Editorial Staff Editor P. RAJAN VARADARAJAN JAMS Department of Marketing Texas A&M University 4112 TAMU College Station, TX 77843-4112 Phone: (979) 862-1019 Fax: (979) 862-1020 E-mail: [email protected]Book Reviews PEGGY H. CUNNINGHAM Queen’s University School of Business 229 Dunning Hall 99 University Avenue Kingston, Ontario K7L 3N6 Canada Phone: (613) 533-2327 Fax: (613) 583-2321 E-mail: pcunningham@ business.queensu.ca Marketing and the Law ANITA CAVA ANN MORALES OLAZÁBAL RENÉ SACASAS Business Law Department School of Business Administration University of Miami P.O. Box 248022 Coral Gables, FL 33124 Phone: (305) 284-4633 Fax: (305) 284-3762 The Academy of Marketing Science is a member of the International Association for Management Education (AACSB). It gratefully acknowledges the financial support of the Mary Kay Cosmetics Excellence in Marketing Fund. The JOURNAL OF THE ACADEMY OF MARKETING SCIENCE is the official journal of the Academy of Marketing Science. It is an international, refereed journal intended to further the science of marketing throughout the world by promoting the conduct of research and the dissemination of research results through the study and improvement of marketing as an economic, ethical, and social force. For manuscript submission information, refer to the inside back cover. *Founding Fellow MANOJ K. AGARWAL Binghamton University CHRISTIE H. AMATO University of North Carolina–Charlotte JONLEE ANDREWS Indiana University KWAKU ATUAHENE-GIMA City University of Hong Kong RICK BAGOZZI Rice University SHARON BEATTY University of Alabama DAN BELLO Georgia State University *HAROLD W. BERKMAN University of Miami SUNDAR BHARADWAJ Emory University DAVID M. BOUSH University of Oregon JAMES R. BROWN Virginia Tech STEVEN P. BROWN Southern Methodist University TOM J. BROWN Oklahoma State University MICHELE BUNN University of Alabama ALAN BUSH University of Memphis ROGER CALANTONE Michigan State University JOSEPH P. CANNON Colorado State University GOUTAM CHAKRABORTY Oklahoma State University GOUTAM CHALLAGALLA Georgia Institute of Technology RAJESH CHANDY University of Minnesota BRUCE CLARK Northeastern University JOSEPH COTE Washington State University DAVID W. CRAVENS Texas Christian University PEGGY H. CUNNINGHAM Queen’s University MICHAEL R. CZINKOTA Georgetown University PRATIBHA DABHOLKAR University of Tennessee PETER A. DACIN Queen’s University PETER DICKSON Florida International University PAM SCHOLDER ELLEN Georgia State University ELLEN GARBARINO Case Western Reserve University JIM GENTRY University of Nebraska RONALD C. GOODSTEIN Georgetown University EVERT GUMMESSON Stockholm University GREG GUNDLACH University of Notre Dame CHRISTIAN HOMBURG University of Mannheim G. TOMAS M. HULT Michigan State University MICHAEL HYMAN New Mexico State University CHARLES A. INGENE University of Mississippi JEAN L. JOHNSON Washington State University SUSAN KEAVENEY University of Colorado at Denver AMNA KIRMANI Southern Methodist University MASAAKI KOTABE Temple University VICKI LANE University of Colorado at Denver MICHAEL R. LEVY Babson College DEBBIE MACINNIS University of Southern California SCOTT B. MACKENZIE Indiana University GREG MARSHALL Oklahoma State University CHARLOTTE MASON University of North Carolina AJAY MENON Colorado State University ANIL MENON IBM Corporation BANWARI MITTAL Northern Kentucky University DAVID B. MONTGOMERY Stanford University MITZI MONTOYA-WEISS North Carolina State University ROBERT M. MORGAN University of Alabama KENT NAKAMOTO Virginia Tech CHERYL NAKATA University of Illinois at Chicago DAS NARAYANDAS Harvard Business School RICHARD C. NETEMEYER University of Virginia DAVID J. ORTINAU University of South Florida AMY L. OSTROM Arizona State University THOMAS J. PAGE Michigan State University A. PARASURAMAN University of Miami ROBERT A. PETERSON University of Texas at Austin S. RATNESHWAR University of Connecticut WILLIAM T. ROBINSON Purdue University SAEED SAMIEE University of Tulsa SANJIT SENGUPTA San Francisco State University VENKATESH SHANKAR University of Maryland JAGDIP SINGH Case Western Reserve University JAMES M. SINKULA University of Vermont K. SIVAKUMAR Lehigh University AMY K. SMITH George Washington University DANIEL C. SMITH Indiana University N. CRAIG SMITH London Business School RICHARD SPRENG Michigan State University DEVANATHAN SUDHARSHAN University of Illinois at Urbana- Champaign DAVID M. SZYMANSKI Texas A&M University STEPHEN S. TAX University of Victoria SHIRLEY TAYLOR Queen’s University GLENN VOSS North Carolina State University BRIAN WANSINK University of Illinois at Urbana- Champaign ROBERT B. WOODRUFF University of Tennessee JOHN WORKMAN Creighton University MANJIT S. YADAV Texas A&M University GEORGE M. ZINKHAN University of Georgia SHAOMING ZOU University of Missouri at Columbia Editorial Review Board For Sage Publications: David Neyhart, Gillian Dickens, Ken Berthel, and Kelli Palma
The Academy of MarketingScience is a member of theInternational Association forManagement Education(AACSB). It gratefullyacknowledges the financialsupport of the Mary KayCosmetics Excellence inMarketing Fund.
The JOURNAL OF THE ACADEMY OFMARKETING SCIENCE is the official journalof the Academy of Marketing Science. It is aninternational, refereed journal intended tofurther the science of marketing throughout theworld by promoting the conduct ofresearch and the dissemination of researchresults through the study and improvement ofmarketing as an economic, ethical, and socialforce. For manuscript submission information,refer to the inside back cover.
MANOJ K. AGARWALBinghamton University
CHRISTIE H. AMATOUniversity of North Carolina–Charlotte
JONLEE ANDREWSIndiana University
KWAKU ATUAHENE-GIMACity University of Hong Kong
RICK BAGOZZIRice University
SHARON BEATTYUniversity of Alabama
DAN BELLOGeorgia State University
*HAROLD W. BERKMANUniversity of Miami
SUNDAR BHARADWAJEmory University
DAVID M. BOUSHUniversity of Oregon
JAMES R. BROWNVirginia Tech
STEVEN P. BROWNSouthern Methodist University
TOM J. BROWNOklahoma State University
MICHELE BUNNUniversity of Alabama
ALAN BUSHUniversity of Memphis
ROGER CALANTONEMichigan State University
JOSEPH P. CANNONColorado State University
GOUTAM CHAKRABORTYOklahoma State University
GOUTAM CHALLAGALLAGeorgia Institute of Technology
RAJESH CHANDYUniversity of Minnesota
BRUCE CLARKNortheastern University
JOSEPH COTEWashington State University
DAVID W. CRAVENSTexas Christian University
PEGGY H. CUNNINGHAMQueen’s University
MICHAEL R. CZINKOTAGeorgetown University
PRATIBHA DABHOLKARUniversity of Tennessee
PETER A. DACINQueen’s University
PETER DICKSONFlorida International University
PAM SCHOLDER ELLENGeorgia State University
ELLEN GARBARINOCase Western Reserve University
JIM GENTRYUniversity of Nebraska
RONALD C. GOODSTEINGeorgetown University
EVERT GUMMESSONStockholm University
GREG GUNDLACHUniversity of Notre Dame
CHRISTIAN HOMBURGUniversity of Mannheim
G. TOMAS M. HULTMichigan State University
MICHAEL HYMANNew Mexico State University
CHARLES A. INGENEUniversity of Mississippi
JEAN L. JOHNSONWashington State University
SUSAN KEAVENEYUniversity of Colorado at Denver
AMNA KIRMANISouthern Methodist University
MASAAKI KOTABETemple University
VICKI LANEUniversity of Colorado at Denver
MICHAEL R. LEVYBabson College
DEBBIE MACINNISUniversity of Southern California
SCOTT B. MACKENZIEIndiana University
GREG MARSHALLOklahoma State University
CHARLOTTE MASONUniversity of North Carolina
AJAY MENONColorado State University
ANIL MENONIBM Corporation
BANWARI MITTALNorthern Kentucky University
DAVID B. MONTGOMERYStanford University
MITZI MONTOYA-WEISSNorth Carolina State University
ROBERT M. MORGANUniversity of Alabama
KENT NAKAMOTOVirginia Tech
CHERYL NAKATAUniversity of Illinois at Chicago
DAS NARAYANDASHarvard Business School
RICHARD C. NETEMEYERUniversity of Virginia
DAVID J. ORTINAUUniversity of South Florida
AMY L. OSTROMArizona State University
THOMAS J. PAGEMichigan State University
A. PARASURAMANUniversity of Miami
ROBERT A. PETERSONUniversity of Texas at Austin
S. RATNESHWARUniversity of Connecticut
WILLIAM T. ROBINSONPurdue University
SAEED SAMIEEUniversity of Tulsa
SANJIT SENGUPTASan Francisco State University
VENKATESH SHANKARUniversity of Maryland
JAGDIP SINGHCase Western Reserve University
JAMES M. SINKULAUniversity of Vermont
K. SIVAKUMARLehigh University
AMY K. SMITHGeorge Washington University
DANIEL C. SMITHIndiana University
N. CRAIG SMITHLondon Business School
RICHARD SPRENGMichigan State University
DEVANATHAN SUDHARSHANUniversity of Illinois at Urbana-
DAVID M. SZYMANSKITexas A&M University
STEPHEN S. TAXUniversity of Victoria
SHIRLEY TAYLORQueen’s University
GLENN VOSSNorth Carolina State University
BRIAN WANSINKUniversity of Illinois at Urbana-
ROBERT B. WOODRUFFUniversity of Tennessee
JOHN WORKMANCreighton University
MANJIT S. YADAVTexas A&M University
GEORGE M. ZINKHANUniversity of Georgia
SHAOMING ZOUUniversity of Missouri at Columbia
Editorial Review Board
For Sage Publications: David Neyhart, Gillian Dickens, Ken Berthel, and Kelli Palma
Salesperson Cooperation:The Influence of Relational, Task, Organizational, and Personal FactorsCengiz Yilmaz and Shelby D. Hunt 335
The Influence of Complementarity, Compatibility,and Relationship Capital on Alliance PerformanceMB Sarkar, Raj Echambadi, S. Tamer Cavusgil, and Preet S. Aulakh 358
Customer Switching Behavior in Online Services:An Exploratory Study of the Role of SelectedAttitudinal, Behavioral, and Demographic FactorsSusan M. Keaveney and Madhavan Parthasarathy 374
Managing Culturally Diverse Buyer-Seller Relationships:The Role of Intercultural Disposition and Adaptive Sellingin Developing Intercultural Communication CompetenceVictoria D. Bush, Gregory M. Rose, Faye Gilbert, and Thomas N. Ingram 391
Guidelines for Conducting Research and Publishing in Marketing:From Conceptualization Through the Review ProcessJohn O. Summers 405
REVIEWS OF BOOKS 416
MARKETING AND THE LAW 424
Journal of the
MarketingScienceFall 2001 Volume 29 Number 4
The ACADEMY of MARKETING SCIENCE
Central Office: School of Business Administration
University of Miami
Published by Sage Publications
Thousand Oaks • London • New Delhi
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001Yilmaz, Hunt / SALESPERSON COOPERATION
Salesperson Cooperation:The Influence of Relational, Task,Organizational, and Personal Factors
Cengiz YilmazGebze Institute of Technology, Turkey
Shelby D. HuntTexas Tech University
Salesperson cooperation has become a crucial issue forthe overall performance of most sales organizations. Theauthors examine the antecedents of task-specific, coopera-tive behaviors of salespersons toward other salespeopleworking in the same organization. The main theses of thestudy are that (1) the four major antecedent categories offactors—relational, task, organizational, and personal—constitute, collectively, the primary determinants of sales-person cooperation and (2) each antecedent category ex-erts, independently, significant influence on the co-operative behaviors of salespersons. The results supportthe main theses and provide useful insights for sales man-agers attempting to foster cooperation among salespeo-ple. The relative impact of each antecedent category, aswell as the effects of specific variables within each, isdiscussed.
Recent decades have witnessed a dramatic change inthe nature of the selling job for many companies. The tra-ditional view of a salesperson—a single, individualistic,persistent person who works independently on a commis-sion basis and who competes fiercely against even fellowsalespersons—has given way to a strikingly different con-ceptualization (Cespedes, Doyle, and Freedman 1989;Weitz and Bradford 1999). Selling in many businessestoday has become an integrated process that requires the
coordinated efforts of salespeople and other participants,both within and across product lines, functional depart-ments, and geographic districts. Cooperation, defined asthe willful contribution of individuals, groups, and so on,to the successful completion of common tasks and/or tothe achievement of mutual objectives (J. Anderson andNarus 1990; Deutsch 1949; Wagner 1995) has become acritical issue in sales management. Many companies seeksales forces composed of cooperative salespersons whocan work effectively in groups. In such sales forces, sales-people share their skills, knowledge, time, and effort withcoworkers to achieve common objectives. This emerging“era of the cooperative salesperson” is manifested in thegrowing use of team selling (Moon and Armstrong 1994),relationship selling (Weitz and Bradford 1999), sellingcenters (Hutt, Johnston, and Ronchento 1985), and keyaccount programs (Cohen 1996).
As a result of the growing importance of cooperativeselling, research in sales force management has begun tofocus on understanding the dynamics of a salesperson’sinterpersonal relationships with coworkers. Issues investi-gated include feedback provided by coworkers (Kohli andJaworski 1994), sales force socialization (Dubinsky,Howell, Ingram, and Bellenger 1986), peer mentoring(Pullins, Fine, and Warren 1996), and altruistic behaviorstoward coworkers as a form of organizational citizenshipbehaviors (e.g., Netemeyer, Boles, McKee, andMcMurrian 1997). Nonetheless, salesperson cooperation,a critical determinant of the effectiveness of selling effortsfor many businesses, has received little attention.
Consider the problem faced by a sales manager whobelieves that salesperson cooperation is important for
sales performance and wants to take action or develop pol-icies to increase such cooperation. The literatures of thedifferent research traditions that have examined coopera-tion give different, sometimes conflicting, advice. As sug-gested by the relationship marketing literature (e.g.,Dwyer, Schurr, and Oh 1987; Morgan and Hunt 1994;J. Smith and Barclay 1997), should the sales managerfocus on taking steps to increase the trust and commitmentof salespeople? Or, should the manager focus on increas-ing the task interdependence of the salespeople, as sug-gested by Deutsch (1973); Van De Ven, Delbecq, andKoenig (1976); and Wageman and Baker (1997)? Or,should the manager simply focus on hiring salespeoplewho have a general proclivity toward cooperativeness, assuggested by the works of Argyle (1991) and Chatman andBarsade (1995)? Answering these questions requiresresearch that crosses disciplinary lines.
Using an interdisciplinary approach, we address thequestion: Why do some salespeople, more than others,cooperate with coworkers? We develop and test a model ofantecedent factors that affect salesperson cooperation,which is viewed as task-specific, cooperative behaviorsamong salespeople. On the basis of a review of themultidisciplinary literature on interpersonal cooperationin organizations and workgroups, we propose that each ofthe antecedent factors suggested by prior research can becategorized into one of four categories: relational, task,organizational, and personal. The main theses of our studyare that (1) the four major antecedent categories constitute,collectively, major determinants of salesperson coopera-tion; (2) each antecedent category exerts, independently,significant influence on cooperative tendencies amongsalespeople; and therefore, (3) sales managers shouldendeavor to address factors in all four categories and notjust focus on one or two. Thus, our study aims to providesales managers with guidance on how to promote coopera-tion among their salespeople.
The article is organized as follows. First, we brieflyreview the literature on interpersonal cooperation in orga-nizations. Next, we describe the four main antecedent cat-egories and develop a structural model that incorporatespredictor variables from each. Third, we test the proposedmodel using a large sample of salespersons (N = 531) from112 different automobile dealerships. The final sectionsinclude implications and suggestions for future research.
K. Smith, Carroll, and Ashford (1995) suggest thatapproaches to the study of cooperation can be grouped intofive broad traditions. First, an influential research traditionexplains the emergence of cooperation based on thecalculative orientations of individuals (e.g., Williamson1975). In this view, individuals will cooperate if and only if
cooperation is in their long-term self-interests based ontheir rational calculations. According to K. Smith et al.(1995), most well-known theoretical explanations ofcooperation belong to this first category (e.g., transactioncost theory and game theory). A second research traditionaddresses the noneconomic aspects of cooperative rela-tionships (e.g., Thibaut and Kelley 1959). Rooted in thesocial exchange literature, research in this traditionfocuses on the effects of interpersonal attraction, psycho-logical attachment, and norms of reciprocity.
A third approach relies heavily on power and conflicttheories (e.g., Emerson 1962). Conflict, the opposite ofcooperation according to some authors and a key conceptin these theories, stems from diversity in individuals’resources, perceptions of injustice, values, and goals. Afourth approach relies on social-structure theories andemphasizes dimensions outside the focal relationship toexplain cooperation (e.g., P. Blau 1974). Social, cultural,and structural aspects of the environment in which therelationship occurs are seen as drivers of cooperation.Finally, the fifth approach involves modeling theories andemphasizes the impact of social learning and imitation oncooperative tendencies (e.g., Bandura 1971). Given thediffering underlying assumptions and units of analysisadopted by each research tradition, the current state ofinquiry on cooperation is replete with explanatory vari-ables (K. Smith et al. 1995).
Differences notwithstanding, at least three similaritiesexist across the research traditions that explore coopera-tion. First, definitions of cooperation in the traditions con-verge on a common conceptual domain, and all include awillful-contribution element and a common task or objec-tive element.1 Second, the resulting outcome for most tasksituations is increased productivity, especially in complextask situations (Tjosvold 1984; Tjosvold and Tsao 1989),because of cooperating individuals tending to (1) provideeach other with necessary information, (2) more willinglyassist and help each other, (3) understand each other’spoints of view, (4) be influenced by each other’s interestsand ideas, and (5) rely on division of labor (Laughlin1978).2 Third, some conceptual overlap exists among theexplanatory variables suggested by each approach, eventhough research in each tradition—true to the “silo” viewof academia—seldom crosses lines (K. Smith et al. 1995).Perhaps this lack of an interdisciplinary approachaccounts for the low variance explained in most studies ofcooperation.
Indeed, research in each of the traditions has (necessar-ily) been limited in scope (i.e., in terms of including allmajor antecedents of cooperation). For example, studiesusing game theory generally emphasize structural andpsychological determinants such as task characteristicsand personalities of the participants (e.g., Murnighan1994), whereas studies based on social-exchange theoryfocus on the aspects of the relationship between
336 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
cooperating parties. Similarly, while social-structure theo-ries focus solely on the broader context in which a cooper-ative relationship occurs, such as the structural and cul-tural environment, modeling theories highlight theinfluence of third parties outside the focal relationship(e.g., managers). However, as Pinto, Pinto, and Prescott(1993) note, factors that act as facilitators of cooperation inorganizations may belong to a broad set of antecedent cate-gories, ranging “from individual factors such as personali-ties of group members, interpersonal relations and trainingand skills . . . to organizational factors such as strategy,structure, reward systems, and cultural norms” (p. 1282).Therefore, using inferences from each of the traditions, weargue that the cooperative behaviors of salespeopleemerge from the combined effects of variables in four dis-tinct categories: (1) the quality of interpersonal relation-ships between organizational members, that is, relationalfactors; (2) specific properties and requirements of the taskat hand, that is, task factors; (3) the structural, cultural, pro-cedural, and managerial dimensions of the organization,that is, organizational factors; and (4) individual charac-teristics of organizational members, that is, personal fac-tors. Table 1 provides a review of the explanatory variablesin the cooperation research. Each antecedent variable usedin the various research approaches can be grouped into oneof the four categories.
A MODEL OF SALESPERSONCOOPERATION
Our model of salesperson cooperation is shown in Fig-ure 1. Although the model incorporates antecedent factorsfrom each main category, it is obvious that not all potentialfactors can be included. Thus, the factors from each cate-gory included in our model are those we propose are mostrelevant to salesperson cooperation in the context of thepresent study. For example, factors such as organizationalcommitment and job satisfaction are included in the modelbecause these factors are frequently used attitudinal vari-ables in the sales management literature in explainingsalesperson behaviors. Similarly, factors such as trust incoworkers and task interdependence are included sincesuch factors are key explanatory factors suggested in atleast one of the research traditions exploring cooperation.We discuss each variable in the four antecedent categoriesand the theoretical and empirical grounds for 15 specifichypotheses.
Relational factors are those that cause salespersons tovalue their relationships with coworkers and developmutually beneficial, long-term orientations in workingrelationships. The social-exchange literature implies that
interpersonal attraction, psychological attachment, andnorms of reciprocity—stimulated by loyalties, friendship,and faithful expectations—affect individuals’ behavioralchoices in relationships. Although such relational vari-ables as communication quality (J. Anderson and Narus1990), shared values (Chatman 1991; Morgan and Hunt1994), cultural differences (McAllister 1995), person-organization fit (Chatman 1991; Netemeyer et al. 1997),and expectations regarding the future behaviors of rolepartners (Wiener and Doescher 1994) have been theorizedto affect cooperative tendencies, the most prominent rela-tional factors are trust and commitment (Achrol 1991;Morgan and Hunt 1994).
Indeed, commitment and trust are considered key fordistinguishing social from purely economic exchange(K. Cook and Emerson 1978; G. McDonald 1981). Coop-eration entails vulnerability, and both commitment andtrust are considered necessary for individuals to value arelationship and to be willing to be vulnerable (Mayer,Davis, and Schoorman 1995; Weitz and Bradford 1999).Morgan and Hunt (1994) theorize that an individual’scommitment to a relationship and trust in the exchangepartner are key determinants of several behavioral tenden-cies in the relationship, including a disposition to cooper-ate. Similarly, we argue that a salesperson’s trust incoworkers and his or her commitment to the organizationare central to understanding how relational factors facili-tate cooperation. Specifically, with respect to salespersoncooperation, we model (1) organizational commitment asmediating the effects of intrinsic and extrinsic job satisfac-tion, (2) trust in coworkers as mediating the effects of pastopportunistic behaviors of coworkers and communicationquality, and (3) both trust and commitment as mediatingthe effect of shared values.
Organizational commitment and cooperation. Organi-zational commitment was originally defined as “thestrength of an individual’s identification with and involve-ment in a particular organization” (Porter, Steers,Mowday, and Boulian 1974:604). Stated this way, highlevels of organizational commitment are characterized bypositive affective responses toward various subgroups, in-cluding coworkers, that form the organization (Becker1992). Thus, a salesperson’s commitment to the organiza-tion should facilitate his or her cooperative tendencies to-ward coworkers. Salespeople who are committed to theorganization should attach more importance to their rela-tionships with coworkers, anticipate future interactionswith coworkers for a longer time horizon, and highly valuetheir associations with coworkers (O’Reilly and Chatman1986). Each of these variables, in turn, positively affectscooperative tendencies (Axelrod 1984; Heide and Miner1992). Supporting this view, organizational commitmenthas been shown to promote several forms of constructiveorganizational behaviors (O’Reilly and Chatman 1986),
Yilmaz, Hunt / SALESPERSON COOPERATION 337
including organizational citizenship (Tompson andWerner 1997) and level of effort exerted for group mainte-nance (G. Blau and Boal 1987). Specifically, Dubinsky,Kotabe, Lim, and Wagner (1997) demonstrate that sales-people who value pro-social behaviors are also more com-mitted to the organization, and MacKenzie, Podsakoff,and Ahearne (1998) show that organizational commitmentis associated strongly in sales force contexts with varioussupportive, extrarole activities, including those directed topeers.
Hypothesis 1: Organizational commitment and salesper-son cooperation are positively related.
Trust in coworkers and cooperation. A salesperson’strust in coworkers stems from his or her perceptions ofsuch trust-generating qualities of coworkers as integrity,reliability, and competence (Larzelere and Huston 1980;Morgan and Hunt 1994; J. Smith and Barclay 1997). Trustexists when the salesperson believes that coworkers pos-sess these major qualities of trustworthiness and is confi-
338 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
TABLE 1Factors Affecting Cooperative Behaviors in Organizations
Factor Effect Exemplar Studies
Relational factorsa. Trust (+) Jones and George (1998); McAllister (1995); Ring and Van De Ven (1994)b. Commitment (+) Dwyer, Schurr, and Oh (1987); Morgan and Hunt (1994)c. Value congruence (+) Chatman (1991); McAllister (1995); Morgan and Hunt (1994)d. Group homogeneity (+) Chatman and Barsade (1995); Kidwell and Bennett (1993)e. Communication quality (+) J. Anderson and Narus (1990); J. Smith and Barclay (1997)f. Communication frequency, modality, direction, and content McAllister (1995); Mohr and Nevin (1990)g. Past opportunistic behaviors of coworkers (–) McAllister (1995); Morgan and Hunt (1994)h. Anticipated future interactions with coworkers (+) Heide and Miner (1992); Kelley and Thibaut (1978)i. Expectations regarding future behaviors of coworkers Seabright (1993); Wiener and Doescher (1994)
Task factorsa. Task interdependence (+) Van De Ven, Delbecq, and Koenig (1976); Wageman and Baker (1997)b. Goal interdependence (+) Tjosvold (1984); Tjosvold and Tsao (1989)c. Outcome interdependence (+) Deutsch (1973); Guzzo and Shea (1992)d. Task complexity (+) Van De Ven et al. (1976); Wageman (1995)e. Costs/benefits of cooperation (–) Deutsch (1973); Wiener and Doescher (1991)f. Task identifiability/visibility (+) George (1992); Wagner (1995)g. Personal accountability (+) Kidwell and Bennett (1993); Wagner (1995)
Organizational factorsa. Organizational design and structure Chatman and Barsade (1995); Pinto, Pinto, and Prescott (1993)b. Organizational culture Chatman and Barsade (1995); J. Smith and Barclay (1993)c. Reward system Axelrod (1984); Drago and Turnbull (1991); Petersen (1992)d. Sales force control system E. Anderson and Oliver (1987)e. Leadership style Podsakoff, MacKenzie, and Bommer (1996)f. Organizational rules and procedures Galbraith and Nathanson (1978); Moenart and Souder (1990)g. Turnover rate (–) Kidwell and Bennett (1993); Spicer (1985)h. Accessibility of coworkers (+) Keller and Holland (1983); Pinto et al. (1993)i. Number of coworkers (–) Steiner (1972); Wagner (1995)
Personal factorsa. Collectivist orientation (+) Jones and George (1998); Wagner (1995)b. Personal cooperativeness (+) Argyle (1991); Chatman and Barsade (1995)c. Agreeableness (+) Chatman and Barsade (1995)d. Extraversion (+) Thorne (1987)e. External locus of control (+) Eby and Dobbins (1997); Vancouver and Ilgen (1989)f. Need for social approval (+) Eby and Dobbins (1997); Hui and Villareal (1989)g. Social competence (+) Argyle (1991); Dodge (1985)h. Empathy (+) Eisenberg and Miller (1987)i. Positive past experience in teams (+) Eby and Dobbins (1997); Loher, Vancouver, and Chajka (1994)j. Self-efficacy for teamwork (+) Eby and Dobbins (1997); Paulhus (1983)k. Age (+) Argyle (1991)l. Gender Colman (1982)m. Education (+) Burke, McKeen, and McKenna (1990)n. Organizational tenure (+) Pullins, Fine, and Warren (1996)
NOTE: Those factors for which the direction of effect was not shown in the table are either higher order, general factors that may influence cooperation indifferent ways through their various dimensions (e.g., leadership style) or categorical variables (e.g., gender).
dent that they will be reflected in future behaviors ofcoworkers. Confidence is crucial because this is whatcauses the most important outcome of trusting relation-ships: the willingness to rely on “the words, actions, anddecisions of the other party” (McAllister 1995:25). Trustreduces perceived uncertainty, facilitates risk-taking be-havior, and fosters a cooperative and/or constructive orien-tation (Mayer et al. 1995; Moorman, Deshpande, andZaltman 1993; Morgan and Hunt 1994). Consistent withits properties, several authors have posited trust as an im-mediate antecedent of cooperation (e.g., Jones and George1998; Ring and Van De Ven 1994) and as a key mediatingconstruct between various relational factors and coopera-tion (Morgan and Hunt 1994).
Hypothesis 2: Trust in coworkers and salesperson coop-eration are positively related.
Trust facilitates organizational commitment. Relation-ships with peers, especially the degree and quality of so-cialization with coworkers, are among the primary driversof commitment to the organization (Hunt, Chonko, andWood 1985; Mottaz 1988). High levels of interpersonaltrust allow mutual respect to prevail, reduce the complex-ity of organizational life, enable organizational members
to develop positive affective responses, and therefore fa-cilitate organizational commitment (Nyhan 1999). Thus, apositive relationship between trust in coworkers and orga-nizational commitment is expected. In support of thisview, Hrebiniak and Alutto (1972) find trust among newemployees as positively related to the subsequent develop-ment of organizational commitment, J. Cook and Wall(1980) report strong correlations between various dimen-sions of trust in peers and organizational commitment, andMorgan and Hunt (1994) find trust to influence relation-ship commitment.
Hypothesis 3: Trust in coworkers and salesperson orga-nizational commitment are positively related.
Intrinsic and extrinsic job satisfaction. Empirical stud-ies in sales force contexts show that job satisfaction andseveral forms of cooperative and/or constructive behav-iors, such as peer mentoring (Pullins et al. 1996) and orga-nizational citizenship (Netemeyer et al. 1997), arepositively related. Similarly, Argyle (1991) notes that jobsatisfaction is higher in cooperative groups. While expla-nations for the relationship between job satisfaction andvarious forms of cooperative and/or constructive behav-iors are based on the premise that those who are satisfied
Yilmaz, Hunt / SALESPERSON COOPERATION 339
PersonalCooperativeness Age Education
FIGURE 1Structural Model of Salesperson Cooperation
NOTE: Relational factors: Trust in Coworkers, Organizational Commitment, Communication Quality, Past Opportunistic Behaviors of Coworkers,Shared Values With Coworkers, Intrinsic Job Satisfaction, Extrinsic Job Satisfaction. Task factor: Task Interdependence. Organizational factors: FinancialRewards, Nonfinancial Rewards, Collectivist Organizational Norms, Number of Coworkers. Personal factors: Personal Cooperativeness, Age, Education,Tenure in Organization.
with their jobs will respond in reciprocation to those whohave contributed to their positive job experience, whetherthis relationship is direct or mediated by organizationalcommitment, or both, is still an issue that warrants furtherresearch (cf. Tompson and Werner 1997). Much researchhas found a positive and strong relationship between jobsatisfaction and organizational commitment (e.g.,Johnston, Parasuraman, Futrell, and Black 1990). Further-more, the preponderance of empirical and conceptual evi-dence (see Brown and Peterson 1993) suggests thatsatisfaction precedes organizational commitment causallyin sales force settings “because it is more specific, less sta-ble, and more rapidly formed” (MacKenzie et al.1998:90). Therefore, we suggest that the satisfaction-cooperation relationship is mediated by organizationalcommitment.
We further distinguish between the intrinsic and extrin-sic aspects of job satisfaction. The former refers to an em-ployee’s satisfaction with the specific nature of the jobitself, while the latter concerns those aspects of the job thatare outside the specific scope but still within the generalcontext of the job (Lucas, Parasuraman, Davis, and Enis1987). Major components of (1) intrinsic job satisfactioninclude the joy of actually performing the job, feelings ofaccomplishment received from the job, and the degree offreedom in the job and of (2) extrinsic job satisfaction in-clude fair pay, financial earnings, work conditions, andbenefit plans (Lucas et al. 1987).
Hypothesis 4: Intrinsic job satisfaction and salespersonorganizational commitment are positively related.
Hypothesis 5: Extrinsic job satisfaction and salespersonorganizational commitment are positively related.
Shared values with coworkers. Shared values are de-fined as “the extent to which [organizational members]have beliefs in common about what behaviors, goals, andpolicies are important or unimportant, appropriate or inap-propriate, and right or wrong” (Morgan and Hunt1994:25). The relationship between shared values and de-velopment of commitment and trust is well documented inthe marketing (Dwyer et al. 1987; Morgan and Hunt 1994)and organizational behavior literatures (Chatman 1991).Shared values positively influence organizational commit-ment because salespeople sharing values with coworkerscan be expected to develop stronger affinities with theiroverall organization. Similarly, shared values positivelyinfluence trust in coworkers because, as Brewer (1979) ob-serves, individuals tend to perceive socially dissimilar in-dividuals as dishonest, untrustworthy, and uncooperative.
Hypothesis 6: Shared values with coworkers and sales-person organizational commitment are positivelyrelated.
Hypothesis 7: Shared values with coworkers and sales-person trust in coworkers are positively related.
Past opportunistic behaviors of coworkers. Empiricalevidence on trust in working relationships suggests thatpeople, when assessing competence and trustworthiness,consider whether partners have carried out role-related re-sponsibilities reliably (J. Cook and Wall 1980). Coworkerswho carry out role responsibilities reliably and in a mannerconsistent with norms of fairness and reciprocity will en-hance partners’ assessments of their trustworthiness(McAllister 1995). In contrast, when coworkers engage inopportunistic behaviors, which Williamson (1975) definesas “self interest seeking with guile” (p. 6) and which John(1984) characterizes as deceitful violations of appropriaterole behavior, the subsequent level of trust placed in co-workers will decrease.
Hypothesis 8: Past opportunistic behaviors of coworkersand salesperson trust in coworkers are negatively re-lated.
Communication quality. Prior research has focused ontwo general aspects of the communication process:(1) mechanistic aspects such as frequency, modality, di-rection, and content (e.g., Churchill, Ford, and Walker1976; Mohr and Nevin 1990) and (2) qualitative aspects(e.g., E. Anderson and Weitz 1989; J. Anderson and Narus1990). Consistent with much research on trusting relation-ships (e.g., Morgan and Hunt 1994; J. Smith and Barclay1997), we limit our discussion to the qualitative aspects ofthe communication process among salespeople.
Communication quality is defined as timely and accu-rate sharing of information through both formal and infor-mal means (E. Anderson and Weitz 1989; J. Anderson andNarus 1990; Morgan and Hunt 1994; J. Smith and Barclay1997). The timely and accurate sharing of information al-lows salespeople to be more confident in their attributionsregarding the trustworthiness of coworkers and enablesthem to better assess the motives and intentions behind theactions of coworkers (Boorom, Goolsby, and Ramsey1998). Thus, communication quality results in increasedtrust (Mayer et al. 1995).
Hypothesis 9: Communication quality with coworkersand salesperson trust in coworkers are positivelyrelated.
Ever since Morton Deutsch published his theory ofcooperation in 1949, task factors have been the most com-monly used explanatory variables in cooperation research.Deutsch’s theory viewed cooperation as a form of socialinteraction that can be characterized by perceptions ofpositive interdependence. That is, Deutsch (1949, 1973,1980) argued that individuals will be more likely to coop-erate if they view (1) one another’s goals as (positively)related and (2) task characteristics as requiring coop-
340 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
eration to achieve those goals (Tjosvold 1984, 1986). Thisnotion of interdependence, further developed by Deutschand Krauss (1960) and Thompson (1967), has resulted inthe extensive interest in structural factors, especially intask factors, among researchers investigating cooperativerelationships. Variables such as task complexity, taskinterdependence, and outcome and goal interdependencehave been posited as key explanatory factors in studies ofcooperation (Kumar, Scheer, and Steenkamp 1995a,1995b; Tjosvold 1984, 1986; Wageman 1995; Wagemanand Baker 1997). Another research stream has investi-gated task characteristics in the context of free riding andsocial loafing. Findings reveal that identifiability of indi-vidual contributions to the task at hand and personalaccountability influence the degree of within-group coop-eration (Kidwell and Bennett 1993; Wagner 1995), espe-cially in reciprocal task-flow situations (i.e., when eachperson acts on the output of the other).
Consistent with Deutsch’s theory, we posit that task in-terdependence, defined as the extent to which salespersonsdepend on one another for information and aid to ac-complish their tasks and improve their performance(Thompson 1967), will have a direct and positive effect onsalesperson cooperation. However, Deutsch viewed inter-dependence as central, or even equivalent, to cooperation—other factors affecting cooperation can do so only indirectlythrough their impact on perceptions of interdependence(Tjosvold 1986). Hence, for example, trust and commit-ment can have no direct effect on cooperation in Deutsch’stheory but can only exert indirect influence by magnifyingperceived interdependence. In contrast, the perspectivetaken in the present study is that variables from each of themajor antecedent categories exert direct influence oncooperation.
Hypothesis 10: Task interdependence and salespersoncooperation are positively related.
The structural, cultural, managerial, and proceduraldimensions of the organization have long been thought toaffect cooperative tendencies among organizational mem-bers (Mintzberg 1979; Shapiro 1977). Within this context,variables such as physical proximity of participants andtheir opportunity to interact (Wagner 1995), organiza-tional cultural norms (Moch and Seashore 1981), leader-ship style (Podsakoff, MacKenzie, and Bommer 1996),and the degree to which organizational control systemsreward cooperative efforts versus individual achievement(E. Anderson and Oliver 1987; Petersen 1992) have beenshown to influence cooperative and/or constructive orga-nizational behaviors. Incorporating organizational factorsinto models explaining cooperation is important becausethey provide managers with actionable guidance on how to
develop and maintain cooperative organizational systems(Pinto et al. 1993).
Three specific organizational factors are hypothesizedin the present study to influence salesperson cooperation:collectivist organizational norms, reward system, andnumber of coworkers. These three variables are thought torepresent major structural, cultural, and proceduraldimensions of the organization affecting cooperative ten-dencies in our sampling context. Research about pro-social organizational behaviors indicates that several man-agerial variables, particularly leadership style and leaderbehaviors, may also influence cooperative tendencies inorganizations (Podsakoff et al. 1996). The rationale for thepotent effects of leadership variables is based on the “mod-eling theories” in K. Smith et al.’s (1995) review of thecooperation literature. Based on this view, a sales managercan promote cooperation among salespeople by (1) actingas a “role model” and/or (2) communicating the appropri-ate behavioral patterns in the form of “guiding principles”(Larson and LaFasto 1989), which further contribute to thedevelopment of organizational norms. The former processinvolves imitation of the leader’s behaviors and thereforeis unlikely to bear a substantive effect in our sampling con-text (i.e., a commission-based, retail selling context wheresalespeople work in a relatively independent manner). Thepotential effects of the latter process is captured largely bythe collectivist organizational norms variable that we dis-cuss next.
Collectivist organizational (cultural) norms. An orga-nization’s internal culture is an important determinant ofhow organizational members interact with each other(Deshpande, Farley, and Webster 1993). Socially sharedrules and acceptable forms of behaviors within an organi-zation, commonly labeled as organizational (cultural)norms, tend to limit the variation across behaviors of orga-nizational members by suppressing or supporting certaintypes of behaviors (Moch and Seashore 1981). As such,the norms embedded in the internal culture of an organiza-tion prescribe behavioral patterns (Kahn, Wolfe, Quinn,Snoek, and Rosenthal 1964). One important dimension oforganizational culture closely relevant to cooperativework environments is the extent to which collectivist ver-sus individualistic norms are embedded within the organi-zation’s culture (Chatman and Barsade 1995).
Individualism-collectivism, as a determinant of coop-eration, has been studied at societal (e.g., Hofstede 1980),individual (e.g., Eby and Dobbins 1997), and organiza-tional (e.g., Chatman and Barsade 1995; Earley 1993) lev-els. As to organizational cultures, individualism-collectivism captures the relative importance organiza-tional members give to the interests of a larger workgroup(i.e., coworkers) as opposed to personal interests (Wagnerand Moch 1986). Specifically, collectivist organizationalcultures encourage the subordination of personal interests
Yilmaz, Hunt / SALESPERSON COOPERATION 341
to the goals of a larger work group and, therefore, put moreemphasis on sharing, cooperation, and harmony (Wagner1995).
Hypothesis 11: Collectivist organizational norms andsalesperson cooperation are positively related.
Reward system. The motivation literature maintainsthat financial rewards (e.g., compensation plans, bonuses,profit sharing plans) and nonfinancial rewards (e.g., hon-ors, opportunities for personal growth, job security, pro-motion) influence the behaviors of organizationalmembers (Pritchard, Jones, Roth, Stuebing, and Ekeberg1988). We define reward system in this study as the degreeto which rewards in the organization, both financial andnonfinancial, encourage cooperation among salespeople.
Petersen (1992) notes that managers should carefullydesign reward systems if certain types of behavioral pat-terns, such as cooperation, are to be developed. Axelrod(1984) suggests that cooperation can be reinforced bymaking cooperative behaviors more attractive through theusage of rewards. Research on team effectiveness showsthat when rewards are linked to group performance, a re-ward system that Campion, Medsker, and Higgs (1993) re-fer to as “interdependent rewards” and Guzzo and Shea(1992) refer to as “outcome interdependence,” group per-formance is facilitated through increased motivation to-ward group-oriented behaviors. Finally, J. Anderson andNarus (1990) and Wiener and Doescher (1991) note thatindividuals will be more likely to cooperate if they believethat the outcome of cooperation is going to be positive. In-deed, the supposed relationship between financial rewardsand all individual behaviors is so strong in the motivationliterature that including financial rewards as an antecedentto cooperation may be considered a control variable. Thatis, once one controls for financial rewards, do other factorsexplain variance in individual cooperation?
Hypothesis 12: The degree to which financial rewardsencourage cooperative behaviors is positively re-lated to salesperson cooperation.
Hypothesis 13: The degree to which nonfinancial re-wards encourage cooperative behaviors is positivelyrelated to salesperson cooperation.
Number of coworkers. Research on work groups hasposited group size as an important predictor of within-group cooperation (Hechter 1987; Wagner 1995). Becauseindividuals’ workplace behaviors and incremental taskcontributions are easier to assess, more visible, and/or“identifiable” in small groups, people in such groups tendto (1) avoid free riding and social loafing and (2) displaycooperative and/or constructive behaviors (George 1992).Furthermore, Pinto et al. (1993) argue that physical prox-imity and accessibility of organizational members may
promote cooperative behaviors by making them morefeasible.
Hypothesis 14: The number of coworkers is negativelyrelated to salesperson cooperation.
Some people are simply more cooperative than others(Argyle 1991). An individual’s disposition to behavecooperatively may stem from such personal factors as per-sonality traits (Baron 1983) and demographic characteris-tics (Argyle 1991). For example, Baron (1983) distin-guishes between “cooperators,” “competitors,” and“individualists” as personality types. Cooperators preferto work in close collaboration with other people and areprimarily interested in the achievement of group objec-tives. Competitors put more emphasis on their personalgoals. Individualists will either cooperate or compete,depending on which best fits their personal needs.
Researchers have used several personality measures asproxies for personal cooperativeness. Examples includecollectivist orientation (Wagner 1995), agreeableness(Chatman and Barsade 1995), extraversion (Thorne 1987),locus of control and need for social approval (Eby andDobbins 1997), social competence (Dodge 1985), andempathy (Eisenberg and Miller 1987). In addition,although empirical evidence is scant, such demographicvariables as age, gender, education, and tenure in the orga-nization have been proposed as predictors of cooperativedispositions (Argyle 1991; Lu and Argyle 1991; Wagner1995). We focus on personal cooperativeness and severaldemographic variables.
Personal cooperativeness. Personal cooperativeness,as examined here, is a personality trait that determines thepredisposition of an individual toward working in closecollaboration with others in all life activities. A salesper-son high in this trait
places priority on associating with others for mutualbenefits, gaining social approval, and working to-gether with others toward a common end or purpose,while a person with low disposition to cooperateplaces priority on maximizing his or her own wel-fare regardless of others’ welfare. (Chatman andBarsade 1995:424)
Hypothesis 15: The personality trait of cooperativenessand salesperson’s cooperative behaviors are posi-tively related.
Demographic differences. While it has been argued thatdemographic differences are indicators of several driversof cooperative behaviors, such as empathy and perspectivetaking (e.g., Davis 1983), several decades of researchhave, in fact, failed to yield conclusive evidence regarding
342 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
the effects of demographic variables on cooperative and/orconstructive tendencies (Podsakoff, MacKenzie, Paine,and Bachrach 2000). Concerning the impact of age, for in-stance, Wagner (1995) reports a positive and significantcorrelation between age and cooperative behaviors, whileLu and Argyle (1991) report a negative correlation. Simi-larly, some studies report significant effects of experience,education, and organizational tenure (e.g., Kidwell andBennett 1993; Pullins et al. 1996; Spicer 1985), andyet others fail to support the view that these variablesare substantively important predictors of cooperation—especially when personality differences are accounted for(Argyle 1991). Given that the literature does not allow usto specify directional hypotheses, we examine the effectsof age, education level, and organizational tenure from anexploratory perspective.
The research setting involved mail surveys of salespeo-ple and sales managers from new-car automobile dealer-ships. Salespeople from the participating dealerships wereasked to respond to self-administered questionnaires inwhich they were instructed to state their opinions regard-ing their coworkers, defined as other salespersons workingin the same dealership. While several “more cooperative”selling contexts (such as those that apply team selling)exist, new-car salespeople represent a pertinent sample forour research for several reasons. First, contrary to the ste-reotype image of the automobile salesperson, cooperativeselling is a rapidly growing practice in this industry. Inresponse to the competition from the Internet and thedemands of the manufacturer firms, many dealershipshave initiated relationship marketing and customer reten-tion programs. Mixed compensation plans (as opposed tofull-commission plans), formal or informal commissionsharing, and year-end bonuses and several forms of manu-facturer incentives based on overall dealership perfor-mance are common practices. Thus, it is not only the casethat some reasonable level of cooperation exists amongnew-car salespeople but also many dealership managersconsider such cooperation desirable for the performanceof the overall firm. Our preliminary interviews with deal-ership managers and salespeople and the data we collectedfor the present research support this view, as we demon-strate in the following sections.
Second, note that our purpose at this initial stage of the-ory testing is to explain variance and explore relationships.Since sales teams are usually composed of people fromdifferent functional areas and with diverse backgrounds(Weitz and Bradford 1999), using such a diverse samplewould have decreased our ability to explore the true natureof the relationships due to substantial amount of extrane-ous variation that cannot be modeled directly. Third, new-
car salespeople have relatively similar task requirements,which eliminates such concerns as “cooperate in whatmanner?” and enables a consistent operational definitionfor the cooperation construct. Fourth, the dealerships inour sample are relatively small organizations (a majorityof them employ less than 10 salespeople), which mini-mizes the possibility of confusion on the part of therespondents as to the question of “cooperate with whom?”Finally, the fact that our sample is drawn from what is gen-erally considered to be a relatively competitive sellingcontext facilitates a strong test of our thesis that each of thefour main antecedent categories exerts a significant anddistinct influence on salesperson cooperation.
Preliminary investigation. The study began with un-structured field interviews with managers and salespeoplefrom four local dealerships. The purpose of the interviewswith managers was to explore whether sales managers inthis sales context regarded salesperson cooperation as im-portant. All four dealership sales managers maintainedthat they wanted their salespeople to cooperate with eachother because they believed such cooperation increasedoverall sales force performance. These interviews alsoprovided useful insights for developing the specific tasksfor measuring the cooperation construct. The interviewswith salespeople provided an on-site pretest of the ques-tionnaire. Ten salespeople from the same four dealershipscommented on items and suggested changes. The finaldraft of the questionnaire was developed after making therequired modifications.
Sampling procedure. A sample frame of 1,181 new-cardealerships in the state of Texas was developed from amailing list provided by an independent research firm.Dealership sales managers were contacted by mail to so-licit their cooperation in return for the summary of results.One hundred and sixty-five dealerships agreed to partici-pate in the study, providing access to 1,975 salespeople.These dealership managers also responded to a short ques-tionnaire designed to measure several organizational-levelvariables. These variables include number of vehicles soldper year, number of employees, number of salespeople,perceived overall degree of cooperation within the salesforce, and importance of cooperation. Ninety percent ofresponses to the question “How important is it for the suc-cess of your dealership that salespersons cooperate witheach other?” were above the midpoint of the scale, rangingfrom 1 (very unimportant) to 7 (very important).3
Four weeks after the initial mailing, the salespersonquestionnaires were mailed to the managers of the 165participating dealerships for distribution to their salespeo-ple. Each questionnaire packet also included a cover letterexplaining the purpose of the study and return envelopes to
Yilmaz, Hunt / SALESPERSON COOPERATION 343
assure respondent anonymity. Five hundred and eighty-five individual salesperson responses from 112 differentdealerships were received. After the elimination of care-less respondents and a listwise deletion of missing cases,531 questionnaires were retained, resulting in an effectiveresponse rate of 27 percent. The mean within-dealershipresponse rate was 50 percent.
Nonresponse bias. Tests for nonresponse bias rely onArmstrong and Overton’s (1977) argument that late re-spondents are similar to nonrespondents (in comparison toearly respondents). Two different tests were conducted:one for the first sampling stage (dealership managers) andone for the second sampling stage (salespeople). For deal-ership managers, we compared late and early respondentson the means of two critical variables, namely, perceivedoverall degree of cooperation within the salesforce and im-portance of cooperation. For individual salespersons, wecompared the two groups on the covariance matrix of con-struct items (Morrison 1976). No significant differenceswere found in either of the tests, suggesting thatnonresponse bias may not be a problem.
Sample characteristics. Our sampling process resultedin a sample that varied greatly on both dealership andsalesperson characteristics. The dealerships vary in size asmeasured by number of employees (M ≅ 40, SD = 49.16),salespeople (M ≅ 12, SD = 9.5), and vehicles sold per year(M ≅ 943, SD = 937.5). Individual respondents vary widelyin age (M = 39.26 years, SD = 11.49), sales experience(M = 10.65 years, SD = 9.78), organizational tenure (M =2.57 years, SD = 3.34), and education (≤ high school di-ploma, 18.15%; some college, 52.45%; college graduate,20.33%; graduate work, 9.07%). Most of the respondentsare male (90.91%) and full-commission salespeople(69.78%).
Constructs are measured using multiple-item mea-sures, whenever applicable. All scales use a 7-point scal-ing format with anchors strongly disagree to stronglyagree, unless otherwise noted. Measurement items areprovided in the appendix. The reliabilities of the multiple-item, reflective measures are presented in Table 2. Thecoefficient alphas, Lisrel-based internal consistency esti-mates (i.e., composite reliability), and the amount of vari-ance captured by each construct in relation to measure-ment error (i.e., average variance extracted) are wellbeyond the acceptable threshold levels suggested byNunnally (1978) and Fornell and Larcker (1981).
Cooperation. For the sake of operational andnomological clarity, we limit the domain of the coopera-tion construct to cooperative behaviors that represent the
“core task” of our respondents, that is, automobile selling.Thus, our conceptualization of salesperson cooperation,based on the work of Laughlin (1978) and Morgan andHunt (1994), requires a measure capturing various formsof task-specific cooperative behaviors that respondents arelikely to display toward their coworkers. Both in-role andextrarole task-specific behaviors (i.e., those that includeand transcend beyond what is formally prescribed by asalesperson’s organizational role) belong to the domain ofcooperation.
