Article Component Papers (Early Contingency Theories Of Effective Leadership)

“SUBSTITUTES FOR LEADERSHIP: THEIR MEANING AND MEASUREMENT”-

CONTEXTUAL RECOLLECTIONS AND CURRENT OBSERVATIONS

John M. Jermier* University of South Florida

Steve Kerr University of Michigan and General Electric Corporation

While in graduate school, I was fortunate enough to have Bob House as a mentor. At that time (1970), he was studying and thinking about what was to become renowned as the Path-Goal Theory of Leadership. One result of our interaction was that, while still a doctoral student, I was able to publish (with Bob and Alan Filley) in Administrative Science Quarterly, and had other research on leadership under review.

Then, as a new faculty member at Ohio State, I was again extremely fortunate to attract as mentor and research colleague another giant of leadership research, Ralph Stogdill. As a result of Ralph’s influence on my thinking and my continuing work with Bob House, I became fascinated by the concept of organizational leadership and power. The positive reinforcement I received from the field, and from Ohio State, for pursuing this topic didn’t hurt either. (I completed my Ph.D. in 1973 and was tenured less than a year later, in no small part for my published research on leadership).

During this period, however, I became increasingly troubled by the low magnitude of results yielded by my and others’ research. As most scholars are aware, statistical significance can be virtually assured by obtaining large samples, and doesn’t depend very much upon whether one’s findings are “significant” by any sane definition of the word. It seemed to me at the time that most of the leadership research, certainly including my own,

* Direct all correspondence to: John M. Jermier, College of Business, University of South Florida, 4202 East Fowler Avenue, Tampa, FL 33620; e-mail: [email protected].

Leadership Quarterly, 8(2), 95.101.

Copyright 0 1997 by JAI Press Inc. All rights of reproduction in any form reserved. ISSN: 1048.9843

96 LEADERSHIP QUARTERLY Vol. 8 No. 2 1997

could by no stretch of the imagination be called “momentous,” “important,” or “full of meaning,” as Webster’s defines the term significance.

After reading about and generating my share of what Fred Fiedler calls “the Irish correlations” (.08, .03., .05, . . .), I became curious about what did make a difference when hierarchical leadership apparently didn’t. I clung to the primordial Psychology assumption that Ability x Motivation = Performance: That for members of a work organization to perform, something has to make them able, and something has to make them want to. I didn’t discount the possibility that the source might be the formal leader, but I wanted to broaden the lens to admit other possibilities. I came to feel, as the article states, that “if we really want to know more about the sources and consequences of guidance and good feelings in organizations, we should be prepared to study these things whether or not they happen to be provided through hierarchical leadership” (p. 401, emphasis in the original). That was the thought process that led me to coin the term “substitutes for leadership.”

I’ve been told by many people (most of them Ph.D. students, possibly for the wrong reasons) that one reason they like the article is because of my use of data drawn from perceptions of popular television subordinates-namely Mary Richards, Hawkeye Pierce, and Archie Bunker. (I must have chosen well. Nearly 20 years later, they remain among the most frequently screened late night television fare). Actually, when I submitted the original manuscript for review, those were the only data in the piece. This led one reviewer to note in his skeptical review that “Kerr has produced a whale of an argument, followed by a mouse of a demonstration.” Thus I was motivated to include in the revision some of the data John Jermier and I were collecting within the Department of Public Safety of the City of Columbus, Ohio, and to welcome John as co-author of the revised version. Let me note specifically here that John, who will now pick up the thread of this narrative, strengthened the revised manuscript considerably, conceptually as well as empirically.

Like Steve, I was fortunate to learn and become inspired about the study of leadership through close contact with some legendary figures in the field. While I was a graduate student at Ohio State University in the mid-1970s Ralph Stogdill and Steve Kerr were on the College of Business Faculty, Chet Schriesheim was an advanced Ph.D. student, Bob House maintained close ties with the program in which I was enrolled (Administrative Science), and there was a continuous flow of distinguished visiting scholars, many of whom studied leadership. My primary interests were in social conflict and change, critical social science, and public bureaucracy, topics which were relevant to the field of administrative science but not specialties of anyone on the Business Faculty. I pursued these interests with scholars outside my own college and university but realized that the field of organizational behavior (the area of specialty of most of the faculty), did not (at that time) overlap appreciably with my more macro interests. It was dominated by the science and method of psychology and by theories of motivation.

