The Impacts of Collective Bargaining on Local Government Services: A Review ofResearch

David T. Methé; James L. Perry

Public Administration Review, Vol. 40, No. 4. (Jul. - Aug., 1980), pp. 359-371.

Stable URL:

http://links.jstor.org/sici?sici=0033-3352%28198007%2F08%2940%3A4%3C359%3ATIOCBO%3E2.0.CO%3B2-V

Public Administration Review is currently published by American Society for Public Administration.

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/aspa.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 support@jstor.org.

http://www.jstor.orgWed Oct 3 14:46:54 2007

The Impacts of Collective Bargaining on Local Government Services: A Review of Research David T. Methe and James L. Perry, University of California, Irvine

T h e past decade's huge growth in local government expenditures, the near bankruptcy of major American cities, and the rising swell of taxpayer discontent are leading indicators of a continuing urban fiscal crisis. A belief of many local government officials, citizens, and scholars is that collective bargaining is partly, and perhaps substantially, t o blame for this fiscal crisis.' The specific impacts of local government employee unioniza- tion and collective bargaining have been debated for the past decade. This debate has yielded occasionally con-flicting and, on some issues, only fragmentary empirical evidence. This paper reviews available evidence as a means of developing generalizations about the impact of unionization and collective bargaining on local services and of identifying future research needs.

Wages and hours worked are probably the inputs most directly affected by collective bargaining. Impacts upon these inputs are translated, through other policy and strategy decisions, into decisions on the size and composition of the work force.

This review focuses primarily on three principal impacts of public employee unions and collective bar- gaining. First, the influence exerted on the inputs used t o produce local government services, especially em-ployee wages, is assessed. Has unionization and collective bargaining led t o public employee wage and benefit gains? The second issue that is explored involves the productivity of resource utilization. For example, has collective bargaining led t o rtductions in labor producti- vity or adjustments in the productivity of other re-sources? The third impact involves the effects of unionization and collective bargaining on local govern- ment expenditure levels. Have local budgets risen with increasing input costs or have local governments held budgets down by compensating, on other ways, for cost

This is a revised version of a paper presented at the annual conference of the American Society for Public Administration, Baltimore, Maryland, April 1979. The research has been sup- ported by a grant from the Faculty Research Committee, University of California, Irvine. The authors would like to thank Naomi Caiden and Ann McWatters for their helpful comments on an earlier draft.

While the growth of public employee unions and collective bargaining has stimulated research about their effects on local governments, the precise impacts remain shrouded in myths and polemics. In this review, empirical research about the effects of collective bargaining on local services is examined. A taxonomy is developed to compare and evaluate the research, the prepon- derance of which has focused on employee wages. Analysis of the research indicates that public employee gains from collective bargaining are not evenly distributed across occupational groups and that collective bargaining has contributed to increased municipal expenditures and fiscal effort. However, research still needs to be undertaken to fill significant gaps in our knowledge about collective bargaining's impacts, particularly with regard to the issues of the efficiency and effectiveness of local government services.

increases attributable t o collective bargaining? These and other issues are explored following a brief discussion of the analytic framework used for selecting and organizing the research literature.

The Analytic Framework

The framework we employ t o organize the relevant literature and, thereby, t o identify the impacts of unionization and collective bargaining on local govern- ment services is a modification of taxonomies developed by Bradford, Malt and Oates and by Burkhead and Hennigan.' The taxonomy is presented in Figure 1.

Environmental variables influence each component of the system, but they are most crucial in stimulating the need for a particular service. The environment subsumes the economic conditions of supply and demand that influence inputs and outputs and the political and legal conditions that influence service delivery. Environmental factors might exacerbate o r moderate the outgrowth of unionization and collective bargaining.

The input category in Figure 1 represents the various factors used in producing a particular service, including

David ~ e t h d received his MPA from Syracuse University and is currently a research assistant and doctoral candidate in the Graduate School of Administration, University of California, Irvine. He is specializing in public financial management.