Measurement items are developed through an interac-tive process with dealership managers and salespeoplewho participated in our preliminary interviews. Theseinformants provided us with valuable insights concerning(1) the nature of cooperation in automobile selling, (2) spe-cific types of cooperative behaviors in various stages of theselling process, and (3) clarity and completeness of theitems in the measure. Relatively higher emphasis is givenin the scale to cooperative behaviors involving relation-ships with customers (e.g., sharing information aboutpotential and current customers, helping one another’scustomers, etc.), based on the unanimous agreementamong our informants that customer-related cooperationis of critical importance for the success of selling effortsand most representative of a cooperative sales force. Otherfacets of salesperson cooperation frequently mentioned bythe informants include assisting coworkers during salespresentations, sharing information about vehicle specifics,and providing support in terms of activities that facilitatethe selling process (e.g., handling of paperwork). Respon-dents rated the extent to which they engage in each type ofcooperative behavior on a 7-point scoring format, rangingfrom very little to very much.
Trust in coworkers and organizational commitment.The scale in Morgan and Hunt (1994) is used for measur-ing trust in coworkers. Based on the Dyadic Trust Scale ofLarzelere and Huston (1980), this measure captures re-spondents’ confidence in the integrity, reliability, compe-tence, and general trustworthiness of relationship partners.An additional item, “I consider my coworkers as peoplewhom I would be willing to let make important job-relateddecisions without my involvement,” was included to putmore emphasis on the competence dimension. Organiza-tional Commitment is measured using the nine-item ver-sion of Mowday, Steers, and Porter’s (1979) Organiza-tional Commitment Scale, which has been used exten-sively in prior research (Mathieu and Zajac 1990).
Measures of exogenous constructs. Shared Values WithCoworkers and Past Opportunistic Behaviors of Cowork-ers use the scales in Morgan and Hunt (1994). The assess-ment of shared values involves a two-stage procedure (cf.Enz 1988): respondents are asked to state the degree towhich (1) they agree and (2) their coworkers would agree
344 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
TABLE 2Descriptive Statistics for the Scales, Reliability Estimates,a and Latent Factor Correlations
10. Collectivist Organizational Norms 5.34 1.15 .41 .44 .44 .42 .30 .27 –.20 .34 .49 .84 .85 .6611. Financial Rewards 5.22 1.60 .44 .40 .39 .42 .41 .24 –.26 .42 .28 .43 NA NA NA12. Nonfinancial Rewards 4.99 1.62 .38 .37 .37 .33 .30 .16 –.21 .42 .22 .39 .61 NA NA NA13. Numbers of Coworkers 16.7 11.4 –.08 –.05 –.05 –.09 –.03 –.01 .05 –.03 –.08 .02 –.02 –.08 NA NA NA14. Personal Cooperativeness 5.17 0.96 .52 .25 .28 .30 .13 .01 –.11 .31 .35 .37 .37 .33 .00 .74 .77 .6515. Age 39.3 11.5 .04 –.03 –.06 –.17 –.16 –.15 .17 –.04 .04 .03 –.05 .04 .03 .07 NA NA NA16. Tenure in Organization 2.57 3.34 .04 .05 –.02 .02 .07 –.04 –.01 .00 .07 .01 –.05 .02 –.07 –.04 .32 NA NA NA17. Education –.05 –.09 .04 –.13 –.11 –.06 .08 –.05 –.04 .01 –.16 –.04 –.07 .07 .07 .01 NA NA NA
NOTE: Discriminant validity is obtained if the variance extracted for a construct is greater than the squared latent factor correlation between a pair of constructs. NA = not applicable because the construct was mea-sured with a formative scale or had fewer than three items.a. Coefficient alphas are reported on the diagonal.
with five statements concerning ethical values. The differ-ences between the two responses (subtracted from 7) arethen used to reflect shared values. For opportunistic be-haviors, we added the following item to the original three-item scale: my coworkers avoid fulfilling their responsibil-ities unless they are watched closely.
Selected items from the marketing practitioner’s JobSatisfaction Scale of Hunt and Chonko (1984) and thesalesperson Intrinsic Job Satisfaction Scale of Lucas et al.(1987) are used to measure intrinsic aspects of therepondents’ job satisfaction. Extrinsic Job Satisfactionitems are drawn from Lucas et al.’s (1987) study. Items inboth scales come from the Job Dimensions Scale (Groves1981; Schletzer 1965). Similarly, for CommunicationQuality, we use selected items from the CommunicationQuality Scales in Morgan and Hunt (1994) and J. Smithand Barclay (1997). Both scales measure the degree oftimely and accurate sharing of information, and both arebased on the Communication/Participation/FeedbackScale of E. Anderson, Lodish, and Weitz (1987).
Reward System, the degree to which the rewards in theorganization encourage (discourage) cooperation betweensalespeople, is operationalized for both financial rewardsand nonfinancial rewards. Single items for both dimen-sions are developed to assess the degree to which suchrewards in the dealership favor cooperative behaviors. A7-point scoring format ranging from strongly discouragecooperation to strongly encourage cooperation is used.For Collectivist Norms embedded within the culture of theorganization, we use the Norms subscale of Individualism-Collectivism, developed in Wagner and Moch (1986) andfurther validated in Wagner (1995). Items of the originalscale were modified slightly to assess organizational-levelcultural norms.
For Task Interdependence, we use the three-item TaskInterdependence Scale in Campion et al. (1993), whichmeasures the degree to which respondents depend on eachother to accomplish their tasks and improve their perfor-mance. While the third item in the scale is a direct measureof interdependence, the first two items tap the degree ofinterdependence from a dyadic perspective in that the firstitem is a measure of the respondent’s dependence oncoworkers and the second item is a measure of the respon-dent’s perception of coworkers’dependence on him or her.For this reason, responses to the first two items are firstaveraged and then combined with the third item to gener-ate a task interdependence score for each respondent.
Finally, Personal Cooperativeness is measured usingitems from the Work-Cooperativeness Scale of Lu andArgyle (1991), the School-Cooperativeness Scale of Rob-erts (1991), and the Acceptance of Cooperation/Teamwork Scale of Oliver and Anderson (1994). Thesescales have been used to determine manifest personalitydifferences across individuals in terms of cooperative
versus competitive behavioral dispositions in specificenvironments. Wordings of the items borrowed from eachscale are altered slightly to develop a measure of GeneralCooperativeness that would apply in all environments—work, school, family, and so on. Thus, as a significant dif-ference from the Cooperation Scale, which is limited totask-specific cooperative behaviors directed towardcoworkers, items in the Personal Cooperativeness Scalemeasure a salesperson’s predisposition toward working inclose collaboration with others in general.
Measure Purification and Validation
Following the two-step procedure recommended byJ. Anderson and Gerbing (1988), we estimate andrespecify the measurement model prior to incorporatingthe structural restrictions. Maximum-likelihood LISREL 8(Jöreskog and Sörbom 1993) is used in the analyses, andthe sample covariance matrix is used as input.4 In addition,because some of the scales in this research are either com-pletely new (e.g., Cooperation) or composed of selecteditems from previously used scales (e.g., Intrinsic Job Satis-faction), it is reasonable to anticipate that several itemswill have to be dropped during respecification of the mea-surement model. Cross validation is recommended forsuch measure purification processes to minimize errorprobability and capitalization on chance. Accordingly,responses were randomly split into two halves so as tocross validate the measurement model.
The initial model, which consisted of all 78 measure-ment items and 17 factors, was estimated using the firstsplit sample. However, several items had high standard-ized residuals and modification indices, making the modelfit not acceptable: χ2
(2,796) = 5,362, Comparative Fit Index(CFI) = .82, Goodness-of-Fit Index (GFI) = .66, root meansquare error of approximation (RMSEA) = .058, standard-ized root mean square residual (SRMR) = .067. Werespecified the model by eliminating three items from theIntrinsic Job Satisfaction Scale, four items from ExtrinsicJob Satisfaction, three from Organizational Commitment,four from Cooperation, two from Trust, three from Com-munication Quality, one from Opportunistic Behaviors,and four from Personal Cooperativeness. Considering thelarge number of constructs and items, the respecifiedmodel fits the data well, χ2
Next, we tested the respecified model on the secondsplit sample. The resulting fit indices indicate that themeasurement model has a good fit to the data. While theGFI is an acceptable .88, the RMSEA value of .044 and theSRMR value of .046 indicate a very good model fit. Simi-larly, in terms of incremental fit, the CFI for the model is.93, which exceeds the recommended .90 acceptance crite-rion (R. McDonald and Marsh 1990). The fit of the model
346 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
is even better when it is estimated using the full sample,χ2
(1,248) = 2,420, CFI = .93, GFI = .88, RMSEA = .041,SRMR = .044. In addition, all items load significantly ontheir respective constructs (with the lowest t-value being11.1), providing support for the convergent validity ofmeasurement items.
Unidimensionality and discriminant validity. Proce-dures for examining the measurement scales forunidimensionality are based on exploratory and confirma-tory factor analyses of scale items, taken one scale at atime, to see if the items in each scale share a single underly-ing factor. Exploratory factor analyses reveal that only onefactor accounts for a major portion of the total variance ineach scale (i.e., only one factor is extracted using aneigenvalue of 1.0 as the cutoff point). Similarly, thegoodness-of-fit indices obtained from one-factor confir-matory factor analyses of the scales are all acceptable (i.e.,GFI > .90, CFI > .90).
Tests for discriminant validity are based on compari-sons of the chi-square statistics obtained from confirma-tory factor analyses of pairwise combinations of the studyconstructs when the correlation between the constructs are(1) constrained to unity and (2) freed for estimation. A sig-nificantly lower chi-square value for the unconstrainedmodel indicates that the two constructs are distinct.Discriminant validity is obtained for all the study con-structs using this test (∆χ2
 > 3.84 for all pairwise compar-isons), as well as the more stringent procedure suggestedby Fornell and Larcker (1981) (see Table 2).
Descriptive statistics for the scales are provided inTable 2. The standard deviations indicate a substantialamount of variance in the responses.6 More important, thelarge standard deviations for the three endogenousconstructs—Cooperation (1.18), Trust in Coworkers(1.37), and Organizational Commitment (1.13)—suggestthat each of these constructs has considerable amount ofvariance to be explained. In addition, most means arewithin one-half point of the scale centers. While the meanfor the Cooperation Scale is 5.38, the dispersion of thisvariable is also reasonably high, indicating that the sampleincludes both cooperative and noncooperative respon-dents (13% of the responses are below the center of thescale). Furthermore, the fact that most of the responses areat the higher end of the Cooperation Scale is not unex-pected. Studies on organizational members commonlyreport similar results (e.g., Chatman and Barsade 1995;Eby and Dobbins 1997). One explanation for this patternof results lies in the very notion of “the organization.”Organizations exist because individuals come together towork for a common purpose. Some level of cooperation is
therefore necessary for sustained membership in theorganization.
Table 3 reports goodness-of-fit indices and standard-ized parameter estimates for the structural model. Theoverall chi-square statistic is significant, χ2
(1,275) = 2530.6,p < .01, as is expected given the large sample size (Bagozziand Yi 1988). All other goodness-of-fit indices are withinthe acceptable ranges (CFI = .93, GFI = .88, RMSEA =.042, SRMR = .051). Taken collectively, these resultsshow that the hypothesized structural relationships fit thedata well. Overall, the hypothesized structural relation-ships explain 45 percent of the observed variance in coop-eration. In addition, 11 of the 15 hypothesized paths aresupported, and at least one factor from each of the fourantecedent categories exerts significant influence on sales-person cooperation.
Also included in Table 3 are the parameter estimatesand associated test statistics of the hypothesized relation-ships adjusted for common method variance. Given thatthe same informants provided the data for most of theexogenous and endogenous constructs in our model, thepossibility exists that common method variance may haveinflated or deflated the magnitudes of the parameter esti-mates for the hypothesized paths. Thus, it is necessary toassess the degree of this form of bias in our results. Theadjusted estimates in Table 3 are obtained after partialingout the portion of variance that is common across all ourobserved variables obtained from the same source (i.e.,salespeople), using the procedure in MacKenzie,Podsakoff, and Paine (1999).
As shown in Table 3, the overall pattern of significantrelationships in the sample is not affected much by com-mon method variance. Of the 11 paths that are significantin the unadjusted analysis, 10 are significant in theadjusted analysis, with the path from collectivist organiza-tional norms to cooperation dropping just slightly to thepoint of being nonsignificant at the traditional .05 level.More important, given that the adjusted estimates havemuch greater standard errors because of the inclusion of anadditional “common method” factor in the model andfewer degrees of freedom, the absolute sizes of the coeffi-cients should be the primary basis of comparison, not thesignificance levels. Note that the magnitudes of theadjusted path coefficients in our results are very close tothe magnitudes of the unadjusted estimates,7 and the corre-lation between the two sets of estimates is .93 (p value <.000). Furthermore, a chi-square difference test indicatesthat the model representing the adjusted estimates is notstatistically different from the (more parsimonious) modelrepresenting the unadjusted estimates (∆χ2
 = 71.2).Accordingly, our discussion in the following paragraphsconcerning the effects of specific antecedents is based onthe unadjusted estimates. We discuss the potential impactof same-source bias in cases where significant deviations
Yilmaz, Hunt / SALESPERSON COOPERATION 347
exist between the adjusted and unadjusted estimates forspecific paths.
Effects of Specific Antecedents
Of the eight constructs hypothesized to exert directinfluence on salesperson cooperation, task interdepen-dence (standardized path coefficient, γi = .30, p < .01) andpersonal cooperativeness (γi = .29, p < .01) have the high-est levels of explanatory power according to both adjustedand unadjusted analyses, providing strong support forHypotheses 10 and 15. Other significant antecedents ofcooperation include financial rewards (Hypothesis 12, |γi =.14, p < .01), trust in coworkers (Hypothesis 2, |βi = .14, p <.01), and collectivist organizational norms (Hypothesis11, |γi = .10, p < .05). However, while the adjusted andunadjusted estimates of the path coefficient linking collec-tivist organizational norms to cooperation are close inmagnitude, the adjusted estimate is slightly below the tra-ditionally accepted .05 significance level because of theinflated standard error value. Finally, the results suggestthat three exogenous relational factors, namely, Com-
munication Quality, Past Opportunistic Behaviors ofCoworkers, and Shared Values With Coworkers, are alsoimportant for cooperation. All three of these constructshave significant indirect effects on cooperative behaviorsof salespeople through their influence on trust incoworkers.
The paths hypothesizing direct effects of organiza-tional commitment (Hypothesis 1), nonfinancial rewards(Hypothesis 13), and number of coworkers (Hypothesis 14)are not supported. The results of the unadjusted analysisalso suggest that none of the three demographic indicators—age, organizational tenure, and education level—are sig-nificant predictors of salesperson cooperation. However,there is a sharp contradiction between the adjusted andunadjusted estimates concerning the potential effects ofage and organizational tenure. While the unadjusted esti-mates for these variables do not bear any form of statisticaland/or substantive significance, the magnitudes of theadjusted estimates are much greater and reach the point ofbeing statistically significant. These results suggest thatsame-source effects may be an explanation for the mixedempirical findings in prior studies that explored the effects
348 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
NOTE: CFI = Comparative Fit Index; GFI = Goodness-of-Fit Index; RMSEA = root mean square error of approximation; SRMR = standardized root meansquare residual; SMC = squared multiple correlation.a. Adjusted estimates are obtained after adding to the model a first-order “common method” factor that has each measure obtained from the same source asan indicator.*p < .05 (one-tailed test). **p < .01.
of such demographic variables in combination with otherpotential antecedents of cooperative behaviors.
Concerning the antecedents of trust in coworkers andorganizational commitment, our findings support stronglyHypotheses 7, 8, and 9, as shared values (γi = .26, p < .01),past opportunistic behaviors (γi = –.30, p < .01), and com-munication quality (γi = .34, p < .01) are significantlyrelated to levels of trust placed in coworkers. Similarly,Hypotheses 3, 4, and 5 are supported because trust incoworkers (βi = .20, p < .01), intrinsic job satisfaction (γi =.65, p < .01), and extrinsic job satisfaction (γi = .21, p < .01)are significant predictors of organizational commitment.Hypothesis 6 is not supported, however, as shared valueswith coworkers are unrelated to organizational commitment.
This study explores the antecedent conditions that pro-mote or inhibit salesperson cooperation. To this end, sev-eral antecedent factors were identified, each factor wascategorized into one of the four broader sets of antecedentconditions, and each factor was tested within a nomologi-cal network for its effect on cooperative behaviors of sales-persons directed toward coworkers, that is, other salespeo-ple. The main thesis of the hypothesized structural model isthat each antecedent category of factors—relational, task,organizational, and personal—exerts significant influenceon cooperation, independently from the effects of others.On the basis of this thesis, the study explores the relativeeffects of each main category on salesperson cooperation.Our findings provide strong support for the main thesis andvaluable insights regarding specific predictors of salesper-son cooperation. First and foremost, the hypothesizedstructural relationships explain 45 percent of the observedvariance in cooperation, which exceeds that found in moststudies within each of the five research traditions explor-ing cooperation. Second, at least one variable from each ofthe four antecedent categories is shown to exert significantinfluence on cooperation. The proportion of variance incooperation accounted for by the significant predictorscaptures almost all of the total variance explained, sincethe proportion of variance explained by nonsignificantpredictors is negligible (less than 1%). Using statisticallysignificant effects only, task factors and personal factorseach explain approximately 15 percent of the observedvariance in salesperson cooperation, while organizationalfactors and relational factors explain 10 percent and 5 per-cent, respectively.8
Consistent with several decades of research, the resultssupport the view that task interdependence is an importantpredictor of cooperation. When salespersons believe thattheir personal success is dependent on the support ofcoworkers, they have a greater tendency to cooperate withcoworkers. However, consistent with our main thesis,
while an interdependent task design will produce morecooperation, focusing merely on task interdependencedoes not seem to guarantee a highly cooperative salesforce. Developing and maintaining a cooperative salesforce requires attention to personal, organizational, andrelational factors as well, since these factors are also foundto exert influence on salesperson cooperation.
Concerning the effects of personal factors, there isstrong empirical support that personal cooperativeness is amajor predictor of salesperson cooperation. At the sametime, while our results suggest that education level is not asignificant predictor, the potential effects of age and orga-nizational tenure are somewhat unclear. Both of these vari-ables have relatively weak zero-order correlations withsalesperson cooperation (see Table 2), and our unadjustedestimates for the effects of these variables are far from hav-ing statistical and substantive significance. However, aftercontrolling for common method variance, age and organi-zational tenure become significant predictors of salesper-son cooperation. This finding is interesting, given thatresearch about constructive employee behaviors in generalhas failed to reveal conclusive evidence regarding theeffects of such demographic factors. Additional researchis needed before this issue can be resolved conclusively.Thus, overall, our results regarding the influence of per-sonal factors highlight both the importance and difficultyof recruitment procedures if a cooperative sales force isdesired. Recruiting salespeople who are cooperators bythe very nature of their personality is crucial, but identify-ing cooperative candidates could be a difficult task. Thecorrelation coefficients relating personal cooperativenessto demographic variables are all small and nonsignificant(see Table 2), suggesting that personal cooperativeness is apersonality trait that is not manifested in demographiccharacteristics. As a result, sales managers who wish todevelop cooperative sales forces, rather than relying solelyon demographic indicators, should attempt to recruitsalespersons who (1) have a history of cooperative behav-iors and/or (2) score high on personality tests of“cooperativeness.”
Regarding organizational factors, organizationalrewards are traditionally seen as one of the most effectivemanagerial tools to influence the behaviors of organiza-tional members—and rightly so. Our findings suggest astrong effect of financial rewards on salesperson coopera-tion. As expected, the degree to which financial rewardsare designed and awarded in a manner that encouragescooperation between salespeople influences cooperativebehaviors. On the other hand, nonfinancial rewards, forexample, honors, opportunities for personal growth, jobsecurity, and promotion, do not seem to affect cooperativetendencies. This latter finding should be interpreted withcaution, however. The fact that the path coefficient con-necting nonfinancial rewards to cooperation is not signifi-cant does not necessarily mean that these two concepts
Yilmaz, Hunt / SALESPERSON COOPERATION 349
have no relationship at all. The correlation between theseconstructs is, in fact, large and significant (r = .38, p value <.000). Nonetheless, the relationship is attenuated in thestructural model (i.e., when other predictor variables arecontrolled for), suggesting that nonfinancial rewards arenot among the primary motivational drivers for ourrespondents. Given the specific nature of our samplingcontext, further research is required to determine theextent to which this finding generalizes to other sellingcontexts.
Our research indicates a moderately strong effect ofcollectivist organizational norms on salesperson coopera-tion. This finding highlights the importance of norm devel-opment and enforcement processes for sales managersattempting to establish a cooperative sales force. Based ontheir standing in the organizational hierarchy, managers inmost organizations have the ability to exert a substantialinfluence on the evolution of organizational norms. This isparticularly true for sales managers who have developedeffective means to communicate the expected behaviorpatterns and who set examples through their own actions(Feldman and Arnold 1983). As Larson and LaFasto(1989) report, members of workgroups are more likely topractice the “guiding principles” suggested by their lead-ers when the leaders themselves live up to the prescribedbehavioral patterns.
Next, in contrast to that hypothesized, we found no sig-nificant relationship between number of coworkers andthe degree of cooperative behaviors directed towardcoworkers. This result contrasts with research aboutworkgroups that suggest a strong effect of group size oncooperative tendencies. One explanation for this unex-pected finding relies on the differences in the types ofinterdependence observed in other workgroups and oursample. According to a typology suggested by Thompson(1967), workgroup members are in reciprocal interdepen-dence when each acts on the output of the other. In recipro-cal interdependence, workgroup size is an important deter-minant of free riding, social loafing, and cooperation(Wagner 1995). On the other hand, the type of interdepen-dence in the present sample is what Thompson (1967)refers to as pooled interdependence, in which each respon-dent is individually responsible for performing his or herjob from the beginning to end and dependent on coworkersfor only certain types of aid and support that enhance per-formance. The number of coworkers may be less impor-tant in pooled interdependence, as cooperation is more of avoluntary act and not required by the flow of interdepen-dent tasks.
Finally, the results show that relational factors, thosethat cause an individual to value his or her association withcoworkers and develop a mutually beneficial, long-termorientation in his or her relationships with coworkers, haveconsiderable effect on salesperson cooperation. This
finding is in line with the growing interest in marketing onrelational variables, particularly on trust. Indeed, a sales-person’s trust in coworkers is not only a significant predic-tor of cooperation even when task, organizational, and per-sonal factors are accounted for but is also a key factormediating the impact of communication quality, pastopportunistic behaviors of coworkers, and shared valueswith coworkers. Several of these exogenous relational fac-tors in the hypothesized model also influence salespersoncooperation indirectly through their effects on trust.9
In contrast, however, the results do not support thepaths from shared values with coworkers to organizationalcommitment and from organizational commitment tocooperation. All three of these constructs are, in fact, posi-tively and significantly correlated (see Table 2), but therelationships become statistically nonsignificant whenother antecedent factors are controlled for in the structuralmodel. Concerning the relationship between organiza-tional commitment and cooperation, for example, a com-mon antecedent, trust in coworkers, seems to be the driv-ing factor. An explanation for these results may lie in thenotion of multiple commitments (Becker 1992; Reichers1985, 1986).
The multiple-commitments view suggests that organi-zational commitment is “a collection of multiple commit-ments to various groups that compromise the organiza-tion” (Reichers 1985:469). Note that the conceptualdomains of shared values and cooperation constructs inour model concern, specifically, the salesperson’s rela-tionships with coworkers. The nomological role that orga-nizational commitment plays in our model depends on thedegree to which salespersons associate coworkers directlywith their overall notion of “the organization.” The moreinfluence coworkers have on one’s affective state regard-ing the organization, the more important should be the roleof organizational commitment. The respondents in oursample may not have viewed their relationships withcoworkers as a strong determinant of what they feel towardtheir respective dealerships and vice versa, thereby yield-ing the result that organizational commitment is unrelatedto both shared values with coworkers and cooperativebehaviors toward coworkers.
Post Hoc Model Respecification
Although the hypothesized model fits the data, onewould not expect a simple model such as Figure 1 to be thebest fit for the data set. Accordingly, in an exploratorymanner, we reviewed LlSREL modification indices andconducted additional analyses to determine whether thereexist additional, nonhypothesized structural paths that arelikely to (1) have statistical significance and (2) improvethe model fit. Two observations that emerged as a result ofthis post hoc specification search deserve further
350 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
discussion. First, all relatively high modification indicessuggest additional paths from some of the exogenous fac-tors, specifically from Collectivist Organizational Norms,Nonfinancial Rewards, Intrinsic Job Satisfaction, and Per-sonal Cooperativeness, to trust in coworkers. Second,when these paths are incorporated into the model, (1) threeof these additional parameters, those linking trust with col-lectivist norms, nonfinancial rewards, and intrinsic satis-faction, are significant; (2) model fit is only marginallyimproved, χ2
(1,271) = 2473.8, CFI = .93, GFI = .88, RMSEA =.041, SRMR = .047; and (3) all previously significantpaths remain significant with only slight changes inparameter estimates. These findings suggest that trust incoworkers might be even more crucial for salespersoncooperation, fully or at least partially mediating the impactof several organizational, personal, and relational factors.However, because exploratory search processes such asthe preceding require cross validation, we urge the readersto be cautious when interpreting these findings.
Limitations and Future Research Directions
Generalizability is a concern for all studies. Eventhough the sample used in the study, due to homogeneityacross respondents, allowed us to control for the back-ground factors and conduct a strong test of the hypothe-sized relationships, caution should be taken when general-izing the results to other selling contexts. In particular, thesample is composed of automobile salespeople, all ofwhom engage in face-to-face, retail selling activities. Mostof the respondents are male (90.91%), full-commissionsalespeople (69.78%), and work in relatively independentworking environments. Researchers might study the theo-retical model in different selling contexts, particularly inteam-selling and industrial-selling contexts.
A closely related issue involves investigating potentialmoderators. Future research could examine the moderat-ing effects of several factors, many of which we control forin this study. A nonexhaustive list of such moderatorsincludes (1) type of sales force and the nature of the sellingjob (team selling versus individual selling, retail sellingversus industrial selling, face-to-face selling versus dis-tance selling, etc.), (2) type of interdependence in the sell-ing task (i.e., whether the task flow generates pooled,sequential, or reciprocal interdependence, etc.), (3) natureof the compensation system (i.e., whether the compensa-tion system is based on individual versus group perfor-mance; whether it is full-commission, a combination sys-tem, or full salary, etc.), and (4) performance-rewardcontingencies (i.e., the degree to which rewards areawarded in proportion to performance).
Another area for future research concerns the potentialeffects of leadership style and leader behaviors in salesforces where salespeople view their manager as a key per-son in their work environment. Podsakoff et al. (1996)show that the effects of leadership variables on pro-socialorganizational behaviors are not only significant but alsoindependent from those of several substitutes for leader-ship. Thus, given the similarities between the literatures onpro-social behaviors and cooperative behaviors, leader-ship variables may bear some distinct influence on sales-person cooperation, particularly in team-selling situations.
Finally, inconsistent with expectations, the studyreveals that organizational commitment is unrelated toboth shared values with coworkers and cooperative behav-iors. Our expectation at the inception of the study, that is,that coworkers constitute a primary group among thosethat form a salesperson’s “overall view of the organiza-tion,” is brought into question. As is often the case, thisunexpected finding suggests fruitful avenues for furtherresearch. Researchers might examine several forms ofconstituency-specific commitments (e.g., commitment tocoworkers, supervisors, top management, union, etc.) tobetter understand the interrelationships between theseconcepts; how they form the global notion of organiza-tional commitment; and how they affect attitudinal andbehavioral dispositions of salespersons toward coworkers,supervisors, and other targets.
In conclusion, in this “era of the cooperative salesper-son,” although many sales managers see overall sales per-formance as being closely linked to the coordinated effortsof their salespeople, getting salespeople to cooperate isoften perceived to be a difficult task. While many “highlycooperative” sales forces exist, it is often difficult to iden-tify the specific factors that contribute to the developmentof cooperation. Our study suggests that each one of thefour major antecedent categories of factors—relational,task, organizational, and personal—is important for acooperative sales force. Specifically, we find that salesmanagers seeking to encourage cooperation should (1) takesteps to increase task interdependence, (2) attempt to hiresalespeople who have a history of cooperative behaviors,(3) develop reward systems that reward cooperative behav-iors, (4) foster trust among their employees, (5) worktoward shared values, (6) discourage opportunistic behav-iors, (7) promote high-quality communication amongsalespeople, and (8) foster collectivist organizationalnorms. Our study, however, is but one step toward under-standing salesperson cooperation.
Yilmaz, Hunt / SALESPERSON COOPERATION 351
StandardizedScale Item Factor Loadings t-Value
Cooperation To what extent do you cooperate with your coworkers by1. Working with them to develop sales presentation techniques.a — —2. Providing support during sales presentations. .74 16.43. Taking care of their customers during their absence. .52 11.54. Assisting them by handling paperwork for them. .73 16.45. Providing feedback for improving their performance. .78 17.76. Handling their customers’ complaints in their absence.a — —7. Assisting them in collecting and storing customer-related data. .75 16.98. Sharing information about vehicle attributes. .71 16.19. Sharing information about competitors.a — —
10. Sharing information about potential customers. .70 15.811. Sharing information about existing customers.a — —
Trust in Coworkers I consider my co-workers as people who(m)1. Cannot be trusted at times (R).b — —2. Are perfectly honest and truthful.b — —3. Can be trusted completely. .85 23.44. Can be counted on to do what is right. .91 29.45. Can be counted on to get the job done right. .87 26.56. Are always faithful. .89 27.87. I have great confidence in. .91 29.88. Have high integrity. .92 29.89. I would be willing to let make important job-related decisions without my involvement. .68 18.2
Organizational These questions concern your feelings toward your dealership.Commitment 1. I am willing to put in a great deal of effort beyond than normally expected to help
this company be successful. .73 16.32. I talk up this company to my friends as a great company to work for. .85 19.93. I would accept almost any type of job assignment to keep working for this organization.b — —4. I find that my values and the company’s values are similar. .77 17.85. I am proud to tell others that I am part of this organization. .83 21.46. This organization really inspires the very best in me in the way of job performance.c — —7. I am extremely glad that I chose this company to work for over others I was considering at
the time I joined. .83 19.48. I really care about the fate of this company. .78 18.19. For me, this is the best of all possible organizations for which to work.c — —
Shared Values Please indicate the degree to which you believe that (1) your coworkers would agree with theWith Coworkers following statements, and (2) you agree with the following statements.
1. To succeed in this business, it is often necessary to compromise one’s ethics (R). .66 11.12. Top management in a business must let it be known in no uncertain terms that unethical
behaviors will not be tolerated. .74 14.53. If an employee is discovered to have engaged in unethical behavior that results primarily
in personal gain (rather than corporate gain), he or she should be promptly reprimanded. .83 15.84. If an employee is discovered to have engaged in unethical behavior that results primarily in
corporate gain (rather than personal gain), he or she should be promptly reprimanded. .82 15.75. Employers should assure that their employees are behaving in a business-like manner. .73 14.5
Past Opportunistic My experience with my coworkers tells me that, to accomplish their personal objectives,Behaviors of they would sometimes, in their interactions with me,Coworkers 1. Alter the facts slightly. .80 18.4
2. Promise to do things without actually doing them later. .90 21.93. Fail to provide me with the support that they are obliged to. .81 20.04. Avoid fulfilling their responsibilities unless they are watched closely.b — —
Intrinsic Job I am satisfied withSatisfaction 1. How interesting my job is. .65 12.1
2. How this job makes good use of my abilities. .77 14.9
352 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
3. The degree of freedom I have in my job.c — —4. The opportunities my job gives me to complete tasks from beginning to end.c — —5. The feelings of accomplishment I get from my job. .79 15.36. The opportunities my job provides me to interact with others. .81 15.67. How my company encourages professional growth.c — —8. The opportunities for independent thought and action in my job. .74 14.6
Extrinsic Job I am satisfied withSatisfaction 1. The compensation plan under which I work. .82 20.7
2. My earnings as a sales consultant. .87 23.33. Fairness of my earnings in relation to efforts I expend. .89 24.04. My probable future earnings in this business.b — —5. The attitude of the public toward my company.a — —6. My benefit plan in general.b — —7. The attitude of the public toward the industry I am working in.a — —
Communication In working relationships with my coworkers,Quality 1. We keep each other informed of new developments. .92 30.1
2. We provide each other with timely information. .94 32.53. We frequently discuss accounts and opportunities. .72 20.94. We sometimes hold back on telling each other what we know about accounts and
opportunities (R).b — —5. We communicate well about our expectations for each other’s performance.a — —6. We provide each other with frequent positive feedback on our performance.a — —
Reward System The financial rewards (compensation plans, bonuses, profit-sharing plans, etc.) andnonfinancial rewards (honors, opportunities for professional growth, job security,promotion, etc.) in auto dealerships influence the behaviors of sales consultants.Some reward systems encourage sales consultants to cooperate with each other, othersencourage competition, and some others may encourage both cooperation and competition.
How would you describe the financial and nonfinancial rewards in your dealership?
The financial rewards in this dealershipStrongly discourage cooperation 1 2 3 4 5 6 7 Strongly encourage cooperation
The nonfinancial rewards in this dealershipStrongly discourage cooperation 1 2 3 4 5 6 7 Strongly encourage cooperation
Collectivist In general, people in this dealership believe thatOrganizational 1. If one is going to be part of this company, then he or she is sometimes going to have to doNorms things that are not in his or her immediate self-interests. .66 12.9
2. Everyone should sometimes make sacrifices for the sake of the larger group of coworkers. .78 14.93. Everyone should realize that one is not always going to get what he or she personally wants. .73 14.24. Everyone should be willing to make sacrifices for the sake of the well-being of people they
work with. .83 15.55. All employees should do their best to cooperate with others instead of trying to work things
out individually. .65 12.3
Task 1. I cannot do my job well without information and assistance from my coworkers.Interdependenced 2. My coworkers depend on me for information and assistance to perform their jobs.
3. Doing a good job in selling cars requires that sales consultants support each other.
Personal Throughout my life, I have alwaysCooperativeness 1. Enjoyed activities that involve a high level of cooperation with other people. .81 16.9
2. Been known as a team player.b — —3. Preferred to work independently rather than in a group (R). .75 14.74. Found more satisfaction working toward a common group goal than working toward my
individual goals.b — —
Yilmaz, Hunt / SALESPERSON COOPERATION 353
StandardizedScale Item Factor Loadings t-Value
5. Found it more difficult to do things with others than by myself (R).b — —6. Found joint projects with other people very satisfying.b — —7. Believed that teamwork is the best way of getting results. .60 12.5
NOTE: (R) denotes a reverse-coded item.a. These items were eliminated because LISREL modification indices indicated that they have high error correlations with other items in their respectivescales.b. These items were eliminated because their squared multiple correlations were less than .2.c. These items were eliminated because LISREL modification indices indicated that they have high shared variance with items in other scales.d. Composite scale.
The authors thank Roy Howell, James B. Wilcox, DaleF. Duhan, Kimberly B. Boal, and Mike Whitman (all ofTexas Tech University) for their helpful comments andassistance in this research. The insightful comments of theeditor and three anonymous reviewers on drafts of this arti-cle are also acknowledged.
1. These two elements distinguish cooperation from other forms ofpro-social workplace behaviors such as peer mentoring and helping be-havior. Unlike such related constructs, the purpose of cooperation is theimprovement of the welfare of all participants (including the cooperatingindividual), not just the other party.
2. See Weitz and Bradford (1999) for an excellent discussion of howthese common threads of cooperation literature apply in this new era of“partnering-oriented” selling.
3. This question was included in the managers’ questionnaires tobuild a priori confidence that cooperation matters in this context. We alsoconducted a post hoc test for the relationship between salesperson coop-eration and a self-reported, long-term performance measure. The partialcorrelation coefficient between salesperson cooperation and perfor-mance (controlling for the influence of all other study variables) is signif-icant (r = .11, p = .019), which implies that highly cooperativesalespeople tend to evaluate themselves as also being high in perfor-mance.
4. Measurement error terms for the composite task interdependencemeasure and other single-item measures are set at 0.1 times the varianceof each measure.
5. It is worthwhile to note that several items across the Intrinsic JobSatisfaction, Extrinsic Job Satisfaction, and Organizational CommitmentScales tend to cross-load on the other constructs even after therespecification. However, the modification indices for these items aremuch smaller in magnitude in comparison with (1) those for the itemsthat were eliminated and (2) total chi-square of the model. Thus, takinginto account the substantive meaning of each item, we decided that themeasurement model has a reasonable level of goodness of fit and stoppedthe respecification process.
6. An interesting issue concerns the sources of variability in the scalesmeasuring organizational and task characteristics, that is, collectivist or-ganizational norms, financial rewards, nonfinancial rewards, and task in-
terdependence. For each of these measures, our study uses perceptions of(multiple) salespersons from each dealership, hence incorporating somelevel of within-dealership variability to the analyses. Ideally, however,the only reason for the variability in these measures should be differencesbetween the dealerships (i.e., across-dealership variability). To assess thedegree to which differences in the perceptions of respondents within eachdealership contribute to the overall variability in these measures, we con-ducted a series of one-way analyses of variance using dealerships as atreatment factor. The results of these analyses reveal that although somewithin-dealership variability exists in the measures of organizational andtask characteristics (on average, less than 40%), most of their total vari-ability is due to differences between the dealerships.
7. Also note that for several paths, the magnitudes of the adjusted esti-mates are actually larger than the unadjusted estimates and therefore con-stitute a stronger case for our overall model.
8. These figures are calculated by multiplying the standardized effectsize of each predictor with the zero-order correlation between the predic-tor and cooperation; therefore, they do not represent the proportion ofvariance in cooperation uniquely attributable to each type of predictor(i.e., incremental variance explained in cooperation when a predictorvariable is “added” to the model). The unique contribution of each pre-dictor variable to the variance explained in cooperation is as follows: taskinterdependence, 5 percent; personal cooperativeness, 5 percent; finan-cial rewards, 2 percent; trust, 1 percent; collectivist organizational norms,1 percent.
9. LISREL modification indices do not suggest direct paths from anyof the exogenous relational factors to cooperation.
Achrol, Ravi. 1991. “Evolution of the Marketing Organization: NewForms for Turbulent Environments.” Journal of Marketing 55 (4): 77-93.
Anderson, Erin and Richard L. Oliver. 1987. “Perspectives on Behavior-Based versus Outcome-Based Salesforce Control Systems.” Journalof Marketing 51 (October): 76-88.
and Barton Weitz. 1989. “Determinants of Continuity in Conven-tional Industrial Channel Dyads.” Marketing Science 8 (Fall): 310-323.
, Leonard M. Lodish, and Barton Weitz. 1987. “Resource Alloca-tion Behavior in Conventional Channels.” Journal of Marketing Re-search 24 (February): 85-97.
Anderson, James C. and David W. Gerbing. 1988. “Structural Modelingin Practice: A Review and Recommended Two-Step Approach.” Psy-chological Bulletin 103 (3): 411-423.
354 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
StandardizedScale Item Factor Loadings t-Value
and James A. Narus. 1990. “A Model of Distributor Firm andManufacturer Firm Working Partnerships.” Journal of Marketing 54(January): 42-58.
Argyle, Michael. 1991. Cooperation: The Basis of Sociability. London:Routledge.
Armstrong, J. Scott and Terry S. Overton. 1977. “EstimatingNonresponse Bias in Mail Surveys.” Journal of Marketing Research14 (August): 396-402.
Axelrod, Robert M. 1984. The Evolution of Cooperation. New York: Ba-sic Books.
Bagozzi, Richard P. and Youjae Yi. 1988. “On the Evaluation of Struc-tural Equation Models.” Journal of the Academy of Marketing Sci-ence 16 (Spring): 77-94.
Bandura, A. 1971. Social Learning Theory. New York: General LearningPress.
Baron, R. 1983. Behavior in Organizations. Newton, MA: Allyn and Ba-con.
Becker, Thomas E. 1992. “Foci and Bases of Commitment: Are TheyDistinctions Worth Making?” Academy of Management Journal35:232-244.
Blau, Gary J. and Kimberly B. Boal. 1987. “Conceptualizing How Job In-volvement and Organizational Commitment Affect Turnover andAbsenteeism.” Academy of Management Review 12 (2): 288-300.
Blau, P. 1974. On the Nature of Organizations. New York: John Wiley.Boorom, Michael, Jerry Goolsby, and Rosemary Ramsey. 1998. “Rela-
tional Communications and Their Effects on Adaptiveness and SalesPerformance.” Journal of the Academy of Marketing Science 26(Winter): 16-30.
Brewer, M. B. 1979. “In-Group Bias in the Minimal Intergroup Situation:A Cognitive-Motivational Analysis.” Psychological Bulletin 86:307-324.
Brown, Steven P. and Robert A. Peterson. 1993. “Antecedents and Conse-quences of Salesperson Job Satisfaction: Meta-Analysis and Assess-ment of Causal Effects.” Journal of Marketing Research 30(February): 63-77.
Burke, Ronald J., Carol A. McKeen, and Katherine S. McKenna. 1990.“Sex Differences and Cross-Sex Effects on Mentoring: Some Prelim-inary Data.” Psychological Reports 67:1011-1023.
Campion, Michael A., Gina J. Medsker, and A. Catherine Higgs. 1993.“Relations Between Work Group Characteristics and Effectiveness:Implications for Designing Effective Work Groups.” Personnel Psy-chology 46:823-850.
Cespedes, Frank V., Stephen X. Doyle, and Robert J. Freedman. 1989.“Teamwork for Today’s Selling.” Harvard Business Review (March-April):44-54.
Chatman, Jennifer A. 1991. “Matching People and Organizations: Selec-tion and Socialization in Public Accounting Firms.” AdministrativeScience Quarterly 36 (September): 459-484.
and Sigal G. Barsade. 1995. “Personality, Organizational Cul-ture, and Cooperation: Evidence From a Business Simulation.” Ad-ministrative Science Quarterly 40:423-443.
Churchill, Gilbert A., Neil M. Ford, and Orville C. Walker Jr. 1976. “Or-ganizational Climate and Job Satisfaction in the Salesforce.” Journalof Marketing Research 13 (November): 323-332.
Colman, A. 1982. Interpersonal Process in Close Relationships.Wokingham, England: Van Nostrand.
Cook, John and Toby Wall. 1980. “New Work Attitude Measures ofTrust, Organizational Commitment, and Personal Need Non-Fullfilment.” Journal of Occupational Psychology 53:39-52.
Cook, Karen S. and Richard M. Emerson. 1978. “Power, Equity, andCommitment in Exchange Networks.” American Sociological Re-view 43 (October): 721-739.
Davis, Mark. 1983. “Measuring Individual Differences in Empathy: Evi-dence for a Multidimensional Approach.” Journal of Personality andSocial Psychology 44:113-126.
Deshpande, Rohit, John U. Farley, and Frederick E. Webster, Jr. 1993.“Corporate Culture, Customer Orientation, and Innovativeness inJapanese Firms: A Quadrant Analysis.” Journal of Marketing 57(January): 23-27.
Deutsch, Morton. 1949. “A Theory of Cooperation and Competition.”Human Relations 2:129-152.
1973. The Resolution of Conflict. New Haven, CT: Yale Univer-sity Press.
. 1980. “Fifty Years of Conflict.” In Retrospections on Social Psy-chology. Ed. L. Festinger. New York: Oxford University Press, 46-77.
and R. M. Krauss. 1960. “The Effect of Threat upon Interper-sonal Bargaining.” Journal of Abnormal and Social Psychology61:181-189.
Dodge, K. A. 1985. “Facets of Social Interaction and the Assessment ofSocial Competence in Children.” In Children’s Peer Relations: Is-sues in Assessment and Intervention. Eds. B. H. Schneider, K. H. Ru-bin, and J. E. Ledingham. New York: Springer-Verlag, 183-218.
Drago, Robert and Geoffrey K. Turnbull. 1991. “Competition and Coop-eration in the Workplace.” Journal of Economic Behavior and Orga-nization 15 (May): 347-364.
Dubinsky, Alan J., Roy D. Howell, Thomas N. Ingram, and DannyBellenger. 1986. “Sales Force Socialization.” Journal of Marketing50 (October): 192-207.
, Masaaki Kotabe, Chae Un Lim, and William Wagner. 1997.“The Impact of Values on Salespeople’s Job Responses: A Cross-Na-tional Investigation.” Journal of Business Research 39 (3): 195-208.
Dwyer, F. Robert, Paul H. Schurr, and Sejo Oh. 1987. “Developing BuyerSeller Relationships.” Journal of Marketing 51 (April): 11-27.
Earley, P. Christopher. 1993. “East Meets West Meets Mideast: FurtherExplorations of Collectivist and Individualistic Workgroups.” Acad-emy of Management Journal 36:319-348.
Eby, Lillian T. and Gregory H. Dobbins. 1997. “Collectivist Orientationin Teams: An Individual and Group Level Analysis.” Journal of Or-ganizational Behavior 18:275-295.
Eisenberg, N. and P. A. Miller. 1987. “The Relation of Empathy toProsocial and Related Behaviors.” Journal of Personality and SocialPsychology 52:91-119.
Emerson, R. E. 1962. “Power-Dependence Relations.” American Socio-logical Review 27:31-41.
Enz, Cathy A. 1988. “The Role of Value Congruity in IntraorganizationalPower.” Administrative Science Quarterly 33 (June): 284-304.
Feldman, Daniel C. and Hugh J. Arnold. 1983. Managing Individual andGroup Behavior in Organizations. Tokyo, Japan: McGraw-Hill.
Fornell, Claes and David F. Larcker. 1981. “Evaluating Structural Equa-tion Models With Unobservable Variables and Measurement Error.”Journal of Marketing Research 18 (February): 39-50.
Galbraith, J. R. and D. A. Nathanson. 1978. Strategic Implementation:The Role of Structure and Process. Dallas, TX: Business Publica-tions.
George, J. M. 1992. “Extrinsic and Intrinsic Origins of Perceived SocialLoafing in Organizations.” Academy of Management Journal35:191-202.
Groves, David L. 1981. “Impact of Industrial Recreation Upon Job Satis-faction and Productivity.” Journal of Environmental Systems 11:187-200.
Guzzo, R. A. and G. P. Shea. 1992. “Group Performance and IntergroupRelations in Organizations.” In Handbook of Industrial and Organi-zational Psychology. Eds. M. D. Dunnette and L. M. Hough. PaloAlto, CA: Consulting Psychologists Press, 269-313.
Hechter, M. 1987. Principles of Group Solidarity. Berkeley: Universityof California Press.
Heide, Jan B. and Anne S. Miner. 1992. “The Shadow of the Future: Ef-fects of Anticipated Interaction and Frequency of Contact on Buyer-Seller Cooperation.” Academy of Management Journal 35 (2): 265-291.
Hofstede, Geert. 1980. Culture’s Consequences. Beverly Hills, CA:Sage.
Hrebiniak, Lawrence G. and Joseph A. Alutto. 1972. “Personal and RoleRelated Factors in the Development of Organizational Commit-ment.” Administrative Science Quarterly 17 (3): 555-572.
Hui, C. H. and Villareal M. J. 1989. “Individualism-Collectivism andPsychological Needs: Their Relationships in Two Cultures.” Journalof Cross-Cultural Psychology 20:310-323.
Hunt, Shelby D. and Lawrence B. Chonko. 1984. “Marketing andMachiavellianism.” Journal of Marketing 48 (Summer): 30-42.
, , and Van R. Wood. 1985. “Organizational Commitmentand Marketing.” Journal of Marketing 49 (Winter): 112-126.
Yilmaz, Hunt / SALESPERSON COOPERATION 355
Hutt, Michael D., Wesley J. Johnston, and John R. Ronchento, Jr. 1985.“Selling Centers and Buying Centers: Formulating Strategic Ex-change Partners.” Journal of Personal Selling and Sales Management5 (May): 33-40.
John, George. 1984. “An Empirical Investigation of Some Antecedents ofOpportunism in a Marketing Channel.” Journal of Marketing Re-search 21 (August): 278-289.
Johnston, Mark W., A. Parasuraman, Charles M. Futrell, and William C.Black. 1990. “A Longitudinal Assessment of the Impact of SelectedOrganizational Influences on Salespeople’s Organizational Commit-ment During Early Employment.” Journal of Marketing Research 17(August): 333-344.
Jones, Gareth R. and Jennifer M. George. 1988. “The Experience andEvolution of Trust: Implications for Cooperation and Teamwork.”Academy of Management Review 23 (3): 531-546.
Jöreskog, Karl G. and Dag Sörbom. 1993. LISREL 8: A Guide to the Pro-gram and Applications. Chicago: Statistical Package for the SocialSciences.
Kahn, Robert L., Donald M. Wolfe, Robert P. Quinn, J. Diedrick Snoek,and Robert A. Rosenthal. 1964. Organizational Stress. New York:John Wiley.
Keller, R. T. and W. E. Holland. 1983. “Communicators and Innovators inResearch and Development Organizations.” Academy of Manage-ment Journal 26:742-749.
Kelley, Harold H. and John W. Thibaut. 1978. Interpersonal Relations: ATheory of Interdependence. New York: John Wiley.
Kidwell, Roland E. Jr. and Nathan Bennett. 1993. “Employee Propensityto Withhold Effort: A Conceptual Model to Intersect Three Avenuesof Research.” Academy of Management Review 18 (3): 429-456.
Kohli, Ajay K. and Bernard J. Jaworski. 1994. “The Influence of Co-Worker Feedback on Salespeople.” Journal of Marketing 58 (4): 82-94.
Kumar, Nirmalya, Lisa K. Scheer, and Jan-Benedict E. M. Steenkamp.1995a. “The Effects of Perceived Interdependence on Dealer Atti-tudes.” Journal of Marketing Research 32 (August): 348-56.
, , and . 1995b. “The Effects of Supplier Fairnesson Vulnerable Resellers.” Journal of Marketing Research 32 (Febru-ary): 54-65.
Larson, Carl E. and Frank M. J. LaFasto. 1989. Teamwork: What Must GoRight/What Can Go Wrong. Newbury Park, CA: Sage.
Larzelere, Robert E. and Ted L. Huston. 1980. “The Dyadic Trust Scale:Toward Understanding Interpersonal Trust in Close Relationships.”Journal of Marriage and the Family 42 (August): 595-604.
Laughlin, P. R. 1978. “Ability and Group Problem Solving.” Journal ofResearch and Development in Education 12:114-120.
Loher, B. T., J. B. Vancouver, and S. Chajka. 1994. “Preferences and Re-actions to Teams.” Paper presented at the meeting of the Society forIndustrial and Organizational Psychology, April, Nashville, TN.
Lu, Luo and Michael Argyle. 1991. “Happiness and Cooperation.” Per-sonality and Individual Differences 12:1019-1030.
Lucas, George H., Jr., A. Parasuraman, Robert A. Davis, and Ben M.Enis. 1987. “An Empirical Study of Salesforce Turnover.” Journal ofMarketing 51 (July): 34-59.
MacKenzie, Scott B., Philip M. Podsakoff, and Michael Ahearne. 1998.“Some Possible Antecedents and Consequences of In-Role andExtra-Role Salesperson Performance.” Journal of Marketing 62(July): 87-98.
, , and Julie Beth Paine. 1999. “Do Citizenship BehaviorsMatter More for Managers Than for Salespeople?” Journal of theAcademy of Marketing Science 27 (4): 396-410.
Mathieu, John E. and Dennis M. Zajac. 1990. “A Review and Meta-Anal-ysis of the Antecedents, Correlates, and Consequences of Organiza-tional Commitment.” Psychological Bulletin 108 (2): 171-194.
Mayer, Roger C., James H. Davis, and David F. Schoorman. 1995. “AnIntegrative Model of Organizational Trust.” Academy of Manage-ment Review 20 (3): 709-734.
McAllister, Daniel J. 1995. “Affect- and Cognition-Based Trust as Foun-dations for Interpersonal Cooperation in Organizations.” Academy ofManagement Journal 38 (1): 24-59.
