Beyond the Hedonic TreadmillRevising the Adaptation Theory of Well-Being
Ed Diener University of Illinois at Urbana–ChampaignRichard E. Lucas Michigan State University
Christie Napa Scollon Texas Christian University
According to the hedonic treadmill model, good and badevents temporarily affect happiness, but people quicklyadapt back to hedonic neutrality. The theory, which hasgained widespread acceptance in recent years, implies thatindividual and societal efforts to increase happiness aredoomed to failure. The recent empirical work outlined hereindicates that 5 important revisions to the treadmill modelare needed. First, individuals’ set points are not hedoni-cally neutral. Second, people have different set points,which are partly dependent on their temperaments. Third,a single person may have multiple happiness set points:Different components of well-being such as pleasant emo-tions, unpleasant emotions, and life satisfaction can movein different directions. Fourth, and perhaps most impor-tant, well-being set points can change under some condi-tions. Finally, individuals differ in their adaptation toevents, with some individuals changing their set point andothers not changing in reaction to some external event.These revisions offer hope for psychologists and policy-makers who aim to decrease human misery and increasehappiness.
Keywords: coping, subjective well-being, life satisfaction,adaptation, happiness
Imagine a world in which the poorest diseased beggarwith no family or friends is as happy as the healthybillionaire who has a surfeit of close and supportive
relationships. Or imagine that individuals living in a crueldictatorship where crime, slavery, and inequality are ram-pant are as satisfied with their lives as people living in astable democracy where crime is minimal. Finally, imaginethat no matter how much effort and care someone put intobeing happy, the long-term effects were no different than ifhe or she lived a profligate and dissolute life. Implausible?These surprising visions are based on a widely acceptedmodel of subjective well-being. Brickman and Campbell(1971) described a hedonic treadmill, in which processessimilar to sensory adaptation occur when people experienceemotional reactions to life events. Just as people’s nosesquickly adapt to many scents and smells thereafter disap-pear from awareness, Brickman and Campbell suggestedthat one’s emotion system adjusts to one’s current lifecirc*mstances and that all reactions are relative to one’sprior experience. Myers described adaptation as a key tounderstanding happiness. In his popular book The Pursuit
of Happiness, David Myers (1992) wrote, “The point can-not be overstated: Every desirable experience—passionatelove, a spiritual high, the pleasure of a new possession, theexhilaration of success—is transitory” (p. 53).
In the original treadmill theory, Brickman and Camp-bell (1971) proposed that people briefly react to good andbad events, but in a short time they return to neutrality.Thus, happiness and unhappiness are merely short-livedreactions to changes in people’s circ*mstances. Peoplecontinue to pursue happiness because they incorrectly be-lieve that greater happiness lies just around the corner inthe next goal accomplished, the next social relationshipobtained, or the next problem solved. Because new goalscontinually capture one’s attention, one constantly strivesto be happy without realizing that in the long run suchefforts are futile.
The hedonic treadmill theory is built on an automatichabituation model in which psychological systems react todeviations from one’s current adaptation level (Helson,1948, 1964). Automatic habituation processes are adaptivebecause they allow constant stimuli to fade into the back-ground. Thus, resources remain available to deal with novelstimuli, which are most likely to require immediate atten-tion (Fredrick & Loewenstein, 1999). The happiness sys-tem is thus hypothesized to reflect changes in circum-stances rather than the overall desirability of thecirc*mstances themselves. This idea was formalized byCarver and Scheier (1990), who maintained that emotionsdepend on the rate of change of important circ*mstances.
In 1978, Brickman, Coates, and Janoff-Bulman of-fered initial empirical support for the treadmill model. Instudies that have become classics in the field, Brickman etal. concluded that lottery winners were not happier thannonwinners and that people with paraplegia were not sub-stantially less happy than those who can walk. Althoughthe empirical support for hedonic adaptation was, in fact,mixed, the studies captured the attention of psychologists.The idea of hedonic adaptation was appealing because it
Ed Diener, Department of Psychology, University of Illinois at Urbana–Champaign; Richard E. Lucas, Department of Psychology, Michigan StateUniversity; Christie Napa Scollon, Department of Psychology, TexasChristian University.
Correspondence concerning this article should be addressed to EdDiener, Department of Psychology, University of Illinois, 603 East DanielStreet, Champaign, IL 61820. E-mail: [emailprotected]
305May–June 2006 ● American PsychologistCopyright 2006 by the American Psychological Association 0003-066X/06/$12.00Vol. 61, No. 4, 305–314 DOI: 10.1037/0003-066X.61.4.305
offered an explanation for the observation that people ap-pear to be relatively stable in happiness despite changes infortune. In addition, the treadmill theory explained theobservation that people with substantial resources aresometimes no happier than those with few resources andthat people with severe problems are sometimes quitehappy. Thus, the research of Brickman and colleaguesbecame central to the way many scientists understandhappiness.
We and many other psychologists readily accepted thetheory of adaptation because evidence frequently supportedthe idea. External conditions are often weak correlates ofreports of happiness. For instance, all demographic vari-ables taken together predict less than 20% of the variancein happiness (Campbell, Converse, & Rodgers, 1976). Die-ner, Sandvik, Seidlitz, and Diener (1993) found that in-come and happiness in the United States correlate only .13,and Diener, Wolsic, and Fujita (1995) similarly found thatobjective physical attractiveness correlated at very lowlevels with reports of well-being. Perhaps even more sur-prising, Okun and George (1984) found that objectivehealth on average correlated only .08 with happiness, andFeinman (1978) found that people who were blind did notdiffer in happiness from those who were able to see.