James Perry is associate professor in the Graduate School of Administration, University of California, Irvine. He is ceauthor or author of numerous articles and ceauthor of Technological Innovation in American Local Governments (Pergamon, 1979).

360 PUBLIC ADMINISTRATION REVIEW

both the amount (e.g., number of employees and/or Review of the Literature hours-worked) and composition (e.g., number of full- time workers versus number of part-time workers) of Summary information on 20 studies that investigated human resources. Public employee unions or collective the impacts of unionization and collective bargaining bargaining might affect managerial decisions about the upon aspects of the delivery of local government services amount and composition of these resources. Wages and is presented in Table 1. We have selected research on hours worked are probably the inputs most directly municipal governments that empllyed large samples from affected by collective bargaining. Impacts upon these which statistical inferences were drawn.' Most of the inputs are translated, through other policy and strategy studies used an ex post facto research design and decisions, into decisions on the size and composition of multiple regression analysis. The studies we selected the work force. originate predominately from the field of labor eco-

The activities category in Figure 1 refers t o the rules nomics, but some come from the fields of public regarding the quality and quantity of services delivered administration and organizational behavior. and the procedures for delivering a designated level of Several conventions were followed in creating Table service. Since these rules influence the productivity of 1. First, the studies are listed chronologically t o high- employees and the efficiency of task accomplishment, light the development of this growing area of research. and because unions and collective bargaining often Second, abbreviated names of variables, rather than the influence these rules, it is expected that the productivity complete variable label, are sometimes used in the table. of the worker will be affected. The direction of the The independent variables reported in the table are effect on productivity, either an increase o r decrease, confined primarily t o those that measured variations in will depend upon the type of rule change. either unionization or collective bargair~ing.~ Finally, the

The output category of the taxonolny presented in "major findings" column in Table 1 summarizes only the Figure 1 refers t o the tasks accomplished or resources general conclusions of each of the studies. expended by the organization. Since outputs are a result Even without a detailed examination of each of the of some combination of inputs and activities, the effects studies in Table 1, several generalizations evolve from of collective bargaining at other points in the input-out- inspection of the summary information. For example, put chain might lead t o decreases or increases of output. perusing the column in Table 1 headed "government

The consequences category is not a simple derivative function studied" reveals that a great deal of research of the output component. Consequences of local services has focused on the uniformed services, i.e., police, fire, involve how citizens perceive matters of concern t o and transit. In fact, these functions are examined in a them. Despite improvements in how efficiently outputs majority of the studies. Research that looks exclusively are produced, environmental factors might hinder a at other local government functions is an exception. For proportionate improvement in the consequences of local example, little or n o research focuses on recreation, services. To the extent that collective bargaining affects library or social services. Of course, there are good outputs directly or how outputs are perceived by reasons for the predominant emphasis on uniformed citizens, the consequences of local services might be services, among them the essential nature of these altered. services, their comparability from city t o city, and the

This input-output taxonomy is used in the next availability of adequate data about these services. section t o classify the impacts of unionization and A review of Table 1 also indicates that most research collective bargaining and t o assess the adequacy and on the impacts of collective bargaining relies on data coverage of previous research. bases drawn from the 1960s and early 1970s. This

ENVIRONMENT

Amount and composi- Rules and procedures Tasks accomplished or Citizens' perceptions of tion of labor, capital and for deploying resources resources expended by changes in matters of other factors of the organization concern production

Police Services: Personnel, cars, Foot vs. mobile Number of city blocks Safety from criminal communications patrol; 1-vs. 2-person patrolled; dollars activity equipment cars expended

Library Services: Personnel, books Central vs. branch Number of books Awareness of cultural libraries loaned; dollars expended developments

FIGURE 1 A Taxonomy for Assessing the Impacts of Collective Bargaining o n Local Services

TABLE 1 'c 0 Summary of Research on the Impacts of Unionization and Collective Bargaining on Local Government Services

H + w m o Study

Year(s)Government Sample Data Were

Scope of Sample Function Studied Size Collected Dependent Variable(s) Independent Variable(s) Major Findings

Melvin Lurie, "The Effect American Transit Transit 250 1914-1949 Basic wage rate per hour Presence (absence) of a 4-10% increase in wages of Unionization on Wages Association member Payments per hour for union 6-7% increase in non-wage in the Transit Industry," organizations "time not worked" and rate earnings Journal of Political premium payments 1.5% increase in fringe Economy, 69 @ecem- Value of fringe benefits benefits ber 1961), 558-572.

