beneficial motives that the other (offending) person attention focused on immediate, concrete stimuli as
may have had. Emotion is typically linked to a partic- a means of keeping it away from some threateningor
ular value judgment about a particularevent or situa- upsetting thoughts.
tion;by transcendingthe situation,one can escape from To be sure,emotionis notinvariablybadfor self-reg-
the emotion thatis linked to thatvaluejudgment. ulation. Some emotions, such as guilt, may even help
Indeed, it is plausible that the contributionof emo- self-control (e.g., Baumeister, 1995; Baumeister,
tional distress to self-regulationfailure is often medi- Stillwell, & Heatherton,1995). Still, these instancesare
ated by effects on transcendence.The interrelations consistent with the general argumentsabout transcen-
among emotion, attention,and self-regulationare not dence becausethey referto cases in which the emotion
well understoodandso commentsmustbe speculative, facilitates self-regulationby actually promoting tran-
but we propose the following: Emotion increases the scendence. A dose of anticipatoryguilt may help the
salience of whateverproducesthe emotion and so at- person realize that what he or she is about to do may
tentionwill tendto focus on whateverhas promptedthe cause damage to important,desired relationships or
emotion.Most commonly, somethingin the immediate have otherunwantedconsequences, and so the person
situationis the cause and so emotion tends to have the may interruptthe pursuitof some short-termgoal or
effect of concentratingattentionin the here andnow- reward. By calling attention to distal outcomes and
therebythwartingtranscendenceand makingself-reg- meaningful implications, guilt helps the individual
ulation more difficult. Violent behavior provides an transcendthe immediatesituationand its temptations,
importantillustrationof such effects of emotion. Vio- therebyaiding self-control.
lence typically results because the person becomes Transcendenceis even relevant to some aspects of
angry at some pressing stimulus-a rival who insults task performance,which is an importantsphere for
one, a child who cries excessively, a spouse who frus- self-regulation.In particular,persistence at difficult,
tratesone's wishes (e.g., Berkowitz, 1989). The anger boring, and unpleasanttasks is a challenge that is en-
keeps attentionconfined to the immediate,provoking demic to many forms of work, and such persistence
situation and so efforts to restrainone's violent im- often requiresthe person to transcendthe immediate
pulses are made more difficult. In their discussion of situation, which on its own merits would seemingly
the role of self-control failure in causing crime, favor quitting. Sansone, Weir, Harpster,and Morgan
Gottfredsonand Hirschi (1990) pointed out thatlong- (1992) showed thatpersistenceon boringtasks is facil-
range considerationswould often militate againstvio- itatedby mentally transformingthem into more inter-
lence. Thus, most murdersbring far more harm than esting processes. Indeed, studies of blue-collar
benefitto the perpetratorsthemselves,andindeedmany manufacturingworkershave shown that such workers

5
BAUMEISTER& HEATHERTON

tend to restructuretheir tedious, repetitioustasks into This question has importantimplications. One set
elaborate games; when they are successful, they be- concerns basic theoreticalquestions of conscious con-
come totallyengrossedin these games to the extentthat trol and intrapsychicconflict. Another concerns legal
they continueto talkaboutthemeven duringbreaksand issues: Are violent crimes the product of irresistible
lunchhours(Burawoy,1979). By extension,when peo- impulses or deliberatechoices? Political issues such as
ple are unable to effect such transcendentreconcep- whetheraddicts,alcoholics, spouse abusers,andothers
tions of these tasks, they are more likely to quit, should be treatedas needy victims or as criminal de-
which can be a severely problematicform of self-reg- generatesalso revolve aroundthis question. Given the
ulation failure. sweep of these implications, it is not surprisingthat
We noted earlierthat alcohol was implicatedas one there are ample argumentson both sides in both the
cause of a great many varieties of self-regulationfail- professional journals and in the popular and mass
ure. Although we suggested thatalcohol's impairment media. We thinkthatan additionalreasonfor the exis-
of self-monitoringmay be one mechanism by which tence of both sides of the argumentis that there is in
alcohol has these effects, it is plausiblethatanotherone fact a large, gray area. In our view, self-regulation
is throughthe impairmentof transcendence.Steele and failure is rarely a matter of deliberate, premeditated
Josephs (1990) coined the term alcohol myopia to choice, but then again it is not often a matterof irresist-
describe the way alcohol limits attentionand restrains ible impulses either.
it to a few proximalstimuli.Theirargumentcan readily Duringthe periodwe spentreadingaboutand study-
be extended to say that alcohol impairs the sort of ing self-regulation,we grew increasingly skeptical of
long-range,abstract,meaningful,or mentally flexible the irresistibleimpulse notion. By definition, such im-
thinkinginvolved in transcendence(and, in fact, alco- pulses cannot be resisted and so they refer to things
hol does seem to increasethe responsivityto immediate people would do even if someone were holding a gun
stimuli rangingfrom violent to sexual to appetitive). to their heads and threateningto kill them if they did
Thus, self-controloften involves seeing the immedi- the forbiddenacts. Despite the popularityof the notion
ate situationin terms of long-rangeconcerns, values, of irresistible impulses in courtroom settings, it is
and goals (see also Carver& Scheier, 1981; Rachlin, readily apparentthat people could and would refrain
1995;Vallacher& Wegner, 1985). The abilityto main- from most behaviorsif their lives dependedon it. The
tainattentionandfocus on these long-termissues is one vast majorityof impulses are resistible.
ingredientof self-regulatorystrength.In general, fac- Thus, the popularimage of the passive victim over-
tors that bind attentionto the immediatesituationand come by powerful, irresistibleimpulses cannot be ac-
pressing stimuli will tend to contributeto self-regula- cepted except in a few rareand extremecases (e.g., the
tion failure. fact that people cannot indefinitely postpone certain
biological functionssuch as falling asleep, urinating,or
breathing-all things that people will eventually do
Acquiescence and Overriding even despite a gun to the head). In reviewing the
empiricalliteratureon self-regulationfailure,we found
One of the most importantyet controversialaspects over and over that there was significant evidence of
of self-regulationfailureis the questionof the extentto deliberate,volitional participationby the individualin
which people acquiesce in it. The question can be the forbiddenactivity. These findings and patternsdo
appreciatedby considering two contrary images of not ruleout the possibilitythattherearepoints at which
self-regulationfailure. Both of them depict a person people feel helpless and passive and are overcome by
who feels an impulse to act in a way thatrunscontrary strongimpulses.They do, however,suggestthatthe full
to his or her normal standardsof proper, desirable episode of self-regulation failure usually involves at
behavior.Self-regulationfailuremeans acting out that least some elements of active acquiescence.
impulse and thus violating the person's standards.In Let us consider some examples in which there is
one image, the well-intentionedpersonis overwhelmed evidence of acquiescence in self-regulation failure
by an irresistibleimpulse that no normalperson could (Baumeisteret al., 1994). Cigarettesmoking is a good
restrain.In the other,the personsimply decides to give example because in the contemporaryUnited States it
in to the impulseratherthango throughthe exertionand is typically inconvenient, if not outright difficult, to
frustrationthat would accompanyself-restraint.Thus, smoke. The would-be smoker must obtain cigarettes
is self-regulation failure a matter of lazy self-indul- and then find a time and place where smoking is still
gence (i.e., heedlessly giving in to temptation)or is it a allowed. The personmust then go throughthe motions
matter of being overcome by powerful, unstoppable of lightingup andinhaling.Smokingis well recognized
forces? as a powerful addiction and as a source of strong

