An Interorganizational Analysis of Power, Conflict, and ...jlpweb/papers/An Interorganizational...

An Interorganizational Analysis of Power, Conflict, and Settlements in PublicSector Collective Bargaining

James L. Perry; Charles H. Levine

The American Political Science Review, Vol. 70, No. 4. (Dec., 1976), pp. 1185-1201.

Stable URL:

http://links.jstor.org/sici?sici=0003-0554%28197612%2970%3A4%3C1185%3AAIAOPC%3E2.0.CO%3B2-M

The American Political Science Review is currently published by American Political Science Association.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available athttp://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtainedprior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content inthe JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained athttp://www.jstor.org/journals/apsa.html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printedpage of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academicjournals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers,and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community takeadvantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.orgWed Oct 3 11:07:20 2007

http://links.jstor.org/sici?sici=0003-0554%28197612%2970%3A4%3C1185%3AAIAOPC%3E2.0.CO%3B2-M

http://www.jstor.org/about/terms.html

http://www.jstor.org/journals/apsa.html

An Interorganizational Analysis of Power, Conflict, and Settlements in Public Sector Collective Bargaining JAMES L.PERRY

University of California, Irvine

The dramatic growth of unionism and collective bargaining in the public sector signi-fies a drastic departure from previous concep- tions of the public employment relationship and previous patterns of public employee ac-tivity.l Concomitant with the rise of public sector unionism has been the increased interest social scientists have exhibited in the determinants of power, conflict, and outcomes in public sector collective bargaining. Despite increased scholarly interest, however, attempts at theory construction have fallen short because there has been a tendency to use conventional language consistent with the common discourse that occurs in the actual coIIective bargaining process; many scholars in the field have viewed idiographic orientations as the only means of validly approaching the subject; and the field has been dominated by scholars seeking rnelio- ristic solutions to employee relations problems in their research rather than focusing on the-ory b~ i ld ing .~

1 One observer notes that in 1955 "the subject of labor relations in public employment could not have meant less to more people, both in and out of government.,, See Ida Klaus, Emerging Relationship,,, Collective Bargaining in the Public Service, ed. D. H . Kruger and Charles F. Schmidt, Jr. (New York: Ran- dom House, 1969), p. 11. Within 15 years, 2.3 million government employees held union membership and an additional 1.7 million government employees were members of associations, often functioning indistin-guishably from unions. Municipal employees are rep-resented for bargaining purposes by unions or associations in more than 1,300 communities. For statistics on the extent of public employee unionization and strike activities, see Tax Foundation, New York, Union and Government Employment (New York: Tax Founda-tion, May 1972), Tables 1 and 2, pp. 8-9, and Table 10, p. 29, and Richard H. Nelson and James L. Doster, "City Employee Representation and Bargaining Pol-icies," Monthly Labor Review, 95 (November, 1972), 43-50. For a comparison of union and association behavior in the field of public education, consult J. Douglas Muir, "The Strike as a Professional Sanction: The Changing Attitude of the National Education Association," Labor Law Journal, 19 (October, 1968), 61 5-627.

2 Scholars have adopted a number of approaches in their efforts to describe and explain conflict and outcomes in public sector collective bargaining. Liston, for example, attempts to explain the causes of public employee strikes merely by asserting that government workers in the lower classifications want higher salaries. See Robert A. Liston, The Limits o f Defiance: Strikes, Rights and Government (New York: Franklin

and CHARLESH. LEVINE Syracuse Unzversity

In order to better understand public sector collective bargaining and to build theory about the phenomena, we need to develop a theo-retical framework that allows for the systematic testing of propositions. Interorganizational theory, one approach that appears fruitful for constructing such a theory, is used in this study to capture major properties of the structure of public sector collective bargaining and to pre- sent and test some derived hypotheses. These hypotheses are tested with data gathered from a sample of 60 case studies of public sector collective bargaining in New York City from 1968 to 1972.

Theoretical Considerations

* public organization may be viewed as a coalition of individuals, "some of them organ- ized into subcoalitions," whose "members in-clude administrators, workers, appointive officials, elective officials, legislators, judges, clien- tele, interest group leaders, e t ~ . " ~Drawing boundaries around the organizational coalition is difficult because organization members play a variety of roles. Nevertheless, the concept of organizational coalition can be simplified by

Watts, 1971). Wellington and Winter, in The Unwns and the Cities (Washington, D.C.: The Brookings In- stitution, 1971), Chapter 1, employ a conceptual framework which assumes "a municipality with an elected city council and an elected mayor who bargains with unions representing the employees of the city." The conceptual framework is for the most part undefined and unaccompanied by indicators or constructs for its operationalization. Drotning, Lipsky, and Foster take a broad positivist approach to public sector collective bargaining. The outcome of interest to Drotning et al. is the mode of conflict resolution -mediation, fact-finding or conciliation. Although the authors fail to operationalize their approach, a more fundamental problem is the absence of any formal explication of the theoretical relationships underlying the variables. See John Drotning, David Lipsky, and Howard Foster, "The Analysis of Impasse Procedures in Public Sector Negotiations," Management o f Personnel Quarterly, 9 (Summer, 1971), 21-30. For one of the more prom- ising recent attempts to explain conflict see Thomas A. Kochan, "A Theory of Multilateral Collective Bar-gaining in City Governments," Industrial and Labor Relations Revzew, 27 (July, 1974), 525-542.

3 Richard M. Cyert and James G. March, A Be-havioral Theory o f the Firm (Englewood Cliffs, N.J.: Prentice-Hall, 1963), p. 27.

1185

The American Political Science Review Vol. 70

identifying coalition members for certain classes of decisions and for specified periods of time.' An organizational coalition may be distinguished from its various subcoalitions by the degree of autonomy it exercises in decision making and goal setting. A subcoalition's decisions are influenced by legitimated authority within a hierarchical structure of which the subcoalition is an integral part. In interorganizational relationships, on the other hand, "the individual decisions of two or more units are made on a more inclusive systemic level which includes these units.'j5

The establishment and perpetuation of an organizational coalition rests on an equilibri-um between member contributions and organizational inducements.6 For individual workers the inducements/contributions contract has both short-term and long-term implications. Thompson writes:

Clearly there is a quid pro quo theme to the inducements/contributions contract, for it defines what is expgcted of individuals in terms of jobs needing to be done, and it defines the rewards which the organization pledges for the appro-priate performance of such jobs. . . . As a unit in his career, the job provides the individual with an arena or sphere of action in which to seek solutions to his career problems, and thus to meet the demands placed on him by the social system.'

March and Simon note that each partici- pant will continue his participation in an or-ganization only so long as the inducements of- fered him are as great or greater than the contributions he is asked to make. They as-sert: "Where the contract is viewed as un-changeable, the only options are 'accept' or ' reje~t ' ."~In cases where the contract is viewed as changeable, however, the motivation of individuals to protect or enhance their "sphere of action" is reflected in the negotiation of inducements/contributions contracts."

Since the action spheres presented by jobs differ, they require different strategies for the

4Cyert and March, A .Behavioral Theory o f the Firm, p. 27.

5 Roland L. Warren, "The Concerting of Decisions as a Variable in Organizational Interaction," Inter-organizational Decision Making, ed. Matthew Tuite, Roger Chisolm, and Michael Radnor (Chicago: Al-dine, 1972), p. 21.

6 Equilibrium theory is presented and discussed in Chester I. Barnard, The Functions o f the Executive (Cambridge, Mass.: Harvard University Press, 1938), and James G. March and Herbert A. Simon, Organiza-tions (New York: John Wiley and Sons, 1958).

