studies of individualism-collectivism: effects on cooperation in groups

8/13/2019 STUDIES OF INDIVIDUALISM-COLLECTIVISM: EFFECTS ON COOPERATION IN GROUPS

1/23

ci idi iny o Manasemenl /ournal Vol. m. Nn, I. [rvi 172.

STUDIES OF INDIVIDUALISM-COLLECTIVISM:EFFECTS ON COOPERATION IN GROUPSJOHN W GNER IIIMichigan State University

Data from 492 college students indicated that group size and individ-uals ' identiability, sense of shared responsibility, and levels of indi-vidualism or collectivism influenced peer-rated cooperation in class-room groups. Levels of individualism or collectivism moderated theeffects of size and identiability on cooperation but not those of sharedresponsibility. These fndings suggest that models of free riding andsocial loafing provide insights into individualistic cooperation ingroups but are limited in their ability to explain the cooperation ofcollectivists.

Cooperation, defined as the willful contribution of personal effort to thecompletion of interdependent jobs, is essential whenever people must coordina te activities amo ng differentiated tasks. Barnard (1938), who was one ofthe first modern organization theorists to recognize this requirementdeemed it crucial to the survival of a social unit that its members induceeach other to behave cooperatively. Other theorists influenced by Barnard(e.g., M arch Sim on, 1958; Sim on, 1947; Th om pson , 1967) later incorporated similar ideas into many of today's archetypal models of organizationand organizational behavior. As a result, cooperation is now a fundamentaconcept in the organization sciences, as evidenced by the articles appearingin this special issue of the Academy 0/ Management Journal.Despite the conceptual importance of cooperation, one must look outside the organization sciences to find long streams of theory and researchconcerning cooperation in the kinds of groups found in current-day organizations. In one such stream, researchers in economics and social psychology have investigated the behavioral patterns termed free riding and socialoafing (Latan, Williams, Harkins, 1979; Sweeney, 1973). Studies in thidomain have sought to identify factors that might curb noncooperative tendencies and instead encourage cooperation in groups (Kerr Bruun, 1983Marwell Ames, 1979). Over a hundred published analyses have resultedIn another area of investigation, a number of social scientists have commented on the individualistic leanings of many current-day theories of social cooperation and human behavior (e.g., Hogan, 1975; Spence, 1985)


2/23

1995 Wagner 153tiescollectivismmight stimulate cooperation in ways not envisioned inmost research (Sampson, 1977, 1978). Others have suggested that modelsbased on individualistic assumptions might prove unable to explain thepersonal or social behaviors of people holding collectivist viewpoints(Lykes, 1985; Shamir, 1990).

During the most recen t decade, a few organization scientists have notedthe relevance of these bodies of research to researchers interested in under-standing the behaviors of individuals and groups in organizations (e.g., Al-banese Van Fleet, 1985; Jones, 1984; Wagner Moch, 1986). In addition,several organizational researchers have borrowed from one or both streamsto pu rsu e various questio ns (e.g., Earley, 1989, 1993; W eldon Gargano,1985).How ever, investigation into cooperation in groups, particularly wherethat cooperation is influenced by variations in group members' levels ofindividualism or collectivism, has yet to develop into a significant domainof organizational research. The present article is intended to help stimulateinterest in such development.

INDIVIDUALISM-COLLECTIVISM AND COOPERATIONIndividualism-collectivism is an analytical dimension that captures therelative importance people accord to personal interests and to shared pur-suits. As defined by Wagner and Moch (1986), individualism is the condi-

tion in w hich personal interests are accorded greater importance than are theneeds of groups. Individualists look after themselves and tend to ignoregroup interests if they conflict with personal desires. The opposite of indi-vidualism, collectivism, occurs when the demands and interests of groupstake precedence over the desires and needs of individuals. CoUectivists lookout for the well-being of the groups to which they belong, even if suchactions sometimes require that personal interests be disregarded.In the social sciences, evidence of the distinction between individual-ism and collectivism can be detected as far back as Aristotle's critique of thecollectivist vision of Plato's Republic, in his own individualist-leaning PoJ-itics (King-Farlow, 1964). Pathbreaking discussions of concepts similar toindiv idua lism and collectivism can also be found in the European sociologyof the late 1800s, whe rein Toenn ies differentiated the geselJscha/t society oftemporary relationships from the gemeinscha/t community of shared obli-gations and irreducible ties (Cahnman, 1973). Similarly, Weber (1947) dif-ferentiated the emergent, associative relationships that flourish in societiesfrom the traditional, communal relationships that thrive in communities. Atabout the same time, Durkheim (1933) distinguished between organic soli-darity, growing out of the necessity for dissimilar specialists to form tempo-rary relationships to perform work requiring broad-ranging skills, and me-


3/23

Aca dem y of Management /ournaJ FebruaryGrowing out of tbese and other classic origins, the distinction betweenself-orientation and collectivity-orientation was introduced to North Amer-ican social scientists by Parsons, who described the former as existing whenindividuals are able to pursue private interests irrespective of their bearingon the interests of others, and the latter as occurring when obligations toward collective well-being are allowed to supersede the pursuit of personagains (Parsons Sh ils, 1951]. Euro pean autho rs also updated tbe d istinctiobetween individualism and collectivism and reintroduced it in the contemporary social sciences to explain behavioral differences among societal cul-tures (Hofstede, 1980] as well as among individuals within a single societacu ltur e (Silverm an, 1971]. In the U nited States, a large body of cross-culturaresearcb developed as researchers compared people from predominately in

dividualistic cultures, such as Australia, Ganada, and the United States, andpredominately collectivist cultures (Japan, Hong Kong, Korea, India, Chinaand Nigeria] in terms of mo tivation (Howard, Sh udo , Um eshima, 1983Hui Villareal, 1989], preferen ces for equity versus equ ality (e.g., BondLeung, Wan, 1982; Hui, Triandis, Yee, 1991], proclivities toward sociinteraction (e.g., Gudykunst, Yoon, Nishida, 1987; Verma, 1985], and similar considerations. A small body of research also developed in which U.Sresea rche rs exam ined th e origins of individ ualistic-co llectivistic differenceswithin a single culture and investigated the effects of these differences onvarious personal and social outcomes (e.g., Breer Locke, 1965; Gox, Lobe McL eod, 1991 ; Lykes, 1985; Wagner Moch, 1986).