When I started to become familiar with leadership theory and research, I noticed that while it had very little to do with social conflict and change or the pressing political issues of the day, it was somewhat integrated with social science and organizational theory. Steve Kerr’s incipient research on leadership was the most friendly to macro considerations of any I encountered. He seemed to make a point of insisting that processes of leadership could be understood best by examining broader contexts of power and influence. This led beyond the level of group dynamics to the possibility of linking organizational-level properties with group and individual factors.

Substitutes for Leadership 97

As was typical of faculty-student relations in the college at the time, students who

wanted to work on research were given every opportunity to do so. When I joined Steve on

this project, I had more enthusiasm and energy than skill but realized I would develop quickly by meeting the challenges involved in sharing responsibility for writing an article

for one of the field’s top journals. I learned about the work Steve had already completed by talking with him as much as by reading the paper presented at the 1976 American Institute

for Decision Sciences meeting. Our field studies were undertaken largely on the strength of

that paper and on what understandings passed between us during conversations about the

construct. Since I was new to publishing, I did not know what to make of the fact that the early draft of the substitutes piece was returned unreviewed by Administrative Science

Quarterly and returned by Organizational Behavior and Human Pe$ormance with the comment Steve noted above. I had heard Ralph Stogdill say that once a paper was rejected

outright by three top scholarly journals, it could be trash but probably had unusual merit. My sense was that the merit of the substitutes framework would eventually be

acknowledged if for no other reason than it offered a more sophisticated conceptualization

of the exercise of power in organizations than did traditional leadership theories. It pointed to unobtrusive and impersonal forces such as technology and task characteristics,

professional standards, and formal regulations (policies, rules and procedures). Thus, it brought aspects of organizational sociology into the field of leadership at a time when

many scholars had tired of the overly psychologized, methodologically-driven approaches that dominated.

As I have followed the published literature on the substitutes for leadership construct over the years, several things have occurred to me. First, there has been very little work

aimed at conceptual refinement and elaboration of the construct. To my knowledge, nearly all researchers have accepted the conceptual domain of substitutes for leadership as

originally formulated (but for interesting adaptations see Manz & Sims, 1980; Sheridan, Vredenburg & Abelson, 1984). There have been several examinations of the original substitutes in combination with a variety of leader behaviors, including some that border

on responding to Staw’s (1984) call to expand the range of outcome variables in organizational behavior research-f. Podsakoff, Niehoff, MacKenzie & Williams (1993).

But, little attention has been given to theorizing new substitutes or linkages between substitutes and various outcome variables (Jermier, 1987; McIntosh, 1988).

This is surprising since we certainly did not intend to imply that the substitutes framework was a closed system or refined to the point that intervening constructs were not needed. Indeed, better specification of the conceptual domain of substitutes and of how substitutes exert their effects on various criteria is long overdue.

Neither did we intend to discourage the investigation of combinations of substitutes. An

intriguing set of questions could be developed focused on how substitutes interact among

themselves to produce effects. Research seems to show that substitutes considered individually are more important than the leader’s interpersonal behaviors in accounting for outcomes (Podsakoff, MacKenzie & Bommer, 1996) and it may be that while effects of many of the substitutes alone are weak, when considered in combination, important thresholds are reached. In order to pursue this line of thinking, good theoretical reasons

would need to be developed explaining why the effects of substitutes in certain combinations are different from their effects when considered individually. In quantitative

98 LEADERSHIP QUARTERLY Vol. 8 No. 2 1997

research, combination effects would need to be specified and detected using nonlinear analysis or clustering algorithms (for the first steps in this direction, see Shoultz, 1991).

Second, the extensive work on substitutes as moderator variables in leadership research

(e.g., Howell, Dorfman & Kerr, 1986), while interesting, has promoted an agenda some

distance from what originally attracted me to the project. Podsakoff et al. (1996, p. 396) conclude their meta-analysis of substitutes research by stating that since few substitutes

actually serve as moderators, “more than 20 years of research on the.. .model has generally

failed to support [its] hypotheses.” In my view, however, the more interesting argument in the OBHP piece is that it is the

main effects that deserve our attention. [Admittedly, we did not emphasize this point

enough to be embraced readily by a field enamored with contingency approaches to

leadership, and it is an argument more in line with my views than those of Steve. Also, given the trajectory proposed by Howell et al. (1986), it is not surprising that researchers

would identify moderator effects as the core hypotheses in the substitutes model]. Still, to me, the greater importance of the substitutes framework lies in its emphasis on individual,

group, task, and organizational processes outside the reach of formal, hierarchical

leadership understood as “superior-subordinate interactions. ” This is what differentiates it from other models of leadership and is what it can stimulate in the way of more