6
SELF-REGULATIONFAILURE

cravings and unpleasantwithdrawalsymptoms, all of changed the plannedroute to pass throughReno, Ne-
which may be beyond the smoker's control;however, vada,which he claimed would be more scenic. (Seattle
smokingis not a matterof simply going limp, becoming is northof San Francisco;Reno is east of it.) In Reno,
passive, and letting it happen. he needed change for a parkingmeter and so entered
Consumingalcohol (or takingotherdrugs)is subject the nearestbuilding,whichjust happenedto be a casino.
to a similaranalysis. Despite the undeniableaddictive- While in the casino, he decided to place a single bet to
ness of alcohol, and despite popularimages thatmany test his luck. The ensuing 3-day gambling binge was
people cannotcontroltheirdrinking,it is clearthatmost perhaps not deliberately planned in advance but the
people who drink alcohol are actively acquiescing in decisions thatbroughthim thereseem disingenuous.In
the process. Orderingor pouring a drinkand raising a similarfashion,people do pick fights in which they lose
glass to one's lips are deliberate,volitional acts. Binge control, manufacturereasons for consuming alcohol,
eaters likewise often describe their eating as out of place themselves in temptingsituations,and engage in
control,yet in many cases the personmustacquiesceto other patternsthat seem as if they were conspiring to
the extent of orderingor preparingfood, puttingit into thwarttheirown self-regulatoryprograms.
one's mouth, and chewing and swallowing it. Apartfrom such extreme cases, it would usually be
Procrastinationis another common self-regulation inappropriateto say that the person plannedand engi-
problem and procrastinatorsmay often feel like pas- neeredthe entirescenarioin advance;in fact, the person
sive, helpless victims, especially during the eventual may often be quitechagrinedby the eventualoutcome.
crisis when the deadlinelooms and the remainingtime To simply say that the self-regulationfailures reflect
is inadequate for the task. Procrastination,however, deliberate free choice would therefore be somewhat
often involves actively doing otherthingsinsteadof the misleading. On the other hand, the stereotype of the
deferredactivity. Back when there was ample time to helpless,passive victimoverwhelmedagainsthis or her
begin workon the task,thepersonwas hardlyovercome will by uncontrollableimpulses is not accurateeither.
by an irresistibleimpulse to go out for a beer or watch The person did participate,more or less freely and
television instead. Rather,the person actively partici- deliberately,in the actionsthatconstitutedthe self-reg-
pated in these other activities. ulationfailure.
If procrastinationinvolves a failure to get started, In orderto resolve the issue of acquiescence,it is first
performanceis also affected by whetherpeople persist necessary to appreciatethat there are often costs and
or quit, and so the matterof deciding when to quit can disadvantagesto self-control.Foregoinganimmediate,
be an importantaspect of the self-regulationof perfor- desiredpleasureis only one of them. Frustration,with-
mance. Althoughthereareoccasionallycases in which drawal, and feelings of deprivation may be acute.
sheer exhaustion forces the person to stop (e.g., when Moreover,if our hypothesisof self-regulatorystrength
marathonrunners collapse and are carried away on is correct,then maintainingself-control and resisting
stretchers),usually the decision to quit is much more temptationcan be a tiringanddrainingexperiencethat
fluid and negotiableandthe personcould have gone on can even consume resources that may be needed for
a little longer.Quittingduringtaskperformanceusually otheracts of self-control.
occurs well before the point of full exhaustion. The Resistingtemptationis thus,in manycases, an ongo-
person somehow selects a point at which to quit and ing (orperennial)andunpleasantexertion.Its difficulty
then goes and does somethingelse. is likely to fluctuateas a function of the strengthand
Delay of gratification is one of the prototypes of salience of the competingimpulse and of the self-reg-
impulse control;yet, in many studiesof delay of grati- ulatorycapacity.An irresistibleimpulse is hardlynec-
fication, the participantmust make some active re- essary for self-regulation failure; rather, a moment
sponse to obtain the immediate reward.Making that duringwhich the impulseis especially strongor attrac-
response is often a matterof deliberateaction. Outside tive, while the self-regulatorystrengthis temporarily
the laboratory,failuresto delay gratificationmay often depleted,maybe sufficient.At some point,perhaps,the
involve even more extensive and obvious forms of costs of exerting control may simply seem too high,
active acquiescence(e.g., when the persondropsout of whereasthe anticipatedbenefits may seem too remote
college or empties a savings account). or uncertain or simply too small and so the person
There are of course instances in which the person's gives in.
acquiescence is even more extreme. People do some- We arethusportrayingthe abrogationof self-control
times seem to arrangeto lose control. Marlatt(1985) as a deliberatechoice, butit is one thatis madein a very
describedthe case of a compulsive gamblerwho was narrowsphere and is strongly influenced by internal
planninga tripfrom San Franciscoto Seattleandat the and external factors, to which we shall return in a
last minute (and following an argumentwith his wife) moment.Apparently,however,people often regardthe