7 James D. Thompson, Organizations in Action (New York: McGraw-Hill, 1967), p. 106.

8 March and Simon, Organizations, p. 83. 9 Thompson, Organizations in Action, p. 106.

negotiation of the inducements/contributions contract.1° Collective bargaining may be viewed as a special (though not uncommon) strategy for the negotiation of the inducements/ contributions contract. Thompson emphasizes that "the content of the inducements/contri-butions contract is determined through power (political) p roce~ses . "~~ Conceived as the dependence of the parties relative to one an-other,12 the power of the individual in "rou- tinized technologies" and "early-ceiling occu-pations" is severely limited in negotiation of the inducements/contributions contract.13 Nonetheless, when there exist large numbers of individuals in similar occupational roles, opportunities develop for collective action as a strategy of negotiation. Under these conditions, the problems of defining and enforcing the employment contract become a matter of concern for all organizational participants:14 "The higher wages, shorter hours, and better working conditions that unions demand are collective goods to the workers."15

Collective bargaining involves a transfor-mation of the basic structure of the public organizational coalition. Rather than the indi-vidual negotiating for the protection and enhancement of his job, negotiation is pursued through the formation of a new and distinct organization.le For purposes of analyzing collective bargaining the employee union can be considered an autonomous organization, free from the decision-making constraints of an overarching organization, and sharing in the formation of policies that determine the public employment relationship. The operation of neither organization, however, can be understood merely as an aggregation of the goals of its members. Coleman writes: "The actions of a trade union and other mass-based collec- tives are not always toward goals collectively held, but the very existence of the collectivity is based upon action toward collective goals."17

To account for collective goals, Thompson suggests conceptualizing goals as the "future domains intended by those in the dominant

10 Ibid., p. 109. 11Ibid.. a. 106. 1 2 lbid.; i. 107. 13 Ibid., p. 110. 1'See March and Simon, Organizations, p. 91. 15 Mancur Olson, Jr., The Logic o f Collective Action

(New York: Schocken Books, 1968), p. 79. Although March and Simon indicate the trade

union occupies an ambiguous position, they do classify it as an extraorganizational group. March and Simon, Organizations, p. 72.

1 7 James S. Coleman, "Foundations for a Theory of Collective Decisions," American Journal o f Sociology, 71 (May, 1966), 615-627, at 616.

1976 1187An Analysis of Public Sector Collective Bargaining

coalition."18 An organization's domain may be viewed as the claims which an organization stakes out for itself in terms of (1) range of products, (2) population served, and ( 3 ) ser-vices rendered.19 The dominant coalition is conceived by Thompson as a subgroup of "interdependent individuals who collectively have sufficient control of the organization's resources to commit them in certain directions and withhold them from others."Z0 In collective bargaining, two dominant coalitions each act to preserve or expand their domains through negotiation over the content of the inducements/contributions contract.

From this interorganizational relationship comes conflict and contractual change, and these are related to the nature of goal divergencies, dependencies, and the degree of sta-bility in the relationship between the two or-ganizations.

Following Cyert and March, the amount of goal divergence in the collective bargaining relationship can be understood in terms of the dimensions of each organization's goals (the future domains they view as important) and the strength of their preference for a particular outcome on a goal dimension.Z1 The strength of an organization's preferences represents its "aspiration level" on any particular goal di-mension. The organization's past goals, its past performance, and the past performance of other "comparable" organizations contribute to the organization's goal dimensions and aspiration levels.

In interorganizational relationships, power resides implicitly in the dependencies of or-ganizations on one another. Pugh et al. focus on organizational dependence in the following manner: "The dependence of an organization reflects its relationships with other organiza-tions in its social environment, such as sup-pliers, customers, competitors, labor unions, management organizations, and political and socia1 organization^."^^ Within the environ-ment of the union and the public organization exist a set of power-dependence relations-what Dill terms "task environment"-those

18 Thompson, Organizations in Action, p. 128. 19 Sol Levine and Paul E. White, "Exchange as a

Conceptual Framework for the Study of Interorgani-zational Relationships," Administrative Science Quar-terly, 5 (March, 1961), 583-601; and Thompson, Organizations in Action, p. 26.

20 Thompson, Organizations in Action, p. 128. 2 1 Cyert and March, A Behavioral Theory o f the

Firm, p. 115. 22 D. S. Pugh, D. J. Hickson, C. R. Hinings, and

C. Turner, "The Context of Organization Structures," Administrative Science Quarterly, 14 (March, 1969), 91-1 14.

actors in the environment which are relevant to goal attainment.23

Net power then results from a set of relationships between the organization and the elements of its pluralistic task environment. This admits the possibility that in collective bargaining the public organization may be powerful in relation to the union but dependent on other actors in its task environment to enforce its claims to domain. Thus, it be- comes important for the organization to judge the amount and sources of support that can be mobilized for its goals, and to arrive at a strategy for their mobilization.

Although the public organization and the union organization negotiate a contract per-iodically, they must deal with each other con- tinuously in handling disputes arising during the administration and interpretation of the contract. Walton and McKersie contend that the pattern of interaction between the organizations during contract administration directly influences the attitudes between the units that later affect contract negotiation^.^^ Confined to the dyadic relationship between the union and public organization, stability refers to the degree to which prior structural features of the relationship expand or contract the capacity for joint decision making during collective bargaining.

In light of this distinction, the degree of stability in the relationship between the union and the public organization may be viewed as a product of the interactions between the two organizations prior to the initiation of contract negotiations. The interactions between the organizations during contract administration af-fect the development of the capacity for collaboration during contract negotiation by influencing the pattern and quality of information exchanged between the partiesz5 Also, the interactions prior to collective bargaining, directly influence attitudes between the parties which in turn determine how the organizations

23 William R Dill, "Environment as an Influence on Managerial Autonomy," Administrative Science Quarterly, 2 (March, 1958), 409-443, at 410. "Or-ganization set" is a term analogous to "task environment" employed by William M. Evan in "The Organi- zation-Set: Toward a Theory of Interorganizational Relations," Approaches to Organizational Design, ed. James D. Thompson (Pittsburgh, Pa.: The University of Pittsburgh Press, 1966), pp. 173-191.

24Richard E. Walton and Robert B. McKersie, A Behavioral Theory o f Labor Negotiations (New York: McGraw-Hill, 1965), chapter 6.

2 5 Richard E. Walton, "Theory of Conflict in Lateral Organizational Relations," ed. J. R. Lawrence, Oper-ational Research in the Social Sciences (London: Tavistock Publications, 1966), pp. 409-426,


view the continuity of the contract administration and contract negotiation phases of the relationship.26

These situational phenomena-goal diver-gence, stability, and organizational dependence -are conceptually distinct from the behaviors of the individual members of the union and public organization and affect the degree to which interorganizational behavior is oppositional, that is, the extent to which conflict takes passive' or aggressive forms. This conceptualization of conflictual behavior is intended to encompass the whole range of oppositional behavior from manifest forms of "strategic maneuvering" to covert forms of passive re-~ i s t a n c e . ~ ~ the de-Within interorganizational cision-making process, oppositional behavior may be triggered by either acts of the other organization or changes in situational contexts.

