Psiholo.ka obzorja / Horizons of Psychology, 9, 1, 7-22 (2000)H Dru.tvo psihologov Slovenije 2000, ISSN 1318-187HZnanstveni teoretsko-pregledni prispevekH On the effects of changes in group status#R RADMILA PRI.LIN Department of Psychology, San Diego State University, US Abstract: Change is a very frequent sociopsychological phenomenon, that we notice wheU transformations from minority to majority, high to low social status, weak to strong socia position and vice versa are being conducted. Causes of change are described in literaturI dealing with attitudes and persuasion, group dynamics, conformity and group developmentp But we lack information on consequences of change. This article describes three subject . related to this problem: (i) assessment and evaluation of personal change, (ii) assessment anB evaluation of social change and (iii) reactions to change-related agents in a group contextp The author introduces a series of experiments based on loss-gain asymmetry (Tversky Kahneman, 1991). The results implicate that assessment and evaluation of change is influN enced by different agents such as the direction of comparison (comparing past to present o . vice versa), desirability of stability or change of attitudes, maintaining positive views abou1 the self or the group, and gain or loss of group status or power. The change of group statu . leads to change in the perception of group-self similarity and group attraction, expectancie . about future group interactions and group self-evaluation. Change of group status and powe . is a central determinant of intergroup relationsp Key words: personal change, social change, loss-gain asymmetry, attitudes, majority-minorit status, intergroup relations, effects of changI Vplivi sprememb polo.aja v skupiniR RADMILA PRI.LIu Oddelek za psihologijo, Dr.avna Univerza v San Diegu, ZD Povzetek: Sprememba je pogost socialnopsiholo.ki pojav, ki ga opazimo pri transformacija . manj.ine v veeino, pri prehodih iz moene v .ibko pozicijo, iz visokega v nizek status in obratno. Vzroki sprememb so opisani v literaturi o preprieanjih, vplivu manj.in, konformnosti in razvoj skupine. Posledice sprememb pa niso tako ob.irno raziskane. Prieujoei prispevek povzema . tem povezane raziskave. Govori o treh temah: (i) o ocenjevanju in vrednotenju osebnostni . sprememb, (ii) o ocenjevanju in vrednotenju socialnih sprememb in (3) o reakcijah skupine n dejavnike, povezane s spremembo. Avtorica navaja serijo eksperimentov, zasnovanih na naeel asimetrienih reakcij na dobitke in izgube (Tversky & Kahneman, 1991). Izsledki raziskav ka.ejo da na ocenjevanje in vrednotenje sprememb vplivajo dejavniki, kot so smer primerjavI (primerjava preteklosti s sedanjostjo in obratno), za.elenost stabilnosti ali spremembe stali.e motivacija, doprinos spremembe k pozitivnemu samovrednotenju ali vrednotenju skupine te . * Naslov / address: prof. dr. Radmila Pri.lin, Department of Psychology, San Diego State University, 5500HCampanile Drive, San Diego, CA 92182-4611, USA, e-mail: rprislin@sunstroke.sdsu.eduH # Prispevek povzema vabljeno plenarno predavanje na 3. Kongresu psihologov Slovenije oktobra 1999 vHPortoro.u dne 6. oktobra 1999.H R. Pri.linH pridobitvev ali izguba statusa in moei skupine. Sprememba statusa skupine vpliva na zaznan 1 podobnost s skupino, prieakovanja o interakciji s skupino, privlaenost skupine in za skupin 1 specifieno kolektivno samovrednotenje. Moe in status skupine sta osrednji determinanto medskupinskih odnosovp Kljuene besede: osebna in socialna sprememba, naeelo asimetrienih reakcij na dobitke in izgube stali.ea, medskupinski odnosi, status manj.ine in veeinI CC=302 1 Although social psychologists have been studying change ever since their disciplinI was founded as a scientific endeavor, they have treated change almost exclusively a . a dependent variable. Domains of attitudes and persuasion (e.g., Eagly & Chaiken 1993), social influence (e.g., Turner, 1991), and group dynamics (e.g., Levine Moreland, 1998) are examples par excellence. As a result, we are much better inN formed about causes than consequences of change. Interestingly, this focus on cause . rather than on consequences characterizes not only social but also other domains o . psychology as well. In a stark contrast, other scientific disciplines, especially thosI focused on social and economic changes are more likely to document not only preN sumed causes (Feng & Zak, 1999) but also consequences of change (Kugler Feng, 199) . This is best illustrated by sociological, demographic, and economic indiN cators of recent socio-political changes that are most profoundly observable in EuN rope but also in other parts of the world: from South Africa to East Timor to Canada etc. As informative as these indicators are, they are mute to the issue o . phenomenological experience of change. Understanding of the issue appears to bI inversely proportional to its relevance, as if confirming an old cynicism that the morI important a phenomenon is the less likely psychologists are to study it. In the pas1 several years, however, cynics have been loosing ground as several researchers including my students and myself, have been examining the phenomenon. In wha1 follows an overview of relevant research will be presented, organized around threI topics: 1) Assessment and evaluation of personal change; 2) Assessment and evaluN ation of social change, and 3) Reactions to change-related agents in a group contextp Assessment and Evaluation of Personal ChangeR .Has your attitude toward death penalty changed in the past ten years?. .Has you . political ideology changed during the same period?. How individuals answer thesI questions has been examined within the social cognition framework (Devine, HamilN ton, & Ostrom, 1994; Fiske & Taylor, 1991) that emphasizes cognitive structures anB processes responsible for arriving to specific answers (Prislin & Ouellette, 1996). A . Effects of changes in group statusH important as they are, cognitive structures, however, are not the sole determinin . factors. In addition, motivational forces impinge on these structures to determinI processes by which they operate (Prislin & Pool, 1996; Prislin, Wood, & Pool, 1998) p Thus, answering questions about personal change involves more than an attempt t 1 recollect one.s past and then comparing it to the present. Assessment of change at personal