7
BAUMEISTER& HEATHERTON

decision as a single event thatis not to be reconsidered, accuratebecause people tend to treat the decision to
at least not untilmuchlater.Once the persondecides to abandoncontrolandindulgethemselvesas irrevocable.
start eating, drinking, smoking, having sex, venting The evidence that people acquiesce in self-regula-
emotion, spendingmoney, or assaultingsomeone, the tion failure,as well as the analysis of self-controlas an
person will often go ahead and participateactively in ongoing innerdebatethatis shapedby perceived costs
the process. andbenefits,has one more importantimplication:Cul-
There is thus an importantasymmetryin the way turalandsituationalfactorscan exert considerablesub-
many people confront internal conflicts surrounding tle influence on self-regulation.To put it anotherway,
self-regulation.Maintainingself-controlis treatedas an the pointat whichpeople lose (or abandon)self-control
ongoing process of negotiation and the fact that one is one that can be moved aroundwithin a wide gray
resisted temptationa few minutes ago does not neces- area,and so many factorscan influence self-controlby
sarily free one from facing a similar decision again. moving thatpoint.
However, abandoning self-control is treated as if it The self-regulationof violent,aggressivebehavioris
were a single decision that is not subject to further a good example.Manyviolent acts areexperiencedand
reconsideration. describedby perpetratorsas episodes of losing control.
Why do people fail to reconsidera decision to go Consistent with this, it is clear that most people are
ahead and indulge themselves, abandoningrestraint? usuallyable to preventangerfromresultingin physical
Several reasons can be suggested. The period of inde- violence. The very high contributionof alcohol to in-
cision is likely to have been one of anxiety and uncer- tensifying violent responses to provocationsis partly
taintyand,in contrast,the decision to go aheadis likely due to the fact that it underminespeople's capacity to
to be markedby relief (and often pleasure).To return regulatetheirbehavior,so they act out violent impulses
voluntarilyfrom a stateof relief andpleasureto one of more frequentlyand extremely (Bushman & Cooper,
anxiety and uncertaintywould certainly be an unap- 1990; Steele & Southwick, 1985).
pealing transition.Moreover,the unpleasantnessof the Despite the appearancethat violent behavior in-
state of denial and innerdebate would be enhancedby volves loss of control, there is evidence of acquies-
guilt or otherformsof anxietyresultingfromthe initial cence: People could control their behavior if they
indulgence. wanted to do so. Most people do stop short of lethal
As an example,one may considera hypotheticalcase violence even when they are extremely angry(Tavris,
of a dieter temptedto enjoy an appealingdessert.The 1989). Among the Malays,the patternof runningamok
phase of confrontingand resistingtemptationis proba- institutionalizeda generalbelief thatprovocationspro-
bly an unpleasantone, markedby the internaleffort of duced anger that led to uncontrollableaggression;but
self-denial and salient thoughtsof the foregone plea- when the Britishtook over and institutedsevere penal-
sure, as well as an ongoing inner debate. Finally the ties for runningamok,the practicediminishedsubstan-
person decides to go ahead and have the dessert after tially, indicating that the young men could control it
all, possibly under the influence of some available afterall (Carr& Tan, 1976). Berkowitz's (1978) study
excuse (e.g., so as not to offend the hostess). This of men in prison for violent assault in Great Britain
decision most likely bringspleasureand relief, and as contained the same mixed message. These men did
the personenjoys the first few bites, the idea of recon- apparentlylose control (often under the influence of
sidering-of returningto self-denial or even of just alcohol) and beat someone else up to their own disad-
renewingthe innerdebate about whetherone ought to vantage(hence theirimprisonment),butthey hadman-
be eatingthis-would be most unappealing.To resume aged to restrainthemselvesfrom going even farther.In
self-denialwhile halfwaythroughthe dessertwould be one memorableanecdote, one of Berkowitz's partici-
unpleasantin severalrespects:It would meanabandon- pants described a violent attack on his wife's lover
ing a very salientpleasurein orderto returnto the state duringwhich he was totallyenragedandseemingly out
of deprivation, it would require a strenuous act of of control.At one point in the attack,he took hold of a
self-regulation,and even if one succeeded in putting bottleby the neck andbrokeit off to use as a weapon-
down the spoon one would alreadyhave earnedsome but then he reconsideredthatif he used thatweaponhe
remorse (e.g., guilt or shame) because of the portion would most likely have killed the other man, which
one alreadyate. would have had serious consequences for him. As a
Self-regulationfailure can thus occur wheneverthe result, he put down the brokenbottle and resumedthe
person experiences even a very brief period in which attackwith his fists, beatingthe othermansenseless but
the costs seem to outweigh the benefits. The popular not killing him.
image in which a momentof weakness can undermine There is thus an undercurrentof control in the loss
months or years of virtuous self-denial is somewhat of control of violent behavior.At some point, people