In addition, longitudinal studies that tracked changesin happiness over time provided more direct evidence thatadaptation can occur. For instance, Silver (1982) found thatindividuals with spinal cord injuries reported strong nega-tive emotions one week after their crippling accident. How-ever, two months later, happiness was their strongest emo-tion. Similarly, Suh, Diener, and Fujita (1996) found thatgood and bad life events affected happiness only if theyoccurred in the past two months. More distant past eventsdid not predict happiness (although many of the events they
studied were relatively mundane). Furthermore, in a num-ber of studies, researchers have traced reactions to thedeath of a spouse, and these studies show that emotionalreactions eventually rebound after this major life event(e.g., Bonanno et al., 2002; Bonanno, Wortman, & Nesse,2004; Lucas, Clark, Georgellis, & Diener, 2003). Thus,parts of the hedonic treadmill model have received robustempirical support (see Fredrick & Loewenstein, 1999, for areview).
Our Research on AdaptationOver the last decade, we and others have tested predictionsderived from the treadmill theory, and our findings suggestthat the model requires important modifications. Althoughthe revisions leave certain core features of the adaptationmodel intact, our research reveals that the idea is in need ofan update. After reviewing these revisions, we describe theimportant implications that they have for psychology.
Revision 1: Nonneutral Set PointsThe original treadmill theory suggested that people returnto a neutral set point after an emotionally significant event.However, decades of research show that this part of thehedonic treadmill theory is wrong. Instead, most people arehappy most of the time (Diener & Diener, 1996). Forinstance, Diener and Diener reviewed studies using a va-riety of methods of assessment, and they concluded thatapproximately three quarters of the samples they investi-gated reported affect balance scores (positive moods andemotions � negative moods and emotions) above neutral.Similarly, Biswas-Diener, Vittersø, and Diener (2005)found that even in such diverse populations as the Amish,African Maasai, and Greenlandic Inughuit, most people areabove neutral in well-being. In the most recent World
Ed Diener
Richard E.Lucas
306 May–June 2006 ● American Psychologist
Values Survey (a large-scale survey in which the nationswith the largest populations are sampled using probabilitymethods; European Values Study Group & World ValuesSurvey Association, 2005), 80% of respondents said thatthey were very or quite happy. Thus, if people adapt andreturn to a baseline, it is a positive rather than neutral one.
A general tendency to experience positive emotionsmay provide the motivation to explore one’s environmentand to approach new goals (Fredrickson, 1998). Lyubomir-sky, King, and Diener (2005) showed that positive moodsfacilitate a variety of approach behaviors and positive out-comes. Thus, the ubiquity of a positive emotional set point,in concert with the less frequent experience of unpleasantemotions, likely results from the adaptive nature of fre-quent positive emotions.
Revision 2: Individual Set PointsThe empirical research that has been conducted since thepublication of Brickman and Campbell (1971) reveals thatif people do have set points, they vary considerably acrossindividuals. These individual differences are due, at least inpart, to inborn, personality-based influences (Diener &Lucas, 1999). Support for this view comes from at leastthree different lines of research. First, research consistentlyshows that one’s level of well-being is reasonably stableover time (e.g., Eid & Diener, 2004). Second, behavioralgenetic studies show that well-being is moderately herita-ble. For instance, Tellegen et al. (1988) found that identicaltwins reared apart were much more similar in their levels ofwell-being than were dizygotic twins who were rearedapart. Finally, research shows that personality factors arestrong correlates of well-being variables. Whereas any sin-gle demographic factor typically correlates less than .20(usually much less) with well-being reports, both self- and
non–self-report measures of personality tend to correlatemuch more strongly with well-being (see Diener & Lucas,1999, for a review). Thus, personality factors may predis-pose individuals to experience different levels ofwell-being.
Revision 3: Multiple Set PointsThe idea of a happiness set point implies that well-being isa single entity with a single baseline. However, work byLucas, Diener, and Suh (1996) indicates that the globalcategory of happiness is composed of separable well-beingvariables. It is important to note that these variables some-times move in different directions over time. Thus, the ideaof a unitary set point is not tenable, because positive andnegative emotions might both decline in tandem or lifesatisfaction might move upward while positive emotionsdecrease.
In Figure 1, we present age trends in positive affect,negative affect, and life satisfaction from the first wave ofthe Victoria Quality of Life Panel Study (see Headey &Wearing, 1989, 1992; Scollon, 2004) and from the 1990World Value Survey (Inglehart & Klingemann, 2000).Both studies are based on probability samples, the formerfrom the state of Victoria in Australia and the latter from 42nations around the world. As can be seen in this figure,there were significant age effects for all five variables (allps � .001). However, these effects varied for the differentwell-being variables. For instance, at the same point in thelife span that positive affect was declining (representing adecrease in overall well-being), negative affect also de-clined (representing an increase in overall well-being).During this same period, both work and home satisfactionincreased. These data indicate that (a) there is not a con-stant global happiness set point that remains stable over thecourse of the entire life span and (b) “happiness” is not aunitary concept with a single set point to which peopleadapt. Instead, these findings suggest that different forms ofwell-being can move in different directions (also see East-erlin, 2005).
