Description
Organization–public relationships is a key concept in public relations research. The emergence of relationship management theory as a new paradigm for public relations increases scholars’ argument about the essence of public relations—what it is and what is its value within the organizations and to the greater society.
Research on Inter-organizational Relationships: Patterns and Motivations
Two connpeting approaches—exchange and powerdependency—have been used in previous studies to conceptualize interorganizational relationships. The basic premise of this paper is that an integrated view of these relationships is needed that incorporates propositions from both of these approaches. Propositions are developed from these two approaches for explaining varia- tions in the frequency of interaction and the nature of interactions between pairs of organizations in an interor- ganizational set. Special emphasis is given to the nature of interactions in an asymmetrical relationship. The propositions are tested using data on interactions between community organizations and local offices ofthe United States Training and Employment Service. The results suggest that interorganizational relationships should be conceptualized as a mixed-motive situation in which each organization behaves in accordance with its own self-interests.*
In the past decade, two approaches to the study of interorganizational relationships have emerged. There are those on the one hand, who view interorganizational relation- ships from an exchange perspective (Levine and White, 1961; Tuite, 1972; White, 1974).''According to this approach, relations form when members of two or more organizations perceive mutual benefits or gains from interacting. The leadership of each organization is internally moti- vated to form the relationship because each perceives that it will be better able to attain its goals by interacting than by remaining autonomous. Aiken and Hage (1968) suggested that this type of relationship forms in periods of scarce or declining resources. In addition to offering an explanation forthe motivations to form the connection, the exchange approach also implies that the nature of the in- teractions between participants in these relationships is characterized by a high degree of cooperation and problem solving as opposed to high levels of conflict and bargaining (Geutzkow, 1966; March and Simon, 1958; Walton, 1965), since the parties are both motivated to coordinate their efforts to maximize joint benefits. Others, however, have adopted a power-dependency approach to interorganizational relations. Here, it is implied that the motivation to interact is asymmetrical. One party is motivated to interact, but the other is not. The interorgani- zational relationship only forms when the motivated party is powerful enough to force or induce the other to interact. The motivation or pressure to interact for at least one of the parties is externally generated, since one party would, if free to choose, prefer to remain autonomous. Bargaining and conflict are the natural form of interaction in this rela- tionship, since each party seeks to attain its own goals at the expense of the other (March and Simon, 1958; Geutzkow, 1966; Schmidt and Kochan, 1972; Aldrich, 1976a).
Unfortunately, these alternative approaches are developing along separate, but parallel, paths. Zeitz (1974), for example, noted that much empirical research has focused on social
220/Administrative Science Quarterly
interorganizationai Reiationshq>s
service delivery systems where the emphasis has been on the goals of achieving coordination and cooperation. Implicitly, therefore, this stream of research has employed the exchange perspective. This has led researchers using this approach to pay little attention to variables that affect the distribution of power within the relationship. In contrast, Benson (1975) focused on the political-economic substructure of interorganizational relations and therefore implicitly adopted a power-dependency approach. Similarly, a recent study of the conditions under which employers are induced to use the United States Training and Employment Service (Schmidt and Kochan, 1976) and another of the conditions under which employers reallocate power within their decision-making structures to deal with trade unions (Kochan, 1975) also fit within the power-dependency perspective. In both studies, such variables as distribution of power, differences in goals, and conflict played a dominant role and organizations with the low intrinsic motivation to interact—in each case the employer—were influenced by the amount of power the other organization brought to bear on the relationship. The separate development of these two competing perspectives is unfortunate, since interorganizational life is not as simple as these distinctions might suggest. No organization is likely to engage in only symmetrical exchange relationships with other organizations or only powerdependency relationships. Rather, an organization is likely to have a mixture of each type within its organization set. Furthermore, no single relationship is likely to be either of these pure or ideal types, but rather one in which there are mixed motives (Walton and McKersie, 1965). The basic premise of this article is that an integrated view of interorganizational relations is needed that is capable of incorporating the basic propositions from both the exchange and the power-dependency models. This view would recognize that relations that are both predominantly symmetrical exchanges and those that are asymmetrical and power-dependency oriented can exist within the same organizational set. Further, overtime, a given relationship between two organizations may shift from being largely one type to the other. Thus, this research attempts to move in the direction of identifying the conditions under which the exchange and the power-dependency approaches can be appropriately applied (Aidrich, 1976). These approaches are incorporated here into a study of the frequency of interactions between community organizations and district offices of the United States Training and Employment Service. The first objective of the study is to determine empirically the characteristics of the interactions between organizations in which each of the above perspectives is predominant. The second objective is to distinguish empirically symmetrical from asymmetrical relationships between the focal organization and members of its organizational set. The third objective is to explore variations in the frequency of interactions which take place under these two conditions, with special emphasis on variations under asymmetrical conditions.
221/ASQ
An explanation of the differences in the frequency of interactions is important not only for the development of theory in this area, but also for public policy. Successful interorganizational relationships are critical to the implementation of almost all public policies (Wamsley and Zaid, 1974). Consequently, before implications can be derived for the design and administration of public policies involving multiple organizations, the factors leading to the development and maintenance of interorganizational relationships must be identified.
HYPOTHESES The following specific hypotheses are tested in this research:
H ^ The lowest frequency of interaction between two organizations will occur when neither organization perceives benefits from interacting—a symmetrical situation. H 2 The greatest frequency of interaction will occur when both organizations perceive benefits from interacting—a symmetrical situation. H 3 In an asymmetrical situation—one organization perceives benefits from interacting and the other does not, the frequency of interaction will be greater than when neither organization perceives benefits and less than when both see benefits. H 4 In an asymmetrical situation the frequency of interaction between two organizations will be higher when the organization perceiving low benefits from interacting perceives the other organization as (a) having goals that are compatible with its own organization, (b) being important to the functioning of its own organization, (c) having greater influence over its organization, and (d) acting aggressively in pursuing its interests by using bargaining and conflict-oriented tactics of influence.
The first two hypotheses are derived from the exchange model of interorganizational relationships (Levine and White, 1961) and argue that the perception of benefits resulting from a relationship will be a good predictor of the frequency of interaction between two organizations. Flowing from the assumption that individuals in organi- zations act in their self-interest, it is predicted that the highest and lowest frequencies of interaction will occur in symmetrical situations in which mutual benefits or mutual lack of benefits are perceived. When members of two organizations perceive benefits from interacting with each other, the mutual motivation for interaction will be great. Similarly, the reverse is posited. Motivations for interaction will be low if parties in neither organization perceive benefits as a result. The third hypothesis suggests that it will be more difficult to initiate and maintain high frequencies of interaction between two organizations when members of only one perceive benefits resulting from the relationship. Likewise, it is more difficult to predict the frequency of interaction when organizations are in this asymmetrical situation. Often, however, this is the rela-
tion which exists between organizations of a given set. Thus, many organizations seek to increase the frequency of interaction with organizations not motivated to do so. Hypotheses 4(a) through (d) focus on situations in which individuals in an organization not perceiving
222/ASQ
Interorganizational Relationships
benefits from interacting with those of another organization who do—^the asymmetrical situation—are most likely to interact anyway. In this situation, it is proposed that a power-dependency framework will be most useful in predicting frequency of interacting. This set of hypotheses focuses on concepts central to a political model of organizational behavior, namely, goals, power, bargaining, and conflict (Wilson, 1973; Benson, 1975). Hypothesis 4 (a) proposes that organizational interaction in the asymmetrical benefit situation will be most probable if the members of the organization perceiving few benefits perceive the other's goals to be compatible with its own. Compatible goals as a condition for interaction has been previously developed by Evan (1966) and Guetzkow (1966). In this situation, those in the organization seeing few benefits also perceive little threat in the relationship; therefore, they do not seek to resist attempts to interact. Essentially, the only costs of interacting are opportunity costs, with time, energy, and other resources being diverted from alternative uses. Hypotheses 4 (b) through (d) are also derived from a power-dependency framework of interorganizational relations in that they consider factors affecting the balance of power between the organizations in the relationship. These hypotheses propose that parties in an organization who perceive low benefits from interacting with those in another may be forced to interact because of a dependency relationship. Thus, hypotheses 4 (b) and 4 (c) specifically propose that the more important (Jacobs, 1974; Aldrich, 1976b; Giiiespie and Perry, 1975) and influential (Emerson, 1962, Thompson, 1967) the other organization is to those perceiving few benefits, the more frequent the interaction. For example, some funding agencies require the formation of practitioner advisory committees as a condition of their research grants. Thus, even though the goals of the researchers and the practitioners may be different, the importance of the practitioners to the researcher requires that such a committee be formed and meet in accordance with the contract terms. Hypothesis 4 (d) proposes that the asymmetrical relationship is analogous to problems faced by pressure groups (Turner, 1968; Lipsky, 1969). That is, people in one organization will interact with those in another in an asymmetrical benefit relationship to the extent that those in the other organization aggressively pursue their interests by engaging in bargaining and pressure-oriented tactics of influence. Additionally, interaction will occur to the extent that the people in the organization receiving the benefit pursue their goals through the use of conflict-oriented strategies (Thompson and McEwen, 1958; Schmidt and Kochan, 1972; Thomas, 1976). While these aggressive tactics are probably not the first means of exerting influence that members of an organization will try, continued frustration with other means of influence will eventually lead to the use of these methods to achieve their goals, especially if the perceived benefits from interaction are strong enough to warrant taking the risks associated with this behavior. Perhaps the best
223/ASQ
examples of interorganizational relationships of this type are trade unions seeking recognition from an employer, inmate associations seeking to negotiate with prison officials, or protest groups seeking to influence poiiticians or community organizations. Hypotheses 1, 2, and 3 are illustrated schematically in the Figure. As shown, there are four possible conditions of interaction contingent on whether the employment service or community organizations perceive high or low benefits from their interaction.