The collective agreement that terminates negotiations between the public and union organizations comprises the inducements/ contributions contract agreed to during the collective bargaining process. The induce-ments/ contributions contract represents a set of decisions of two individual organizations at a more inclusive systemic level because the organizations jointly establish their domains with regard to the public employment rela-tionship until the negotiation of a new agree-ment. Contractual change refers only to the changes in the content of the inducements/ contributions contract negotiated from the conclusion of one agreement to the conclusion of the subsequent agreement through the collective bargaining process. Inducements in-clude such items as salary levels, monetary fringe benefits, union security provisions, and the control of the physical and health-related environment of the workplace. Although often implicit in the employment relationship itself and only suggested in the management rights clause, contributions items frequently specified in the contract include the length of the work- week, the residual rights of management, the grounds for disciplinary actions, and descrip- tions of job duties.

Table 1. Theoretical Definitions

Goal Divergence: The degree of incompatibility between the goals of the union and public organization.

Stability: The degree to which prior structural features

26 Walton, "Theory of Confict in Lateral Organiza- tional Relations"; see also Walton and McKersie, A Behavioral Theory o f Labor Negotiations.

27 Louis R. Pondy, "Organizational Conflict: Con- cepts and Models," Administrative Science Quarterly, 12 (September, 1967), 296320, at 312.

of the relationship expand or contract the capacity for joint decision making during collective bargaining.

Organizational Dependence: The degree of support from the task environ- ments of the negotiating organizations.

Conflictual Behavior: The degree of behavioral opposition between the union and public organizations during collective bargaining.

Contractual Change: The degree of change in the content of the inducements/contributions contract.

Propositions and Operational Linkages With the groundwork laid through the in-

troduction of the interorganizational analytic framework, some of the concepts may be in- terrelated via formal theoretical statements. Preceding each proposition is a statement of the theoretical linkages and premises, and following each proposition is an exposition of supportive empirical evidence and the hy-pothesized operational linkages.

I. The choice of tactics for pursuing in-compatible goals is likely to be influenced by the nature and radicalism of each organization's goals.28 While goal divergence identifies the state of incompatibility or opposition between forces, the degree of conflictual behavior characterizes the behavioral outcome of the interactions between the organizations. The relationship between these concepts is not merely unidirectional since the escalation of conflict can be expected to influence the nature of the goal divergencies. As conflictual behavior es-calates, there is a greater tendency for participants to associate only with those similarly oriented to the issues. This form of social support strengthens individual organizational members, highlights the incompatibility of goals, and solidifies the divergence of positions.

These relationships between goal divergence and conflictual behavior are particularly relevant for understanding strike behavior.29 The union views the strike costs as costs to be re- couped and revises its aspirations upward in an attempt to offset the costs of the strike. In response to the union action, management's resolve to avoid concessions on certain goal dimensions is likely to increase, and management may even be inclined to withdraw earlier concessions.

28 See Lewis A. Coser, The Functions o f Social Conflict (Glencoe, Ill.: The Free Press, 1956), and Louis Kriesberg, The Sociology o f Social Conflicts (Englewood Cliffs, N.J.: Prentice-Hall, 1973).

29 See Walton and McKersie, A Behavioral Theory o f Labor Negotiations, pp. 56-57 for an extended discussion of this point.

An Analysis of Public Sector Collective Bargaining

Proposition I. In public sector collective bargaining, the greater the goal divergence, the greater the conflictual behavior; as conflictual behavior increases, goal divergence also increases.

Empirical and experimental evidence from Axelrod, Bell, and Deutsch and Kraus is supportive of the plausibility of the two-direc-tional theoretical relationship between goal divergence and conflictual behavior.S0 The operational linkage for Proposition I is hy-pothesized to be a positive, S-curve relationship. When the degree of goal divergence is minimal, some conflictual behavior may exist, but not a sufficient amount to produce feedback consequences. At intermediate levels of goal divergence, a nearly linear relationship between goal divergence and conflictual behavior is anticipated. Goal divergence may further increase while the degree of conflictual behavior has approached or reached an upper limit, thus accounting for a leveling off in the operational linkage.

11. Schelling writes that bargaining neces-sitates some kind of collaboration or mutual accommodation even if only in the avoidance of mutual disaster.31 The time and effort re-quired by the public and union organizations to form mutually consistent expectations, the nature of the expectations, and the capacity to communicate or to destroy communication in collective bargaining depends on the stability of the relationship developed during interaction in administering the collective agreement.

In analyzing interunit conflict in organizations, Walton proposes that more flexible and informal decision rules and less frequent re-course to the use of outside arbiters is likely to increase the capacity for collaboration in joint decision making between the units.32 The development of mutual attitudes toward one another is a product of the degree of positive or negative affect between the organizations throughout their relationship. It is during interaction in contract administration that the or-ganizations develop mutual expectations, atti-

30 Robert Axelrod's investigation of a series of sit-uations in politics confirmed his hypothesis that the greater the conflict of interest in a strategic interaction, the greater the conflictual behavior. See Robert Axelrod. Conflict o f Interest (Chicago: Markham Pub- lishing, 1970); see also Inge Powell Bell, CORE and the Strategy of Non-Violence (New York: Random House, 1968), p. 59; and Morton Deutsch and Robert M. Kraus, "The Effects of Threat Upon Interpersonal Bargaining," Journal of Abnormal and Social Psy-chology, 61 (September, 1960), 210-225.

31 Thomas C. Schelling, The Strategy o f Conflict (New York: Oxford University Press, 1960).

32 Walton, "Theory of Conflict in Lateral Organi- zational Relations."

tudes, and decision rules relevant to their later behavior during collective bargaining. It is this structure of interorganizational interactions which expedites agreement or, conversely, hinders agreement and the control of conflictual behavior during collective bargaining.

Proposition 11. In public sector collective bargaining, the lesser the stability during contract administration, the greater the conflictual behavior during negotiations.

Newcomb; Loomis; Bixenstine, Levitt, and Wilson; and Hammond et al. provide experimental results that support this p r o p ~ s i t i o n . ~ ~ Lincoln's case analysis of the 1966 New York City Transit Authority negotiations identifies instability in the system as an explanatory variable, for the strike which subsequently occurred.34 Pruitt, in a theoretical analysis of stability between pairs of national decision-making units,35 suggests that the operational linkage between stability and conflictual behavior may be a step function; i.e., when the level of stability reaches a critical state, it causes the likely degree of conflictual behavior to change radically. The operational linkage at the interorganizational level of analysis, however, is more likely to be a power function.

111. Within the context of public sector collective bargaining, the choice of negotiation strategies and the employment and timing of particular commitment tactics for the attainment of new organizational domains will depend upon the support each organization ob- tains within its task environment. The amount of asymmetry between the dependence positions of the organizations influences the suc-cess of strategic choices and in turn the degree of behavioral opposition.

Small imbalances in relative dependence continually reinforce the need for cooperation by highlighting the ir~terdependencies of the

33 T. M. Newcomb, "An Approach to the Study of Communicative Acts," Psychological Review, 60 No-vember, 1953), 393-404; J. L. Loomis, "Cornmunica- tion, the Development of Trust and Cooperative Be-havior," Human Relations, 12 (November, 1959), 305- 315; V. E. Bixenstine, C. A. Levitt, and K. V. Wilson, "Collaboration Among Six Persons in a Prisoner's Dilemma Game," Journal of Conflict Resolution, 10 (December, 1966), 488-496; K. R. Hammond, F. J. Todd, M. Wilkins, and T. 0 . Mitchell, "Cognitive Conflict Between Persons: Application of the Lens-Model Paradigm," Journal of Experimental Social Psychology, 2 (October, 1966), 343-360.