McDonald, Gerald W. 1981. “Structural Exchange and Marital Interac-tion.” Journal of Marriage and the Family (November):825-839.
McDonald, Roderick and Herbert W. Marsh. 1990. “Choosing aMultivariate Model: Noncentrality and Goodness of Fit.” Psycholog-ical Bulletin 107 (2): 247-255.
Mintzberg, H. 1979. The Structure of Organizations. Englewood Cliffs,NJ: Prentice Hall.
Moch, M. and S. E. Seashore. 1981. “How Norms Affect Behaviors inand of Corporations.” In Handbook of Organizational Design, Vol. 1.Eds. P. C. Nystrom and W. H. Starbuck. New York: Oxford UniversityPress, 534-565.
Moenart, R. K. and W. E. Souder. 1990. “An Analysis of the Use ofExtrafunctional Information by R&D and Marketing Personnel: Re-view and Model.” Journal of Product Innovation Management 7:91-107.
Mohr, Jakki and John R. Nevin. 1990. “Communication Strategies inMarketing Channels: A Theoretical Perspective.” Journal of Mar-keting 54 (October): 36-51.
Moon, Mark A. and Gary M. Armstrong. 1994. “Selling Teams: A Con-ceptual Framework and Research Agenda.” Journal of PersonalSelling and Sales Management 14 (Winter): 17-30.
Moorman, Christine, Rojit Deshpande, and Gerald Zaltman. 1993. “Fac-tors Affecting Trust in Market Research Relationships.” Journal ofMarketing 57 (January): 81-101.
Morgan, Robert M. and Shelby D. Hunt. 1994. “The Commitment-TrustTheory of Relationship Marketing.” Journal of Marketing 58 (July):20-38.
Morrison, Donald. 1976. Multivariate Statistics. New York: McGraw-Hill.
Mottaz, Clifford J. 1988. “Determinants of Organizational Commit-ment.” Human Relations 41 (6): 467-482.
Mowday, Richard T., Richard M. Steers, and Lyman W. Porter. 1979.“The Measurement of Organizational Commitment.” Journal of Vo-cational Behavior 14:224-227.
Murnighan, J. K. 1994. “Game Theory and Organizational Behavior.” InResearch in Organizational Behavior. Eds. B. M. Staw and L. L.Cummings. Greewich, CT: JAI, 323-347.
Netemeyer, Richard G., James S. Boles, Daryl O. McKee, and RobertMcMurrian. 1997. “An Investigation Into the Antecedents of Organi-zational Citizenship Behaviors in a Personal Selling Context.” Jour-nal of Marketing 61 (July): 85-98.
Nunnally, Jum C. 1978. Psychometric Theory. New York: McGraw-Hill.Nyhan, Ronald C. 1999. “Increasing Affective Organizational Commit-
ment in Public Organizations.” Review of Public Personnel Adminis-tration 19 (3): 58-70.
Oliver, Richard, L. and Erin Anderson. 1994. “An Empirical Test of theConsequences of Behavior- and Outcome-Based Sales Control Sys-tems.” Journal of Marketing 58 (October): 53-67.
O’Reilly, C. A. and Jennifer Chatman. 1986. “Organizational Commit-ment and Psychological Attachment: The Effects of Compliance,Identification, and Internalization on Prosocial Behavior.” Journal ofApplied Psychology 71 (3): 492-499.
Paulhus, D. 1983. “Sphere-Specific Measures of Perceived Control.”Journal of Personality and Social Psychology 44:1253-1265.
Petersen, Trond. 1992. “Individual, Collective, and Systems Rationalityin Workgroups: Dilemmas and Market-Type Solutions.” AmericanJournal of Sociology 98:469-510.
Pinto, Mary Beth, Jeffrey K. Pinto, and John E. Prescott. 1993. “Anteced-ents and Consequences of Project Team Cross-Functional Coopera-tion.” Management Science 39 (October): 1281-1297.
Podsakoff, Phillip M., Scott B. MacKenzie, and William H. Bommer.1996. “Transformational Leader Behaviors and Substitutes for Lead-ership as Determinants of Employee Satisfaction, Commitment,Trust, and Organizational Citizenship Behaviors.” Journal of Man-agement 22 (2): 259-298.
, , July Beth Paine, and Daniel G. Bachrach. 2000. “Orga-nizational Citizenship Behaviors: A Critical Review of the Theoreti-cal and Empirical Literature and Suggestions for Future Research.”Journal of Management 26 (3): 513-552.
Porter, Lyman W., Richard M. Steers, Richard T. Mowday, and P.Boulian. 1974. “Organizational Commitment, Job Satisfaction, andTurnover Among Psychiatric Technicians.” Journal of Applied Psy-chology 59:603-609.
Pritchard, Robert D., Steven D. Jones, Philip L. Roth, Karla K. Stuebing,and Steven E. Ekeberg. 1988. “Effects of Group Feedback, Goal Set-
356 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
ting, and Incentives on Organizational Productivity.” Journal of Ap-plied Psychology 73 (2): 337-358.
Pullins, E. Bolman, Leslie M. Fine, and Wendy L. Warren. 1996. “Iden-tifying Peer Mentors in the Sales Force: An Explanatory Investiga-tion of Willingness and Ability.” Journal of the Academy ofMarketing Science 24 (2): 125-136.
Reichers, A. E. 1985. “A Review and Reconceptualization of Organiza-tional Commitment.” Academy of Management Review 10:465-476.
. 1986. “Conflict and Organizational Commitments.” Journal ofApplied Psychology 71:508-514.
Ring, P. S. and A. Van de Ven. 1994. “Developmental Processes of Coop-erative Interorganizational Relationships.” Academy of ManagementReview 19:90-118.
Roberts, Brent. 1991. “The School Cooperativeness Scale.” Working Pa-per. University of California, Berkeley.
Seabright, Paul. 1993. “Managing the Local Commons: Theoretical Is-sues in Incentive Design.” Journal of Economic Perspectives 7 (Fall):113-134.
Schletzer, V. 1965. “A Study of the Predictive Effectiveness of the StrongVocational Interest Blank for Job Satisfaction.” Unpublished doctoraldissertation. University of Minnesota, Duluth.
Shapiro, B. P. 1977. “Can Marketing and Manufacturing Co-Exist?”Harvard Business Review 55:104-114.
Smith, J. Brock and Donald W. Barclay. 1993. “Team Selling Effective-ness: A Small Group Perspective.” Journal of Business-to-BusinessMarketing 1 (2): 3-31.
and . 1997. “The Effects of Organizational Differencesand Trust on the Effectiveness of Selling Partner Relationships.”Journal of Marketing 61 (January): 3-21.
Smith, Ken G., Susan J. Carroll, and Stephen J. Ashford. 1995. “Intra-and Interorganizational Cooperation: Toward a Research Agenda.”Academy of Management Journal 38 (1): 7-23.
Spicer, M. W. 1985. “A Public Choice Approach to Motivating People inBureaucratic Organizations.” Academy of Management Review10:518-526.
Steiner, I. D. 1972. Group Processes and Productivity. Orlando, FL: Aca-demic Press.
Thibaut, John W. and Harold H. Kelley. 1959. The Social Psychology ofGroups. New York: John Wiley.
Thompson, J. D. 1967. Organizations in Action. New York: McGraw-Hill.
Thorne, A. 1987. “The Press of Personality: A Study of Conversation Be-tween Introverts and Extroverts.” Journal of Personality and SocialPsychology 53:718-726.
Tjosvold, Dean. 1984. “Cooperation Theory and Organizations.” HumanRelations 37 (9): 743-767.
. 1986. “The Dynamics of Interdependence in Organizations.”Human Relations 39 (6): 517-540.
and Yuan Tsao. 1989. “Productive Organizational Collaboration:The Roles of Values and Cooperation.” Journal of Organizational Be-havior 10 (2): 189-195.
Tompson, Holly B. and Jon M. Werner. 1997. “The Impact of Role Con-flict/Facilitation on Core and Discretionary Behaviors: Testing a Me-diated Model.” Journal of Management 23 (4): 583-601.
Vancouver, J. B. and D. R. Ilgen. 1989. “Effects of Interpersonal Orienta-tion and the Sex-Type of the Task on Choosing to Work Alone inGroups.” Journal of Applied Psychology 74:927-934.
Van De Ven, Andrew H., Andre L. Delbecq, and Richard Koenig. 1976.“Determination of Coordination Modes Within Organizations.”American Journal of Sociology 41 (April): 322-338.
Wageman, Ruth. 1995. “Interdependence and Group Effectiveness.” Ad-ministrative Science Quarterly 40 (1): 145-180.
and George Baker. 1997. “Incentives and Cooperation: The JointEffects of Task and Reward Interdependence on Group Perfor-mance.” Journal of Organizational Behavior 18:139-158.
Wagner, John A. 1995. “Studies of Individualism-Collectivism: Effectson Cooperation in Groups.” Academy of Management Journal 38 (1):152-172.
and M. K. Moch. 1986. “Individualism-Collectivism: Conceptand Measure.” Group and Organization Studies 11 (September):280-304.
Weitz, Barton and Kevin D. Bradford. 1999. “Personal Selling and SalesManagement.” Journal of the Academy of Marketing Science 27 (2):241-254.
Wiener J. Lyle and Tabitha A. Doescher. 1991. “A Framework for Pro-moting Cooperation.” Journal of Marketing 55 (April): 38-47.
and . 1994. “Cooperation and Expectations of Coopera-tion.” Journal of Public Policy and Marketing 13 (Fall): 259-270.
Williamson, Oliver E. 1975. Markets and Hierarchies: Analysis and An-titrust Implications. New York: Free Press.
ABOUT THE AUTHORS
Cengiz Yilmaz is an assistant professor of marketing at GebzeInstitute of Technology, Turkey. He obtained his Ph.D. in market-ing from Texas Tech University in 1999. His research interestsfocus on sales management, distribution channels and relation-ship marketing, and strategic issues concerning intra- andinterfirm aspects in marketing systems and their links with busi-ness performance. His research has been published in variousconference proceedings.
Shelby D. Hunt is the J. B. Hoskins and P. W. Horn Professor ofMarketing at Texas Tech University, Lubbock, Texas. A past edi-tor of the Journal of Marketing (1985-1987), he is the author ofModern Marketing Theory: Critical Issues in the Philosophy ofMarketing Science (South-Western, 1991) and A General Theoryof Competition: Resources, Competences, Productivity, Eco-nomic Growth (Sage Publications, 2000). He has written numer-ous articles on competitive theory, macromarketing, ethics,channels of distribution, philosophy of science, and marketingtheory. Three of his Journal of Marketing articles, “The Natureand Scope of Marketing” (1976), “General Theories and Funda-mental Explananda of Marketing” (1983), and “The Compara-tive Advantage Theory of Competition” (1995) (with Robert M.Morgan) won the Harold H. Maynard Award for the “best articleon marketing theory.” His 1985 Journal of Business Research ar-ticle with Lawrence B. Chonko, “Ethics and Marketing Manage-ment,” received the 2000 Elsevier Science Exceptional Qualityand High Scholarly Impact Award. His 1989 article, “Reificationand Realism in Marketing: In Defense of Reason,” won the Jour-nal of Macromarketing Charles C. Slater Award. For his contri-butions to theory and science in marketing, he received the 1986Paul D. Converse Award from the American Marketing Associa-tion, the 1987 Outstanding Marketing Educator Award from theAcademy of Marketing Science, and the 1992 American Mar-keting Association/Richard D. Irwin Distinguished MarketingEducator Award.
Yilmaz, Hunt / SALESPERSON COOPERATION 357
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001Sarkar et al. / ALLIANCE PERFORMANCE
The Influence of Complementarity,Compatibility, and RelationshipCapital on Alliance Performance
MB SarkarRaj EchambadiUniversity of Central Florida
S. Tamer CavusgilMichigan State University
Preet S. AulakhTemple University
Value creation through alliances requires the simulta-neous pursuit of partners with similar characteristics oncertain dimensions and different characteristics on otherdimensions. Partnering firms need to have different re-source and capability profiles yet share similarities intheir social institutions. In this article, the authors empiri-cally examine the impact of partner characteristics on theperformance of alliances. In particular, they test hypothe-ses related to both direct impact of partner characteristicson alliance performance and indirect effects through rela-tional capital aspects of the alliance. Empirical resultsbased on a sample of alliances in the global constructioncontracting industry suggest that complementarity inpartner resources and compatibility in cultural and opera-tional norms have different direct and indirect effects onalliance performance. Accordingly, organizational rou-tines aimed at partner selection need to be complementedby relationship management routines to maximize the po-tential benefits from an alliance.
The rise of hybrid organizational forms, or interfirmalliances (Borys and Jemison 1989), reflects attempts by
organizations to cope with the discontinuities created by avolatile, interdependent, and information-intensive globaleconomy. The quest for sustainable competitive advantagethrough competing and collaborating simultaneously inglobal markets (Lado, Boyd, and Hanlon 1997) has pro-pelled the formation of collaborative relationships thatexhibit governance structures very distinct from tradi-tional vertically integrated forms. However, along with theproliferation of alliances has come the realization thatmany alliances underperform and fail to deliver results rel-ative to expectations or potential (Madhok and Tallman1998).
Recent research suggests that the success of bothdomestic and cross-border collaborations may be a func-tion of partner characteristics (Hitt, Dacin, Levitas,Arregle, and Borza 2000; Madhok 1995; Saxton 1997).However, as Hitt et al. (2000) note, further research isrequired in this area of inquiry. In this article, we addresshow different types of interfirm diversity among partners(Parkhe 1991) affect the performance of alliances. We sug-gest that the issue of partner selection presents firms with apotential paradox, wherein seemingly contradictory ele-ments need to coexist and be simultaneously achieved toconvey a more illuminating “insight into truth than eithercan muster in its own right” (Slaate 1968:4). Specifically,we suggest that collaborative value creation requires thepursuit of partners who possess similar characteristics on
certain dimensions and dissimilar and/or complementarycharacteristics on other dimensions. Value generated fromalliances is enhanced when partners have differentresource and capability profiles yet share similarities intheir social institutions. These partner characteristics areimportant since they help in the formation of relationshipcapital or the behavioral and sociopsychological aspects ofan alliance that find expression in relational dynamicssuch as mutual trust, commitment, and informationexchange (Cullen, Johnson, and Sakano, 2000; Heide andJohn 1992).
Our empirical study is set in the global constructioncontracting industry. This industry is particularly appro-priate since project-based alliances play a crucial role inthis industry. The literature on interorganizational collabo-rations has been criticized for its relatively narrow focuson equity-based joint ventures and for ignoring looselystructured alliances where separate legal entities are notformed (Cullen, Johnson, and Sakano 1995). In such alli-ances, partner characteristics and relationship-based gov-ernance mechanisms are likely to assume greater salienceas coordinating mechanisms since bureaucratic lines ofcontrol are typically absent. In this regard, the constructioncontracting industry offers an appropriate context.
The layout of the article is as follows. We start with a lit-erature review and then present our case for consideringtwo types of interfirm diversity. Thereafter, we present ourmodel and our hypotheses, describe data collection con-text and procedures, and discuss our methods. We then testour model and end with a Discussion and Implicationssection.
The importance of interfirm collaborations is reflectedin a virtual explosion of interdisciplinary research on thistopic. Existing research has examined structural andsociopsychological aspects of collaborations to betterunderstand issues related to performance of alliances(Aulakh, Kotabe, and Sahay 1996; Parkhe 1993). Struc-tural aspects focus on the ex ante aspects of partnerships.This includes investigating why firms enter into alliances(Hagedoorn 1993), partner selection criteria, and owner-ship control issues. While this research stream providesimportant insights into the structuring of partnerships(Aulakh et al. 1996), an underlying premise here is thateffective interorganizational alliances are associated withselection of appropriate partners since choosing partnerswho possess necessary resources and with whom strategicand economic incentives can be aligned is a critical deter-minant of partnering success. This perspective suggeststhat the resource-based interdependence between partnerfirms, as well as their social compatibility, creates an
environment that facilitates the achievement of joint goalsand objectives (Aulakh et al. 1996; Parkhe 1993).
Interactive theorists (Heide and Miner 1992), on theother hand, focus on the “pattern of interaction that facili-tates and allows for the effective functioning of the alli-ance on a day-to-day basis” and various sociopsychologi-cal factors that help create relationship capital (Cullenet al. 2000:224). Complementing the structural approachby explicitly considering the sociopsychological dimen-sions, this stream highlights behavioral issues relevant tothe development and maintenance of relationships(Bradach and Eccles 1989; Johnson, Cullen, Sakano, andTakenouchi 1996; Madhok 1995). The key premise here isthat cooperative behavior springs from the development ofrelational capital between partners, which is critical intransforming the potential value of an alliance into actual-ized collaborative economic rents (Madhok and Tallman1998).
In essence, extant research housed in multiple disci-plines and using different theoretical lenses suggests anassociation between structural aspects of partners (theirdiversity in resource profiles and social compatibility),sociopsychological issues (relationship capital), andeffective collaborations (performance of alliances). Somepast research has shown (Heide 1994; Johnson et al. 1996)that structural and relational aspects are related, suggest-ing the possibility that structural aspects affect allianceperformance both directly and indirectly through thesociopsychological aspects of a relationship. However, theinterrelationships between these variables and how theyaffect performance are less clear. For example, it is unclearwhether interfirm diversity affects performance, and if itdoes, whether it influences performance directly, indi-rectly through relationship capital, or both. Furthermore,do different aspects of partner characteristics have animpact on different performance dimensions differently?
We investigate this nexus between interfirm diversityand alliance performance. In our conceptualization, wedraw on Parkhe’s (1991) conceptualization of interfirmdiversity in terms of Type I diversity (complementaryresources and capability profiles) and Type II diversity(social dimensions). Type I diversity deals with the recip-rocal strengths of the partners and relates to differences intheir skills, resources, and capabilities that generate alli-ances in a search for synergy. The complementarity ofresources, which is the raison d’être of any alliance, cre-ates mutual interdependency and “facilitates the forma-tion, development and collaborative effectiveness” of alli-ances (Parkhe 1991:580). On the other hand, alliances are“socially contrived mechanisms for collective action,which are continually shaped and restructured by actionsand symbolic interpretations of the parties involved”(Ring and Van de Ven 1994:96). Accordingly, Type IIdiversity refers to interorganizational cultural and pro-
Sarkar et al. / ALLIANCE PERFORMANCE 359
cessual differences between social actors participating inthe alliance (Parkhe 1991). Dissimilarities between socialactors can negatively affect the quality of interactions in apartnership and thus hinder the complex integration andtransformation of disparate pools of tacit know-how intovalue creation.
In addition, given that “the production of a collectivegood is inextricably intertwined with the underlyingdynamics of exchange” (Madhok and Tallman 1998:327),behavioral aspects of alliances need to be considered. Infact, the genesis of interfirm cooperation is based on thepremise that competitive advantage accrues to firms thatcan successfully transcend transaction-based exchangeand develop long-term cooperative relationships. A grow-ing body of relationship marketing literature has con-cluded similarly. Researchers have questioned the domi-nant paradigm of the discrete transaction and have positedthat interfirm exchanges take place in a context of continu-ity where relational constructs such as trust and commit-ment are key (J. Anderson and Narus 1990; Bucklin andSengupta 1993; Gundlach, Achrol, and Mentzer 1995;Heide and John 1992; Morgan and Hunt 1994). Consistentwith this literature, we believe that relational aspects medi-ate the relationship between interfirm diversity and collab-oration performance.
In this section, we first articulate the two types ofinterfirm diversity, namely, resource complementarity (orParkhe’s Type I interfirm diversity) and cultural and opera-tional compatibility (or Parkhe’s Type II interfirm diver-sity) and examine the rationale for their direct effects onperformance.1 Thereafter, we examine the mediatingeffects of relational-capital variables (mutual trust, recip-rocal commitment, and bilateral information exchange) onthe relationship between interfirm diversity and perfor-mance. Our conceptual model is presented in Figure 1.
Relationships Between InterfirmDiversity and Performance
Performance. While researchers have used financial,survival (Killing 1983), duration (Kogut 1988), and own-ership instability (Gomes-Casseres 1987) as measures ofcollaborative performance, Geringer and Hebert (1989)note the lack of consensus regarding an appropriate defini-tion and measure of the construct. Given the multi-dimensionality of the concept and the empirical setting,we use two perceptual measures of performance—one re-lated to the economic performance of the venture, namely,project performance, and the other related to the strategicaspects of the relationship, namely, strategic performance.
Our rationale is as follows: it has been noted that alliancesprovide a focal firm with two streams of economic rents.First, alliances create common or shared benefits that ac-crue collectively to alliance participants (Khanna 1998).Second, collaborations create rents indirectly, as when afirm picks up skills from a specific partner and appliesthem to its operations in areas unrelated to the activities ofthe specific alliance. Conceptualized as private benefits,these accrue when externalities from an alliance generatevalue in a firm’s operational domain that falls outside thefocal collaboration (Khanna, Gulati, and Nohria 1998).The first measure, project performance, relates to the for-mer, while the latter (i.e., strategic performance) refers tothe private benefits that accrue to a focal firm through stra-tegic learning-related benefits.
Type I interfirm diversity: Resource complementarity.Research indicates that resource complementarity is cru-cial to collaborative success (Bleeke and Ernst 1991;Harrigan 1985). As noted by Johnson et al. (1996), re-source complementarity involves both uniqueness andsymmetry. On one hand, complementarity determines themix of unique and valuable resources available to achievestrategic objectives (Killing 1983), thus enhancing com-petitive viability of the alliance. On the other, comple-mentarity implies strategic symmetry, wherein a balancedshare of unique strengths creates partner interdependence(Harrigan 1985). We conceptualize resource comple-mentarity as the extent to which each partner brings inunique strengths and resources of value to the collabora-tion (Johnson et al. 1996).
From a resource-based view (RBV) (Wernerfelt 1984),differential firm performance is due to heterogeneity andimperfect mobility of resource and capabilities (Barney1991). The two perspectives within the RBV, namely, thestatic and dynamic views (Lado et al. 1997), differ in theirexplanation regarding how economic rents are generatedand sustained. The static view emphasizes sustainabilityas accruing from unique firm resources in a state of equi-librium. For example, Barney’s (1991) argument thatresources need to be rare, inimitable, valuable, andnonsubstitutable for sustained competitive advantage isrooted in the static view. On the other hand, arguing thatchanging environments reduce the rent-creating ability ofmost resources, the dynamic RBV emphasizes flow andthe dynamic accumulation of capabilities, rather thanstatic resource stocks (Dierickx and Cool 1989). Together,they suggest that both stock and flow of resources andcapabilities are crucial to performance.
Furthermore, it has been noted that many skills andresources required for sustained competitiveness and sur-vival are transorganizational in nature, in that they are resi-dent outside a focal firm’s boundaries and direct control(Achrol 1997) and are thus accessible only through
360 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
collaborations. Therefore, extending the atomistic firmview that is prevalent in much of RBV literature, it isargued that organizations that are able to access and usecomplementary transorganizational strategic assets2
(TSA) through alliances would realize strategic advan-tages in both the stock of resources that they possess andthe flow of learning that can enhance their capabilities(Dyer and Singh 1998; Lado et al. 1997).
First, by pooling complementary resources and capa-bilities, firms can initiate and perform competitively onprojects that they could not have done alone (Harrigan1985). Accessing complementary resources through mar-ket mechanisms is not always feasible, nor is internal de-velopment (Chung, Singh, and Lee 2000; Sarkar,Echambadi, and Harrison 2001). For example, in the pres-ent context of the construction industry, no single com-pany has the complete array of resources to individuallydevelop and deliver projects and to fully absorb the magni-tude of risk. Second, interactive learning opportunities thathelp firms add to their capabilities and know-how arelikely to be greater in cases where there is diversity and
nonredundancy in knowledge bases. Accordingly, weargue that the potential for partners to synergistically le-verage the pooled resources and capabilities in the market-place would increase with resource complementarity. Inother words, when partners bring in unique and valuablestrengths and resources, both the learning aspects of the al-liance, as well as the performance of the project for whichthe alliance has been created, are likely to be enhanced.Thus,
Hypothesis 1a-b: Resource complementarity betweenpartners will be positively associated with (a) pro-ject and (b) strategic performance.
Type II interfirm diversity: Cultural and operationalcompatibility. The effect of partner compatibility on creat-ing value through alliances has been noted (Madhok1995). Compatibility, or the congruence in organizationalcultures and capabilities between alliance partners, influ-ences the extent to which partners are able to realize thesynergistic potential of an alliance (Madhok and Tallman1998). While alliance formation is an outcome of per-
Sarkar et al. / ALLIANCE PERFORMANCE 361
H1a(+), H2a(+), H3a(+)
Relationship Capital Performance
H1b(+), H2b(+), H3b(+)
FIGURE 1A Conceptual Model of the Role of Complementarity, Compatibility,
and Relationship Capital on Alliance Performance
ceived recognition of potential benefits that can accrue ifresources and capabilities are pooled, the actualization ofthis collaborative potential is generated through the dy-namic process of interaction and integration of the part-ners’ resource bases and the effectiveness with which thepartners succeed in moving away from a market-based ex-change toward a mutually oriented cooperative relation-ship (Koza and Lewin 1998).
In this context, it has been noted that compatibility, or asimilarity in outlook and objectives, rather than ownershiparrangements, makes for success in equity-based alliances(Friedmann and Beguin 1971). Similar organizational val-ues reduce coordination costs between collaborating orga-nizations and serve as a means for behavioral control (Dasand Teng 1998) and expectation management (Chunget al. 2000). On the other hand, incompatibility amongpartners may lead to a counterproductive working rela-tionship characterized by strife and suspicion. Socialincompatibility may lead to an inability on the part of thepartners to develop a harmonious relationship and thusnegatively influence collaborative effectiveness (Sarkar,Cavusgil, and Evirgen 1997). For example, it has beennoted that cultural clash has caused many mergers andacquisitions to fail due to the inability of the two entities towork seamlessly (Wilkof, Brown, and Selsky 1995).Higher levels of stress result when organizations that areessentially incompatible in their values, norms, and capa-bilities attempt to blend their organizational cultures in analliance (Das and Teng 1998). Organizational differenceshinder role socialization (Smith and Barclay 1997), thusmaking it more difficult for interfacing managers to worktogether.
Therefore, there appears theoretical and empirical sup-port behind the idea that organizational compatibility invarious domains has a positive effect on alliance perfor-mance. We conceptualize Type II diversity or organiza-tional compatibility along two dimensions, namely,cultural compatibility and operational compatibility. Cul-tural compatibility refers to the congruence in organiza-tional philosophies, goals, and values. The second,operational compatibility, addresses the extent of congru-ence in the partners’ procedural capabilities. Therefore,while the first dimension addresses broad issues related toorganizational norms and value systems, operational com-patibility relates to status similarity on capability andprocessual issues that assume salience in the context of aworking relationship. Thus, we posit that both forms ofcompatibility will directly affect performance.
Hypothesis 2a-b: Cultural compatibility of the partnerswill be positively associated with (a) project and (b)strategic performance.
Hypothesis 3a-b: Operational compatibility of the part-ners will be positively associated with (a) projectand (b) strategic performance.
The Mediating Role of Relationship Capital
Besides the direct impact of structural factors (resourcecomplementarity and partner compatibilities) on allianceperformance, there is evidence that partner characteristicsindirectly affect performance through certain mediatingbehavioral variables. Researchers have argued and foundempirical support for (a) the effect of relationship-capitalvariables on alliance outcomes (e.g., Aulakh et al. 1996;Bradrach and Eccles 1989) and (b) links between partnercharacteristics and relationship capital (Cullen et al. 1995;Morgan and Hunt 1994; Stump and Heide 1996). Thesociopsychological aspects embodied in relationship capi-tal are important since they act as coordinating mecha-nisms and determine the quality of the relationship in thecollaboration. In fact, it has been suggested that interfirmcooperation can lead to competitive advantage only whenfirms transcend transaction-based exchange and developlong-term relationships (Dyer and Singh 1998). We con-sider three key aspects of relationship capital, namely,mutual trust, mutual commitment, and informationexchange, which have been highlighted in the literature asfactors that differentiate relationship-based practices fromarm’s-length exchange (Heide and John 1992; Morgan andHunt 1994).
Mutual trust. Interfirm trust, which has been describedas a “fundamental relationship building block” (Wilson1995:337) and as a critical element of economic exchange(Ring 1996), is argued to be essential to the developmentof enduring alliances (Aulakh et al. 1996; Johnson et al.1996). In economic exchange, trust implies a general ex-pectation of good faith efforts by parties to honor commit-ments, to be honest in negotiations, and to decryopportunistic behavior (see Hosmer 1995). An extensiveliterature (see Ring 1996) points toward two issues rele-vant to this study: (1) that interfirm trust needs to be char-acterized by the “property of bilateral expectations”(Aulakh et al. 1996:1008) and mutuality (E. Anderson andWeitz 1989) and (2) that trust has both cognitive and be-havioral components (Moorman, Zaltman, and Deshpande1992). Consistent with this literature, we conceptualizemutual trust in a partnership as the degree of confidenceshared by the partners regarding each other’s integrity(Aulakh et al. 1996).
Reciprocal commitment. Commitment, or “an ex-change partner believing that an ongoing relationship withanother is so important as to warrant maximum efforts atmaintaining it . . . to ensure that it endures indefinitely”(Morgan and Hunt 1994:23), is a critical element of rela-tionship capital (Madhok 1995). This “enduring desire tomaintain a relationship” (Moorman et al. 1992:316) in-volves a long-term orientation such that partners restricttheir search for alternatives and forego better short-termoptions in favor of strengthening an ongoing relationship
362 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
(Dwyer, Schurr, and Oh 1987). These continuity expecta-tions influence partners to make relationship-specific in-vestments that, on one hand, demonstrate their reliabilityand commitment to their exchange partner, and on theother, enhance the competitiveness of the alliance (E. An-derson and Weitz 1992). Consistent with literature, we be-lieve that it is not the act of commitment alone but alsorather the structure of commitment that fashions relation-ship quality (E. Anderson and Weitz 1992). Accordingly,we conceptualize reciprocal commitment as the degree towhich both partners are willing to invest requisite re-sources into the alliance (Gulati, Khanna, and Nohria1994).
Bilateral information exchange. Collaborative commu-nication serves as a pseudointegrating device since it helpsalign partners’ interests and values (Mohr, Fisher, andNevin 1996). Acting as a bonding mechanism betweenpartners, the exchange of timely, quality, and participativecommunication is argued to be vital to successful collabo-rations (Mohr and Nevin 1990). In fact, in the marketingliterature, it has been described as “the glue that holds to-gether a channel relationship” (Mohr and Nevin 1990:36)and that facilitates the realization of mutual benefits by al-lowing exchange of necessary information and by reduc-ing misunderstandings and uncertainty (Dwyer et al.1987). We conceptualize reciprocal information exchangein terms of formal and informal communication of mean-ingful and timely information (J. Anderson and Narus1990). In doing so, we are consistent with existing litera-ture that has conceptualized collaborative communicationin terms of its quality, extent, and participation (Mohr andSpekman 1994).
Relationship capital is likely to be fostered when part-ners perceive a high level of complementarity and compat-ibility. The principle of reciprocity implies that actions arecontingent on the potential of rewarding reactions fromothers. Thus, in relationships where partners need eachother’s resources and where reciprocal needs exist, part-ners are less likely to resort to opportunism. The resourceinterdependence created through Type I diversity is likelyto result in reciprocity and thus reduce incentives foropportunistic behavior, as both partners perceive value intheir relationship (Morgan and Hunt 1994; Stump andHeide 1996). Resource-interdependent partners are morelikely to be motivated to create relationship capital byengaging in trustworthy acts that increase their vulnerabil-ity to each other, signaling their expectations of continuityand solidarity to the relationship by committing relation-ship-specific resources and maintaining open andparticipative lines of communication.
Furthermore, symmetric dependence motivates bothparties to jointly show forbearance (Williamson 1991) orflexibility in response to changing circumstances, whichin turn has the effect of preserving the relationship through
the generation of relational norms (Macneil 1980). Thecondition of dependence that results from Type I diversity,where both partners need each other’s resources and capa-bilities, represents a mutual safeguard and thus a jointincentive to create relationship capital in the collaboration(Oliver 1990). The resulting strategic symmetry facilitatesthe formation and development of cooperative norms(Madhok 1995).
Evans’s (1963) similarity hypothesis suggests that in adyadic relationship, the degree of similarity is positivelyassociated with favorable relationship outcomes. Similar-ity in values serves as a base for social relationships, whichlie at the heart of social interaction processes. The resul-tant bonds create stability in the relationship and greaterlevels of tolerance through a “social ‘glue’ [that helps] totide over temporary periods of disequilibrium” (Madhok1995:121). Type II diversity or organizational compatibil-ity, by facilitating a sense of unity and congeniality in therelationship, is thus likely to foster relational-capital-building behaviors among partners. This would be espe-cially pertinent in international alliances where culturaldifferences are likely to exist. Accordingly,
Hypothesis 4a-c: Resource complementarity will be pos-itively associated with (a) mutual trust, (b) recipro-cal commitment, and (c) bilateral informationexchange.
Hypothesis 5a-c: Cultural compatibility will be posi-tively associated with (a) mutual trust, (b) reciprocalcommitment, and (c) bilateral information ex-change.
Hypothesis 6a-c: Operational compatibility will be posi-tively associated with (a) mutual trust, (b) reciprocalcommitment, and (c) bilateral information ex-change.
Relationship Between RelationalCapital and Performance
Existing literature identifies various interrelated waysin which mutual trust affects interfirm exchanges. Mutualtrust acts as a substitute for hierarchical governance andassumes added significance where formal ownership-based governance is absent (Dwyer et al. 1987). Trust al-lows for bilateral governance through joint accomplish-ments, shared beliefs, and mutual concern (Heide 1994).Mutual trust also deters opportunistic behavior (Bradachand Eccles 1989) in favor of long-term gains. The motiva-tion for opportunistic behavior is reduced because “behav-ioral repertoires are biased toward cooperation” (Hill1990:511). Finally, “trust has efficiency implications, andits potential cost reduction and value enhancing propertiesneed to be recognized” (Madhok 1995:126). There is evi-dence that trust has important implications for market per-formance and efficiency (Aulakh et al. 1996; Bleeke andErnst 1991). This occurs because of reduced costs of mon-
Sarkar et al. / ALLIANCE PERFORMANCE 363
itoring and the ability of partners to engage in the complexprocess of integrating their disparate tacit resources andcapabilities effectively when the relationship is character-ized by mutual trust (Dyer and Singh 1998). Accordingly,
Hypothesis 7a-b: Mutual trust will be positively associ-ated with (a) project and (b) strategic performance.
The positive effect of commitment on collaborativeperformance has been widely reported (Badaracco 1991;Gundlach et al. 1995). Given that the competitive advan-tage of any alliance derives from the mix of skills and otherresources that it commands, it is likely that commitment ofresources by partners has a positive effect on financial per-formance. Furthermore, Gulati et al. (1994) argue that bi-lateral commitment of resources moves alliances fromwin-lose situations to win-win situations, thus suggestingthat reciprocal commitment is likely to enhance partners’perceptions of how successful the relationship has been.Reciprocal commitment of inputs leads to stable long-term relationships through aligning incentive structuresand enhancing confidence in each other (Williamson1985). By reducing the threat of opportunistic behaviorand increasing the cost of dissolution, commitments byboth parties act as powerful signals of relationship quality.Long-term relationships reduce search and start-up costsof frequently dealing with new parties. Along with result-ing economies of learning costs and experience effects,they also require simpler governance structures, monitor-ing systems, and “provide a host of efficiencies”(Gundlach et al. 1995:80). Also, the “lock-in” effect of re-ciprocal commitment (Katz 1989) promotes behavior thatensures the continuance of the relationship since the quasi-rents that are generated through the relationship would belost to both parties in the event of termination (Heide1994). Accordingly,
Hypothesis 8a-b: Reciprocal commitment will be posi-tively associated with (a) project and (b) strategicperformance.
It has been noted that healthy interfirm collaborationsare characterized by open communication, accessibility,availability, information flows, and a sense of participa-tion and involvement in the relationship (Mohr and Nevin1990; Mohr et al. 1996). These attributes create transpar-ency in the relationship and signal a mutual willingness toincrease vulnerability to each other. Furthermore, commu-nication facilitates the realization of mutual benefits by al-lowing exchange of necessary information and byreducing misunderstandings and uncertainty (Dwyer et al.1987; Mohr and Nevin 1990). Mutual disclosure ensuesfrom a norm of information exchange (Heide and John1992) and helps volitional compliance between partners. Ithighlights shared interests and common goals (Mohr et al.
1996) and thus positively affects collaboration perfor-mance (Badaracco 1991). Information asymmetry andparticipatory imbalance create an environment prone toopportunism and power imbalances, whereas sharedpower and participative decision-making are characteris-tics of successful alliances (Bucklin and Sengupta 1993).In other words, participative and frequent exchange of in-formation and maintaining open-door policies to eachother result from a willingness of the partners to createtransparency in the relationship.
Hypothesis 9a-b: Bilateral information exchange will bepositively associated with (a) project and (b) strate-gic performance.
Sample Selection and Data Collection
The sampling frame included 561 firms in the interna-tional construction contracting industry. EngineeringNews Record (ENR), a McGraw-Hill published weeklytrade journal for the construction contracting industry,periodically collects and publishes industry data. Theuniqueness of the construction industry lies in its charac-teristics. First, each construction product is different interms of specifications, scope, and requirements. Thepotential to standardize complex construction projects islimited due to the individualized preferences of the projectowners, dissimilarities of the project locations, and theadvancement of technology. As a corollary, most construc-tion firms tend to be specialized in certain product niches.Second, there is no centralized or fixed point of production(Eccles 1981). The construction product is spatially boundto a single site, which happens to be both the point of pro-duction and consumption of the product. Third, variabilityin location leads to the requirement of resource mobility.Equipment, vehicles, materials, and labor need to bemobile to be able to move from one site to another for dif-ferent projects. Fourth, both the degree of competition andthe competitive structure of this industry have dramati-cally changed over recent years as U.S. dominance isthreatened by firms from within the Triad and from thenewly industrialized countries (NICs) and some develop-ing countries (Yates, Mukherjee, and Njos 1999).
International alliances are very common in this indus-try, where the complexity of projects makes it imperativefor firms to collaborate on various modules with other spe-cialist firms. Typically, the alliances are loosely structuredand not equity based, thus making it particularly appropri-ate in the wake of Cullen et al.’s (1995) critique that muchof the collaboration literature focuses on equity-basedjoint ventures and ignores alliances where separate legalentities are not formed. The globalized nature of this
364 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
industry, its reliance on international collaborations, andits understudied nature motivated us to set the empiricalpart of this study here.
We followed a systematic approach toward pretesting,refining, and validating our scales. In-depth semistruc-tured interviews, lasting about 1 hour each, were con-ducted with 12 experts from industry associations and con-struction contracting firms to get insights into industrydynamics and to ensure that conceptual constructs weregrounded in reality. Following the process of adaptingexisting measures and adapting developing new ones, wepilot tested the survey instrument on these industry execu-tives to eliminate ambiguous scale items. Using this feed-back, a revised instrument was developed, and feedbackwas obtained from academic experts regarding item clar-ity. The questionnaire was subsequently modified andfinalized through an iterative process.
We derived our sampling frame from the 1993 list of thetop 225 international contractors and the top 400 U.S. con-tractors. Since some of the U.S. contractors were includedin both lists, the total number of firms that emerged was561. Data were collected using a self-administered ques-tionnaire prepared in English. The survey cover page elab-orated the fact that the survey was about nonequity collab-orative ventures with other contractors, and not aboutrelationships involving contractors and project owners.Furthermore, respondents were requested to answer rele-vant questions in the context of a collaborative venture thatthey were most familiar with and that involved an interna-tional partner. It was elaborated that the project should beone that either has been recently completed or is nearingcompletion. Sixty-eight usable questionnaires werereturned, making the effective response rate 12.3 percent.We received responses from firms in 18 countries otherthan the United States.3 To assess nonresponse bias, theresponses were divided into two groups based on the dateon which they were received. The two groups of early andlate responders were then compared on their annual bill-ings (Armstrong and Overton 1977). The results indicateno significant difference between these groups.
All the items used to measure the constructs wereclosed-ended with 5-point Likert-type scales of stronglyagree to strongly disagree.
Type I and II interfirm diversity. Resource comple-mentarity was measured through a three-item scale thatwas adapted from J. Anderson and Narus (1990) into a pro-ject setting. The items tapped the level of resource interde-pendence in the relationship by measuring the extent towhich both partners perceived the value of resources andcapabilities that the other brought to the relationship. Cul-tural compatibility was operationalized through a four-
item scale that measured the perceived levels of similarityand congruence in organizational norms and values(Dwyer et al. 1987; Heide and John 1992; Morgan andHunt 1994) and mutual appreciation of each other’s goalsand objectives. Operational compatibility was operation-alized through a three-item scale that measured the level ofcongruence in the partners’ managerial skills, organiza-tional procedures, and technical capabilities (Sarkar et al.1997; Wilson 1995).
Relationship capital. Mutual trust was measuredthrough a four-item scale that assessed the perceived levelof moral integrity, fairness, and dependability in the rela-tionship. Reciprocal commitment was measured through athree-item scale adapted from E. Anderson and Weitz(1992) and the organizational-commitment literature. Thescale measured the mutual willingness of each partner toinvest required resources into the relationship. Bilateralinformation exchange was measured through a four-itemscale adapted from Heide and John (1992). This constructtapped the extent to which partners exchanged and sharedinformation through face-to-face and mediated interaction.
Performance. We developed a three-item scale for stra-tegic performance and a four-item scale for project perfor-mance. We measured strategic performance through itemsthat assessed the degree to which the relationship met stra-tegic and learning objectives. The latter were measuredthrough perceived profitability, efficiency, client satisfac-tion, and quality of the project.
We used Partial Least Squares (PLS) version 3.0 to esti-mate our causal model. PLS, also called “soft modeling”(Lohmöller 1989), estimates latent variables as exact lin-ear combinations of observed measures and thereforeassumes that all measured variance is useful variance to beexplained. PLS makes minimal demands on sample size(Barclay and Smith 1997), thus making it especiallyappropriate for testing structural models with relativelysmaller sample sizes.4
Although PLS estimates both factor loadings and struc-tural paths simultaneously, we followed the procedureadvocated by Hulland (1999) in evaluating PLS models.The estimated model was analyzed and interpreted in twostages: (a) the assessment and reliability of the measure-ment model and (b) the testing of the structural model. Weassessed the adequacy of the measurement model throughexamining individual-item reliabilities, the convergentvalidity of the measures associated with each construct,and assessing their discriminant validity.
We first assessed individual-item reliabilities by exam-ining loadings of the measures on their respective con-structs. A rule of thumb is to check for loadings of .70 ormore (which implies a shared variance of 50% or greater
Sarkar et al. / ALLIANCE PERFORMANCE 365
between the item and the construct). An examination ofthe initial measurement model revealed that of the 28items, 20 had loadings greater than .7, 5 items had loadingsgreater than .65, and 3 items had loadings of less than .5.These 3 items with poor loadings were removed from sub-sequent analysis. Table 1 provides the final list of individ-ual items used in the analysis and their loadings. Overall,these statistics are above the cutoff suggested by Hulland(1999) and indicate that all of our items demonstrate goodindividual-item reliabilities.
Next, we focused on assessing the construct validity ofour constructs by computing the composite reliabilities.We used the internal consistency measure developed byFornell and Larcker (1981),5 who argue that their measureof internal consistency is superior to Cronbach’s alphasince the loadings estimated within the causal model areused in its computation. The internal consistency valuesfor the constructs are reported in Table 2. All constructsexhibit composite reliabilities of .7 or more, thus indicat-ing that the reliabilities of all the constructs are adequate(Hulland 1999).
Finally, to complete the psychometric assessment ofour model, we examined the discriminant validity, whichrepresents the extent to which measures of a given con-struct differ from measures of other constructs in the samemodel. Fornell and Larcker (1981) suggest the use of aver-age variance (i.e., the average variance shared between aconstruct and its measures) extracted to assess discrim-inant validity.6 As shown in Table 2, the average variancesextracted in all the constructs were all at least or greaterthan .50, which is indicative of convergent validity(Barclay and Smith 1997). Also, the overall model pro-vided reasonable evidence of discriminant validity in thatthe variance shared between any two constructs was lessthan the average variance extracted by the constructs, andall measures loaded higher on intended constructs than onother constructs (Hulland 1999). Overall, these statisticsindicate that the psychometric properties of the model aresufficiently strong to enable interpretation of structuralestimates.
Since PLS does not attempt to minimize residual itemcovariance, there is no summary statistic to measure theoverall fit of models as in the case of SEM techniques.Variance explained (R2) and the sign and significance ofpath coefficients are used to assess nomological validity. Abootstrapping method of “sampling with replacement”was used to assess the statistical significance of the param-eter estimates. Standard errors were computed on the basisof 500 bootstrapping runs. Results of the structural modelare given in Tables 3 and 4. Table 3 provides the results ofthe direct effects of interfirm diversity on relationship-capital constructs. Table 4 presents the results of the
overall comprehensive causal model: interfirm diversityleading to relationship capital, subsequently leading toperformance. The direct effects of the various exogenousconstructs on performance are given in the first column.The indirect effects (reported in column 2) of interfirmdiversity on performance through relationship-capitalvariables were computed using a hand-calculable signifi-cance test proposed by Baron and Kenny (1986).7
Direct effects. An examination of the R2 values revealthat the variance explained in endogenous constructsranges from .15 to .55. The results indicate that resourcecomplementarity is related to project performance (β =.22, p < .05) but not to strategic performance (β = .09, p >.05). Hypothesis 1a is thus supported, but not Hypothesis1b. Results further indicate that while the direct effect ofcultural compatibility on project performance is not statis-tically significant (β = .03, p > 0.05), the effect on strategicperformance is (β = .50, p < .05). Hypothesis 2b is thussupported, but not Hypothesis 2a.
Contrary to expectations, operational compatibility isnot significantly related to project performance (β = .13,p > .05), thus indicating lack of support for Hypothesis 3a.Regarding the impact of operational compatibility on stra-tegic performance (β = –.25, p < .05), the path coefficientis statistically significant; however, the sign is reversed,thereby failing to support Hypothesis 3b.8
Hypotheses 4a, 5a, and 6a respectively hypothesize thatresource complementarity will positively affect mutualtrust, reciprocal commitment, and bilateral informationexchange. The results reflect interesting differences in theimpact of resource complementarity on various relationship-capital constructs. Contrary to expectations, resourcecomplementarity is not significantly related to trust (β =.07, p > .05) or bilateral information exchange (β = .01, p >0.05), thereby failing to support Hypotheses 4a and 6a.However, resource complementarity has a significant rela-tionship with reciprocal commitment (β = .27, p < .05),supporting Hypothesis 5a.
Hypotheses 4b, 5b, and 6b respectively suggest thatcultural compatibility enhances relationship capital. Asexpected, cultural compatibility is significantly related tomutual trust (β = .40, p < .05), reciprocal commitment (β =.42, p < .05), and bilateral information exchange (β = .39,p < .05), thus supporting all three hypotheses.
Hypotheses 4c, 5c, and 6c respectively hypothesize thepositive impact of operational compatibility on trust, com-mitment, and bilateral information exchange. Asexpected, operational compatibility is significantly relatedto trust (β = .37, p < .05), and commitment (β = .19, p <.05), thereby supporting Hypotheses 4c and 5c. However,operational compatibility is not significantly related to
366 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
bilateral information exchange (β = –.02, p > .05), therebyfailing to support Hypothesis 6c.
Hypotheses 7a and 7b state that trust will be positivelyassociated with both project performance and strategicperformance. Trust is significantly related to project per-formance (β = .17, p < .05), thereby supporting Hypothesis 7a.However, contrary to expectations, trust is not signifi-cantly related to strategic performance (β = –.15, p > .05),
thereby failing to support Hypothesis 7b. Commitment issignificantly related to both project performance (β = .39,p < .05) and strategic performance (β = .30, p < .05),thereby supporting Hypotheses 8a and 8b. Furthermore,results indicate that reciprocal information exchange is notsignificantly related to project performance (β = .07, p >.05), thereby failing to support Hypothesis 9a. However,reciprocal information exchange is positively related to
Sarkar et al. / ALLIANCE PERFORMANCE 367
TABLE 1Measurement Model
Construct Itema Loading
Resource Both firms needed each other’s resources to accomplish their goals and responsibilities .66Complementarity The resources contributed by both firms were significant in getting the bid .91
Resources brought into the venture by each firm were very valuable for the other .93Cultural The organizational values and social norms prevalent in the two firms were congruent .77
Compatibility Executives from both firms involved in this project had compatible philosophies/approaches to business dealings .76The goals and objectives of both firms were compatible with each other .86The chemistry was right between the two firmsb —
Operational Technical capabilities of the two firms were compatible with each other .89Compatibility The organizational procedures of the two firms were compatible .82
Employees of both firms had similar professional or trade skills .67Mutual Trust Both firms were generally honest and truthful with each other .83
Both firms treated each other fairly and justly .80Both firms found it necessary to be cautious in dealing with each other (R) .80Relying on each other was risky for both firms (R)b —
Reciprocal There was frequent communication between the two firms (e.g., visits to each other’s firms, meetings, writtenInformation and telephone communications) .81Exchange Exchange of information in this relationship took place frequently and informally .91
Making contact with people from the other firm was hard for both firms (R) .69Decisions regarding the project were made unanimously in joint meetings with managers from both firmsb —
Reciprocal Both firms were willing to dedicate whatever people and resources it took to make this project a success .83Commitment Both firms provided experienced and capable people to the project .90
Both firms were committed to making this project a success .89Strategic The collaboration provided a very effective medium of learning .68
Performance Collaborating with this partner was a wise business decision .91Our strategic objectives going into the venture were achieved .85
Project The owner’s objectives (in terms of specifications, schedule, quality) were met .68Performance A quality job was done on the project .80
Overall, the project was efficiently carried out .84The venture was profitable for our firm .71
NOTE: (R) Indicates items that were reverse-coded.a. Scale ranging from 1 (strongly disagree) to 5 (strongly agree).b. Items that were deleted after initial tests.
TABLE 2Internal Consistency, Square Roots of Average Variance Extracted, and Correlation Matrix
NOTE: The diagonal (in italics) shows the square root of the average variance extracted for each construct.
strategic performance (β = .18, p < .05), thereby providingsupport for Hypothesis 9b.
Indirect effects. Preliminary tests suggested that condi-tions stipulated by Baron and Kenny (1986) for a media-tion model were satisfied,9 indicating that relationship-capital variables mediated the relationship between part-ner characteristics and alliance performance. We examinedthe contribution of the relationship-capital variables to theexplanatory power of the model. Specifically, we exam-ined the increase in R2s of the alliance performance con-structs when the relationship-capital variables wereincluded. The R2 of project performance increases from.45 to .55, while that of strategic performance increases
from .24 to .41. The increases for both project performance,F(3, 61) = 4.52, Fcrit = 2.76, and strategic performance,F(3, 61) = 6.20, Fcrit = 2.76, are significant at p < .05, thusindicating that relationship-capital variables contributesubstantially to the explanatory power of the model.