We also used the longitudinal component of the Vic-toria Quality of Life Panel Study to examine change inwell-being within persons over time. Specifically, we mod-eled change in work satisfaction and marital satisfactionover an eight-year period using growth curve modeling.Significant individual differences in change emerged onboth variables, indicating that different people changed atdifferent rates and in different directions (Scollon & Die-ner, 2005). It is important to note that the correlationbetween changes in the two variables was substantially lessthan 1 (r � .48), even at the latent level. This shows thatthe two variables do not always change in unison. Not allindividuals who increased in work satisfaction increased inmarital satisfaction. At best, only one quarter of the vari-ance in change could be accounted for by the correspond-ing amounts of change in another variable. Thus, not onlydo the various well-being components change in differentways over the course of the life span, but changes in onedomain do not fully correspond to changes in otherdomains.
Christie NapaScollon
307May–June 2006 ● American Psychologist
As a final test of the separability of well-being com-ponents, we examined the stability of positive and negativeaffect over time in the Victoria Quality of Life Panel Study.We found that, consistent with the idea that there is nosingle set point, the various components exhibited differ-ential stability. Specifically, long-term levels of negativeaffect were substantially more stable than were long-termlevels of positive affect. In addition, the stability of positiveaffect and life satisfaction declined with longer time peri-ods, whereas the stability of negative affect did not (seeTable 1). These findings suggest that stable individualbaselines might be more characteristic of negative affectthan positive affect. However, over a period of a few years,life satisfaction was most stable.
Revision 4: Happiness Can ChangePerhaps the most controversial aspect of Brickman andCampbell’s (1971) hedonic treadmill model is the idea thatpeople cannot do much to change their long-term levels ofhappiness and life satisfaction. If the hedonic treadmillmodel is correct, adaptation is inevitable, and no change inlife circ*mstance should ever lead to lasting changes inhappiness. Although the work cited at the beginning of thisarticle was suggestive of such an effect, until recently verylittle evidence has been available to provide longitudinaltests of this hypothesis. Thus, questions have remainedabout the extent to which important life events can perma-nently alter individuals’ happiness set points.
One type of evidence demonstrating that life circum-stances matter comes from well-being differences acrossnations. If there are strong national differences in well-being and these differences can be predicted from objectivecharacteristics of those nations, then this would suggestthat the stable external circ*mstances that vary across na-tions have a lasting impact on happiness. The first columnof Table 2 presents affect balance scores (reported between1981 and 1984) for several nations that differed markedlyin affluence and human rights. The right column of Table 2
Figure 1Age Trends in Subjective Well-Being
Note. WVS � 1990 World Value Survey; VQOL � first wave of the Victoria Quality of Life Panel Study.
Table 1Stability of Subjective Well-Being Measures in theVictoria Quality of Life Panel Study and the GermanSocio-Economic Panel Study
Time period betweenmeasurements
Positiveaffecta
Negativeaffecta
Lifesatisfactiona
Lifesatisfactionb
2 years .37 .44 .61 .514 years .32 .40 .50 .456 years .32 .42 .44 .418 years .23 .48 .43 .37
a Data are from the Victoria Quality of Life Panel Study. b Data are from theGerman Socio-Economic Panel Study.
308 May–June 2006 ● American Psychologist
shows that these nations also differed in life satisfaction.Because the objective conditions in these countries re-mained consistent for many years, the cross-national dif-ferences in happiness suggest that people do not alwayscompletely adapt to conditions. Perhaps more important,these mean-level differences can be predicted from objec-tive characteristics of the nations. For instance, Diener,Diener, and Diener (1995) found that the wealth and thehuman rights of nations were strong predictors of averagenational well-being. Similarly, researchers at The Econo-mist found that 85% of the variance in national levels ofwell-being could be explained by nine objective character-istics, including gross domestic product per person, lifeexpectancy at birth, political stability, and divorce rates(Economist Intelligence Unit, 2005). Furthermore, if peo-ple adapted to conditions, only change in conditions andnot the long-term level of conditions would influence feel-ings of well-being. However, Diener and Biswas-Diener(2002) reviewed studies showing that national levels ofwealth strongly predict the subjective well-being of na-tions, whereas change in wealth is inconsistent in its effectsacross studies.
This cross-sectional evidence that circ*mstances mat-ter is supported by more definitive longitudinal studiesexamining individuals over time. For instance, Fujita andDiener (2005) used longitudinal data to determine whetherlong-term average levels of happiness ever change. Theyexamined changes in baseline levels of well-being over a17-year period in a large and representative sample fromGermany. Although there was considerable stability inhappiness reports, 24% of respondents changed signifi-cantly from their early baseline, comprising the first fiveyears of the study, to the last five years. Nine percentchanged by approximately two standard deviations ormore. Thus, long-term levels of happiness do change forsome individuals. The more intriguing question, then, is
why happiness set points change for some individuals morethan for others.
Using the same sample of Germans, we have exam-ined the ways that specific life events influence happiness.In support of the initial adaptation model, people do seemto adapt to some life events. For instance, Lucas et al.(2003) showed that, on average, Germans did not getlasting boosts in happiness after marriage. Instead, theyreported short-term increases in happiness that were fol-lowed by relatively quick adaptation. However, the extentof adaptation varies for different life events. Lucas et al.(2003) showed that widows and widowers, people whowere laid off from work (Lucas, Clark, Georgellis, &Diener, 2004), and individuals who divorced (Lucas,2005b) all reported long-lasting changes in life satisfactionafter these life events. The widows showed the greatestamount of adaptation (at least in terms of the absoluteincrease from their lowest level of happiness), but even thistook about eight years and was not quite complete. Figure2 shows life satisfaction levels before and after these fourimportant life events.