sociologically inclined analysis of organizational behavior. In figure-ground imagery, it

elevates the usual background to the center of attention and challenges the view that interpersonal leadership should be seen as the primary theoretical category. If the purpose

of the substitutes framework is merely to propose moderators of the leader behavior- subordinate criterion variable relationships, as argued by Podsakoff et al. (1996), it is difficult to see what is added other than a rather minor extension to the Path-Goal Theory

of Leadership (see House & Mitchell, 1974). The moderator variable interpretation of substitutes was developed furthest by Howell

and Dorfman (198 1) and Howell et al. (1986) who proposed a three-part analytical strategy

that required (to begin) a significant correlation between the leader behaviors and the criterion of interest. I always questioned the necessity of limiting tests for potential

substitutes to situations where there is a significant cross-sectional correlational effect involving leader behaviors. I question this procedure because implicit theories of

leadership held by people bias upward their estimation of the importance of interactional

leadership. I question it even more basically because at the heart of the matter of leader substitutes

is the typical situation where leader behavior and outcome variables are weakly related or

unrelated. The substitutes model offers an explanation for this situation along the lines that leadership has already had its effects through the substitutes such that formal, face-to-face interactions would be superfluous. The fact that you cannot predict outcomes using leadership behaviors in many situations is what led to the development of the substitutes model, which proposes a wide range of variables that do impact criteria independent of the

leader’s interpersonal behaviors. It is not the study of interpersonal interactions between managerial leaders and

subordinate followers that is the core of the substitutes for leadership framework, but the idea that managerial leadership works through technological, structural, and other

impersonal processes in the organization to achieve its effects. That is, formal leaders do

Substitutes for Leadership 99

attempt to control the organization, but they do so by making decisions that minimize the need for the face-to-face exercise of power.

Another central part of the substitutes framework is that many employees do not need structuring and emotional support from their organizational leaders to do their jobs, and many will rely more on people other than formal leaders in deciding how to organize the work and in deriving satisfactions from participating in organizational life. The substitutes framework drew heavily on classic group dynamics theory and research, which always heavily contextualized the roles of formal leaders, foreshadowing the contemporary movement toward self-managed teams.

The point is that we still do not have much research on the processes through which the substitutes themselves exert their effects. For example, although one of the master trends of the past two decades has been the move toward “post-bureaucratic organizations” (Pinchot & Pinchot, 1994; Heckscher & Donnellon, 1994), a strong case can be made that a more trenchant pattern is the “McDonaldization of Society” (Ritzer, 1996), complete with well specified procedures and processes leaving little need for formal managers. In such systems, formalized practices are probably crucial in affecting organizational behavior, but little is known about how rules, policies and standard operating procedures are designed by managerial leaders or used by employees. It is not that we could not benefit from more studies of interpersonal leadership in combination with substitutes (but see point three below), but that the fields of organizational behavior and leadership would probably develop faster if attention were turned to conducting detailed studies of the substitutes themselves.

Third, leadership research (including my own work) has relied far too much on cross- sectional designs and questionnaire surveys. There have been a few studies on leader substitutes that utilized other methods (e.g., Jermier & Berkes, 1979; Pitner, 1986), but the remainder can be described as increasingly rigorous field surveys that have not paid much attention to contextual factors and have not probed very deeply into the subjective worlds of actors in the contexts. In my opinion, we have reached the point of diminishing returns from conducting and studying the type of research reviewed in Pod&off et al. (1996). I believe that my comments from ten years ago (Jermier, 1987), advocating longitudinal field research which utilizes both qualitative and quantitative methods, are still relevant today concerning tests of the substitutes model. We need to understand how substitutes get created and how leadership interactions change as a result of the introduction of the substitutes (cf. Sheridan et al., 1984). We also need to know how substitutes might differ in different types of organizations (Tosi, 1991). And, we need to know more about how people respond to various substitutes. Thus, detailed fieldwork which examines the development of substitutes over a period of time and the subjective meaning the substitutes are given by people in the setting would be informative, as would be work aimed at understanding how substitutes affect a variety of reactions by participants in the settings. Extending the time frame of observation is more important than whether the researcher uses qualitative or quantitative methods.

Another missing piece is experimental research on substitutes. Podsakoff et al. (1996) call for experimental or quasi-experimental research on the substitutes construct because it would allow causal inferences to be drawn more easily. In addition, it would allow for more rigorous examination of combinations of substitutes in relation to outcome variables and in interaction with leadership variables.