Next, based on the approach suggested by Baron andKenny (1986), we assessed the mediation by examiningthe size and significance of the indirect effects. Interest-ingly, for both performance constructs, all three antecedents—namely, resource complementarity, cultural compatibility,and operational compatibility—had significant indirecteffects. The standardized effect sizes were small tomedium (J. Cohen 1988), ranging from .06 to .22. Taken in
368 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
TABLE 3Effect of Interfirm Diversity on Relationship Capital: Standardized PLSa Coefficients
Relationship capitalMutual trust H7b + –.15 NA —Reciprocal commitment H8b + .30*** NA —Bilateral information exchange H9b + .18*** NA .21
NOTE: NA = not applicable.a. PLS = Partial Least Squares.b. Only statistically significant indirect effects were included in the computation.c. Only statistically significant effects (direct or indirect) were included.*** Significant at p < .05.
tandem, the statistical significance of the increase in R2sand the significant indirect effects together indicate theimportant role of relationship capital in explaining therelationship between partner characteristics and perfor-mance. Also, the presence of these indirect effects sug-gests that any omission of these variables from a theoreti-cal model could lead to an underestimation of the totaleffects of partner characteristics on performance. Spe-cifically, resource complementarity affects both projectperformance (β = .10) and strategic performance (β = .08),cultural compatibility affects both project performance (β =.22) and strategic performance (β = .20), and operationalcompatibility seems to influence both project performance(β = .13) and strategic performance (β = .06) through rela-tionship capital.10 In summary, the results indicate theimportance of considering both direct and indirect effectson alliance performance, thus giving further credence tothe theoretical rationale behind integrating both structuraland relational perspectives into an explanation of allianceperformance.
DISCUSSION AND IMPLICATIONS
Recent scholarship on international alliances has artic-ulated the need for more research partner selection issues(Hitt et al. 2000), especially because of their impact onalliance performance (Johnson et al. 1996; Saxton 1997).Although strategic alliances have proliferated throughoutthe world, a substantial proportion of these underperform(Madhok and Tallman 1998). In examining determinantsof alliance performance, we focus on a unique aspect asso-ciated with the characteristics of partners involved in analliance, namely, interfirm diversity (Parkhe 1991). Wesuggest that performance is likely to be enhanced whenfirms are able to manage the paradox involved in choosinga firm that is different, yet similar. Complementaryresource and capability profiles enhance the value gener-ated in alliances, as do similarity in the social institutionsof the partners. We therefore focus on three constructsrelated to interfirm diversity. Drawing on Parkhe’s (1991)work, we develop a multidimensional treatment ofresource complementarity, cultural compatibility, andoperational compatibility. We integrate extant interna-tional alliance literature that has traditionally examinedthe structural and sociopsychological aspects of alliancesseparately and develop a theoretical framework that sug-gests that the diversity-related characteristics of partnersaffect performance directly and indirectly through theireffects on relationship capital or sociopsychological vari-ables that are the focus of interactive theorists (Cullen et al.2000; Heide and Miner 1992). Our results provide aunique contribution to the understanding of how partnercharacteristics affect alliance-related performancebecause they suggest that various types of interfirm
diversity differentially affect performance and that modelsthat integrate both structural and relationship-capital vari-ables have greater explanatory power over the vexingquestion of alliance performance. We now summarize themain implications of what this research has found.
With regard to complementarity, our analysis indicatesthat the synergy that results when alliance partners pooltogether complementary resources and capabilitiesenhances performance. First, it enhances the economicefficiency and qualitative effectiveness of the task beingjointly carried out both directly and indirectly. While thedirect effect is stronger, there is a substantive indirecteffect, primarily through reciprocal commitment. It thusappears that when firms can partner with firms that cancomplement their weaknesses, not only is there a directeffect on project performance, but it also has the addedeffect of increasing the commitment of each partner to therelationship wherein they are willing to invest requisiteresources into the relationship to make it a success. Thisserves as a powerful signaling mechanism that reduces thethreat of opportunism, aligns incentive structures, and pro-vides a host of efficiencies. Second, although comple-mentarity does not have a direct effect on strategic perfor-mance (i.e., a perception that the relationship iscompetency enhancing and strategically beneficial), itdoes have an indirect effect, although weak, through themediating effect of relationship-capital variables. Thus,Type I interfirm diversity appears to have important impli-cations for performance. It enhances the efficiency andeffectiveness of the joint task being performed, or the com-mon benefits from the alliance, as well as more strategicprivate benefits that a focal firm may take out from therelationship. Furthermore, while its effect on the commonbenefits is primarily direct, its effect on private benefits ismediated through relationship capital. This appears plau-sible since mere complementarity may not lead to learningand knowledge transfer, which requires a certain depth ofinteraction and relationship quality for tacit know-how tobe transferred.
The effects of Type II diversity, namely, cultural andoperational compatibility, on performance are intriguing,thus highlighting the dilemma and complexity surround-ing partner selection issues. Cultural compatibilityenhances both project and strategic performance. It affectsproject performance indirectly through relationship-capi-tal variables and influences strategic performance bothdirectly and indirectly (with the direct effect being stron-ger). Also, the total effect on strategic performance isstronger than on project performance, while relative tocomplementarity, cultural compatibility has a strongereffect on strategic performance than on project perfor-mance. The results suggest that when partners in an alli-ance share similar organizational cultures, they are likelyto enjoy a better quality of relationship, which in turn willfacilitate an effective intermingling of skills and
Sarkar et al. / ALLIANCE PERFORMANCE 369
competencies and ensure that the project is efficiently andeffectively carried out. On the other hand, the strategiclearning that a focal firm can achieve from its partner or theprivate rents that it generates from the focal alliance isdirectly affected by the cultural congruence of thepartners.
On the other hand, operational compatibility has a posi-tive indirect effect on project performance, implying thatcompatibility in procedural capabilities enhances the qual-ity of the relationship and thus increases the efficiency andeffectiveness of the project execution. However, it has asurprising negative direct impact on strategic perfor-mance, which is partly mitigated by a weaker positiveeffect through the relational mediators. This counter-iintuitive result is intriguing in that it suggests the follow-ing: although common benefits accruing from an allianceare enhanced when levels of professional skills, technicalcapabilities, and operational procedures of two partnersconverge, private benefits from the alliance may be ham-pered. We offer a speculation based on the learning raceanalogy of Hamel (1991). Operating at similar levels oftechnology and skills (albeit the specific technology,skills, and capabilities may be different) is likely toincrease the absorptive capacity of a firm and enhance itsability to recognize, assimilate, and commercialize exter-nal information (W. Cohen and Levinthal 1990) from itspartner. Accordingly, to protect itself from redundancy, afocal firm may be wary of passing information and know-how that it considers critical to partners that possess highlevels of absorptive capacity. In response to the perceivedthreat, they may clamp down with procedures that limittransfer of information and know-how beyond what isimmediately relevant for project execution to their part-ners. Thus, although operational compatibility translatesinto better project management and execution, it has a neg-ative effect on strategic performance. However, thisintriguing result needs to be examined in future research.Interestingly, this research suggests that different types ofcompatibility related to the social institutions of the part-ners could have very different effects on various aspects ofalliance performance.
Furthermore, the indirect effects suggest the impor-tance of the mediating relationship-capital variables.Taken individually, the results suggest that reciprocalcommitment and mutual trust enhance the level of directcommon benefits from an alliance, while reciprocal com-mitment and bilateral information exchange increase theprivate benefits that accrue from an alliance. At a moremacro perspective, however, our research suggests that theissue of partner compatibility, which is influenced by orga-nizational routines related to partner selection, by itself can-not maximize benefits that can emerge from an alliance.The results indicate the strong impact of relationship-
capital variables on alliance performance and thus high-light the importance of alliance management capabilities.In other words, the performance-enhancing effect of com-patibility is enhanced even further when firms make con-scious efforts to create relationship capital and embed therelationship within certain sociopsychological statesthrough an interaction process designed to specificallyimprove the quality of the relationship.
The findings from this study have important manage-rial and theoretical implications. From a normative per-spective, the results suggest the relative importance of var-ious aspects in choosing the appropriate alliance partner.In particular, allying with firms with complementaryresources is likely to ensure success of the particular pro-ject (or shared benefits), while finding partners with simi-lar cultural norms is more important to achieve strategicprivate benefits. Thus, criteria used in partner choice canbe guided by specific firm objectives in forging cross-bor-der alliances. From a theoretical perspective, our findingsclarify the relative importance and interrelationshipsbetween partner characteristics and relational aspects inexplaining alliance performance. Much of the priorresearch has generally considered these two aspects asalternative ways to improve alliance performance. Ourfindings, by examining the direct and indirect effects ofpartner characteristics on performance, highlight the needto examine both simultaneously.
While the study makes important contributions to thealliance literature, several potential limitations should benoted. First, we undertook a cross-sectional approach todata collection. This prohibits studying the temporalaspects of a relationship. Second, we collected informa-tion from only one side of the dyad. To what extent percep-tions would have converged is unknown. Third, we usedsingle informants, a procedure less rigorous than the use ofmultiple informants (Kumar, Stern, and Anderson, 1993).Fourth, our sample size is relatively small. Although theclosed nature of the industry partly explains the reasonbehind the low response rates, the sampling frame is alsolimited. The industry-specific nature of the study makes usbelieve that at the very least, our data do not suffer from thewide heterogeneity that characterizes most survey data.
Several intriguing questions remain. For example, vari-ous contingencies may exist that moderate the relationshipbetween various partner characteristics and performancevariables. Do complementarity, cultural compatibility, andoperational compatibility always play an equal role? Arethey compensatory? Is one or two more important in cer-tain aspects of the relationship than the other(s)? Do theproposed relationships hold in all conditions and contexts?The scope of our article and the small sample size limit ourability to test these questions, and we leave it to futureresearch to unravel these conundrums.
370 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
1. Parkhe’s Type I and II diversities are related to differences be-tween partner firms. In this study, we examine differences in resources(which we call resource complementarity) and differences in cultural andoperational norms. However, since it is proposed that firms need to besimilar in terms of social norms, we conceptually operationalize Type IIdiversity as reverse, that is, compatibility (or similarity in cultural and op-erational norms).
2. Strategic assets are defined as “set of difficult to trade and imitate,scarce, appropriable, and specialized resources and capabilities that be-stow the firm competitive advantage” (Amit and Schoemaker 1993:36).
3. Details of respondent countries are available from authors on re-quest.
4. Chin’s (1998) rule of thumb suggests that the sample size for aPartial Least Squares (PLS) study be equal to the larger of the following:(1) 10 times the scale with the largest number of formative (i.e., emer-gent) indicators or (2) 10 times the largest number of structural paths di-rected at a particular construct in the structural model. These conditionsare satisfied in our study.
5. Internal consistency = ((Σλyi)2
+ Σ var(εi)), wherevar(εi) = 1 – λyi
6. Average variance extracted = Σλyi
2+ Σ var(εi) e var(εi) = 1 –
7. Indirect effects that include three variables (X1 → X2 → X3) canbe tested as follows: a and b are the path coefficients for the direct effectsof X1 → X2 and X2 → X3, respectively. SEa and SEb are the standard er-rors. The product ab represents the indirect effect of X1 on X3. The stan-dard error for the indirect effect is given as follows: SEab = sqrt [(b
2 × SEb
8. We estimated multiple PLS models, by dropping different inde-pendent variables, to investigate the possibility that the negative influ-ence of operational compatibility on alliance performance may beattributed to suppressor effects. PLS results are robust and indicate thatthe impact of operational compatibility on alliance performance is indeednegative when we control for the effects of other exogenous constructs.
9. To establish that a mediation model exists, Baron and Kenny(1986) stipulate that four conditions must hold: (1) the antecedent vari-able must be related to the mediator, (2) the antecedent variable must berelated to the dependent variable, (3) the mediator must be related to thedependent variable, and (4) the relationship between the antecedent andthe dependent variable must be less in (3) than in (2). All conditions weresatisfied. Tables 2, 3, and 4 show the results of all the steps.
10. The indirect effect of resource complementarity on a specific per-formance variable equals the sum of statistically significant indirect ef-fects through trust, commitment, and reciprocal information exchange onthat particular performance type. Similar computations were made forcultural and operational compatibility variables.
Achrol, Ravi S. 1997. “Changes in the Theory of Interorganizational Re-lations in Marketing: Toward a Network Paradigm.” Journal of theAcademy of Marketing Science 25:56-71.
Amit, Raphael and Paul J. H. Schoemaker. 1993. “Strategic Assets andOrganizational Rent.” Strategic Management Journal 14 (1): 33-46.
Anderson, Erin and Barton Weitz. 1989. “Determinants of Continuity inConventional Industrial Channels Dyads.” Marketing Science 8(Fall): 310-323.
and . 1992. “The Use of Pledges to Build and SustainCommitment in Distribution Channels.” Journal of Marketing Re-search 29:18-34.
Anderson, James C. and James A. Narus. 1990. “A Model of DistributorFirm and Manufacturer Firm Working Partnerships.” Journal of Mar-keting 54 (1): 42-58.
Armstrong, J. Scott and Terry S. Overton. 1977. “EstimatingNonresponse Bias in Mail Surveys.” Journal of Marketing Research14 (August): 396-402.
Aulakh, Preet S., Masaaki Kotabe, and Arvind Sahay. 1996. “Trust andPerformance in Cross Border Marketing Partnerships: A BehavioralApproach.” Journal of International Business Studies 27 (5): 1005-1032.
Badaracco, Joseph L. Jr. 1991. The Knowledge Link: How Firms Com-pete Through Strategic Alliances. Bostons: Harvard BusinessSchool.
Barclay, Donald and Brock Smith. 1997. The Effects of OrganizationalDifferences and Trust on the Effectiveness of Selling Partner Rela-tionships.” Journal of Marketing 61 (1): 3-21.
Barney, Jay B. 1991. “Firm Resources and Sustained Competitive Ad-vantage.” Journal of Management 17 (1): 99-120.
Baron, Reuben M. and David A. Kenny. 1986. “The Moderator-MediatorVariable Distinction in Social Psychological Research: Conceptual,Strategic and Statistical Considerations.” Journal of Personality andSocial Psychology 15 (6): 1173-1182.
Bleeke, Joel and David Ernst. 1991. “The Way to Win in Cross-BorderAlliances.” Harvard Business Review 69 (6): 127-135.
Borys, Bryan and David B. Jemison. 1989. “Hybrid Arrangements asStrategic Alliances: Theoretical Issues in Organizational Combina-tions.” Academy of Management Review 14:234-249.
Bradach, Jeffrey L. and Robert G. Eccles. 1989. “Price, Authority, andTrust: From Ideal Types to Plural Forms.” American Review of Soci-ology 15:97-118.
Bucklin, Louis P. and Sanjit Sengupta. 1993. “Organizing Successful Co-Marketing Alliances.” Journal of Marketing 57 (April): 32-46.
Chin, Wynne W. 1998. “The Partial Least Square Approach to StructuralEquation Modeling.” In Modern Methods for Business Research. Ed.G. A. Marcoulides. Mahwah, NJ: Lawrence Erlbaum, 295-336.
Chung, Seungwha, Harbir Singh, and Kyungmook Lee. 2000.“Complementarity, Status Similarity and Social Capital as Drivers ofAlliance Formation.” Strategic Management Journal 21:1-22.
Cohen, Jacob. 1988. Statistical Power Analysis for the Behavioral Sci-ences. New York: Academic Press.
Cohen, Wesley M. and David A. Levinthal. 1990. “Absorptive Capacity:A New Perspective on Learning and Innovation.” Administrative Sci-ence Quarterly 35 (September): 128-152.
Cullen, John B., Jean L. Johnson, and Tomoaki Sakano. 1995. “Japaneseand Local Partner Commitment to IJVs: Psychological Conse-quences of Outcomes and Investments in the IJV relationship.” Jour-nal of International Business Studies 26 (1): 91-116.
, , and . 2000. “Success Through Commitment andTrust: The Soft Side of Strategic Alliance Management.” Journal ofWorld Business 35 (3): 223-240.
Das, Tarun K. and Bing-Sheng Teng. 1998. “Between Trust and Control:Developing Confidence in Partner Cooperation in Alliances.” Acad-emy of Management Review 23 (3): 491-512.
Dierickx, Ingemar and Karel Cool. 1989. “Asset Stock Accumulationand Sustainability of Competitive Advantage.” Management Science35 (12): 1504-1513.
Dwyer, Robert F., Paul H. Schurr, and Sejo Oh. 1987. “DevelopingBuyer-Seller Relationships.” Journal of Marketing 51 (April): 11-27.
Dyer, Jeff H. and Harbir Singh. 1998. “The Relational View: CooperativeStrategy and Sources of Inter-Organizational Competitive Advan-tage.” The Academy of Management Review 23 (4): 660-679.
Eccles, Robert. 1981. “The Quasifirm in the Construction Industry.”Journal of Economic Behavior and Organization 2:335-357.
Evans, F. B. 1963. “Selling as a Dyadic Relationship: A New Approach.”American Behavioral Scientist 6 (May): 76-79.
Fornell, Claes and David F. Larcker. 1981. “Evaluating Structural Equa-tion Models With Unobservable Variables and Measurement Error.”Journal of Marketing Research 18 (1): 39-50.
Friedmann, Wolfgang G. and John-Pierre Beguin. 1971. Joint Interna-tional Business Ventures in Developing Countries. New York: Co-lumbia University Press.
Geringer, J. Michael and Louis Hebert. 1989. “Control and Performanceof International Joint Ventures.” Journal of International BusinessStudies 20 (2): 235-254.
Gomes-Casseres, Benjamin. 1987. “Joint Venture Instability: Is It aProblem?” Columbia Journal of World Business 22 (2): 97-102.
Sarkar et al. / ALLIANCE PERFORMANCE 371
Gulati, Ranjay, Tarun Khanna, and Nitin Nohria. 1994. “Unilateral Com-mitments and the Importance of Process in Alliances.” Sloan Man-agement Review 35 (3): 61-69.
Gundlach, Gregory T., Ravi S. Achrol, and John T. Mentzer. 1995. “TheStructure of Commitment in Exchange.” Journal of Marketing 59(January): 78-92.
Hagedoorn, John. 1993. “Understanding the Rationale of Strategic Tech-nology Partnering: Interorganizational Modes of Cooperation andSectoral Differences.” Strategic Management Journal 14 (5): 371-385.
Hamel, Gary. 1991. “Competition for Competence and InterpartnerLearning Within International Strategic Alliances.” Strategic Man-agement Journal 20 (Special Issue): 83-103.
Harrigan, Kathryn R. 1985. Strategies for Joint Ventures. Lexington,MA: Lexington Books.
Heide, Jan B. 1994. “Interorganizational Governance in MarketingChannels.” Journal of Marketing 58 (1): 71-85.
and George John. 1992. “Do Norms Matter in Marketing Rela-tionships?” Journal of Marketing 56:32-44.
and Anne S. Miner. 1992. “The Shadow of the Future: Effects ofAnticipated Interaction and Frequency of Contact on Buyer-SellerCooperation.” Academy of Management Journal 2:265-300.
Hill, Charles W. L. 1990. “Cooperation, Opportunism, and the InvisibleHand: Implications for Transaction Cost Theory.” Academy of Man-agement Review 15 (3): 500-513.
Hitt, Michael A., Tina M. Dacin, Edward Levitas, Jean-Luc Arregle, andAnca Borza. 2000. “Partner Selection in Emerging and DevelopedMarket Context: Resource-Based and Organizational Learning Pro-spective.” Academy of Management Journal 43 (3): 449-467.
Hosmer, LaRue Tone. 1995. “Trust: The Connecting Link Between Or-ganizational Theory and Philosophical Ethics.” Academy of Manage-ment Review 20 (2): 370-403.
Hulland, John. 1999. “Use of Partial Least Squares (PLS) in StrategicManagement Research: A Review of Four Recent Studies.” StrategicManagement Journal 20 (2): 195-204.
Johnson, Jean L., John B. Cullen, Tomoaki Sakano, and HideyukiTakenouchi. 1996. “Setting the Stage for Trust and Strategic Integra-tion in Japanese-U.S. Cooperative Alliances.” Journal of Interna-tional Business Studies 27 (5): 981-1004.
Katz, Michael L. 1989. “Vertical Contractual Relations.” In Handbook ofIndustrial Organizations, Vol 1. Eds. R. Schmalensee and R. D.Willig. Reading, NY: Elsevier, 655-721.
Khanna, Tarun. 1998. “The Scope of Alliances.” Organization Science 9(3): 340-355.
, Ranjay Gulati, and Nitin Nohria. 1998. “The Dynamics ofLearning Alliances: Competition, Cooperation, and Relative Scope.”Strategic Management Journal 19 (3): 193-210.
Killing, J. Peter. 1983. Strategies for Joint Venture Success. London:Croom Helm.
Koza, Mitchell P. and Arie Y. Lewin. 1998. “The Co-Evolution of Strate-gic Alliances.” Organization Science 9 (3): 255-264.
Kumar, Nirmalya, Louis W. Stern, and James G. Anderson. 1993. “Con-ducting Interorganizational Research Using Key Informants.” Acad-emy of Management Journal 36 (6): 1633-1651.
Lado, Augustine A., Nancy G. Boyd, and Susan C. Hanlon. 1997. “Com-petition, Cooperation, and the Search for Economic Rents: A Syn-cretic Model.” Academy of Management Review 22 (1): 110-141.
Lohmöller, Jan-Bernd. 1989. Latent Variable Path Modeling With PartialLeast Squares. New York: Springer-Verlag.
Macneil, Ian R. 1980. The New Social Contract. New Haven, CT: YaleUniversity Press.
Madhok, Anoop. 1995. “Revisiting Multinational Firms’ Tolerance forJoint Ventures: A Trust-Based Approach.” Journal of InternationalBusiness Studies 26 (1): 117-138.
and Stephen B. Tallman. 1998. “Resources, Transactions, andRents: Managing Value Through Interfirm Collaborative Relation-ships.” Organization Science 9 (3): 326-339.
Mohr, Jakki and John R. Nevin. 1990. “Communication Strategies inMarketing Channels: A Theoretical Perspective.” Journal of Mar-keting 50 (October): 36-51.
and Robert Spekman. 1994. “Characteristics of Partnership Suc-cess: Partnership Attributes, Communication Behavior, and ConflictResolution Techniques.” Strategic Management Journal 15 (2): 135-152.
, Robert J. Fisher, and John R. Nevin. 1996. “Collaborative Com-munication in Interfirm Relationships: Moderating Effects of Inte-gration and Control.” Journal of Marketing 60 (July): 103-115.
Moorman, Christine, Gerald Zaltman, and Rohit Deshpande. 1992. “Re-lationships Between Providers and Users of Market Research: TheDynamics of Trust Within and Between Organizations.” Journal ofMarketing Research 29 (3): 314-328.
Morgan, Robert M. and Shelby D. Hunt. 1994. “The Commitment-TrustTheory of Relationship Marketing.” Journal of Marketing 58 (3): 20-38.
Oliver, Christine. 1990. “Determinants of Interorganizational Relation-ships: Integration and Future Directions.” Academy of ManagementReview 15:241-265.
Parkhe, Arvind 1991. “Interfirm Diversity, Organizational Learning, andLongevity in Global Strategic Alliances.” Journal of InternationalBusiness Studies 20:579-601.
. 1993. “Strategic Alliance Structuring: A Game Theoretic andTransaction Cost Examination of Interfirm Cooperation.” Academyof Management Journal 36 (4): 794-829.
Ring, Peter Smith. 1996. “Fragile and Resilient Trust and Their Roles inEconomic Exchange.” Business and Society 35 (2): 148-175.
and Andrew Van de Ven. 1994. “Developmental Processes of Co-operative Interorganizational Relationships.” Academy of Manage-ment Review 19 (1): 90-118.
Sarkar, Mitrabarun, S. Tamer Cavusgil, and Cuneyt Evirgen. 1997. “TheCommitment-Trust Framework in International Collaborative Ven-tures.” In Cooperative Strategies: North American Perspectives. Eds.Paul W. Beamish and J. Peter Killing. San Francisco: The NewLexington Press, 255-285.
, Raj Echambadi, and Jeffrey Harrison. 2001. “Alliance Entrepre-neurship and Firm Market Performance.” Strategic ManagementJournal 22 (6/7): 701-711.
Saxton, Todd. 1997. “The Effects of Partner and the Relationship Charac-teristics on Alliance Outcomes.” Academy of Management Journal40:443-461.
Slaate, H. A. 1968. The Pertinence of the Paradox. New York: Human-ities Press.
Smith, J. Brook, and Donald W. Barclay. 1997. “The Effect of Organiza-tional Differences and Trust on the Effectiveness of Selling PartnerRelationship.” Journal of Marketing 61 (1): 3-21.
Stump, Rodney, and Jan B. Heide. 1996. “Controlling Supplier Oppor-tunism in Industrial Relationships.” Journal of Marketing Research33 (November): 431-441.
Wernerfelt, Birger. 1984. “A Resource-Based View of the Firm.” Strate-gic Management Journal 5 (2): 171-180.
Wilkof, M. V., D. W. Brown, and J. W. Selsky. 1995. “When Stories AreDifferent: The Influence of Corporate Culture Mismatches inInterorganizational Relations.” Journal of Applied Behavioral Sci-ence 30:373-388.
Williamson, Oliver E. 1985. The Economic Institutions of Capitalism.New York: Free Press.
Wilson, David T. 1995. “An Integrated Model of Buyer-Seller Relation-ships.” Journal of the Academy of Marketing Science 23 (4): 335-345.
Yates, J. K., S. Mukherjee, and S. Njos. 1999. Anatomy of ConstructionIndustry Competition in the Year 2000. Austin, TX: Construction In-dustry Institute.
ABOUT THE AUTHORS
MB Sarkar (Ph.D., Michigan State University) is an assistantprofessor of marketing at the University of Central Florida. Hiscurrent research includes strategic alliances, innovation and en-trepreneurship, knowledge management, and electronic markets.
372 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
His research has been published in the Strategic ManagementJournal, the Journal of International Business Studies, and theJournal of Business Research, among others.
Raj Echambadi (Ph.D., University of Houston) is an assistantprofessor of marketing at the University of Central Florida. Hiscurrent research interests include investigation of territorial loy-alty issues, management of innovations, and estimation issuespertaining to structural equation modeling and Partial LeastSquares. His research has been published in the Strategic Man-agement Journal, Multivariate Behavioral Research, and theJournal of Product Innovation Management.
S. Tamer Cavusgil (Ph.D., University of Wisconsin) is Univer-sity Distinguished Faculty and serves as the John WilliamByington Endowed Chair in global marketing at Michigan StateUniversity (MSU). He is also the executive director of MSU’sCenter for International Business Education and Research, a na-
tional resource center. His teaching, research, and administrativeactivities have focused on international business and marketing.His research has been published in the Journal of Marketing, theJournal of Marketing Research, and the Journal of InternationalBusiness Studies, among others. His specific interests include theinternationalization of the firm, global marketing strategy, andinternationalization of business education. He was the foundingeditor of the Journal of International Marketing, now publishedby the American Marketing Association.
Preet S. Aulakh (Ph.D., University of Texas at Austin) is an as-sociate professor of strategy and international business at the FoxSchool of Business and Management, Temple University. His re-search focuses on international technology licensing, cross-bor-der joint ventures and strategic alliances, and strategies of firmsfrom developing economies. His research has been published inthe Academy of Management Journal, the Journal of Marketing,and the Journal of International Business Studies.
Sarkar et al. / ALLIANCE PERFORMANCE 373
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001Keaveney, Parthasarathy / ONLINE SERVICES
Customer Switching Behavior inOnline Services: An Exploratory Studyof the Role of Selected Attitudinal,Behavioral, and Demographic Factors
Susan M. KeaveneyMadhavan ParthasarathyUniversity of Colorado at Denver
With a quarter of a billion Internet users worldwide andestimates of more than one-half billion people online bythe year 2003, growth in the online services industry hasbeen exponential. With this growth has come concernabout customer “churn,” a concern that parallels issues ofcustomer switching behavior in services industries in gen-eral. This manuscript reports results of two field studies,conducted among two randomly selected samples of on-line service users, that investigate the degree to which se-lected behavioral (information that customers used whenmaking the online service decision, their service usage),attitudinal (risk-taking propensity), and demographic (in-come and education) factors are effective in discriminat-ing between continuers and switchers. The research inStudy 1 is replicated in Study 2 and extended to consideradditional attitudinal factors of satisfaction and involve-ment. Implications for managers and researchers arediscussed.
During the past decade, the online service industry haswitnessed tremendous growth, much of it spurred by theInternet revolution (cf. Cyberatlas.com 2000e; Datamonitor.com 1999). Online service providers such as AmericaOnline (AOL), CompuServe, Prodigy, Delphi, andMicrosoft Network (MSN), which offer content (e.g.,news, weather, sports, health care, entertainment), features(e.g., software downloads, financial research data), e-
commerce access (e.g., retail merchandise, travel), andinteractive services (e.g., e-mail, bulletin boards, chatrooms), have grown at phenomenal rates—some report-edly growing at rates of 200 percent per year (AOL.com2000; Prodigy.com 2000). Rapid growth has attracted theinterest of the major Internet service providers (ISPs),such as IBM, AT&T’s Worldnet, WorldCom’s UUnet, orPSINet, whose original focus was dial-up Internet access.Industry sources observe that the major ISPs now competefor markets claimed by the traditional online service pro-viders, with value-added and premium services (Borland1998).
A key issue for online service providers as a result ofthis increased competition is “churn,” or customer move-ment in and out of the marketplace. Some churn is onlineservice discontinuance, where individuals try a service(e.g., AOL) but subsequently decide to stop using the ser-vice category (e.g., online services) altogether. For exam-ple, one study reported that 9 million people in the UnitedStates had tried using the Internet during 1997 but had dis-continued use by 1998 (Kingsley and Anderson 1998).Some of the churn is “customer service switching behav-ior” (Keaveney 1995), where customers continue to usethe service category (e.g., online services) but switch fromone service provider (e.g., Prodigy) to another (e.g., AOL).For example, a 1995 report stated that only 2.5 of 6.7 mil-lion online users were still with their original online ser-vice provider (Carl 1995), and a 1997 report stated thatonline service providers had lost 3 percent to 5 percent oftheir customers in a 90-day window (Levin 1997). Morerecent reports state that 14 percent of Internet users indi-cated a “strong intent to switch ISPs in the next 12 months,”
and an additional 22 percent said that they might considerswitching providers at some point (Cyberatlas.com2000c). While industry reports of numbers of Internetusers, customer migration, and churn can be rather fluid inthis fast-changing market, it is clear that churn is an impor-tant issue for online service providers.
Concern about churn in the world of electronic com-merce parallels a general and widespread concern withcustomer retention and customer switching behavior inservices industries. Identification of “defection-prone”customers is a “pivotal requirement for companies and aripe area for research” (Zeithaml 2000:77). The purpose ofthis research is to begin to identify defection-prone cus-tomers of online services. Study 1 explores the effective-ness of selected attitudinal, behavioral, and demographicvariables in discriminating between switchers andcontinuers of online services. Study 2 replicates Study 1and extends the research to include measures of customersatisfaction and involvement.
BACKGROUND: SERVICE SWITCHING
Service Switching and the Firm
The benefits of customer retention and the costs of cus-tomer switching behavior have received much attention inthe literature. A decrease in customer switching createsbenefits on both sides of the income statement in the formof higher revenues and lower costs and has been shown tobe effective from both defensive and offensive strategicmarketing perspectives (Fornell and Wernerfelt 1987;Zeithaml 2000). Consider first the defensive strategic per-spective: on the revenue side, studies have shown that con-tinuing customers purchase higher volumes at higher mar-gins (Grant and Schlesinger 1995; Heskett, Sasser, andSchlesinger 1997; Reichheld 1993, 1996; Reichheld andKenny 1990; Reichheld and Sasser 1990) and increasetheir usage of a service even when prices increase (Boltonand Lemon 1999). Firms whose customers report high lev-els of satisfaction also enjoy “insulating” effects on theirreputations (Anderson and Sullivan 1993). On the costside, studies have shown that selling costs of serving con-tinuing customers are lower and operating efficiencies arehigher (Heskett et al. 1997; Reichheld and Sasser 1990).
From an offensive strategic perspective, on the revenueside, retained customers attract new customers throughpositive word of mouth, thereby increasing market share(Heskett et al. 1997; Reichheld and Sasser 1990). Con-tinuing customers may also help the firm to sustain pre-mium pricing tactics and validate the firm’s good reputa-tion to new customers (Zeithaml 2000). On the cost side,the positive word of mouth associated with increased cus-tomer retention can lower marketing costs to acquire newcustomers (cf. Bolton and Bronkhorst 1995; Peters 1988).
Customer switching behavior can be particularly dam-aging for subscription- or membership-based servicefirms where customers commit to ongoing relationshipsand services are continuously provided. For example,insurance, banking, public utilities, health care, telecom-munications, cable, and other subscription-based servicesdepend on customers to pay all or part of their charges on afixed-fee continuous basis. Service firms in industriessuch as banking, health care, and telecommunicationscharge fixed fees for access and basic services, with addi-tional charges for increased usage. Others, like public util-ities, may estimate overall costs for high and low periodsof usage and charge customers prorated but relatively sta-ble periodic fees, with increased (decreased) billing forhigher-than-estimated (lower-than-estimated) averageuse. Still others, like cable companies, health clubs, andonline services, charge a single periodic fee for accesswith unlimited usage. For each of these service firms thatallocate fixed costs across large numbers of customers anddepend on receipt of fixed membership or access fees on acontinuous basis, customer switching behavior can have aparticularly devastating effect on the bottom line.
Service Switching and the Customer
Because of the arguments presented above, the identifi-cation of those customers prone to service switchingbehavior is a high priority. Prior work has examined rea-sons for customer service switching (Bolton andBronkhurst 1995; Keaveney 1995), differences in satisfac-tion and loyalty of service switchers versus stayers(Ganesh, Arnold, and Reynolds 2000), cognitive modelsof service switching (Bansal and Taylor 1999), and pro-cess models of customer service switching (Roos 1999).Few studies to date have identified which, if any, customercharacteristics might be effective in predicting customerswitching behavior.
A substantial body of research has supported the rela-tionships of two attitudinal variables, satisfaction and ser-vice quality, to customers’ switching intentions (e.g.,Anderson 1994; Anderson and Sullivan 1993; Boulding,Kalra, Staelin, and Zeithaml 1993; Cronin and Taylor1992; Oliva, Oliver, and Bearden 1995; Reichheld andSasser 1990; Zeithaml, Berry, and Parasuraman 1996).Fewer studies have examined these relationships in thecontext of actual switching behavior, but they too suggestthat dissatisfaction explains at least some customerswitching (Bansal and Taylor 1999; Bolton andBronkhorst 1995; Keaveney 1995).
Other research has supported the need to considercauses of service switching beyond dissatisfaction. In anexploratory study of customer switching behavior in ser-vices industries, Keaveney (1995) identified eight majorcauses. Some could be associated with feelings of dissatis-faction with the service (e.g., core service failures, failed
Keaveney, Parthasarathy / ONLINE SERVICES 375
service encounters, poor service recoveries), but otherswere extrinsic or situational factors (e.g., price, inconve-nience, ethics, competition, and involuntary situations).Anderson (1996) reported the importance of price toler-ance in addition to satisfaction when predicting customerswitching. Dabholkar and Walls (1999) found that out-come-related, extrinsic, and service process factors wereall important in predicting customers’ intentions to switchproviders of an experiential service. Bolton andBronkhurst (1995) found that customers of cellular ser-vices who had complained to the firm were more likely toswitch than customers who had not complained. Thus, itseems important to augment studies of satisfaction andintentions—which, although quite important, have beenwell studied—to include other attitudinal, behavioral, anddemographic variables.
In light of the preceding discussion, the researchreported in this article contributes to the services market-ing literature in a number of ways: (1) this research beginsthe important work of identifying and profiling defection-prone customers; (2) this research examines the actualswitching behavior of online service customers, not justtheir self-reported repurchase intentions; (3) this studygoes beyond satisfaction and intentions to explore hereto-fore unexamined attitudinal, behavioral, and demographiccharacteristics of customers as possible predictors ofswitching behavior; and (4), this research focuses on cus-tomers of online services—a rapidly growing industry,and one seemingly plagued by churn.
CONCEPTUAL DEVELOPMENTAND RESEARCH HYPOTHESES
A basic tenet of consumer behavior theory holds thatwhen consumers make purchase decisions, they use differ-ent types of information sources to help them identify pos-sible alternatives, to evaluate those alternatives, and tomake choices. Information sources can be characterized ina variety of ways—internal versus external, personal ver-sus impersonal, marketer-oriented versus third-party, andso forth. Research on consumer information search tells usthat consumers differ in their preferences for types ofinformation (cf. Furse, Punj, and Stewart 1984). Prefer-ences for different types of information sources have alsobeen found to vary between goods versus services deci-sions (Murray 1991).
The consumer satisfaction/dissatisfaction literature(CS/D) and the service quality literature both state that theinformation used by buyers in their choice decisions has adirect effect on the formation of predictive expectations ofperceived performance (Boulding et al. 1993; Oliver 1997;
Zeithaml, Berry, and Parasuraman 1993). Predictiveexpectations affect consumers’ judgments of satisfactionand service quality, which in turn affect key behavioraloutcomes such as repurchase, switching behavior, and loy-alty (Oliver 1997; Yi 1990; Zeithaml et al. 1996).
Thus, the types of information that customers use whenmaking choice decisions will, directly or indirectly, affecttheir subsequent assessments of the choice. Differenttypes of information sources, when each is typically used,and its anticipated effects on predictive expectations, satis-faction, and switching behavior are discussed in the nextsections.
External sources of information. External informationsearch is a purposeful decision by the consumer to scan theenvironment for new and decision-relevant information(Berning and Jacoby 1974; Moore and Lehman 1980).Consumers may engage in external information search byreading about products or services in mass media, usingmarketer-produced print information, and researchingthird-party reviews of products and services in articles andbooks. (Note that these sources are also characterized asimpersonal, as opposed to personal, information sources.)
Some consumers are more likely than others to rely onexternal, impersonal sources of information when makingpurchase decisions. In a study of purchasers of new auto-mobiles, Furse et al. (1984) identified two segments ofconsumers, representing 44 percent of respondents, thatwere characterized by their above-average use of external,impersonal sources such as brochures, pamphlets, adver-tisements, magazines, reviews, and ratings. Moreover,some purchase situations are more (or less) likely to trig-ger the desire for external, impersonal sources of informa-tion. In a study of durable goods, consumers preferredadvertising and other external sources when they believedthey were capable of drawing their own conclusions aboutproduct attributes and judging the merits of the productthemselves (Houston 1979). Consumers also prefer exter-nal, impersonal sources more when choosing goods, butless when choosing services (Murray 1991).
Consumers who seek external, impersonal sources areinterested in gaining factual, objective information aboutproduct or service attributes. A preponderance of factualinformation about attributes should lead to more accuratepredictive expectations about the future performance ofthe product or service (Boulding et al. 1993). Accuratepredictive expectations mean an absence of discon-firmation, as is demonstrated by the use of “accurate: itwas just as I thought” as the central or neutral value inscales measuring positive and/or negative disconfirmation(Oliver 1997:103). Absence of disconfirmation, combinedwith a positive valence of the predictive expectation for thechosen good or service, means that the positive predictiveexpectation is confirmed by perceived performance,
376 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
resulting in satisfaction. Barring any unpleasant surprisesin performance, continued use by consumers shouldresult.
In the case of a thoughtful purchase decision, asdescribed above, contradictory information from actualexperience with a product or service may be cognitivelyresisted and confirming evidence sought. “Deviation froman attitude that is supported by a complex cognitive struc-ture is said to involve a high psychological cost to the indi-vidual” (Pritchard, Havitz, and Howard 1999:335).Rethinking information in a complex informationalschema is effortful, disruptive, and dissonant; consumersbecome committed to the decision they worked so hard toreach (Pritchard et al. 1999). Strong expectations can evenact to edit or filter perceived performance in such a waythat consumers see what they expect to see and overlookdiscrepant information (Hoch and Deighton 1989). Again,barring any clear and unambiguous unpleasant surprises inperformance, consumers will be motivated to continuetheir satisfied use of the product or service (Oliva et al.1995.)
Finally, evidence from the advertising literature pro-vides yet further support for this line of reasoning. Whenconsumers learn about a product and make product deci-sions under the influence of advertising, the advertising in-formation may be said to prime or “transform the productexperience” (Hoch and Deighton 1989). Unless perceivedproduct performance is unambiguously negative, prior ex-pectations based on advertising information predominateand perceptions of performance are assimilated (Deighton1984). In summary, on the basis of theory and research inthe satisfaction, information-processing, experiential-learning, and advertising literatures, we propose the fol-lowing hypothesis:
Hypothesis 1: Online service continuers were influencedby external sources of information when makingtheir subscription decisions more than online ser-vice switchers.
Interpersonal sources of information. Other customersrely on word-of-mouth opinions of others more than ontheir own decision-making processes when making prod-uct or service decisions. Consumers who have little priorexperience with a product or service, who find the deci-sion-making process difficult, or who have little confi-dence in their own abilities to judge the product or servicemay prefer to ask advice from others perceived as knowl-edgeable (Furse et al. 1984). Consumers also tend to usemore, and to have more confidence in, interpersonalsources of information when purchasing services thanwhen purchasing goods (Murray 1991). This is thought tobe because of the experiential nature of services, whichrenders services difficult to evaluate prior to purchase anduse. While information from other individuals is both sub-
jective and evaluative, it does provide a semblance ofvicarious experience.
With a heuristic mode of processing, such as relying onthe opinions or judgments of others, decisions are based ona more superficial assessment of information (Furse et al.1984). Without the well-informed expectations that resultfrom independent decision processing, consumers are vul-nerable to negative disconfirmation if services do not per-form for them in the way that their friends have described(Oliver 1997). Moreover, we know that service consump-tion can be highly individualized and difficult to compre-hend until personally experienced. Thus, consumers whorely on vicarious experience to make service choices mayrisk disappointment and dissatisfaction if their personalexperiences differ from others’ experiences. When expec-tations are violated by contrary experience, motivation tocorrect the problem—by switching services or discontinu-ing use—is high (Hoch and Deighton 1989).
Hypothesis 2: Online service continuers were influencedby interpersonal sources of information when mak-ing their subscription decisions less than online ser-vice switchers.
Experiential sources of information. When faced with apurchase decision, consumers first engage in internalsearch, examining information in memory about past ex-periences and product-relevant knowledge (Bettman1979). The more perceived risk associated with the pur-chase situation, as in the case of services, the more con-sumers prefer their own observations and experiences assources of information (Murray 1991).
Hoch and Deighton (1989) observed that “consumerstend to grant special status to conclusions drawn fromexperience.” They theorized that motivation and involve-ment in the decision tend to be higher, consumers havecontrol over the pace and content of the learning, andsource credibility is high because the interests of sourceand consumer are one. Research suggests that informationfrom experience is likely to have a greater influence onbehavior than information from other sources (Smith andSwinyard 1982).
Having already used a service, or having service-rele-vant knowledge in memory, means that the consumer notonly has personal knowledge about service attributes butalso has personal experience about how this service worksfor him or her and satisfies his or her needs. As a respon-dent in one major study of customer expectations of ser-vice was quoted as saying, “My expectations are definitelyinfluenced by my past experience . . . my expectations aremore realistic because of the knowledge I’ve gained”(Zeithaml et al. 1993:10). Thus, predictive expectationsbased on experience are quite accurate and probabilities ofdisconfirmation low, as long as subsequent experiencesare not surprisingly different.
Keaveney, Parthasarathy / ONLINE SERVICES 377
Consumers learn from their experiences with productsand services and update what they already know(Boulding et al. 1993). Consumers who already have priorexperience with a product or service, or who have product-related knowledge, learn quickly from experience (Hochand Deighton 1989). If what experience has to teach is un-ambiguously positive or negative, experienced consumerswill quickly adjust their subsequent evaluations in thesame direction. If, however, continued experience is onlyslightly different from past experience, consumers will ei-ther ignore or assimilate the new information. In this case,there is little to learn from experience and subsequent eval-uations will not change. Thus, barring clear and unex-pected negative surprises in performance, we expect thefollowing relationship to hold:
Hypothesis 3: Online service continuers were influencedby experiential sources of information when makingtheir subscription decisions more than online ser-vice switchers.
No studies to date have directly investigated the rela-tionship between levels of service usage and switchingbehavior. However, a number of related studies lead us tohypothesize that heavier users of the service will be lesslikely to switch. First, the disconfirmation paradigm pre-dicts that frequent usage should provide customers withrelatively accurate and realistic performance expectations,thereby decreasing disconfirmation and increasing satis-faction and repurchase intentions (Anderson and Sullivan1993). Results from the Swedish Customer SatisfactionBarometer (SCSB) project indicate that higher levels ofcustomer usage are associated with lower incidences ofdisconfirmation, somewhat higher levels of satisfaction,and higher repurchase intentions (Anderson 1994).
From an overall satisfaction perspective, when a cus-tomer uses a service frequently or intensely, he or shedevelops a fairly strong and—as inferred by continueduse—presumably positive attitude about it. A consumerwith multiple positive service experiences will be increas-ingly insensitive to discrete instances of service failure andmay endure a number of problems before positive satisfac-tion is revised downward (Oliva et al. 1995; Oliver 1997).In a study of frequently purchased consumer goods,LaBarbera and Mazursky (1983) found that the longer thesequence of repeat purchase behavior, the more that expe-rience with the brand accounted for repurchase behaviorand the lower the role of satisfaction. Bolton and Lemon(1999) found evidence of a dynamic relationship betweencustomers’ prior usage, cumulative satisfaction evalua-tions, and subsequent usage: Customers with positiveevaluations of their prior usage experience higher cumula-
tive satisfaction; higher overall satisfaction in turn waslinked to increased subsequent usage.
A third explanation is offered by theories of consumerlearning. Because consumers act as though the cost oflearning is greater than the cost of making a mistake (Hochand Deighton 1989), heavy users of a particular online ser-vice may feel that they have developed nontransferable,provider-specific skills. Having invested time and energyto become facile at one online service, consumers may beunwilling to learn how to use an alternative product (Albaand Hutchinson 1987; Hoch and Deighton 1989). Basedon the above discussion,
Hypothesis 4a: Online service continuers use the onlineservice more frequently than online serviceswitchers.
Hypothesis 4b: Online service continuers use the onlineservice more intensively (i.e., on average, their on-line sessions continue for longer periods of time)than online service switchers.
Hypothesis 4c: Online service continuers exhibit higheroverall service usage than online service switchers.
Propensity for Risk-Taking Behavior
We are aware of no studies that examine the direct rela-tionship between propensity for risk-taking behavior andactual service-switching behavior. However, there are anumber of reasons why we expect the two to be related.Both propensity for risk-taking behavior and brand-switching behavior have been empirically shown to be cor-related with a third, hypothetically upstream variableknown as “optimum stimulation level” (Raju 1980). The-ories of optimum stimulation level (OSL) suggest thatindividuals prefer a certain level of stimulation, whichmay be termed their “optimum stimulation level.” Whenan individual’s stimulation level is below optimum, he orshe will take action to raise it; when it is above optimum,he or she will act to reduce it. One way in which consumersraise or lower their optimum stimulation levels is throughtheir behavior in the marketplace. Thus, individuals withhigher OSLs reported that they had both higher risk-takingand brand-switching tendencies than individuals withlower OSLs (Raju 1980). The question becomes, if theoptimum stimulation level is positively associated with apropensity for risk-taking behavior, and the optimum stim-ulation level is positively associated with brand-switchingbehavior, is a propensity for risk-taking behavior associ-ated with brand- (or service-) switching behavior?
Both theory and evidence in the services marketing lit-erature suggest that customers perceive the selection andpurchase of services to be riskier than goods (Guseman1981; Murray 1991; Murray and Schlacter 1990). It wouldbe reasonable to expect that individuals with higher risk-taking propensities might be more likely than their lower-
378 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
risk-taking counterparts to select and purchase these “risk-ier” services (particularly if their stimulation levels were tofall below optimum levels). Let us suppose, for the sake ofargument, that the consumer views switching from oneservice provider to another as analogous to selecting andpurchasing a new service—one must engage in a similarprocess of selecting and choosing the new service pro-vider. Like selecting and choosing services, service-switching behavior can be seen as an inherently risky ac-tivity. By extension, then, we might expect that individualswith higher propensities for risk taking would be morewilling to engage in the activity of switching among ser-vice providers should the need arise. Thus, we propose thefollowing:
Hypothesis 5: Online service continuers have a lowerpropensity for risk-taking behavior than online ser-vice switchers.
Customers with higher incomes and education levelsmay be capable of developing sophisticated and probablyaccurate estimates of what to expect from a service. Forexample, customers with higher incomes may have experi-ence with more frequent usage of services or usage of agreater variety of services. In the online world, higher-income users tend to be more experienced “surfers”(Cyberatlas.com 2000b). Higher incomes also allow cus-tomers to afford to choose “buy” in potential “make-ver-sus-buy” service decisions, potentially giving them moreexperience with a wider variety of services.
Higher education levels may also provide consumerswith greater skills in forming hypotheses about future ser-vice performance. Because services are intangible, theycan require somewhat more imagination, vision, orhypothesizing to predict what actually using the servicewould be like. This is particularly true of services havingexperiential or credence properties where, even after expe-riencing the service, many consumers find evaluation ofservice performance to be complex and difficult (Murray1991; Nelson 1970).
In contrast, lower income and less-educated service us-ers may find that their expectations were ambiguous, theirability to learn from experience limited (Hoch andDeighton 1989), their assessments uncertain, and theirevaluations of the service more vulnerable to instances ofdissatisfaction. For example, the Internet’s older, less-edu-cated, and lower income users tend to lack confidence intheir computer skills (Cyberatlas.com 2000a). Older, less-educated, and lower income computer users reported thatthey felt doubtful about their computer proficiency signifi-cantly more often than younger, more highly educated,higher income users. Thus, less-educated, lower incomecustomers may not be as skilled at forecasting the use of
online services or at evaluating online service perfor-mance.
Hypothesis 6: Online service continuers have higher av-erage income levels than online service switchers.
Hypothesis 7: Online service continuers have higher av-erage education levels than online service switchers.
Sampling frame. To test these hypotheses, data werecollected from present and past subscribers of online ser-vices (e.g., America Online, Delphi, Prodigy, MSN, andCompuServe). A major national online service providerprovided a list of all customers who had ever subscribed tothe service since the company’s inception. The enormityof this list necessitated reducing the sampling frame. Acomputer program generated a random number between 1and 26, with the random number used to indicate a letter ofthe alphabet. The number 13, corresponding to the letter Mwas selected. The company then generated a list of all cur-rent and former customers whose user names began withthe letter M. Thus, some people on the list were still sub-scribers, some had switched services, some had switchedservices a number of times, and some no longer subscribedto any online service.
As the list consisted of more than 200,000 individuals,it was further reduced in size by a computer program writ-ten to randomly select 40,000 records. After cleaning thedata by eliminating duplicate user records and recordswith missing data, a base list of 28,217 records wasobtained. Each record contained the user name, date of ini-tial subscription with the service, date of termination of theservice (if any), name, and address.
Sample size and selection. Sample size was estimatedusing both power analysis and consideration of recom-mended procedures for discriminant function analysis.Power analysis at an alpha level of .01 and a low-to-moder-ate anticipated effect size of .35 requires a sample size of200 respondents; a minimum alpha level of .10, holdingeffect size constant, requires only 100 respondents,whereas a midrange alpha of .05 requires 130 respondents(Cohen and Cohen 1980). Discriminant function analysisis sensitive to sample size as well; some experts suggestthat 20 observations for each predictor variable are ideal(Hair, Anderson, Tatham, and Black 1995; Tabachnickand Fidell 1983). Since our study has eight independentvariables plus one interaction term, this suggested a sam-ple of about 180 observations. Taken together, a samplesize of around 200 respondents was deemed desirable.Based on a reasonably conservative estimate of a 20 per-
Keaveney, Parthasarathy / ONLINE SERVICES 379
cent response rate for a consumer mail survey, 1,000 indi-viduals were sampled.