Employment Service Benefits High Community Organization Benefits High
A
Low
B
Low
C
D
Cell A = Mutual benefit voluntary interaction. High level of interaction predicted. Cell D = No or low interdependence; therefore, no or low interaction. Cell B = Asymmetrical condition where community group benefits, but employment service does not. Intermediate kind of interaction. Interaction is dependent on the community group's ability to force employment service to interact. Cell C = Reverse condition of Cell B.
Figure. Conditions of Interaction.
METHODOLOGY Focal Organizations The data for this study were obtained from district offices of the United States Training and Employment Service and those community organizations which had an ongoing reiationship with them. The State Employment Service is a part of the United States Training and Employment Service of the United States Department of Labor and is the primary public manpower delivery system in the state. Although it receives most of its operating funds and primary policy direction from the federal agency, it is administered by the state. The district offices perform a number of activities which essentially involve (1) matching individuals seeking employment with job vacancies filed with the office by employers, (2) assisting job applicants become employable, (3) assisting employers fill job vacancies, and (4) providing labor market information. This study is based on the relationships between 23 employment service district offices in a midwestern state and the primary organizations interacting with each district office. These data were collected during the winter of 1972-1973. Members of the Organizational Sets The composition of the employment service district offices' organizational set was determined on the basis of commonality of clients with community organizations in a district and relationships between these organizations and the
224/ASQ
Interorganizational Relationships
Table 1
Organizational Set Categories Type of Organization AFL-CIO affiliated unions Non-AFL-CIO affiliated unions Chamber of commerce Community action program Concentrated employment program office Division of vocational rehabilitation Equal employment opportunity commission M i nority i nterest organizations National Alliance of Businessmen Probation and parole office Public and vocational schools Social welfare service Veterans of Foreign Wars Total Number
2 7 19 15 3 16 7 8 2 20 12 15 20 146
Percent of Sample
1.4 4.8
13.0 10.3
2.1
11.0
4.8 5.5 1.4
13.7
8.2
10.3 13.7 100.0
district offices. From initial interviews with directors of the district offices, a set of 13 organizations significant to the functioning of these offices was specified. They are shown in Table 1. The criteria for including a community organization in the district office organizational set were twofold: (a) clients of each of the set organizations were or could be recipients of services offered by the district offices, and (b) each organization in the set had some degree of interaction with the district offices. If both of these criteria were met, the organization was included in the organizational set of the employment service district offices.^ In most of the districts, the 13 organizations shown in Table 1 interacted with the district office in that area.
Data
The resource limitations of this study did not allow for a sampling of a sufficient number of employers in each district to include in the analysis.
Two basic sets of data were obtained to test the hypotheses in this study. First, from the community organizations, data were obtained which described their interactions with employment service district offices. Second, a data set was obtained from the district offices describing interactions with the community organizations. Directors of each community organization and corresponding district office received a self-report questionnaire measuring the 10 dimensions of an organizational relationship used in this study. Each of the questionnaire items focused on an aspect of the relationship between a community organization and district office. The respective directors indicated their responses to the items on seven-point scales anchored at the extremes with polar adjectives. The specific dimensions measured were derived from Marrett (1971) and supplemented with additional items designed to capture dimensions of conflict processes, namely, incompatible goals, tension, conflict, and influence (Schmidt and Kochan, 1972; Kochan, Huber, and Cummings, 1975). For the specific items used in this study, see Appendix A. The use of perceptual data was deemed appropriate in keeping
225/ASQ
with similar research in the area of interorganizational relationships (Aiken and Hage, 1968). Additionally, the data were collected from top decision makers in each of the organizations sampled and their behavior is important in guiding the direction of their respective organizations. Since it is assumed that their behavior is a function of their perceived environment (Forehand and Gilmer, 1964; Stern, 1970), perceptual data is meaningful in this context. Overall, the response rates for the survey questionnaires sent to the directors of the community organizations were respectable. After a single mail follow-up, 157 usable questionnaires were obtained, representing a 78.8 percent response rate. Additionally, all 23 personally administered questionnaires from the employment service district office directors were completed. These directors responded to questions concerning each of the 13 types of community organizations in their districts. To have only one response set for each type of community organization in an employment service district, the scores for the same type of organizations in a district were averaged. This averaging occurred in 21 cases. To determine if the averaged scores were significantly different from the nonaveraged scores across community organizations, a comparison of the two were made. This comparison between the single organizational reports (A/ = 136) and the subsample of averaged reports (A/ = 10) showed no significant difference between mean scores on the dimensions examined (see Appendix B). Therefore, the averaged reports and nonaveraged reports were combined. The final data set consisted of 146 paired relationships between community organizations and employment service district offices. Dimensions of these 146 paired relationships were used forthe analyses in this study.
RESULTS Perceptions of the Relationships
The mean and standard deviations on each of the 10 dimensions measured across the 146 paired relationships are reported in Table 2. On 5 of the dimensions there is a significant difference between an employment service district office and respective community organization. The four most significant reported differences between the organizations exist over (a) the influence of the other organization, (b) the degree of compatible goals, (c) the extent of conflict and, (d) the importance of the other organization. The district office perceived the community organization to be less important to its operation than the reverse. Likewise, employment service district office directors reported that community organizations had less influence on the office's operation. The district office directors, however, reported a lower level of compatible goals and more conflict than did the community organization directors. In addition, employment service directors perceived fewer benefits to their organizations from interacting with the community organizations than vice versa; the difference, though, is not statistically significant at the conventional level. Thus, overall, the employment service is not only the
226/ASQ
Interorganizational Relationships Table 2 Perceptions of Paired Relationships Dimension As Perceived by Community organization N= 146 Mean 4.24 4.36 1.82 2.99 1.90 4.38 1.66 4.59 5.42 2.47 Standard deviation 1.76 1.88 .65 .70 1.54 1.53 1.20 2.05 1.62 1.91 Employment service/V= 146 Mean 4.42 4.01 2.12* 2.94 1.92 3.39* 2.08'* 3.89** 4.98** 2.84— Standard deviation 1.65 1.88 .72 .46 1.17 1.70 1.35 1.88 1.86 1.74
Frequency of interaction Benefits from interaction Formalization of agreements Setting terms of interaction Extent of tension Influence of other organization Extent of conflict Importance of other organization Compatibility of goals Bargaining in relationship
p=.OO1.
p=.O1
p = .O8.
more powerful and influential organization in these relationships, but also more resistant to dealing with these community organizations than the reverse. These data are consistent with the data reported on another sample of employment service-community organization relationships (Aldrich, 1976a). Dimensions of Interorganization Relations Before variations in the frequency of interaction across these relationships are examined, the results of a series of factor analyses are presented. These analyses were performed to assess the validity of a distinction used in the literature between cooperative and conflictful relationships. Although Simmel (1955) discussed the duality of conflict and cooperation in all relations, Guetzkow's (1966) classic article spoke about this distinction in connection with interorganizational relationships. No empirical evidence, however, has yet been presented to assess its utility for distinguishing among relationships. As noted earlier, prior empirical research has emphasized either cooperative interactions or those marked by conflict in the models used to study relations within an organizational field. If the central arguments made in Guetzkow (1966), Aldrich (1976 b), and in this study are correct, both cooperative and conflictual patterns of interaction within a single organizational set should emerge. Thus, the factor analysis reported here addresses two issues: (1) does the distinction between cooperative exchanges involving mutual benefits versus conflict-power perspectives make sense empirically as a typology, and (2) can both types of relationships be observed within the same organizational set? The factor analytic procedure consisted of (1) constructing a correlation matrix for all of the interactions, (2) factoring with a principle axis solution, and (3) rotating the resulting factors with a varimax procedure after rotation with Kaiser normalization.