Together these results suggest that happiness can anddoes change. What then should be made of the classicempirical findings of Brickman et al. (1978)? First, itshould be noted that when Brickman et al.’s results areexamined closely, the evidence for adaptation is not nearlyas strong as many psychologists have tended to assume. Inthe case of individuals with spinal cord injuries, Brickmanet al. did find that the participants who were disabledreported significantly less happiness than did controls. Infact, when we calculated standardized mean differences ingeneral happiness from Brickman et al.’s data, we foundthat the difference between the spinal cord–injured andcontrol groups was about 0.75 standard deviations—aneffect that most psychologists would consider large. Sim-ilar effects have been found numerous times: Authors of anumber of recent reviews have concluded that individualswith spinal cord injuries are less happy than are people inthe general population, with effect sizes in the moderate tolarge range (Dijkers, 1997, 2005; Hammell, 2004). How-ever, the studies cited in these reviews are often publishedin rehabilitation journals and are rarely cited in psycholog-ical literature on adaptation.
Finally, Lucas (2005a) used two large, nationally rep-resentative panel studies to examine adaptation to the onsetof disability. Participants in this study (who were followedfor an average of seven years before and seven years afteronset) reported moderate to large drops in satisfaction andvery little evidence of adaptation over time. For instance,those individuals who were certified as being 100% dis-abled reported life satisfaction scores that were 1.20 stan-dard deviations lower than their nondisabled baseline lev-els. Thus, although people with paraplegia and otherindividuals with disabilities usually are not subjectivelymiserable, happiness levels do seem to be strongly affectedby this important life circ*mstance. When compared withthe actual variability between individuals in happinessrather than with the extreme endpoints of the scale, manyof the group differences in happiness are substantial.
Table 2The Happiness of Selected Nations
NationAffect balance (PA � NA),
1981–1984Life satisfaction,
1999–2001
Canada 2.33 7.85United States 2.23 7.66China 1.46 6.53West Germany 1.45 7.42Mexico 1.38 8.14India 0.72 5.14Turkey 0.62 5.61Russia 0.33 4.65
Note. Mean scores are taken from the World Value Survey, the BradburnAffect Balance Scale, where affect balance can vary from 5 to �5, with 0 as theneutral point. The national differences in both positive affect (PA) and negativeaffect (NA) in the full sample are highly significant, p � .001. Life satisfactionscores, with a range of 1 to 10, were taken from the European Values StudyGroup and World Values Survey Association (2005) Data Wave 1999–2001.
309May–June 2006 ● American Psychologist
It should no longer come as a surprise that peopleliving in negative circ*mstances report well-being scoresthat are above neutral. This well-documented fact is inter-esting and theoretically important, but it should not be usedas evidence that people inevitably adapt. Furthermore, it isnot enough for researchers interested in adaptation to showthat people who have experienced a negative life circum-stance report well-being scores that are higher than whatother people would think they should report (e.g., Brick-man et al., 1978; Riis et al., 2005). Such research findingstell more about the average person’s affective forecastingerrors than about adaptation itself (Gilbert & Wilson,2000). To determine whether adaptation has occurred, it isnecessary to compare individuals who have experienced anevent or life circ*mstance with those who have not, ideallyfollowing the same individuals over time.
Revision 5: Individual Differences inAdaptationAn implicit assumption of the hedonic treadmill theory isthat adaptation to circ*mstances occurs in similar ways forall individuals. If adaptation results from automatic andinevitable homeostatic processes, then all individualsshould return to neutrality or at least to their own uniquebaseline. But we have found individual differences in therate and extent of adaptation that occurs even to the sameevent. In our longitudinal studies, the size and even the
direction of the change in life satisfaction varied consider-ably across individuals. For example, Lucas et al. (2003)found adaptation to marriage at the aggregate level, butthere was a great deal of variability in these effects. Indi-viduals who reacted most positively to their marriagetended to be above their baseline many years after theevent, but these individuals were counterbalanced by thosewho experienced a lasting decline in satisfaction after theirmarriage. In fact, the standard deviation for the amount ofchange that occurred after the event was almost as large asthe standard deviation for baseline levels.
Understanding individual differences in adaptationwill help illuminate when and why adaptation does or doesnot occur. For example, in our study on reaction andadaptation to marriage (Lucas et al., 2003), we relied onlaboratory studies of emotional reactivity (e.g., Larsen &Ketelaar, 1991) to predict that the happiest individualsshould react most strongly to positive life events. However,the results showed—somewhat surprisingly—that less sat-isfied individuals were more likely to benefit from marriagein the long run. These individuals with initially low base-lines reported more positive reactions to marriage, andthese positive reactions persisted long into the marriage.One explanation for this effect is that the most satisfiedindividuals are more likely than less satisfied individuals tohave strong social support even before the marriage. Peoplewho chronically experience many positive events may have
Figure 2Adaptation to Good and Bad Events
5.5
6
6.5
7
7.5
8
-5 -4 -3 -2 -1 0 1 2 3 4 5
Year
Lif
e S
ati
sfa
ct i
on
Widowhood Divorce Unemployment Marriage
Ye
ar
of E
ven
t
310 May–June 2006 ● American Psychologist
less to gain from one more positive event. Likewise, peoplewho chronically experience many bad events may not bestrongly affected by the addition of one more negative lifeevent. Therefore, deviations from a person’s typical lifeevents might produce the greatest changes in happiness setpoints (Headey & Wearing, 1992; Oishi, Diener, Choi,Kim-Prieto, & Choi, 2005).