level is an active, constructive process, guided as much by motivationa forces and information-processing schemas as by the objective state of affairs. Thi . assumption is the crux of Michael Ross. model of the assessment of personal changep According to Ross, when assessing whether their attitudes (opinions, beliefs, part affiliations, etc.) have changed, individuals undergo a multi-stage process (Ross, 1989) p The process commences with the assessment of current attitudes. Next step, howN ever, does not involve the assessment of past attitudes. Rather, there is an intermediN ary side-step whereby an implicit theory of change is invoked. This theory, in turn determines how attitudes held in the past compare to those currently espoused. ThI function of an implicit theory, therefore, is to organize memories into a coherent patN tern of information supportive of the theory. The organizational function is regulateB by motivational forces in that memories are organized to support conclusions abou1 stability or change depending on whether the former or the latter yields favorablI self-evaluation and/or social evaluationp In the domain of attitudes and other socially debatable dimensions, stabilit appears to be preferred over change. Compared to those who tend to move along thI attitudinal continuum, individuals who have stable attitudes over time are evaluateB more positively by others and tend to think more highly of themselves. Because o . this functionality of attitude stability for winning social rewards and maintaining posiN tive views about the self, people tend to exaggerate stability of their attitudes (Niemi Katz, & Newman, 1980). Even when their attitudes change over time, individual . maintain an illusion of stability by revising their past. Accordingly, their recollection . of their past standings on social issues appear to be more heavily influenced by thei . contemporary attitudes than by their past attitudes. Indeed, in a longitudinal study o . attitudes toward various social issues, Markus (1986) found that recollection of pas1 attitudes correlated more strongly with measures of current attitudes that with measN ures of attitudes obtained in the pastp As mentioned earlier, Ross. model postulates that implicit theories may emN phasize stability as well as change. Which one is emphasized depends on the resultN ant implications for the self. Thus when change rather than stability yields positivI implications for the self, recollection of the past may be distorted in the service o . maintaining an illusion of change. That is, when change appears functional for achievin . favorable self- or social evaluation, individuals will detect change even when nonI has actually occurred. They will typically do so by recollecting their personal past a . worse than the present, or, if feasible, by distorting their personal present as bette . than the past. Empirical support for the former was obtained in a study in whic . 10RR. Pri.linH participants were assigned in either a study-skill improvement program or a contro group (Conway & Ross, 1985). The program turned out to be less efficient thaU participants had been led to expect, resulting in a virtually no improvement in thei . study-skills. Yet, participants in the treatment condition, who were highly motivated t 1 believe that their skills would change for the better maintained their illusion of changI by recollecting their initial skills as being substantially worse than they had actuall reported at the onset of the study. Indeed, participants in this study exhibited reacN tions typical of many among us. For example, many of us are prone to believe that wI were fatter in the past . it makes us look thinner now even if we have not lost a gram that we teach better today than we did years ago even if nothing has changed in ou . teaching efficiency, that we are better parents to our second child than to our firstborU even if none of parenting skills has improved or worse yet, even if we have we los1 much of our initial patience as parentsp Overall, biases reflecting implied (desired) consistency appear more pervasivI than biases reflecting implied (desired) change. Yet a word of caution is in orderp Although understanding biases in recall is important for understanding how individuN als assess change, it is equally important not to exaggerate the extent of the bias. Bia . occurs only when individuals. implicit theories of stability and change contradict realN ity. To the extent that these theories are accurate, individuals. assessment of changI is accurate. Moreover, even when their implicit theories of change are inaccurate individuals are not completely at liberty to reconstruct their past or to reinterpret thei . present (Kunda, 1989; Lundgren & Prislin, 1998). To the extent that external indiN cants of stability (change) are salient, they will correct theory-driven biases. ThesI external indicants may reflect an .objective. reality or shared social interpretations which may be so powerful as to correct not only an individual.s illusion of change bu1 also her or his actual personal change (Nadler, 1993) p Assessment and Evaluation of Social ChangeR .Have things changed for you?. .If they have, has the change been for the better o . for the worse?. .How satisfied are you with the present that was brought about b change of the past?. Answering these and similar questions requires that individual . infer whether they live in a different social reality from those in which they lived iU the past and if so, whether the current social reality is better or worse than it used t 1 bep Just as the assessment and evaluation of personal change are shaped by motiN vational forces and information-processing schemas (Silka, 1988), so are the assessN ment and evaluation of social change (e.g., Chiu & Hong, 1999; Wieczorkowska Burnstein, 1999) This postulate implies that (social) psychology cannot rely on .obN jective. indicators of social change to infer how the change will be experienceB Effects of changes in group statusH11R phenomenologically. Indeed, to a surprise of many, initial enthusiastic reactions t 1 recent profound political and socio-economic changes have subsided both in EuropI (Boehnke, Hagan, Hefler, 1998; Breakwell & Lyons, 1996; Haeyrynen, 1999; Macek Flanagan, Gallay, Kostron, Botcheva, & Czapo, 1998; Sloutsky & Searle-White, 1993) and