8
SELF-REGULATION
FAILURE

allow themselves to lose control.The determinationof Misregulation


thatpoint is subjectto a greatmany subtle influences.
Theoriesaboutaggressiononce exploredthe notion We turn now to examine a very different type of
of a "subcultureof violence."Accordingto thattheory, self-regulation failure, namely misregulation. Al-
certainsubculturesplaced a positive value on aggres- thoughunderregulationmay provide the most familiar
sive behaviorand so people sought to gain esteem and and vivid instances of self-regulatoryfailure, not all
prestige by acting aggressively. This view was largely instancesfit in thatcategory.In underregulation, people
discreditedby accumulatingevidence thatviolent peo- end up being unableor unwilling to exert the requisite
ple did not apparentlyseek to win approvalor esteem control over themselves. In misregulation,however,
by violent acts (e.g., Berkowitz, 1978) and that mem- the cause of failure lies in the use to which the efforts
bersof the supposedlyviolentsubculturesdid notreport are directed.The person may even be quite successful
that they placed positive values on violent acts (see at exerting control over him- or herself but the end
Tedeschi & Felson, 1994). resultis failurebecausethe efforts aremisguidedor are
We think,however,thatthe notionof a subcultureof wasted in otherways.
violence may deserve to be reconsidered in another Ourreview of the empirical literatureyielded three
form:Subcultures(orindeedcultures)caninfluencethe main causes of misregulation: (a) misunderstood
point at which people believe it is appropriateto lose contingencies, (b) quixotic efforts to control the un-
control over aggressive impulses. Such collective be- controllable, and (c) giving too much priority to
liefs can exert considerableinfluence over the point at affect regulation. Let us examine each of these in
which people believe it is appropriate,reasonable,or turn.
even desirable to abandon self-control. Thus, many The first cause involves false beliefs about the self
assaults and homicides occur in direct response to and the world (particularlyabout the contingencies
verbalinsults, but most insults do not lead to physical between them). Well-intentioned and well-executed
violence. It takes cultural norms to prescribe which efforts at self-regulationmay end in futility because
insults, in which settings, will cause the person to they werebasedon false assumptionsaboutwhatwould
retaliate with physical aggression. Studies and inter- yield desirable results. Thus, under the influence of
views with teen gang members, for example, often inflatedegotism andemotionaldistress,people may set
reportthatthe young men and women say that violent unrealisticallyhigh goals that will increase the likeli-
retaliation is appropriateand even necessary in re- hood or costliness of failure (Baumeister,Heatherton,
sponse to certaininsults (e.g., Anderson, 1994; Bing, & Tice, 1993;Ward& Eisler, 1987;Wright& Mischel,
1991; Currie, 1991; Jankowski, 1991). Likewise, the 1982). As HeathertonandAmbady(1993) argued,peo-
American South has higher homicide rates than other ple who are prone to overly optimistic self-views may
parts of the countrybut only for homicides related to be especially vulnerableto this form of self-regulation
arguments,which suggests that Southernculturesup- failure.
ports the view thatcertainprovocationsrequireone to Unwarrantedoptimism may also cause excessive
lose control of violent, retaliatoryimpulses (Nisbett, persistence in futile endeavors and although the
1993). chances of success were minimal all along, the persis-
Indeed, moving the point at which one loses con- tence increasesthe costs (e.g., time, effort, andmoney)
trol may be a major way that a culture can influence that accompanythe failure (Rubin & Brockner,1975;
self-regulation. From our perspective, various forces Staw, 1976). Increasedfrustrationandotheremotional
in modern American culture have exerted a broad costs may result from such failures due to excessive
influence to shift this point in ways that make people persistence; indeed, in unrequitedlove, people often
more likely to abandon self-control. The pervasive- persistpastthe pointof rationalor optimalhope andthe
ness of self-regulation problems in modern America resultsof suchpersistenceincludeconsiderabledistress
may be less a result of characterflaws or deficiencies and inconvenience for both the aspiringlover and the
than a result of a social climate that encourages target (Baumeister,Wotman, & Stillwell, 1993). One
people to regardmany situations as ones in which an study showed that futile persistence is often mediated
average, reasonable person would supposedly lose by false expectations; when people were educated
control. The notion of irresistible impulses may be about common patternsof excessive, fruitless persis-
weak and dubious as a scientific hypothesis but as a tence, they were less likely to make the same mistake
social doctrine (and as a legal defense strategy) it themselves (Nathansonet al., 1982). Anothershowed
may be powerful and influential. Once it becomes that if people are encouragedto make careful calcula-
widely accepted, it is likely to operate as a self-ful- tions aboutthe probabilities,contingencies, and likely
filling prophecy. payoffs, they are less likely to fall into the trap of

9
BAUMEISTER& HEATHERTON

excessive persistence (Conlon & Wolf, 1980), which out of theirminds.Researchhas shown thatsuchefforts
also indicates that false assumptions and misguided at thoughtsuppressionareat best only partlysuccessful
expectations play a crucial role in this form of and they createstrongvulnerabilitiesto resurgencesof
misregulation. the unwantedthought (Wegner, Schneider, Carter,&
False assumptionscontributeto anotherpatternof White, 1987); indeed, efforts to suppress undesired
misregulationin the task-performancerealminvolving thoughtsmay ironicallycreatea "syntheticobsession"
speed-accuracytradeoffs.On many tasks, speed is in- with those thoughts(Wegner, 1992, 1994).
creased at the expense of accuracyand vice versa, but Performance can be impaired by this form of
the relationis far from linearand thereare many cases misregulationtoo and indeed one of the most familiar
in which reducingspeed will fail to yield greateraccu- and frustratingkinds of performancefailure-chok-
racy. Moreover, people may assume falsely that they ing under pressure-is a classic case of it. Choking,
can increase speed without substantiallosses of accu- which is defined as performing below the level of
racy. Heckhausenand Strang(1988) showed that ath- one's ability despite situational incentives and sub-
letes attemptingto achieve a recordperformanceon an jective wishes and efforts to do one's best, arises
experimental task tended to increase speed dramati- because the person consciously overrides well-
cally but the loss of accuracy outweighed the gains learned patterns of skilled response in the hope of
broughtby the increased speed. The role of false as- maximizing performance-but then finds that the
sumptionswas evident: The athletes in that study be- deliberate, controlled processes cannot perform as
lieved thatthey could maintainhigh accuracyat higher efficiently and effectively as the overlearned, auto-
speeds. matic ones (Baumeister, 1984). In a typical case, the
Misregulationcan also resultfromfalse assumptions person has achieved a level of overlearning (i.e.,
aboutemotions. Many people believe thatit is helpful skill) so that performancecan flow with a minimum
to vent theirangeror otherforms of emotionaldistress of conscious direction. However, on a particularly
butthey find thatsuchacts oftenmakethemmorerather important occasion, the pressure and desire to do
than less upset (e.g., Tavris, 1989). Affect misregula- well cause the person to want to pay special attention
tion is marked by many patternsof misregulationin and therefore to oversee the performance process
which people incorrectlyassumethatwhat worksonce consciously. This conscious oversight overrides the
or with one emotion will work with others too. Thus, automaticquality of skilled performance;sadly, con-
consuming alcohol often makes people feel good and trolled processes cannot match the automatic skills
so they may drinkas a way of self-medicatingfor their for either speed or accuracy. For example, the typist
own depression;however,they often find thatintoxica- or pianist who under pressure seeks to consciously
tion makes the depression worse rather than better monitor every finger movement quickly discovers
(Doweiko, 1990). Likewise, because socializing with that both speed and accuracy suffer.
friends is often effective at curing a sad or depressed Chokingis thus a paradigmaticinstance of this sec-
mood, people may try it to cure angry moods, but in ond form of misregulation.The person successfully
many cases they end up reciting their grievances or overrides the normal, habitual, overlearnedor auto-
problems to these friends and rekindling their own matic response but the person cannot make him- or
anger(Tice & Baumeister,1993). herself performeffectively without using those skills.
The secondgeneralpatternof misregulationinvolves The resultis thatthe personends up performingworse
the quixotic effort to controlthingsthatarebeyondthe thanusual as a directresultof efforts to performbetter
scope of potentialcontrol.Therearemanyautomaticor thanusual.
innatelypreparedprocesses thatpeople simply cannot The third broad patternof misregulation involves
alterandtheirefforts to controlthemdirectlyarelikely aiming one's self-regulatory efforts at a tangential,
to backfire. One rather clear example is that most peripheral,or irrelevantpart of the problem. Many
emotionalandmood statescannotbe altereddirectlyby problems that confront people have multiple aspects
sheer act of will (hence the pervasiveness of indirect and self-regulatoryefforts can be focused on any part
strategies for affect regulation).If people try to alter of them.If the personselects the wrong aspect of his or
theirmoods directly,they arelikely to be unsuccessful herbehaviorto regulate,the problemwill not be solved
and indeed the failure of their efforts may make them and may even get worse.
feel worse. The most commonpatternof misregulationinvolves
Thought suppression is a good example of such emphasizing (short-term)affect regulation at the ex-
quixotic misregulation. People often seem to believe pense of some other, more lasting and substantiveas-
that they can directly control their thoughts and so pect. Often a particular problem consists of both
they believe that unwanted thoughts can be driven practicalobstacles or difficulties and subjective,emo-