Orley Ashenfelter, "The Cities with populations Fire fighters 20 1 1961-1 966 Average hourly wage Presence (absence) of a 6-16% increase in average Effect of Unionization of 25,000-100,000 Average annual salary union hourly wages on Wages in the Public Average weekly duty 3-9% decrease in average Sector: The Case of Fire hours annual duty hours Fighters," Industrial 0-10% increase in average and Labor Relations annual salary Review, 24 (January 1971), 191-202.

- -

Ronald G. Ehrenberg, Cities with populations Fire fighters 270 1969 Entrance and maximum Presence (absence) of a Presence of union has virtual- "Municipal Government of 25,000-250,000 annual salary for fire union ly no effect on hours or Structures, Unioniza- fighters Presence (absence) of a earnings tion, and the Wages of Average annual salary written contract In cities with union contracts: Fire Fighters," Indus- for all department 2-1 8%increase in hourly trial and Labor Rela- employees wages tions Review, 27 Fire fighters' annual 2-9% decrease of annual (October 1973), 3648. hours of work hours

Entrance and maximum 0-9% increase in annual hourly wage for fire earnings fighters

Average hourly wage for all department employees

Roger W. Schmenner, 11 large cities (Boston, Police and f i e 57-80 1962-1970 Minimum police and fire Percent employees High level of unionization "The Determinants of Providence, New York, fighters salaries unionized increased police-fire Municipal Employee Philadelphia, Baltimore, Teachers Average monthly earn- Presence (absence) of salaries 15%; formal collec- Wages," Review of Washington, St. Louis, General municipal ings for general formal collective tive bargaining decreased Economics and Statis- New Orleans, Denver, employees bargaining police-fire salaries tics, 60 (February San Francisco, and Work stoppages per General municipal employees 1973), 83-90. Honolulu). employee achieved no salary gains

from unionization or bar- gaining.

TABLE 1 (continued)

Year(s) Government Sample Data Were

Study Scope of Sample Function Studied Size Cdlected Dependent Variable(s) Independent Variable(s) Major Findings

James L. Fruend, "Mar- Cities with a population All 40-80 196 5-1 97 1 Percentas change in - - Percent unionization of The relationships between ket and Union Influ- over 50,000 average weekly earn- city work force wage changes and union ences on Municipal ings, 1965-1971 Public sector strike power and aggressiveness Wages," Industrial and activity are negligible Labor Relations Re- view, 27 (April 1974), 391404.

Daniel S. Hamermesh, Cities with population Unionized bus 2 3 4 8 1963-1 972 Entry wage of union- Presence (absence) of a Only in one of three data "The Effect of Govern- between 100,000- drivers ized bus drivers rela- union sets (transit) did the earn- men t Ownership on 1,000,000 Construction tive to manufacturing ings of unionized govern- Union Wages," in Daniel trades Wage of government ment employees exceed S. Hamermesh (ed.), Random sample of craftsmen relative to those of private unionized Labor in the Public and occupations that of unionized pri- employees in the same Non-Profit Sectors vate craftsmen in the same occupation (by 9-(Princeton, N.J.: Prince- same trade 12%) ton University Press, Annual earnings 1975). PP. 227-263. - --