34 Albert A. Lincoln, "The New York City Transit Strike: An Explanatory Approach," Public Policy, 16 (1967), 271-292.

35 Dean G. Pruitt, "Stability and Sudden Change in Interpersonal and and International Affairs," Journal of Conflict Resolution, 13 (March, 1969), 18-38.

The American Political Science Review

organizations. As the asymmetry of relative dependence increases, the likelihood increases that persuasion as a means for reaching agreement will fail, stimulating greater behavioral opposition in the modes of transaction be-tween the organizations. Large differences in relative dependence invite the stronger of the organizations to use coercion against the weaker organization to expand its domain. On the other hand, refusal by the stronger organization to make changes requested by the weaker organization stimulates greater con-flictual activity by the weaker organization.

Proposition 111. In public sector collective bargaining, the greater the asymmetry of relative dependence between the negotiating organizations, the greater the conflictual behavior.

Narrative and empirical support of the proposition is provided in studies by Terhune, Kerr and Siegel, and Zinnes.3" negatively skewed, positive curvilinear operational link-age is expected because beyond a certain level of asymmetry of relative dependence the likelihood of conflictual behavior is considerably reduced. Where an organization is so limited by its dependence position that it very nearly ceases to operate as an independent unit, the likelihood of conflictual behavior rapidly de- clines.

IV. The more substantial or encompassing a conflict, the greater the pressures from third parties for its s~lut ion.~ ' In the public sector, as conflictual behavior persists or increases, efforts to arrive at agreement and achieve finality in the collective bargaining process generally lead to mediation or factfinding mechanisms. Concurrent with this phenomenon is the fact that as conflictual behavior between the union and public organization increases, agreement is less likely to be achieved by voluntary concession and reduction in demands. The anticipa- tion of third party intervention may stimulate marginal concessions by the organizations and actual third party intervention may legitimate or require further concessions.38 Thus, as con-

36 Kenneth Terhune, "Motives, Situation, and Inter- personal Conflict Within Prisoner's Dilemma," Journal o f Personality and Social Psychology, 8 (March, 1968), 1-24; Clark Kerr and Abraham Siegel, "The Inter- industry Propensity to Strike -An International Com- parisop," Industrial Conflict, ed. Arthur Kornhauser, Robert Dubin, and Arthur M. Ross (New York: Mc-Graw-Hill, 1954), pp. 189-212; and Dina A. Zinnes, "An Analytical Study of the Balance of Power The- ories," Journal o f Peace Research, 4 (1967), 270-288.

37 Bernard Berelson and Gary A. Steiner, Human Behavior: An Inventory o f Scientific Findings (NewYork: Harcourt, Brace, and World, 1964), pp. 619-620.

38 Schelling, The Strategy o f Conflict, and Sidney

flictual behavior increases, the process is likely to produce greater change in the inducements/ contributions contract. Proposition IV. In public sector collective bargaining, the greater the conflictual behavior, the greater the contractual change.

Little empirical research has been under-taken to assess the plausibility of this proposition. However, one study of the effect of impasse during collective bargaining on ne-gotiated salaries of teachers found better mean salary gains in some impasse districts.39 The proposed relationship is expected to be clearly evident if all dimensions of the collective agreement are analyzed. The operational linkage between conflictual behavior and contractual change should be positive and linear. The slope of the function 1s expected to be substantially less than unity.

V. When it comes to a test of strength, the organization that can mobilize the most re-sources and trade on relationships within its task environment should achieve a more advan- tageous outcome. Even the expectation of los- ing at some future time may encourage the weaker organization to search for an immediate solution or offer a favorable concession to the stronger organization. The degree of im-balance in their relative strength is likely to affect the amount of the concessions that the weaker organization will have to offer in or-der to extract lesser concessions from the stronger organization or to reach final agreement. Since demands are almost always initiated by the union, changes in the contract will usually take the form of concessions to the union. Proposition V. In public sector collective bargaining, the greater the asymmetry of relative dependence favoring the public organization, the lesser the contractual change; the greater the asymmetry of relative dependence favoring the union, the greater the contractual change.

This proposition is consistent with Gam-son's theoretical analysis40 and is supported in Rapoport's and Orwant's review of experimental games, particularly those of ~ lood-and Cap10w.~~Lammers likewise found a strong

Siegel and Lawrence E. Fouraker, Bargaining and Group Decision Making (New York: McGraw-Hill, 1960).

39 Robert Arnold Mickler, "A Study of the Effect of Impasse During Collective Bargaining on NegotiatedSalaries of Teachers" (Ph.D. dissertation, Maxwell School, Syracuse University, 1972).

40 William A. Gamson, Power and Discontent (Homewood, Ill.: The Dorsey Press, 1968).

4 1 Anatol Rapoport and Carol Orwant, "Experi- mental Games: A Review," Behavioral Science, 7 (January, 1962), 1-37.

1976 An Analysis of Public Sector Collective Bargaining

Goal Divergence L - 1

Contractual Change

Theoretical Linkages

----- Interperiod Feedback

Figure 1:An Overview of the Model

relationship between relative strength of the parties and ou t~ome .~ ' The trace line for the operational linkage between relative depend-ence and contractual change should resemble

42 Cornelius J. Lammers, "Strikes and Mutinies: A Comparative Study of Organizational Conflicts Be-tween Rulers and Ruled," Administrative Science Quarterly, 14 (December, 1969), 558-572. A laboratory simulation of interparty exchange during labor negotiations tested the effects of varying the level of social cues during negotiations on the outcome of bargaining. Increases in the level of social cues favorable to the union positively influenced the level of contractual change; social cues favorable to management led to a reverse outcome. See Ian R. Morley and Geoffrey M. Stephenson, "Interpersonal and Inter- party Exchange: A Laboratory Simulation of an In-dustrial Negotiation at the Plant Level," British Journal of Psychology, 60 (November, 1969), 543-545.

a negatively sloping S-curve; the relationship is also expected to exhibit both upper and lower limits.

A diagrammatic summary of the theoretical relationships is presented in Figure 1; Ta-ble 2 presents a summary of the operational linkages. The broken lines in Figure 1 depict feedback between the outcome of the collective bargaining process in one period and its effect upon the variables in the succeeding period." Because conflictual behavior is caus-

43 For an exploration of feedback processes in labor relations systems see A. Dale Allen, Jr., "A Systems View of Labor Negotiations," Personnel Journal, 50 (February, 1971), 103-114; and John T.Dunlop, In-dustrial Relations Systems (New York: Henry Holt, 1958).

Table 2. Summary of Operational Linkages

Relationship Direction Limit

Goal Divergence and Conflictual Behavior positive

Stability and Conflictual Behavior negative power none

Asymmetry of Relative Dependence and Conflictual positive negative skew, none Behavior curvilinear

Conflictual Behavior and Contractual Change positive linear none

Asymmetry of Relative Dependence favoring one of the negative upper and organizations and Contractual Change lower limits

1192 The American Political Science Review Vol. 70

ally dependent on goal divergence, stability, and asymmetry of relative dependence but causally prior to contractual change, it is de- picted as an intervening variable in Figure 1.

Methodology

The practice of collective bargaining in the public sector has not triggered a bounty of empirical studies which have accumulated large bodies of comparative data. The dearth of research in this field partly stems from the realities of public sector collective bargaining and from the nature of measurement problems associated with the most relevant variables. These problems indicate that relatively inex-pensive data collection procedures like the analysis of available aggregate data sets would be inadequate to test the propositions pre-sented in this paper. To provide an appropriate setting to test the model, we had to se-lect a field site with an adequate population of cases for the selection of a sample of cases, adequate archival records for the measurement of the variables, and a uniform legal and political framework within which collective bargaining takes place. For these reasons, the tests of the hypotheses focused on a cross-sectional analysis of public sector collective bargaining in one large municipality, New York City.