10
SELF-REGULATIONFAILURE

tional distress, and when people respond by focusing regulatorypatterns,it is importantto realize that the
their efforts on emotional regulationthey neglect the majority of such violations are inherently trivial. A
more fundamental,practical aspects, thereby leaving single cookie may violate a weight-loss plan, but the
the problem unsolved or even compounding it. By impactof that cookie on the diet is probablyminimal.
giving priorityto affect regulation,they allow the cause The socially importantinstancesof self-regulationfail-
of the problemto get worse and so in the long runthey ure tend to involve large-scale breakdowns such as
end up worse off. Oftenthey end up feeling worse even binges. To be sure, a binge may begin with a single
thoughaffect regulationwas theirtop priority. lapse, but to understandthe lapse is not sufficient to
This form of misregulation can be seen in some explain the binge.
patternsof procrastination.A personmayhavea project Our review concluded that in many cases a second
deadline but working on the project causes anxiety, and importantset of causes of self-regulation failure
possibly because the projectis importantand because only entersthe pictureafteran initial lapse, and indeed
the person wants to do very well. Puttingoff working as a resultof thatlapse.We use the termlapse-activated
on theprojectthusbecomes aneffective meansof affect causes to describe these factors. This concept was
regulationin the short run because one escapes from anticipatedin addiction research by Marlatt (1985),
anxiety each time one elects not to work on the task; whose term abstinence violation effect referredto the
the cumulativeeffect of such decisions makestheprob- tendencyfor people to respondto an initial indulgence
lem considerably worse because the time until the in alcoholor otheraddictivebutforbiddensubstanceby
deadline grows shorter,making it ever harderto do a consuming more. The category of lapse-activatedre-
good job. As the deadline looms, the panic response sponses includesabstinenceviolationeffects as well as
becomes ever betterjustified. other, conceptually similar patternsthat are not con-
Giving top priorityto affect regulationmay also be cernedwith abstinence.
a factor behind many destructive patterns of failed An early clear demonstration of lapse-activated
impulse control. Many consummatoryresponses are misregulationwas by Hermanand Mack (1975), who
affectively pleasantand so people will indulge in them termedtheir effects counterregulatoryeating. In their
as a way of regulating their emotions. People may study,dieterswho had been preloadedwith food actu-
smoke cigarettes, abuse alcohol, take drugs, go on ally went on to eat more thandieters who had not had
shoppingsprees,engage in promiscuoussex, or gamble such a preload, contraryto what nondieters do (and
away their money as a way of escaping from a bad what common sense would prescribe).Subsequentre-
mood, butthe consequencesof such actionscan be even searchhas demonstratedthata person's beliefs are the
worse thanwhatcaused the badmood in the firstplace. primarydeterminantsof this disinhibited eating. For
Thus, eating or drinkingbinges may occur becausethe instance, dieters will engage in counterregulatoryeat-
person thinks that eating or drinkingwill remedy the ing when they have eaten very small amountsof per-
emotional distress. Shilts (1987) cited some survey ceived high-calorie foods (e.g., a small bite of
evidence that when the AIDS epidemic was first chocolate)butwill be able to maintaintheirdiets if they
spreading,many gay men became distraughtand upset believe they have not brokentheir diets (even if they
over the dangerand respondedby going out to engage have consumed an incrediblyfatty Caesar salad). The
in promiscuous,unprotectedmale-male sex to get their dieter's initial minor transgression leads to such
minds off those stressful thoughts. Although that re- thoughtsas, "Whatthe hell, I have blown it, so I may
sponse may have been effective as self-distraction,it as well eat the whole darnthing."The irony is thatthe
tendedto increasethe underlyingproblem. small amountof fatteningfood in the initial lapse does
Thus, the category of misregulationencompasses not constitute a serious threatto the dieter's goal of
several forms of the misuse or ineffective use of self- weight loss-but the subsequent binge eating does
control.People may fail at self-regulationbecausethey sabotagethatgoal.
are tryingto controlthe wrong aspect of the process or Marlatt(1985) documentedthis lapse-activatedpat-
because they are trying to control something that is tern across a number of addictive and problematic
essentially immune to control. False beliefs and as- behaviors,including smoking, alcoholism, and heroin
sumptionsaboutthe contingenciesbetween one's own addiction.Marlatt'smodel suggests that lapses often
acts and one's outcomes often play an importantrole. arise in high-risk situations in which a person has
difficulty coping. Marlattargues that a lapse becomes
Lapse-Activated Responses a relapse largely because of the person's commitment
to complete and absolute abstinence. Performingthe
Althoughconsiderableresearchhas focusedon what forbiddenbehaviorleads to unpleasantdissonance and
causes people to violate their standardsor other self- self-attributionsof weakness and failure. The attribu-