227/ASQ
As shown in Table 3, the factor structures of the paired relationships as perceived by community organizations and employment service district offices are virtually equivalent. On this basis, the factor structure of the combined organizational data is interpreted. Only factor loadings of .4 and above on the dimensions are considered. The dimensions loading on Factor 1 include frequency of interaction, benefiting from the interaction, extent of formalized agreements, influence of the other organization, importance of the other organization, and the compatibility of goals between the organizations. This factor describes a type of cooperative relationship in which members of the organization are intrinsically motivated to interact, because they perceive benefits from the interaction. Factor 1 accounts for 33.5 percent of the total variance. Factor 2, on the other hand, describes a conflictual relationship between district offices and community organizations. This relationship is indicated by loadings on tension, conflict, bargaining, and formalization of agreements. Factor 2 accounts for 22.4 percent of the total variance. Factor 3 accounts for 9.9 percent of the total variance. Only one variable, however, loads at the .4 level on this factor, namely, the amount of reciprocity in setting the terms of interaction between the district office and the community organization. These three factors account for 66 percent of the total variance forthe combined sample. Thus, these findings indicate that the cooperation-conflict distinction is useful. Relationships, however, can be characterized as being cooperative and marked by conflict within the same interorganizational network. The degree to which agreements are formalized is associated with both conflict and cooperation. Therefore, the concepts of formalization and reciprocity do not discriminate between these two dimensions and are not included in the remainder of the analysis.
Frequency of Interaction
It was predicted that the lowest frequency of interaction
Table 3
Factor Structure of Paired Relationships
Item Factor 1 Community Organizations r/V=146) Frequency of interaction Benefits fram interaction Formalization of agreements Setting terms of interaction Extent of tension Influence of other organizations Extent of conflict Importance of other organizations Compatibility of goals Bargaining in relationship Percent of total variance Eigenvalue .784 .829 .488 .067 -.167 .675 .-.168 .836 .787 .207 34.3 3.43 Factor 2 Community EmployOrganizament Service Combined tions (yv=146) (/V=292) .791 .793 .511 .021 Oil .729 -.178 .863 .848 .268 37.1 3.71 .836 .836 .619 .043 -.226 .629 -.053 .784 .627 .192 33.5 3.35 .184 -.004 .503 .108 .807 -.059 .830 -.026 -.112 .643 20.7 2.07 Employment Service Factor 3 Community Organizations Employment Service .036 -.109 -.023 .997 .048 .057 -.116 .021 .070 .017 10.1 1.01
Combined .129 -.133 .467 .153 .848 -.011 .847 -.135 -.404 .552 22.4 2.24
Combined -.104 .162 -.131 .979 .043 -.008 .083 .185 .175 .340 9.9 1.00
.120 .024 .423 -.031 .824 -.073 .830 .052 .066 .636 19.0 1.90
-.106 -.014 -.176 .959 .076 .094 -.033 .103 .069 .146 10.0 1.00
228/ASQ
interorganizational Relationshqis
between members of two organizations would occur where neither perceived benefits from their interaction. On the other hand, the greatest frequency of interaction would take place when members of both organizations had mutual perceptions of benefits resulting from their interaction. The community organization and employment service scores on the dimension of frequency of interaction were classified according to whether the organizations reported high or low benefits from the interaction. The dependent variable, frequency of interaction, is the average of the responses of the community organization and employment service in each relationship. By averaging scores across both organizations, perceptual bias of either respondent is reduced. The scores on the interaction dimension were divided at the median, with the results shown in Table 4. A Chi square analysis of these four cells indicates a systematic variation in the data (X2=9.29, 1; p <.002).
Table 4 Frequency of Interaction by Perception of Who Benefits
Community Organization Perceives
Employment Service Perceives High benefits X = 5.31,S.D.=1.14 Low benefits X=2.21,S.D. =1.69 (B) X=4.58 S.D. =1.26 N =19 X = 3.02
S.D.= 1.07 N =40
A/=59
High benefits X=6.04,S.D.=.82 Low benefits X=2.78,S.D.=1.01
X = 5.28 S.D.= .93 N =50 X= 4.31 S.D.= 1.40 N -34 /V=84 58.7%
(A)
N=7A 48.3%
A/=74
(C)
(D)
51.7%
A; = 1 4 3
41.3%
100%
X2=9.29with 1 df.p =.0023.
The mean value for frequency of interaction in the mutually high benefits condition is 5.28, while the mean value on this same dimension in the mutually low benefits condition is 3.02. These mean values are significantly different from each other and are in the direction predicted by hypotheses H 1 and H 2 (F=29.18; 3,138; p <.OOO1). Likewise, the mean value of Cell A (5.28) is significantly different from the mean values of Cells B (4.58) or C (4.31). The mean value Cell D (3.02) is significantly different than the mean values of Cells B and C. Therefore, these data support hypotheses H ^ through H 3. Interactions under Asymnnetrical Conditions Hypotheses 4 (a) through (d) predicted that in the asymmetrical benefits condition, frequency of interaction would be high when the low benefit organization perceived the other organization as having compatible goals, high influence and importance, and engaging in aggressive pressure tactics of bargaining and conflict. The correlations used to test these hypotheses are presented in Tables 5 and 6. These correlations are based on the data from the relationships that were classified into Cells B and C, the asymmetrical rela22 9/ASQ
Table 5 Zero-Order Correlations between Frequency off Interaction and Independent Variables When Employment Service Perceives Low Benefits and Community Organization Perceives High Benefits (Cell B) Dimension As Perceived by Employment Service /V=19 .64* .06 .56*» .60* .31 p<.05. Community Organizations /V=19 .09 -.72* .02 .56»* -.15
Influence of other organization Extent of conflict Importance of other organization Compatibility of goals Bargaining in relationship 1.
tionships. In Table 5, the correlations pertaining to the interactions that were classified in Cell B are presented. This is the situation in which the employment service perceived low benefits to itself from interacting and the community organization perceived high benefits. If the hypotheses are supported here, positive correlations would be expected between the employment service's perceptions of (1) compatible goals (r = .60; p <.O1), (2) importance of the other organization (r = .56; p <.O5), (3) the amount of bargaining taking place (r = .31; n.s.), and (4) the amount of conflict in the relationship (r = .06; n.s.). Similarly, the relevant correlations for testing the hypotheses in Table 6 are the community organization's perceptions of (1) compatibility of goals (r = .38; p <.O5), (2) importance of the employment service (r = .28; n.s.), (3) influence of the employment service (r = .36; p <.O5), (4) the amount of bargaining that takes place (r = .33; p <.O5), and (5) the amount of conflict that occurs (r = - . 1 7 ; n.s.). Overall, the hypotheses are generally supported for the variables measuring influence, importance, bargaining, and goal compatibility; the measure of conflict, however, shows a much more complex relationship with frequency of interactions.
Table 6 Zero-Order Correlations between Frequency of Interaction and Independent Variables When Employment Service Perceives High Benefits and Community Organization Perceives Low Benefits (Cell C) Dimension As Perceived by Employment Service /V=34
.05
Community Organizations A/=34 .36** -.17
.28
1 nfluence of other organization Extent of conflict Importance of other organization Compatibility of goals Bargaining in relationship 1. 230/ASO
.52*
.15 .12 .15
.38*' .33*«
p<.05.
Interorganizational Relationships
One possible explanation for the lack of support for the hypotheses concerning the role of conflict in inducing interactions may be that the hypotheses failed to account adequately for the importance of differences in the balance of power within the employment service and community organization relationships. The data in Table 2 showed the employment service to be the more dominant organization in these relationships. The service is perceived by the community organizations as having more influence and being more important, and the service perceives less goal compatibility and more conflict in the relationships. If these characteristics are considered in terms of conflict theory (Schmidt and Kochan, 1972; Thomas, 1976; Robbins, 1974), it can be argued that the employment service is in a position to suppress conflict and therefore make conflict a useless strategy to the other organizations. This may occur when it is in the low benefit condition and also when it is in a high benefit condition and is able to force its will on the community organizations. The correlations on the conflict variable in Tables 5 and 6 are consistent with this interpretation. For example, there is a strong negative correlation between the community organization's perceptions of conflict and frequency of interaction when the employment service is in the low benefit condition (r = - . 7 2 ; p <.O1). This suggests that the employment service either suppresses conflict here or the community organization sees conflict as being futile, since it lacks sufficient power to achieve its goals through conflict. In Table 6, in the situation in which the employment service benefits from interaction, on the other hand, there is a strong positive correlation between the employment service's perception of conflict and frequency of interaction (r = .52; p <.O1). This would be consistent with the employment service using a forcing strategy since it generally is in the dominant power position. While these interpreta- tions are both plausible, given these correlations, and con- sistent with previous models of conflict behavior, it must be cautioned that they are post hoc interpretations of unanticipated findings and thus should be viewed only as hypotheses for further investigation. CONCLUSIONS AND IMPLICATIONS This article argues for the need to incorporate both the exchange and the power-dependency approaches of interorganizational relations into a single framework. While previous research has shown the validity of each of these approaches in different interorganizational fields, this analysis shows that both types of relationships can exist within the same interorganizational field or organizational set. Instead of focusing on dichotomous models of interorganizational relations, it is more fruitful to distinguish between patterns of relationships as determined by the relative strength of various dimensions of the relationship. The implication is that those interested in developing a theory of interorganizational relations from the perspective of the total field (Van de Ven, Emmett, and Koenig, 1974) might do well adopting a game theory perspective to in231/ASQ
teractions within the field that takes into account the presence of mixed motives (Walton and McKersie, 1965; Luce and Raiffa, 1957). Specifically, in attempting to predict behavior of organizations within the field, researchers will need to assume that members of organizations have selfinterests and they should focus on the costs and benefits to interactions with other organizations in terms of the organization's specific goals rather than the goals of the total field.3 This approach is consistent with Thompson's (1967) argument that organizations segment their relevant environments and adopt unique strategies for coping with each. If this view is correct, policy makers concerned with regulating or changing interorganizational relations will need to not only focus on the problems of coordination, but also on the distribution of power among organizations within the field. Furthermore, they may need to take a more active role in intervening in the processes of interactions between organizations to increase the probability of achieving policy results desired. In general, the data presented here suggest that asymmetrical relations pose difficult theoretical and policy problems for those concerned with interorganizational interactions and therefore warrant greater attention in future theoretical and empirical research.