Two important research traditions shed light on whenpeople adapt or do not adapt to negative events. The first ofthese traditions focuses on the utility of specific copingstrategies. The second focuses on personality characteris-tics that influence the specific coping strategies that peopleuse. From these literatures, it is known that certain copingstrategies are more effective than others and that individ-uals vary in their preferred strategies. For example, indi-viduals who tend to use reappraisal strategies experiencemore positive emotions and fewer negative emotions thando individuals who use strategies such as suppression(Gross & John, 2003). Using reappraisal is also associatedwith having better interpersonal relationships, which arelikely to translate into increased social support. Similarly,among older people, the endorsem*nt of coping styles suchas using humor, seeking information, and “keeping going”predicts adjustment to old age (Staudinger & Fleeson,1996).
Personality researchers have shown that a number ofstable individual differences predispose people to use cer-tain coping strategies (Carver, Scheier, & Weintraub,1989). For example, neurotic individuals often choose in-effective strategies for coping, which can lead to greaterreactivity to a stressful event (Bolger & Zuckerman, 1995)and possibly a slower return to baseline levels of happiness.Similarly, Ferguson (2001) found that neuroticism andintroversion were associated with relatively ineffectivecoping behaviors such as denial. However, optimistic in-dividuals tend to engage in active coping or strategies thatcan actually change the situation that is causing negativeaffect (Aspinwall & Taylor, 1997; Chang, 1998; Scheier,Weintraub, & Carver, 1986). Such strategies often pay offby leading to a resolution of the stressful situation. Opti-mistic people also tend to seek out social support, engage inpositive reappraisal of adverse events, and feel as if theyhave the resources to overcome stressful situations (Scheieret al., 1986)—all factors that help buffer against the long-term effects of negative life circ*mstances. For example,optimism has been shown to predict problem-focused cop-ing and quicker recovery from surgery (Scheier et al.,2003). Thus, there appear to be individual differences ineffective coping and adaptation to stressful events.
Research on individual differences in adaptation raisesquestions about the processes that underlie adaptation ef-fects. If adaptation is an inevitable and automatic process,it should occur in similar ways for most people, much ashomeostatic processes work to return all people to theirbody temperature set point. The fact that substantial indi-vidual differences in these effects exist argues against thistype of inevitable habituation model. It also suggests thatresearch into these individual differences may help psy-chologists understand exactly how adaptation occurs. How-
ever, at this point, it is unclear whether there is a relativelyautomatic core habituation process that can be modified bycoping and other variables or whether adaptation and cop-ing are synonymous. Future research must incorporatemeasures of coping (along with other potential moderatorsand process variables) into sophisticated longitudinal stud-ies that allow for strict tests of adaptation effects.
Recent research has provided a much stronger test ofthe hedonic treadmill than earlier studies did because ofmethodological refinements. First and most important, byrelying on very large samples, researchers in recent studieshave been able to track individuals from before an eventhappens to the time of the event to many years after theevent. By contrast, earlier researchers drew conclusionsfrom cross-sectional data in which preevent levels of lifesatisfaction of groups such as lottery winners or peoplewith paraplegia were unknown. Second, large longitudinaldesigns allow for more precise measurement of changes inhappiness over time and more powerful statistical methodsthat go beyond examinations of group means to revealindividual differences in adaptation. Finally, recent studieshave used large and often representative samples of partic-ipants, unlike early studies that frequently used small ac-cidental samples.
Implications of the Revised ModelIf revisions must be made to the original hedonic treadmillmodel, is adaptation still an important concept for psycho-logical research? We answer with a resounding “yes.”Although recent studies have challenged the idea that ad-aptation is inevitable, people do adapt to many life events,and they often do so within a relatively short period oftime. Thus, adaptation processes can explain why manyfactors often have only small influences on happiness.People tend to adapt to these conditions over time.
However, recent findings do place limits on the typesof psychological processes that can account for the adap-tation that does occur. For instance, initial models thatrelied on automatic physiological systems to account forhedonic adaptation will likely not be able to fully accountfor all existing data. Instead, more flexible processes arelikely involved, and these processes may vary across eventsand individuals or even within the same individual overtime. The research on coping with adversity will be a usefulstarting point for investigations of adaptation. However,processes related to adaptation to positive events must alsobe explored.
Newer theories of adaptation (e.g., Kahneman & Tha-ler, in press; Wilson & Gilbert, 2005) rely on individuals’attention to particular life circ*mstances in explaining thechanges. Kahneman and Thaler, for instance, posited thatvarious features of a specific life circ*mstance might in-fluence whether it draws a person’s attention. It is thisattention that determines whether an individual can adapt.Thus, Kahneman and Thaler predicted that conditions thatcontinue to draw attention can influence well-being but thatthe novelty of certain circ*mstances wears off and there-fore they draw less attention over time. Wilson and Gilbertfurther suggested that people naturally seek to explain and
311May–June 2006 ● American Psychologist
make sense of life events and circ*mstances. Features ofone’s life that cannot be explained continue to draw atten-tion and thereby affect one’s emotions and overall well-being. Experience-sampling reports over time of the stimulito which people attend are needed to test attention theories.Whether these attention theories can predict greater andless habituation has not yet been rigorously tested.