South Africa (Duckitt & Mphuthing, 1998; Finchilescu & Dawes, 1998) p To explain evaluations of social change, Norbert Schwartz and his colleague . (Schwartz, Wnke, & Bless, 1994) developed the feature matching model that emN phasizes the interpretative nature of remembrance and perceptions as two key deterN mining factors. Developed within the social cognition framework, the model postuN lates that the assessment of social change starts with the construal of mental repreN sentations of the past and the present, which are then compared. The comparisoU process, however, does not follow the rules of formal logic. Departure from forma logic is reflected in the fact that the outcome of the comparison depends not only oU the features of the past and the present but also on the direction of comparison. Thus present social reality will be evaluated differently depending on whether the assessN ment involves comparison of the present to the past or the reversed comparison o . the past to the presentp The relevance of the direction of comparison stems from the fact that it deterN mines whichof the many features of the present and the past will be involved in thI comparison (Tversky, 1977). As an illustration, consider a case of an individual wh 1 in the past had a secure job paying $1000 per month. In addition, the job benefit . included a guaranteed pension after a certain age. The individual.s representation o . the past likely would include the following features: A) being employed, B) makin . $1000 per month, C) having job securing, and D) having guaranteed pension. Let. . further assume that as a result of a social change, the individual has lost job securit and the guaranteed pension but has gained E) the voice within his or her organization F) prospects for professional development, G) performance-based criteria for proN motion, and I) lower tax on the same monthly salary. If the individual assesses thI change by comparing the past to the present, he or she will examine whether feature . A to D also are part of the present. The individual is unlikely to consider features E t 1 I, which are part of the present but not of the past. As a result, the individual wil conclude that the social change resulted in a loss of job security and pension anB likely will evaluate the change negatively. In contrast, if the individual assesses thI change by comparing the present to the past, he or she will examine whether feature . E to I also were part of the past while largely neglecting features A to D. BecausI features E to I are largely positive and were not present in the past, the individua likely will evaluate the change positivelyp In contrast to this illustrative example where the past and the present arI represented by the same number of features, our typical representation of the presen1 is richer in that includes more unique features than our representation of the past whose many features are usually forgotten (Dunning & Madey, 1995). As a conseN quence, the comparison of the present to the past should make salient more uniquI 12RR. Pri.linH features than would be the case when comparing the past to the present. Thus, thI answer to the question .How much has social reality changed?. is likely to be .A lot0 when we compare the present to the past, and .A little. when we compare the pas1 to the present. Generally, more change is inferred from the present-to-the past comN parison than from the past-to-the present comparisonp The direction of comparison influences not only inferences about amount o . change but also evaluations of the inferred change. To predict whether people wil evaluate social change as positive or negative, we need to know which features the use in representing the present and the past and which direction of comparison the will chose. Although theoretically, we are free to use either direction, in most case . our spontaneous assessments of social change is triggered by some current problemp This further implies that in assessing and evaluation change, we are likely to concenN trate on the present as the subject of comparison and on the past as the object o . comparison. If indeed, a current problem is in the focus of our attention, it probably i . so because it is novel and was not part of the past. Thus, when comparing the presen1 to the past, the current problem will stand out as a distinctive feature of the presentp On the other hand, problems that we had in the past likely will escape our attentioU because of the psycho-logic (rather than logic) nature of the comparison proces . (see above). In the end, the resultant evaluation of the present likely is to be negativI while the past likely is to be looked at as the time of the .good old days. 0 Although a current problem is most likely to trigger the present-past compariN son, it is not always so. Theoretically, at least, a positive feature of the present ma start the comparison process. When that happens, that is, when some positive featurI of the present triggers the comparison process, will the outcome be different? Shal we in this case conclude that the past was .gloom and doom,. and that we live in better world? Not necessarily. According to prospect theory (Tversky & Kahneman 1991), when the comparison process leads to the conclusion that the positive featurI of the present did not exist in the past and therefore represents a gain due to socia change, this still does not guarantee positive evaluation of the change. If the compariN son process also leads to conclusion that the change resulted in some losses, that is, i1 caused a loss of some positive aspects of the past, then the principle of loss-gaiU asymmetry applies. Losses loom larger than gains. For example, losing $100 (as in $100 tax increase) is felt more intensively than gaining $100 (as in a $100 salar increase) p Another important implication is that gains and loses of an equal magnitude d 1 not result in .zero net change.. Rather, changes with negative outcomes have greate . impacts than changes with positive outcomes (Tversky, 1994; Tversky & Kahneman 1991). Thus, a change is unlikely to be evaluated as positive unless the resultant gain . exceed the resultant losses. Moreover, because of salience of losses, even a smal loss may outweight a relatively large gain. Thus, if a positive feature of the present i . to lead to the conclusion that the present is