For an example of how this self-interest approach can be useful in developing propositions in a specific interorganizational context, see the Kochan and Dyer (1976) model of organizational change in a union-management relationship.
Stuart M. Schmidt is an assistant professor in the School of Business Adnninistration at Temple University. Thomas A. Kochan is an assistant professor in the New York State School of Industrial and Labor Relations at Cornell University.
REFERENCES Aiken, Michael, and Jerald Hage 1968 "Organizational interdependence and intraorganizational structure." American Sociological Review, 33: 912-930. Aldrich, Howard E. 1976a "An interorganizational dependency perspective on relations between the employment service and its organization set." In Ralph H. Kilman, Louis R. Pondy, and Dennis P. Slevin (eds.). The Management of Organizational Design: 231-266, New York: North Holland. 1976b " Resou rce depe nd e nee a nd interorganizational relations." Administration and Society, 7: 419-455. Benson, J. Kenneth 1975 "The interorganizational network as a political economy." Administrative Science Quarterly, 20: 229-249. Blau, Peter M. 1964 Exchange and Power in Social Life. New York: Wiley. Emerson, Richard M. 1967 "Power-dependence relations." American Sociological Review, 27: 3 Evan, William 1966 "The organization set: toward a theory of interorganizational relations." In James D. Thompson (ed.), Approaches to Organizational Design, 173-191. Pittsburgh: University of Pittsburgh Press. Forehand, G. A., and B. Von Haller Gilmer 1964 "Environmental variations in the study of organizational behavior." Psychological Bulletin 62: 361-382. Giiiespie, David F., and Ronald W. Perry. 1975 "The influence of an organizational environment on interorganizational relations." American Journal of Economics and Sociology, 34: 21-A2. Guetzkow, Harold 1966 "Relations among organizations." In Raymond V. Bow232/ASQ ers (ed.). Studies on Behavior in Organizations: 13-44. Athens: University of Georgia Press. Hall, Richard H. 1972 Organizations: Structure and Process. Englewood Cliffs, N.J.: Prentice-Hall. Jacobs, David 1974 "Dependence and vulnerability: an exchange approach to the control of organizations." Administrative Science Quarterly, 19: 45-59. Kochan, Thomas A. 1975 "Determinants of the power of boundary units in an interorganizational bargaining relation." Administrative Science Quarterly, 20: 434-452. Kochan, Thomas A., George P. Huber, and L. L Cummings. 1975 "Determinants of intraorganizational conflict in collective bargaining in the public sector." Administrative Science Quarterly, 20: 10-23.
Interorganizational Relationships Kochan, Thomas A., and Lee Dyer. 1976 "A model of organizational change in the context of union-management relations." Journal of A|:^lied Behavior Science, 12: 59-78. Levine, Sol, and Paul White 1961 "Exchange as a conceptual framework for the study of interorganizational relationships." Administrative Science Quarterly, 5: 583-601. Lipsky, Michaei 1968 "Protest as a political resource." American Political Science Review, 2: 1144— 1158. Luce, R. Duncan, and Howard Raiffa 1957 Games and Decisions. New York: Wiley. March, James G., and Herbert A. Simon 1958 Organizations. New York: Wiley. Marrett, Cora B. 1971 "On the specification of interorganizational dimensions." Sociology and Social Research, 56: 83-99. Robbins, Stephen P. 1974 Managing Organizational Conflict. Englewood Cliffs, N.J.: PrenticeHall. Schmidt, Stuart M., and Thomas A. Kochan 1972 "The concept of conflict." Administrative Science Quarterly, 17:359-370. 1975 "An application of a'political economy' approach to effectiveness: employment service-employer exchanges." Administration and Society, 7: 455-^73. Simmel, Georg 1955 Conflict and the Web of Group-Affiliations. Trans, by K. H.Wolff and R.Bendix. New York: The Free Press. Stern, George G. 1970 People in Context: Measuring Person-Environment Congruence in Education and Industry. New York: Wiley. Thomas, K. W. 1976 "Conflict and conflict management." In M. D. Dunnette (ed.) The Handbook of Industrial and Organizational Psychology: 889-936. Chicago: Rand McNally. Thompson, James D. 1967 Organizations in Action. New York: McGraw-Hill. Thompson, James D., and William J. McEwen 1958 "Organizational goals and environment: goal-setting as an interaction process." American Sociological Review, 23: 23-31. Tuite, Matthew F. 1972 "Toward a theory of joint decision-making." In Matthew Tuite, Roger Chisholm, and Michael Radnor, Interorganizational Decision Making: 9-19. Chicago: Aldine. Turner, Ralph H. 1969 "The public perception of protest." American Sociological Review. 34: 815-836. Van de Ven, Andrew H., Dennis Emmett, and Richard Koenig, Jr. 1974 "Frameworks for interorganizational analysis." Organization and Administrative Science, 1: 113-129. Walton, Richard E. 1965 "Two strategies of social change and their dilemmas." Journal of Applied Behavioral Science. 1: 167-179. Walton, Richard E., and Robert B. McKersie 1965 A Behavioral Theory of Labor Negotiations. New York: McGraw-Hill. Wamsley, G. L., and M. N. ZaId 1976 "The political economy of public organizations." Public Administration Review, 33: 62-72. White, Paul E. 1974 "Intra-and interorganizational studies." Administration and Society, 6: 105-152. Wilson, James Q. 1973 Political Organization. New York: Basic Books. Zeitz, Gerald 1974 "Interorganizational relationships and social structure: a critique of some aspects of the literature." Organization and Administrative Sciences, 5: 131-139.
APPENDIX A Study of the Employment Service These questions examine the nature of the relationship between the State Employment Service local office and your organization. For each question, circle the number which best describes your opinion. 1. How frequent is there contact between your organization and the employment service local office? Almost never Daily 1 2 3 4 5 6 7 2. To what extent do you feel that your organization benefits in attaining its goals as a result of its interaction with the employment service local office? Not at all 1 Very great extent 7
2
3
4
5
6
3. To what extent are there agreements between your organization and the employment service local office? There is no agreement 1 233/ASQ Informal agreement 2 Formal agreement 3
4. How are the terms of your organization's interaction with the employment service local office reached? The terms are set: Completely Mostly by Mostly Completely by them them MutuaHy by us by us 1 2 3 4 5 5. To what extent does tension exist between your organization and the employment service local office? None A great deal 1 2 3 4 5 6 7 6. When representatives of your organization meet with the employment service to discuss issues of mutual concern, how much influence does the employment service have on the decisions reached? None Very Much 1 2 3 4 5 6 7 7. To what extent has the employment service local office attempted to interfere with the attainment of the goals of your organization i n areas of overlapping interest? Never Frequently 1 2 3 4 5 6 7 8. How important is the employment service local office for the operations of your organization? Not Extremely important important 1 2 3 4 5 6 7 9. To what extent are the goals of the employment service local office (as you understand them) compatible with the goals of your organization? Not Extremely compatible compatible 1 2 3 4 5 6 7 10. To what extent to you see your organization's relationship with the employment service as a form of bargaining? Not at Very all much 1 2 3 4 5 6 7 rhankyou for your cooperation. APPENDIX B Comparison of Sample with Subsample of Data Averaged by Organization Type Frequency of Interaction Subsample Data Averaged by Organization Type in a District Single Organization Data A/ = 136 X 4.57 S.D. 1.28 Community Organization Benefits X 4.38 S.D. 1.48 Employment Service Benefits X 4.50 S.D. 1.51
4.31
1.44
4.36
1.91
4.00
1.91
F=.327,p=.57 F=.002,p=.97
F=.716,p=.4O
234/ASQ
doc_959400427.docx
Organization–public relationships is a key concept in public relations research. The emergence of relationship management theory as a new paradigm for public relations increases scholars’ argument about the essence of public relations—what it is and what is its value within the organizations and to the greater society.