Our revisions to the hedonic treadmill model suggestthat interventions to increase happiness can be effective,and research supports this conclusion. These changes mightbe targeted at the individual, organizational, or even soci-etal level. For instance, in an early set of studies, Fordyce(1977, 1983) demonstrated in seminal studies that a mul-tipronged program successfully raised individuals’ happi-ness for an extended period of time. These gains in well-being persisted over a period of a year or more. Perhapsbecause of the widely accepted view that happiness couldnot be changed, however, few rigorous studies have beenconducted to follow up on this work. Very recently, this hasbegun to change. For instance, Sheldon and Lyubomirsky(2004) demonstrated that changes in activities raised peo-ple’s happiness. They found that when individuals per-formed several random acts of kindness on one day eachweek, their happiness improved. Seligman, Steen, Park,and Peterson (2005) reported a series of happiness inter-ventions that were implemented via the Internet. Theyfound that several of these interventions led to changes inhappiness that persisted for at least six months. Finally,Emmons and McCullough (2003) found that interventionsto increase thoughts of gratitude increased levels of posi-tive affect. Although these experimental intervention stud-ies are in the initial stages, they indicate that levels ofhappiness can be raised. Again, this contradicts the idea ofan unchangeable baseline for happiness.
If interventions can cause lasting changes among in-dividuals, it may also be possible for organizations to adoptmacrolevel policies that raise well-being for larger groups.For instance, organizational psychologists strive to makethe workplace engaging and interesting. These benefitsmight be worthwhile in themselves, or the increased hap-piness that they provide may lead to increases in organi-zational citizenship and productivity. Similarly, commu-nity psychologists strive to enhance the quality of lifewithin neighborhoods and cities. Our findings that baselinehappiness can change, along with new studies showing thatinterventions can raise levels of happiness, provide anoptimistic foundation for the various fields of appliedpsychology.
Finally, if organizational policies can have an impacton the happiness of large groups, it may be possible tochange the happiness of a society as a whole. Philosopherssuch as J. S. Mill and Jeremy Bentham maintained that thebest society is one where the greatest numbers of citizensexperience the most happiness. Echoing this sentiment,Diener and Seligman (2004) called for a system of nationalaccounts of well-being in which people’s happiness, mean-ing, and engagement are assessed over time and in varioussituations. The goal of such a program would be to helppolicymakers understand when and why people are miser-
able and when and why they are happy. This informationwould then allow policymakers to develop programs toreduce misery and enhance happiness. Furthermore, it ishoped that national accounts of well-being might lead topolicies that would heighten the engagement, joy, trust, andaffection of ordinary citizens who do not have extraordi-nary problems. Fortunately, our findings indicate that thegoal of creating a happier society is not doomed by thehedonic treadmill.
Although the research reviewed in this article providesan optimistic picture of the possibility for change, theprocesses of adaptation must still be carefully consideredwhen designing and assessing well-being interventions.People might initially react positively to interventions justas they do to naturally changing conditions, but over timethey may adapt to the intervention and return to theirformer levels of well-being. Thus, effective interventionsmust change people’s baseline well-being, and measure-ments must be repeated over a long period of time to ruleout the possibility that the effectiveness of the interventionis only temporary. A strong understanding of adaptationtheories will enable researchers to develop programs with agreat likelihood of long-term success.
Future Research and Conclusions
Although researchers have made progress in understandingadaptation, several key issues remain. First, an overarchingquestion concerns the factors that lead to lasting change.Why do adaptation effects appear to vary across differentevents and circ*mstances? Although some theories (e.g.,Kahneman & Thaler, in press) offer predictions about thedifferential adaptation across varying events, these theoriesdo not seem to explain the full set of results. For instance,it is unclear why people seem to exhibit a lasting effect ofunemployment on well-being even after they become re-employed. A corollary question concerns how much con-trol people have over adaptation: Can people slow adapta-tion to good events and speed recovery from bad events?Another important challenge is differentiating passive ac-ceptance of negative circ*mstances versus active copingand a positive resolution of events. Finally, our studiesraise the issue as to whether some components of well-being adapt more readily than others. For instance, docognitive evaluations such as satisfaction adapt moreslowly than moods and emotions? These are exciting un-answered questions about adaptation, questions that willneed to be answered before fully effective interventions canbe designed.
The treadmill model of happiness posited by Brick-man and Campbell (1971) represents a milestone in psy-chologists’ understanding of happiness, and our longitudi-nal findings on marriage support the treadmill idea. Ourfindings also indicate that different types of well-being maychange at different rates or even in different directions.Furthermore, both experimental and longitudinal studiesnow show that the strong form of the adaptation theory isuntenable. Adaptation may proceed slowly over a period ofyears, and in some cases the process is never complete.
312 May–June 2006 ● American Psychologist
Finally, there are individual differences in the rates ofadaptation.
Those who provide interventions aimed at improvingsubjective well-being need to understand the patterns in-volved in adaptation so that successful interventions can bedesigned. Although some of the studies we described in-volve changes in life circ*mstances that are extreme, otherstudies suggest that smaller interventions can make a dif-ference. Adaptation is a powerful force, but it is not socomplete and automatic that it will defeat all efforts tochange well-being. The exciting research challenge is todiscover the factors that control the adaptation process.Fortunately, research on coping, personality traits, and theeffectiveness of interventions all offer clues about factorsthat influence adaptation. With the understanding that ad-aptation may be incomplete and varies across persons, theefforts to understand adaptation should be amplified.
REFERENCES
Aspinwall, L. G., & Taylor, S. E. (1997). A stitch in time: Self-regulationand proactive coping. Psychological Bulletin, 121, 417–436.
Biswas-Diener, R., Vittersø, J., & Diener, E. (2005). Most people arepretty happy, but there is cultural variation: The Inughuit, the Amish,and the Maasai. Journal of Happiness Studies, 6, 205–226.
Bolger, N., & Zuckerman, A. (1995). A framework for studying person-ality in the stress process. Journal of Personality and Social Psychol-ogy, 69, 890–902.