Collective Bargaining in New York City. In 1958, with the issuance of Executive Order No. 49, Mayor Robert Wagner initiated a program to implement collective bargaining practices prevailing in the private sect0r.4~ Horton writes: "The executive order quickly was dubbed the 'Little Wagner Act' and the city's civil service leaders called it their 'Magna Carta.' In retrospect, however, the City's new labor relations program hardly deserved these appellation^."^^

Collective bargaining procedures were al-tered in 1967 by Executive Order No. 52 which established independent, impartial tri- bunals to settle disputes arising in contract negotiations, contract administration, and representation elections.4~imultaneously,the New York State Legislature passed the Public Em-

44 New York City, Office of the Mayor, Executive Order No. 49, March 31, 1958.

45 Raymond D. Horton, Municipal Labor Relations in New York City (New York: Praeger, 1973), p. 30.

4 6 New York City, Office of the Mayor, Executive Order No. 52, September 29, 1967. See Horton, Mu-nicipal Labor Relations in New York City for an extended discussion of the evolution of New York City's collective bargaining law.

47 New York State Sessions Laws (1967), Chapter 392.

ployees Fair Employment Act (Taylor Law) which exempted New York City from the jurisdiction of the State Public Employment Rela- tions Board (PERB) on the condition that its system for collective bargaining be "substantially e q ~ i v a l e n t . " ~ ~ wereThus, City policies modified in accordance with the provisions of the Taylor Law to parallel State guidelines more

Sayre and Kaufman's contention that "no single ruling elite dominates the political and governmental system of New York City"49 is characteristic of its labor relations activities. A multiplicity of public officials and agencies are involved in New York City government labor relations: the mayor, agency personnel, the Office of Labor Relations, the Office of Collective Bargaining, the Bureau of the Budget, the Department of Personnel, the City Council, and the Board of E ~ t i m a t e . ~ ~ The gov- ernor and the State legislators are also occa- sional participants in City labor relations policy making.51

While the Office of Labor Relations provides the professional expertise in contract negotiation^,^^ it is the Office of Collective

Bargaining (OCB) that can be considered the governmental "core The OCB is a tripartite board consisting of two members appointed by the mayor, two members ap-pointed by a committee of municipal union leaders, and three members selected by unani- mous vote of the mayoral and union appoint- ees. The primary function of the OCB is the resolution of bargaining impasses. The Board of Certification, composed of three of the seven OCB members, is responsible for cer-tifying employee bargaining representatives. In addition, arbitrability of grievances is de-

48These procedures are spelled out in the New York City Collective Bargaining Law, New York City, Administrative Code, chapter 54. For the establish-ment of the Office of Collective Bargaining, see: New York City Charter, chapter 54, section 1173.

49 Wallace Sayre and Herbert Kaufman, Governing New York City: Politics in the Metropolis, 2nd ed. (New York: W. W. Norton, 1965), p. 710.

50 Horton, Municipal Labor Relations in New York City, p. 12.

51 Zbid. 52 The Office of Labor Relations was created by

Mayor Lindsay in 1966. Its primary function is to represent the City government and units within its jurisdiction in collective bargaining. For an inside view of its operations see Anthony C. Russo, "Man- agement View of the New York City Experience," Unionization of Municipal Employees, ed. Robert H . Connery and William V. Farr, (New York: The Academy of Political Sciepce, Columbia University, 1971, pp. 80-93.

53 Sayre and Kaufman, Governing New York City, p. 710.

1976 An Analysis of Public Sector Collective Bargaining 1193

termined by all OCB members in their role as the Board of Collective Bargaining.j4

When the Office of Collective Bargaining started operations in January, 1968, there were approximately 400 bargaining units in exist-e n ~ e . 5 ~Through consolidation of units, the number of bargaining units was reduced to 325 by 1971.56

The choice of New York City as the locus of research creates some dilemmas because New York's city government often is considered sui generis by researchers. It is larger and possibly more complex than any other United States city government.j7 The weight of several other factors, however, outweighed these considerations in the selection of New York City as the research site. First, a limited geographic area provided the opportunity to gather in-formation that, dispersed over a larger region, would have been too costly and time consum- ing to collect. Second, testing the model on a sample of New York City cases did provide an opportunity to disconfirm the propositions, a consideration that took precedence over concern with external validity. Finally, where a comprehensive legal framework for public employment collective bargaining has been adopted by state and local jurisdictions, there tends to be marked similarities in statutory provi~ions.5~Therefore, no compelling reason exists for contending a priori that an explanatory model tested on cases of public sector collective bargaining in New York City would be less valid for other settings with a com-prehensive statutory framework.

Sample Selection. Specification of the population was achieved through the collection of comprehensive lists of collective bargaining

5 4 Horton, Municipal Labor Relations in New York City, pp. 77-79.

55 New York Citv. Office of Collective Bargaining, Annual Reports, 1970-1971, pp. 6-7.

56 Ibid. 57 Sayre and Kaufman, Governing New York City,

p. 709, characterize New York City as "imperial" among American cities.

58 On this point, compare the 1971 and 1973 ver- sions of the following publications: U.S. Department of Labor, Labor Management Services Administration, Division of Public Employee Labor Relations, Sum-mary o f State Policy Regulations for Public Sector Labor Relations: Statutes, Attorney General's Opin-ions, and Selected Court Decisions (Washington, D.C.: U.S. Government Printing Office, November, 1971 and February, 1973). For a general discussion of the diffusion of similar institutional innovations, see Jack L. Walker, "The Diffusion of Innovation Among the American States," American Political Science Re-view, 63 (September, 1969), 880-889. For an analysis of public employment bargaining laws see Thomas A. Kochan, "Correlates of State Public Employee Bar-

settlements in New York City. A set of master lists,j"repared by the New York City Office of Labor Relations (OLR), contained identifying and descriptive information on collective bargaining settlements between New York City governmental units and various unions representing public employees from 1968 through 1972. Three sources, containing lists of public employee work stoppages during this period, were also consulted to assure that these cases were included on the master lists.GO AS a result of this process, the population encompassed 535 collective bargaining settlements.

The next step toward selecting the 60-case research sample was to identify the level of conflict during negotiations so that the population could be stratified. T o achieve this objec- tive, several sets of documents were consulted. First, the lists of public employee work stoppages were again utilized.61 Second, all cases in which impasse, fact-finding, or mediation occurred were identified.62 Third, The New York Times Index was utilized to identify negotiations, strikes, settlements, and other activities that may have been reported in The New York Times. Fourth, several civil service and local union publications proved to be valuable sources. The Chief and the Civil Service Leader, publications covering topics of general interest to civil service employees, provided information on a wide range of cases within the population; and the local publication of the AFSCME, Public Employee Press, yielded valuable information pertaining to many additional cases in the population. Stratification was based on a four category scale. Different

gaining Laws," lndustrial Relations, 12 (October, 1973), 322-337.

59 New York City, Office of Labor Relations, Col-lective Bargaining Settlements by the Ofice of Labor Relations, 1968-1972 (mimeographed).