Records in the base list were sorted on the basis of pres-ence or absence of a termination date to ensure that a mixof continuers and switchers would be sampled. Because arelatively equal distribution between groups was desiredfor the most robust estimation of the discriminant function(Tabachnick and Fidell 1983), 500 names were selectedrandomly from each group (with and without terminationdates).
In the first wave of the data collection process, the ini-tial sample of 1,000 was sent a postcard by mail and askedwhether they would be willing to participate in the study.They were also asked (1) if they currently subscribed to apaid computer online service; (2) whether they had ever, atany time in the past, subscribed to an online service formore than 3 months that they later dropped; and (3) theirnames and addresses. The 3-month adoption period wasspecified in an effort to eliminate spurious switchingamong trial offers and focus on service switching after areasonable adoption-and-use period. One hundred forty-three postcards were returned by the post office asundeliverable.
Of the 857 remaining postcards, 443 were completedand returned. Of the 443 respondents, 125 reported thatthey were currently subscribed to an online service andhad not stopped using any online service provider to whichthey had subscribed for more than 3 months (“yes” to item1 and “no” to item 2). This group, designated the“continuers,” was composed primarily of people who hadeither (1) been continuously subscribed with the list pro-vider or (2) tried the list provider but continuously sub-scribed with a different provider. Two hundred and eigh-teen respondents reported that they had stopped using anonline service provider after 3 months at least once (“yes”to item 2). This group, designated the “quitters” group,was composed primarily of people whose most recentonline service switch was either (1) from the list providerto a new online service provider, (2) between other onlineservice providers, or (3) from another service provider tothe list provider.
Questionnaires were sent by mail to the 443 respon-dents who had agreed to participate in the study. The ques-tionnaires differed somewhat, depending on the aboveclassifications: the “continuer” questionnaire asked abouttheir current continuous subscription, while the “quitter”questionnaire focused on the online service provider thatthey had most recently stopped using. Two reminder post-card mailings were sent at 15-day intervals. At the end ofthe 45-day survey period, a total of 205 questionnaireswere returned, for a 46 percent response rate among the
study’s participants or a 20.5 percent response rate relativeto the initial sample. Of the 205 respondents, 72 returned“continuer” and 133 returned “quitter” surveys.
Following Armstrong and Overton’s (1977) method forchecking possible nonresponse bias, early respondents(defined as those who responded within the first 2 weeks ofthe mailing date) were compared with later respondents oneach of the measured variables. No significant differencesin means were found between the two groups, at p < .05, onany measured variables.
Measures of Variables
Relative sources of informational influence. Sources ofinfluence were conceptualized as the relative influence ofdifferent sources of information on the consumers’ sub-scription decisions. The concept was operationalized us-ing three categories of consumer information adoptedfrom Kotler (1997). External sources of information weredefined as heterophilous sources, or sources from outsidethe customer’s social system, including mass media, ad-vertising, articles, brochures, pamphlets, reviews, andother impersonal sources. Thus, in this study, externalsources included both commercial sources and publicsources of information. Interpersonal sources of informa-tion were defined as homophilous sources, or interper-sonal communication sources within the social system,also known as word-of-mouth communications. Experien-tial sources of information were defined as personal expe-rience with the service or similar services, and generalservice-relevant knowledge.
As Kotler (1997) notes, “Of key interest to the marketerare the major information sources to which the consumerwill turn and the relative influence each will have on thesubsequent purchase decision [italics added]” (p. 193).Thus, respondents were asked to distribute 100 pointsamong the three types of information sources according tohow much each had influenced the online subscriptiondecision. Items for all variables are reported in theappendix.
Usage level. The consumer’s overall usage of onlineservices was conceptualized as a function of two issues:frequency and intensity (Ram and Jung 1990). Frequencyof use was defined as the number of times that customersengaged the online service, operationalized as the numberof times they logged on during a specified time period. In-tensity was defined as the depth of involvement in the on-line activity, operationalized as the amount of time inhours they interacted with the service each time theylogged on.
Following Ram and Jung (1990), the product of the twoitems was used to compute overall service usage. Forexample, the consumer who used the service every day but
380 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
for less than 15 minutes per occasion (the type of personwho only checks e-mail) would have lower overall serviceusage than the consumer who used the service only once aweek but spends more than 2 hours per occasion.
Propensity for risk-taking behavior. Propensity forrisk-taking behavior is defined as the propensity to enjoy,take advantage of, or otherwise seek new and potentiallyrisky activities and experiences. The eight-item scale de-veloped by Raju (1980) was used. Reliability figures be-tween .80 and .83 were reported in the original study;reliability of the scale in the present study was .91.
Income and education. In addition to the above, incomeand education were measured as categorical variables.Nine categories, ranging from less than $20,000 to morethan $90,000, were used to measure income, while eightcategories from grade school to advanced degree wereused to measure education.
DATA ANALYSIS AND RESULTS
Categorization of Groups:Switchers Versus Continuers
Further classification of the “quitter” respondents as“service discontinuers” or “service switchers” was con-ducted based on more detailed information provided byseveral items in the questionnaire. As discussed earlier,service discontinuers are those former customers whohave stopped using a service altogether—in this case, theyhad stopped using online services. Service switchers arethose customers who stopped using the services of one ser-vice provider and switched to become the customers of adifferent service provider (Keaveney 1995). In the “quit-ter” questionnaire, respondents were asked to name theonline service they had most recently stopped using and toprovide the start and stop dates (month and year) of thesubscription. They were also asked to name the online ser-vice provider to which they currently subscribed and thestart date (month and year) of the subscription. Based onthis analysis, 82 respondents had discontinued use of anonline service and were currently subscribed to a differentonline service provider. This group was classified as ser-vice switchers. The remaining 51 respondents had discon-tinued use of an online service provider but had not sub-scribed with a new service. These service discontinuerswere not used in the present study because this group couldnot be said to have switched from one service provider to adifferent service provider. In summary, 69 usablecontinuer surveys and 82 switcher surveys were used in thediscriminant analysis.
Descriptive statistics for the sample are reported inTable 1. The sample is somewhat higher in age, income,
and education and is more predominantly male than theglobal Internet population demographics reported at pub-lication date (Cyberatlas.com 2000d).
Discriminant Function Analysis and Results
To test the ability of the hypothesized variables to dis-criminate between switchers and continuers, two-groupdiscriminant function analysis was used. The independentvariables were external influence, interpersonal influence,experiential influence, utilization rate (frequency, inten-sity, and overall), risk-taking behavior, income, and educa-tion. To check the validity of the models, a proportional50-50 split sample validation procedure was used. Resultsare shown in Tables 2, 3, and 4. Results reported in thetables are for the analysis sample, as recommended byHair et al. (1995).
The discriminant function was significant (Wilks’slambda = .66, p < .02), thereby confirming overall differ-
Keaveney, Parthasarathy / ONLINE SERVICES 381
TABLE 1Characteristics of the Two Samples
Sample 1 Sample 2
Variable Frequency Percentage Frequency Percentage
EducationGrade school 0 0 1 0.3Some high school 5 3.5 8 2.7High school graduate 8 5.6 30 10.0Trade/tech graduate 1 0.7 2 0.7Some college 23 16.2 43 14.4College graduate 58 40.8 127 42.5Advanced degree 47 33.1 88 29.4
ences in characteristics between switchers and continuers.Predictive accuracy of the discriminant function wasassessed by comparing the hit ratios of the analysis(68.1%), holdout (62.7%), and overall (67.6%) sampleswith chance. Because sample sizes in each of the groupsare about equal, simple, proportional, and maximumchance criteria are all about 50 percent. Hair et al. (1995)recommended that classification accuracy should be atleast 125 percent of that achieved by chance, making thedesired cutoff for this study to be 62.5 percent. Hit ratiosfor each sample exceed the recommended value. Signifi-cance of the discriminatory power of the classificationmatrix, when compared to chance, is indicated by Press’sQ statistic. Press’s Q-values were significant for the analy-sis (9.06, p < .01), holdout (4.31, p < .05), and overall(16.94, p < .01) samples.
Seven of the nine hypotheses had discriminant loadingswith a significance level of p < .05; an eighth was sup-ported at p < .07. Each of the hypotheses about informa-tional influence was supported at the p < .04 level or better.Hypothesis 1, which proposed that online servicecontinuers relied more on external sources of informationwhen making their subscription decisions than did onlineservice switchers, was supported by a structure coefficientof .37 (p < .03). Hypothesis 2, which proposed that onlineservice continuers relied less on interpersonal sources ofinformation when making the subscription decision thandid switchers, was supported with a structure coefficient of–.67 (p < .00). Hypothesis 3, which proposed thatcontinuers relied more on experiential sources of informa-tion than did switchers, was supported by a structure coef-ficient of .36 (p < .04).
Hypotheses 4a and 4c were supported, providing evi-dence that online service continuers can be discriminated
from switchers by their higher levels of some types of ser-vice usage. Online service continuers can be discriminatedfrom switchers by higher frequency of use (.65, p < .00)and overall usage (.41, p < .02), but not by intensity of use(.22, p < .20).
Hypothesis 6, which predicted that online servicecontinuers had higher average income levels than switch-ers, was supported (.38, p < .03). Hypothesis 7 was sup-ported only at p < .07, indicating somewhat qualified sup-port that online service continuers have higher averageeducation levels than switchers (coefficient .32).
Hypothesis 5 proposed that online service continuerswould have a lower propensity for risk-taking behaviorthan online service switchers. However, results indicatedthat continuers had a higher propensity for risk-takingbehavior than switchers (coefficient .39, p < .03). Groupmeans are reported in Table 5.
STUDY 2: REPLICATION AND EXTENSION
Involvement and satisfaction are consistent themesunderlying hypothesis development in Study 1, but theywere neither explicitly measured nor tested. Study 2 repli-cates the first study in a new sample and extends the set ofindependent variables to include measures of satisfactionand involvement.
Conceptual Developmentand Research Hypotheses
Prior studies have found that dissatisfaction is associ-ated with increased brand- (product-) switching behavior
382 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
TABLE 2Results of the Discriminant Analysis
Discriminant Loadings Study 2 Study 2
Study 2 Study 2Study 1 Replication Extension
Hypothesis Attribute Study 1 Replication Extension Ratio p Ratio p Ratio p
(LaBarbera and Mazursky 1983), service-switching be-havior (Bansal and Taylor 1999; Bolton and Bronkhorst1995; Keaveney 1995; Roos 1999), and both product- andservice-switching intentions (e.g., Anderson 1994; Ander-son and Sullivan 1993; Boulding et al. 1993; Cronin andTaylor 1992; Oliva et al. 1995; Oliver 1997; Reichheld andSasser 1990; Yi 1990; Zeithaml et al. 1996). Based onprior work, we, too, expect the following:
Hypothesis 8: Online service continuers are more satis-fied with the service than online service switchers.
More highly involved consumers have been found toreport higher levels of positive disconfirmation and satis-faction (Oliver and Bearden 1983; Richins and Bloch1991). These higher levels of satisfaction tend to be endur-ing. Involved consumers engage in greater prepurchasesearch (Beatty and Smith 1987) and greater deliberation inchoice (Celsi and Olson 1988). Higher levels of involve-ment have been associated with higher levels of commit-ment to a decision and resistance to belief change (Prit-chard et al. 1999).
Theories of cognitive consistency and assimilation pre-dict that under conditions of high product and/or serviceinvolvement, satisfied customers will be motivated to pre-serve their satisfied state across a wide range of perfor-mances, even when performance is somewhat discrepant.Researchers have found that satisfied customers tolerategreater degrees of poor performance as their product and/or service involvement increases (Oliva et al. 1995). Thisperseverance may continue until negative performance in-formation is sufficiently discrepant from the cognitiveschema as to demand attention and subsequent reprocess-ing (Fiske and Taylor 1984). Thus,
Hypothesis 9: Online service continuers experiencegreater involvement with the service than online ser-vice switchers.
The research methods used for Study 1 were generallyfollowed, with a few modifications. The procedure isbriefly reviewed here and changes noted.
The sampling frame of 28,217 names provided by themajor online service provider and used in Study 1 wasonce again used as the base list, after removing the recordsused in Study 1. In an effort to increase the number ofresponses, a randomly selected sample of 2,000 individu-als was generated. The base list was sorted into quittersand continuers based on the presence or absence of a
Keaveney, Parthasarathy / ONLINE SERVICES 383
TABLE 4Classification Results
Sample Type Actual Membership Continuers Switchers Total
service termination date, and 1,000 individuals were ran-domly selected from each group.
The 2000 individuals were sent postcards asking fortheir participation in the study. Once again, they wereasked (1) if they currently subscribed to an online service,(2) whether they had ever stopped using an online serviceprovider after using it for at least 3 months, and (3) theirnames and addresses.
One thousand one hundred and six postcards were com-pleted and returned. Of the 1,106 respondents, 495 weresent “continuer” questionnaires and 611 were sent “quit-ter” questionnaires, following the classification systemdescribed in Study 1. One reminder mailing was sent 15days following the initial mailing. At the end of the 30-daysurvey period, a total of 390 questionnaires were returned,for a 35 percent response rate among the study’s partici-pants or a 19.5 percent response rate relative to the initialpostcard mailing. Of the 390 respondents, 136 returned“continuer” and 254 returned “quitter” surveys.
Once again, to check for possible nonresponse bias,early respondents (those who responded within the first 2weeks) were compared with later respondents on each ofthe measured variables (Armstrong and Overton 1977).No significant differences in means of any measured vari-ables were found between the two groups (p < .05).
Measures of Variables
Measures of sources of informational influence, utili-zation level, risk-taking propensity, income, and educationwere the same as in Study 1. Reliability of the risk-takingpropensity scale was computed again for Study 2; reliabil-ity was .93. Measures of satisfaction and involvement arediscussed below. Items are reported in the appendix.
Satisfaction. Satisfaction was defined as an overall cog-nitive and affective state of happiness and contentment(Oliver 1997). Three items were developed for use in this
study, based on typical scales of satisfaction (see Yi 1990or Oliver 1997 for excellent reviews). Items are shown inthe appendix. Reliability for the satisfaction scale was .75.
Product involvement. Product involvement was definedas interest and expertise in the product class, followingZinkhan and Locander (1988). Their scale was selectedbecause, like the present study, it focused on customers’in-volvement and interest in a technology-related area (i.e.,calculators). The seven-item scale included measures ofinvolvement, interest, expertise, and experience with theproduct class. Reliability was .80 in this study, comparedwith .87 for Zinkhan and Locander’s (1988) study. Itemsare shown in the appendix.
Data Analysis and Results
Categorization of Groups
Further classification of the “quitter” respondents asservice discontinuers or service switchers was conductedas in Study 1. Of 390 respondents, 136 were continuers.The remaining 254 quitter responses were composed of172 service switchers and 82 service discontinuers. Again,service discontinuers were not used in the study. In sum-mary, the analysis included 136 usable continuer surveysand 172 usable switcher surveys. Characteristics of thesample are reported in Table 1.
Discriminant Function Analysis and Results
Replication. The first phase of data analysis examinedwhether the results of Study 1 were replicated in a differentsample. Following the procedures applied in Study 1, ninevariables were subjected to two-group discriminant func-tion analysis to determine their effectiveness in discrimi-nating between switchers and continuers. Also as inStudy 1, a proportional 50-50 split sample validation pro-cedure was used to check the classification accuracy of the
384 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
discriminant function. As reported below and in the tables,Study 2 replicates the results of Study 1.
The independent variables were (Hypothesis 1) exter-nal informational influence; (Hypothesis 2) interpersonalinformational influence; (Hypothesis 3) experientialinformational influence; (Hypotheses 4a, 4b, and 4c) fre-quency, intensity, and overall usage; (Hypothesis 5) pro-pensity for risk-taking behavior; (Hypothesis 6) income;and (Hypothesis 7) education. As before, the discriminantfunction was significant (Wilks’s lambda = .67, p = .00),again confirming overall differences in characteristicsbetween switchers and continuers. Hit ratios of the analy-sis (73.1%), holdout (72.6%), and overall (70.6%) sampleswere well in excess of the 62.5 percent cutoff rate recom-mended for a 25 percent improvement over chance, againconfirming the predictive accuracy of the discriminantfunction. Press’s Q-value was significant for the analysis(27.7, p < .01), holdout (27.6, p < .01), and overall (44.83,p < .01) samples. Discriminant function and classificationresults are shown in Tables 2, 3, and 4.
As in Study 1, continuers were discriminated fromswitchers by their use of information sources. Continuerswere significantly more likely than switchers to use exter-nal information sources (coefficient .26, p < .04), were sig-nificantly less likely than switchers to use interpersonalinformation sources (coefficient –.67, p < .00), and weresignificantly more likely than switchers to use experientialinformation sources (coefficient .56, p < .00) when makingthe online subscription decision. Once subscribed,continuers used the online service significantly more fre-quently than switchers (coefficient .51, p < .00) and exhib-ited greater overall usage (coefficient .25, p < .05), but nodifferences in the intensity of usage were detected (coeffi-cient .15, p < .25). Continuers were again characterized bysignificantly higher income (coefficient .44, p < .00) butonly somewhat higher education levels (coefficient .24,p < .06). The surprising finding that continuers have a sig-nificantly higher propensity for risk-taking behavior thanswitchers was also replicated (coefficient .42, p < .00).Group means are shown in Table 5.
Extension. Study 2 then extended the scope of Study 1by including measures of satisfaction and involvement inthe discriminant function analysis. As shown in Tables 2,3, and 4, classification results are improved with the addi-tion of the two attitudinal variables.
The discriminant function was once again significant(Wilks’s lambda = .61, p < .00). Hit ratios of the analysis(83.8%), holdout (77.8%), and overall (79.6%) sampleswere substantially improved and are now well above the62.5 percent cutoff rate suggested for a 25 percentimprovement over chance. Classifications based on thediscriminant function are now 67 percent, 56 percent, and59 percent improvements over chance, respectively.Press’s Q-values were significant for the analysis (59.57,
p < .01), holdout (41.67, p < .01), and overall (93.02, p <.01) samples.
Both new hypotheses, Hypotheses 8 and 9, were sup-ported at the p < .02 level or better. Hypothesis 8, whichproposed that online service continuers would be discrimi-nated from online service switchers by higher levels of sat-isfaction with the service, was supported by a structurecoefficient of .42 (p < .00). Hypothesis 9 proposed thatonline service continuers would be discriminated fromonline service switchers by greater involvement with theservice. Hypothesis 2 was supported with a structure coef-ficient of .27 (p < .02). The significance of the other vari-ables in the discriminant function equation did not change.Thus, the improvement in the classification accuracy ofthe discriminant function appears to be due entirely to theaddition of satisfaction and involvement to the model.Implications of Study 1 and Study 2 (replication andextension) for managers and researchers are discussed inthe following sections.
The goal of this research was to explore whetherselected attitudinal, behavioral, and demographic charac-teristics of consumers might be effective in discriminatingbetween switchers and continuers of online services. Twoempirical studies, conducted in field settings, supportedthe general proposition that online service switchers havecharacteristics that are identifiably and significantly dif-ferent from those of online service continuers.
Contributions of the Research
The research contributes to the services marketing liter-ature by identifying defection-prone customers and pro-viding further explanation of customer service-switchingbehavior. In summary, the profile of an online serviceswitcher is that of an individual who was influenced tosubscribe to the service through word of mouth, ratherthan through research or previous experience; who usedthe service less; who was less satisfied and less involvedwith the service; and who had a lower income and educa-tion level, as well as a lower propensity for taking risks.
This research makes a significant contribution to theword-of-mouth literature (WOM) with the interestingfinding that switchers were more likely to have relied onWOM sources when making their subscription decisions.Researchers in the services marketing area recognize theimportance of WOM communications in the context ofservices (Murray 1991; Zeithaml 2000), with some sug-gesting that WOM may be an even more important com-munication tool for services firms than is advertising(Danaher and Rust 1996a, 1996b; Kordupleski, Rust, andZahorik 1993; Murray 1991). The present research clearly
Keaveney, Parthasarathy / ONLINE SERVICES 385
identifies some possible downside risks heretoforeunidentified in the literature.
Some of our findings support, and contribute to, theexperiential learning literature (Hoch and Deighton 1989).First, continuers had more prior experience with the ser-vice or with related services than did switchers. Second,continuers had more experience with the service after pur-chasing, in the form of increased usage, than did switchers.We propose that these results are explained by experientiallearning theory, but testing of the explicit linkages is agood topic for future research.
Our findings that switchers had significantly lower ser-vice usage than continuers also contributes to the rela-tively new body of services research investigating the rela-tionships between service usage and customer satisfaction,repurchase intentions, and switching (cf. Anderson 1994;Anderson and Sullivan 1993; Bolton and Bronkhurst1995; Bolton and Lemon 1999). While the results reportedhere provide support for the relationship between serviceusage and continuance, future research should explore fur-ther explanations for the link.
The findings of this study also extend services market-ing literature in the area of consumer information-processingtheory. Specifically, services literature has revealed thatconsumers prefer to use external, impersonal sources ofinformation less when making purchasing decisions forservices than they do when making purchase decisions forgoods (Murray 1991). The research reported here indi-cates that there are repercussions to those preferences, thatis, lower use of external, impersonal sources of informa-tion when making an online service choice was related tohigher service switching. Services research has alsoshown that consumers are less likely to make an outrightpurchase or use direct observation or trial as informationstrategies when purchasing services relative to goods(Murray 1991). The results of this study indicate thatreduced reliance on these experiential sources is associ-ated with increased service-switching behavior. Hence,service customers’ least preferred modes of informationacquisition are, contrarily, the two modes best suited toservice continuance.
Finally, this study provides two new contributions tothe satisfaction literature. First, and perhaps most interest-ingly, the study contributes to recent work on informationsatisfaction. Spreng, MacKenzie, and Olshavsky (1996)found that customers’ overall satisfaction was composedof both product satisfaction and information satisfaction,or satisfaction with the information provided to customersas they were making their decisions. (Their study focusedonly on external, impersonal information sources.) Wehypothesized that the differential effects of types of infor-mation were due to the accuracy of predictive expecta-tions. An alternative explanation for our results might be
that consumers experienced satisfaction or dissatisfactionwith the different information sources and that informa-tion satisfaction mediated the information-switching rela-tionship. This question provides another interesting areafor future research. Second, the research reported here pro-vides additional evidence of the behavioral outcomes(here, customer service-switching and continuance behav-iors), of satisfaction and/or dissatisfaction, and serviceinvolvement.
Implications for Managers
A major contribution of this article is that it empiricallyidentifies and profiles defection-prone customers ofonline services that will enable managers to develop “earlywarning systems” to target potential switchers and takeaction to retain them. One goal of this research was toselect variables that were not only theoretically sound butmanagerially implementable. For example, results indi-cating that switchers and continuers could be identified bydifferences in levels of income and (to some degree) edu-cation is a particularly useful finding since income andeducation are two of the more accessible demographicsegmentation variables. Service usage data in generalshould be available given good internal marketing infor-mation systems, and online service usage data are readilyavailable. Measuring the relative importance of differenttypes of information used when making the subscriptiondecision requires only the addition of a brief question atthe time of enrollment, a variation of the question alreadyasked by many firms (“How did you hear about us?”). Sat-isfaction is another variable already collected by many ser-vices firms. Items for the remaining variables, risk-takingpropensity and service involvement, could be added toonline companies’ data collection efforts.
On the basis of our results, we recommend that manag-ers target customer retention strategies at both pre- andpostpurchase stages of the consumer’s decision process.During prepurchase phases, marketers should increasecustomer involvement in the decision-making process,increase trial and other experiential opportunities, makemarketer-generated sources of information more accessi-ble or more appealing, and generally help potential cus-tomers to engage in active learning about the service. Dur-ing postpurchases phases, marketing activities should bedesigned to increase customer satisfaction, involvement,and service usage (especially frequency), thereby reduc-ing the likelihood of customer switching. In particular, ser-vice usage results were consistent with satisfaction the-ory’s proposition that more incidences of satisfied usageshould lead to the formation of attitudes that are resilient todiscrete incidences of service failure. Since service fail-ures are inevitable, service firms might find that
386 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
encouraging more usage, greater familiarity, and moreexperience with the service, both before and after the pur-chase decision, would lead their customers to form thesomewhat more resilient positive attitudes evidenced bythe continuers.
Implications for Future Research
The research reported here contributes another set offindings to the study of customer service-switching behav-ior. Much work remains. Some ideas for future researchare offered.
The study was conducted among a field sample of for-mer and current customers of a major online service pro-vider. Future research should also examine the general-izability of results across other subscription-based servicessuch as telecommunications, cable, insurance, public utili-ties, health care, and other membership-based services. Inaddition, the study was concerned with the differencebetween two broadly defined groups—switchers versuscontinuers—and did not further classify switchers by thefrequency of their switching behavior or the completenessof the switch. Future research work might examine the dif-ference between frequent and infrequent switchers (e.g.,as characterizes some customers in the long-distance tele-phone industry) and between complete and “partial” or“incomplete” switchers (those customers who reduce uti-lization or transfer part, but not all, of their accounts).
One limitation of discriminant function analysis is that,like many multivariate analysis techniques, it permits oneto determine the overall effects of a set of variables on thedependent variable but does not allow for examination ofrelationships between the variables. The results can bethought of as providing a broad-brush outline of thedomain—variables x, y, z are good predictors; variables a,b, c are not—with the investigation of causal relationshipsthe work of subsequent research stages. This research gen-erates many interesting causal questions regarding theunderlying mechanisms through which the various attitu-dinal, behavior, and demographic variables ultimatelyaffect switching behavior.
The unexpected finding that continuers, rather thanswitchers, had higher risk-taking propensities was puz-zling and is a good topic for future investigation. As dis-cussed when generating the hypothesis, both theory andevidence suggest that customers perceive services to beriskier than goods. Perhaps the risk-taking aspect of thechoice was in the adoption of the risky new service, not inswitching from one service provider to another within anow-familiar service class. Another possibility, in the fast-changing world of technology, is that the riskier choice isto remain loyal to a service provider in the face of manynew market entrants. Alternately, perhaps the risk-prone
customer’s decision to purchase a risky service stimulatedincreased information search (Murray 1991), which led tomaking a more involved, or at least a more sound, deci-sion—qualities associated with continuance rather thanswitching behavior.
Finally, future research should examine not only reve-nue generation but also profitability of service switchersand continuers. Recent articles have begun to emphasizethe importance of segmenting the “right” customers(Blattberg and Deighton 1996; Dowling and Uncles 1997;Jeffrey and Franco 1996). It will be important for servicemanagers to determine who among the potential switchersshould be targeted for retention.
Sources of Influence
How much did each of these sources influence you to sub-scribe to [this/that]a service? Please make sure that the totalequals 100 percent.
1. [External sources of influence]: _____Articles, reviews, advertising, or other activities of
2. [Interpersonal sources of influence]: _____Opinions of friends, colleagues, relatives, or others
3. [Experiential sources of influence]: _____My own personal experience and general computer
1. Frequency: Typically, how often [do/did] you use this ser-vice?
_____ Less than once a month_____ Once a month_____ Once a week_____ 2 to 3 times a week_____ Most days
2. Intensity: Typically, how much time [do/did] you spend us-ing the service each time you logged on?
_____ Less than 15 minutes_____ 15 to 30 minutes_____ 31 minutes to 1 hour_____ 1 to 2 hours_____ More than 2 hours
Keaveney, Parthasarathy / ONLINE SERVICES 387
Risk-Taking Scale (Raju 1980)
1 = strongly disagree, 7 = strongly agree1. When I eat out, I like to try the most unusual items the res-
taurant serves even if I am not sure I would like them.2. I am the kind of person who would try any new product
once.3. When I go to a restaurant, I find it safer to order dishes I am
familiar with (reverse).4. I am cautious in trying new/different products (reverse).5. Even for an important date or dinner, I wouldn’t be wary of
trying a new or unfamiliar restaurant6. I would rather stick to a brand I usually buy than try some-
thing I am not very sure of (reverse).7. I never buy something I don’t know about at the risk of
making a mistake (reverse).
1 = strongly disagree, 7 = strongly agree1. On the whole, I [am/was] satisfied with my experience with
[this/that] service.2. Overall, my negative experience [outweighs/outweighed]
my positive experience with [this/that] service (reverse).3. In general, I [am/was] happy with the service experience.
1 = strongly disagree, 7 = strongly agree1. I am very interested in online services.2. My level of involvement with online services is high.3. I am particularly engaged in the online service environ-
ment.4. I consider myself an expert on the online electronic envi-
ronment.5. I consider myself an Internet expert.6. I purchase products from online vendors regularly.7. My level of expertise regarding personal computers is high.
1. $20,000 or less2. $20,001-$30,0003. $30,001-$40,0004. $40,001-$50,0005. $50,001-$60,0006. $60,001-$70,0007. $70,001-$80,0008. $80,001-$90,0009. More than $90,000
1. Grade school2. Some high school
3. High school graduate4. Trade/tech graduate5. Some college6. College graduate7. Advanced degree
a. The brackets refer to [continuer/switcher] verisons of the surveys.Continuers answered with regard to the service they currently subscribeto; switchers answered with regard to the service they most recentlyswitched from.
Alba, Joseph W. and J. Wesley Hutchinson. 1987. “Dimensions of Con-sumer Expertise.” Journal of Consumer Research 13 (March): 411-454.
Anderson, Eugene W. and Mary Sullivan. 1993. “The Antecedents andConsequences of Customer Satisfaction for Firms.” Marketing Sci-ence 12 (Spring): 125-143.
Armstrong, J. Scott and Terry S. Overton. 1977. “EstimatingNonresponse Bias in Mail Surveys.” Journal of Marketing Research14 (August): 399-403.
Bansal, Harvir S. and Shirley F. Taylor 1999. “The Service ProviderSwitching Model (SPSM): A Model of Consumer Switching Behav-ior in the Services Industry.” Journal of Service Research 2 (Novem-ber): 200-218.
Beatty, Sharon E. and Scott M. Smith. 1987. “External Search Effort: AnInvestigation Across Several Product Categories.” Journal of Con-sumer Research 1:83-95.
Berning, Carol A. and Jacob Jacoby. 1974. “Patterns of Information Ac-quisition in New Product Purchases.” Journal of Consumer Research1 (September): 18-22.
Bettman, James R. 1979. An Information Processing Theory of Con-sumer Choice. Reading, MA: Addison-Wesley.
Blattberg, Robert C. and John Deighton. 1996. “Manage Marketing bythe Customer Equity Test.” Harvard Business Review 74 (July-Au-gust): 136-144.
Bolton, Ruth N. and Tina M. Bronkhorst. 1995. “The Relationship Be-tween Customer Complaints to the Firm and Subsequent Exit Behav-ior.” Advances in Consumer Research 22:94-100.
and Katherine N. Lemon. 1999. “A Dynamic Model of Cus-tomers’Usage of Services: Usage as an Antecedent and Consequenceof Satisfaction.” Journal of Marketing Research 36 (May): 171-186.
Boulding, William, Ajay Kalra, Richard Staelin, and Valarie A.Zeithaml. 1993. “A Dynamic Process Model of Service Quality:From Expectations to Behavioral Intentions.” Journal of MarketingResearch 30 (February): 7-27.
Carl, Jeremy. 1995. “Online Service Users: Loyal as Alley Cats?” WebWeek 1 (November).
Celsi, Richard L. and Jerry C. Olson. 1988. “The Role of Involvement inAttention and Comprehension Processes.” Journal of Consumer Re-search 15:210-224.
Cohen, Jacob and P. Cohen. 1980. Applied Multiple Regression/Correla-tion Analysis for the Behavioral Sciences. New York: LawrenceErlbaum.
Cronin, Joseph J. Jr. and Steven A. Taylor. 1992. “Measuring ServiceQuality: A Reexamination and Extension.” Journal of Marketing 56(July): 55-68.
Dabholkar, Pratibha and Simon Walls. 1999. “Service Evaluation andSwitching Behavior for Experiential Services: An Empirical Test ofGender Differences Within a Broader Conceptual Framework.” Jour-nal of Consumer Satisfaction, Dissatisfaction, and Complaining Be-havior 12:123-137.
Danaher, Peter J. and Roland T. Rust. 1996a. “Determining the OptimalReturn on Investment for an Advertising Campaign.” European Jour-nal of Operational Research 95 (3): 511-521.
388 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
and . 1996b. “Indirect Financial Benefits From ServiceQuality.” Quality Management Journal 3 (2): 63-85.
Deighton, John. 1984. “The Interaction of Advertising and Evidence.”Journal of Consumer Research 11 (December): 763-770.
Dowling, Grahame R. and Mark Uncles. 1997. “Do Customer LoyaltyPrograms Really Work?” Sloan Management Review (Summer): 71-82.
Fiske, Susan T. and Shelley E. Taylor. 1984. Social Cognition. New York:Random House.
Fornell, Claes and Birger Wernerfelt. 1987. “Defensive Marketing Strat-egy by Customer Complaint Management: A Theoretical Analysis.”Journal of Marketing Research 24 (November): 337-346.
Furse, David H., Girish N. Punj, and David W. Stewart. 1984. “ATypology of Individual Search Strategies Among Purchasers of NewAutomobiles.” Journal of Consumer Research 10 (March): 417-431.
Ganesh, Jaishankar, Mark J. Arnold, and Kristy E. Reynolds. 2000. “Un-derstanding the Customer Base of Service Providers: An Examina-tion of the Differences Between Switchers and Stayers.” Journal ofMarketing 64 (July): 65-87.
Grant, Alan W. and Leonard A. Schlesinger. 1995. “Realize Your Cus-tomer’s Full Profit Potential.” Harvard Business Review 73 (Septem-ber-October): 59-72.
Guseman, Dennis S. 1981. “Risk Perception and Risk Reduction in Con-sumer Services.” In Marketing of Services. Eds. James Donnelly andWilliam George. Chicago: American Marketing Association, 200-204.
Hair, Joseph F. Jr., Rolph E. Anderson, Ronald L. Tatham, and William C.Black. 1995. Multivariate Data Analysis. Englewood Cliffs, NJ:Prentice Hall.
Heskett, James L., W. Earl Sasser Jr., and Leonard A. Schlesinger. 1997.The Service Profit Chain. New York: Free Press.
Hoch, Stephen J. and John Deighton. 1989. “Managing What ConsumersLearn From Experience.” Journal of Marketing 53 (April): 1-20.
Houston, Michael J. 1979. “Consumer Evaluation of Product Informa-tion Sources.” In Current Issues and Research in Advertising. Eds.James H. Leigh and Claude R. Martin, Jr. Ann Arbor: University ofMichigan, 135-144.
Jeffrey, Jaclyn R. and John J. Franco. 1996. “Identifying a Company’sMost Profitable Customers: The First Step in Intelligent CustomerRetention.” Information Strategy: The Executive’s Journal, Summer,pp. 15-21.
Keaveney, Susan M. 1995. “Customer Switching Behavior in Service In-dustries: An Exploratory Study.” Journal of Marketing 59 (2): 71-82.
Kingsley, Paul and Terry Anderson. 1998. “Facing Life Without theInternet.” Internet Research: Electronic Networking Applicationsand Policy 8 (4): 303-312.
Kordupleski, Raymond E., Roland T. Rust, and Anthony J. Zahorik.1993. “Why Improving Quality Doesn’t Improve Quality (or What-ever Happened to Marketing?).” California Management Review 35(3): 82-95.
Kotler, Philip R. 1997. Marketing Management: Analysis, Planning, Im-plementation, and Control. Englewood Cliffs, NJ: Prentice Hall.
LaBarbera, Priscilla A. and David Mazursky. 1983. “A Longitudinal As-sessment of Consumer Satisfaction/Dissatisfaction: The DynamicAspect of the Cognitive Process.” Journal of Marketing Research 20(November): 393-404.
Levin, Rich. 1997. “The Ups and Downs of ISPs.” Information Week 657(November 17).
Moore, William L. and Donald R. Lehman. 1980. “Individual Differ-ences in Search Behavior for a Nondurable.” Journal of ConsumerResearch 7 (December): 296-307.
Murray, Keith B. 1991. “A Test of Services Marketing Theory: ConsumerInformation Acquisition Activities.” Journal of Marketing 55 (Janu-ary): 10-25.
and John L. Schlacter. 1990. “The Impact of Services VersusGoods on Consumers’ Assessments of Perceived Risk and Variabil-ity.” Journal of the Academy of Marketing Science 18 (1): 51-65.
Oliva, Terence A., Richard L. Oliver, and William O. Bearden. 1995.“The Relationships Among Consumer Satisfaction, Involvement,and Product Performance: A Catastrophe Theory Application.” Be-havioral Science 40:104-132.
Oliver, Richard L. 1997. Satisfaction. New York: McGraw-Hill. and William O. Bearden. 1983. “The Role of Involvement in Sat-
isfaction Processes.” In Advances in Consumer Research. Eds. Rich-ard P. Bagozzi and Alice M. Tybout. Ann Arbor, MI: Association forConsumer Research, 250-255.
Peters, Thomas J. 1988. Thriving on Chaos. New York: Knopf.Pritchard, Mark P., Mark E. Havitz, and Dennis R. Howard. 1999. “Ana-
lyzing the Commitment-Loyalty Link in Service Contexts.” Journalof the Academy of Marketing Science 27 (3): 333-348.
Raju, P. S. 1980. “Optimum Stimulation Level: Its Relationship to Per-sonality, Demographics, and Exploratory Behavior.” Journal of Con-sumer Research 7 (December): 272-282.
Ram, S. and Hyung-Shik Jung. 1990. “The Conceptualization of ProductUsage.” Journal of the Academy of Marketing Science 18 (Winter):67-76.
Susan M. Keaveney (Ph.D., University of Colorado) is a profes-sor of marketing at the University of Colorado at Denver. Sheteaches marketing management, international marketing, andservices marketing for the Graduate School of Business and Ex-ecutive MBA Programs and has received numerous awards forExcellence in Teaching. Prior to commencing her academic ca-reer, she worked in the retailing, financial services, and healthcare industries. Her research in services marketing focuses oncustomer retention and service switching and has been publishedin such journals as the Journal of Marketing, the Journal of theAcademy of Marketing Science, the Journal of Retailing, Mar-keting Management, the Journal of Marketing Education, Psy-
chology and Marketing, the Journal of Marketing Channels, theJournal of Promotion Management, the Journal of Marketing inHigher Education, and the Journal of Business and Psychology.In 1996, she received the American Marketing Association’sSERVSIG award for the Best Services Article in 1995. She is co-author, with Philip R. Cateora, of Marketing: An InternationalPerspective, published both in English and in Japanese transla-tion and was actively involved in the “internationalization” ofbusiness schools in the United States.
Madhavan Parthasarathy (Ph.D., University of Nebraska atLincoln) is an assistant professor of marketing at the Universityof Colorado in Denver. He teaches Internet Marketing, NewProduct Development, Marketing Management, and MarketingResearch for the Graduate School of Business. His research in-terests focus on customer retention, switching and discontinu-ance behavior, and the diffusion of new products. He haspublished in such journals as Information Systems Research, In-ternational Marketing Review, the Journal of Macromarketing,and the Journal of Business and Industrial Marketing, amongothers.
390 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001Bush et al. / INTERCULTURAL COMMUNICATION
Managing Culturally DiverseBuyer-Seller Relationships: The Role ofIntercultural Disposition and AdaptiveSelling in Developing InterculturalCommunication Competence
Victoria D. BushGregory M. RoseFaye GilbertUniversity of Mississippi
Thomas N. IngramColorado State University
Given the increase in cultural diversity within marketingorganizations as well as within current and potential cus-tomer bases, possessing the appropriate communicationskills becomes crucial to success in managing culturallydiverse relationships. Although marketing researchershave recognized the importance of adaptive selling behav-ior for successful buyer-seller relationships, the explora-tion of the intercultural aspects of these relationships hasonly recently begun. This article examines how adaptiveselling behaviors and intercultural dispositions of market-ing executives contribute to their perceived interculturalcommunication competence. Results show that in additionto being adaptive, the intercultural disposition of a mar-keter is of key importance in developing intercultural com-munication competence. Theoretical and practicalimplications for incorporating intercultural communica-tion into the development of successful buyer-seller rela-tionships are discussed.
In recognition of an increasingly diverse workforce andcustomer base in the United States, marketers haveacknowledged the importance of studying cultural differ-ences to better understand, develop, and manage culturallydiverse buyer-seller relationships. Webster (1992) stressesthat marketers could benefit further by working with mem-bers from other disciplines such as cultural anthropologyand sociology for a more in-depth understanding of thedifferences in values, beliefs, or decision-making pro-cesses, among other constructs, that contribute to manag-ing culturally diverse relationships (Deshpande and Web-ster 1989; Montgomery 1991; Montgomery and Weiss1991; Webster and Deshpande 1991). Intercultural com-munication is one such discipline that can contribute to thedevelopment of successful buyer-seller relationships.
The purpose of this article is to discuss and investigatethe role that intercultural communication and adaptiveselling behavior play in developing the intercultural com-munication competence of marketing executives. First,relevant literature in marketing and buyer-seller relation-ships concerning adaptive selling and communication isdiscussed. Second, intercultural communication is intro-duced with a discussion of intercultural disposition and
competence. Third, a model is developed and testedamong marketing executives. Finally, implications formarketing managers and academic researchers arediscussed.
Cultural diversity is defined as differences in age, gen-der, race, nationality, and ability (Loden and Rosener1991). The proportion of diverse groups in the workforcehas been steadily increasing (Seal 1991). For example, inpersonal selling, the growth rate of women, African Amer-icans, Hispanics, and Asians has increased tremendouslyin recent years (Comer, Nicholls, and Vermillion 1998). Inaddition, according to the American Advertising Federa-tion, the annual spending power of Asians, Hispanics, andAfrican Americans in the United States is approximately$1 trillion (Bertagnoli 2001). This increasingly diversecustomer base is one of the top reasons for marketers tovalue diversity whether it be in the United States or whenexpanding into global markets (Bertagnoli 2001). Accord-ing to Thansee Mustafa, chair of College Relations atJohnson & Johnson, “A diverse workforce ensures theinnovative thought necessary within a company to remaincompetitive” (Bertagnoli 2001:9).
Because of these trends, marketing managers are facedwith two challenges. First, they will need to expend moreeffort to recruit, retain, and manage culturally diverse mar-keting executives. Second, to remain competitive, theymust recognize the diversity of their present and futurecustomer base to develop and maintain successful buyer-seller relationships.
Cultural diversity in buyer-seller relationships has onlyrecently been recognized as an important and under-researched issue (Comer and Nicholls 2000; Comer et al.1998). Most of the existing research has focused onwomen in the sales force (e.g., Comer and Jolson 1991).Much less has been done on other groups such as AfricanAmericans, Hispanics, and Asians (Comer and Nicholls2000; Comer et al. 1998). Equally important is researchinvestigating culturally diverse customers. Some initialresearch has been done on similarities between buyers andsellers in terms of age and gender (Dwyer, Richard, andShepherd 1998; Kang and Hillary 1998). However, moreis needed on the impact of diverse communication styles inthe buyer-seller interaction (Comer et al. 1998). Further-more, cultural diversity research would be strengthened ifit could be incorporated into an overall theoretical frame-work. It is suggested here that the theoretical foundationsof adaptive selling behavior and intercultural communica-tion can provide the basis for incorporating cultural diver-sity issues into a viable stream of research for buyer-sellerrelationship management.
Adaptive Selling Behavior
A genuine customer orientation requires that marketerslearn as much about their customers’ communicationstyles, behaviors, and needs as possible. Accordingly,adaptive selling has been defined as “the altering of salesbehaviors during a customer interaction or across cus-tomer interactions based on perceived information aboutthe nature of the selling situation” (Weitz, Sujan, andSujan 1986:175).
Researchers have noted that adaptive selling can bepracticed in effective and ineffective ways (e.g., Spiro andWeitz 1990; Weitz et al. 1986). Adaptive selling is onlyeffective in the long term when benefits outweigh the costsof selecting and training salespeople to practice thisapproach, such as in the development of a long-termbuyer-seller relationship. The effectiveness of adaptiveselling is moderated by (1) the variety of customer needsand types encountered by the sellers, (2) the importance ofthe typical buying situation to the seller, (3) the resourcesprovided by the company, and (4) the seller’s skills andcapabilities (Weitz et al. 1986).
Cultural diversity is relevant when considering thepotential impact of these four moderators. Adaptive sell-ing is thought to be more effective when encountering aheterogeneous customer base. With increases in ethnicbusinesses, customers, and foreign opportunities, market-ers have and will increasingly encounter a wide variety ofcustomers from a cultural perspective.
The importance of the buying situation as discussed byWeitz et al. (1986) relates to the economic significance,that is, size of the order. While larger orders may indeed beworth the extra effort to practice adaptive selling, theimportance of an individual order may not completelycharacterize the full significance of the buying situation.Recent emphasis on the value of a customer over anextended period of time, even a lifetime, suggests that thesignificance of buying situations might be defined byorder size in the short run or by the economic value of acustomer during an extended period of time. To build andsustain customer relationships during extended time hori-zons, marketers must recognize that as the workforcebecomes more diverse, so will the population of buyersand purchasing agents. Thus, understanding cultural dif-ferences among buyers and practicing adaptive selling areexpected to become even more crucial for success in thefuture.
With these changes in the cultural composition of theworkforce, companies are committing more resources tobetter understand the economic impact of culturallydiverse workforces, to pursue diverse markets, and toselect and train employees to be more effective with
392 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
diverse customer bases (Wentling 2000). Finally, market-ing executives must have the capability to adapt to cultur-ally diverse buyer-seller interactions. Hence, someresearch has recognized, to a limited extent, cultural diver-sity in the adaptive selling framework. For example, Levyand Sharma (1994) found that age and gender interact andhave an impact on adaptive selling. However, littleresearch exists on the diversity of customers in the adap-tive selling framework.
Communication in Buyer-Seller Relationships
Communication can be defined as occurring whenevermeaning is attributed to behavior (Porter and Samovar1991). Communication has been recognized as an impor-tant construct in effective buyer-seller and channel rela-tionship management (Anderson and Narus 1990;Boorom, Goolsby, and Ramsey 1998; Mohr and Nevin1990; Williams, Spiro, and Fine 1990). The role of com-munication in relationship marketing has historically beenreduced to the assertion that “more communication” or“improved communication” is needed. Mohr and Nevin(1990) state that this view is not only simplistic but alsoinaccurate. For example, more communication in a rela-tional situation where a high degree of conflict exists is notalways the answer if that communication is in the form ofthreats. Furthermore, simply stating that communicationneeds improvement does not give marketers any input onhow to improve communication (Bush and Ingram 2001).
Williams et al. (1990) addressed the need to investigatebuyer-seller relationships from a more detailed communi-cation perspective. Marketing researchers have investi-gated a number of communication skills and their potentialeffect on performance. Such abilities as gathering infor-mation and asking questions (Schuster and Danes 1986),listening or the lack thereof (Ingram, Schwepker, andHutson 1992), and recognizing different communicationstyles (Williams and Spiro 1985) have been advocated ascritical to successful buyer-seller interactions. However,little research in this area has been conducted using adap-tive selling behavior as a theoretical framework (e.g.,Boorom et al. 1998).
An exception to this observation is a recent study byBoorom et al. (1998). These authors demonstrated thatcertain communication traits affect the adaptiveness ofsalespeople. Citing research conducted in the communica-tion competence literature (e.g., Spitzberg and Cupach1984), the authors state that “communication effectivenessis viewed as arising from a set of dispositions and traits thatfacilitate receiving, deciphering, and responding appropri-ately to information in dyadic interactions” (Boorom et al.1998:17). It is suggested here that these dispositions andtraits become even more crucial in a culturally diverse set-ting. Specifically, if marketing executives possess such
dispositions and traits, their communication strategy willbe more appropriate in a culturally diverse buyer-sellerinteraction. Thus, the adaptive selling framework providesa bridge for incorporating intercultural communicationinto the marketing literature.
a symbolic, interpretive, transactional, contextualprocess in which the degree of difference betweenpeople is large and important enough to create dis-similar interpretations and expectations about whatare regarded as competent behaviors that should beused to create shared meanings. (Lustig and Koester1993:58)
In other words, intercultural communication occurs when-ever a message in one culture must be processed in anotherculture (Porter and Samovar 1991). In the marketing con-text, intercultural communication occurs when a messagefrom a buyer (seller) from one culture must be processedby a seller (buyer) from another (Bush and Ingram 1996).
Research in intercultural communication has uncov-ered dispositions and skills that can affect buyer-sellerinteractions. It is posited here that these constructs can beincorporated into a theoretical framework that integratescultural diversity into buyer-seller relationship manage-ment. Furthermore, previous research in the interculturalcommunication and education literature has focused pri-marily on the intercultural dispositions of college stu-dents, military personnel, and the international relocationof corporate personnel (Bradford, Allen, and Beisser1997). To date, no research has specifically investigatedthe intercultural disposition of marketing executives inrelation to the adaptive selling framework. Figure 1depicts a model of intercultural communication and adap-tive selling. The components of the model are discussed asfollows.
Of key importance to successful intercultural commu-nication is to possess what Gudykunst, Wiseman, andHammer (1977) call a general “cross-cultural attitude” orintercultural disposition. An individual’s intercultural dis-position has been described as the “core” of effectiveintercultural communication. Gudykunst et al. (1977)developed a framework for intercultural disposition thatincludes (1) the ability to empathize with people fromother cultures, (2) being astute noncritical observers oftheir own and other people’s behavior, (3) being lessethnocentric, and (4) being accurate in perceiving differ-ences and similarities between the sojourner’s own culture
Bush et al. / INTERCULTURAL COMMUNICATION 393
and the host culture. Researchers have conceptualized andmeasured these characteristics in an attempt to predictintercultural effectiveness or cross-cultural adaptation(e.g., Cui and Van Den Berg 1991; Hammer, Gudykunst,and Wiseman 1978). Some of these capabilities have alsobeen investigated in the marketing literature (e.g., McBane1995; Spiro and Weitz 1990) and are most relevant to thebuyer-seller interaction.
Empathy. An empathic tendency refers to the capacityto clearly project an interest in others and to obtain and re-flect a reasonably complete and accurate sense of an-other’s thoughts, feelings, and experiences. Past researchhas shown that persons who tend to empathize with othersalso tend to be more interculturally effective (Cleveland,Mangone, and Adams 1960; Ruben 1976). A certain de-gree of empathy will allow an individual to further under-stand a member of another culture’s point of view.
The importance of empathizing with others has beenrecognized in the marketing literature (e.g., Comer andDrollinger 1999; McBane 1995; Spiro and Weitz 1990;Weitz et al. 1986). In more specific terms, Davis (1983)proposed that empathy refers to the reaction of individualsto the observed experiences of others. These observers canhave a variety of empathetic reactions that consist of per-spective taking, fantasy, empathetic concern, and personaldistress. The cognitive component of empathy, perspectivetaking, has been the most widely accepted theoretical per-spective in the marketing literature (McBane 1995), andconventional wisdom holds that this dimension shoulddirectly help the marketing executive involved in a buyer-seller interaction (Ingram et al. 1992). Thus, the cognitiveaspect of empathy may aid in understanding culturalaspects of the buyer’s decision-making process.
Worldmindedness. According to Gudykunst et al.(1977), part of one’s intercultural disposition includes be-
ing an astute noncritical observer of one’s own and otherpeople’s behavior. Thus, an individual must be receptive tonew ideas and ways of looking at the world—even enjoy-ing new experiences. Worldmindedness refers to an indi-vidual’s frame of reference, apart from knowledge about,or interest in, cultural diversity (Sampson and Smith1957). This “cosmopolitan” perspective toward the worldhas been well documented in its ability to facilitateintercultural communication (Wiseman, Hammer, andNishida 1989) in a variety of settings such as study-abroadprograms (Carlson and Widaman 1988) and internationalrelocation (Bhawuk and Triandis 1996). Furthermore,worldmindedness has been shown to increase knowledgeof one’s own culture as well.