Bonanno, G. A., Wortman, C. B., Lehman, D. R., Tweed, R. G., Haring,M., Sonnega, J., et al. (2002). Resilience to loss and chronic grief: Aprospective study from preloss to 18-months postloss. Journal of Per-sonality and Social Psychology, 83, 1150–1164.
Bonanno, G. A., Wortman, C. B., & Nesse, R. M. (2004). Prospectivepatterns of resilience and maladjustment during widowhood. Psychol-ogy and Aging, 19, 260–271.
Brickman, P., & Campbell, D. T. (1971). Hedonic relativism and planningthe good society. In M. H. Appley (Ed.), Adaptation level theory: Asymposium (pp. 287–302). New York: Academic Press.
Brickman, P., Coates, D., & Janoff-Bulman, R. (1978). Lottery winnersand accident victims: Is happiness relative? Journal of Personality andSocial Psychology, 36, 917–927.
Campbell, A., Converse, P. E., & Rodgers, W. L. (1976). The quality ofAmerican life. New York: Russell Sage Foundation.
Carver, C. S., & Scheier, M. F. (1990). Origins and functions of positiveand negative affect: A control-process view. Psychological Review, 97,19–35.
Carver, C. S., Scheier, M. F., & Weintraub, J. K. (1989). Assessing copingstrategies: A theoretically based approach. Journal of Personality andSocial Psychology, 56, 267–283.
Chang, E. C. (1998). Dispositional optimism and primary and secondaryappraisal of a stressor: Controlling for confounding influences andrelations to coping and psychological and physical adjustment. Journalof Personality and Social Psychology, 74, 1109–1120.
Diener, E., & Biswas-Diener, R. (2002). Will money increase subjectivewell-being? A literature review and guide to needed research. SocialIndicators Research, 57, 119–169.
Diener, E., & Diener, C. (1996). Most people are happy. PsychologicalScience, 7, 181–185.
Diener, E., Diener, M., & Diener, C. (1995). Factors predicting thesubjective well-being of nations. Journal of Personality and SocialPsychology, 69, 851–864.
Diener, E., & Lucas, R. E. (1999). Personality and subjective well-being.In D. Kahneman, E. Diener, & N. Schwarz (Eds.), Well-being: Thefoundations of a hedonic psychology (pp. 213–229). New York: RussellSage Foundation.
Diener, E., Sandvik, E., Seidlitz, L., & Diener, M. (1993). The relation-ship between income and subjective well-being: Relative or absolute?Social Indicators Research, 28, 195–223.
Diener, E., & Seligman, M. E. P. (2004). Beyond money: Toward aneconomy of well-being. Psychological Science in the Public Interest, 5,1–31.
Diener, E., Wolsic, B., & Fujita, F. (1995). Physical attractiveness andsubjective well-being. Journal of Personality and Social Psychology,69, 120–129.
Dijkers, M. (1997). Quality of life after spinal cord injury: A meta-analysis of the effects of disablement components. Spinal Cord, 35,829–840.
Dijkers, M. P. J. M. (2005). Quality of life of individuals with spinal cordinjury: A review of conceptualization, measurement, and research find-ings. Journal of Rehabilitation Research and Development, 42, 87–110.
Easterlin, R. (2005, June). Life cycle happiness and its sources: Whypsychology and economics need each other. Paper presented at theInternational Conference on Capabilities and Happiness, Milan, Italy.
Economist Intelligence Unit. (2005). The Economist Intelligence Unit’sQuality-of-Life Index. Retrieved July 17, 2005, from http://www.economist.com/media/pdf/QUALITY_OF_LIFE.pdf
Eid, M., & Diener, E. (2004). Global judgments of subjective well-being:Situational variability and long-term stability. Social Indicators Re-search, 65, 245–277.
Emmons, R. A., & McCullough, M. E. (2003). Counting blessings versusburdens: An experimental investigation of gratitude and subjectivewell-being in daily life. Journal of Personality and Social Psychology,84, 377–389.
European Values Study Group & World Values Survey Association.(2005). European and World Values Surveys Integrated Data File,1999–2002, Release I (2nd ICPSR version) [Computer file]. AnnArbor, MI: Inter-University Consortium for Political and Social Re-search.
Feinman, S. (1978). The blind as “ordinary people.” Journal of VisualImpairment and Blindness, 72, 231–238.
Ferguson, E. (2001). Personality and coping traits: A joint factor analysis.British Journal of Health Psychology, 6, 311–325.
Fordyce, M. W. (1977). Development of a program to increase personalhappiness. Journal of Counseling Psychology, 24, 511–520.
Fordyce, M. W. (1983). A program to increase happiness: Further studies.Journal of Counseling Psychology, 30, 483–498.
Fredrick, S., & Loewenstein, G. (1999). Hedonic adaptation. In D. Kah-neman, E. Diener, & N. Schwarz (Eds.), Well-being: The foundations ofa hedonic psychology (pp. 302–329). New York: Russell SageFoundation.
Fredrickson, B. L. (1998). What good are positive emotions? Review ofGeneral Psychology, 2, 300–319.
Fujita, F., & Diener, E. (2005). Life satisfaction set point: Stability andchange. Journal of Personality and Social Psychology, 88, 158–164.
Gilbert, D. T., & Wilson, T. D. (2000). Miswanting: Some problems in theforecasting of future affective states. In J. P. Forgas (Ed.), Feeling andthinking: The role of affect in social cognition (pp. 178–197). NewYork: Cambridge University Press.
Gross, J. J., & John, O. P. (2003). Individual differences in two emotionregulation processes: Implications for affect, relationships, and well-being. Journal of Personality and Social Psychology, 85, 348–362.