Although collectivism as defined in this research might seem similar tocohesiveness, commitment, or conformity, the latter concepts rest on theassumption that person-group relationships are temporary and based onmomentary agreement or passing attraction. In contrast, collectivism is anorientation toward person-group relationships in which such relationshipsare looked at as being far more permanent and central. A heuristic I suggested in earlier research (Wagner, 1982) is useful in appreciating the depthof this di stinc tion : an ind ivid ual ist acts as thoug h he or she defines self as anentity consisting of a single person, bounded by his or her skin, but a collectivist acts as if he or she defines self as an entity extending beyond theindividual to include a particular group of others, bounded by the sociaperimeter of that group. Thus, selfishness for an individualist implies attention to personal pursuits and inattention to group interests, but selfishnesdefined in the manner of a collectivist connotes attention to group interestand inattention to personal desires.Thinking about the implications of these differences in self-definitionsuggests that variations in individualism-collectivism should influence per


4/23

1995 W agner 155more personally satisfying pursuits. Under these circumstances, individu-alists are likely to prefer to avoid cooperation and instead devote their at-tention to the pu rsuit of person al gains. In contrast, cooperation is consistentwith the self-definitions of collectivists who favor the pursuit of group in-terests. In attending to group performance and well-being, collectivists arelikely to seek out and contribute to cooperative endeavors that benefit theirgroups, irrespective of the imm ediate personal implication s of these endeav-ors (Spence, 1985; Wagner, 1982].Evidence supporting this line of reasoning can be found in four cross-cultural an alyses, two by Gabrenya, Latan, and Wang (1981,1983 ], and twoby Earley (1989, 1993). Both studies by Gabrenya and colleagues comparedthe performance of U.S. and Ghinese students, and both studies also com-pared the performance of Chinese students working alone and in groups. Inthe first study, Gabrenya and colleagues (1981) found that transfer studentsin the United States who came from China, a collectivist nation, producedmore working together than working alone. In their second study, however,the researchers discovered that stud ents from China who were tested in theirhom eland prod uced more working alone than working together (Gabrenya etal., 1983]. In con trast to these mixed findings, b oth of Earley s studiesshowed no significant reduction of productivity in groups of Chinese col-lectivists, and a noticeab le am ount of reduction in groups of people from theUnited States, an individualistic nation. Earley attributed these findings tothe positive effects of cultural collectivism on cooperation in groups.Overall, the strongest evidence indicates that members of collectivistnational cultures frequently opt to cooperate in groups, especially whencultural traditionalism is favored over modernity (Hsu, 1970; Yang, 1981)and when they are working in in-groups of close associates (Earley, 1993).but that members of individualistic national cultures show a marked ten-dency to avoid cooperation (Gabrenya et al., 1981, 1983]. To the extent thata similar distinction exists among individuals in a single society, the find-ings of cross-cultural research suggest that ypothesis 1: IndividuaJism-coJIectivism wiiJ influencecooperation in groups in such away that collectivists willcooperate more than will individualists.

FREE RIDING SOCIAL LOAFING AND COOPERATIONFree riding is a choice indiv idu als som etimes make to avoid cooperating

in the pursuit of rewards to be shared by the members of a group, organi-zation, or society, while expecting to derive personal benefit from those


5/23

Academy ofM anagement ournal Februaryriding in that the latter grows out of rational calculation but social loafingcan occur without conscious awareness. As indicated below, however, re-search has suggested that the primary origins of social loafing are motiva-tional. Therefore, free riding and social loafing are analogous in source andeffect: both grow out of the same choice to withhold cooperative effort fromgroup endeavors, and both have the same potential to jeopardize group per-formance and well-being (Kidwell Bennett, 1993).Free Rid ing Effects of Gr oup Size

Analyses of free riding can be traced to Samuelson (1954), who firstidentified various problems inherent in providing societies with collectiveco nsu m ptio n goods, now called public goods. Such goods can be consumedby all societal members if they are available and denied to none because oftheir intrinsic indivisibility. Public goods include clean air and water, acommon language, and a shared government. They also include the rewardstructure and stable employment shared by the members of an organization,the friendships and interd epe nde nce shared by the members of a group, thesense of security and common identity shared by the members of a family,and so forth.As Olson (1965) noted, a chief implication of Samuelson's work is that

pu blic goods in vite free ridin g, since their in divisibility makes it possible forfree riders to derive benefit without personal cost. Yet this property threat-ens the provision of public goods, because people who might otherwisedevote effort to their acquisition are likely to choose instead to free ride. Forsome individuals, this choice is a matter of rational maximization, sincebeing able to con sum e pu blic goods withou t con tributing personal resourcesenables them to direct untapped resources elsewhere and gain even more forthemselves. For other individuals, the choice to free ride grows out of aph en om en on tha t Kerr (1983) labeled the suck er effect : if surrou nde d byothers likely to free ride, an individual not otherwise inclined may alsochoose to free ride in order to avoid the inequity of contributing more thanthe others for the same share of public goods. For both of these reasonswithout corrective inducements public goods are unlikely to be obtainedand everyone must do without (e.g., Olson, 1965).