60 New York City, Office of Collective Bargaining, Annual Report - 1972, pp. 23-24; New York State Public Employment Relations Board, Detailed Listing o f Public Employee Work Stoppages, New York State, 1968-1972. (Mimeographed); Economic Development Council of New York City, New York City's Public Sector Collective Bargaining, Looking Ahead in New Yonk City, No. 4 (New York: Economic Development Council of New York City, 1973), pp. 101-103.

6lNew York City, Office of Collective Bargaining, Annual Report-1972; New York State Public Em- ployment Relations Board, Detailed Listing o f Public Employee Work Stoppages; and Economic Develop-ment Council of New York City, New York City's Public Sector Collective Bargaining.

62 New York City, Office of Collective Bargaining, Annual Report - 1968-1972. The bulk of information was contained in these reports. Additional information on impasses was retrieved from other unpublished documents within the Office of Collective Bargaining.


conflict acts served to define minimum inclusion levels for each group. Group 1 included all cases where no conflictual acts were ob-served. Group 2 included cases in which impasses occurred but in which no overt conflictual acts were observed. Group 3 included all cases in which overt acts short of a strike were observed. Group 4 included all cases in which a strike did occur. Once stratification of the population had been completed, the final step in specifying the sample was to select 15 cases randomly from each of the four groupsG3

Operational Measures

The indicators for each of the concepts defined in Table 1 are presented below. Be-cause most of the indicators were developed specifically for this study, little attention can be given to the empirical validity of the indicators as they have been used in previous studies.

Goal Divergence. The number of demands sub- mitted by the union at the initiation of negotiations was selected as the indicator of goal divergence. This indicator was chosen over others primarily for advantages of data avail- ability, ease of accessibility, and completeness. Using an unobtrusive indicator to measure goal divergence also circumscribed many of the difficulties which could have been encountered in attempting to measure the variable with reac-tive instruments.

Ideally, the number of initial public organization demands should have been considered also. For the population utilized here, how-ever, there were approximately thirty demands general to all collective bargaining negotiations in New York City.64 Since the content of the demands was not published or available from a sufficient number of the involved organizations so that further analysis could be performed, the number of initial union demands was considered adequate for cross-sectional a n a l y ~ i s . ~ ~

evidence for assessing its empirical validity.66 From an appendix of data from Hammer-mesh,6' a median test was performed on the divergence of wage proposals. Dichotomizing the cases on the basis of whether a strike did or did not occur, and computing the prob-ability for the marginal totals in a 2 x 2 con-tingency table produced a significant relationship at alpha = .05. Although this evidence is not substantial, it does point to the empirical validity of using initial bargaining demands to measure goal divergence.

Relative Dependence. In Pugh et a1.68 the size of the organizational unit relative to its parent organization was utilized to measure depend-ence. Relative size was found to intercorrelate substantially with other dimensions of dependence: status of organization unit (.68), organizational representation on policy-making body (.SO), specializations contracted out (-.60), and vertical integration (-.40).09 Pugh et al. concluded from these results, and from low intercorrelations with the dimensions of other concepts, that this indicator does tap "aspects of the dependence of the organization, particularly its dependence on external resources and power. . . ."70

A second consideration was how to operationalize the variable to measure the difference in the amount of dependence between the organizations. The operationalization selected rested on an additive assumption. The dependence score for the union organization was subtracted from the dependence score for the public organization. Under this approach, asymmetry of relative dependence favoring one of the organizations theoretically ranges from a lower limit of -1 (most favorable to the union organization) to an upper limit of + 1 (most favorable to the public organization). This indicator was used to test Proposition V. The absolute value of this indicator was used to test Proposition 111 since the amount of

While this indicator was assessed on an a priori basis for its face validity, there was little

63 An appendix containing the cases analyzed, the affiliations of the employee groups, and the year in which negotiations were conducted is available upon request from the authors.

64 Communication from John T. Burnell, Director, Office of Labor Relations, March 23, 1974. The initial number of union demands was also supplied by the OLR.

65 The content of the demands would have been extremely useful for further analysis, but the lack of uniformity in maintaining these as records within or-ganizations and the often sensitive nature of their content militated against obtaining them.

asymmetry alone, and not which organization was favored, was hypothesized to be the significant determinant of conflictual behavior.

Stability. In order to measure this concept, data on the number of grievance arbitrations filed and the number of employees in the unit were

00 Daniel S. Hammermesh, "Who Wins in Wage Bargaining," Industrial and Labor Relations Review, 26 (July, 1973), 1146-1149, at 1147.

67 Ibid. 68 D. S. Pugh, D. J. Hickson, C. R. Hinings, and C.

Turner, "The Context of Organizational Structures." 69 Ibid., p. 106, 70 Ibid., p, 107,


collected. Most of this information was avail-able from the New York City Office of Collec- tive Bargaining and Office of Labor Relations." In cases where these data were not available from these sources, the data were obtained from union publications.

A time frame for isolation of the grievance arbitration data had to be selected so as to be compatible with the conceptualization of sta-bility. The period beginning one year prior to the initiation of contract negotiations appeared to be an appropriate time span. Collecting grievance arbitration data for one year pro-vided a sufficiently long, continuous span of time prior to negotiations to observe variance in the stability of relationship patterns.

The decision to utilize grievance arbitration data for the indicator of stability stemmed from several factors. First, grievance arbitra-tian is generally recognized as an interorganizational mechanism for concerting the action of two or more organizations. Second, regarding the conduct of labor relations in New York City, Executive Order No. 52 required that "an employee organization certified for the unit of which the grievant is a member shall have the right to bring grievances unresolved at Step 2 of the general procedure . . . to impartial arbi- t r a t i ~ n . " ~ ~Third, previous research by Derber, Chalmers, and Stagner7"uggested that the grievance process serves as a valid indicator of the stability of the labor-management relationship. Finally grievance arbitration data were easily accessible and inexpensive to collect.

Conflictual Behavior. Dubin points to the diffi- culty of developing adequati empirical indicators of conflictual behavior since "conflict behavior includes nonhomogenous acts . . ." and ". . . it is a major task to develop some measuring scale by which it is possible to equate a given number of units of personal abuse with the calling of a strike."74 Dubin notes, however, that "some acts are more criti-

71 The number of employees in the bargaining unit was contained on the master lists used in sample selection. Most of the information on grievance arbitrations filed was obtained from internal records of the OCB.

7 2 New York City, Office of the Mayor, Executive Order No. 52, p. 19.

7 3 Milton Derber, W. Ellison Chalmers, and Ross Staaner, The Local Union-Management Relationship

cal to the union-management relationship than other^.";^

Ivo and Rosalind Feierabend approach the measurement of conflictual behavior on the basis of observationally defined events repre-senting different points on a scale.;(; Similarly, to analyze conflictual behavior in public sector collective bargaining, we chose to use a four-point scale, ranging from 1 (denoting absence of conflictual behavior) through 4 (denoting extreme conflictual behavior) to stratify the cases. Once the cases were selected, they were also ranked within the groups when distinc-tions could be made in the frequency or inten- sity of conflictual acts. The final scores for the cases ranged from 1 to 32.

Contractual Change. In scaling items within the contracts, weights were given to contract provisions on the basis of their increasing cost to the public organization or their favor- ableness to the union.77 In the case of some contract provisions, the item was assigned a value of 0 or 1 depending on its presence or absence in the contract. For the remainder of the contract provisions, however, more specific ordinal scales were identifiable. By coding suc- cessive contracts on this basis, an index of con- tractural change was obtained by subtracting the values of the previous contract from those of the current contract. Despite an attempt to formulate an exhaustive set of provisions, it was unreasonable to assume that all provisions could be anticipated prior to analysis. Further- more, some judgment was likewise exercised by the coder in assigning values to idiosyn-cratic provisians and deviations in particular contracts.