11
BAUMEISTER& HEATHERTON

tion of failureto the self diminishesthe person's sense of skilled performance (i.e., they may choke). The
of self-controland he or she abandonsattemptsto rein result of this impairmentmay be to increase the pres-
in subsequentbehavior.Thus, a minortransgressionis sure on them to perform well so as to overcome the
seen as a catastropheratherthan a small slip and this problemscausedby the initialchoking.As the pressure
perceptioninduces the person to abdicateall self-con- increases,they may choke even more. Schlenker,Phil-
trol. Marlatt'sresearchhas led to a therapyknown as lips, Boniecki, andSchlenker(1995) showed thathome
relapseprevention,which consists primarilyof cogni- teams in championshipfinal baseball games tend to
tive restructuringto help the addictcope with high-risk make errorswhen they fall behind,presumablyin part
situationsand with lapses. becausethey aretryingto overcometheirinitialdeficit.
For ourpurposes,the key point is thatseveralcausal Although more systematicdata are needed, the recent
factors come into play as a result of an initial lapse in SuperBowl games have providedvivid illustrationsof
self-controland these can undermineself-controlsub- such spiralingfailures,as the Buffalo teams have made
sequently.Moreover,it is often the subsequentbreak- moreandmoremistakesonce they beganto fall behind.
down in self-control that has the most severe and Likewise, test anxiety seems to conform to the pattern
disastrous results. There are several mechanisms of in which the personbecomes preoccupiedby ruminat-
lapse-activatedpatterns,as follows. ing over an initial failure (to know an answer) and
One important mechanism is that people may because of this preoccupationbecomes unableto con-
cease monitoring themselves after an initial lapse, centrateon subsequentquestions(see Wine, 1971).
possibly because it would be distressing to attendto Destructive patterns of persistence also have ele-
their behavior when they have already failed to live mentsof lapse-activatedcausality.In manycases, peo-
up to standards,and possibly because the initial lapse ple must invest time and energy as well as other
may provide such pleasure or intense sensation that resources(e.g., money or prestige)in some decision. If
they focus narrowly on it (i.e., loss of transcen- it goes bad,people arereluctantto cut theirlosses, and
dence). Polivy (1976) showed that dieters who had indeed the more they invest the more difficult it be-
been preloaded with food apparentlyceased to keep comes for them to acceptthatcourse of action is futile,
track of how much they ate, as indicated by errorsin and so the eventual losses continue to mount (e.g.,
retrospective self-reports of subsequent consump- Bazerman,Giuliano,& Appelman, 1984; Staw, 1976).
tion. More generally, eating binges seem to be In Teger's (1980) phrase, people become "too much
markedby an immersion in sensation and a cessation invested to quit" and so they invest-and lose-con-
of monitoring one's own behavior (Heatherton & siderablymore.
Baumeister, 1991). There are of course also interpersonalaspects to
Spiralingpatternsof distress may also be a form of manyself-controlsituations,andthese can be activated
lapse-activated causes of misregulation. An initial by lapses so as to contributeto escalating failures of
lapse may occur becausethe personwas sufferingfrom self-regulation. The most obvious example would
some form of distress.The lapse may, however,gener- probablyinvolve violent episodes. An initial aggres-
ate guilt, fear, anxiety, or other forms of distress, sive outburstmay be a momentarylapse in self-control
therebymaking the person feel worse. The escalating butit mayhavelastingeffects if someoneelse is harmed
distress may contributeto a furtherabandonmentof orprovoked.An aggressiveresponseby others(or even
self-control. the formation by others of an expectation that the
Distress is of course not the only emotionthatcan be individual is prone to violent outbursts)may lead to
activated by a lapse and contributeto furtherbreak- furtherviolence.
downs in self-regulation. Lawson (1988) noted that Ironically,some factorsthataid self-regulationup to
manypeople will initiallyengagein extramaritalsex on the point of an initial lapse may turn into factors that
the assumptionthatit will be a casual, isolatedepisode produce misregulationas the result of such a lapse.
that will not affect or threatentheir marriage.Some Most prominentamong these are zero-tolerancebe-
find, however, that they begin to experience love or liefs. Suchbeliefs, which arecommonin some spheres,
otherformsof intimateattachmentto theirillicit partner catastrophizethe initiallapse as a way of preventingit.
and these feelings may cause the extramaritalinvolve- People are encouragedto believe that having a single
ment to escalate, even to the point at which it does drink,committinga single sexual indiscretion,or tak-
become a threatto the marriage. ing a single dose of a drugon one occasion will lead to
As we noted, some lapse-activatedpatterns have disaster (see also Marlatt, 1985). Undoubtedly such
little to do with abstinence violations. Performance beliefs discourage people from allowing a lapse to
effects may provide one instance. Under pressureto happen. If a lapse does occur, however, such beliefs
performwell, people may experiencesome impairment may help producelapse-activatedincreases in the un-