In marketing, Weitz and Jap (1995) emphasize thatrelationship development within channels of distributionis affected by the manner in which intentions, expecta-tions, and information are communicated. In a culturallydiverse interaction between a buyer and a seller, the sellermust be receptive to the cultural background of the buyer(Bush and Ingram 1996; Honeycutt and Ford 1995). AsSharma, Shimp, and Shin (1995) state, “Individuals differin terms of their experience with and openness toward thepeople, values, and artifacts of other cultures” (p. 28).Thus, the marketing executive who is worldminded (i.e.,an astute observer of his or her own culture as well as oth-ers) may be in a better position to develop a relationshipwith a buyer from a different cultural background.
Ethnocentrism. The third aspect of intercultural dispo-sition that Gudykunst et al. (1977) refer to is being lessethnocentric. Ethnocentrism is the tendency to identifywith one’s in-group (ethnic, racial, cultural, etc.) and eval-uate out-groups according to its standards (Gudykunst andKim 1984). A high level of ethnocentrism can distort themeaning of an individual’s behavior. In turn, a low level ofethnocentrism, termed cultural relativity, has been identi-
394 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
• Attributional Complexity
Perceived InterculturalCommunication Competence
• Ability to Deal With Stress
• Interpersonal Relationships
• Communication Styles
H1 (+) H3 (+)
FIGURE 1Intercultural Disposition, Adaptive Selling Behavior, and Perceived
Intercultural Communication Competence: A Conceptual Model
fied as a critical variable in facilitating effective intercul-tural communication (Porter and Samovar 1991).
The concept of ethnocentrism has been investigated inthe marketing literature from several perspectives. From astrategic perspective, marketers have looked at the effectof ethnocentrism on global planning and decision-making.The ethnocentrism, polycentrism, regiocentrism,geocentrism (EPRG) framework has served as a guidingforce in explaining the structure and perspective of inter-national organizations (Shoham, Rose, and Albaum1995). Consumer behavior researchers have developedand empirically tested the construct of consumerethnocentrism (e.g., Shimp and Sharma 1987). Here,ethnocentrism results from the love for one’s own country,implies an intention to not purchase foreign productsbased on moral concerns rather than economic issues, andrefers to a personal level of prejudice against imports.Since the overall level of consumer ethnocentrism in asocial system is the aggregation of individual tendencies(e.g., Sharma et al. 1995), ethnocentric marketing execu-tives may not realize the implications of these tendenciesin a culturally diverse buyer-seller interaction. In such aninteraction, the success of a culturally diverse interactionbetween a buyer and a seller may be unknowingly inhib-ited on the seller’s part by his or her ethnocentrism (Bushand Ingram 1996).
Attributional complexity. Finally, intercultural disposi-tion includes being accurate in perceiving differences andsimilarities between one’s own culture and the host culture(Gudykunst et al. 1977). Palmer and Pickett (1999) statethat before individuals “can take corrective action regard-ing a discrepancy between a referent and their perfor-mance, they must first generate a theory as to why thediscrepancy occurred” (p. 26). Since intercultural interac-tions may involve more complex culturally embedded ver-bal and nonverbal signals, individuals must be able toaccurately identify more potentially complex causes of theinteraction. Attributional complexity refers to an individ-ual’s tendency to attribute complex causes to other’s be-haviors (Fletcher, Danilovics, Fernandez, Peterson, andReeder 1986). These authors suggest that when formingcausal attributions, “attributionally complex people tendto notice and use information garnered from a behavioralinteraction to a greater extent than attributionally simplepeople” (Fletcher et al. 1986:876). This results, as sug-gested here, in a more accurate explanation of another’sbehavior. This tendency has been shown to be associatedwith lower anxiety and greater effectiveness in intercul-tural interactions (Gudykunst and Kim 1984).
Thus, if a person is adept or accurate at figuring out“what makes people tick,” that person may be better able tocommunicate in an unfamiliar setting such as departing toa foreign country (Stephan and Stephan 1992) or enteringa culturally diverse market/sales territory.
In summary, it is posited that a marketer must have theappropriate intercultural disposition—that is, empathy,worldmindedness, low ethnocentrism, and attributionalcomplexity—to effectively communicate with diversebuyers. The central argument of adaptive selling is that theability to create and modify messages through interactivecommunication with customers will result in better perfor-mance (Spiro and Weitz 1990; Sujan, Weitz, and Kumar1994; Weitz 1978; Weitz et al. 1986). Based on the previ-ous discussion of intercultural communication, it is pos-ited that a key contributor to this adaptability, or “workingsmarter,” in a culturally diverse interaction is theintercultural disposition of a marketing executive. Thus,the following is hypothesized (refer to Figure 1):
Hypothesis 1: There is a positive relationship betweenthe intercultural disposition and adaptive selling be-havior of marketing executives.
Perceived InterculturalCommunication Competence
Perceived intercultural communication competence isan impression that message behavior is appropriate andeffective in a given context (Spitzberg 1991). Appropriate-ness means that the valued rules and norms of either cul-ture in an interaction are not violated significantly. Thus,similar to Williams et al.’s (1990) descriptions, appropri-ateness refers to the communication codes and rules thatare valued in each culture. Effectiveness refers to theaccomplishment of the goal of the interaction.
Extensive research has been conducted in the field ofintercultural communication to define the subdimensionsof perceived intercultural communication competence(e.g., Brislin and Yoshida 1994; Gudykunst 1994). Theseinclude the ability to handle stress, to establish and main-tain a meaningful relationship, and to deal with differentcommunication styles (Hammer et al. 1978). Since a cul-turally diverse experience can be composed of many newor novel situations, an individual must be capable of han-dling the psychological stress that can result when tryingto cope with those differences. The ability to establish andmaintain a relationship is also an important factor whenaddressing someone perceived as culturally different.Here, an individual must not only be comfortable in theinitial phases of relationship development but also be ableto maintain the relationship, such as in a long-term buyer-seller relationship. Finally, an individual must perceivehimself or herself as being able or comfortable with com-municating with a culturally diverse person who may beused to communicating in a different style and/or socialsystem.
Bush et al. / INTERCULTURAL COMMUNICATION 395
Research has shown that the intercultural disposition ofan individual can affect the perceived intercultural com-munication skills or competence of that individual in a cul-turally diverse interaction (cf. Abe and Wiseman 1983;Wiseman et al. 1989). For example, Wiseman et al. (1989)found that more ethnocentric individuals had less under-standing of a Japanese host culture. In a review of the liter-ature on intercultural effectiveness, Hannigan (1990) con-cluded that cross-cultural attitude (e.g., Gudykunst et al.1977) is an important and widely researched factor in pre-dicting cross-cultural success—one of its outcomes beingperceived intercultural communication competence.
Based on the previous discussion, an individual’sintercultural disposition is a key contributor to the successof a culturally diverse interaction. This disposition can in-fluence the perceived intercultural communication com-petence of the individual (refer to Figure 1). Thus, thefollowing is hypothesized:
Hypothesis 2: There is a positive relationship betweenthe intercultural disposition of marketing executivesand their perceived intercultural communicationcompetence.
Finally, the relationship between adaptive selling andintercultural communication competence is examined. In-tuitively, if a marketing executive is adaptive to the buyer-seller interaction, then he or she should be competent in aculturally diverse buyer-seller interaction (see Figure 1).In other words, an adaptive seller should naturally be ableto handle the stress of new situations, be able to establishand maintain a relationship with a buyer from a culturallydiverse background, and be able to deal with differentcommunication styles. Thus, adaptive selling behaviorsshould result in better intercultural communication skills.However, this relationship has not been empirically exam-ined within the context of cultural diversity nor within themarketing literature.
Relevant research has been conducted in the inter-cultural literature that may parallel the concept of adaptiveselling behavior and its impact on perceived interculturalcommunication competence. Cleveland et al. (1960)suggested that a sojourner to a foreign culture must pos-sess flexibility. This is characterized by the ability tochange goals and methods for reaching goals as needed.This is similar to a seller being flexible in the sellingapproach used (Weitz et al. 1986). Thus, the following ishypothesized:
Hypothesis 3: There is a positive relationship betweenthe adaptive selling behavior of marketing execu-tives and their perceived intercultural communica-tion competence.
The target population for this study was defined asmembers of organizations involved in marketing interac-tions with nonretail customers. Our goal was to recruit adiverse sample of marketing executives across variousindustries in a major southern metropolitan city. Membersof a regional sales and marketing association along withexecutive MBA students involved in marketing courses ata major state university in the same city were solicited bythe researchers. The letter stressed that participants bemarketing executives involved in the selling process withnonretail customers and that the issue of cultural diversityamong their clientele was a real concern to them.
A total of 30 letters were sent to marketing managersand professors—serving as key contact points for theirorganization or class. A total of six groups (20% responserate) agreed to participate in the survey. Three groups con-sisted of marketing organizations (51 respondents, 42%)and three consisted of executive MBA marketing classes(71 respondents, 58%). In total, 122 respondents partici-pated in the study. All respondents, including executiveMBAs, were employed in marketing organizations, hadregular interactions with customers, and were concernedabout cultural diversity in their clientele. Fifty-six percentof the respondents listed sales as their occupation, another20 percent stated they were in marketing, and 24 percentstated that they were in marketing management. Themajority of the participants in this study were male(69.7%), married (70.8%), and between 30 and 50 years ofage. More than 80 percent of the respondents were Cauca-sian and most of the respondents were college graduates(85%).
The marketing executives in this study also representeda broad spectrum of industries. Participants stated theywere involved in such industries as health care products(20%), financial planning (18%), business consulting(15%), manufacturing (25%), transportation (7%), com-puter technology (5%), and others (10%). In a recentworldwide survey conducted by Sales and MarketingExecutives International (SMEI) in 1998, 14 percent of themarketing executives surveyed stated they were in finance,13 percent in manufacturing, 2 percent in computer tech-nology, 2 percent in medical, and 1 percent in transporta-tion (SMEI 2000). The marketing executives in the presentstudy are similar in representation to those in the SMEIstudy in some categories (i.e., finance, manufacturing,computer technology). However, our study also representsthe metropolitan location with a higher percentage of mar-keting executives in health care, manufacturing, and trans-portation (U.S. Census Bureau 2000). In sum, our objec-tive was to purposely recruit marketing executives who
396 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
were involved in culturally diverse marketing interactionsacross a variety of industries.
To further validate the selection of participants, severalquestions regarding diversity issues were included in thesurvey. Respondents were asked on a scale of 1 (stronglyagree) to 6 (strongly disagree) their level of agreement tothe following statement: “The issue of cultural diversity inthe workplace is not relevant to me and my organization.”The mean response was 5.01 on a 6-point scale, indicatingthat respondents were indeed concerned about culturaldiversity. Furthermore, respondents were asked how oftenthey interact with culturally diverse customers. Using ascale of 1 (frequently), 2 (occasionally), 3 (rarely), and 4(never), 87.6 percent of the respondents stated that theyfrequently (53.3%) or occasionally (34.3%) interact withdiverse customers. Thus, the relevance of diversity and fre-quent or occasional interaction with diverse customersrevealed that the sample was involved in diverse marketinginteractions.1
Because of the nature and content of this study, two pre-cautionary steps were included in analyzing the data. First,the measures of interest to this study have the potential forsocially desirable responses. To address this issue, a corre-lation analysis between the constructs and social desirabil-ity (i.e., the shortened Marlow-Crowne Scale) was con-ducted (Strahan and Gerbasi 1972). No significantcorrelations at the .05 level were found. Second, the stabil-ity of the responses for this study was also assessedbetween each company or class from which the data werecollected. A series of one-way ANOVAs were conductedon the constructs of interest. No significant difference wasfound between the groups before the data were combinedfor analysis.
To test the hypotheses generated for the present study,measures from the intercultural communication and mar-keting literature were used. Where possible, existingscales, or slightly modified versions of these scales, havebeen used to maintain scale integrity and enhance compar-isons to other studies (e.g., Hensel and Bruner 1992). Spe-cific scale items and their reliability in the present studyare presented in the appendix.
Empathy. Davis (1980, 1983) developed and validateda multidimensional Interpersonal Reactivity Index that fo-cuses on empathetic tendencies. The seven-item Perspec-tive Taking subscale measures the “tendency to adopt thepsychological point of view of others” (Davis 1983:114)and has demonstrated sufficient reliability among men(α = .75) and women (α = .78). The Perspective Takingscale has been used in previous marketing studies. For ex-ample, Spiro and Weitz (1990) achieved a reliability of α =
.77 for this scale. For the present study, the seven itemsconsisting of perspective taking were measured using a 6-point scale, ranging from 1 (describes me very well) to 6(does not describe me well).
Worldmindedness. The Worldmindedness Scale(Sampson and Smith 1957) consists of 32 items in totalthat measure one’s “frame of reference” toward the worldon a variety of issues such as immigration, race relations,economics, and education. Some researchers (e.g., Wise-man et al. 1989) suggest that six items of this scale aremore culture-general and thus more applicable to variouscross-cultural situations. A typical item is, for example, “Itwould be better to be a citizen of the world than of any par-ticular nation” (p. 101). Wiseman et al. (1989) found thatthese items loaded on one factor. All six items were mea-sured on a 6-point scale, ranging from 1 (strongly agree) to6 (strongly disagree).
Ethnocentrism. Ethnocentrism was measured usingitems developed by Stephan and Stephan (1992). Thisscale includes such statements as “Americans have beenvery generous in teaching other people how to do things inmore efficient ways” (p. 92). Reliability for this scale haspreviously been reported to be as high as α = .83. Two re-lated items were also included from Shimp and Sharma’s(1987) revised Patriotism Scale, which was based on ear-lier work by Levinson (1950). These items are the follow-ing: “America may not be perfect, but the American wayhas brought us about as close as human beings can get to aperfect society” and “the main threat to basic American in-stitutions this century has come from the infiltration of for-eign ideas, doctrines and agitators” (p. 287). These itemswere included to reflect the fact that the study is focused onAmerican respondents. All items from both scales weremeasured using a 6-point scale, ranging from 1 (stronglyagree) to 6 (strongly disagree).
Attributional complexity. Five items were used from theFletcher et al. (1986) scale and consist of such statementsas “I really enjoy analyzing the reasons or causes for peo-ple’s behavior.” These items have been used in previousstudies measuring abilities before departing to a foreigncountry (Stephan and Stephan 1992). In addition, thesefive items were chosen because they exhibited the highestcorrelations with the overall scale in two previous studies(Fletcher et al. 1986; Stephan and Stephan 1992).Attributional complexity was measured on a 6-point scale,ranging from 1 (describes me very well) to 6 (does not de-scribe me well).
Adaptive selling. Adaptive selling was measured usingthe ADAPTS Scale developed by Spiro and Weitz (1990).This scale measures the degree to which salespeople per-ceive themselves as “working smarter” (Spiro and Weitz1990; Weitz 1978; Weitz et al. 1986). In other words,
Bush et al. / INTERCULTURAL COMMUNICATION 397
ADAPTS measures the adaptability of salespeople to theircustomers. The reliability of this scale has been well docu-mented in the marketing literature with coefficient alphasranging from .85 (Boorom et al. 1998, Spiro and Weitz1990) to .88 (Sujan et al. 1994). Again, all items for thepresent study were measured on a 6-point scale, rangingfrom 1 (strongly agree) to 6 (strongly disagree).
Previous research has found that adaptive selling iscomposed of adaptive selling beliefs and adaptive sellingbehaviors (Marks, Vorhies, and Badovick 1996). Since thetheoretical foundation of our study focuses on inter-cultural disposition (i.e., a trait) and the resulting skills(i.e., behaviors) of adaptive selling and perceived inter-cultural communication competence, we decided to focuson adaptive selling behaviors. Ten items were used to mea-sure this scale with all model fit indices indicating that thescale was unidimensional (with Comparative Fit Index[CFI], Incremental Fit Index [IFI], and Goodness-of-FitIndex [GFI] above .90 for a one-factor model).
Perceived intercultural communication competence.The Intercultural Communication Competence Scale(Hammer et al. 1978) is a “culture-general” as opposed to“culture-specific” method to assess perceived skills whencommunicating with different cultures (e.g., Harris andMoran 1991). The scale consists of three dimensions:(1) ability to deal with psychological stress (five items),(2) ability to establish and maintain interpersonal relation-ships (five items), and (3) ability to deal with differentcommunication styles (two items). These dimensions havebeen used in previous studies (e.g., Cui and Van Den Berg1991; Hammer et al. 1978) and have been shown to load onthree factors. Respondents were asked to rate themselveson each ability from 1 (high ability) to 6 (low ability). Abil-ity to deal with psychological stress included questions re-garding the respondents’ self-rating on such issues asdealing with “frustration,” “pressure to conform,” “inter-personal conflict,” and “anxiety” in communicating withdifferent cultures. Ability to establish interpersonal rela-tionships included such items as the ability to “initiate in-teraction with a stranger” and “develop satisfyinginterpersonal relationships with others.” The final dimen-sion included the ability to deal with “different communi-cation styles” and to “accurately understand the feelings ofanother person.”
The finalized questionnaire was pretested among 25graduate business students. Adjustments were made to thewording and length based on their suggestions. The ordersequence of some of the items was also revised.
The reliability of each measure was assessed via coeffi-cient alphas (Cronbach 1951). With the exception ofworldmindedness (α = .63), which was included for con-ceptual reasons despite its low reliability, the scales used inthis study were at or above the .70 level (Nunnally 1978).
In summary, the specific scales used in this study werebased on previous research, demonstrate unidimen-sionality and multidimensionality where appropriate, andshow sufficient reliability.2
Model Parameter Estimatesand Goodness of Fit
All hypotheses were tested in a structural-equationsmodel (see Figure 2). A structural-equations analysis wasemployed because it allows for the simultaneous estima-tion of a series of interdependent relationships and allowsfor an assessment of the overall fit of a model (Hair,Anderson, Tatham, and Black 1992). The correlationmatrix for model estimation is presented in Table 1.Intercultural disposition and intercultural competencewere examined as multidimensional constructs. Empathy,worldmindedness, ethnocentrism, and attributional com-plexity were analyzed as indicators of intercultural dispo-sition, while the ability to deal with stress, to establish andmaintain interpersonal relationships, and to manage dif-ferent communication styles were used as indicators ofintercultural competence. Summated scales were used foreach of the indicators of intercultural disposition andintercultural competence. Adaptive selling was repre-sented by a single-item indicator, with each of the 10 itemsgiven equal weight in forming the summated scale. Theerror term of the indicator was set to one minus the reliabil-ity multiplied by the observed variance to adjust for themeasurement error in the construct (Price, Arnould, andTierney 1995; Shoham, Rose, and Kahle 1998).
Overall model fit was acceptable, as evidenced by anonsignificant chi-square value (χ2 = 23.09 with 18degrees of freedom, p = .19), which indicates the lack of asignificant discrepancy between the hypothesized modeland the observed results. The GFI and Bollen’s IFI werealso above .90 (GFI = .942, IFI = .934 ), further indicatingan acceptable model fit (Bollen 1989).
Hypothesis 1, which posited that intercultural disposi-tion and adaptive selling would be positively related, wasnot supported. Intercultural disposition and adaptive sell-ing were not significantly related (γ = .301, t = 1.63, p >.05). Hypotheses 2 and 3, which directionally examine theinfluence of intercultural disposition and adaptive selling,respectively, on intercultural competence, were supported.Both intercultural disposition (γ = .522, t = 1.79, p < .05)and adaptive selling (β = .352, t = 2.03, p < .05) were posi-tively related to intercultural competence in a one-tailedtest. Furthermore, 51 percent of the variance is explainedby intercultural communication competence. Thus,intercultural disposition and adaptive selling appear to be
398 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
relatively independent factors, both of which contribute tointercultural competence.
The overall purpose of this study was to investigate therole of intercultural disposition and adaptive selling indeveloping perceived intercultural communication com-petence. The results show that intercultural dispositionand adaptive selling both affect the perceived interculturalcommunication competence of marketing executives.However, adaptive selling and intercultural dispositionseem to be separate unrelated constructs. Thus, programsdesigned to increase the adaptiveness of marketing
executives alone may not increase the effectiveness of aculturally diverse buyer-seller interaction. It should be rec-ognized that the intercultural disposition or “cross-cul-tural attitude” of marketing executives may also have aneffect on this relationship. Marketers may perceive them-selves as quite adaptive to their customers. However, whileadaptive selling within a marketing executive’s own cul-ture has been shown to affect outcomes, this constructalone may not be sufficient to ensure successfulintercultural interactions.
The findings of our study have several implications formarketing managers. First, managers must be able torecruit, retain, and manage culturally diverse marketingexecutives. Second, managers must also recognize thattheir existing employees will be interacting with more
Bush et al. / INTERCULTURAL COMMUNICATION 399
4) .352 (2.03)
.522 (1.79).301 (1.63)
FIGURE 2Intercultural Disposition, Adaptive Selling Behavior, and Perceived
Intercultural Communication Competence: Empirical Findings
NOTE: All parameter estimates are standardized. t-values are given in parentheses.*No t-values are reported for Worldmindedness and Stress because these items were used to set the metric for the construct.
* Significant at the .05 level. ** Significant at the .01 level.
diverse coworkers and customers. To address these twoissues, selection and training methods are needed thatfocus on recognizing and developing intercultural skills.
The importance of training for adaptive selling has beenrecognized by academic researchers and practitioners. Intheir framework for adaptive selling, Weitz et al. (1986)incorporated training as a key variable to influence theknowledge and skills of marketers to practice adaptiveselling. As such, practitioners have used methods of train-ing that focus on adaptation. For example, the communica-tion style matrix (Merrill and Reid 1981) has been used totrain marketers to classify customers on the basis of threecommunication style dimensions: (1) responsiveness,(2) assertiveness, and (3) flexibility. However, marketersmust also be trained to recognize and interact with custom-ers’various intercultural communication styles. Given thecurrent trends of an increasingly diverse workforce,intercultural communication training may help marketersgain a competitive advantage in understanding culturallydiverse markets.
Selecting the appropriate intercultural training pro-gram and recruiting new employees require methods forassessing marketing executives’ intercultural dispositionand perceived intercultural communication competence.The measures used in this study can provide an initial start-ing point in identifying individuals who possess thesecapabilities. Other methods of assessment include per-sonal inventories, interview questions, and simulationgames that help identify the reactions of individuals todiverse cultures (cf. Bush and Ingram 2001; Cushner andBrislin 1996; Shirts 1973).
LIMITATIONS AND DIRECTIONSFOR FUTURE RESEARCH
As with any study, there are limitations to the resultsfrom this work. First, external validity may be somewhatlessened because of the use of marketing professionalsfrom executive MBA courses and from actual corpora-tions. However, individuals were solicited from a broadbase of industries who were concerned about the issue ofdiversity in their clientele as well as within their own orga-nization. Second, the findings of this study are based onself-reported data. However, the constructs of interest tothis study were not significantly correlated with socialdesirability. Third, the reliability for the WorldmindednessScale was relatively low, which may have attenuated someof the relationships reported. Fourth, the results of thisstudy could vary with such issues as individuals’ experi-ence and travel abroad. While beyond the scope of thepresent study, future research could examine these issues.
The constructs selected for this study are not exhaustivebut rather serve as a starting point for relating adaptiveselling concepts to intercultural domains. For example,
since “other variables such as allocentrism and dogmatismmay be related to cultural ethnocentrism” (Sharma et al.1995:35) and “cultural similarity, historical associationsbetween countries, and present political-economic rela-tions between countries [and cultures] may moderate theeffect of consumer-ethnocentric tendencies” (p. 35), theconcept of intercultural disposition provides a plethora ofresearch opportunities. Furthermore, this study has taken aculture-general approach, whereas more investigationneeds to be done with specific cultures and ethnicities.Such issues as social distance, distrust, and culture shockwhen interacting with a specific cultural or ethnic group ofbuyers or sellers are viable avenues to pursue.
The area of intercultural communication in marketingis ripe for future research. One research path involves theidentification of individuals who may or may not be com-petent in dealing with culturally diverse individuals. In thepresent study, intercultural disposition and competencewere investigated. Individuals who scored well on theseconstructs may be better able to deal with cultural differ-ences. It would also be of interest to investigate how theseconstructs relate to other variables such as relationshiptrust, performance, turnover, and so on. In other words, arecertain individuals more predisposed to successful inter-action with culturally diverse individuals than others?Such findings could affect the type of diversity recruitmentand training needed in marketing organizations.
Within the framework of adaptive selling, futureresearch could investigate how intercultural dispositionand competence ultimately affect performance. His-torically, however, research results that provide clear link-ages between adaptive selling skills and performance out-comes have been rare. Goolsby, Lagace, and Boorom(1992) only found one relationship between the psycho-logical adaptiveness trait of self-monitoring and perfor-mance. Those authors concluded that “the relationshipbetween adaptive selling and . . . performance is likely tobe moderated by factors such as the nature of the sales situ-ation and/or the nature of the product itself” (p. 62). Whenthe buyer-seller situation crosses cultural boundaries,intercultural disposition may be one of the factors that hasmediated the effect of adaptiveness on subsequentperformance.
Another research path concerns the customer’s evalua-tion of the marketing executive’s intercultural skills. Simi-lar arguments can be found in the intercultural literaturethat advocate investigating the host culture’s reactions toforeigners/sojourners into their culture (Gudykunst 1994).Thus, future research can be conducted by investigatingbuyers’ perceptions of sellers’ intercultural skills.
Investigating various types of training programs fordeveloping intercultural competence could also providehelpful insights. Such programs as exposure to differentethnic neighborhoods (i.e., market territories), movies thatemphasize cultural differences, classes on cross-cultural
400 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
communication, and simulation exercises could be investi-gated and compared. Long-term study is also needed totrack the overall effectiveness of such training programson such issues as performance, retention, and upwardmobility of culturally diverse employees within the mar-keting organization.
Finally, measures more specific to the marketing envi-ronment could be developed. Some of the measures usedin this study were borrowed from the educational andcross-cultural communication disciplines. This may havebeen the cause of one of the measures’ (i.e., worldminded-ness) somewhat lower reliability. Insights from the presentstudy can be used to develop better measures to gauge theintercultural skills of marketing executives.
While marketing academics and practitioners recog-nize the importance of adapting to the customer, adaptingto a culturally diverse customer is a relatively new issue.However, more traditional marketing functions are evolv-ing toward relationship management. According toJohnston, Lewin, and Spekman (1999), the
changing global economic and business environ-ment is forcing firms to move increasingly closer totheir exchange partners, to form international alli-ances, and to participate in complex multifirm/mul-tinational networks. . . . Indeed, many practitionersbelieve that a company’s competitive prowess nolonger depends upon the company itself, but on itsability to establish and maintain high-quality busi-ness partnerships. (P. 269)
Thus, traditional forms of communication in marketingmust be viewed from an integrated marketing perspective,where highly personalized, targeted communication be-comes a key focus in the development of relationships.When these relationships cross cultural or ethnic bound-aries, the necessity of training for adaptive, interculturalcommunication becomes more apparent.
In conclusion, successful relationship managementskills should be emphasized in today’s culturally diverseenvironment. People possessing these skills will becomevaluable assets to the firm (Thorelli 1986; Webster 1992).According to Gronroos (1999), “A true transition toward arelationship marketing strategy requires a focus onresources and competencies in the relationship” (p. 334).One of those resources are the communication competen-cies exhibited by the marketing organization. As globalcompetition and cultural diversity in the workplace con-tinue to increase, people possessing intercultural commu-nication competence will be of extreme importance inmanaging relationships that cross cultural boundaries.
APPENDIXScale Items and Reliabilities for Study
Empathic Tendency—Perspective Taking ( = .71)
Range: 1 (describes me very well) to 6 (does not describe mewell)
1. I believe that there are two sides to every question andtry to look at them both.
2. I try to look at everybody’s side of a disagreement be-fore I make a decision.
3. When I’m upset at someone, I usually try to “put my-self in his or her shoes’ for a while.
4. If I’m sure I’m right about something, I don’t wastemuch time listening to other people’s arguments. (R)
5. I sometimes find it difficult to see things from the“other guy’s” point of view. (R)
6. I sometimes try to understand my friends better byimagining how this looks from their perspective.
7. Before criticizing somebody, I try to imagine how Iwould feel if I were in their place.
Worldmindedness ( = .63)
Range: 1 (strongly agree) to 6 (strongly disagree)
1. It would be better to be a citizen of the world than ofany particular nation.
2. Our responsibility to people of other races ought to beas great as our responsibility to people of our ownarea.
3. Any healthy individual, regardless or race or religion,should be allowed to live wherever he or she wants toin the world.
4. Our schools should teach the history of the worldrather than of our own nation.
5. Our country should permit the immigration of foreignpeoples even if it lowers our standard of living.
Ethnocentrism ( = .74)
Range: 1 (strongly agree) to 6 (strongly disagree)
1. In many countries, people do not place a high value onhuman life—to them, life is cheap.
2. English should be accepted as the international lan-guage of communication.
3. Americans have been very generous in teaching otherpeople how to do things in more efficient ways.
4. Primitive people have unsophisticated social and po-litical systems.
5. Minority groups within a country should conform tothe customs and values of the majority.
Bush et al. / INTERCULTURAL COMMUNICATION 401
1. America may not be perfect, but the American wayhas brought us about as close as human beings can getto a perfect society.
2. The main threat to basic American institutions duringthis century has come from the infiltration of foreignideas, doctrines, and agitators.
Attributional Complexity ( = .78)
Range: 1 (describes me very well) to 6 (does not describe mewell)
1. I really enjoy analyzing the reasons or causes of peo-ple’s behavior.
2. I am very interested in how my own thinking workswhen I make judgments about people or attach causesto their behavior.
3. To understand a person’s personality or behavior, Ihave found it important to know how that person’s at-titudes, beliefs, and character traits fit together.
4. I think a lot about the influences that society has onother people.
5. I enjoy learning about other cultures through readingand movies.
Adaptive Selling Behaviors ( = .88)
Range: 1 (strongly agree) to 6 (strongly disagree)
1. When I feel that my sales approach is not working, Ican easily change to another approach.
2. I like to experiment with different sales approaches.3. I am very flexible in the selling approach I use.4. I don’t change my approach from one customer to an-
other. (R)5. I can easily use a wide variety of selling approaches.6. It is easy for me to modify my sales presentation if the
situation calls for it.7. Basically, I use the same approach with most custom-
ers. (R)8. I vary my sales style from situation to situation.9. I try to understand how one customer differs from an-
other.10. I feel confident that I can effectively change my
planned presentation when necessary.
Perceived InterculturalCommunication Competence
Range: 1 (high ability) to 6 (low ability)
The following is a list of qualities that have been identified asbeing associated with success in sales encounters with peoplefrom different cultural backgrounds. Please rate your:
Intercultural Communication Stress Skills (α = .82):
1. ability to deal with interpersonal conflict2. ability to deal with stress3. ability to deal with social alienation4. ability to deal with anxiety5. ability to deal with communication misunderstand-
ings between yourself and others
Intercultural Communication Relationship Skills (α = .75):
1. ability to initiate interaction with a stranger2. ability to develop satisfying interpersonal relation-
ships with others3. ability to maintain satisfying interpersonal relation-
ships with others4. ability to accurately understand the feelings of an-
other person5. ability to empathize with another person
Intercultural Communication Style Skills (Kuder-Richardson =.73):
1. ability to deal with different communication styles2. ability to deal with different social systems
NOTE: (R) = reverse coded.
1. The model was also run excluding the 12.4 percent of respondentswho indicated that they rarely (7.7%) or never (4.7%) interact with di-verse customers. This analysis produced similar results to those using thefull sample with no changes in the significance of any of the paths and allfit indices suggesting an acceptable model fit.
2. All constructs were examined via factor loadings and item-to-totalcorrelations. Any item-to-total correlation that fell below .30 was de-leted. Subsequent confirmatory factor analysis revealed that all first-order scales were unidimensional, with all constructs fitting adequatelyon a one-factor model. The Comparative Fit Index (CFI), the IncrementalFit Index (IFI), and the Goodness-of-Fit Index (GFI) for all constructs inthis study were above .90 for a one-factor model. More details regardingreliability and validity of the measures are available from the authorsupon request.
Abe, H. and R. Wiseman. 1983. “A Cross-Cultural Confirmation of theDimensions of Intercultural Effectiveness.” International Journal ofIntercultural Relations 7 (1): 53-67.
Anderson, J. and J. Narus. 1990. “A Model of Distributor Firm and Manu-facturer Firm Working Relationships.” Journal of Marketing 54 (Jan-uary): 42-58.
Bertagnoli, Lisa. 2001. “Culture Club: Executives Get Boost From Di-versity Recognition.” Marketing News, January 29, p. 9.
402 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
Bhawuk, Dharm P. S. and Harry C. Triandis. 1996. “The Role of CultureTheory in the Study of Culture and Intercultural Training.” In Hand-book of Intercultural Training. Eds. Dan Landis and Rabi Bhagat.Thousand Oaks, CA: Sage, 17-34.
Bollen, Kenneth A. 1989. Structural Equations With Latent Variables.New York: John Wiley.
Boorom, Michael L., Jerry R. Goolsby, and Rosemary P. Ramsey. 1998.“Relational Communication Traits and Their Effect on Adaptivenessand Sales Performance.” Journal of the Academy of Marketing Sci-ence 26 (1): 16-30.
Bradford, Lisa, Mike Allen, and Kevin Beisser. 1997. “An Evaluation andMeta-Analysis of Intercultural Communication Competence Re-search.” Working Paper. University of Wisconsin–Milwaukee.
Brislin, R. and T. Yoshida. 1994. Intercultural Communication Training:An Introduction. Thousand Oaks, CA: Sage.
Bush, Victoria D. and Thomas N. Ingram. 1996. “Adapting to DiverseCustomers: A Training Matrix for International Marketers.” Indus-trial Marketing Management 25 (5): 373-383.
and . 2001. “Building and Assessing Cultural DiversitySkills: Implications for Sales Training.” Industrial Marketing Man-agement 30 (1): 65-76.
Carlson, J. and K. Widaman. 1988. “The Effect of Study Abroad DuringCollege on Attitudes Toward Other Cultures.” International Journalof Intercultural Relations 12:1-17.
Cleveland, H., G. Mangone, and J. Adams. 1960. The Overseas Ameri-cans—A Report on Americans Abroad. New York: McGraw-Hill.
Comer, Lucette B. and Tanya Drollinger. 1999. “Active Empathetic Lis-tening and Selling Success: A Conceptual Framework.” Journal ofPersonal Selling and Sales Management 19 (Winter): 15-29.
and Marvin A. Jolson. 1991. “Perceptions of Gender StereotypicBehavior: An Exploratory Study of Women in Selling.” Journal ofPersonal Selling and Sales Management 11 (Winter): 17-32.
and J.A.F. Nicholls. 2000. “Communication Between HispanicSalespeople and Their Customers: A First Look.” Journal of PersonalSelling and Sales Management 20 (Summer): 121-127.
, , and Leslie J. Vermillion. 1998. “Diversity in the SalesForce: Problems and Challenges.” Journal of Personal Selling andSales Management 18 (Fall): 1-20.
Cronbach, L. 1951. “Coefficient Alpha and the Internal Structure ofTests.” Psychometrika 16 (September): 297-334.
Cui, G. and S. Van Den Berg. 1991. “Testing the Construct Validity ofIntercultural Effectiveness.” International Journal of InterculturalRelations 15:227-241.
Cushner, Kenneth and Richard W. Brislin. 1996. Intercultural Interac-tions: A Practical Guide. Thousand Oaks, CA: Sage.
Davis, M. H. 1980. “A Multidimensional Approach to Individual Differ-ences in Empathy.” JSAS Catalog of Selected Documents in Psychol-ogy 10:85.
. 1983. “The Effects of Dispositional Empathy on Emotional Re-actions and Helping: A Multidimensional Approach.” Journal of Per-sonality 51:114-174.
Deshpande, Rohit and Frederick Webster. 1989. “Organizational Cultureand Marketing: Defining the Research Agenda.” Journal of Mar-keting 53 (January): 3-15.
Dwyer, Sean, Orlando Richard, and C. David Shepherd. 1998. “An Ex-ploratory Study of Gender and Age Matching in the Salesperson-Prospective Customer Dyad: Testing Similarity-Performance Predic-tions.” Journal of Personal Selling and Sales Management 18 (Fall):55-69.
Fletcher, G., P. Danilovics, G. Fernandez, D. Peterson, and G. Reeder.1986. “Attributional Complexity: An Individual Differences Mea-sure.” Journal of Personality and Social Psychology 51:875-884.
Goolsby, Jerry, Rosemary Lagace, and Michael Boorom. 1992. “Psycho-logical Adaptiveness and Sales Performance.” Journal of PersonalSelling and Sales Management 12 (Spring): 51-63.
Gronroos, Christian. 1999. “Relationship Marketing: Challenges for theOrganization.” Journal of Business Research 46 (November): 327-335.
and Y. Y. Kim. 1984. Communicating With Strangers: An Ap-proach to Intercultural Communication. New York: Random House.
, R. L. Wiseman, and M. R. Hammer. 1977. “Determinants of aSojourner’s Attitudinal Satisfaction: A Path Model.” In Communica-tion Yearbook. Ed. B. Ruben. New Brunswick, NJ: Transaction Pub-lishing, 415-425.
Hair, J. F., R. E. Anderson, R. L. Tatham, and W. C. Black. 1992.Multivariate Data Analysis. New York: Macmillan.
Hammer, M., W. Gudykunst, and L. Wiseman. 1978. “Dimensions ofIntercultural Effectiveness: An Exploratory Study.” InternationalJournal of Intercultural Relations 2 (4): 383-393.
Hannigan, Terence P. 1990. “Traits, Attitudes, and Skills That Are Re-lated to Intercultural Effectiveness and Their Implications for Cross-Cultural Training: A Review of the Literature.” International Journalof Intercultural Relations 14:89-111.
Harris, Phillip R. and Robert T. Moran. 1991. Managing Cultural Differ-ences. Houston, TX: Gulf.
Hensel, Paul and Gordon Bruner. 1992. “Scaling and Measurement:Multi-Item Scaled Measures in Sales Related Research.” Journal ofPersonal Selling and Sales Management 12 (3): 77-82.
Honeycutt, Earl and John Ford. 1995. “Guidelines for Managing an Inter-national Sales Force.” Industrial Marketing Management 24 (2):135-144.
Ingram, Thomas N., Charles H. Schwepker, Jr., and Don Hutson. 1992.“Why Salespeople Fail.” Industrial Marketing Management 21 (Au-gust): 225-230.
Johnston, Wesley J., Jeffrey E. Lewin, and Robert E. Spekman. 1999.“International Industrial Marketing Interactions: Dyadic and Net-work Perspectives.” Journal of Business Research 46 (November):259-271.
Kang, Jikyeong and Julie Hillary. 1998. “Older Salespeople’s Role in Re-tail Sales Encounters.” Journal of Personal Selling and Sales Man-agement 18 (Fall): 39-47.
Levinson, Daniel J. 1950. “The Study of Ethnocentric Ideology.” In TheAuthoritarian Personality. Eds. T. W. Adorno, Else Frenkel-Brunswik, Daniel J. Levinson, and R. Nevitt Sanford. New York:Harper & Row, 108-163.
Levy, Michael and Arun Sharma. 1994. “Adaptive Selling: The Role ofGender, Age, Sales Experience, and Education.” Journal of BusinessResearch 31:39-47.
Loden, M. and J. Rosener. 1991. Workforce America! Managing Em-ployee Diversity as a Vital Resource. Homewood, IL: Irwin.
Lustig, Myron and Jolene Koester. 1993. Intercultural Competence. NewYork: HarperCollins.
Marks, Ronald, Douglas W. Vorhies, and Gordon J. Badovick. 1996. “APsychometric Evaluation of the ADAPTS Scale: A Critique and Rec-ommendations.” Journal of Personal Selling and Sales Management16 (4): 53-65.
McBane, Donald A. 1995. “Empathy and the Salesperson: A Multidi-mensional Perspective.” Psychology and Marketing 12 (July): 349-370.
Merrill, David and Roger Reid. 1981. Personal Styles and Effective Per-formance. Radnor, PA: Chilton.
Mohr, J. and J. Nevin. 1990. “Communication Strategies in MarketingChannels: A Theoretical Perspective.” Journal of Marketing 54 (Oc-tober): 36-51.
Montgomery, David. 1991. “Understanding the Japanese as Customers,Competitors, and Collaborators.” Japan and the World Economy 3(1): 61-91.
and Allan M. Weiss. 1991. “Managerial Preferences for StrategicAlliance Attributes.” Research Paper 1134. Stanford University,Graduate School of Business, Stanford, CA.
Nunnally, J. 1978. Psychometric Theory. 2d ed. New York: McGraw-Hill.
Palmer, Timothy and Gregory Pickett. 1999. “The Role of MentalModels in Control Theory: Understanding Cognitive Factors Influ-encing the Behaviors of Salespeople.” Journal of Marketing Theoryand Practice 7 (1): 17-29.
Porter, R. and L. Samovar. 1991. “Basic Principles of Intercultural Com-munication.” In Intercultural Communication. Eds. L. Samovar andR. Porter. Belmont, CA: Wadsworth, 5-22.
Bush et al. / INTERCULTURAL COMMUNICATION 403
Price, Linda L., Eric J. Arnould, and Patrick Tierney. 1995. “Going to Ex-tremes: Managing Service Encounters and Assessing Provider Per-formance.” Journal of Marketing 59 (2): 83-97.
Ruben, B. 1976. “Assessing Communication Competency for Intercul-tural Adaptation.” Group and Organizational Studies 1:334-354.
Sales and Marketing Executives International, Sales and Marketing Man-agement Resource Center. 2000. “Sales and Marketing Compensa-tion and Trends Survey” [Online]. Available: http://www.sell. org
Sampson, D. and H. Smith. 1957. “A Scale to Measure World-MindedAttitudes.” Journal of Social Psychology 45:99-106.
Schuster, Camille P. and Jeffrey E. Danes. 1986. “Asking Questions:Some Characteristics of Successful Sales Encounters.” Journal ofPersonal Selling and Sales Management 6 (May): 17-27.
Seal, K. 1991. “Know thy Neighbor.” American Way, August 15, p. 34.Sharma, Subhash, Terence Shimp, and Jeongshin Shin. 1995. “Consumer
Ethnocentrism: A Test of Antecedents and Moderators.” Journal ofthe Academy of Marketing Science 23 (1): 26-37.
Shimp, Terence A. and Subhash Sharma. 1987. “ConsumerEthnocentrism: Construction and Validation of the CETSCALE.”Journal of Marketing Research 24 (August): 280-289.
Shirts, Gary. 1973. BAFA BAFA. La Jolla, CA: Simili II.Shoham, Aviv, Gregory M. Rose, and Gerald Albaum. 1995. “Export
Motives, Psychological Distance, and the EPRG Framework.” Jour-nal of Global Marketing 8 (3/4): 9-38.
, , and Lynn R. Kahle. 1998. “Risky Sports Participation:From Intention to Action.” Journal of the Academy of Marketing Sci-ence 26 (4): 307-321.
Spiro, R. and B. Weitz. 1990. “Adaptive Selling: Conceptualization,Measurement and Nomological Validity.” Journal of Marketing Re-search 27 (February): 61-69.
Spitzberg, Brian H. 1991. “Intercultural Communication Competence”In Intercultural Communication. Eds. L. Samovar and R. Porter.Belmont, CA: Wadsworth, 353-365.
and William R. Cupach. 1984. Interpersonal CommunicationCompetence. Beverly Hills, CA: Sage.
Stephan, Cookie and Walter Stephan. 1992. “Reducing InterculturalAnxiety Through Intercultural Contact.” International Journal ofIntercultural Relations 16:89-106.
Strahan, Robert and Kathleen Gerbasi. 1972. “Short, Homogeneous Ver-sions of the Marlow-Crowne Social Desirability Scale.” Journal ofClinical Psychology 28:191-193.
Sujan, Harish, Barton Weitz, and Nirmalya Kumar. 1994. “Learning Ori-entation, Working Smart, and Effective Selling.” Journal of Mar-keting 58 (July): 39-52.
Thorelli, Hans. 1986. “Networks: Between Markets and Hierarchies.”Strategic Management Journal 7:37-51.
U.S. Census Bureau, CenStats DataWell. 2000. “County Business Pat-terns, Shelby County, TN, Major Industry Group” [Online]. Avail-able: http://tier2.census.gov/cgi-win/cbp/sector.exe
Webster, Frederick. 1992. “The Changing Role of Marketing in the Cor-poration.” Journal of Marketing 56 (4): 1-17.
and Rohit Deshpande. 1991. “Analyzing Corporate Cultures inApproaching the Global Marketplace.” Report No. 90-111. Mar-keting Science Institute, Cambridge, MA.
Weitz, Barton A. 1978. “The Relationship Between Salesperson Perfor-mance and Understanding of Customer Decision Making.” Journalof Marketing Research 15 (November): 501-516.
, and Sandy Jap. 1995. “Relationship Marketing in DistributionChannels.” Journal of the Academy of Marketing Science 23 (4): 305-320.
, Harish Sujan, and Mita Sujan. 1986. “Knowledge, Motivationand Adaptive Behavior: A Framework for Improving Selling Effec-tiveness.” Journal of Marketing 50 (October): 174-191.
Wentling, Rose Mary. 2000. “Current Status of Diversity Initiatives inSelected Multinational Corporations.” Human Resource Develop-ment Quarterly 11 (1): 35-61.
Williams, Kaylene C. and Rosann L. Spiro. 1985. “Communication Stylein the Salesperson-Customer Dyad.” Journal of Marketing Research22 (November): 434-442.
, , and Leslie M. Fine. 1990. “The Customer-SalespersonDyad: An Interaction/Communication Model and Review.” Journalof Personal Selling and Sales Management 10 (Summer): 29-43.
Wiseman, R. L., M. R. Hammer, and H. Nishida. 1989. “Predictors ofIntercultural Communication Competence.” International Journal ofIntercultural Relations 13:101-357.
ABOUT THE AUTHORS
Victoria D. Bush (Ph.D., University of Memphis) is an associateprofessor of marketing at the University of Mississippi. Her re-search has appeared in such journals as the Journal of Advertis-ing, the Journal of Advertising Research, Industrial MarketingManagement, the Journal of Public Policy and Marketing, theJournal of Business Ethics, and the Journal of Services Mar-keting. Her research interests are in diversity, advertising, andethics.
Gregory M. Rose (Ph.D., University of Oregon) is an associateprofessor of marketing at the University of Mississippi. His re-search interests include consumer socialization and cross-cul-tural consumer behavior. He has published or has forthcomingarticles in the Journal of Consumer Research, the Journal of theAcademy of Marketing Science, the Journal of Business Re-search, the Journal of Consumer Psychology, the Journal of Ad-vertising, the Journal of Marketing, and other journals andproceedings.
Faye Gilbert (Ph.D., University of North Texas) is an associateprofessor of marketing at the University of Mississippi. She haspublished in the Journal of Business Research, Psychology andMarketing, the Journal of Health Care Marketing, the Journal ofResearch in Pharmaceutical Economics, the Journal of AppliedBusiness Research, the Journal of Marketing Management, theJournal of Marketing Theory and Practice, and the Journal ofMarketing Education, among others. Her work emphasizes theapplication of consumer behavior theory to health care and tochannel relationships.
Thomas N. Ingram (Ph.D., Georgia State University) is a pro-fessor of marketing at Colorado State University. He has beenhonored as the Marketing Educator of the Year by Sales and Mar-keting Executives International (SMEI) and as a recipient of theMu Kappa Tau National Marketing Honor Society RecognitionAward for Outstanding Scholarly Contributions to the Sales Dis-cipline. He has served as the editor of the Journal of PersonalSelling and Sales Management and is the current editor of theJournal of Marketing Theory and Practice. His primary researchis in personal selling and sales management. His work has ap-peared in the Journal of Marketing, the Journal of Marketing Re-search, the Journal of Personal Selling and Sales Management,and the Journal of the Academy of Marketing Science, amongothers. He is the coauthor of three textbooks: ProfessionalSelling: A Trust-Based Approach, Sales Management: Analysisand Decision-Making, and Marketing: Principles and Perspectives.
404 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001Summers / RESEARCH IN MARKETING
Guidelines for Conducting Researchand Publishing in Marketing:From Conceptualization Throughthe Review Process
John O. SummersIndiana University
A primary mission of institutions of higher learning is thegeneration and dissemination of knowledge. The low ac-ceptance rates at the leading research journals in market-ing, typically in the single digits to low teens, suggests theneed to increase the quality of the research manuscriptsproduced. This article presents a set of guidelines for re-searchers aspiring to do scholarly research in marketing.Discussed are issues such as developing the necessary re-search skills, conceptualizing the study, constructing theresearch design, writing the manuscript, and respondingto reviewers. Also presented are the author’s personal ob-servations concerning the current state of research inmarketing.
This article is intended for doctoral students and thoseresearchers who are beginning or are early in their careersand would like to increase their journal acceptance rates.The experienced author with several major publicationsand years of reviewing experience will find little, if any-thing, “new” to them. What follows are the author’s reflec-tions on more than a quarter century of guiding doctoralstudents and reviewing for, and publishing in, some of theleading journals in marketing. The author’s remarks pri-marily relate to research that involves the collection andanalysis of primary data (e.g., case studies, surveys, andexperiments). Not addressed are such things as reviewpapers, theory development not based on empiricalresearch, and quantitative marketing models.
Manuscript Acceptance Ratesat Leading Marketing Journals:From Single Digit to Low Teens
The acceptance rate at the leading research journals iscurrently averaging around 10 percent. Because editorsare limited in the number of pages they can have in eachissue, a journal’s acceptance rate is constrained by thenumber of manuscripts submitted and the average lengthof the manuscripts accepted. Hence, as the overall qualityof the manuscripts received by a journal increases overtime, its standards for acceptance also rise.
For most top journals, there isn’t a dramatic drop inquality between the top 10 percent of manuscripts receivedand the next best 10 percent, and most of the manuscriptssubmitted to the leading journals are reasonably well-done. About 80 percent of the manuscripts submitted arerejected on the initial round of reviews. There are severalbasic reasons for rejecting manuscripts reporting on em-pirical studies. These include the following:
1. The research questions being investigated arenot very interesting (e.g., studies that are mainlydescriptive and lack theoretical implications).
2. The research, although well executed, does notappear to make a sufficiently large contributionto the literature (e.g., the study largely replicatespast research with minor modifications).
3. The conceptual framework is not well developed(e.g., lacks precise conceptual definitions of theconstructs and/or compelling theoretical ratio-nale for the hypotheses).
4. The methodology is seriously flawed (e.g., thesample is inappropriate for the research ques-tion, the validity of one or more key measures is
suspect, and/or the experiment lacks experimen-tal realism).
5. The writing is so confused that an invitation torevise and resubmit is considered unlikely to re-sult in an acceptable manuscript.
For a detailed discussion of the weaknesses in manu-scripts cited by the reviewers of one leading journal alongwith some guideposts for authors, see Varadarajan (1996).
To be published in a respected peer-reviewed journal, astudy must be judged as meeting the currently acceptedstandards for scholarly research. Moreover, the study mustbe judged as more worthy than others competing for thesame journal space. What should researchers do toincrease the chances that their studies will make a signifi-cant contribution to marketing knowledge and be amongthose that are eventually published by one of the leadingresearch journals? Answering this question is the majorfocus of this article.
SCHOLARLY RESEARCH ONSUBSTANTIVE ISSUES IN MARKETING
This section presents a set of 12 guidelines for research-ers aspiring to do scholarly research in marketing. Theseguidelines deal with developing the necessary set ofresearch skills and the research process.