Findings

The operational relationships between the variables are examined graphically in Figures

75 Ibid. 76 IVOK. Feierabend and Rosalind L. Feierabend,

"Aggressive Behaviors Within Politics, 1948-62:- A Cross-National Study," Journal o f Conflict Resolution, (September, 1966), 249-271.

7; The coding scheme employed here was con-structed from a contract analysis instrument originally developed by Thomas A. Kochan and Hoyt N. Wheel- er. See Contract Analysis o f Fire Fighter Labor Agree- ments (Washington, D.C.: U.S. Department of Labor, Labor Management Services Administration, Division of Public Employee Relations, Contract No. L-72-157, February, 1973). Some additions and modifications

( ~ r b a n a , Ill.: Institute of ~ a b o rand Industrial ~ e - were made in their original instrument to reflect pro- lations, 1960). visions across employee groups and categories peculiar

74 Robert Dubin, "A Theory of Conflict and Power to New York City contracts. An appendix containing in Union-Management Relations," Industrial and the clauses which were coded and the values assigned Labor Relations Review, 13 (July, 1960), 501-508, to the clauses is available upon request from the au-at 501. thors.


5 120 Deciles of Goal Divergence

Figure 2. Operational Linkage Between Goal Divergence and Conflictual Behavior

.01 .99 Deciles of Asymmetry of Relative Dependence

Figure 4. Operational Linkage Between Asymmetry of Relative Dependence

and Conflictual Behavior

-.99 .87 Deciles of Asymmetry of Relative Dependence

Favoring One of the Organizations Figure 6. Operational Linkage Between Asymmetry

of Relative Dependence Favoring One of the Organizations and Contractual Change

0 Quintiles of Stability .002

Figure 3. Operational Linkage Between Stability and Conflictual Behavior

1 32 Deciles of Conflictual Behavior

Figure 5. Operational Linkage Between Conflictual Behavior and Contractual

Change

L . " " l l l t . -.99 .87 Deciles of Asymmetry of Relative Dependence

Favoring One of the Organizations Figure 7. Operational Linkage Between Asymmetry

of Relative Dependence Favoring One of the Organizations and Conflictual Behavior

1976 An Analysis of Public Sector Collective Bargaining 1197

Table 3. Correlates of Conflictual Behavior and Contractual Changea

Variable 1 2 3 4 5 6

1. Goal Divergence

2. Stability

3. Asymmetry of Relative Dependence

4. Asymmetry of Relative Dependence Favoring One of the Organizations

5. Conflictual Behavior

6. Contractual Change

"Pearsonian correlation coefficients rounded to the nearest one-thousandth. The figures in parentheses are levels of significance for n = 60. Variable values have been transformed to standard scores.

2 through 7. Only the relationship between stability (decreasing) and conflictual behavior asymmetry of relative dependence and cod ic - also exhibits a substantial positive correlation. tual behavior deviates from the hypothesized When a transformation based upon the revision relationships. On the basis of the graphic anal- of the hypothesized operational linkage sug-ysis, the relationships between conflictual be- gested by Figure 4 is employed, the relation- havior and stability and asymmetry of relative ship between asymmetry of relative depend-dependence were linearized, and a correlation ence and conflictual behavior is found to be matrix was computed for the entire set of var- significant. The hypothesized relationships be- iables. The correlation coefficients are dis- tween conflictual behavior, relative dependence played in Table 3. and contractual change also accord with our

Expectations about the relationship between predictions. the initial number of union demands and con- In addition to the lack of support for the flictual behavior are borne out by the substan- hypothesized relationship between asymmetry tial positive correlation with the indicator of of relative dependence and conflictual behavior conflictual behavior. The relationship between indicated by the zero-order correlations, the

Table 4. Selected Bivariate Relationships Controlled for Goal Divergence and Conflictual Behavior

Relationship Partial r Control Variable

Stability and Conflictual Behavior - .205 Goal Divergence ( .OW)

Asymmetry of Relative Dependence and Conflictual .I22 Goal Divergence Behavior ( .179)

Asymmetry of Relative Dependence Favoring One of the - .280 Goal Divergence Organizations and Conflictual Behavior (.010)

Goal Divergence and Contractual Change .537 Conflictual Behavior

Asymmetry of Relative Dependence Favoring One of the - .291 Conflictual Behavior Organizations and Contractual Change (.013)

Stability and Contractual Change ,097 Conflictual Behavior ( ,233)


r2= .2-i

Divergence

r2= .04 - r2= .34 r2= .05 - Contractual Stability A Conflictual

Behavior -)Conflictual Change

Behavior r2= .43

/ /-/

rZ= .06 r2=.15 Asymmctry of Asymmetry of Relative Dependence Relative Dependence Favoring One of Favoring one of the Organizations the Organizations

a. Determinants of Con~¶ictual Behavior b. Determinants of Contractual Change

Figure 8. Determinants of Conflictual Behavior and Contractual Change

first-order partial correlations (Table 4) indi-cate further difficulties in the original formulation of the relationships. The revised relationship between asymmetry of relative depend-ence and conflictual behavior and the relationship between stability and conflictual behavior are reduced below the .05 level of significance when controlled for goal divergence. Although stability remains significant a t the .06 level, the reduction in the significance of the relationship between asymmetry of relative dependence and conflictual behavior is substantial. On the other hand, controlling for goal divergence does not substantially reduce the significance of the relationship between asymmetry of relative dependence and conflictual behavior when the direction of asymmetry is specified (i.e., the relationship betwee,n the actual rather than the absolute value of the asymmetry of the relative dependence variable and the conflictual behavior variable). In addition, when conflictual behavior is utilized as the control variable to test the indepe,ndent effects of goal divergence and asymmetry favoring one of the organizations, both relationships remain signifi-cant well below the .05 level. We had expected the relationships between goal divergence and contractual change t o be substantially reduced according to the original formulation.

Based upon these findings, Figure 8 (a and b) presents revised models of the determinants of conflictual behavior and contractual change.

The major changes in the models are repre-sented by the specification of the direction of the asymmetry in explaining conflictual behavior and the addition of the direct linkage between goal divergence and contractual change. The operational linkage of Figure 7 indicates that the level of conflictual behavior decreases as a linear function of management's dependence on the union. That goal divergence ac-counts for a substantial amount of the con-tractual change in the revised model can be explained in part by inadequacies of the indicator used to measure goal divergence. The initial number of union demands appears t o represent not only the divergence, but also the convergence of goals. It is also not a good measure of what was termed the aspirational o r strength of preference dimension of the concept. The relationship between goal divergence and contractual change may be inexplicably strong for these reasons.

Another interesting phenomenon surfaces when the contractual change index is broken down into the three dimensions of salary adjustment, nonsalary cost changes, and noncost changes.78 Table 5 displays the correlations be-

78 These categories correspond to the frequentlyused empirical distinctions among wages, fringe benefits, and other terms and conditions of employment. Different terminology is used here to reflect more clearly the inclusion criteria for each category. Those provisions other than wages and salaries which either


Table 5. Zero-Order Correlations Between the Explanatory Variables and Three Contractual Change Dimensions

Explanatory Variables Contractual Change Dimensions

Salary Nonsalary Noncost Adjustment Cost Items Items

Goal Divergence

Asymmetry of Relative Dependence Favoring One of the - .I04 - .407 - .227 Organizations

Conflictual Behavior

tween the explanatory variables and these three dimensions of the inducements/contributions contract. None of the explanatory variables correlates at a significant level with salary adjustment, although they all exhibit significant correlations for the 'other two dimensions of contractual change. At least two complemen-tary interpretations may be posited to account for these relationships.