12
SELF-REGULATIONFAILURE

wantedbehavior.The person may feel that a catastro- would be excessive to say thatpeople freely choose to
phe has occurred and that there is no use in making lose control,they do seem to show considerableactive
furtherefforts at self-control.Alternatively,the person participationand acquiescence in the behaviors that
may find that the predictedcatastrophicconsequences constitute self-regulatoryfailure. We suggested that
have not materializedand conclude that the fears and self-regulationoften involves an unpleasantinnercon-
warningswere entirely unfounded.Zero-tolerancebe- flict markedby competing wishes and uncertainty.If
liefs can be comparedto a militarystrategyof putting the person decides even briefly to relax self-control,
all troopsin the frontline, which will indeedstrengthen typically he or she will not consider reinstatingit and
the front line but will leave the army with no reserves so a brief abdicationof self-regulatoryeffort can lead
to use if the frontline is breached. to a serious,protractedbreakdown.In colloquialterms,
the popularimage of a moment of weakness is more
Conclusion accurate than the image of the irresistible impulse.
Moreover,culturecan exert considerableinfluence by
Self-regulation is a complex mechanism that can teachingpeople whichcircumstancesmakeit appropri-
break down in many different ways. Underregulation ate to abandoncontrol.
occurs because people lack stable, clear, consistent Unfortunately,the norms and forces that currently
standards,becausethey fail to monitortheiractions,or dominatemodernWesterncultureseem generallycon-
becausethey lack the strengthto overridethe responses ducive to weakeningself-control.As long as this is the
they wish to control.Misregulationoccursbecausethey case, it seems likely that our society will continue to
operateon the basis of false assumptionsaboutthem- suffer from widespreadand even epidemic problems
selves and aboutthe world, because they try to control thathave self-regulatoryfailureas a common core.
things that cannot be directly controlled, or because
they give priorityto emotions while neglecting more
importantand fundamentalproblems. Note
We have proposedthat the evidence aboutself-reg-
ulatoryfailures conforms to a strengthmodel; that is, Roy F. Baumeister,Departmentof Psychology, Case
the capacityto regulateoneself is a limited, renewable Western Reserve University, Cleveland, OH 44106-
resource. When stress or fatigue depletes an 7123.
individual's strength,self-regulatoryfailures become
morelikely. Capacitiesfor self-controlareanimportant References
realmof stable, long-termindividualdifferences.
Anderson, E. (1994). The code of the streets. Atlantic Monthly,
The control of attentionis centralto self-regulation 273(5), 81-94.
andloss of attentionalcontrolis a decisive precursorof Bandura, A. (1977). Self-efficacy: Toward a unifying theory of
many forms of self-regulation failure. In particular, behavioralchange.Psychological Review,84, 191-215.
effective self-regulationoftenrequiresthe individualto Baumeister,R. F. (1984). Choking underpressure:Self-conscious-
ness and paradoxicaleffects of incentives on skillful perfor-
be able to transcendthe immediatesituationby consid-
mance. Journal of Personality and Social Psychology, 46,
ering long-termconsequencesandimplications.When 610-620.
transcendenceis weakenedby anythingthat binds at- Baumeister, R. F. (1995). Transcendence,guilt, and self-control.
tentionto the here and now, the chancesof self-regula- Behavioraland Brain Sciences, 18, 122-123.
tion failure are increased. Baumeister,R. F., Heatherton,T. F., & Tice, D. M. (1993). When
Many spheres of self-regulationfailure show signs ego threats lead to self-regulation failure: Negative conse-
quences of high self-esteem. Journal of Personalityand Social
of lapse-activatedcauses. That is, an initial and seem- Psychology,64, 141-156.
ingly minorbreakdownin self-controlmay set off other Baumeister,R. F., Heatherton,T. F., & Tice, D. M. (1994). Losing
causes and factors that preventthe reassertionof self- control:How and whypeoplefail at self-regulation.San Diego:
control and cause the breakdownto snowball. Indeed, Academic.
the initial lapse may often be trivial,whereasthe binge Baumeister, R. F., Stillwell, A. M., & Heatherton,T. F. (1995).
Personalnarrativesaboutguilt:Role in actioncontrolandinter-
is catastrophic,andso these lapse-activatedfactorsthat
personalrelationships.Basic and Applied Social Psychology,
producea snowballingeffect are what deserve empha- 17, 173-198.
sis in theory,research,and intervention. Baumeister,R. F., Wotman,S. R., & Stillwell, A. M. (1993). Unre-
The degree of volition and acquiescencein self-reg- quited love: On heartbreak,anger, guilt, scriptlessness, and
humiliation.Journalof Personalityand Social Psychology, 64,
ulatoryfailureis a controversialissue with implications
377-394.
that go far beyond psychology. Ourreview has led us Bazerman,M. H., Giuliano,T., & Appelman,A. (1984). Escalation
to reject the model that self-regulatoryfailure is typi- of commitmentin individualand group decision making. Or-
cally the result of irresistible impulses. Although it ganizationalBehaviorand HumanPerformance,33, 141-152.

13
BAUMEISTER& HEATHERTON

Berkowitz,L. (1978). Is criminalviolence normativebehavior?Hos- Social Psychology,45, 935-942.