100 LEADERSHIP QUARTERLY Vol. 8 No. 2 1997

The summary point is that there is now nearly 20 years of experience with research on

the substitutes construct, and nearly all studies published fall within one genre: the cross-

sectional field survey. While this is understandable for many reasons and has elevated the

standards by which such work is produced and evaluated, the development of any model or

conceptual framework should not be dependent on only one method. According to Meindl(1993, p. 92), the original OBHP article was one of four key studies

that “dealt severe blows to [the] conventional leadership wisdom.” Today, it is often cited

in introductory organizational behavior textbooks as an example of a contingency theory

of leadership, perhaps more radical in the way it portrays the effects of interpersonal

leadership than other models but, nevertheless, part of the contemporary conventional

wisdom. While it is gratifying to have a place in any sphere of wisdom, especially along

with Steve Kerr (a central figure in leadership research who also is surely one of the field’s

most facilitative and developmentally-minded mentors), often I have thought the

substitutes framework should be seen less as a contingency model of leadership and more

(a) as a screen to be used by researchers to determine whether interpersonal leadership

should even be part of a particular study; and (b) as a framework for stimulating and

organizing research on the substitutes themselves.

REFERENCES

Heckscher, C.C. & Donnellon, A.M. (Eds.). (1994). 7’/te post-bureaucratic orgunizutinn: new

perspectives on organiz,ational change. Thousand Oaks, CA: Sage. House, R.J. & Mitchell, T.R. (1974). Path-goal theory of leadership. JnurnaE qf C~~nfernp~~ra~

Business, 3, 8 l-98. Howell, J.P. & Dorfman, P.W. (1981). Substitutes for leadership: test of a construct. Academy of

Management Journal, 24,714-728.

Howell, J.P., Dorfman, P.W., & Kerr, S. (1986). Moderator variables in leadership research.

Asu~~~~~ of ~an~~enze?zt Review, II, 8%102. Jermier, J.M. f 1987, November). Suhsiiturrs for leadership: theory development. Paper presented at

the national meeting of the Decision Sciences institute, Boston, MA. Jermier, J.M. & Berkes, L. (1979). Leader behavior in a police command bureaucracy: a closer look

at the quasi-military model. Admini.~tr~ti~l~ Science Quarterly, 24, l-23, Manz, CC. & Sims, H.P., Jr. (1980). Self-management as a substitute for leadership: a social

learning theory perspective. Academy ofManagement Review, 5, 361-368. McIntosh, N.J. (1988). Substitutes for leadership: review, critique, and suggestion. Paper presented

at the national meeting of the Academy of Management, Anaheim, CA. Meindl, J.R. (1993). Reinventin, 0 leadership: a radical. social psychological approach. In J.K.

Mumighan (Ed.), So&l P.~~c~~~l~~g~ in t~r~a~zi~at~#n,s: Advances in Theory and Research (pp.

X9- 118). Englewood Cliffs, NJ: Prentice Hall. Pinchot, G. & Pinchot, E. (1994). The end of bureuucracy and the rise of the intelligent organization.

San Francisco: Berrett-Koehler. Pitner. N. (1986). Substitutes for principal leader behavior: an exploratory study. ~du~ati~~nu~

Adn?jni,~trati~n Quarter&, 22, 23-42. Podsakoff, P.M., MacKenzie, S.B., & Bommer, W.H. (1996). Meta-analysis of the relationships

between Kerr and Jermier’s substitutes for leadership and employee job attitudes, role

perceptions, and performance. Journal ofApplied Psychology, 81, 380-399.

Substitutes for Leadership 101

Podsakoff, P.M., Niehoff, MacKenzie, S.B., & Williams, M.L. (1993). Do substitutes for leadership really substitute for leadership? An empirical examination of Kerr and Jermier’s situational leadership model. Organizational Behavior and Human Decision Processes, 54, l-44.

Ritzer, G. (1996). The McDonaldization of society: an investigation into the changing character of contemporary social life. Thousand Oaks, CA: Pine Forge Press.

Sheridan, J., Vredenburgh, D., & Abelson, M. (1984). Contextual model of leadership influence in hospital units. Academy of Management Journal, 27, 57-78.

Shoultz, D.C. (1991). Substitutes for leadership in a manufacturing support environment. Unpublished Ph.D. thesis, University of South Florida.

Staw, B.M. (1984). Organizational behavior: a review and reformulation of the field’s outcome variables. Annual Review of Psychology, 35,627666.

Tosi, H. (1991). The organization as a context for leadership theory: a multilevel approach. Leadership Quarterly, 2, 205-228.