Olson's observation stimulated research aimed at learning how socialunits might best stimulate cooperation, thereby controlling free riding andinsuring the continued availability of public goods. Much of this researchhas focused on the effects of group size on free riding, because the presenceof large numbers of co-actors can shield an individual's free riding andeliminate fears of corrective retributions but few co-actors make behaviors


6/23

1995 W agner 157

Am es, 1979; Tillock & M orrison, 1979), and question s have been raisedabout the conceptual framework underlying Olson's original prediction(Cham berlin, 1974; Sm ith, 1975; Sweeney, 1973). Despite these reservations,research on free riding is normally interpreted as showing that group sizeplays at least a modest role in shaping an individual's choice to engage incooperative behaviors (e.g., Alban ese Van F leet, 1985; Jones, 1984). Th us,Hypothesis 2: Group size will influence cooperation ingroups in such a way that the members of small groupswiJJ cooperate more than will the members of largegroups

Soc ial Loafing Effects of Identifiability and Shared R espon sibilityResearch on social loafing originated with an experiment by Ringel-mann (1913) in which individuals were compared with groups on a rope-pulling task (Kravitz & M artin, 1986). The R ingelmann experiment laterstimulated a study by Ingbam, Levinger, Craves, and Peckham (1974) thatprompted additional interests among social psychologists. In the resultinganalyses (e.g., Ha rkins, Latan, W illiams, 1980; Kerr Bruun, 1981;Latanet al., 1979) researchers found that increasing the number of workers per-forming a task reduced the average individual effort devoted to task perfor-

mance . The label *'social loafing was applied to this effect.Because Ingham and colleagues (1974) ruled out coordination difficul-ties as a primary source of social loafing, later research has focused insteadon mo tivational origins. Some of this research has suggested that tenden ciestoward social loafing are influenced by identifiability (also called observ-ability, anonymity, accountability, and task visibility), which involves thedegree to which others can observe and assess an individual's behaviors(Ceorge, 1992; Harkins & Szym anski, 1988; Szymanski & Hark ins, 1987).Researchers have considered a variety of masking agents, including groupsize, as discussed above, personal deceit, and the absence of a reliable per-formance assessment mechanism. Like those investigating free riding, theyhave found that the ability to mask personal behavior encourages socialloafing (Ceorge, 1992; Harkins & Petty, 1982; W eldon & Cargano, 1988).Thus,

Hypothesis 3: Identifiability will influence cooperation ingroups in such away thatmembers perceiving themselvesas having a high level of identi/iabiJity wilJ cooperatemore than those perceiving a low level of identi/iabiJity.Research has also suggested that shared responsibility (also called felt


7/23

Acade my of Management Journal Februaof the responsibility for group performance. Members with feelings of rduced personal responsibility thus feel personally dispensable, believinthat the group can succeed at its work without their personal input (Weldo Mustari, 1988). In contrast, members retaining a sense of personal resposibility feel that their behaviors can make the difference between succesand failure for their group. In this manner, strong feelings of shared responsibility reduce personal proclivities to engage in cooperation, and weak feeings of shared responsibility may encourage such proclivities (Fleishman1980;W eldon Cargano, 1985; W eldon Mustari, 1988). Thus,

Hypothesis 4: Shared respo nsibility will influence co op-eration in groups in such a way that members perceivingthemselves to have a low level of shared responsibiJitywill cooperate more than mem bers perceiw ng a high Jevelof shared responsibiJiiy.

A MODEL OF COOPERATION IN GROUPSTo recap itula te, researc h suggests that differences among people in levof individualism or collectivism are likely to affect their cooperation igrou ps, with greater collectivism stim ulating greater cooperation. Studies ofree riding and social loafing further suggest a model in which small grou

size, high identifiability, and low shared responsibility encourage cooperation.Speculation also suggests that differences described by the individuaism-collectivism dimension will moderate the degree to which group sizidentifiability, and shared responsibility influence cooperation becausthos e variables are all expec ted to affect coop eration by curtailing peopleproclivities to pursue personal interests, a behavioral tendency likely to bexh ibited by ind ivid ua lists but not by coUectivists. CoUectivists are likely cooperate for reasons of collectivismthe definition of self they adopt anthe priority they give to group well-beingirrespective of the effects ogroup size, identifiability, or shared responsibility. Consistent with this oservation, I suggestHypothesis 5: ndividuaJism-coJJecivism will exert mod-erating effects in such a way that group size, identi/iabiJ-ity and shared responsibility will have greater influenceon the cooperation of individualists than on the cooper-ation of coUectivists.

The purpose of this study was to assess the model, diagrammed


8/23

1995 Wagner 159FIGURE

Model of Hypothesized Relationships^

Individualism-Collectivism + )Group Size (-]Identifiability (+)SharedResponsibility (-)

Cooperation

aTbe signs in parentheses indicate the expected direction of zero-order relationshipsbetween independent variables and cooperation.

three assessing individualist-coUectivist values, and four tapping individ-ualist-collectivist norms. The second measure, by Erez and Earley (1987), isbased on the work of Hofstede (1980] and is a single scale made up of fouritems measuring individualist-coUectivist cultural values. The third mea-sure, by Triand is and colleagues (Triandis, Bontempo, Villareal, Asai, &Lucca, 1988), includes a first dimension of 12 items assessing self-relianceand competitiveness, a second dimension of 10 items from Hui (1988) thattap concern for others in an in-group (e.g., friends, family, community), anda third dimension of 7 items assessing differentiation between individualsand their in-groups.No single study has compared the three measures. Therefore, it is un-clear whether they overlap enough to be considered synonymous, or if in-stead they access distinctly different aspects of individualism-collectivismand should be interpreted independently. If the measures are independent,it is not readily apparent which of them taps the aspects of individualism-collectivism having the kinds of effects hypothesized here. To deal withmeasurement issues of this sort, I conducted a factor analysis of the threemeasures and derived a multidimensional measure that was then used toconduct hypothesis tests.