better than the past, the feature must fa . outweigh both negative features of the present and positive features of the pastp Effects of changes in group statusH13R Comparing the present to the past (or vice versa) is not the only way of evaluN ating social change. Alternatively, evaluation may be based on how one.s presen1 situation compares to that of others (Festinger, 1954). This implies that even wheU social change brings about considerable improvements, it is not necessarily positivel evaluated if it simultaneously increases relative deprivation or the feeling that onI compares unfavorably to others (Williams, 1975). In support of this proposition, it wa . found that even a substantial increase in personal income is not accompanied with thI correspondent increase in subjective well-being. For example, although the real valuI of personal income in the United States has more than doubled between 1960 anB 1990, the percentage of people considering themselves as very happy has remaineB unchanged at 30 percent (Myers, 1993). Similar conclusion was reached in a stud that examined material wealth and subjective well-being among more than 170,00 1 individuals in sixteen nations (Inglehart, 1990). This, of course, does not imply a tota lack of relationship between material and psychological well-being; rather, it indicate . that the relationship is tenuous at best (Campbell, 1981; Inglehart, 1990 If both, comparisons over time and social comparisons operate to influencI evaluation of social change, then they may operate in tandem, leading to the samI conclusion but also, they may lead to contradictory conclusions. For example unfavorable outcomes of social comparisons may potentially override favorable outN comes of the comparison of change over time. This may explain, for example, widespread dissatisfaction of East Germans, for whom unification changed standN ards of comparison, which may be stronger determinants of their evaluation of changI than any possible improvement over time. Similar process may be responsible for thI observed lack of long-term increase in happiness following enormous lottery winN nings (Inglehart, 1990) p There are some important practical lessons to be learned from social cognitivI research on the assessment and evaluation of change. All processes discussed so fa . suggest that the assessment and evaluation of a social change can be significantl influenced by, among other things, a) individuals. motivation in evaluating change, b their focus of attention, and c) their comparison standards. Interestingly, politician . wishing to convince their electorate that things have changed appear to have beeU aware of this as evident by their many strategies used to this goal. For example, the work hard to motivate people to believe in change (improvement). Motivating peoplI to believe in improvement may not be as difficult as it may seem, even if there arI few objective indicators than any change, much less positive, has occurred. Majorit of those who voted for politicians arguing that there has been some social improveN ment, are typically motivated to justify their (voting) behavior. Their dissonance reN duction typically results in their believing in social improvements. Of course, thi . strategy would work to a certain extent. If, however, elected officials. claims abou1 social improvement are supported by media and claims from other social sources they may combine to validate the picture of reality as significantly improved. T 1 contradict such a shared vision may be psychologically difficult, as Achs (1955) taugh1 14RR. Pri.linH us long time ago. It even may be unwise if those who do not do not see a (positive change are socially stigmatized (Crocker, Major, & Steel, 1998). When these force . combine, they may cause a person to see a change where none has occurredp If controlling motivation proves to be too demanding, it may be easier to guidI temporal comparisons that people make. For example, to the extent that the electorN ate makes comparisons in the .right. direction and on the .right. features, socia reality may seem substantially improved. For example, nothing makes the presen1 social reality look better than negative features of the past social reality. This likely i . a reason for the incumbent politicians trying to focus their electorates. attention to thI negative aspects of the past and away from the negative aspects of the present. IU contrast to this strategy typically used by incumbents, challengers try to focus thI electorate.s attention to the negative aspects of the present. Both often attempt t 1 boost this .social improvement. program by guiding social comparison toward downN ward, which results in a positive evaluation of change. Reactions to Change-Related Agents in a Group ContextR Assessment and evaluation of change, accurate or biased, is only part of the storyp An important aspect of understanding change is understanding reactions to changeN related agents, especially in a group context. In a program of research that my stuN dents and I have been following for the past several years, we are trying to underN stand how people react to others when their own position within a group change . (Prislin, Limbert, & Bauer, 1999). More specifically, we.ve been studying reaction . to others following changes in majority/minority status within a group and followin . changes in power in inter-group contextp To understand individuals. reactions to changes in their status within a group we developed the gain-loss asymmetry model of changes in majority and minorit status. The basic premise of the model is change away from majority status is expeN rienced as loss, whereas changes toward majority status is experienced as gainp Supporting evidence for this premise is abundant. With an exception of elite minoriN ties (Moscovici & Faucheux, 1972), all other minorities .they have been pigeonholed pathologized, deprecated, stigmatized, and dismissed in a countless way. (Moscovici 1994, p. 239). In contrast, majorities enjoy many benefits, both tangible (Sidanius 1999) and intangible (Festinger, 1950; Moscovici, 1976; Simon, 1998; Tajfel & Turner 1986). It therefore appears to be more advantageous to be in majority than in minorN ity, except in highly circumscribed conditions that ensure elitist status to minoritiesp This imbalance in advantages likely is a