tile and instrumentalaggressionin violent incidents.Journalof Kirschenbaum,D. S. (1987). Self-regulatoryfailure:A review with
Research in Crimeand Delinquency,15, 148-161. clinical implications.Clinical Psychology Review, 7, 77-104.
Berkowitz,L. (1989). Frustration-aggressionhypothesis:Examina- Lawson, A. (1988). Adultery:An analysis of love and betrayal. New
tion and reformulation.AmericanPsychologist, 106, 59-73. York:Basic.
Bing, L. (1991). Do or die. New York:HarperCollins. Marlatt,G. A. (1985). Relapse prevention:Theoreticalrationaleand
Burawoy,M. (1979). Manufacturingconsent. Chicago:Universityof overview of the model. In G. A. Marlatt& J. R. Gordon(Eds.),
Chicago Press. Relapseprevention(pp. 3-70). New York:Guilford.
Bushman,B. J., & Cooper,H. M. (1990). Effects of alcohol on human Mischel, W. (1974). Processes in delay of gratification. In L.
aggression:An integrativeresearchreview. Psychological Bul- Berkowitz (Ed.), Advances in experimentalsocial psychology
letin, 107, 341-354. (Vol. 7, pp. 249-292). San Diego: Academic.
Carr,J. E., & Tan, E. K. (1976) In searchof the trueAmok:Amok as Mischel, W. (in press). From good intentions to willpower. In P.
viewed within the Malay culture.AmericanJournalof Psychi- Gollwitzer & J. Bargh (Eds.), The psychology of action. New
atry, 133, 1295-1299. York:Guilford.
Carver,C. S. (1979). A cyberneticmodel of self-attentionprocesses. Mischel, W., Shoda, Y., & Peake, P. K. (1988). The nature of
Journal of Personalityand Social Psychology, 37, 1251-1281. adolescentcompetenciespredictedby preschool delay of grati-
Carver,C. S., & Scheier,M. F. (1981). Attentionandself-regulation: fication. Journal of Personality and Social Psychology, 54,
A control theory approach to human behavior. New York: 687-696.
Springer-Verlag. Nathanson,S., Brockner,J., Brenner,D., Samuelson, C., Country-
Carver, C. S., & Scheier, M. F. (1982). Control theory: A useful man,M., Lloyd, M., & Rubin,J. Z. (1982). Towardthe reduction
conceptualframeworkforpersonality-social,clinical andhealth of entrapment.Journalof AppliedSocial Psychology, 12, 193-
psychology. Psychological Bulletin, 92, 111-135. 208.
Conlon,E. J., & Wolf, G. (1980). The moderatingeffects of strategy, Nisbett, R. E. (1993). Violence and U.S. regional culture.American
visibility, andinvolvementon allocationbehavior:An extension Psychologist,48, 441-449.
of Staw's escalation paradigm.OrganizationalBehavior and Polivy, J. (1976). Perceptionof calories and regulationof intake in
HumanPerformance,26, 172-192. restrainedand unrestrainedsubjects. Addictive Behaviors, 1,
Currie,E. (1991). Dope and trouble:Portraitsof delinquentyouth. 237-243.
New York:Pantheon. Prochaska,J. O., & DiClemente, C. C. (1984). The transtheoretical
Doweiko, H. E. (1990). Concepts of chemical dependency.Pacific approach: Crossing traditionalboundaries of change. Home-
Grove, CA: Brooks/Cole. wood, IL: Irwin.
Emmons, R. A., & King, L. A. (1988). Conflict among personal Prochaska,J. O., & DiClemente,C. C. (1986). Towardacomprehens-
strivings:Immediateandlong-termimplicationsfor psycholog- ive model of change. In W. Miller and N. Heather (Eds.),
ical and physical well-being. Journalof Personalityand Social Treatingaddictivebehaviors: Processes of change (pp. 3-27).
Psychology, 54, 1040-1048. New York:Plenum.
Funder,D. C., Block, J. H., & Block, J. (1983). Delay of gratification: Rachlin, H. (1995). Self-control: Beyond commitment.Behavioral
Some longitudinalpersonalitycorrelates.Journalof Personality and Brain Sciences, 18, 109-121.
and Social Psychology, 44, 1198-1213. Rodriguez, M. L., Mischel, W., & Shoda, Y. (1989). Cognitive
Glass, D. C., Singer, J. E., & Friedman,L. N. (1969). Psychic cost of person variables in the delay of gratification of older children
adaptationto an environmentalstressor.Journalof Personality at risk. Journal of Personality and Social Psychology, 57,
and Social Psychology, 12, 200-210. 358-367.
Gottfredson,M. R., & Hirschi,T. (1990). A general theoryof crime. Rubin, J. Z., & Brockner, J. (1975). Factors affecting entrapment
Stanford,CA: StanfordUniversityPress. in waiting situations: The Rosencrantz and Guildensternef-
Heatherton,T. F., & Ambady, N. (1993). Self-esteem, self-predic- fect. Journalof Personality and Social Psychology, 31, 1054-
tion, and living up to commitments. In R. Baumeister(Ed.), 1063.
Self-esteem:Thepuzzle of low self-regard(pp. 131-145). New Sansone, C., Weir, C., Harpster,L., & Morgan, C. (1992). Once a
York:Plenum. boring task, always a boring task? Interestas a self-regulatory
Heatherton,T. F., & Baumeister,R. F. (1991). Binge eatingas escape mechanism.Journalof Personalityand Social Psychology, 63,
from self-awareness.Psychological Bulletin, 110, 86-108. 379-390.
Heckhausen,H., & Strang,H. (1988). Efficiency underrecordper- Schachter,S. (1971). Some extraordinaryfacts aboutobese humans
formancedemands:Exertioncontrol-An individualdifference and rats.AmericanPsychologist, 26, 129-144.
variable? Journal of Personality and Social Psychology, 55, Schachter,S. (1982). Recidivism and self-cure of smoking and obe-
489-498. sity. AmericanPsychologist, 37, 436-444.
Herman, C. P., & Mack, D. (1975). Restrainedand unrestrained Schlenker,B. R., Phillips, S. T., Boniecki, K. A., & Schlenker,D. R.
eating. Journalof Personality,43, 647-660. (1995). Championshippressures: Choking or triumphing in
Hull, J. G. (1981). A self-awarenessmodel of the causes andeffects one's own territory?Journalof Personalityand Social Psychol-
of alcohol consumption.Journalof AbnormalPsychology, 90, ogy, 68, 623-643.
586-600. Shilts, R. (1987). And the bandplayed on: Politics, people, and the
James,W. (1950). Theprinciples of psychology (Vol 2). New York: AIDSepidemic.New York:Viking.
Dover. (Originalwork published1890) Shoda,Y., Mischel, W., & Peake,P. K. (1990). Predictingadolescent
Jankowski,M. S. (1991). Islands in the street: Gangs and American cognitive and self-regulatory competencies from preschool
urbansociety. Berkeley: Universityof CaliforniaPress. delay of gratification:Identifyingdiagnosticconditions.Devel-
Kanfer, F. H., & Karoly, P. (1972). Self-control: A behavioristic opmentalPsychology, 26, 978-986.
excursion into the lion's den. BehaviorTherapy,3, 398-416. Staw, B. M. (1976). Knee-deepin the big muddy:A studyof escalat-
Karniol,R., & Miller,D. T. (1983). Why notwait?:A cognitivemodel ing commitmentto a chosen course of action. Organizational
of self-imposed delay termination.Journal of Personalityand Behaviorand HumanPerformance,16, 27-44.

14
SELF-REGULATIONFAILURE

Steele, C. M., & Josephs, R. A. (1990). Alcohol myopia: Its prized the self-concept:Motivationalconsequencesof discrepantself-
and dangerouseffects. AmericanPsychologist,45, 921-933. guides.Journalof Personalityand Social Psychology, 55, 625-
Steele, C. M., & Southwick, L. (1985). Alcohol and social behavior 633.
I:Thepsychology of drunkenexcess. Journalof Personalityand Ward,C. H., & Eisler, R. M. (1987). Type A behavior,achievement
Social Psychology, 48, 18-34. striving,and a dysfunctionalself-evaluationsystem. Journal of
Tavris, C. (1989). Anger: The misunderstoodemotion. New York: Personalityand Social Psychology, 53, 318-326.
Simon & Schuster. Wegner, D. M. (1992). You can't always think what you want:
Tedeschi, J. T., & Felson, R. B. (1994). Violence,aggression, and Problemsin the suppressionof unwantedthoughts.In M. Zanna
coercive actions. Washington, DC: American Psychological (Ed.),Advancesin experimentalsocial psychology (Vol. 25, pp.
Association. 193-225). San Diego, CA: Academic.
Teger, A. I. (1980) Too much invested to quit. New York: Per- Wegner, D. M. (1994). Ironicprocesses of mental control. Psycho-
gamon. logical Review, 101, 34-52.
Tice, D. M., & Baumeister,R. F. (1993). Controllinganger:Self-in- Wegner,D. M., Schneider,D. J., Carter,S. R., & White, T. L. (1987).
duced emotion change. In D. M. Wegner & J. W. Pennebaker Paradoxicaleffects of thoughtsuppression.Journal of Person-
(Eds.), Handbookof mentalcontrol (pp. 393-409). Englewood ality and Social Psychology,53, 5-13.
Cliffs, NJ: Prentice-Hall. Wine, J. (1971). Test anxiety and directionof attention.Psychologi-
Vallacher, R. R., & Wegner, D. M. (1985). A theory of action cal Bulletin, 76, 92-104.
identification. Hillsdale, NJ: Lawrence Erlbaum Associates, Wright,J., & Mischel, W. (1982). Influence of affect on cognitive
Inc. social learning person variables. Journal of Personality and
Van Hook, E., & Higgins, E. T. (1988). Self-relatedproblemsbeyond Social Psychology,43, 901-914.

15

You might also like