Research on Inter-organizational Relationships: Patterns and Motivations
Two connpeting approaches—exchange and powerdependency—have been used in previous studies to conceptualize interorganizational relationships. The basic premise of this paper is that an integrated view of these relationships is needed that incorporates propositions from both of these approaches. Propositions are developed from these two approaches for explaining varia- tions in the frequency of interaction and the nature of interactions between pairs of organizations in an interor- ganizational set. Special emphasis is given to the nature of interactions in an asymmetrical relationship. The propositions are tested using data on interactions between community organizations and local offices ofthe United States Training and Employment Service. The results suggest that interorganizational relationships should be conceptualized as a mixed-motive situation in which each organization behaves in accordance with its own self-interests.*
In the past decade, two approaches to the study of interorganizational relationships have emerged. There are those on the one hand, who view interorganizational relation- ships from an exchange perspective (Levine and White, 1961; Tuite, 1972; White, 1974).''According to this approach, relations form when members of two or more organizations perceive mutual benefits or gains from interacting. The leadership of each organization is internally moti- vated to form the relationship because each perceives that it will be better able to attain its goals by interacting than by remaining autonomous. Aiken and Hage (1968) suggested that this type of relationship forms in periods of scarce or declining resources. In addition to offering an explanation forthe motivations to form the connection, the exchange approach also implies that the nature of the in- teractions between participants in these relationships is characterized by a high degree of cooperation and problem solving as opposed to high levels of conflict and bargaining (Geutzkow, 1966; March and Simon, 1958; Walton, 1965), since the parties are both motivated to coordinate their efforts to maximize joint benefits. Others, however, have adopted a power-dependency approach to interorganizational relations. Here, it is implied that the motivation to interact is asymmetrical. One party is motivated to interact, but the other is not. The interorgani- zational relationship only forms when the motivated party is powerful enough to force or induce the other to interact. The motivation or pressure to interact for at least one of the parties is externally generated, since one party would, if free to choose, prefer to remain autonomous. Bargaining and conflict are the natural form of interaction in this rela- tionship, since each party seeks to attain its own goals at the expense of the other (March and Simon, 1958; Geutzkow, 1966; Schmidt and Kochan, 1972; Aldrich, 1976a).
Unfortunately, these alternative approaches are developing along separate, but parallel, paths. Zeitz (1974), for example, noted that much empirical research has focused on social
220/Administrative Science Quarterly
interorganizationai Reiationshq>s
service delivery systems where the emphasis has been on the goals of achieving coordination and cooperation. Implicitly, therefore, this stream of research has employed the exchange perspective. This has led researchers using this approach to pay little attention to variables that affect the distribution of power within the relationship. In contrast, Benson (1975) focused on the political-economic substructure of interorganizational relations and therefore implicitly adopted a power-dependency approach. Similarly, a recent study of the conditions under which employers are induced to use the United States Training and Employment Service (Schmidt and Kochan, 1976) and another of the conditions under which employers reallocate power within their decision-making structures to deal with trade unions (Kochan, 1975) also fit within the power-dependency perspective. In both studies, such variables as distribution of power, differences in goals, and conflict played a dominant role and organizations with the low intrinsic motivation to interact—in each case the employer—were influenced by the amount of power the other organization brought to bear on the relationship. The separate development of these two competing perspectives is unfortunate, since interorganizational life is not as simple as these distinctions might suggest. No organization is likely to engage in only symmetrical exchange relationships with other organizations or only powerdependency relationships. Rather, an organization is likely to have a mixture of each type within its organization set. Furthermore, no single relationship is likely to be either of these pure or ideal types, but rather one in which there are mixed motives (Walton and McKersie, 1965). The basic premise of this article is that an integrated view of interorganizational relations is needed that is capable of incorporating the basic propositions from both the exchange and the power-dependency models. This view would recognize that relations that are both predominantly symmetrical exchanges and those that are asymmetrical and power-dependency oriented can exist within the same organizational set. Further, overtime, a given relationship between two organizations may shift from being largely one type to the other. Thus, this research attempts to move in the direction of identifying the conditions under which the exchange and the power-dependency approaches can be appropriately applied (Aidrich, 1976). These approaches are incorporated here into a study of the frequency of interactions between community organizations and district offices of the United States Training and Employment Service. The first objective of the study is to determine empirically the characteristics of the interactions between organizations in which each of the above perspectives is predominant. The second objective is to distinguish empirically symmetrical from asymmetrical relationships between the focal organization and members of its organizational set. The third objective is to explore variations in the frequency of interactions which take place under these two conditions, with special emphasis on variations under asymmetrical conditions.
221/ASQ
An explanation of the differences in the frequency of interactions is important not only for the development of theory in this area, but also for public policy. Successful interorganizational relationships are critical to the implementation of almost all public policies (Wamsley and Zaid, 1974). Consequently, before implications can be derived for the design and administration of public policies involving multiple organizations, the factors leading to the development and maintenance of interorganizational relationships must be identified.
HYPOTHESES The following specific hypotheses are tested in this research:
H ^ The lowest frequency of interaction between two organizations will occur when neither organization perceives benefits from interacting—a symmetrical situation. H 2 The greatest frequency of interaction will occur when both organizations perceive benefits from interacting—a symmetrical situation. H 3 In an asymmetrical situation—one organization perceives benefits from interacting and the other does not, the frequency of interaction will be greater than when neither organization perceives benefits and less than when both see benefits. H 4 In an asymmetrical situation the frequency of interaction between two organizations will be higher when the organization perceiving low benefits from interacting perceives the other organization as (a) having goals that are compatible with its own organization, (b) being important to the functioning of its own organization, (c) having greater influence over its organization, and (d) acting aggressively in pursuing its interests by using bargaining and conflict-oriented tactics of influence.
The first two hypotheses are derived from the exchange model of interorganizational relationships (Levine and White, 1961) and argue that the perception of benefits resulting from a relationship will be a good predictor of the frequency of interaction between two organizations. Flowing from the assumption that individuals in organi- zations act in their self-interest, it is predicted that the highest and lowest frequencies of interaction will occur in symmetrical situations in which mutual benefits or mutual lack of benefits are perceived. When members of two organizations perceive benefits from interacting with each other, the mutual motivation for interaction will be great. Similarly, the reverse is posited. Motivations for interaction will be low if parties in neither organization perceive benefits as a result. The third hypothesis suggests that it will be more difficult to initiate and maintain high frequencies of interaction between two organizations when members of only one perceive benefits resulting from the relationship. Likewise, it is more difficult to predict the frequency of interaction when organizations are in this asymmetrical situation. Often, however, this is the rela-
tion which exists between organizations of a given set. Thus, many organizations seek to increase the frequency of interaction with organizations not motivated to do so. Hypotheses 4(a) through (d) focus on situations in which individuals in an organization not perceiving
222/ASQ
Interorganizational Relationships
benefits from interacting with those of another organization who do—^the asymmetrical situation—are most likely to interact anyway. In this situation, it is proposed that a power-dependency framework will be most useful in predicting frequency of interacting. This set of hypotheses focuses on concepts central to a political model of organizational behavior, namely, goals, power, bargaining, and conflict (Wilson, 1973; Benson, 1975). Hypothesis 4 (a) proposes that organizational interaction in the asymmetrical benefit situation will be most probable if the members of the organization perceiving few benefits perceive the other's goals to be compatible with its own. Compatible goals as a condition for interaction has been previously developed by Evan (1966) and Guetzkow (1966). In this situation, those in the organization seeing few benefits also perceive little threat in the relationship; therefore, they do not seek to resist attempts to interact. Essentially, the only costs of interacting are opportunity costs, with time, energy, and other resources being diverted from alternative uses. Hypotheses 4 (b) through (d) are also derived from a power-dependency framework of interorganizational relations in that they consider factors affecting the balance of power between the organizations in the relationship. These hypotheses propose that parties in an organization who perceive low benefits from interacting with those in another may be forced to interact because of a dependency relationship. Thus, hypotheses 4 (b) and 4 (c) specifically propose that the more important (Jacobs, 1974; Aldrich, 1976b; Giiiespie and Perry, 1975) and influential (Emerson, 1962, Thompson, 1967) the other organization is to those perceiving few benefits, the more frequent the interaction. For example, some funding agencies require the formation of practitioner advisory committees as a condition of their research grants. Thus, even though the goals of the researchers and the practitioners may be different, the importance of the practitioners to the researcher requires that such a committee be formed and meet in accordance with the contract terms. Hypothesis 4 (d) proposes that the asymmetrical relationship is analogous to problems faced by pressure groups (Turner, 1968; Lipsky, 1969). That is, people in one organization will interact with those in another in an asymmetrical benefit relationship to the extent that those in the other organization aggressively pursue their interests by engaging in bargaining and pressure-oriented tactics of influence. Additionally, interaction will occur to the extent that the people in the organization receiving the benefit pursue their goals through the use of conflict-oriented strategies (Thompson and McEwen, 1958; Schmidt and Kochan, 1972; Thomas, 1976). While these aggressive tactics are probably not the first means of exerting influence that members of an organization will try, continued frustration with other means of influence will eventually lead to the use of these methods to achieve their goals, especially if the perceived benefits from interaction are strong enough to warrant taking the risks associated with this behavior. Perhaps the best
223/ASQ
examples of interorganizational relationships of this type are trade unions seeking recognition from an employer, inmate associations seeking to negotiate with prison officials, or protest groups seeking to influence poiiticians or community organizations. Hypotheses 1, 2, and 3 are illustrated schematically in the Figure. As shown, there are four possible conditions of interaction contingent on whether the employment service or community organizations perceive high or low benefits from their interaction.
