The Role of City Managers

• Published date: 01 June 1999
• Author: Jeffrey L. Brudney,Sally Coleman Selden,Gene A. Brewer
• Date: 01 June 1999
• DOI: 10.1177/02750749922064319
• Subject Matter: Articles

ARPA/June1999Seldenetal./THE ROLE OF CITY MANAGERS

THE ROLE OF CITY MANAGERS

Are They Principals, Agents, or Both?

SALLY COLEMAN SELDEN

Syracuse University

GENE A. BREWER

JEFFREY L. BRUDNEY

The University of Georgia

This study uses a surveyof 1,135 city managers to evaluate the contemporary city manager role and

to evaluate its consistency with principal-agent theory.The findings are mixed. It appears that city

councils can control city managers as principal-agenttheory suggests, but that most city councils

optforlesscomplexsolutions involving trust and role sharing with the city manager.In other words,

principal-agent theory does not fully explain the council-manager relationship. In an effort to

understand this relationship morethoroughly, this study develops a typology of city manager roles

based on the extent of their involvement in policy and the degreeof autonomy exercised.

The creation of the city manager role during the Progressiveera was founded

on strengthening the relationship between elected officials and professional

administrators (Svara, 1998, p. 53).1In the 1920s, scholars and elected officials

began to voice doubts about the power accruing to city managers in the policy

process; confusion arose concerning the appropriate role of the city manager

(Svara, 1998). As a result, more narrow definitions of the city manager role

began to emerge, and efforts were made to apply the politics-administration

dichotomyto the council-manager relationship (Golembiewski&Gabris, 1994,

1995; Nalbandian, 1989; Stillman, 1974; Svara, 1998). At an abstract level, the

dichotomy and its application to the city manager role are consistent with

principal-agent theory. Both spring from the idea that elected officials should

control public administrators (e.g., see Emmet Redford’s [1969] classic discus-

sion of overhead democracy). The problem is that the politics-administration

dichotomy does not hold up in practice; politics and administration are messily

entwined. This point is illustrated by recent research showing that city manag-

ersare activelyinvolvedin policymaking andpolitical processes such as broker-

ing council and community interests (Ammons & Newell, 1988; Berman, 1997;

AUTHORS’NOTE: This is a revisedversion of a paper presented at the 1997 Annual Meeting of the

American PoliticalScience Association, Washington, D.C., August 28-31, 1997. The authors thank

the editors and reviewers of ARPA for their comments and suggestions.

AMERICAN REVIEW OF PUBLIC ADMINISTRATION,Vol. 29 No. 2, June 1999 124-148

© 1999 Sage Publications, Inc.

124

Nalbandian, 1989, 1991; Svara, 1991, 1998). What, then, are the implications

for principal-agent theory?

This study uses principal-agent theory as a conceptual lens to view the city

manager role. A guiding question is whether city managers are principals,

agents, or both. After all, city managers are responsible to elected officials and

they are responsible for subordinate employees. These dual responsibilities cre-

ate potential for role conflict. In this and several other respects, the city manager

rolestrains principal-agent theory.If principal-agent theory does notadequately

describe the role of the city manager, what does? This forms our second guiding

question: How do city managers perceive their role?

To evaluate these issues, we examine data from the 1996 Council-Manager

Plan TaskForce Survey administered by the International City/County Manage-

mentAssociation (ICMA). One limitationof this research isthat it presents only

the agent’s perspectiveof this relationship, which may not reflect or correspond

with the views or perceptions of city council members. Nevertheless, findings

from this study provide a foundation for understanding this relationship, at least

fromthe city manager’s point of view.The analysis firstpresents descriptivesta-

tistics on five potential sources of conflict between the city manager role and

principal-agent theory: (a) the city manager’s appointment and removalauthor-

ity; (b) the city manager’s role in formulating the budget; (c) the city manager’s

involvementin the political arena; (d) communication between the city council,

city manager, and subordinate employees; and (e) the nature and extentof coun-

cil control over the city manager. The analysis then provides estimates from

ordinary least squares regression models that try to explain two of these tension

points (appointment/removal authority and communication). Finally,the article

introduces a two-by-two typology of city manager roles based on managers’

involvement in the policy process and the degree of autonomy they exercise in

directing administrative activities. The findings show that few city managers

appear to act in strict accordance with principal-agent theory.

PRINCIPAL-AGENT THEORY IN

POLITICAL-BUREAUCRATIC RELATIONS

In an organizational context, principal-agent theory suggests a hierarchical

or superior-subordinate relationship between the principal and the agent. The

theorycenters around how the principal can maintain controlovertheagent in an

efficient manner (White, 1985). In the context of political-bureaucratic rela-

tions, the theory posits that elected officials—the principals—have political

incentives to control public agencies and public employees—their agents. As in

othertheories of political control, public administrator agentsare accountable to

their elected principals. Because principals and agents are self-interested actors

with divergent interests, however, elected officials must try to maintain strict

Selden et al. / THE ROLE OF CITY MANAGERS 125