Develop a Broad Setof Methodological Skills
Developing a broad set of methodological skills (e.g.,qualitative research methods, survey research methodol-ogy, and experimental design) is critical to becoming aproductive researcher. Those with a limited set of method-ological tools are restricted in what they can study andwhat they can learn from their research. For example,someone with weak or no training in qualitative researchmethods is very limited with regard to developinggrounded theory in his or her research area of interest, andresearchers without a background in experimental designare likely to use surveys to test causal hypotheses.Developing a broad set of methodological skills early inone’s career provides long-term benefits because one canrely on this same set of skills for many years. Many of theresearch techniques used today were developed severaldecades ago. For example, much of the most importantwork on reliability and validity was published during the1950s and 1960s.
Learn to Be a CriticalReader of the Literature
It is important to become practiced in reading the litera-ture in a critical manner. When researchers take an
“accepting point of view” in reading the literature andfocus on the conclusions of these studies, it will seem tothem as if everything has been done, and they will feel dis-appointed that they had not thought to do these studiesfirst. It is only when researchers look for flaws and/or limi-tations in the research they read that they begin to developideas for building on this research. For example, withregard to the conceptual framework, readers should con-cern themselves with whether the conceptual definitionsare sufficiently unambiguous and whether the theoreticalrationale provided for each of the hypotheses is convinc-ing. With regard to survey research methodology, theymight consider whether there is a serious problem withshared method variance and/or whether the measures usedvalidly capture the constructs of interest. The limitationsidentified in existing research alert the researcher toopportunities for making contributions to the research areaof interest.
Focus on DevelopingHypotheses to Be Tested
As researchers start reading the literature, it is impor-tant that they begin thinking about identifying the hypoth-eses they might want to test. This will help them developsome structure for their conceptual frameworks and con-struct boundaries for their empirical studies. This, in turn,will allow them to determine which articles in their generalarea of interest are most central to the empirical study theyplan to design. In deciding what hypotheses to investigatein the empirical study, thought should be given to thepotential contribution to the literature and the feasibility ofdeveloping a rigorous research design for testing them.Researchers who fail to focus on developing hypotheses asthey review the literature often end up spending manymonths or even a year reading the literature without havingidentified a single hypothesis they want to test.
Use the Literature toStimulate Your Thinking
It is critical that the existing literature be used to stimu-late one’s thinking beyond that of merely understandingwhat is covered in each of the individual articles reviewed.In this regard, researchers need to consider such things aswhy different studies may have produced what seem to beconflicting results and what overall inferences one candraw from the studies as a group. They also should concernthemselves with how existing conceptual frameworksmight be improved. For example, have previous research-ers overlooked important antecedents or consequences?Have past studies failed to consider potential mediators ormoderators? Researchers must avoid allowing the litera-ture to constrain their thinking. One aid for doing this is forresearchers to constantly ask themselves what they
406 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
personally believe about the phenomenon of interest.These are issues that researchers should concern them-selves with as they are reviewing the literature rather thanonly after all of the literature has been read.
Put It on Paper
Researchers should write down their ideas as they occurto them and maintain a file. Failure to immediately commitone’s ideas to paper means that time will be wasted tryingto rediscover old ideas, and some ideas may be lost forever.The mere act of writing down their ideas often makesresearchers more aware of ambiguities in their thinking.Frequently, arguments that seem so clear in their headsbecome unraveled when they write these down. This per-mits researchers to identify the problems in their currentthinking and work to resolve them. Finally, committingone’s thoughts to writing makes it much easier to get con-structive feedback from others.
Don’t Work in Isolation
It is difficult for most researchers to conceptualize atight research study without interacting with others, if forno other reason than that it is difficult for people to evalu-ate their own work. This is particularly true for less-experi-enced researchers. Doctoral students who have infrequentinteraction with their dissertation committees almostalways take a long time to complete their dissertations. It isoften the case that researchers clarify their own thoughts,identify problems with their conceptual framework, anddiscover new ideas solely as a result of communicatingtheir current thinking to others. The mere process of orallyexplaining their thoughts to others forces researchers toexamine their ideas more deeply. Hence, it is almostalways a mistake for researchers to wait until they feel theirconceptual frameworks are very well developed beforeexposing them to others. Although almost anyone willingto listen and read what has been written can be helpful, par-ticularly valuable are those who constantly ask for clarifi-cation and question the researcher’s assumptions, concep-tual definitions, and theoretical rationale. These inter-actions are especially beneficial when researchers havepreviously committed their ideas to writing.
Develop Precise ConceptualDefinitions for the Constructs
The conceptual definitions of the constructs of interestwarrant special attention. Constructs are the buildingblocks of theory. Without well-developed conceptual defi-nitions for the constructs, it is impossible to develop acoherent theory. For example, we cannot develop a
meaningful theoretical rationale for why Construct Ashould be related to Construct B if the exact meaning ofeach of these two constructs has not been established.Moreover, it is impossible to develop a valid measure of aconstruct that is not precisely defined.
Avoid developing pseudodefinitions. Some authorswill talk about some Construct A being a result of or thecause of some other Construct B. However, one cannotdefine a construct in terms of its antecedents or its conse-quences. Moreover, trying to do so means that the pro-posed theoretical linkage between A and B would not beempirically testable (i.e., it could not be falsified); rather, itwould be true by definition. Another type of pseudodefini-tion one finds in the literature involves merely givingexamples of what is included in a construct (e.g., Con-struct A includes such things as . . . ). These pseudodefini-tions invariably provide an incomplete listing of the con-struct’s content and fail to indicate what is not included inthe construct. The central role of constructs requires thatresearchers make reasonably certain that their constructsare well defined before moving on to other aspects of theirconceptual framework or to their research designs.
Evaluate the Hypotheses
The hypotheses to be tested also need to be evaluatedbefore designing the empirical study.
• Are the hypotheses clearly written?• Is each of the hypotheses falsifiable?• Do any of the hypotheses involve truism or tautolo-
gies?• Are any of the hypotheses trivial in the sense that
others would be likely to question the methodologyof any study that reported negative results?
• Is the theoretical rationale provided for each hypoth-esis compelling?
• Are there any additional theoretical arguments thatwould strengthen the conceptual support for the hy-potheses?
• Do the hypotheses to be tested represent a cohesiveset?
It is important for researchers to aggressively solicitcriticism of all aspects of their conceptual framework. It isonly when continued exposure of the conceptual frame-work to criticism ceases to uncover serious flaws and allnecessary revisions have been made that researchersshould move to the design phase. The time to revise theconceptual framework is before the data are collected. Af-ter the data are collected, researchers are severely re-stricted by the available measures as to what changes theycan make in their conceptual frameworks.
Summers / RESEARCH IN MARKETING 407
Identify the Intended Contributions
At this point, it is important to make explicit the in-tended contributions of the study and to evaluate them.The contributions of a study can be conceptual, empirical,or methodological in nature. Conceptual contributionscould involve such things as:
1. improved conceptual definitions of the originalconstructs;
2. the identification and conceptual definition ofadditional constructs to be added to the concep-tual framework (e.g., additional dependent, in-dependent, mediating, and/or moderatorvariables);
3. the development of additional theoretical link-ages (i.e., research hypotheses) with their ac-companying rationale; and
4. the development of improved theoretical ratio-nale for existing linkages.
Empirical contributions would include such things as:
1. testing a theoretical linkage between two con-structs that has not previously been tested,
2. examining the effects of a potential moderatorvariable on the nature of the relationship be-tween two constructs,
3. determining the degree to which a variable me-diates the relationship between two constructs,and
4. investigating the psychometric properties of animportant scale.
When field studies are being used, methodological con-tributions might involve changes in the design of past stud-ies that:
1. reduce the potential problems with sharedmethod variance through the insightful use ofmultiple methods of measurement,
2. increase the generalizability of the researchthrough more appropriate sampling procedures,
3. allow the investigation of the plausibility of“third-variable explanations” for the results ofpast studies, and/or
4. enhance the construct validity of key measuresthrough the use of refined multiple-item mea-sures and/or the use of measurement approachesthat do not rely on self-reports.
With respect to laboratory experiments, methodologi-cal contributions might involve such things as modifica-tions in the experimental procedures that serve to:
1. increase the internal, ecological, and/or externalvalidity of the experiment;
2. improve the construct validity of the putativecauses and effects (e.g., through the develop-ment of improved manipulations of the inde-pendent variables and/or the improvement ofmultiple-item scales for the dependent vari-ables);
3. enhance statistical conclusion validity;4. increase the experimental realism of the experi-
ment; and/or5. decrease the plausibility of demand artifacts.
Not infrequently, less-experienced researchers try todesign their studies to contain many such contributions inan attempt to make certain that the overall contribution oftheir research will be sufficiently high. Pursuit of this ap-proach is often associated with the risk of the researcher’stime and effort getting so spread out among many tasksthat every aspect of the study is poorly done. The impor-tant issue is not how many contributions a study will makebut rather the significance of each contribution. Oneshould be concerned with such things as the degree towhich a proposed contribution fills some important gap inthe literature. For example, a study could make a very sub-stantial contribution by demonstrating that a previouslyunidentified moderator variable could explain what previ-ously appeared to be conflicting results in past research.Feedback from successful researchers with a reputationfor being candid is very helpful in pruning the list of in-tended contributions to those likely to have the greatest im-pact on the research area of interest.
Designing the Empirical Study
When the conceptual framework has been set and theintended contributions of the study determined, it is timeto consider the details of the research design. Althoughpast research in an area can serve as a valuable guide, it isimportant to recognize that no study is without method-ological shortcomings. One should always be cognizant ofthe methodological weaknesses and/or limitations of pub-lished research and attempt to overcome these limitationsin one’s own work. For example, to the degree that previ-ous measures appear to lack content validity, consider-ation should be given to revising some of the items used inthese scales and developing new items to add.
The time for researchers to get critical feedback on theirresearch designs is before they collect their data. Althoughresearchers can make some modifications to their concep-tual frameworks (e.g., clarify conceptual definitions, pro-vide additional theoretical rationale for some of thehypotheses) even while their manuscripts are under reviewat a research journal, nothing can be done to improve theresearch methodology once the data have been collected.Moreover, if the data are seriously flawed, no amount ofrewriting of the manuscript can overcome this fact.
408 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
Experts on the particular research methods being usedshould be solicited to critique the research design beforethe data are collected. Moreover, they should be encour-aged to be as critical and detailed as they are when review-ing manuscripts for a journal.
A rigorous pretest of the questionnaire can almostalways provide valuable information on how it might beimproved. Unfortunately, many pretests are not very rigor-ous and only give the researcher a false sense of security.For example, when conducting a pretest of a question-naire, many researchers will ask a small sample from thepopulation of interest to complete the questionnaire andwhen they are finished ask them if they noticed any prob-lems. If those in the pretest sample complete all items onthe questionnaire and do not report any problems with anyof the items, these researchers conclude that the question-naire is without serious flaws. However, this conclusion isseldom justified. Participants often mark responses to themost confusing questionnaire items and never questionwhat these items were intended to measure. When askedafter completing a questionnaire whether any part of it wasconfusing, participants typically say little, if anything,even when many of the questions are confusingly worded.There are several plausible reasons for this situation. First,pretest participants may be constrained in the time andthought they are willing and able to devote to filling outquestionnaires. Second, they may not be sufficientlyskilled and/or experienced at detecting and articulatingproblems with questionnaire items. Finally, they may bereluctant to be critical, even when asked to.
Pretesting of the questionnaire is especially critical ifnew scales are being constructed or previous scales havebeen significantly revised. To determine what pretest par-ticipants really think about their questionnaires, research-ers must be very aggressive in extracting this information.For example, as the pretest participants complete the ques-tionnaire, the researcher might ask these participantswhether they can think of more than one way to interpretwhat each item is asking and to report these interpreta-tions. This should be done separately with each participantone question at a time. The researcher might also ask theseparticipants to explain why they responded the way theydid on each item. However, this approach will work only ifthe participants are perceptive and willing to devote a sig-nificant amount of time thinking about each item. Oneinsightful and articulate pretest participant who is commit-ted to providing constructive criticism is worth more than20 reluctant pretest participants.
Whenever feasible, it is a good idea to use multiple-item scales because these scales are usually more reliablethan single-item scales and their reliability can be easily
measured when the scales are reflective. When buildingmultiple-item, reflective scales, it is useful to administerthe questionnaire to a small sample (e.g., approximately30 participants) after the initial pretest has been conductedand revisions made. This allows researchers to determineif their items are producing the anticipated pattern of cor-relations. When this pattern is not achieved, the samplecorrelation matrix can be used to identify problem items.These items can then be revised or discarded based on acareful analysis of the content of each item.
Experiments involving human subjects are even moredifficult to design and pretest than are surveys. Whendeveloping a new experimental design, it is critical that anextensive evaluation of the design be undertaken. In addi-tion to pretesting the measures, researchers need to be con-cerned with whether (1) the experiment has a sufficientamount of experimental realism, (2) the experiment con-tains demand artifacts, (3) the manipulations provide theintended variance in the independent variables, and (4) themanipulations might be causing unintended variance inother variables that might have an impact on the dependentvariables of interest. After evaluating their own initialexperimental designs and making the necessary revisions,researchers should ask one or two individuals with specialexpertise in experimental design (e.g., those who routinelyreview manuscripts reporting experimental studies for theleading research journals) to examine their experimentaldesigns and materials and to comment on what they feelthe weaknesses of the designs might be. After revisingtheir designs, researchers should recruit three or fourinsightful and articulate individuals to serve as initial pre-test participants. These participants should be asked toprovide a verbal protocol as they proceed through theexperiment in a thoughtful manner. After all necessaryrevisions have been made, a pretest using participantsfrom the population of interest should be conducted. Theprimary purpose of this pretest is to collect manipulationand confounding check measures. This will tell research-ers whether their manipulations are working as planned. Ifthe dependent variables are assessed during this pretest,they should be measured after the manipulation and con-founding checks. Given a sufficient sample size for thepretest, it will not be necessary to include manipulationand confounding checks in the main experiment.
Unless a behavioral experiment largely replicates a pastresearch design, failure to identify several significantproblems in the initial design is reason for concern. It israrely, if ever, the case that a newly developed researchdesign does not contain several serious methodologicalproblems. Hence, when the initial pretest does not revealserious defects in the research design, the researcher
Summers / RESEARCH IN MARKETING 409
should strongly consider conducting a second, more rigor-ous pretest.
CRAFTING MANUSCRIPTS FORSCHOLARLY JOURNALS IN MARKETING
When researchers do an excellent job of conceptualiz-ing their studies, developing and executing their researchdesigns, and analyzing their data, the most difficult part oftheir work is behind them. Researchers need not be tal-ented or creative writers to report the results of well-con-ceptualized and executed studies. They only need to beorganized, accurate, and concise in their writing. All well-written manuscripts have three characteristics in common:(1) an introduction that “sells” the study; (2) tight logic,clarity, and conciseness throughout all sections; and (3) acreative and insightful Discussion and Conclusionssection.
Introduction—Selling the Study
To convince readers of the importance of their studies,authors need to accomplish the following four goals in theindicated order:
1. Establish the importance of the general area ofinterest.
2. Indicate in general terms what has been done inthis broad area.
3. Identify important gaps, inconsistencies, and/orcontroversies in the relevant literature.
4. Provide a concise statement of the manuscript’spurpose(s), the contributions the manuscriptmakes to the literature.
The contributions noted should relate back to the gaps,inconsistencies, and controversies identified earlier.
In establishing the importance of the general area of in-terest, one need not develop long and complicated argu-ments or discuss the detailed results of several articles.Establishing the importance of the topic area can often beaccomplished rather quickly and easily as the followingsample text suggests:
_____________ researchers have devoted consider-able attention to developing and testing models of___________________ (e.g., cite several promi-nent articles in the area).1
Next, the author should indicate in general terms whathas been done in the broad area. A lot of journal space neednot be devoted to achieving this goal. It is not expected ordesirable that authors report the detailed findings of indi-vidual studies. For example, consider the following sam-ple text:
Previous research has addressed several aspects of___________ : (1) ______________ (cite two to threerelevant articles), (2) _____________ (cite two tothree relevant articles), and (3) _____________ (citetwo to three relevant articles).
The results of the studies cited need not be reviewedwhen the current article focuses on different issues thanthose covered in the studies cited.
Then, researchers need to identify important gaps, in-consistencies, and/or controversies in the literature. Thisserves to establish the need for additional research in thetopic area of interest. This task, like those that precede it,can be achieved in a concise manner. For example, con-sider the following sample text:
However, in addition, ___________________ en-compasses several unexplored dimensions thatlately have attracted research attention in other dis-ciplines (cite two to three relevant articles).
Some of these unexplored ________ appear to beimportant and worthy of investigation in the contextof _____________________________.
An investigation of these issues is important be-cause ___________________________________.
Furthermore, previous empirical research has fo-cused primarily on ___________________ . Verylittle research has been done on ______________ .
Finally, and most important, the researcher must pro-vide a concise statement of the manuscript’s purposes, thecontributions the manuscript makes to the literature. Thisstatement should follow logically from the text that identi-fies gaps, inconsistencies, and/or controversies in the liter-ature. For example, consider the following sample text:
In this study we seek to extend _______________by addressing the gaps in ________________ . Thestudy investigates the impact of four ___________ :(1) __________ , (2) ___________ , (3) ____________ ,and (4) _______________ . In addition, interrela-tionships among __________________________are examined.
Researchers should avoid trying to develop a long listof contributions (conceptual, empirical, and methodologi-cal). Inevitably, several of these “contributions” will be oflow importance and will divert the reader’s attention fromthe major focus of the study. Researchers must make clearwhat major contributions their studies make and explainwhy these contributions are important. It is a mistake to as-sume that readers will decipher the importance of the studyfrom a description of what was done. The failure to clearlyspecify the importance of the study in the introduction isoften the result of not having given enough thought to thisissue before the study was conducted.
410 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
Writing quality is often a reflection of the clarity of theauthor’s thoughts. Overly vague ideas invariably lead toconfused writing or the lack of any writing. It is generallythe case that when authors have trouble writing, the prob-lem lies primarily with the clarity of their thoughts asopposed to their ability to phrase their ideas properly. Assuch, authors should first question their understanding ofwhat they want to communicate when they are having dif-ficulty writing.
The manuscript must be clearly written, concise, andcharacterized by tight logic. When evaluating their ownwriting, authors will often ask themselves whether the textis consistent with their ideas. This is far too low a standardto use because it does little, if anything, to ensure that thereader will understand the author’s message. Instead, oneshould adopt Stevenson’s standard: “Don’t write merely tobe understood. Write so that you cannot possibly bemisunderstood.”
Authors need to ask themselves whether it is possible toderive either unintended meaning or no meaning at allfrom what they have written. The aggressive search foralternative interpretations of one’s text is a key to identify-ing ambiguous and confusing passages.
Jargon, the specialized vocabulary of a discipline, canbe useful by adding precision and conciseness to research-ers’writings. However, it is frequently misused (overused)in an attempt to make a manuscript appear more sophisti-cated. Unfortunately, it typically achieves the oppositeeffect. All such terms should be defined where they firstappear unless their meaning is (1) invariant and (2) well-known to most readers.
Conciseness in writing is a virtue, particularly whenpublishing in research journals. Since journal space isscarce and costly, the contribution-to-length ratio is an im-portant consideration in a journal’s decision as to whetheror not to accept a manuscript for publication. While writ-ing in a succinct manner can be a daunting task for first-time authors, examining particularly well-written articlesin the target journal can be very helpful. For example, con-sider the following passages that deal with conceptual def-initions, theoretical rationale for hypotheses, and researchmethodology:
Conceptual definitions: “__________________ isdefined as _______________________________ .”(If borrowed, cite the source.)Rationale for hypotheses: “Considerable evidencefrom previous research suggests that __________________________________ .” (Cite two to threekey articles.)“Furthermore, _____________ (cite “leading ex-perts”) argue that _______________ , they hypothe-size ___________________________ .”
Research methodology: “Data were obtainedthrough self-administered questionnaires from_________________ in three ________________ .”“A total of ______ usable responses were obtainedfor an overall response rate of ________ .”“ ________ was measured by an ___ item instru-ment based on the research of ______________(cite key article).”
Each of the above passages contains a lot of informa-tion while using very few words.
Another way to keep the length of a research manu-script reasonable is to be parsimonious in the use of refer-ences. Often, two or at most three well-chosen referenceswill provide sufficient support for a position. Moreover,too many references may make the manuscript difficult toread.
Sections involving reviews of the literature deservespecial attention. It is unsatisfactory to provide a series ofsummaries of individual studies when reviewing pastresearch. These consume journal space without addinganything to our understanding of the literature. As Chur-chill and Perreault (1982) observe, a review should“advance the field by virtue of its insightful, integrative,and critical evaluation of the state of work in a subjectarea.” A good review section will provide a synthesis of theliterature and make clear what is “known” with a fairamount of certainty and where the gaps are.
A Creative and InsightfulDiscussion and Conclusions Section
The Discussion and Conclusions section is the lastthing readers see, and it can have a large impact on theirimpressions of the research being reported. This sectionshould build on the Introduction section. In this regard, itneeds to reaffirm the importance of the study by showinghow the study reported fits into the literature (e.g., whatgaps in the literature it fills). The study’s contributions andtheir importance should be made clear by communicatingthe study’s implications for theory and practice. To merelysummarize the empirical results is an inappropriatestrategy.
It is important to clearly distinguish between conclu-sions and speculation when writing the Discussion andConclusions section. Conclusions must be clearly sup-ported by the data. However, authors may have valuable,informed speculation to share. As Churchill and Perreault(1982) observe, “Good science and good ‘speculation’arenot incompatible, but each should be clearly labeled sothat the two are not confused” (p. 286). A few interestingideas can go a long way here. While the Discussion andConclusions section should not be dominated by specula-tion, authors should identify new issues raised by the
Summers / RESEARCH IN MARKETING 411
study’s findings and/or provide insightful (nonobvious)directions for future research.
Self-Edit the Manuscript
The initial draft of even the most carefully preparedmanuscript can always be significantly improved. Assuch, the initial draft should be revised prior to submittingthe manuscript to others for their evaluation. It is difficultfor authors to edit their own writing. In addition to theproblem of being critical of one’s own work, authors knowwhat they wanted to communicate. This makes it difficultfor them to notice ambiguities and omissions in theirmanuscripts. However, there are things writers can do toreduce these problems. Laying their manuscripts aside fora few weeks reduces writers’ familiarity with their papers.This can help them develop a fresh perspective and bemore open to changes. Another strategy involves analyz-ing the manuscript from the point of view of someone whoknows little or nothing about the topic area. This wouldinclude such things as checking to see whether the special-ized terms have been clearly defined and whether the logicunderlying each of the arguments made and the positionstaken are readily apparent. Finally, authors should askthemselves whether their students would be likely tounderstand most of what they have written. If not, themanuscript needs to be reworked.
Solicit Critical FeedbackBefore Submission
“A colleague who will read what is written, thenquestion its assumptions, ask what’s new, and quibbleabout its language is a person to be cultivated” (Markland1983:142).
Getting feedback from colleagues before a manuscriptis submitted to a journal can significantly increase thechances of the manuscript being ultimately accepted forpublication, but only if the feedback solicited is highlycritical and authors respond to this feedback in a positivefashion. Authors should select critics with extensivereviewing experience and ask them to treat their manu-scripts like they would if they had received these manu-scripts from a journal editor for review. It is not essentialthat these critics be experts in the topic area of interest. Astrong reviewer can usually provide excellent feedback onmanuscripts dealing with a wide range of topics. The feed-back writers receive from their colleagues on variousaspects of the manuscript (e.g., conceptual definitions,theoretical rationale, measurement of the constructs, andwriting) can provide valuable guidance as to how authorsmight improve their manuscripts.
Should the “reviews” received from an initial set of col-leagues contain few substantive criticisms and/or
suggestions for major changes, authors should considersoliciting critiques from a second set of colleaguesbecause it is unlikely that the first set of colleagues werebeing sufficiently critical. Almost all of the approximately10 percent of the manuscripts that are eventually acceptedfor publication at leading research journals are the subjectof substantial reviewer criticism and go through at leastone major revision. Anyone who spends the time to givehighly critical, constructive feedback to an author is doingthe author an enormous favor.
Responding to the Reviewers
Authors are seldom pleased by the reviewers’ reactionsto their manuscripts. After their initial reading of thereviewers’ comments, authors are frequently angered and/or depressed because they feel the reviewers have notfairly judged their work, some reviewers more so than oth-ers. There is a natural tendency for authors to want to provethe most critical reviewers wrong, an approach that is dys-functional to the goal of getting their manuscripts pub-lished. Authors need to pause to recover from their initialemotional reaction and develop a pragmatic approach todealing with the reviews. They need to keep in mind thateven the most critical reviewers are not vindictive andmost of what they say is valid criticism. Reviewers for theleading research journals tend to be very successfulresearchers, and they typically spend from 1 to 2 days pre-paring their reviews for a single manuscript. The manu-script revision process must be guided by a careful consid-eration of the suggestions and critical comments of thereviewers and the editor.
When, even after careful consideration, the specificcontent of a reviewer’s comment appears to be unjustified,authors should examine whether the comment is the prod-uct of some other problem with the manuscript. For exam-ple, authors may sometimes feel the reviewers are askingabout issues already covered in their manuscripts or thatthe reviewers do not understand what the authors aredoing. When this happens, it is best for authors to considerhow they organized and explained things in their manu-script. It may be that the authors need to better communi-cate what was done. Reviewers spend considerable timeand effort reading each manuscript. If they are confused, itis likely that the journal’s readers will also be confused.
In addition to carefully studying the reviewers’ individ-ual comments, authors should look for trends in eachreviewer’s comments. It may be that several of a reviewer’scomments are all related to a single basic problem.Reacting to the comments individually may not fix thisproblem and could even create additional problems byproducing a disjointed manuscript. Authors should alsolook for recurring themes across reviewers. Studying therelated comments as a group may give authors a better
412 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
understanding of the underlying problem and lead to astronger paper than would a piecemeal approach. More-over, any shortcomings that are noted by more than onereviewer deserve special attention.
Authors should try to respond to all of the reviewers’comments in a positive fashion. It is always in the author’sbest interest to set a tone for courtesy when responding toreviewers. The accepted norm is professionalism andcourteousness even when communicating disagreementswith the reviewers and the editor.
After making the necessary revisions to their manu-scripts and formulating their responses to the reviewers,authors should prepare a thorough set of revision notes thataddress both the major themes included in each review andthe reviewers’ individual comments. A separate set ofresponses should be prepared for each reviewer. The revi-sion notes are easiest for reviewers to follow when each oftheir individual comments is followed by the authors’detailed responses.
THE STATE OF RESEARCH IN MARKETING:SOME PERSONAL OBSERVATIONS
While it is easy for an experienced reviewer to be criti-cal of any study, research in marketing has greatlyimproved during the past two decades. Researchers aregiving increased attention to providing a solid theoreticalbase for their studies. Theories developed in other disci-plines have been widely used for this purpose. Purelydescriptive studies have all but disappeared. More thoughtis also being given to how a given study fits into the exist-ing literature and what contribution it makes. Becausetoday’s research studies are more theory based and tightlylinked to the literature, the results of these individual stud-ies are more easily generalized to other contexts.
Today’s quantitative studies are more rigorouslydesigned than past research. More attention is being givento the development and/or use of multiple-item measuresof the central constructs and to providing evidence regard-ing the psychometric properties of the measures used inthe study, primarily internal-consistency measures of reli-ability (e.g., coefficient α). Greater attention is being paidto selecting subjects that are appropriate for the researchquestion of interest. There is less reliance on collegeundergraduate student samples. Finally, the results oftoday’s studies are less open to alternative interpretationsthan past studies.
However, there are areas that are in need of improve-ment. These include (1) theory building research; (2) claimsregarding convergent and discriminant validity; (3) use ofsingle-source, self-report data; and (4) experimentalrealism.
Lack of Theory-Building Research
Marketing researchers have devoted little attention totheory-building research. It is difficult to think of manyempirical articles in marketing whose primary purpose isto develop theory as opposed to merely introducing mar-keters to theories developed in other disciplines (e.g., psy-chology and sociology) and/or testing existing theories.As a discipline, marketing has become content with bor-rowing theory from other disciplines. Several factors maycontribute to this situation. First, most of our doctoral pro-grams do not do a good job of teaching the qualitativeresearch methods (e.g., conducting field interviews andcase studies) that are essential to developing grounded the-ory.2 Many doctoral programs devote very little time tothese methods even though one could argue that rigorousqualitative research is more difficult to conduct, analyze,and report than are surveys or experiments. As a result,most graduates are not skilled at theory-building research.Second, many in our discipline appear to believe that qual-itative research is inherently not as rigorous or prestigiousas quantitative research (e.g., surveys and experiments)and, therefore, the results are difficult to publish. Thisbelief seems to be reinforced by the fact that few doctoraldissertations are based on qualitative research, and oneseldom sees a rigorous qualitative research study pub-lished in any of the leading research journals in marketing.It may also be due, in part, to the negative reactions ofsome researchers to those qualitative researchers whoseem to feel that their research findings do not need to beobjectively verifiable. For too many of the qualitative stud-ies published in the past two decades, it is difficult, if notimpossible, for other researchers to determine whether theauthors’ conclusions are adequately supported by the datacollected and/or to replicate the authors’ findings.
Psychometric Properties of Measures
The vast majority of authors’ claims regarding the con-vergent validity of their measures are unwarranted (i.e.,maximally different methods of measurement are rarelyused), tests for discriminant validity are typically veryweak, and test-retest reliability is rarely examined.3
Although authors often claim to have provided evidenceregarding the convergent validity of their measures, it isusually the case that they use the same interitem correla-tions as evidence of both reliability and convergent valid-ity. Furthermore, in many studies, it appears that theresearchers have sacrificed the content validity of some oftheir measures by deleting items in their initial scales todevelop unidimensional scales.4 Often, the remainingitems reflect a much narrower construct than that origi-nally contemplated. Researchers need to give more con-sideration to using formative scales (i.e., scales for which
Summers / RESEARCH IN MARKETING 413
the observed measures are considered to form the abstractunobserved construct) in those situations where attemptsto develop unidimensional reflective scales (i.e., scaleswhose item scores are considered to be caused by, orreflective of, the construct of interest) fail to result in mea-sures with acceptable content validity. When this occurs, itis often the case that the construct is composed of severaldifferent aspects or dimensions that are not highlycorrelated.
Single-Source Self-Report Data
A long-standing issue regarding studies employing sur-veys is that many involve self-reports and/or key-infor-mant reports from a single source.5 Data are never col-lected from any other source, and the survey respondentsprovide measures for both the independent and thedependent variables. The single-source issue is less of aconcern when several of the variables are objective and/orfactual in nature (e.g., the respondent’s age and corporateprofits as a percentage of sales) and, therefore, more likelyto be independently verifiable from other sources. How-ever, when most or all of the measures involve summaryjudgments of an attitudinal or perceptual nature, commonmethod variance becomes a serious concern in interpretingthe correlations between these measures. Another relatedproblem with single-source data involving self-reportsand/or key informants relates to the consistency motif. Agreat deal of past research on cognition and attitudes hasshown that respondents have an urge to provide answersthat they feel are logically consistent. This creates prob-lems because respondents will often have lay theories ofhow the variables of interest should be related.
Perhaps the most frequent and serious problem withexperiments in marketing is the lack of experimental real-ism (i.e., the degree to which the experiment involves theparticipants, forces them to take it seriously, and has animpact on them). 6 Experiments that ask the participants torole-play without previously having had similar task-related experiences and/or for which there are no meaning-ful consequences for the participant tend to lack experi-mental realism. In these situations, the respondents are mostlikely to tell the experimenter what they feel is a reason-able response. Unfortunately, participants are not alwaysable to predict how they would behave in a given situation.
REVIEWING FOR SCHOLARLYJOURNALS IN MARKETING
Although they are frequently the targets of authors’ an-ger, reviewers provide an indispensable service to the dis-
cipline. Without them, no top research journal could oper-ate. Most reviewers are among the most prolific authors inthe field. They serve as reviewers because they want tohelp the discipline advance, because they feel they owe itto their discipline, because of the prestige of being a mem-ber of an editorial board, and/or because they enjoy the re-viewing process. How reviewers perform their jobs has ahuge impact on how manageable editors’positions and au-thors’ tasks are likely to be. Below are some guidelines forreviewers that help editors and/or authors fulfill their re-sponsibilities.
1. Clearly identify all of the major problems withthe manuscript that are within the reviewer’s ar-eas of expertise. Reviewers should avoid takingstrong positions on issues that are not withintheir areas of expertise.
2. When making global evaluations (e.g., the writ-ing is unclear, the theoretical rationale for thehypotheses are weak, etc.), provide specific ex-amples supporting these evaluations.
3. Indicate which problems are major and whichare minor.
4. Indicate which flaws appear to be correctableand which are not.
5. For correctable flaws, indicate what might bedone to fix them.
6. For uncorrectable flaws, indicate which shouldbe discussed in the Limitations section.
7. If the manuscript is considered to be potentiallypublishable with revisions, clearly indicate whatmust be done to make the article acceptable.
8. When recommending rejection of an article,specify the specific reasons (e.g., uncorrectableflaws). Provide a convincing argument as to whythese flaws justify rejecting the manuscript.
9. Be tactful in writing the Comments to the Au-thors. Start these comments with some positivestatements about the manuscript. Avoid makingpersonal comments and using words with nega-tive connotations (e.g., naive and hopelesslyconfused).
10. When not too time-consuming, direct the au-thors to articles or books that may be useful tothem in revising their manuscripts and/or de-signing their next study. For example, if the the-oretical rationale provided for a hypothesis isweak, cite previous research that might help theauthors develop stronger rationale.
11. Avoid suggesting that the authors cite literaturethat is only loosely related to the research issuesof interest.
12. Avoid asking the authors to cite the reviewer’sarticles unless they are central to the research.
13. Be open to alternative paradigms for studyingthe research questions of interest.
14. Allow the authors some flexibility to write thearticle they want to write.
15. Provide timely reviews (i.e., within 30 days).
414 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
A major key to getting one’s research accepted for pub-lication and dissemination in a leading journal is payingcareful attention to doing the best job possible at every stepof the research and publication process, starting withdeveloping the research idea through preparing the finalrevision of the manuscript. The success of each step isdependent on the steps that preceded it (e.g., it is impossi-ble to develop valid measures of constructs without havingdeveloped precise conceptual definitions of these con-structs). Hence, it is important for researchers to check theadequacy of each completed aspect of their studies beforeproceeding to the next stage. Too frequently, researchersdo not seek feedback from their colleagues until they havewritten the first draft of their manuscript. Moreover, feed-back is only helpful when it is solicited from those withhigh levels of expertise, those providing the feedback aremotivated to be highly critical, and those receiving thefeedback are receptive to constructive criticism. Beingresponsive to criticism is especially critical when goingthrough the review process at a major journal. Not infre-quently, a publishable study never gets in print because theauthor chooses to argue with the reviewers, ignores thereviewers’comments, and/or otherwise fails to adequatelyaddress the reviewers’ and editor’s concerns and incorpo-rate their suggestions in the revised manuscript.
Research in marketing has improved greatly both con-ceptually and methodologically during the past quartercentury. However, much remains to be done. Theory-building research is lacking in marketing. Surveyresearchers should reduce their reliance on single-source,self-report data and use maximally different methodswhen trying to assess convergent validity. Finally, experi-menters need to be more concerned with the experimentalrealism of their studies.
The author thanks A. Parasuraman, Thomas Hustad,Scott MacKenzie, Cheryl Jarvis, and the editor for theirconstructive comments on previous drafts of this article.
1. This sample text is based on material found in Kohli (1985), as aremost all sample texts presented in this section. Basically, the verbiagespecific to Kohli’s study was stripped from Kohli’s article to provide a
sample text appropriate for a wide range of studies. This basic approachcan and should be used with other particularly well-written articles.
2. For an excellent discussion of building theories from case study re-search, see Eisenhardt (1989).
3. For the most authoritative treatments of convergent anddiscriminant validity, see Campbell and Fisk (1959) and Campbell(1960).
4. For an authoritative discussion of content validity, see Cronbach(1971).
5. For an excellent discussion of the problems associated with single-source, self-report data, see Podsakoff and Organ (1986).
6. For an authoritative discussion of experimental realism, seeAronson and Carlsmith (1968).
Aronson, Elliot and J. Merrill Carlsmith. 1968. “Experimentation in So-cial Psychology.” In The Handbook of Social Psychology. 2nd ed.Vol. 2. Eds. Gardner Lindzey and Elliot Aronson. Reading, MA: Ad-dison-Wesley, 1-79.
Campbell, Donald. 1960. “Recommendations for APA Test StandardsRegarding Construct, Trait, or Discriminant Validity.” American Psy-chologist 15 (August): 546-553.
and Donald W. Fisk. 1959. “Convergent and Discriminant Vali-dation by the Multitrait-Multimethod Matrix.” Psychological Bulle-tin 56 (March): 81-105.
Churchill, Gilbert A., Jr. and William D. Perreault, Jr. 1982. “JMR Edito-rial Policies and Philosophy.” Journal of Marketing Research 19 (Au-gust): 283-287.
Cronbach, L. J. 1971. “Test Validation.” In Educational Measurement. 2ded. Ed. R. L. Thorndike. Washington, DC: American Council on Edu-cation, 443-507.
Eisenhardt, Kathleen M. 1989. “Building Theories From Case Study Re-search.” Academy of Management Review 14 (4): 532-550.
Kohli, Ajay K. 1985. “Some Unexplored Supervisory Behaviors andTheir Influence on Salespeople’s Role Clarity, Specific Self-Esteem,Job Satisfaction, and Motivation.” Journal of Marketing Research 22(November): 424-433.
Markland, Murry F. 1983. “Taking Criticism—And Using It.” ScholarlyPublishing: A Journal for Authors and Publishers 14 (February):139-147.
Podsakoff, Philip M. and Dennis W. Organ. 1986. “Self-Reports in Orga-nizational Research: Problems and Prospects.” Journal of Manage-ment 12 (4): 531-544.
Varadarajan, P. Rajan. 1996. “From the Editor: Reflections on Researchand Publishing.” Journal of Marketing 60 (October): 3-6.
ABOUT THE AUTHOR
John O. Summers (Ph.D., Purdue University, 1968) is a profes-sor of marketing in the Kelley School of Business at Indiana Uni-versity. His work has appeared in the Journal of MarketingResearch, the Journal of Marketing, the Journal of ConsumerResearch, the Journal of the Academy of Marketing Science, theJournal of Business Research, the Journal of Business Adminis-tration, and the Journal of Advertising Research. He served onthe Editorial Review Board of the Journal of Marketing Researchfrom 1972 through 1998.
Summers / RESEARCH IN MARKETING 415
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001REVIEWS OF BOOKS
This edition of JAMS features reviews of five eclecticbut important books that may represent must-reads formarketing academics, practitioners, and graduate stu-dents. The section begins with reviews of two books thataddress questions of the social responsibility of business.First is Jay M. Handleman’s review of the Handbook ofMarketing and Society edited by Paul N. Bloom and Greg-ory T. Gundlach. This is an important collection of articlesfor those interested in interplay between marketing andsociety. Articles by top scholars in the field clearly showhow marketing can affect various aspects of consumerwelfare. Next is Edwin R. Stafford and Cathy L.Hartman’s review of NGOs Engaging With Business: AWorld of Difference and a Difference to the World, writtenby Simon Heap. This book shows how nongovernmentorganizations (NGOs) are collaborating with the globalbusiness community to address social and environmentalproblems through environmentally sustainable practicesand ethical codes of conduct.
Marketing has long adopted and adapted theories andmethods originating in other disciplines. Alf H. Walle III’sbook, Rethinking Marketing: Qualitative Strategies andExotic Visions, is a new addition to this tradition. It isreviewed by June Cotte. Both the author and the reviewerstress that while new methods and insights can be gainedfrom the humanities, simple appropriation of the methodswithout true mastery is dangerous.
The last two books reviewed in this volume come fromtwo marketing titans—Philip Kotler and Shelby D. Hunt.As A. Coskun Samli notes in his review of Kotler on Mar-keting: How to Create, Win and Dominate Markets, PhilipKotler needs no introduction to either marketing practitio-ners or marketing academics. Kotler on Marketing is muchmore than an abbreviated version of Marketing Manage-ment. It opens up new horizons in teaching and practicingmarketing. Robert A. Peterson and Ashutosh Prasad tookon the task of reviewing Shelby D. Hunt’s A General The-ory of Competition: Resources, Competences, Product-
ivity, Economic Growth. In this book, Hunt synthesizesextant theories of competition to formulate what he termsthe resource-advantage theory of competition.
I hope you enjoy reading these insightful reviews andthe books they cover. While many of the books reviewed inthis section in the past have been written by Americanauthors, we hope to have books written by authors fromother countries included in this section soon. Thus, Iencourage you to submit suggestions for books to beincluded in the future in this section. I also invite you tocontact me if you are interested in reviewing a text in yourparticular area of expertise.
Handbook of Marketing and SocietyVolume Editors: Paul N. Bloom and Gregory T.GundlachThousand Oaks, CA: Sage, 2001, 543 pages, $94.95(hardcover)
It is hard not to notice the heightened interest by mar-keting scholars and practitioners in the interplay betweenmarketing and society. This handbook’s compilation ofarticles by some of the field’s top scholars represents atimely effort by editors Paul N. Bloom and Gregory T.Gundlach to provide a venue through which readers canassess where the field has been and where it is going. Theeditors have organized the collection of articles around aneasy-to-follow framework that represents their view ofhow marketing can affect various aspects of consumerwelfare. Their framework represents an effective structurearound which the literature within this growing field canbe organized.
There are, no doubt, many challenges that the editorsand contributors faced in putting together this volume ofwork. However, there are two challenges that particularlystand out in my mind. First, the intended audience for thishandbook is quite diverse. Marketing practitioners, publicpolicymakers and regulators, and marketing scholars allrepresent important potential target audiences for thishandbook. Providing a collection of work that all audi-ences will find valuable is a daunting task. The secondchallenge lies in attempting to do justice to the diversity of
scholarship that examines marketing’s impact on societalwell-being. For me, this handbook succeeds on the firstchallenge but falls short on the second.
First, dealing with the challenge of a diverse range oftarget audiences, this handbook provides a series of well-written articles that almost anyone who is interested inunderstanding the interplay between marketing and soci-ety will find of at least some value. For example, the publicpolicy practitioner will find this handbook useful with anumber of chapters related to public policy and marketingissues. As a brief illustration of this point, chapter 2 byGregory Gundlach is useful in thinking about how market-ing thought and research can be applied to consumer pro-tection and antitrust issues. Chapter 3, by Paul Bloom,Julie Edell, and Richard Staelin, is an excellent follow-uparticle that provides a useful set of criteria by which publicpolicymakers and regulators can assess the application ofacademic research to regulatory and public policy deci-sions. Corporate marketers interested in public policyissues will find Mary Jane Sheffet’s essay (chapter 6)important in thinking about antitrust issues in the contextof developing marketing strategy.
For marketing academics, this handbook provides twoprimary benefits. First, there are a number of chapters thatprovide a preliminary empirical examination (usually byway of descriptive statistics) and a literature review thatgives a solid grounding of the current state of knowledgefor a given topic. For example, researchers interested inexamining the effects of warnings such as product hazardwarnings (chapter 15), food labels (chapter 16), and envi-ronmental claims (chapter 17) will find good literaturereviews and preliminary descriptive statistics on theseissues. Brenda Derby and Alan Levy (chapter 16) presentsome preliminary descriptive statistics on the consumerand marketplace impacts from the Nutrition Labeling andEducation Act that may provide the basis for more exten-sive academic research. Robert Mayer, Linda Lewis, andDebra Scammon (chapter 17) provide a superb review ofthe issues surrounding environmental marketing claimsand then lay out the basis for interesting future researchdirections. Identifying future research directions repre-sents the second main benefit of this handbook for market-ing academics. For example, Andrew Abela and PaulFarris (chapter 9) provide an excellent review of studiesthat have examined the impact of advertising on competi-tion, prices, and consumer welfare. The authors clearlyoutline the conflicting results from previous work on thistopic, lay out some solid assumptions on which to conductfuture research, and then provide specific research direc-tions for those interested in examining this topic. Most ofthe chapters in this handbook provide a section on futureresearch.
Marketing academics may find a couple of other “sec-ondary” benefits from this handbook. Some of the chap-ters, particularly those that detail regulatory issues, mayprovide useful definitions and illustrations for classroomuse. In addition, those researchers who might be calledupon to serve as expert witnesses by regulators and publicpolicymakers will find three of the chapters (chapters 1, 2,
and 14) useful in detailing the role of such research and thepotential “land mines” that researchers might face if cross-examined.
One area in which the handbook falls short, however, isin its relatively narrow representation of what is otherwisea wonderfully diverse field of scholarship. This handbookaffords some issues numerous chapters that provide seem-ingly infinite refinement to the given topic, while otherimportant issues are either glossed over or not given avoice at all.
Nearly three-quarters of the chapters of this handbookcan be classified as dealing with primarily two issues. Thefirst are regulatory issues such as antitrust, consumer pro-tection, deceptive advertising regulations, product safetyregulations, and other governmental remedies. The secondis the effect of some specific (advertising, pricing, fran-chising) or general aspect of marketing strategy on con-sumer welfare defined primarily from an economicorientation.
Three very important topic areas in this field are onlyafforded the remaining one quarter of the handbook—three chapters on social marketing, two chapters related tocorporate social responsibility, and only one chapter onconsumer resistance (boycotting). Such a distribution ofchapters leaves three important gaps. First, social market-ing and corporate social responsibility, which arguablyrepresent two of the greatest growth areas of the field, aregiven very little attention, particularly when compared toissues such as warning labels, for instance. A second gaprelates to our understanding of “societal welfare.” Whileunderstanding societal welfare from the marketing per-spective is a primary motivation underlying much of thework in the field, the reader is left without much of a senseof what is meant by this construct beyond the relativelynarrow definitions of consumer welfare defined primarilyfrom an economic perspective. This shortcoming directlystems from the third gap being that important streams ofliterature within the marketing and consumer behaviordisciplines that are directly related to marketing and soci-ety are given either no or very limited attention. In particu-lar, critical theory (cf. Murray and Ozanne 1991) andsubsequent applications of this theoretical frameworkhave been given no attention, even though such researchdirectly considers the interplay between marketing andaspects of societal welfare that go beyond consumer wel-fare defined from the economic perspective. Furthermore,the stream of research that examines various aspects ofconsumer resistance to marketing practice (cf. Klein 2000;Penaloza and Price 1993—as a small sample) is given noattention except for the essay by Craig Smith (chapter 7)that examines consumer boycotts. The absence of atten-tion given to these streams of research not only limits abroadening of our understanding of societal welfare butalso casts a blind eye to what represents great areas offuture research potential in the field of marketing and soci-ety. The hope is that this edition of the handbook repre-sents Volume 1 of what will be many future volumes thatprovide readers with a more complete scope of researchthat captures the interplay between marketing and society.
REVIEWS OF BOOKS 417
Klein, Naomi. 2000. No Logo: Taking Aim at the Brand Bullies. Toronto,Canada: Knopf.
Murray, Jeff B. and Julie L. Ozanne. 1991. “The Critical Imagination:Emancipatory Interests in Consumer Research.” Journal of Con-sumer Research 18 (2): 129-144.
Penaloza, Lisa and Linda Price. 1993. “Consumer Resistance: A Concep-tual Overview.” In Advances in Consumer Research. Vol. 20. Eds.Leigh McAlister and Michael L. Rothschild. Provo, UT: Associationfor Consumer Research, 123-128.
Jay M. HandelmanMcGill University
NGOs Engaging With Business: A World ofDifference and a Difference to the World
By Simon HeapOxford, UK: INTRAC, 2000, 309 pages, £15.95(paper)
Increasingly, market globalization is shifting the bal-ance of power from national governments to multinationalcorporations. Because global businesses are able to oper-ate and move people, capital, resources, and informationefficiently around the world, their collective influence onpeople’s lives and the planet requires an increased obliga-tion for social and environmental responsibility. This trendis changing how development, human rights, and environ-mental nongovernmental organizations (NGOs) addressthe world’s social and environmental problems. NGOs arerecognizing the opportunities for leveraging the globalbusiness community as collaborative partners rather thanas adversaries to advance environmentally sustainablepractices and ethical codes of corporate conduct. AsNGOs pursue strategic relationships with the private sec-tor, however, their new business development models andcollaborative implications for industry are largelyunderresearched (for exceptions in marketing, seeDrumwright, Cunningham, and Berger 2000; Milne, Iyer,and Gooding-Williams 1996). Simon Heap’s bookattempts to fill this gap in the literature by examining thechanging nature of NGO–private sector relationships anddetailing several varied case histories of NGO–private sec-tor partnerships. Heap’s well-written book delivers muchof what it promises.
A historian by training, Heap is a researcher for theInternational Non-Governmental Organisation Trainingand Research Centre (INTRAC) based in Oxford, UnitedKingdom. His book is the fruit of a two-phased investiga-tion. The first was to explore and identify issues relevant tothe emerging dynamics between NGOs and private sec-tors, and the second was to study those issues empiricallyacross several NGO-business partnership case studiesconducted by INTRAC. Following this pattern, the firsthalf of the book provides a comprehensive review ofemerging literature on NGO-business collaboration, andthe second half details INTRAC’s case studies, which
draw primarily from depth interviews of key participants.Methodological detail is kept to a minimum so that thebook will appeal to business and NGO practitioners, aswell as academics and students.
Heap describes several win-win motivations drivingNGO-private sector collaboration. From the NGOs’ per-spective, NGOs are growing disenchanted with govern-ments as a provider and enforcer for social, ethical, andenvironmental solutions to global business challenges.NGOs see the power of markets and corporate access tosupply chains and distribution channels to facilitatechange. As such, NGOs have added market-oriented tac-tics, such as product endorsements and technology advo-cacy to their arsenals of traditional protest and activismstrategies. The case histories illustrate the advantages anddisadvantages of different tactics. Although the privatesector has been slower to recognize the opportunities,businesses engaged in constructive relationships withNGOs can head off negative public confrontations, andcorporate social and environmental initiatives derivedfrom NGO collaboration enjoy enhanced public credibil-ity. Heap describes the interesting case of Greenpeace’sendorsement and collaborative marketing of “green-freeze,” an ozone- and climate-safe refrigerant that helpedGerman appliance manufacturers establish the technologyas an industry standard throughout Europe. Manufacturerstapped into Greenpeace’s influential networks of support-ers, consumers, scientists, government agencies, develop-ment organizations, and other stakeholders to advance thetechnology. Greenpeace even helped to broker joint ven-tures between its corporate partners and factories in devel-oping nations to further diffuse greenfreeze. In short,NGO–private sector collaboration can facilitate sociallyand environmentally responsible global development toachieve both market and socioenvironmental goals.
In practice, however, NGO-business partnerships facesignificant challenges, particularly in integrating profitand socioenvironmental values (Crane 1998). Heap drawsthe following conclusion from his analysis of a product en-dorsement partnership between the Rainforest Allianceand Chiquita Brands in Cost Rica: while product endorse-ments provide a valuable signal to consumers who lackenvironmental expertise, they can open the NGO to criti-cism from stakeholders who are not involved in the en-dorsement process and whose interests are not addressedproperly.
If the NGO has a narrow set of objectives which donot cover all the major issues related to the productor service sector in question, then working on a one-to-one basis is very risky and likely to prove an inef-ficient programme in the long run. (P. 189)
Close corporate ties can threaten NGO credibility shouldthe partnership fail to meet its social or environmentalobjectives (cf. Westley and Vredenburg 1991).