Employment Service Benefits High Community Organization Benefits High
A
Low
B
Low
C
D
Cell A = Mutual benefit voluntary interaction. High level of interaction predicted. Cell D = No or low interdependence; therefore, no or low interaction. Cell B = Asymmetrical condition where community group benefits, but employment service does not. Intermediate kind of interaction. Interaction is dependent on the community group's ability to force employment service to interact. Cell C = Reverse condition of Cell B.
Figure. Conditions of Interaction.
METHODOLOGY Focal Organizations The data for this study were obtained from district offices of the United States Training and Employment Service and those community organizations which had an ongoing reiationship with them. The State Employment Service is a part of the United States Training and Employment Service of the United States Department of Labor and is the primary public manpower delivery system in the state. Although it receives most of its operating funds and primary policy direction from the federal agency, it is administered by the state. The district offices perform a number of activities which essentially involve (1) matching individuals seeking employment with job vacancies filed with the office by employers, (2) assisting job applicants become employable, (3) assisting employers fill job vacancies, and (4) providing labor market information. This study is based on the relationships between 23 employment service district offices in a midwestern state and the primary organizations interacting with each district office. These data were collected during the winter of 1972-1973. Members of the Organizational Sets The composition of the employment service district offices' organizational set was determined on the basis of commonality of clients with community organizations in a district and relationships between these organizations and the
224/ASQ
Interorganizational Relationships
Table 1
Organizational Set Categories Type of Organization AFL-CIO affiliated unions Non-AFL-CIO affiliated unions Chamber of commerce Community action program Concentrated employment program office Division of vocational rehabilitation Equal employment opportunity commission M i nority i nterest organizations National Alliance of Businessmen Probation and parole office Public and vocational schools Social welfare service Veterans of Foreign Wars Total Number
2 7 19 15 3 16 7 8 2 20 12 15 20 146
Percent of Sample
1.4 4.8
13.0 10.3
2.1
11.0
4.8 5.5 1.4
13.7
8.2
10.3 13.7 100.0
district offices. From initial interviews with directors of the district offices, a set of 13 organizations significant to the functioning of these offices was specified. They are shown in Table 1. The criteria for including a community organization in the district office organizational set were twofold: (a) clients of each of the set organizations were or could be recipients of services offered by the district offices, and (b) each organization in the set had some degree of interaction with the district offices. If both of these criteria were met, the organization was included in the organizational set of the employment service district offices.^ In most of the districts, the 13 organizations shown in Table 1 interacted with the district office in that area.
Data
The resource limitations of this study did not allow for a sampling of a sufficient number of employers in each district to include in the analysis.
Two basic sets of data were obtained to test the hypotheses in this study. First, from the community organizations, data were obtained which described their interactions with employment service district offices. Second, a data set was obtained from the district offices describing interactions with the community organizations. Directors of each community organization and corresponding district office received a self-report questionnaire measuring the 10 dimensions of an organizational relationship used in this study. Each of the questionnaire items focused on an aspect of the relationship between a community organization and district office. The respective directors indicated their responses to the items on seven-point scales anchored at the extremes with polar adjectives. The specific dimensions measured were derived from Marrett (1971) and supplemented with additional items designed to capture dimensions of conflict processes, namely, incompatible goals, tension, conflict, and influence (Schmidt and Kochan, 1972; Kochan, Huber, and Cummings, 1975). For the specific items used in this study, see Appendix A. The use of perceptual data was deemed appropriate in keeping
225/ASQ
with similar research in the area of interorganizational relationships (Aiken and Hage, 1968). Additionally, the data were collected from top decision makers in each of the organizations sampled and their behavior is important in guiding the direction of their respective organizations. Since it is assumed that their behavior is a function of their perceived environment (Forehand and Gilmer, 1964; Stern, 1970), perceptual data is meaningful in this context. Overall, the response rates for the survey questionnaires sent to the directors of the community organizations were respectable. After a single mail follow-up, 157 usable questionnaires were obtained, representing a 78.8 percent response rate. Additionally, all 23 personally administered questionnaires from the employment service district office directors were completed. These directors responded to questions concerning each of the 13 types of community organizations in their districts. To have only one response set for each type of community organization in an employment service district, the scores for the same type of organizations in a district were averaged. This averaging occurred in 21 cases. To determine if the averaged scores were significantly different from the nonaveraged scores across community organizations, a comparison of the two were made. This comparison between the single organizational reports (A/ = 136) and the subsample of averaged reports (A/ = 10) showed no significant difference between mean scores on the dimensions examined (see Appendix B). Therefore, the averaged reports and nonaveraged reports were combined. The final data set consisted of 146 paired relationships between community organizations and employment service district offices. Dimensions of these 146 paired relationships were used forthe analyses in this study.
RESULTS Perceptions of the Relationships
The mean and standard deviations on each of the 10 dimensions measured across the 146 paired relationships are reported in Table 2. On 5 of the dimensions there is a significant difference between an employment service district office and respective community organization. The four most significant reported differences between the organizations exist over (a) the influence of the other organization, (b) the degree of compatible goals, (c) the extent of conflict and, (d) the importance of the other organization. The district office perceived the community organization to be less important to its operation than the reverse. Likewise, employment service district office directors reported that community organizations had less influence on the office's operation. The district office directors, however, reported a lower level of compatible goals and more conflict than did the community organization directors. In addition, employment service directors perceived fewer benefits to their organizations from interacting with the community organizations than vice versa; the difference, though, is not statistically significant at the conventional level. Thus, overall, the employment service is not only the
226/ASQ
Interorganizational Relationships Table 2 Perceptions of Paired Relationships Dimension As Perceived by Community organization N= 146 Mean 4.24 4.36 1.82 2.99 1.90 4.38 1.66 4.59 5.42 2.47 Standard deviation 1.76 1.88 .65 .70 1.54 1.53 1.20 2.05 1.62 1.91 Employment service/V= 146 Mean 4.42 4.01 2.12* 2.94 1.92 3.39* 2.08'* 3.89** 4.98** 2.84— Standard deviation 1.65 1.88 .72 .46 1.17 1.70 1.35 1.88 1.86 1.74
Frequency of interaction Benefits from interaction Formalization of agreements Setting terms of interaction Extent of tension Influence of other organization Extent of conflict Importance of other organization Compatibility of goals Bargaining in relationship
p=.OO1.
p=.O1
p = .O8.
more powerful and influential organization in these relationships, but also more resistant to dealing with these community organizations than the reverse. These data are consistent with the data reported on another sample of employment service-community organization relationships (Aldrich, 1976a). Dimensions of Interorganization Relations Before variations in the frequency of interaction across these relationships are examined, the results of a series of factor analyses are presented. These analyses were performed to assess the validity of a distinction used in the literature between cooperative and conflictful relationships. Although Simmel (1955) discussed the duality of conflict and cooperation in all relations, Guetzkow's (1966) classic article spoke about this distinction in connection with interorganizational relationships. No empirical evidence, however, has yet been presented to assess its utility for distinguishing among relationships. As noted earlier, prior empirical research has emphasized either cooperative interactions or those marked by conflict in the models used to study relations within an organizational field. If the central arguments made in Guetzkow (1966), Aldrich (1976 b), and in this study are correct, both cooperative and conflictual patterns of interaction within a single organizational set should emerge. Thus, the factor analysis reported here addresses two issues: (1) does the distinction between cooperative exchanges involving mutual benefits versus conflict-power perspectives make sense empirically as a typology, and (2) can both types of relationships be observed within the same organizational set? The factor analytic procedure consisted of (1) constructing a correlation matrix for all of the interactions, (2) factoring with a principle axis solution, and (3) rotating the resulting factors with a varimax procedure after rotation with Kaiser normalization.