primary reason for most minorities. attempt . to become majoritiesp Social psychological literature has documented the strategies that minoritie . use in their attempts to become majorities (e.g., Wood, Lundgren, Ouellette, Busceme & Blackstone, 1994) as well as the strategies that majorities use to maintain thei . Effects of changes in group statusH15R positions (e.g., Cialdini & Trost, 1998). In a sharp contrast, social psychological literaN ture is almost mute to the question about consequences of change; that is, about intraN group and inter-group dynamics following a successful attempt of minorities to beN come majorities. Our model assumes that when minority becomes majority, comN parative advantages of majority status result in the change being experienced as gain and change from majority to minority being experienced as a loss. This furthe . implies a substantial change in intra-group dynamics. Because losses loom large . than gains, negative reactions to losing majority status should be stronger than posiN tive reactions to gaining majority status. An important implication is that immediatel following changes in majority/minority positions, the group will be especially fragilI with dividing (centrifugal) forces being stronger than unifying (centripetal) forcesp In first of a series of studies that tested these assumption, we examined thI effects of initial position within a group (majority or minority) and subsequent changI (no change, partial, complete) on members liking for the group, their perception o . similarity between themselves and the group, and strength of attitudes (Prislin, 1996 toward the issue that defined majority/minority positions within the group. In fourN member groups, initial position was created by having two out of the three confederN ates support the participant.s opinion on a socially relevant issue (majority position) p Conversely, to create a minority position, all the three confederates opposed the parN ticipant.s opinion. In the course of group interaction, either all 3 confederates mainN tained their positions thereby creating no change in participant.s position, or one conN federate changed her opinion thereby creating partial change in participant.s positioU or 2 confederates changed their minds, thereby creating a complete change in thI participant.s position. Thus, in the complete change condition, the participant.s posiN tion was changed from majority to minority or vice versa, in the partial change, thI members were left evenly divided, and in the change groups, the participant.s positioU remained constantp Initial position and subsequent change interacted to significantly affect particiN pants. perception of similarity between the group and themselves. Specifically, therI was a dramatic decrease in perception of similarity in participants who lost thei . majority status. In contrast, no significant difference was found among participant . who gained majority status. Virtually identical results were obtained for evaluativI reactions to the group (i.e., group attraction. This pattern of results painted a clea . picture: .New. minorities (majorities who lost their initial position within the group reacted very much like .old/consistent. minorities (minorities who were in that posiN tion from the beginning to the end of group interaction). In a sharp contract, ne . majorities (minorities whose initial position improved) did not react like .old/consistN ent. majorities (majorities who were in that position from the beginning to the end o . group interaction. To the contrary, new majorities. reactions were much alike thosI of .old/consistent. minoritiesp Although gaining majority position did not improve former minorities. reaction . toward the group, it did significantly strengthen their attitudes. New majorities beN 16RR. Pri.linH came less tolerant of opposing views then they had been, and at the same time, the came to believe that the attitudinal issue was more important. Taken together, thesI findings suggest that new majorities may less than amicable, if not hostile toward ne . minorities due to their increased intolerance for opposing opinions and lack of attachN ment for the group as a wholep The obtained asymmetry in reactions to gaining and losing majority status whereas supportive of our model, might be perceived as somewhat artificial becausI participants did not have much control over their position in the group nor did the depend on other group members. for anything but validation of their opinions. T 1 address these concerns, an additional study was conducted. It was a conceptua replication of the first study except that the participants were led to believe that the had control over change in their positions. Specifically, they were told that the stud was about political campaigns and that they, as political candidates, should do thei . best to win support of 3 confederates, who posed as voters. In addition, to make thei . position within the group more consequential, participants were promised an extr reward for achieving majority status, that is, for being electedp Results replicated previous findings in that losing majority position significantl decreased perception of group-self similarity and group attraction whereas gainin . majority positions did not cause any increase in these reactions to the group. AddiN tional measures revealed that participants initially in majority developed clear expecN tations that the group would be supportive, both actively by offering its help, anB passively by refraining from hostility. These positive expectations, however, changeB dramatically as participants. position in the group changed so that at the end, thosI who lost majority position to become new minorities reacted very much like minoriN ties whose position never changed. In contrast to these substantial changes in exN pectancies due to losing majority position, there virtually were no changes in expectN ancies due to gaining majority position: Those becoming a majority still held expectaN tions about the group as unhelpful though not actively hostile. Interestingly, expectanN cies about the group.s likely behavior toward them were largely reciprocated in parN ticipants. expectancies about their likely behavior toward the groupp The robustness of our findings about asymmetry in reactions to gaining anB losing majority position suggests