One of the interpretations involves the tem- poral aspect of the collective bargaining process in the public sector. Downing suggests that it makes "good common sense -when it comes to the question of wages and fringe benefits, you do not want to have seniority problems hanging fire."s3 The key to adducing the importance of timing in relation to the composition of the items in the inducements/ contributions contract is suggested by Stevens:

The agenda-related point of interest is that even though a number of items are on the "same" agenda, they may in a sense be accorded separate "forums." . . . The significance of the attempt to distribute items on the "same" agenda among separate forums by pressing for seriatim treat-ment is conditioned by a provision strongly im- plied by the package settlement rule: Any agree- ments or proposals relating to individual items made during the course of negotiations are understood to be provisional and may be withdrawn

entailed a direct fiscal outlay or for which cost calcu- lations could be made were vlaced in the nonsalarv cost category. Items such as* tuition reimbursemeni, uniform allowances, and meal allowances were in-cluded in this category. Among the items included in the noncost category were provisions covering con-tracting out, transfers, physical working conditions, and posting of vacancies.

7 s Thomas G. Downing, "Strategy and Tactics at the Bargaining Table," Personnel, 37 (January/Feb-ruary, 1963), 44-45.

(.214) (.001) ( .040)

- ,065 .385 .368 ( .309) ( . o w ( .@2)

if final agreement upon all items in dispute is not reached. . .

The importance of these statements for the discussion at hand rests on several assumptions. First, collective bargaining in the public sector may be partitioned into at least three time periods, each associated with an appropriate forum: (1) pre-impasse, two-party bargaining; (2) impasse, mediated bargaining; (3) post-mediation, power bargaining. Second, the grouping of agenda items for seriatim treat- ment depends on the organizations involved and varies from situation to situation. Noncost items may, for example, normally be agreed upon first in negotiations, but the second as-sumption suggests that there is sufficient varia- tion from case to case so that the items which represent subordinate differences will vary. Finally, any package settlement should be viewed as having an upper bound above which management will refuse to agree and a lower bound below which the union will refuse to agree.

Within this framework, the statistical re-lationships between the dimensions of the contractual change index and the explanatory variables can be more easily understood. For example, take the case in which the agenda orderings of the organizations are consonant and salary is agreed upon in the first time period. Assume that agreement on this item brings the organizations close to their settlement points. E~~~ if bargaining proceeds. beyond this forum, it seems unlkely that adjustments in the other time periods will sub- itantially the focus ofsettlement in the

period' agreement On

80 Carl M. Stevens, Strategy and Collective Bargain- ing Negotiations (New York: McGraw-Hill, 1963), pp. 44-45.


salary does not preclude the importance of the explanatory variables in determining nonsalary cost and noncost outcomes in succeeding forums and time periods.

A second explanation relates to the visibility of salary adjustments and the proportion of the settlement allocated to each of the items in the inducements/contributions contract. The magnitude of the salary adjustment may vary inversely with the explanatory variables be-cause of the need for the organizations to maintain or enhance future task-environment support. A weak union, for example, may pursue a large wage increase in preference to other inducements because of the immediate and visible gains its leaders are able to provide its rank-and-file members. Conversely, once the salary demands of strong unions become publicly visible it may be necessary for them to accept other, less visible inducements so that their current gains are not translated into loss of support in the future.

Conclusions

A summary of findings on the original theoretical propositions and operational linkages is presented in Table 6. Although these results do not advance our understanding of public sector collective bargaining processes suffi-ciently to draw any firm conclusions beyond the scope of the research, three areas of practical implication are suggested to us by this research: the need to give greater emphasis to public interest considerations in the collective

bargaining process; the consequences of col-lective bargaining for internal resource alloca-tion in public organizations; and the utility of the model for evaluating alternative policies or policy changes.

The research indicates that perhaps the scope of the public interest is too narrowly defined in the collective bargaining process. In the bi- lateral structure of negotiations the broader needs of the community tend to be ignored. As Kochan notes: "The issue which now must be confronted is how to devise a bargaining system for municipal employees which insures them the rights of 'first class citizenship' and still provides a voice for the pluralistic inter-ests of the c ~ m r n u n i t y . " ~ ~

A second area of practical implication, re-lated to public interest considerations, is the consequences of collective bargaining for the allocation of resources among subunits of the public organization. When the outcome of public sector collective bargaining tends to allocate scarce governmental resources according to union power, severe wage and benefit imbal-ances can develop which may have serious programmatic and personnel consequences. For example, program outputs may be reduced in one agency to maintain wage and benefit levels in another. Also, the betterment of working conditions for employees in some units may be

81 Thomas A. Kochan, City Employee Bargaining With a Divided Management (Madison, Wis.: Indus- trial Relations Research Institute, 1971), p. 68.

Table 6. Summary of Findings of the Theoretical and Operational Relationships

Relationship

Goal Divergence and Conflictual Behavior

Stability and Conflictual Behavior

Asymmetry of Relative Dependence and Conflictual Behavior

Conflictual Behavior and Contractual Change

Asymmetry of Relative Dependence Favoring One of the Organizations and Contractual Change

Confirmation Status

Confirmed

Confirmed

Disconfirmed

Confirmed

Confirmed

Expected Operational Linkage

Direction Type Limits

Positive S-curve Upper

Negative Power None

Positive Negative None skew, curvi-linear

Positive Linear None

Negative S-curve Upper and lower limits

Observed Operational Linkage

Direction Type Limits

Positive S-curve Upper

Negative Power None

-Negative Curvi- None

linear

Positive Linear None

Negative S-curve Upper and lower limits


reflected in the resentment and lower morale of employees in other units.

The model could also be useful for evaluating alternative policy prescriptions. For ex-ample, a policy change designed to control the scope of bargaining could be evaluated on the basis of how it affects the values of the goal divergence variable over time. Similarly, the comparative effectiveness of state collective bargaining policies for state and local government employees could be evaluated on the basis both of how the policies affect the distri- butions of the explanatory variables and of how they limit the degree of conflictual behavior.

The greater part of the impact of collective bargaining on government operatians and the public employment relationship most certainly lies in the future. It will probably be another decade before the rapid growth of public em-

ployee unianism slows appreciably. In the meantime, collective bargaining is bound to continue growing in use and evolving with experience. A disconcerting prospect is that collective bargaining in the public sector will continue to evolve toward the private sector model without adequate consideration being given to the differences in the two setting~.~z

82For a discussion of some of the contextual dif-ferences between collective bargaining in the public and private sectors, see George H. Hildebrand, "The Public Sector," Frontiers o f Collective Bargaining, ed. John Dunlop and Neil Chamberlain ('New York: Harper and Row, 1967, pp. 125-154; David T. Stan-ley, Managing Local Government Under Union Pres- sure (Washington, D.C.: The Brookings Institute, 1972); Harry Wellington and Ralph K. Winter, Jr., The Unions and the Cities; and Gerald M. Pops, "A Decisional Analysis of Grievance Arbitration in the Public Sector" (Ph.D. dissertation, Maxwell School, Syracuse University, 1974).

An Interorganizational Analysis of Power, Conflict, and ...jlpweb/papers/An Interorganizational...

Documents

Transcript of An Interorganizational Analysis of Power, Conflict, and ...jlpweb/papers/An Interorganizational...