Hammell, K. W. (2004). Exploring quality of life following high spinalcord injury: A review and critique. Spinal Cord, 42, 491–502.
Headey, B., & Wearing, A. (1989). Personality, life events, and subjectivewell-being: Toward a dynamic equilibrium model. Journal of Person-ality and Social Psychology, 57, 731–739.
Headey, B., & Wearing, A. (1992). Understanding happiness: A theory ofsubjective well-being. Melbourne, Victoria, Australia: LongmanCheshire.
Helson, H. (1948). Adaptation-level as a basis for a quantitative theory offrames of reference. Psychological Review, 55, 297–313.
Helson, H. (1964). Current trends and issues in adaptation-level theory.American Psychologist, 19, 26–38.
Inglehart, R., & Klingemann, H.-D. (2000). Genes, culture, democracy,and happiness. In E. Diener & E. M. Suh (Eds.), Culture and subjectivewell-being (pp. 165–184). Cambridge, MA: MIT Press.
Kahneman, D., & Thaler, R. H. (in press). Anomalies: Attention andexperienced utility. Journal of Economic Perspectives.
Larsen, R. J., & Ketelaar, T. (1991). Personality and susceptibility to
313May–June 2006 ● American Psychologist
positive and negative emotional states. Journal of Personality andSocial Psychology, 61, 132–140.
Lucas, R. E. (2005a). Happiness can change: A longitudinal study ofadaptation to disability. Manuscript submitted for publication, Michi-gan State University, East Lansing.
Lucas, R. E. (2005b). Time does not heal all wounds: A longitudinal studyof reaction and adaptation to divorce. Psychological Science, 16, 945–950.
Lucas, R. E., Clark, A. E., Georgellis, Y., & Diener, E. (2003). Reexam-ining adaptation and the set point model of happiness: Reactions tochanges in marital status. Journal of Personality and Social Psychol-ogy, 84, 527–539.
Lucas, R. E., Clark, A. E., Georgellis, Y., & Diener, E. (2004). Unem-ployment alters the set point for life satisfaction. Psychological Science,15, 8–13.
Lucas, R. E., Diener, E., & Suh, E. (1996). Discriminant validity ofwell-being measures. Journal of Personality and Social Psychology, 7,616–628.
Lyubomirsky, S., King, L. A., & Diener, E. (2005). The benefits offrequent positive affect: Does happiness lead to success? PsychologicalBulletin, 131, 803–855.
Myers, D. (1992). The pursuit of happiness. New York: Morrow.Oishi, S., Diener, E., Choi, D. W., Kim-Prieto, C., & Choi, I. (2005). The
dynamics of daily events and well-being across cultures: The decliningmarginal utility of daily events. Manuscript submitted for publication,University of Virginia, Charlottesville.
Okun, M. A., & George, L. K. (1984). Physician- and self-ratings ofhealth, neuroticism, and subjective well-being among men and women.Personality and Individual Differences, 5, 533–539.
Riis, J., Loewenstein, G., Baron, J., Jepson, C., fa*gerlin, A., & Ubel, P. A.(2005). Ignorance of hedonic adaptation to hemodialysis: A study usingecological momentary assessment. Journal of Experimental Psychol-ogy: General, 134, 3–9.
Scheier, M. F., Matthews, K. A., Owens, J. F., Magovern, G. J. S.,Lefebvre, R. C., Abbott, R. A., et al. (2003). Dispositional optimismand recovery from coronary artery bypass surgery: The beneficial
effects on physical and psychological well-being. In P. Salovey & A. J.Rothman (Eds.), Social psychology of health (pp. 342–361). New York:Psychology Press.
Scheier, M. F., Weintraub, J. K., & Carver, C. S. (1986). Coping withstress: Divergent strategies of optimists and pessimists. Journal ofPersonality and Social Psychology, 51, 1257–1264.
Scollon, C. N. (2004). Predictors of intraindividual change in personalityand well-being. Unpublished doctoral dissertation, University of Illi-nois at Urbana-Champaign.
Scollon, C. N., & Diener, E. (2005). Love, work, and changes in extra-version and neuroticism over time. Manuscript submitted forpublication.
Seligman, M. E. P., Steen, T. A., Park, N., & Peterson, C. (2005). Positivepsychology progress: Empirical validation of interventions. AmericanPsychologist, 60, 410–421.
Sheldon, K., & Lyubomirsky, S. (2004). Achieving sustainable newhappiness: Prospects, practices, and prescriptions. In P. A. Linley & S.Joseph (Eds.), Positive psychology in practice (pp. 127–145). Hoboken,NJ: Wiley.
Silver, R. L. (1982). Coping with an undesirable life event: A study ofearly reactions to physical disability. Unpublished doctoral disserta-tion, Northwestern University, Evanston, IL.
Staudinger, U. M., & Fleeson, W. (1996). Self and personality in old andvery old age: A sample case of resilience? Development and Psycho-pathology, 8, 867–885.
Suh, E., Diener, E., & Fujita, F. (1996). Events and subjective well-being:Only recent events matter. Journal of Personality and Social Psychol-ogy, 70, 1091–1102.
Tellegen, A., Lykken, D. T., Bouchard, T. J., Wilcox, K. J., Segal, N. L.,& Rich, S. (1988). Personality similarity in twins reared apart andtogether. Journal of Personality and Social Psychology, 54, 1031–1039.
Wilson, T. D., & Gilbert, D. T. (2005). Making sense: A model of affectiveadaptation. Manuscript submitted for publication, University of Vir-ginia, Charlottesville.
314 May–June 2006 ● American Psychologist