227/ASQ
As shown in Table 3, the factor structures of the paired relationships as perceived by community organizations and employment service district offices are virtually equivalent. On this basis, the factor structure of the combined organizational data is interpreted. Only factor loadings of .4 and above on the dimensions are considered. The dimensions loading on Factor 1 include frequency of interaction, benefiting from the interaction, extent of formalized agreements, influence of the other organization, importance of the other organization, and the compatibility of goals between the organizations. This factor describes a type of cooperative relationship in which members of the organization are intrinsically motivated to interact, because they perceive benefits from the interaction. Factor 1 accounts for 33.5 percent of the total variance. Factor 2, on the other hand, describes a conflictual relationship between district offices and community organizations. This relationship is indicated by loadings on tension, conflict, bargaining, and formalization of agreements. Factor 2 accounts for 22.4 percent of the total variance. Factor 3 accounts for 9.9 percent of the total variance. Only one variable, however, loads at the .4 level on this factor, namely, the amount of reciprocity in setting the terms of interaction between the district office and the community organization. These three factors account for 66 percent of the total variance forthe combined sample. Thus, these findings indicate that the cooperation-conflict distinction is useful. Relationships, however, can be characterized as being cooperative and marked by conflict within the same interorganizational network. The degree to which agreements are formalized is associated with both conflict and cooperation. Therefore, the concepts of formalization and reciprocity do not discriminate between these two dimensions and are not included in the remainder of the analysis.
Frequency of Interaction
It was predicted that the lowest frequency of interaction
Table 3
Factor Structure of Paired Relationships
Item Factor 1 Community Organizations r/V=146) Frequency of interaction Benefits fram interaction Formalization of agreements Setting terms of interaction Extent of tension Influence of other organizations Extent of conflict Importance of other organizations Compatibility of goals Bargaining in relationship Percent of total variance Eigenvalue .784 .829 .488 .067 -.167 .675 .-.168 .836 .787 .207 34.3 3.43 Factor 2 Community EmployOrganizament Service Combined tions (yv=146) (/V=292) .791 .793 .511 .021 Oil .729 -.178 .863 .848 .268 37.1 3.71 .836 .836 .619 .043 -.226 .629 -.053 .784 .627 .192 33.5 3.35 .184 -.004 .503 .108 .807 -.059 .830 -.026 -.112 .643 20.7 2.07 Employment Service Factor 3 Community Organizations Employment Service .036 -.109 -.023 .997 .048 .057 -.116 .021 .070 .017 10.1 1.01
Combined .129 -.133 .467 .153 .848 -.011 .847 -.135 -.404 .552 22.4 2.24
Combined -.104 .162 -.131 .979 .043 -.008 .083 .185 .175 .340 9.9 1.00
.120 .024 .423 -.031 .824 -.073 .830 .052 .066 .636 19.0 1.90
-.106 -.014 -.176 .959 .076 .094 -.033 .103 .069 .146 10.0 1.00
228/ASQ
interorganizational Relationshqis
between members of two organizations would occur where neither perceived benefits from their interaction. On the other hand, the greatest frequency of interaction would take place when members of both organizations had mutual perceptions of benefits resulting from their interaction. The community organization and employment service scores on the dimension of frequency of interaction were classified according to whether the organizations reported high or low benefits from the interaction. The dependent variable, frequency of interaction, is the average of the responses of the community organization and employment service in each relationship. By averaging scores across both organizations, perceptual bias of either respondent is reduced. The scores on the interaction dimension were divided at the median, with the results shown in Table 4. A Chi square analysis of these four cells indicates a systematic variation in the data (X2=9.29, 1; p <.002).
Table 4 Frequency of Interaction by Perception of Who Benefits
Community Organization Perceives
Employment Service Perceives High benefits X = 5.31,S.D.=1.14 Low benefits X=2.21,S.D. =1.69 (B) X=4.58 S.D. =1.26 N =19 X = 3.02
S.D.= 1.07 N =40
A/=59
High benefits X=6.04,S.D.=.82 Low benefits X=2.78,S.D.=1.01
X = 5.28 S.D.= .93 N =50 X= 4.31 S.D.= 1.40 N -34 /V=84 58.7%
(A)
N=7A 48.3%
A/=74
(C)
(D)
51.7%
A; = 1 4 3
41.3%
100%
X2=9.29with 1 df.p =.0023.
The mean value for frequency of interaction in the mutually high benefits condition is 5.28, while the mean value on this same dimension in the mutually low benefits condition is 3.02. These mean values are significantly different from each other and are in the direction predicted by hypotheses H 1 and H 2 (F=29.18; 3,138; p <.OOO1). Likewise, the mean value of Cell A (5.28) is significantly different from the mean values of Cells B (4.58) or C (4.31). The mean value Cell D (3.02) is significantly different than the mean values of Cells B and C. Therefore, these data support hypotheses H ^ through H 3. Interactions under Asymnnetrical Conditions Hypotheses 4 (a) through (d) predicted that in the asymmetrical benefits condition, frequency of interaction would be high when the low benefit organization perceived the other organization as having compatible goals, high influence and importance, and engaging in aggressive pressure tactics of bargaining and conflict. The correlations used to test these hypotheses are presented in Tables 5 and 6. These correlations are based on the data from the relationships that were classified into Cells B and C, the asymmetrical rela22 9/ASQ
Table 5 Zero-Order Correlations between Frequency off Interaction and Independent Variables When Employment Service Perceives Low Benefits and Community Organization Perceives High Benefits (Cell B) Dimension As Perceived by Employment Service /V=19 .64* .06 .56*» .60* .31 p<.05. Community Organizations /V=19 .09 -.72* .02 .56»* -.15
Influence of other organization Extent of conflict Importance of other organization Compatibility of goals Bargaining in relationship 1.
tionships. In Table 5, the correlations pertaining to the interactions that were classified in Cell B are presented. This is the situation in which the employment service perceived low benefits to itself from interacting and the community organization perceived high benefits. If the hypotheses are supported here, positive correlations would be expected between the employment service's perceptions of (1) compatible goals (r = .60; p <.O1), (2) importance of the other organization (r = .56; p <.O5), (3) the amount of bargaining taking place (r = .31; n.s.), and (4) the amount of conflict in the relationship (r = .06; n.s.). Similarly, the relevant correlations for testing the hypotheses in Table 6 are the community organization's perceptions of (1) compatibility of goals (r = .38; p <.O5), (2) importance of the employment service (r = .28; n.s.), (3) influence of the employment service (r = .36; p <.O5), (4) the amount of bargaining that takes place (r = .33; p <.O5), and (5) the amount of conflict that occurs (r = - . 1 7 ; n.s.). Overall, the hypotheses are generally supported for the variables measuring influence, importance, bargaining, and goal compatibility; the measure of conflict, however, shows a much more complex relationship with frequency of interactions.
Table 6 Zero-Order Correlations between Frequency of Interaction and Independent Variables When Employment Service Perceives High Benefits and Community Organization Perceives Low Benefits (Cell C) Dimension As Perceived by Employment Service /V=34
.05
Community Organizations A/=34 .36** -.17
.28
1 nfluence of other organization Extent of conflict Importance of other organization Compatibility of goals Bargaining in relationship 1. 230/ASO
.52*
.15 .12 .15
.38*' .33*«
p<.05.
Interorganizational Relationships
One possible explanation for the lack of support for the hypotheses concerning the role of conflict in inducing interactions may be that the hypotheses failed to account adequately for the importance of differences in the balance of power within the employment service and community organization relationships. The data in Table 2 showed the employment service to be the more dominant organization in these relationships. The service is perceived by the community organizations as having more influence and being more important, and the service perceives less goal compatibility and more conflict in the relationships. If these characteristics are considered in terms of conflict theory (Schmidt and Kochan, 1972; Thomas, 1976; Robbins, 1974), it can be argued that the employment service is in a position to suppress conflict and therefore make conflict a useless strategy to the other organizations. This may occur when it is in the low benefit condition and also when it is in a high benefit condition and is able to force its will on the community organizations. The correlations on the conflict variable in Tables 5 and 6 are consistent with this interpretation. For example, there is a strong negative correlation between the community organization's perceptions of conflict and frequency of interaction when the employment service is in the low benefit condition (r = - . 7 2 ; p <.O1). This suggests that the employment service either suppresses conflict here or the community organization sees conflict as being futile, since it lacks sufficient power to achieve its goals through conflict. In Table 6, in the situation in which the employment service benefits from interaction, on the other hand, there is a strong positive correlation between the employment service's perception of conflict and frequency of interaction (r = .52; p <.O1). This would be consistent with the employment service using a forcing strategy since it generally is in the dominant power position. While these interpreta- tions are both plausible, given these correlations, and con- sistent with previous models of conflict behavior, it must be cautioned that they are post hoc interpretations of unanticipated findings and thus should be viewed only as hypotheses for further investigation. CONCLUSIONS AND IMPLICATIONS This article argues for the need to incorporate both the exchange and the power-dependency approaches of interorganizational relations into a single framework. While previous research has shown the validity of each of these approaches in different interorganizational fields, this analysis shows that both types of relationships can exist within the same interorganizational field or organizational set. Instead of focusing on dichotomous models of interorganizational relations, it is more fruitful to distinguish between patterns of relationships as determined by the relative strength of various dimensions of the relationship. The implication is that those interested in developing a theory of interorganizational relations from the perspective of the total field (Van de Ven, Emmett, and Koenig, 1974) might do well adopting a game theory perspective to in231/ASQ
teractions within the field that takes into account the presence of mixed motives (Walton and McKersie, 1965; Luce and Raiffa, 1957). Specifically, in attempting to predict behavior of organizations within the field, researchers will need to assume that members of organizations have selfinterests and they should focus on the costs and benefits to interactions with other organizations in terms of the organization's specific goals rather than the goals of the total field.3 This approach is consistent with Thompson's (1967) argument that organizations segment their relevant environments and adopt unique strategies for coping with each. If this view is correct, policy makers concerned with regulating or changing interorganizational relations will need to not only focus on the problems of coordination, but also on the distribution of power among organizations within the field. Furthermore, they may need to take a more active role in intervening in the processes of interactions between organizations to increase the probability of achieving policy results desired. In general, the data presented here suggest that asymmetrical relations pose difficult theoretical and policy problems for those concerned with interorganizational interactions and therefore warrant greater attention in future theoretical and empirical research.