METHODSData for this study were collected from undergraduate students enrolled

in an introductory management course at a large midwestern university. All541 students in the course were contacted at the beginning of a ten-week


9/23

l ** cademy o Management Journal Februaryiors, 92 were college seniors, and 6 did not classify themselves; and theiraverage self-reported grade point averagewas 2.9 (4.0 was the highest levels.d. - 0.4].Task ndMeasuresAll participants signed waivers permitting access to all performancemeasures and grades recorded during the term. Participants also completedan initial questionnaire consisting of individualism-collectivism items anddemographic questions. Fulfilling a normal part of the management coursestudents then formed case analysis groups andprepared anoral case reporfor presentation during the term.

Preparing for the presentation typically took 18-20 hours of s tudentsworking together closely as a group outside of class to interpret the caseapply course material, formulate a problem statement, anddevise a suitablesolution. Substantial interaction among group members was required tocomplete the preparat ion, and s tudents developed strong expectationsamong themselves thatall group members would participate in this processPresenting the case report required further interaction, although of limitedduration and intensity.

Immediately following presentation of their g roup s report, each groupof students completed peer assessments that werean established part of thegrade structure of the course. Near the end of the term, participants com-pleted a second q uestion naire that contained individualism-collectivismitems plu s questions about identifiability, shared responsibility,and variousgroup attributes. From these sources, the following data were obtained.Cooperation Cooperation was measured by a peer assessment instrument consisting of a single page beginning with this paragraph:

You have 100 points to allocate among the members of yourgroup according to their performance as group members duringthe preparation and presentation of your group s case analysis.Please allocate pointsina manner reflecting the degree of coop-erative effort exerted by each individual during the entire periodof time your group worked on its presentation.

Following this introduction were blank lines that permitted students to listhe members of their group, including themselves, and to allocate pointamong listed members.W ithin groups, agreement about each indiv idu al s ratingwas quite highas indicated by a reliability estimate of .91 (James, Dem aree, Wolf, 198Nunnally, 1978].I standa rdized this m easure using z-transformation to compensate for the effects of different groups sizes (before standardization, ascore of 50 indicated average performance by a member of a two-person


10/23

1995 WagnerGroups ranged from two to eight members, and sizes were coded withoutfurther transformation.Identifiability. Three items in the second questionnaire measured selfreported iden tifiability: M y behav iors as a group member were readily ob-servable to others in the gro up , Oth ers in the group could not tell wh etherI was doing wha t I was suppose d to do , and In the group, each membercould tell wh ether other mem bers were doing their fair share. Participantsrecorded their responses on seven-point Likert scales ranging from 1, strongly disagree, to 7, strongly agree. The second item was reverse-coded during scaling so that high ratings indicated high identifiability.Shared responsibility. Three items in the second questionnaire mea-sured perc eptio ns of shared resp onsibility: Th e mem bers of the groupshared the responsibility for getting things do ne , I felt persona lly respon -sible for the produ ctivity of the gro up , and Mem bers of the group some-times didn't feel individually responsible for the performance of the groupas a w ho le. Participants recorded their responses on the same Likert scaleused for identifiability, and the second item was reverse-coded so that highratings indicated high shared responsibility.Individualism collectivism. Measures of personal differences in indi-vidualism-collectivism were constructed from 43 items that appeared inboth questionnaires. Of these items, 10 were from Wagner and Moch (1986),4 were from Erez and Earley (1987), and 29 were from Triandis and col-leagues (^Triandis et al., 1988; Hui, 1988). Participan ts reco rded responses toall items on the same seven-point Likert scale described above. Item re-sponses were reversed as needed so that high ratings indicated strongercollectivism.Factor analysis of data from the first questionnaire revealed an 11-factorsolution that was reduced to 5 factors on the basis of a scree test of factoreigenvalues. Varimax rotation produ ced the factor weights shown in Table 1.As the table shows, factor 1 consisted of one item from Erez and Earley andfour items from Triandis and colleagues that assessed personal indepen-dence and self-reliance; factor 2 incorporated five items from T riand is andcolleagues that addressed the importance accorded to competitive success;factor 3 included two items from Wagner and Moch and one item from Erezand Earley that conc erned the value a ttached to working alone; factor 4 wasmade up of the four items from Wagner and Moch that measured espousal ofnorms about the subordination of personal needs to group interests; andfactor 5 consisted of the three items from Wagner and Moch that assessedbeliefs about the effects of personal pursuits on group productivity.