For businesses, NGO collaboration can make corpora-tions vulnerable when their operations and practices are
418 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
scrutinized by NGO activists. Most NGOs use both coop-erative “carrot” and traditional adversarial “stick”approaches in their advocacy campaigns, and NGOs arenot averse to working with companies on one issue andprotesting the same companies on another issue simulta-neously. Partnering NGOs and businesses need to learn to“agree to disagree” when engaged in ongoing relation-ships. Heap quotes Milliman, Clair, and Mitroff (1994) tosummarize how business managers and NGO campaign-ers need to view one another in collaboration: “For corpo-rate officials, this means trying to understand theemotional and spiritual views of [NGOs]. For [NGOs], thismeans having the willingness to temper their idealism andmoral convictions to pursue cooperative, often compro-mised, solutions” (p. 42).
Heap finds that environmental NGOs appear to be fur-ther along the learning curve with corporate collaborationthan their human rights and development NGOs. This islargely because industry is increasingly recognizing thatits long-term future is dependent on environmental andresource sustainability; by contrast, industry has yet toconsider its long-term market or financial sustainability asbeing threatened by social inequity and domestic civilsociety. As such, while many NGOs may desire to workwith the private sector, social- and development-orientedNGOs have more difficulty convincing businesses thatthey are worthwhile partners.
The book’s key strengths lay in its chapters that over-view past research and detail INTRAC’s primary researchof various NGO-business partnerships. Heap organizeskey managerial implications from past research for select-ing partners, managing relations, and strategically posi-tioning such partnerships among stakeholders and broadersociety. The descriptions and analyses of INTRAC’s pri-mary research cover five NGO-private sector relationshipsinvolving sustainable business practices (e.g., the MarineStewardship Council) and Third World development (e.g.,Living Earth and Shell; Rainforest Alliance and Chiquita).The cases provide good contextual detail to understandeach partnership’s processes, challenges, and outcomes.
Given the book’s breadth and depth, the four-page con-clusion and summary chapter misses the opportunity todraw significant parallels and themes between INTRAC’sdetailed cases and research from other scholars to advancean integrated “process model” or a comprehensive set of“lessons learned.” Such a synthesis could have also beenused as an organizing framework to present the cases andimprove the book’s overall contribution to understandingNGO-business collaboration. Moreover, Heap providesonly brief discussion concerning how collaboration mayaffect or undermine the effectiveness of more traditionaladversarial strategies used against business by other NGOswho have not bought into dialogue or partnership. Thebroader question of whether collaboration is more effec-tive than confrontation to bring corporations in line withsocial responsibility and sustainability is not addressedfully. Some critics and researchers have charged thatNGO-business dialogue and collaboration have fallenshort of their heralded promises (e.g., Livesey 1999;
Rowell 1996). Some research suggests that collaborationcan only work when corporations perceive a crediblethreat of confrontation should the business fail to actresponsibly (cf. Harrison 1999).
Indeed, marketing researchers should find the growingtrend of NGO-business collaboration as an important con-text with which to investigate relationships, social market-ing, and image management. Given the depth and clarityof Heap’s case descriptions, marketing academics willfind his book to be a useful starting point to explore howNGO-private sector relationships are affecting marketingpractice.
In short, we would categorize the book as a good refer-ence tracing the increasing demands for global corporatesocial responsibility and how NGOs are responding to thetrend. Relationship marketing scholars, in particular, mayfind the book worthwhile for stimulating thought on howto extend the concepts of trust, commitment, compatibil-ity, and conflict resolution into the unique context ofNGO–business collaboration. The book may be most suit-able as a supplementary text for graduate-level coursesrelated to sociology, political science, corporate environ-mentalism, stakeholder management, and natural resourceand environmental policy. Marketing instructors, how-ever, may find the book’s illustrative case examples ofNGO–business relationships useful for embellishing lec-tures and class discussions on social marketing, ethics, andglobalization.
Crane, Andrew. 1998. “Exploring Green Alliances.” Journal of Mar-keting Management 14 (August): 559-579.
Drumwright, Minette E., Peggy H. Cunningham, and Ida E. Berger.2000. “Social Alliances: Company/Nonprofit Collaboration.” ReportNo. 00-101. Marketing Science Institute, Cambridge, MA.
Harrison, Kathryn. 1999. “Talking With the Donkey: Cooperative Ap-proaches to Environmental Protection.” Journal of Industrial Ecol-ogy 2 (3): 51-72.
Livesey, Sharon. 1999. “McDonald’s and the Environmental DefenseFund: A Case Study of a Green Alliance.” Journal of Business Com-munications 36 (January): 5-39.
Milliman, John, Judith A. Clair, and Ian Mitroff. 1994. “EnvironmentalGroups and Business Organizations: Conflict or Co-Operation?” So-ciety for Advanced Management Journal 59 (1991): 41-47.
Milne, George R., Easwar S. Iyer, and Sara Gooding-Williams. 1996.“Environmental Organization Alliance Relationships Within andAcross Nonprofit, Business, and Government Sectors.” Journal ofPublic Policy and Marketing 15 (2): 203-215.
Rowell, Andrew. 1996. Green Backlash: Global Subversion of the Envi-ronmental Movement. New York: Routledge.
Westley, Frances and Harri Vredenburg. 1991. “Strategic Bridging: TheCollaboration Between Environmentalists and Business in the Mar-keting of Green Products.” Journal of Applied Behavioral Science 27(March): 65-90.
Edwin R. StaffordCathy L. HartmanUtah State University
By Alf H. Walle IIIWestport, CT: Quorum Books, 2001, 211 pages,$67.50
The central question Walle addresses in his book is,“Do marketers and consumer researchers have anything togain from the humanities and qualitative social science?”Although I thoroughly enjoyed this book (more on thatlater), this question is immaterial. Most consumerresearchers have already answered this question with aresounding “yes, of course!” Evidence that they have isclear in the Journal of Consumer Research, as well as else-where (e.g., Sherry 1995; Stern 1990, 1992). The con-sumer researchers who answer “no, there is nothingwhatsoever to be gained from the humanities,” if indeedany of them still exist, would be hard-pressed to find ratio-nal reasons to justify their answers. Many researchers whowould never use humanities perspectives or methodslikely still can see the value of more, and diverse, sourcesof information on consumers and marketing strategy.
Marketing managers may be a different story, but if theplethora of media stories on the increasing use of anthro-pological and other “qualitative social sciences” methodswithin businesses is an indication, I doubt that they needmuch convincing either (e.g., “Sending Ethnographers”2000; “Storytelling” 1997). Walle’s stated goal of inter-preting humanistic and qualitative social sciences perspec-tives so that managers can use them is laudable, althoughperhaps slightly condescending.
Having said all that, let me explain why I did enjoy thebook. Or perhaps I won’t yet, because I haven’t yet dealtwith the book’s title. The title does an extreme disserviceto the book itself. First, the main “Rethinking Marketing”conjures up an image of a critical perspective on the fieldof marketing, which this book is not. In fact, my impres-sion from this main title was that the book would point outall of the problems with marketing as a discipline, perhapswith the goal of debunking some of the central tenets ofcurrent marketing thought, à la Stephen Brown (1995).(This anticipation is partly why I initially agreed to reviewWalle’s book.) It doesn’t. In fact, this book reaches intohistory and literature to show the seemingly universalapplicability of core marketing strategy issues such as tar-geting, segmentation, globalized versus localized strate-gies, and so on to “more firmly establish links between thehumanities/qualitative social sciences and the strategicorientation of contemporary marketing” (p. ix). Thus, the“rethinking marketing” in the title could more accuratelybe called “buttressing marketing.” A second disappoint-ment in the title is the phrase “qualitative strategies.” Per-haps the reader of this review would expect, as I did, thatthis subtitle indicates the book would offer relatively con-crete how-to suggestions for using qualitative methods to“rethink marketing.” It doesn’t. It does, however, offerbroad suggestions for the sorts of strategies (in general)that could be used. For example, in Part 1, Walle uses avery broad humanities approach to studying great works in
literature to “cull wisdom” marketers can use. He linksthemes in The Aeneid and The Virginian to modern adver-tising and uses the New Testament to demonstrate local-ized marketing strategies, niche marketing, internationalfranchises and multinationals, and positioning.
In the second part of the book, Walle starts to point thereader more toward the source material in the social sci-ences and humanities. He makes the very valid point thatsome of the approaches he discusses are controversial andcome with intellectual baggage due to infighting withinthe humanities. (If any interested reader wants to see someexamples of this infighting, simply check out CamillePaglia’s columns in Salon.com—read any reference toFoucault and you’ll see what I mean.) For an example ofthe format of the book, consider the following. Walle fol-lows the development of literary criticism from Aristotleto Freud to New Criticism and neo-Freudian LeslieFielder. He does this to set up his analysis of the Joe Camelcharacter, which takes the form of compare/contrast/destroy Derrida’s deconstructionist method (as practicedby Barbara Stern) to his own structural “myth and symbol”approach. (Although he stays relatively polite, it is obvi-ous that Walle harbors little or no respect for Stern’s work;if you have a morbid, voyeuristic fascination with intellec-tual carnage, be certain to read chapter 6.) Other methodsand approaches that Walle discusses include “AmericanStudies,” and Walle also makes an impassioned plea formore attention to be paid to Edward T. Hall’s work in lin-guistic anthropology. Because Hall’s work is the basis forsome of my own work, I wholeheartedly concur.
A main theme throughout the book is that marketersand consumer researchers should look to the humanitiesfor new methods and new insights. However, Walle cau-tions that a simple appropriation of the methods withouttrue mastery is dangerous. In demonstrating this point, hetakes on the work of Barbara Stern and describes MorrisHolbrook’s (1997) analysis of Edward Bellamy’s work asfundamentally flawed—“The basic premise of this chap-ter is that Holbrook significantly misread Bellamybecause he lacked substantive knowledge of the literaryheritage of the late nineteenth century; therefore, his workis compromised” (p. 132). Personally, I have never seenMorris Holbrook described in print as lacking in knowl-edge—I expect Walle should expect a lively reaction fromHolbrook himself. Walle’s analysis of Bellamy shows par-allels between nineteenth-century utopian literature andGeorge Fisk’s writings in marketing. Although I under-stand why Walle cautions us against uninformed appropri-ation, the somewhat depressing conclusion appears to bethat unless you already have a Ph.D. in the humanities orare willing to spend years of your life retraining yourself,these methods aren’t for you. I should note that the authornever says this, but I drew this conclusion.
While pointing out how far he feels we as consumer andmarketing researchers have come, Walle also makes thepoint that we have much further to go in our understandingand applications of humanities perspectives and methods.At the end of the book, he tries to explain that his aim is notto attack what has already been done by groundbreakerslike Stern and Holbrook. However, whenever he does
420 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
mention a specific consumer researcher, he does more than“point out certain limitations.” These attacks (yes, that’swhat I’d call them, contrary to his request that they not beseen as such) detract from the book. Walle could havemade his very eloquent argument about the need for eclec-tic but informed use of the humanities in marketing with-out the harsh polemical style he uses when talking aboutexisting consumer research.
Pointing out some of the limitations of a book is, Ibelieve, part of the purpose of an informed review. Regard-less of the objections I have to certain parts of the book, itwas an extremely enjoyable read. If there are academicianswho dread submerging themselves into difficult and com-pletely new disciplines without a primer or a guidepost,then this is the book for you. It stimulates new thinking andpoints out “the way in,” while cautioning about the needfor a much fuller immersion in the basic disciplines. Theanalogy I drew is that this is like a map of a forest wilder-ness area without details like rivers and mountains. Itshows only the borders of the area, and outlines them verywell, but does not provide the detail needed to survive inthe forest. The book is (I think) written for marketing man-agers; it is very accessible to any reader, but I do think thatthe abstract nature and far-reaching implications of thebook are likely more interesting to marketing academics.
Having a taste for reading widely across disciplinesmyself, I highly recommend this “quick read” to others.After all, who among the eclectic readers out there couldfail to be intrigued by a book that links Milton Friedmanwith Friedrich Nietzsche, interprets their work using thewritings of William James, and that contains a sectiontitled “Witches, Whores, and Ethics: Illusions or Reality?”
tember 25, p. 32.Sherry, John F. 1995. “Marketing and Consumer Behavior: Into the
Field.” In Contemporary Marketing and Consumer Behavior. Ed.John F. Sherry. Thousand Oaks, CA: Sage, 3-44.
Stern, Barbara B. 1990. “Literary Criticism and the History of MarketingThought: A New Perspective on ‘Reading’ Marketing Thought.”Journal of the Academy of Marketing Science 18 (4): 329-336.
. 1992. “Historical and Personal Nostalgia in Advertising Text:The Fin de Siècle Effect.” Journal of Advertising 21 (4): 11-22.
“Storytelling: A New Way to Get Close to Your Customer.” 1997. For-tune, February 3, p. 102.
June CotteUniversity of Western Ontario
Kotler on Marketing: How to Create, Win andDominate Markets
By Philip KotlerNew York: Free Press, 1999, 257 pages, $27.50
Philip Kotler single-handedly has done more for mar-keting education than most of us combined who are in this
profession. Kotler on Marketing is another notch in thatoutstanding tradition.
This is a good book—well written and full of ideas. It iscomposed of four parts: strategic marketing, tactical mar-keting, administrative marketing, and transformationalmarketing.
The first part is particularly dynamic. It articulates awell-known concept by marketing people quite ade-quately. By emphasizing marketing, a company can bevery profitable. This implies the ability of that company tolearn. Perhaps a vice president of learning is in the making.This position may become more critical in the organiza-tional structure than even a chief financial officer. What islearned here can be put to work to create and deliver betterconsumer value. These values are created by “winningpractices.” These must be put in proper segment-positioncombinations. All of these identify where the companycan create more value and can reflect this value in its brandequity position.
Part 2 gets down to more specific areas of market intel-ligence, development, designing marketing mixes, creat-ing customer loyalty, and delivering customer value,which in reality distinguishes marketing from selling andmakes it the essential competitive force in a marketeconomy.
Part 3 dwells upon administration or implementation ofmarketing plans. Two very key concepts are essential here.First, marketing plans must be developed. There are anumber of those, such as brand plans, product categoryplans, new product plans, market segment plans, and oth-ers. These marketing plans must have five key compo-nents: situation analysis, marketing objectives and goals,marketing strategy, marketing action plan, and marketingcontrols. Second, these plans must be implemented by themarketing organization. This organization is also in chargeof evaluation and control. Again, this is a good sectioneven though it deals mostly with the conventional wisdom.
Finally, transformational marketing (a new and a some-what nebulous concept since marketing by definition istransformational) is the last section that deals with primar-ily electronic marketing, which in some ways is an oxymo-ron in the sense that marketing is a personal valuegeneration, whereas electronic marketing at this point intime is more like communicating with the markets in animpersonal or dehumanized manner. Granted, e-com-merce has an important role to play; however, it is not quiteclear how it is generating additional consumer value.
As for me, the first three parts are more interesting andpresent more depth. The last section should have gone waybeyond the cyberspace marketing. Value marketing, rela-tionship marketing, and mass customization are all reali-ties and are likely to change or transform marketing to asubstantial extent.
In some ways, Kotler on Marketing is a summary ofMarketing Management, the millenium edition, but it is agood summary. I certainly would give my right arm if myweekend MBAs were to read this book before they start thecourse with me. In fact, I think this book was written withthis market in mind, anyway.
REVIEWS OF BOOKS 421
Over the years, I have approached marketing manage-ment by analyzing four distinct approaches: negative mar-keting, inactive marketing, reactive marketing, andproactive marketing. The book can be placed betweenreactive and proactive; however, I consider it to be a littlebit short on proactivity. To be proactive, the firm mustknow, must detect, and react early to major market changesin its markets. There are at least three areas that need to beconnected to the contents of this book to make it moreproactive. All three are related to markets and marketchanges. First, market turbulence or a more extreme ver-sion of it, chaos, necessitates understanding dramatic mar-ket changes and converting them into possible marketingopportunities. Thus, early detection of market changes isextremely important. Second, learning organizations andknowledge maps. Although at the beginning of the book aninteresting discussion regarding learning organizations ispresented, it does not go far enough as to determining howthe learning and/or knowledge are distributed in the orga-nization and are connected to the organizational processflow. If knowledge is not located in the critical areas wherekey decisions are made, the company cannot be quiteproactive. Finally, with the development of more andmore advanced software, data warehousing and data min-ing are critical tools that would facilitate the marketingproactivity.
Perhaps one key point that I raise is related to the title ofthe book. Professor Kotler says, “how to create, win anddominate markets.” There is no problem with winning anddominating markets, but I maintain that marketers do nothave enough supernatural powers to create markets (howwe wish we did). We can tap into untapped markets, wecan articulate unsatisfied consumer needs, we can addressto tacit needs, we can stimulate demand, but we cannotcreate needs and hence, we cannot create markets. Thebook may give some marketing managers the wrong ideathat they could single-handedly create markets.
To summarize, this is a very good book, particularly ifone is not familiar with Kotler’s monumental book, Mar-keting Management. It will certainly open up new horizonsin teaching and practicing marketing. It is only hoped thatit will be read by our colleagues who are not in marketingand know nothing about it but either by teaching or bypracticing business subjects they are influencing market-ing directly or indirectly.
A. Coskun SamliUniversity of North Florida
A General Theory of Competition: Resources,Competences, Productivity, Economic GrowthBy Shelby D. HuntThousand Oaks, CA: Sage, 2000, 303 pages
Ceteris paribus, the understanding of any phenomenonincreases as an increasingly general and unifying theory ofit evolves. In A General Theory of Competition, ShelbyHunt has attempted to construct a general and unifying
theory of competition. Specifically, through a series ofsuccessive analyses, Hunt has attempted to reduce extanttheories of competition and “traditions,” including neo-classic economic theory, differential advantage theory,competence-based tradition, and economic sociology, towhat he terms the resource-advantage theory of competi-tion. Such reductionism, in the words of Causey (1977),requires successfully explaining the laws found in theextant theories and traditions in terms of the laws found inresource-advantage theory. Has Hunt been successful? IsA General Theory of Competition a true tour de force, aquixotic attempt to slay neoclassic economics, both, orneither?
Shelby Hunt has produced, in A General Theory ofCompetition, a tightly written (indeed, at times evendense) tome that is simultaneously interesting, infor-mative, and important. According to Hunt, the monograph“develops the structure, foundations, and implications ofthe resource-advantage (‘R-A’) theory of competition.This new theory, it is argued, is a general, interdisciplinary,evolutionary, disequilibrium-provoking, process theory ofcompetition” (p. xii). Much of what is contained in thismonograph reflects both a distillation and a synthesis ofmore than a dozen articles that Hunt has authored orcoauthored during the past few years in journals found insuch disparate disciplines as economics, management,and, of course, marketing. Thus, the substance of themonograph is neither unique nor novel for those who havefollowed Hunt’s recent work. Even so, A General Theoryof Competition represents a prodigious scholarly effort,one intended, in part, to challenge existing dogma regard-ing “perfect” competition (hence the earlier reference to aquixotic attempt). Consequently, it is likely to be contro-versial and subjected to a plethora of deconstruction aswell as reconstruction endeavors.
A major strength of the book is that it articulates a clearinterpretation and integration of existing theories of com-petition and related traditions. As such, it reminds thereader of Hunt’s (1991) Modern Marketing Theory in itsinsightful yet concise treatment of oftentimes complexphenomena. This articulation alone makes A General The-ory of Competition worthwhile reading. Hunt possesses anenviable knack for coalescing what often seem to be unre-lated perspectives into a coherent and parsimoniouswhole, and the frequent comparisons made betweenresource-advantage theory and extant theories and tradi-tions are most enlightening (and often persuasive).
Perhaps the main “take-away” from the monograph isthe continually voiced allegation regarding the limitationsand failures of neoclassic economics to accurately accountfor and predict real-world competition, especially thedisequilibrating nature of competition. Economists andbusiness pundits, though, are increasingly making this al-legation as well. For example, Noble laureate North (1995,p. 7) opined that
it is not clear where economics is going. But the di-rection is suggested by two glaring shortcomings ofneoclassic theory: it is a frictionless theory in a
422 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
world in which the frictions are where the action is,and it is static in a world in which dynamic change isgoing on at an unprecedented rate. Remedying thesedefects requires that economics . . . modifies the un-realistic assumptions . . . and incorporates time.
Moreover, many of the critical notions in the mono-graph, such as the treatment of intangible assets asresources, are already diffusing into the marketing litera-ture, with Srivastava, Shervani, and Fahey’s (1998)award-winning conceptualization of market-based assetsbeing but one example.
By definition, A General Theory of Competition is awork in progress. Thus, it is appropriate at this time to raisea few issues that need to be addressed, the most importantof which is whether resource-advantage theory qualifies asa theory. According to Hunt (1991), a theory is “a system-atically related set of statements, including some lawlikegeneralizations, that is empirically testable” (p. 4). With-out question, resource-advantage theory consists of sys-tematically related statements, some of which can beconstrued as laws. However, is resource-advantage theoryempirically testable? As a general, unifying theory,resource advantage is at times necessarily nonspecific andeven unnecessarily vague. (Indeed, chapter 5 can be inter-preted as implying that resource-advantage theory can bebroadly considered a theory of human behavior.) Whatconstitutes, for example, “superior financial performance”or “constrained self interest seeking,” and how can theseconstructs be subjected to rigorous empirical testing? If itis not possible to adequately define an industry, how then is
it possible to define competitors? If a firm obtains newmanagement or a new name (essentially new resources), isit still the same firm?
Resource-advantage theory is offered as a general the-ory of competition, with neoclassic economic theorybeing a special case. However, since resource-advantagetheory seems to reject several tenets of neoclassic eco-nomic theory, such as utility maximization by consumers,it is difficult to accept the conclusion that neoclassic eco-nomic theory reduces to resource-advantage theory. Thetwo theories would intuitively seem to be either incom-mensurate or rivals. In brief, to reiterate Hunt’s closingstatement in the monograph, “There is still a lot of work tobe done—a lot of work” (p. 259).
Causey, Robert L. 1977. Unity of Science. Boston, MA: D. Reidel.
Hunt, Shelby D. 1991. Modern Marketing Theory: Critical Issues in thePhilosophy of Marketing Science. Cincinnati, OH: South-Western.
North, Douglass C. 1995. “Economic Theory in a Dynamic EconomicWorld.” Business Economics 30 (January): 7-12.
Srivastava, Rajendra K., Tasadduq A. Shervani, and Liam Fahey. 1998.“Market-Based Assets and Shareholder Value: A Framework forAnalysis.” Journal of Marketing 62 (January): 2-18.
Robert A. PetersonUniversity of Texas at Austin
Ashutosh PrasadUniversity of Texas at Dallas
REVIEWS OF BOOKS 423
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001MARKETING AND THE LAW
Marketing and the Law
Ann Morales Olazábal, Anita Cava, and René Sacasas, EditorsUniversity of Miami
Trade Dress Protection Does Not ExtendExpired PatentTrafFix Devices, Inc. v. Marketing Displays, Inc., 121S. Ct. 1255 (2001)
When is it safe to copy an expired patent? Your com-pany has waited a number of years for your competitor’sprofitable patent to expire. As the date draws near, youplan your strategy—how to lure customers away from thepatent-holding company. You have pondered whether tochange the design, examined the various methods for mar-keting the item, explored pricing strategies, and attemptedto gauge your competitor’s reaction to the patent expira-tion and your launch of a similar product.
Once the patent expires, confident you are within yourrights to copy the patented feature, you begin selling thecopied product. To your dismay, the holder of the expiredpatent sues you for trade dress infringement under theLanham Act. How can this be? Can the patent holder, ineffect, extend the life of the patent by claiming trade dressprotection?
This is the scenario presented by TrafFix Devices, Inc. v.Marketing Displays, Inc., a recent United States SupremeCourt case. Marketing Displays, Inc. (MDI) manufacturedand marketed sign stands, which were built to stand erectdespite adverse wind conditions. The stands used a pat-ented dual-spring design visibly situated at the base of thesign. When the patents expired, TrafFix, a competitor,began to sell sign stands with identical spring mecha-nisms. MDI sued TrafFix under the Lanham Act, claimingthat TrafFix had infringed on its trade dress by copying thevisible spring mechanism. MDI asserted that its springstand design should be afforded trade dress protectionbecause it was visible and recognizable to buyers and usersas being MDI’s product.
Trade dress is the total image or appearance of a prod-uct that serves to identify the product with its manufactureror source. It can include distinctive ways products arepackaged, styled, shaped, or even their color. Althoughtrade dress can be registered as a trademark with the U.S.Patent and Trademark Office, the design can still be
protected without such a filing, as long as it has acquired a“secondary meaning.” That is, the design serves its pri-mary purpose but also serves to identify the product withthe manufacturer. Trade dress infringement occurs whenanother party adopts the same or a similar feature that islikely to cause confusion for a potential purchaser.
Federal law clearly protects against trade dressinfringement. The Lanham Act requires the party seekingtrade dress protection to establish (1) that the trade dress isdistinctive, either inherently or because of a “secondarymeaning” in the marketplace; (2) that the trade dress of thetwo competing products is confusingly similar; and (3) thatthe feature of the trade dress is primarily nonfunctional.Because the party claiming trade dress infringement mustprove each of these elements, failure to prove any one ofthe three will result in a denial of trade dress protection.
Whether a manufacturer can obtain protection fromcompetition through a trade dress infringement claim is ofextreme importance. Patents restrict competitors for a lim-ited time only, while trade dress is protected for as long asit is used.
In the TrafFix case, the trial judge held that MDI had notproven that consumers associated the dual-spring design’slook with MDI and so had not established the necessary“secondary meaning.” The lower court also found thatMDI’s spring design would be afforded trade dress protec-tion only if the feature were shown to be nonfunctional.
Upon review, the Sixth Circuit Court of Appealsreversed, observing that in this case, it would take “littleimagination” on the part of a competitor to hide the dual-spring mechanism or to add an additional spring or two toavoid infringing on MDI’s trade dress. Basing its holdingon the availability of alternative designs, the appellatecourt held that trade dress protection can be denied onfunctionality grounds only when the exclusive use “putscompetitors at a significant non-reputation-relateddisadvantage.”
The United States Supreme Court disagreed, rejectingthis standard. Focusing on the functionality requirement,the justices unanimously concluded that the two expiredpatents on the dual-spring design were strong evidencethat MDI’s design is functional and therefore not eligiblefor trade dress protection.
The appropriate way to protect a functional componentis by use of a patent. Congress has legislated a limitedperiod of exclusivity for patents, and once that time has
expired, the law allows competitors to copy the design.This encourages innovation by rewarding the creator witha short-term monopoly to recoup development costs and toincrease profitability. Once the patents expire, furtherrestrictions from competition would impede the free mar-ket process. Congress did not intend that patent law be cir-cumvented by the use of trade dress protection provided inthe Lanham Act. The Court pointed out that for it to decideotherwise would allow companies to use trade dress pro-tection to effectively extend a patent indefinitely, creatinga monopoly situation and unacceptably hindering freemarket competition.
The nonfunctional-feature requirement for trade dressprotection ensures that it will not be used to extend a pat-ent. The Court’s opinion noted that while MDI had reliedon the functionality of the design in the past, by this suit itwas seeking trade dress protection—which protects non-functional features—after the patent had expired. Not onlydid MDI promote the functionality of the dual-springdesign as evidence to support the patent application, butMDI also had successfully defended its exclusive right tothe design under the patent, even when the competitorplaced the springs in different positions than those in theMDI patent. The focus taken at the time was on the overallfunction of the springs, not the look of the mechanism.
What lessons can a company take away from this case,either as the patent holder or the competitor? Advanceplanning and preparation can determine the outcome of atrade dress infringement lawsuit. A company that may ulti-mately claim trade dress infringement can strengthen itsfuture case in several ways. The firm should seek federaltrademark registration of the trade dress feature/design, ifit qualifies. Management should note that trademark regis-tration does not necessarily guarantee trade dress protec-tion because of the possible differences in interpretationsof this opinion by the U.S. Patent and Trademark Office asopposed to court interpretations.
In anticipation of future trade dress battles, careful pat-ent application preparation is also necessary. The manu-facturer should include descriptions of functional featuresonly, being careful to exclude ornamental, incidental, andarbitrary features. Instead, these features should be regis-tered for trade dress protection.
Moreover, to eventually succeed in demonstrating thenecessary secondary meaning (i.e., acquired distinctive-ness), producers should regularly collect and preserve evi-dence to support possible future trade dress infringementclaims. Consumer surveys, sales data, and advertisingexpenses, as well as actual ads are examples of evidence tosupport acquired distinctiveness attributed to the feature inquestion.
Finally, manufacturers should remember that even ifthe feature was not previously patented, a party seekingtrade dress protection also will have to prove thenonfunctionality of the feature.
The TrafFix case makes it clear that patent holders can-not extend expired patents by making a trade dressinfringement claim under the Lanham Act. Once the pat-ent on a functional feature expires, competitors can now be
assured that they can copy the feature without fear of tradedress infringement lawsuits.
Linda ChristiansenIndiana University Southeast
Supreme Court Expands Federal Power toRegulate the Availability and Use of Data
Condon v. United States, 528 U.S. 141 (2000)
In the era of Internet communication and e-commerce,privacy of one’s personal data has emerged as a major con-cern among individuals and government regulators. Eventhe popular NBC television program The West Wing,which has been known for mirroring current politicalissues, chimed in on the matter when one of the fictionalWhite House staff members asserted that privacy would bethe legal issue of the twenty-first century.
Against this political and social backdrop, the UnitedStates Supreme Court recently articulated a broad state-ment of authority for Congress to regulate the availabilityand use of personal data. The Court’s decision in Reno v.Condon and its potential impact for privacy regulation arediscussed herein.
The constitutional “right of privacy” in the UnitedStates does not provide consumers any protection againstuse or abuse of personal data by private sector organiza-tions. The U.S. Constitution only guarantees that it willprotect individuals or organizations from governmentintrusion, not from invasions by private sector parties,such as businesses or nonprofit organizations. Thus, anyprotection individuals or organizations enjoy from eachother comes through legislation passed by the states orCongress. In the United States, the federal government canonly regulate those areas of national life that are specifi-cally enumerated in the Constitution. One such federalenumerated power is the authority to regulate interstatecommerce.
Historically, the interstate commerce clause—andCongress’s power to promulgate legislation thereunder—has been interpreted quite broadly by the courts, especiallyin the twentieth century. As a matter of fact, under currentinterpretations, all commercial enterprises, no matter howsmall or localized, can be regulated by Congress becausethey “affect” interstate commerce. Still, the regulatedactivity must be commercial, as opposed to purely charita-ble, recreational, or educational. Arguably, however, theSupreme Court in Condon expressed a broader view ofcongressional authority to regulate data privacy in tradi-tionally noncommercial settings.
The genesis of the Condon case is the Driver’s PrivacyProtection Act (DPPA), a federal law passed by Congressin 1994. This statute regulates the states’disclosure of per-sonal information contained in motor vehicle department(DMV) records, including the driver’s photograph, socialsecurity number, driver identification number, name,address, telephone number, and medical or disability
MARKETING AND THE LAW 425
information. Prior to the passage of DPPA, such informa-tion that drivers provided to their states to receive a driver’slicense and to register their vehicles was widely sold toindustry for surveys, marketing, and other purposes. TheSupreme Court noted that, for example, the state of Wis-consin received almost $8 million annually from the saleof this information.
The DPPA restricts the states’ right to sell DMV infor-mation without the prior approval of the driver. Such a sys-tem of affirmative approval is described as an “opt in”system. Under DPPA, the states must ask the subjects ifthey want their data included in the information disclosed,and the subjects must affirmatively agree. Presumably, inthis era of increasing concern over privacy, not many citi-zens would agree to this sale of their information.
The DPPA does make exceptions for all state use of theinformation to carry out governmental functions. The stat-ute also allows for limited disclosure to industry for pur-poses such as product recalls, performance monitoring,and research. The reuse or resale of information receivedunder these exceptions, however, is restricted. The statuteimposes criminal penalties for intentional violators andcreates a civil action for drivers whose information iswrongfully disclosed.
In Condon, the DPPA was challenged by SouthCarolina because its state disclosure provisions directlycontravened the federal mandate. South Carolina’s DMVrecords were available to anyone who filled out a requestform and who confirmed that the records would not beused for telephone solicitations. South Carolina law onlyrestricted use of driver information by way of a systemcommonly described as “opt out,” which requires driversto proactively request that their information be excludedfrom disclosure.
As the Condon challenge made its way through thecourts, it was widely watched by the direct marketingindustry because motor vehicle information was com-monly purchased for targeted solicitations. Industryobservers were concerned about the impact the federallimitations would have on the availability of this oft-usedinformation. State governments and states’ rights advo-cates were interested in the case because it representedissues of federalism versus states’ rights.
In deciding Condon, the United States Supreme Courtpointed out that the relevant data were used by both publicand private parties engaged in interstate commerce. Assuch, the Court declared this type of information to be “anarticle of commerce” and therefore that “its sale or releaseinto the interstate stream of business is sufficient to sup-port congressional regulation.” Thus, the Court held thatthe DPPA is constitutional, and it follows that SouthCarolina’s or any other state’s sale of driver informationwithout drivers’ affirmative “opt in” consent violatesDPPA and could subject the states and other users thereofto sanctions.
As noted above, the Condon decision was important tomany interest groups. Presumably, none of these courtwatchers anticipated that the Supreme Court would articu-late a per se congressional power to regulate the use andavailability of data. Yet, that is the practical implication ofthe Court’s conclusion when it declared the DMV data “anarticle of commerce.” The Court’s description of the DMVdata and its use by commercial and noncommercial parties“in the stream of interstate commerce” seem to apply toany data that are systematically collected and reused byany organization. The commercial nature of the users nolonger is important.
Moreover, the Court’s broad statement of authorityappears not limited to the sale of data, even though that wasthe activity at issue in Condon. For example, the Courtstated that these particular data were “released” for mat-ters related to “interstate motoring.” Here, the Court seemsto be saying that dissemination of data in relation to anyactivity that can be characterized as “interstate” leads tothe conclusion that such data are “an article of commerce.”Sale of the information is not required. As noted above,even small, local activity has been declared “interstate” inour modern jurisprudence. Regardless of whether the reg-ulated entity is commercial, governmental, nonprofit, orrecreational, it appears Congress now can regulate any andall of an entity’s uses of data because information “is anarticle of commerce.”
After Condon, the states will continue to be the focus ofthis congressional power. For example, a recent Senate billproposed a unified tax collection scheme for states toimpose sales tax obligations on Internet and other remotesellers. In crafting such a uniform state scheme, the billrequires that the states protect consumer privacy in the taxcollection process. The states have long sought the abilityto collect state taxes on these remote sales without disputesover whether such processes impose undue burdens oninterstate commerce. Now, as Congress contemplates pav-ing the way for states to accomplish this tax collectiongoal, it may also flex its power to impose on states the obli-gation to protect the corresponding data that will be gener-ated in the process.
Thus, it appears Congress is taking its federalism vic-tory in Condon and continuing to impose privacyprotections on the states. Presumably, new legislation inall fields, such as education, health care, or “faith-based”community services may also include privacy mandatesregarding information obtained in the course of the regu-lated activity. Watch for privacy protection to be a compo-nent of all new federal regulations.
Rita Marie CainUniversity of Missouri at Kansas City
426 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001INDEX
JOURNAL OF THE ACADEMY OF MARKETING SCIENCE
Number 1 (Winter 2001) pp. 1-112Number 2 (Spring 2001) pp. 113-208Number 3 (Summer 2001) pp. 209-332Number 4 (Fall 2001) pp. 333-432
AULAKH, PREET S., see Sarkar, MB.BARRY, ANN MARIE, “How Advertising Works: The
Role of Research, edited by John Philip Jones” [Re-views of Books], 103.
BERTHON, PIERRE, LEYLAND F. PITT, and MICHAEL T.EWING, “Corollaries of the Collective: The Influenceof Organizational Culture and Memory Developmenton Perceived Decision-Making Context,” 135.
BLODGETT, JEFFREY G., LONG-CHUAN LU,GREGORY M. ROSE, and SCOTT J. VITELL, “Ethi-cal Sensitivity to Stakeholder Interests: A Cross-Cul-tural Comparison” [Research Note], 190.
BRAVO, SANDRA J., “Creative Strategy in Direct Mar-keting, 2d ed., by Susan K. Jones” [Reviews of Books],107.
BURKE, DEBRA, “Put ‘Tony the Tiger’ in Your Tank?”[Marketing and the Law], 203.
BURTON, SCOT, see Mason, K.BUSH, VICTORIA D. GREGORY M. ROSE, FAYE
GILBERT, and THOMAS N. INGRAM, “ManagingCulturally Diverse Buyer-Seller Relationships: TheRole of Intercultural Disposition and Adaptive Sellingin Developing Intercultural Communication Compe-tence,” 391.
CAIN, RITA MARIE, “Supreme Court Expands FederalPower to Regulate the Availability and Use of Data”[Marketing and the Law], 425.
CARLSON, LES, RUSSELL N. LACZNIAK, and ANNWALSH, “Socializing Children About Television: AnIntergenerational Study,” 276.
CARR, CHRIS, see Tietje, B.CAVUSGIL, S. TAMER, see Sarkar, MB.CHALLAGALLA, GOUTAM, see Venkatesh, R.CHRISTIANSEN, LINDA, “Trade Dress Protection Does
Not Extend Expired Patent” [Marketing and the Law],424.
CLARKE, IRVINE, III, and MARGARET OWENS,“Internet Marketing: Metatags May Create TrademarkLiability” [Marketing and the Law], 204.
COMPEAU, LARRY D., see Nicholson, C. Y.CORNWELL, T. BETTINA, “Advances in International
Marketing, edited by S. Tamer Cavusgil and Tage KoedMadsen” [Reviews of Books], 318.
COTTE, JUNE, “Rethinking Marketing: QualitativeStrategies and Exotic Visions, by Alf H. Walle III” [Re-views of Books], 420.
COULTER, RON, “High Visibility: The Making and Mar-keting of Professionals into Celebrities, by Irving Rein,Philip Kotler, and Martin Stoller” [Reviews of Books],105.
CRAVENS, DAVID W., see Grant, K.CURRAN, CATHARINE M., and MICHAEL R.
HYMAN, “Selling to Newly Emerging Markets, byRussell R. Miller” [Reviews of Books], 324.
DE R. BARONDES, ROYCE, “Excessive Access to WebInformation Can Be Tortious” [Marketing and theLaw], 327.
DEL CASTILLO, DIAMELA, “Rumors & Lies: The Pa-rameters of Liability for Commercial Speech” [Mar-keting and the Law], 328.
DICKSON, PETER R., PAUL W. FARRIS, and WILLEMJ.M.I. VERBEKE, “Dynamic Strategic Thinking,”216.
ECHAMBADI, RAJ, see Sarkar, MB.ESTELAMI, HOOMAN, and DONALD R. LEHMANN,
“The Impact of Research Design on Consumer PriceRecall Accuracy: An Integrative Review,” 36.
EWING, MICHAEL T., see Berthon, P.FARRIS, PAUL W., see Dickson, P. R.GILBERT, FAYE, see Bush, V. D.GRANT, KEN, DAVID W. CRAVENS, GEORGE S.
LOW, and WILLIAM C. MONCRIEF, “The Role ofSatisfaction With Territory Design on the Motivation,Attitudes, and Work Outcomes of Salespeople,” 165.
HANDELMAN, JAY M., “Handbook of Marketing andSociety, edited by Paul N. Bloom and Gregory T.Gundlach” [Reviews of Books], 416.
HARTMAN, CATHY L., see Stafford, E. R.HENARD, DAVID H., see Szymanski, D. M.HULT, G. THOMAS M., “International Marketing Strat-
egy: Contemporary Readings, by Isobel Doole andRobin Lowe” [Reviews of Books], 322.
HUNT, SHELBY D., see Yilmaz, C.HYMAN, MICHAEL R., see Curran, C. M.INGRAM, THOMAS N., see Bush, V. D.JASSAWALLA, AVAN R., see Sashittal, H. C.JENSEN, THOMAS, see Mason, K.KEAVENEY, SUSAN M., and MADHAVAN
PARTHASARATHY, “Customer Switching Behaviorin Online Services: An Exploratory Study of the Roleof Selected Attitudinal, Behavioral, and DemographicFactors,” 374.
KEITH, JANET E., see Wagner, J. A.KLEIN, NOREEN M., see Wagner, J. A.KOHLI, AJAY K., see Venkatesh, R.LACZNIAK, RUSSELL N., see Carlson, L.LEHMANN, DONALD R., see Estelami, H.LEWIN, JEFFREY E., “The Effects of Downsizing on Or-
ganizational Buying Behavior: An Empirical Investi-gation,” 151.
LIDDELL, PEARSON, MELISSA MOORE, and ROB-ERT MOORE, “Just Sign on the Electronic Line”[Marketing and the Law], 110.
LOW, GEORGE S., see Grant, K., and JAKKI J. MOHR, “Factors Affecting the Use
of Information in the Evaluation of Marketing Commu-nications Productivity,” 70.
LU, LONG-CHUAN, see Blodgett, J. G.MACKENZIE, SCOTT B., PHILIP M. PODSAKOFF, and
GREGORY A. RICH, “Transformational and Trans-actional Leadership and Salesperson Performance,”115.
MASON, KEVIN, THOMAS JENSEN, SCOT BUR-TON, and DAVE ROACH, “The Accuracy of Brand
and Attribute Judgments: The Role of Information Rel-evancy, Product Experience, and Attribute-RelationshipSchemata” [Research Note], 307.
MCCUBBINS, TIPTON F., “A Trade Dress Tale: GangAbducts Princess, Enhances Ransom Through Internet,Court Approves” [Marketing and the Law], 109.
MOHR, JAKKI J., see Low, G. S.MONCRIEF, WILLIAM C., see Grant, K.MOORE, MELISSA, see Liddell, P.MOORE, ROBERT, see Liddell, P.NAKATA, CHERYL, and K. SIVAKUMAR, “Instituting
the Marketing Concept in a Multinational Setting: TheRole of National Culture,” 255.
NICHOLSON, CAROLYN Y., LARRY D. COMPEAU,and RAJESH SETHI, “The Role of Interpersonal Lik-ing in Building Trust in Long-Term Channel Relation-ships,” 3.
OWENS, MARGARET, see Clarke, I.PARKER, RICK, “Personalities and Products: A Histori-
cal Perspective on Advertising in America, by EddApplegate” [Reviews of Books], 102.
PARTHASARATHY, MADHAVAN, see Keaveney, S. M.PETERSON, ROBERT A., and ASHUTOSH PRASAD,
“A General Theory of Competition: Resources, Compe-tences, Productivity, Economic Growth, by Shelby D.Hunt” [Reviews of Books], 422.
PITT, LEYLAND F., see Berthon, P.PODSAKOFF, PHILIP M., see MacKenzie, S. B.PRASAD, ASHUTOSH, see Peterson, R. A.RICH, GREGORY A., see MacKenzie, S. B.ROACH, DAVE, see Mason, K.ROSE, GREGORY M., see Blodgett, J. G., see Bush, V. D.SACASAS, RENÉ, “The ‘Pizza Wars’” [Marketing and
the Law], 205.SAMIEE, SAEED, “Globalization, Privatization, and
Free Market Economy, edited by C. P. Rao” [Reviewsof Books], 319.
SAMLI, A. COSKUN, “Kotler on Marketing: How toCreate, Win and Dominate Markets, by Philip Kotler”[Reviews of Books], 420.
SAMUELS, JEFFREY M., see Samuels, L. B.SAMUELS, LINDA B., and JEFFREY M. SAMUELS,
“Supreme Court Makes Protection of Product DesignTrade Dress More Difficult” [Marketing and the Law],108.
SARKAR, MB, RAJ ECHAMBADI, S. TAMERCAVUSGIL, and PREET S. AULAKH, “The Influenceof Complementarity, Compatibility, and RelationshipCapital on Alliance Performance,” 358.
428 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001
SETHI, RAJESH, see Nicholson, C. Y.SIVAKUMAR, K., see Nakata, C.STAFFORD, EDWIN R., and CATHY L. HARTMAN,
“NGOs Engaging With Business: A World of Differenceand a Difference to the World, by Simon Heap” [Re-views of Books], 418.
SUMMERS, JOHN O., “Guidelines for Conducting Re-search and Publishing in Marketing: From Conceptual-ization Through the Review Process,” 405.
SZYMANSKI, DAVID M., “Modality and Offering Ef-fects in Sales Presentations for a Good Versus a Ser-vice” [Research Note], 179.
, see Troy, L. C.
, and DAVID H. HENARD, “Customer Satisfac-tion: A Meta-Analysis of the Empirical Evidence,” 16.
TIETJE, BRIAN, and CHRIS CARR, “The Latest in In-surance Liability Coverage Determinations: ‘MarketOverflow’ Claims” [Marketing and the Law], 330.
TROY, LISA C., DAVID M. SZYMANSKI, and P.RAJAN VARADARAJAN, “Generating New ProductIdeas: An Initial Investigation of the Role of Market In-formation and Organizational Characteristics” [Re-search Note], 89.
VARADARAJAN, P. RAJAN, see Troy, L. C.VENKATESH, R., GOUTAM CHALLAGALLA, and
AJAY K. KOHLI, “Heterogeneity in Sales Districts:Beyond Individual-Level Predictors of Satisfaction andPerformance,” 238.
VERBEKE, WILLEM J.M.I., see Dickson, P. R.VITELL, SCOTT J., see Blodgett, J. G.WAGNER, JUDY A., NOREEN M. KLEIN, and JANET E.
KEITH, “Selling Strategies: The Effects of Suggestinga Decision Structure to Novice and Expert Buyers,”289.
WALSH, ANN, see Carlson, L.WHITE, D. STEVEN, “Behind the Success and Failure of
U.S. Export Intermediaries: Transactions, Agents andResources, by Mike W. Peng” [Reviews of Books ],323.
YILMAZ, CENGIZ, and SHELBY D. HUNT, “Salesper-son Cooperation: The Influence of Relational, Task,Organizational, and Personal Factors,” 335.
“Corollaries of the Collective: The Influence of Organiza-tional Culture and Memory Development on PerceivedDecision-Making Context,” Berthon et al., 135.
“Customer Satisfaction: A Meta-Analysis of the Empiri-cal Evidence,” Szymanski and Henard, 16.
“Customer Switching Behavior in Online Services: AnExploratory Study of the Role of Selected Attitudinal,
Behavioral, and Demographic Factors,” Keaveney andParthasarathy, 374.
“Dynamic Strategic Thinking,” Dickson et al., 216.“The Effects of Downsizing on Organizational Buying
Behavior: An Empirical Investigation,” Lewin, 151.“Factors Affecting the Use of Information in the Evalua-
tion of Marketing Communications Productivity,” Lowand Mohr, 70.
“Guidelines for Conducting Research and Publishing inMarketing: From Conceptualization Through the Re-view Process,” Summers, 405.
“Heterogeneity in Sales Districts: Beyond Individual-Level Predictors of Satisfaction and Performance,”Venkatesh et al., 238.
“The Impact of Research Design on Consumer Price Re-call Accuracy: An Integrative Review,” Estelami andLehmann, 36.
“The Influence of Complementarity, Compatibility, andRelationship Capital on Alliance Performance,” Sarkaret al., 358.
“Instituting the Marketing Concept in a Multinational Set-ting: The Role of National Culture,” Nakata andSivakumar, 255.
“Managing Culturally Diverse Buyer-Seller Relation-ships: The Role of Intercultural Disposition and Adap-tive Selling in Developing Intercultural Communica-tion Competence,” Bush et al., 391.
“Marketing Implementation in Smaller Organizations:Definition, Framework, and Propositional Inventory,”Sashittal and Jassawalla, 50.
“The Role of Interpersonal Liking in Building Trust inLong-Term Channel Relationships,” Nicholson et al.,3.
“The Role of Satisfaction With Territory Design on theMotivation, Attitudes, and Work Outcomes of Sales-people,” Grant et al., 165.
“Salesperson Cooperation: The Influence of Relational,Task, Organizational, and Personal Factors,” Yilmazand Hunt, 335.
“Selling Strategies: The Effects of Suggesting a DecisionStructure to Novice and Expert Buyers,” Wagner et al.,289.
“Socializing Children About Television: An Intergenera-tional Study,” Carlson et al., 276.
“Transformational and Transactional Leadership andSalesperson Performance,” MacKenzie et al., 115.
Reviews of Books:
“Advances in International Marketing, edited by S. TamerCavusgil and Tage Koed Madsen,” Cornwell, 318.
“Behind the Success and Failure of U.S. Export Intermedi-aries: Transactions, Agents and Resources, by Mike W.Peng,” White, 323.
“Creative Strategy in Direct Marketing, 2d ed., by SusanK. Jones,” Bravo, 107.
“A General Theory of Competition: Resources, Compe-tences, Productivity, Economic Growth, by Shelby D.Hunt,” Peterson and Prasad, 422.
“Globalization, Privatization, and Free Market Economy,edited by C. P. Rao,” Samiee, 319.
“Handbook of Marketing and Society, edited by Paul N.Bloom and Gregory T. Gundlach,” Handelman, 416.
“High Visibility: The Making and Marketing of Profes-sionals into Celebrities, by Irving Rein, Philip Kotler,and Martin Stoller,” Coulter, 105.
“How Advertising Works: The Role of Research, edited byJohn Philip Jones,” Barry, 103.
“International Marketing Strategy: Contemporary Read-ings, by Isobel Doole and Robin Lowe,” Hult, 322.
“Kotler on Marketing: How to Create, Win and DominateMarkets, by Philip Kotler,” Samli, 421.
“NGOs Engaging With Business: A World of Differenceand a Difference to the World, by Simon Heap,”Stafford and Hartman, 418.
“Personalities and Products: A Historical Perspective onAdvertising in America, by Edd Applegate,” Parker,102.
“Rethinking Marketing: Qualitative Strategies and ExoticVisions, by Alf H. Walle III,” Cotte, 420.
“Selling to Newly Emerging Market, by Russell R. Miller,”Curran and Hyman, 324.
Marketing and the Law:
“Excessive Access to Web Information Can Be Tortious,”de R. Barondes, 327.
“Internet Marketing: Metatags May Create Trademark Li-ability,” Clarke and Owens, 204.
“Just Sign on the Electronic Line,” Liddell et al., 110.“The Latest in Insurance Liability Coverage Determina-
tions: ‘Market Overflow’Claims,” Tietje and Carr, 330.“The ‘Pizza Wars,’” Sacasas, 205.“Put ‘Tony the Tiger’ in Your Tank?” Burke, 203.“Rumors & Lies: The Parameters of Liability for Com-
mercial Speech,” del Castillo, 328.“Supreme Court Expands Federal Power to Regulate the
Availability and Use of Data,” Cain, 425.“Supreme Court Makes Protection of Product Design
Trade Dress More Difficult,” Samuels and Samules,108.
“Trade Dress Protection Does Not Extend Expired Pat-ent,” Christiansen, 424.
“A Trade Dress Tale: Gang Abducts Princess, En-hances Ransom Through Internet, Court Approves,”McCubbins, 109.
“The Accuracy of Brand and Attribute Judgments: TheRole of Information Relevancy, Product Experience,and Attribute-Relationship Schemata,” Mason et al.,307.
“Ethical Sensitivity to Stakeholder Interests: A Cross-Cultural Comparison,” Blodgett et al., 190.
“Generating New Product Ideas: An Initial Investigationof the Role of Market Information and OrganizationalCharacteristics,” Troy et al., 89.
“Modality and Offering Effects in Sales Presentations fora Good Versus a Service,” Szymanski, 179.
430 JOURNAL OF THE ACADEMY OF MARKETING SCIENCE FALL 2001