For an example of how this self-interest approach can be useful in developing propositions in a specific interorganizational context, see the Kochan and Dyer (1976) model of organizational change in a union-management relationship.
Stuart M. Schmidt is an assistant professor in the School of Business Adnninistration at Temple University. Thomas A. Kochan is an assistant professor in the New York State School of Industrial and Labor Relations at Cornell University.
REFERENCES Aiken, Michael, and Jerald Hage 1968 "Organizational interdependence and intraorganizational structure." American Sociological Review, 33: 912-930. Aldrich, Howard E. 1976a "An interorganizational dependency perspective on relations between the employment service and its organization set." In Ralph H. Kilman, Louis R. Pondy, and Dennis P. Slevin (eds.). The Management of Organizational Design: 231-266, New York: North Holland. 1976b " Resou rce depe nd e nee a nd interorganizational relations." Administration and Society, 7: 419-455. Benson, J. Kenneth 1975 "The interorganizational network as a political economy." Administrative Science Quarterly, 20: 229-249. Blau, Peter M. 1964 Exchange and Power in Social Life. New York: Wiley. Emerson, Richard M. 1967 "Power-dependence relations." American Sociological Review, 27: 3 Evan, William 1966 "The organization set: toward a theory of interorganizational relations." In James D. Thompson (ed.), Approaches to Organizational Design, 173-191. Pittsburgh: University of Pittsburgh Press. Forehand, G. A., and B. Von Haller Gilmer 1964 "Environmental variations in the study of organizational behavior." Psychological Bulletin 62: 361-382. Giiiespie, David F., and Ronald W. Perry. 1975 "The influence of an organizational environment on interorganizational relations." American Journal of Economics and Sociology, 34: 21-A2. Guetzkow, Harold 1966 "Relations among organizations." In Raymond V. Bow232/ASQ ers (ed.). Studies on Behavior in Organizations: 13-44. Athens: University of Georgia Press. Hall, Richard H. 1972 Organizations: Structure and Process. Englewood Cliffs, N.J.: Prentice-Hall. Jacobs, David 1974 "Dependence and vulnerability: an exchange approach to the control of organizations." Administrative Science Quarterly, 19: 45-59. Kochan, Thomas A. 1975 "Determinants of the power of boundary units in an interorganizational bargaining relation." Administrative Science Quarterly, 20: 434-452. Kochan, Thomas A., George P. Huber, and L. L Cummings. 1975 "Determinants of intraorganizational conflict in collective bargaining in the public sector." Administrative Science Quarterly, 20: 10-23.
Interorganizational Relationships Kochan, Thomas A., and Lee Dyer. 1976 "A model of organizational change in the context of union-management relations." Journal of A|:^lied Behavior Science, 12: 59-78. Levine, Sol, and Paul White 1961 "Exchange as a conceptual framework for the study of interorganizational relationships." Administrative Science Quarterly, 5: 583-601. Lipsky, Michaei 1968 "Protest as a political resource." American Political Science Review, 2: 1144— 1158. Luce, R. Duncan, and Howard Raiffa 1957 Games and Decisions. New York: Wiley. March, James G., and Herbert A. Simon 1958 Organizations. New York: Wiley. Marrett, Cora B. 1971 "On the specification of interorganizational dimensions." Sociology and Social Research, 56: 83-99. Robbins, Stephen P. 1974 Managing Organizational Conflict. Englewood Cliffs, N.J.: PrenticeHall. Schmidt, Stuart M., and Thomas A. Kochan 1972 "The concept of conflict." Administrative Science Quarterly, 17:359-370. 1975 "An application of a'political economy' approach to effectiveness: employment service-employer exchanges." Administration and Society, 7: 455-^73. Simmel, Georg 1955 Conflict and the Web of Group-Affiliations. Trans, by K. H.Wolff and R.Bendix. New York: The Free Press. Stern, George G. 1970 People in Context: Measuring Person-Environment Congruence in Education and Industry. New York: Wiley. Thomas, K. W. 1976 "Conflict and conflict management." In M. D. Dunnette (ed.) The Handbook of Industrial and Organizational Psychology: 889-936. Chicago: Rand McNally. Thompson, James D. 1967 Organizations in Action. New York: McGraw-Hill. Thompson, James D., and William J. McEwen 1958 "Organizational goals and environment: goal-setting as an interaction process." American Sociological Review, 23: 23-31. Tuite, Matthew F. 1972 "Toward a theory of joint decision-making." In Matthew Tuite, Roger Chisholm, and Michael Radnor, Interorganizational Decision Making: 9-19. Chicago: Aldine. Turner, Ralph H. 1969 "The public perception of protest." American Sociological Review. 34: 815-836. Van de Ven, Andrew H., Dennis Emmett, and Richard Koenig, Jr. 1974 "Frameworks for interorganizational analysis." Organization and Administrative Science, 1: 113-129. Walton, Richard E. 1965 "Two strategies of social change and their dilemmas." Journal of Applied Behavioral Science. 1: 167-179. Walton, Richard E., and Robert B. McKersie 1965 A Behavioral Theory of Labor Negotiations. New York: McGraw-Hill. Wamsley, G. L., and M. N. ZaId 1976 "The political economy of public organizations." Public Administration Review, 33: 62-72. White, Paul E. 1974 "Intra-and interorganizational studies." Administration and Society, 6: 105-152. Wilson, James Q. 1973 Political Organization. New York: Basic Books. Zeitz, Gerald 1974 "Interorganizational relationships and social structure: a critique of some aspects of the literature." Organization and Administrative Sciences, 5: 131-139.
APPENDIX A Study of the Employment Service These questions examine the nature of the relationship between the State Employment Service local office and your organization. For each question, circle the number which best describes your opinion. 1. How frequent is there contact between your organization and the employment service local office? Almost never Daily 1 2 3 4 5 6 7 2. To what extent do you feel that your organization benefits in attaining its goals as a result of its interaction with the employment service local office? Not at all 1 Very great extent 7
2
3
4
5
6
3. To what extent are there agreements between your organization and the employment service local office? There is no agreement 1 233/ASQ Informal agreement 2 Formal agreement 3
4. How are the terms of your organization's interaction with the employment service local office reached? The terms are set: Completely Mostly by Mostly Completely by them them MutuaHy by us by us 1 2 3 4 5 5. To what extent does tension exist between your organization and the employment service local office? None A great deal 1 2 3 4 5 6 7 6. When representatives of your organization meet with the employment service to discuss issues of mutual concern, how much influence does the employment service have on the decisions reached? None Very Much 1 2 3 4 5 6 7 7. To what extent has the employment service local office attempted to interfere with the attainment of the goals of your organization i n areas of overlapping interest? Never Frequently 1 2 3 4 5 6 7 8. How important is the employment service local office for the operations of your organization? Not Extremely important important 1 2 3 4 5 6 7 9. To what extent are the goals of the employment service local office (as you understand them) compatible with the goals of your organization? Not Extremely compatible compatible 1 2 3 4 5 6 7 10. To what extent to you see your organization's relationship with the employment service as a form of bargaining? Not at Very all much 1 2 3 4 5 6 7 rhankyou for your cooperation. APPENDIX B Comparison of Sample with Subsample of Data Averaged by Organization Type Frequency of Interaction Subsample Data Averaged by Organization Type in a District Single Organization Data A/ = 136 X 4.57 S.D. 1.28 Community Organization Benefits X 4.38 S.D. 1.48 Employment Service Benefits X 4.50 S.D. 1.51
4.31
1.44
4.36
1.91
4.00
1.91
F=.327,p=.57 F=.002,p=.97
F=.716,p=.4O
234/ASQ
doc_959400427.docx