Factor analysis of data from the second questionnaire replicated thesame five-factor solution. The five individualism-collectivism measures


11/23

cademy ofManagement Journal February

TABLE 1Factor Analysis of Individualism-Collectivism ItemsCollectivism Factors

Items 1 2 3 4 51. Only those who depend on themselves getahea d in life*' .7922. To be sup erior a person mu st stan d alone'^ .4583. If you want son:iething done right, you've gotto do it yourself .6194. W hat ha pp en s to me is my ow n doing*^ .4075. In the long run the only person you can counton isyourself .6906. W inn ing is everything*^ .7387. I feel that winn ing is imp ortant in hoth work

and games' .7388. Succ ess is the most imp ortant thing in life^ .6959. It annoys me when other people perform betterthan 1 do'' .50310. Doing your hest isn't enough; it is important towin' ' .68111 . I prefer to work with others in a group rathertha n wo rkin g alone*^ .79912 . Given the choice, I would rather do a johwhere I can work alone rather than doing a joh

wh ere I have to work with others in a group .80513 . Working with a group is hetter than workingalone' ' .80414 . People should be made aware that if they aregoing to be part of a group then they aresometimes going to have to do things theydo n't wa nt to do^ .68415 . People who belong to a group should realizethat they're not always going to get what theyperso nally want^ .65216. People in a group should realize that theysometimes are going to have to make sacrifices

for the sake of the group as a whole^ .75217. People in a group should be willing to makesacrifices for the sake of the group'swell-being*' .70418 . A group is more productive when its membersdo what they want to do rather than what thegroup wa nts them to do^ .7319. A group is most efficient when its memhers dowhat they think is best rather than doing whatthe group wa nts them to do^ .7520 . A group is more productive when its membersfollow their own interests and concerns* .75

Facto r eig env alue 5.35 2.63 2.53 1.83 1.62


12/23

1995 Wagner 163year in school (junior or senior)were included as control variables tosuppress the effects of prospective rater biases on peer-rated cooperation.Also includ ed was the grade po int average earned in prior classes, to controlfor gross differences in ability. Table 2 reports means, standard deviations,reliability coefficients, and correlations pertaining to all the measures usedin this study.Design and nalysis

I performed mo dera ted regression proced ures (Stone Hollenbeck,1984, 1989) to control for the effects of exogenous demographic differencesarising while assessing the main and moderator effects on cooperation ofgroup s ize, identif iabil i ty , shared responsibil i ty , and individualism-collectivism. In these procedures, the control variables of age, gender, race,year in school, and grade point average were regressed against cooperationin the first step of a hierarchical regression analysis. Next, I entered groupsize, identifiability, shared responsibility, and the five dimensions of indi-vidualism-collectivism to assess the main effects remaining. Then, first-order interactions were entered in a third step to examine the simple mod-erator effects present after the elimination of control and main effects. Theregression analysis concluded with the entry of higher-order interactionsthat might have additional explanatory power. I assessed the statistical sig-nificance of the block of variables entered in each step with an F-test of thechange in R^. Within significant blocks of variables, I appraised the statis-tical significance of the effect of each variable with a t-test.

RESULTSResults of the m oderated regression analysis, shown in Table3,revealedthat the five control variables had a statistically significant joint effect andthat threeage, race, and grade point averagehad their own statistically

significant effects on ratings of cooperation. Whether because of rater biasesor actual differences in behavior, participants rated older students, whitestude nts, and s tud ents with higher grade point averages as more cooperative.Regression analysis results pertinent to Hypotheses 1-4 indicated thatthe block of main effects had a statistically significant joint effect on coop-eration. Within this block, group size, identifiability, shared responsibility,and one of the five measures of individualism-collectivism, collectivism 1,had statistically significant effects on cooperation after effects of the controlvariables were accounted for. In support of Hypotheses 1-4, all these effectswere in the expected direction: high collectivism, group smallness, highidentifiability, and low shared responsibility all contributed to high coop-


13/23

164 AcademyofManagement ournal February

u

COa;t-ioU

CO C l

m If) r-t COo O o o

S

5CD

5o

JS *M o o CO O 12C^ in CO o rH o

cT eno o 1 o cvo < ; o

q q\ \ I

CO lOo o

CDo

O

OO

CO

CMCM

C O C M C O C O r H C M r H Oo o q q ' - | q q ^I T Io o o o o o o o

CM

o i-iCM r iI I

cn oo t-t

CO m

O oo o

eo

eo o

CO T-o o

o m

es] oen p

1 1

eu

aid

O-1comccd

a3 o


14/23


15/23

166 Academy of Management ournal FebruaryFIGURE 2

Plots of Statistically Significant First-Order InteractionsCollectivism 1 by Grou p Size

Cooperation 0

5Small Group Size

Collectivism by Identifiability5

Gooperation 0

- . 5Large Low

C

Identifiability High

Collectivism 2 by Identifiability.5

Cooperation 0

- . 5

Gollectivism 3 by Group Size.5

Cooperation 0

- . 5

N. C

I\Low HighIdentifiability Small LargeGroup Size

In each figure, the individualist trend is marked 1 and the collectivist trend is marked C .dence of statistically significant effects. The reduced model consisting onlyof statistically significant control variables, main effects, interactions, andstatistically nonsignificant main effects appearing in statistically significantinteractions, yielded anR of .19 (F^g478 ^ 7.98, p < .01).