that the asymmetry may be functional. Whereas it i . important for all those whose positions within a group change to accurately asses . and adequately evaluate the change, this task may be more pressing for those whosI position within the group deteriorates than for those whose position within the grouo improves. The former face the task of self-protection, which arguably is more urgen1 than the task of bearing the fruits of newly gained position, faced by the latter. InN deed, it appears that the asymmetry in reactions to social changes is widespread a . indicated by minorities less enthusiastic reactions to apparently improved social conN ditions (Finschilescy & Dawes, 1998; Pettigrew and Tropp; 1999) p Cautiousness observed in minorities. reactions to improvements in their status however, does not imply insensitivity to changes. Their reactions may, however, deN Effects of changes in group statusH17R pend on the mode of change. Indeed, in a recently completed study, our participants. majority or minority status was changed either by other group members. changin . their opinions (as in previous studies) or by new members supporting minority withiU the group (and opposing majority within the group) joining the group. The latter, .deN mographic. mode of change moderated initial minorities. reactions to change. MiN norities who gained majority status by virtue of new, like-minded members joining thI group reacted significantly more positively to the change than minorities who gaineB majority status by virtue of existing members changing their opposition into supportp On the other hand, majorities losing their status reacted equally intensively and negaN tively to the loss, irrespective of whether their loss was due to new, opposing memN bers joining the group or to old members withdrawing their support. It appears tha1 reactions to losing majority status are invariably negative irrespective of how the los . occurs whereas reactions to gaining majority status vary from extremely cautious t 1 positive, depending on how the gain occursp There are some important implications of these findings. If indeed gaining maN jority status by virtue of new members joining a group triggers more positive reacN tions than gaining majority status via old members changing their positions, then aU important question is whether the former mode of change is preferred to the latter. I . given a choice, how would people pursue their goal of gaining majority status? WoulB they prefer evolution that involves trying to change existing members. opinions, o . .revolution,. that involves either recruiting new supportive members (as in our study or trying to get rid of existing opposing members? These are important questions fo . the future researchp Changes in majority/minority position are rarely, if ever, just numerical changesp As indicated previously, among the benefits of majority status is power or the abilit to control one.s own and others. fates (Jones, 1972). Changes in power are espeN cially important for intergroup relations because groups in power control outcome . for the powerless groups. Existing theories of intergroup relations (e.g., elite theory Prewit & Stone, 1973; 5-stage model of intergroup relation, Taylor & McKirnan 1987) posit that intergroup relations are cyclical. Than is, when powerless group . arise to power, they treat others in the same, usually discriminatory way they werI treated, motivating others to seek change and thereby initiate a new cyclep This account of intergroup relations that emphasizes reciprocity (of typicall discriminatory behavior) may be too simplistic. My students and I argue that the usI of newly gained power is determined not only by past relations between groups bu1 also achievements of a newly powerful group. Past relations, certainly, are importan1 in establishing motivational basis for the use of power. There are, however, othe . motives guiding groups. behavior that are presumably general and impinge on behavio . of all groups. One such a motive is positive social identity postulated by highly influenN tial social identity theory (Tajfel, 1982). To establish their raison d.etre groups strivI to be different from other groups and to be different in a positively evaluated direcN tion. They strive to establish and maintain positively distinct social identity. This moN 18RR. Pri.linH tive presumably is responsible for discrimination against out-groups so frequentl observed in many studies but especially in studies using minimal group paradigm. I1 appears that power is used to create new or to exaggerate existing positively valueB differences between in-group and relevant out-groups (see Brewer, 1979; Hogg Abrams, 1988) p What happens when this motive for positively distinct social identity is paireB with the motive for reciprocity that is triggered once a powerless group arises t 1 power? My research team reasoned that use of power would be different dependin . on whether or not newly powerful groups have already satisfied their needs for posiN tive social identity. That is, groups that are successful in their line of activity (hig . status groups) have already satisfied their (presumably primary) motive for positivI social identity. Thus, they should act on their secondary motive . the motive fo . reciprocity. When they gain power, they should treat out-groups the way thosI outgroups treated them in the past. If however, groups have not satisfied their motivI for positive social identity, that is, if they are unsuccessful in their line of activity anB therefore have low status, they should use power to satisfy their unsatisfied motivep Thus, they should use power to discriminate against outgroups irrespective of ho . those outgroups treated them in the past. By discriminating against outgroups, newl powerful groups should establish differences between themselves in a positively valN ued direction. In short, we hypothesized that upon rising to power, high status group . would use the power to reciprocate past intergroup behavior and low status group . would abuse the power to indiscriminately discriminate against othersp This hypothesis was tested in a study in which two 4-member groups were tolB that the goal was to examine how organizational groups work under conditions creN ated by supervisors and how supervisors make decisions about work conditions fo . their