DISCUSSIONConsistent with the findings of most prior research on free riding and


16/23

1995 W agner 167reliance has a direct effect: individualists who feel independent and selfreliant are less apt to engage in cooperative behavior, and collectivists whofeel interdependent and reliant on groups are more likely to behave coop-eratively. The same aspect of individualism-collectivism also moderates re-lationships between group size, identifiability, and cooperation in such away that group size and identifiability have greater effects on the coopera-tion of individualists than they do on the cooperation of collectivists. Asecond aspect of individualism-collectivism, differences in the importanceattached to competition and personal success, also moderates the identifi-ability-cooperation relationship, and a third aspect, differences in the valueplaced on working alone, moderates the relationship between group size andcooperation. In both of these interactions, high collectivism reduces theinfluence of identifiability and size on cooperation.In contrast to these findings supporting predictions, the present studyfailed to show that differences in individualism-collectivism among U.S.college students moderate the relationship between shared responsibilityand cooperation. One interaction, between collectivism 3 and shared re-spon sibility, cam e close to attain ing statistical significance p < .10), sug-gesting that improved measurement or perhaps other methods might pro-duce findings supportive of hypothesized expectations. However, it is alsopossible that the null results of this study signal the existence of an addi-tional, untapped dimension of individualism-collectivism that has the ex-pected moderator effects. Or the results of this study might accurately reflectthe absence of similar effects among Americans in general. Further analysisis required to weigh these possibilities.Although college students served as the participants in the presentstudy, the use of an ordina ry classroom assignment a graded presentation],a measure of cooperation that was perceived as routine [a customary peerevaluation), and a task that required sustained in terdep end ence a groupproject) seem to permit cautious generalization to other settings in whichinterdependent tasks are performed in groups. Allowed such generalization,the results of this study hold important implications for research on indi-vidualism -collectivism and g roup coop eration. One of these grows out of theresults of the factor analyses, which indicate that the three measures byWagner and M och 1986), Erez and Earley 1987), and Triandis and col-leagues 1988) are neith er entirely synonym ous nor completely indepen-dent. At a minimum, this finding suggests that researchers should exercisecaution when attempting to reach cumulative conclusions on the basis ofempirical studies that have used different measures of individualism-collectivism. It also provides support for the recommendation that multipledimensions be measured in future research intended to assess the anteced-


17/23

Academy o M anagement Journal Februarybehavior. Central to Sampson's criticism has been the assertion that currentsocial theories developed in the United States only partially represent thefull range of human variability. Consistent with this assertion, the presentstudy indicates that conceptualizations of cooperation that are based on freerider or social loafing models or on similar treatments of the social psychol-ogy of group behavior are limited in their applicability to collectivists.

This finding suggests the development of a more balanced stream ofresearch on cooperation, perhaps one growing out of the realization that freeriding and social loafing in the purest form represent individualism pursuedto an extreme. What would a model of group cooperation look like if it alsoincorporated a description of collectivism pursued to an extreme? Perhaps itwould include the possibility that overzealous collectivist concerns aboutgroup well-being could stimulate overcontribution to group endeavors, pre-maturely exhausting the resources of extremist members. If widespreadamong the membership of a group, such overcontribution and exhaustionmight undermine the group's ability to look after its members, ultimatelythreatening the group's performance and continued survival. In such situa-tions, the cooperative response might not involve making personal contri-butions to group endeavors but might instead consist of withholding someminimal reserve of effort in order to conserve one's resources and ensurelong-term group viability.

Redirected research on cooperation might also deal with the paradoxthat collectivists whose cooperation outpaces the contributions of others intheir groups are distinguishing themselves from their colleagues and becom-ing less coU ectivistic as a result. Paraph rasing a collectivist adage, being thenail that sticks ou t, even for reasons inten de d to benefit group well-being,carries the liability of need ing to be ha m m ered back in to preserve thecollectivist sense of self and group. Analyzing how groups of coUectivistsdeal with this problem might suggest varieties of surveillance, control, andsense-making that influence collectivist cooperation in much the same waythat factors such as group size, identifiability, and shared responsibilityshape individualist behaviors.In conclusion, this study substantiates the idea that variation in indi-vidualism-collectivism can have effects, within a single societal culture, oncooperation in groups, and that these effects can extend and modify theinfluence of factors often analyzed in research on free riding and socialloafing. Do collectivist self-definitions ind uce a significant n um ber of work-ing Am ericans to cooperate w ith one another? Might work place collectivismthus supplement or even supplant many of the management procedurespresently used to encourage cooperation in U .S. organizations? Do problems


18/23

1995 Wagner 169REFERENCES

Albanese, R., Van Fleet, D. D. 1985. Rational behavior in groups: The free riding tendency.Academy of Management Review 10: 244255.Alfano. G., & Ma rwell, G. 1980. Exp erimen ts on tbe provision of public goods III: Non-divisibility and free riding in rea l groups. Social Psychology Quarterly 43: 300-309.Barnard, C. I. 1938. hefunctions of the executive. Cambridge, MA: Harvard University Press.Bond, M . H.. Leung, K., Wan, K. C. 1982. How does cultural collectivism operate? The impactof task and maintenance contributions on reward distribution. Journal of Cross-CulturalPsychology 13: 186-200.Breer, P . E., & Locke, E. A. 1965. Task experience as a source of attitudes. Homewood, IL:Dorsey Press.Cahnm an, W. J. 1973.Eerdinand Toennies: A new evaluation. Leiden, Netherlands: Brill.Chamberlin, J. R. 1974. Provision of public goods as a function of group size.AmericanPolit-ical Science Review 68: 707-716.Chamberlin. J. R. 1978. The logic of collective action: Some experimental results. BehavioralScience 23: 441-4 45.Cox, T. H., Lobel, S. A., McLeod, P. L. 1991. Effects of ethnic group cultural differences oncooperative and com petitive behavior on a group task.Acodemy of M anagement Journal34:827-847 .Darley, J. M., Latan, B. 1968. Bystander interven tion in emergencies: Diffusion of respons i-bility. Journal of Personality and Social Psychology 8: 377-38 3.Durkheim, E. 1933.The division of labor in society. New York: Free Press.Earley, P. C. 1989. Social loafing and collectivism: A comparison of the United States and thePeople's Republic of China.Adm inistrative Science Quarterly 34:565-581.Earley, P. C. 1993. East meets West meets Mideast: Further explorations of collectivistic andindividualistic work groups.Academy of ManagementJournal 36: 319-348 .Erez, M., Earley, P. C. 1987. Comparative analysis of goal setting strategies across cultures.Journal of Applied Psychology 7 :658-665 .Fleish ma n. J. A. 1980. Collective action as helping bebav ior: Effects of responsib ility diffusionon contributions to a public good. Journal of Personality and Social Psychology 38:

629-637 .Gabrenya, W. K., Jr., Latan, B., Wang, Y. E. 1981.Social loafing am ong Ch inese overseasand U.S. students. Paper presented at the Second Asian Conference of the InternationalAssociation for Cross-Cultural Psychology, Taipei, Taiwan, R.O.C.Gabrenya, W. K., Jr., Latan, B., Wang, Y. E. 1983. Social loafing in cross-cultural perspective.Journal of Cross-Cultural Psychology 14: 368-384.George, J. M. 1992. Extrinsic and intrinsic origins of perceived social loafing in organizations.Academy of Management Journal 35: 191-202 .Gudykun st, W. B.. Yoon, Y. C , Nishida, T. 1987. The influence of individualism-collectivism

on perceptions of communication in ingroup and outgroup relations. CommunicationMonographs 54: 295-306.


19/23

170 Academy ofManagement /ourna] FebruaryHofstede, C. 1980.Cultures consequ ences: International differences in work related values.Beverly H ills, CA: Sage.Hogan, R. 1975. Theoretical egocentrism and the problem of compliance. American Psychol-ogist 30: 533-54 0.How ard. A., Shud o, K., & Um eshim a. M. 1983. Motivation and values among Japanese andAmerican managers. Personnel Psychology 36: 883-898.Hsu F. L. 197 0.Am ericans and C hinese: Purp ose an d fulfillment in great civilization. Garden City. NY: Natural History Press.Hui. C. H. 1988, Measurement of individualism-collectivism. Journal of Research in Person-ality 22: 17-36 .Hui.C. H., Triand is. H. C, Yee, C. 1991, Cultural differences in reward allocation: Is collectivism the explanation? British Journal of Social Psychology 30: 145-157.Hui, C. H. Villareal. M. J. 1989. Individualism-collectivism and psychological needs. ournof Cross-Cultural Psychology 20: 310 -323.Ing ham . A.. Levinger, G., Graves, J., & Peck bam , V. 1974. Th e R ingelm ann effect: Studies ofgroup size and group performance. Journal of Personality and Social Psychology 10:371-384 .Isaac,R. M,. Walker, J. M. 1988. Croup size effects in public goods provision: The voluntarycontributions mechanism. Quarterly Journal ofEconomics 103: 179-199.Jam es. L. R.. Demaree, R. G., Wolf G. 1984. Estimating w ithin-group interrater reliability w ithand without response bias. Journal of Applied Psychology 69: 85-98.Jones, G. R. 1984. Task vis ibility, free riding , and sh irking: Exp laining the effect of structure andtechnology on employee behavior. Academy of Management Journal 9: 684-695 .Kerr. N, 1983. Motivation losses in small groups: A social dilemma analysis. Journal of Per-sonality and Social Psychology 45: 819-82 8.Kerr, N.. Bruu n. S. 1981. Ringelm ann revisited: A lternative explanation s for tbe social loafineffect. Personality and Social Psychology Bulletin 7: 224-231.Kerr, N. L., & Bru un, S. E. 1983. The dis pen sab ility of m ember effort and group motivationlosses: Free rider effects. Journal of Personality and Social Psychology 44: 78-94 .Kidwell, R. E., Jr., Bennett, N. 1993. Employee propensity to withhold effort: A conceptua

model to intersect three avenues of research. Academy of Management Review 18: 429 -456.King-Farlow, J. 1964. The concept of m in e. Inquiry :268-276.Kravitz, D. A., & M artin. B. 1986. Ringelmann rediscovered: The original article. Journal ofPersonality and Social Psychology 50:936-941 .Latan, B .. W illiams, K., & Hark ins, S. 1979. Many ha nds make light work: Tbe causes andconsequences of social loafing. Journal of Personality and Social Psychology 37:8 2 2932.Lykes, M. B. 1985. Gender and individualistic vs. collectivist bases for notions about theself.

Journal of Personality 53: 356 -383 .March, J. B,, Simon, H. A. 1958. Organizations. New York: Wiley.


20/23

1995 Wagner 171Nunnally, J. G. 1978. Psychometric theory {2 ed.}. New York: McG raw-Hill.Olson, M. 1965.The log ic af collective action: Public good s and the theo ry of group s. Cam-bridge, MA: Harvard University Press.Parsons, T., Shils, E. A. 1951. Toward a general theory of action.Cambridge, MA: HarvardUniversity Press.Ringelmann, M. 1913. Recherches sur le moteurs anims: Travail de l'homme. Annales de