subordinates (Rothgerber & Prislin, 1999). Each group was ostensibly randoml chosen to be supervised (powerless) and therefore believed that the other group . was assigned the role of supervisors. After a waiting period during which the superN visor group ostensibly made its decision, each group was required to work on a tas under either unfavorable, fair, or favorable conditions, presumably set for them by thI supervisor group. Following a completion of the task, half of the groups were told tha1 their performed excellently, thereby enjoying high status. For another half, the lo . status condition was created by informing them that they performed very poorlyp When it was their turn to act as supervisors (powerful group), former powerN less groups that had achieved high status reciprocated the treatments they had reN ceived: They set up better than average conditions for groups that had treated the . favorably, worse than average conditions for those who had discriminated agains1 them, and about average conditions for those who had treated them fairly. In conN trast, low status groups did not base their responses to their former supervisors oU how those supervisors than treated them. These groups used power in aU undiscriminating discriminatory way in that they uniformly set less than optimal conN ditionsp Effects of changes in group statusH19R Our findings support the social identity theory contention about power anB status differential as central determinants of intergroup relations. Moreover, the clarify the theory by specifying conditions under which power is used to harm anB benefit others. Whereas standard formulations of the theory clearly predict that powe . is used to harm hostile out-groups in defense of social identity, standard formulation . are not sufficiently specified to allow predictions about reactions to benevolent groupsp This research further develops the theory by specifying that beneficent treatment i . reciprocated when primary motive postulated by the theory, positive social identity i . satisfiedp In closing, this program of research on the effects of change in group statu . indicate that past matters. Majority of natural groups have a history of prior relationse moreover, these relations are dynamic and they tend to change: majorities becomI minorities, powerful become powerless and vice versa. Failure to take these dynami aspects of intergroup behavior into account necessarily leads to poor understandin . of the etiology of intergroup phenomena. As this review suggests, my own and sevN eral other programs of research take change as an important socio-psychologica variable. We have just started answering important questions. Among numerous othN ers waiting to be addressed are questions about potential moderating effects of thI ways in which changes are brought about. We.ve seen that groups react differentl depending on how they achieve majority position. We suspect that modes of rising t 1 power matter (arbitrary vs. merit-based), too. Moreover, change and its effects oU intergroup behavior may well depend on the scarcity of resources groups are comN peting for: Is power used differently in .good. and .bad. times? Also, changes tha1 are experienced at a group level are likely to be publicly debated. If so, then we ma expect that groups (societies) develop shared theories about the changes they arI undergoing and that there should be a high level of agreement in the reconstruction . of the past among members of the same group (society). This agreement should bI functionally important as it provides apparent validation for the reconstruction of thI past. An important implication of this line of reasoning is that collective theories o . change may be potentially more powerful than individual (implicit) theories of changep This issue of .collective. memories, along with those previously mentioned, provide . an exciting research agenda for the future. Reference Asch, S. E. (1995). Opinions and social pressure. Scientific American, 193, 31-35p Boehnke, K., Hagan, J., & Hefler, G. (1998). On the development of xenophobia in Germany The adolescent years. Journal of Social Issues, 54, 585-601p Breakwell, G. M., & Lyons, E. (1996). Changing European identities: The social psychological analysis of change. Bodmin: Butterworth Heinemann. Brewer, M. (1979). Ingroup bias in the minimal group situation: A cognitive-motivationa 20RR. Pri.linH analysis. Psychological Bulletin, 86, 307-324p Campbell, A. (1981). The sense of well-being in America. New York: McGraw-Hillp Chiu, C.Y., & Hong, Y.Y. (1999). Social identification in a political transition: The role o . implicit beliefs. International Journal of Intercultural Relations, 23, 297-318p Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms, conformity, and compliN ance. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), The handbook of socialHpsychology (pp. 151-192). Boston, MA: McGraw-Hillp Conway, M., & Ross, M. (1984). Getting what you want by revising what you had. JournalHof Personality and Social Psychology, 47, 738-748p Crocker, J., Major, B., & Steele, C. (1998). Social stigma. In D. T. Gilbert, S. T. Fiske, & Gp Lindzey (Eds.), The handbook of social psychology (pp. 504-595). Boston, MA McGraw-Hillp Devine, P. G., Hamilton, D.L., & Ostrom, T.M. (1994). Social cognition: Impact on socialHpsychology. San Diego, CA: Academic Pressp Duckitt, J., & Mphuthing, T. (1998). Political power and race relations in South Africa African attitudes before and after the transition. Political Psychology, 19, 809-832p Dunning, D., Madey, S. F. (1995). Comparison process in counterfactual thought. In N. Jp Roese et al. (Eds.). What might have been: The social psychology of counterfactua thinking (pp. 103-131). Mahwah, NJ: Erlbaump Eagly, A., & Chaiken, S. (1993). Psychology of attitudes. San Diego, CA: Academic Pressp Feng, Y., & Zak, P. (1999). The determinants of democratic transitions. Journal of ConflictH Resolution, 43, 162-177p Festinger, L. (1950). Informal social communication. Psychological Review, 57, 271-282p Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117-140p Finchilescu, G., Dawes, A. (1998). Catapulted into democracy: South African adolescents. sociopolitical orientations following rapid social change. Journal of Social Issues,H 54, 563-583p Fiske, S. T., & Taylor, S. E. (1991). Social cognition. New