nstitute National Agronomique 2(12): 1-40.Sam pson, E. E. 1977. Psychology an d the Am erican ideal. Journal of Personality and SocialPsychology 35: 767-782.Sampson, E. E. 1978. Scientific paradigms and social values: WantedA scientific revolution.Journal of Personality and Social Psychology 36: 1332 -1343.Sampson, E. E. 1988. The debate on individualism: Indigenous psychologies of the individualand their role in personal and societal functioning. American Psychologist 43: 15-22.Samuelson, P. 1954. The pure theory of public expenditure. Review of conomics and Statis-tics 37: 387-389 .Shamir, B. 1990. Calculations, values, and identities: The sources of coUectivistic work moti-vation. Human Relations 43: 313-332.Silverman, D. 1971.The theory of organizations. New York: Basic Books.Simon, H. A. 1947.Adm inistrative behavior. New York: Free Press.Smith, J. 1975. Com mun ities, associations, and the sup ply of collective goods.Am ericanJournal of Sociology 82: 291-308.Spence, J. T. 1985. Achievement American style: The rew ards and costs of individualism .American Psychologist 40: 1285-1295.Stone, E. F., Hollenbeck, J. R. 1984. Some issues associated with the use of moderated regres-sion. Organizational Behavior and Human Performance 34: 195-21 3.Stone, E. F., Hollenbeck, J. R. 1989. Clarifying some controversial issues surrounding statis-tical procedures for detecting moderator variables: Empirical evidence and related matters.Journal of Applied Psychology 74: 3-10.Sweeney, J. W. 1973. An experimental investigation of the free-rider problem. Social Science

Research 2: 277-292.Sweeney,J W. 1974. Altruism, the free rider problem, and group size. Th eory and Decision 4:

259-275 .Szymanski, K., cH arkins , S. G. 1987. Social loafing and self-evaluation with a social standa rd.Journal of Personality and Social Psychology 53: 891-8 97.

Thom pson, J. D. 1967. Organizations in action.New York: McGraw-Hill.Tillock, H., Morrison, D .E. 1979. Group size and contribution s to collective action: Anexamination of Olson's theory using data from Zero Population, Inc. In L. iCriesberg [Ed.),Research in social move ments conflicts and chang e vol. 2: 131-158. Greenwich, CT:JAI Press.Tria ndis , H. G., Bo ntem po, R., Villareal, M., Asai, M., Lucca, N. 1988. Indiv idualism andcollectivism: G ross-cultural perspectives on self-ingroup relationships.Journ al of Person-ality and Social Psychology 54: 323-338 .Verma, J. 1985. The ingroup and its relevance to individual behavior: A study of collectivism


21/23

Academy ofManagement Journal FebruaryWfbor. M. n)'17. The theory of social and economic organization. New York: Free Press.Weber, M. 1958. The protestant ethic and the spirit of capitalism. New York: Scribner's.W eld on . E., Ca rgan o. G. M. 1985. Cog nitive effort in additive task s: The effects of sharedresponsibility on tbe quality of multiattribute judgments. Organizational Behavior and

Human Decision Processes 36 :3 4 8 3 6 1 .Weldon, E.. Gargano, G. M. 1988. Cognitive loafing: The effects of accountability and sharedresponsibility on cognitive effort. Personality and Social Psychology Bulletin 14: 159-171.Weldon, E., M usta ri, E. L. 1988. Felt dis pen sab ility in groups of coactors : The effects of sharedresponsibility and explicit anonymity on cognitive effort. Organizational Behavior andHuman Decision Processes 41:330-351 .Yang. K. S. 1981. Social orientation and indiv idua l mod ernity among Chinese studen ts inTaiwan. Journal of Social Psychology 113: 159-170 .

John A Wagner III received his Ph.D. degree in business administration from theUniversity of Illinois, Urbana. He is an associate professor of management at MichiganState University. His research interests concern tbe effects of differences in direction-participation and individualism-collectivism on coordination and behavior in groupsand organizations.


22/23

1995 Volume 38 Number 1 173ACKNOWLEDGMENT OF REVIEWERS FORSPECIAL RESEARCH FORUM ON COOPERATION

Deborah AnconaPhilip AndersonSamuel BacharachWayne BakerMax BazermanSchon BeechlerJanice BeyerAllan BirdWarren BoekerDaniel BrassRudi BresserJoel BrocknerPhilip BromileyJohn ButlerAlbert CannellaPaul CollinsEdward ConlonRichard D AveniAndre DelbecqGregory DessRobert DrazinIrene DuhaimeJane DuttonJeffrey EdwardsKathleen EisenhardtMiriam ErezSydney FinkelsteinRobert FolgerCharles FombrunJames FredricksonMartin GannonRaghu GarudMichael Geringer

Marilyn GistCurtis GrimmAnil GuptaRichard GuzzoJerry HageKathryn HarriganJan HeideGharles HillRobert HoskissonDuane IrelandVirod JainJose Carlos JarilloKatherine KleinRichard KopelmanDavid KrackhardtRoderick KramerRobert KrapfelGuido KricKxRajesh KumarMichael LawlessGarrie LeanaDavid LeiDorothy Leonard BartonJames LincolnEdwin LockeJohn MichelGeorge MilkovichAlex MillerAneil MishraTerence MitchellRichard MowdayMargaret NealeGregory Northcraft

Judy OlianChristine O liverBen OviattArvind ParkheRobin PinkleyRichard PriemKeith ProvanVasudevan RamunajamDavid RavenscraftElaine RomanelliHarry SapienzaCarl ScheragaBenjamin SchneiderClaudia Bird SchoonhovenMichael SchusterSaniit SenguptaPaul ShrivastavaHenry SimsKen A. SmithCharles SnowRobert SpekmanGretchen SpreitzerTimothy StearnsCindy StevensDon StockdaleSusan TaylorDean TjosvoldBeverly TylerStefan WallvJames WalshGregory YoungEdward ZajacDale Zand


23/23

Copyright of Academy of Management Journal is the property of Academy of Management and its content may

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.

studies of individualism-collectivism: effects on cooperation in groups

Documents