York, NY: McGraw-Hillp Haeyrynen, Y.P. (1999). Collapse, creation, and continuity in Europe: How do peoplI change. In Y. Engestroem et al. (Eds.). Perspectives on activity theory. Learning inHdoing: Social, cognitive and computational perspectives (pp. 115-132). NY, Ne . York: Cambridge University Pressp Hogg, M.A., & Abrams, D. (1988). Social identifications: A social psychology ofHintergroup relations and group processes. London: Routledge & Kegan Pauop Inglehart, R. (1990). Culture shifts in advanced industrial society. Princeton, NJ: PrincetoU University Pressp Jones, J.M. (1972). Prejudice and racism. Reading, MA: Addison-Wesleyp Kugler, J., & Feng, Y. (1999). Explaining and modeling democratic transitions. Journal ofHConflict Resolution, 43, 139-146p Levine, J. M., & Moreland, R. L. (1998). Small Groups. In D. T. Gilbert, S. T. Fiske, & Gp Lindzey (Eds.), The handbook of social psychology (pp. 415-469). Boston, MA McGraw-Hillp Lundgren, S. R., & Prislin, R. (1998). Motivated cognitive reasoning and attitude changep Personality and Social Psychology Bulletin, 24, 715-726p Macek, P., Flanagan, C., Gallay, L., Kostron, L., Botcheva, L., & Csapo, B. (1998) p Postcommunist societies in times of transition: Perceptions of change amon . Effects of changes in group statusH2b adolescents in Central and Eastern Europe. Journal of Social Issues, 54, 547-561p Markus, G. B. (1986). Stability and change in political attitudes: Observed, recalled anB explained. Political Behavior, 8, 21-44p Moscovici, S. (1976). Social influence and social change. New York, NY: Academic Pressp Moscovici, S. (1994). Three concepts: Minority, conflict, and behavioral style. In Sp Moscovici, A. Mucchi-Faina, & A. Maas (Eds.), Minority influence (pp. 233-251) p Chicago: Nelson-Hallp Moscovici, S., & Faucheux, C. (1972). Social influence, conformity bias, and the study o . active minorities. In L Berkowitz (Ed.), Advances in experimental social psycholog (Vol. 6, pp. 149-202). New York, NY: Academic Pressp Myers, D. G. (1993). The pursuit of happiness. New York: Avonp Nadler, A. (1993). Deviance in primary groups: The social negotiation of personal changep In W. Stroebe & M. Hewstone (Eds.). European Review of Social Psychology (Volp 4, pp. 187-222). Chichester: Wileyp Niemi, G., Katz, R. S., Newman, D. (1980). Reconstructing past partisanship: The failure o . party identification recall questions. American Journal of Political Science, 24, 633N 651p Pettigrew, T., and Tropp, L. R. (1999, February). Does intergroup contact reduce prejudiceE Recent meta-analytic findings. Paper presented at the 16th Annual Claremont SympoN sium on Applied Psychology. Claremont, CAp Prewit, K., & Stone, A. (1973). The ruling elites: Elite theory, power, and American democ-Hracy. New York: Harper & Rowp Prislin, R. (1996). Attitude stability and attitude strength: One is enough to make it stablep European Journal of Social Psychology, 26, 447-472p Prislin, R., Limbert, W.M., & Bauer, E. (1999). From majority to minority and vice versa: TheHasymmetrical effects of losing and gaining majority position with a group. ManuN script under reviewp Prislin, R., & Ouellette, J. (1996). When it is embedded, it is potent: Effects of genera attitude embeddedness on formation of specific attitudes and behavioral intentionsp Personality and Social Psychology Bulletin, 22, 933-948p Prislin, R., & Pool, G.J. (1996). Behavior, consequences, and the self: Is all well that end . well? Personality and Social Psychology Bulletin, 22, 845-861p Prislin, R., Wood, W., & Pool, G.J. (1998). Structural consistency and the deduction of nove from existing attitudes. Journal of Experimental Social Psychology, 34, 66-89p Ross, M. (1989). Relation of implicit theories to the construction of personal historiesp Psychological Review, 96, 341-357p Rothgerber, H., & Prislin, R. (1999). Intergroup dynamics following a change in power.HManuscript under reviewp Schwartz, N., Wnke, M., & Bless, H. (1994). Subjective assessment and evaluations o . change: Some lessons from social cognition research. In W. Stroebe, & M. hewstonI (Eds.). European Review of Social Psychology (Vol. 5, pp. 181-210. Chichester Wileyp Sidanius, J. (1999, February). Social dominance: The interaction between gender andHethnic discrimination. Paper presented at the 16th Annual Claremont Symposium oU Applied Social Psychology, Claremont, CAp Silka, L. (1988). Intuitive judgments of change. New York: Springer-Verlagp 22RR. Pri.linH Simon, B. (1998). Individuals, groups, and social change: On the relationship betweeU individual and collective self-interpretations and collective action. In C. Sedikides, Jp Schloper, & C. A. Insko (Eds.), Intergroup cognition and behavior (pp. 257-282) p Mahwah, NJ: Earlbaup Sloutsky, .M., & Seartle_White, J. (1993). Psychological responses of Russians to rapiB social change in the former U.S.S.R. Political Psychology, 14, 511-526p Tajfel, H. (Ed.) (1982). Social Identity and Intergroup Relations. Paris: Maison des Science . de l.Homme and Cambridge: Cambridge University Pressp Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup conflict. In Wp Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33N 48). Pacific Grove, CA: Brooks/Colep Taylor, D.M., & McKirnan, D.J. (1984). A five-stage model of intergroup relations. BritishH Journal of Social Psychology, 23, 291-300p Turner, J. C. (1991). Social influence. Pacific Grove, CA: Brooks/Colep Tversky, A. (1977). Features of similarity. Psychological Review, 84, 327-352p Tversky, A. (1994). Contingent preferences: Loss aversion and tradeoff contrast in decisioU making. Japanese Psychological Research, 36, 3-9p Tversky, A., & Kahneman, D. (1991). Loss aversion in riskless choice: A reference-dependN ent model. Quarterly Journal of Economics, 107, 1039-1061p Wieczorkowska, G., & Burnstein, E. (1999). Adapting to the transition from socialism t 1 capitalism in Poland: The role of sreening strategies in social change. PsychologicalHScience, 10, 98-105p Williams, R. M., Jr. (1975). Relative deprivation. In L. A. Coser (Ed.), The idea of socialHstructure: Papers in honor of Robert K. Merton (pp. 355-378). New York: Harcour1 Brace Jovanovichp Wood, W., Lundgren, S., Ouellette, J. A., Busceme, S., & Blackstone, T. (1994). Minorit influence: A meta-analytic review of social influence processes. PsychologicalHBulletin, 115, 323-345p