Ronald G. Ehrenberg and Cities with a population AU 284-478 1967 Average monthly earn- Percent employees 2-16%increase in average Gerald S. Goldstein, "A over 25,000 ings represented by unions monthly earnings Model of Public Sector or employee associa- Local union/nonunion wage Wage Determination," tions differentials tended to be Journal of Urban Geographic factors larger for each function Economics, 2 (July Occupational factors than the national union/ 1975), 223-245. National union non-union differentials

affiia tion

Thomas A. Kochan and Sample of 380 cities Fire fighters 121 1972 A cumulative index of Union characteristics The legal environment had Hoyt N. Wheeler, "Muni- cipal Collective Bargain-

that were known to have engaged in collective

scores on 53 contract items; progressively

Management characteristics

the greatest effect on the contract index

ing: A Model and Andy- bargaining with a local of higher scores were as- Degree of multilateral Degree of decision-making sis of Bargaining Out- the International Asso- signed an item accord- bargaining power of the management comes," Industrial and ciation of Fire Fighters ing to the degree it Environmental negotiator had a positive Labor Relations Review, approached union bar- characteristics association with the con- 29 (October 1975), gaining goals tract index 46-66.

Paul F. Gerhart, "Deter- Sample of cities, All except 26 2 1968 A cumulative index of Statutory bargaining Statutory bargaining obliga- minants of Bargaining counties, special districts teachers and scores on 158 contract obligation tion and public policy en- Outcomes in Local Gov- and school districts of transit items; progressively Anti-strike legislation vironment were related to emment Labor Negotia- varying (but unspeci- higher scores were as- Lnternational union higher bargaining outcomes

4.. 9 C"

tions," Industrial and Labor Relations Review, 29 (April 1976),

fied) population range signed an item accord- ing to the degree it ap- proached union bar-

affiiiation Union political activity Public policy environ-

Union political activity and international union affilia- tion had no influence on

9 + w

331-351. gaining goals ment (voter sympathy) bargaining outcomes Automatic penalties against

strikers lessened union m 0

power.

0

<.> C 0 6,4-

TABLE 1 (continued)

Study Scope of Sample Government Function Studied

Sample Size

Year(s) Data Were Collected

Dependent Variable(s) Independent Variable(s) Major Findings

--

r m

22 m>

w 03 0

David Lewin and John H. Keith, Jr., "Mana- gerial Responses to

Cities with populations over 250,000

Police 46-55 197 1-1972 Minimum and maximum annual salary for police patrolmen

Presence (absence) of a union

Unionization was related to lower police salaries

$ 5 z

Perceived Labor Short- Z ages," Criminology. 14 (May 1976), 65-93.

0-30 i 5

James L. Perry and Charles H. Levine, "An Interorganizational Analy-

New York City All 60 1968-1 972 A cumulative index of changes of scores on contract items and

Union power Relative power

Union power was related to higher amounts of con- tractual change, s ea t e r

8 $

sis of Power, Conflict and separate indices for the improvements in nonsalary g Settlements in Public Sector Collective Bar-

salary, non-salar y cost, and non-cost compo-

cost items, and greater improvement in noncost

P 5

gaining," American Po- litical Science Review.

nents of the contract items Union power was not re-

'1tn cn

70 (December 1976), lated to salary adjustments 1185-1201.

W. Clayton Hall and Cities with populations Police 141 1973 Minimum and maximum Presence (absence) of a Existence of a written labor Bruce Vanderporten, over 50,000 annual salary for a written labor contract contract was not significant "Unionization, Monop- police private Presence (absence) of Formal collective negotia- sony Power and Police Average annual salary formal collective tions are related to small to Salaries," Industrial Re- for all police depart- negotiations moderate pay increases lations. 16 (February ment personnel 1977), 94-100.

Richard B. Victor, The Cities with populations Police and fire 187-209 1975 Average annual salary Percentage of the muni- The three measures of union Effects of Unionism on over 50,000 fighters Per capita full-time cipal labor force in the power exhibit remarkable the Wage and Employ- equivalent police and function who are union internal consistency ment Levels o f Police fire fighters members 7.8-1 2.3% increase for police and Firefighters (Santa Presence (absence) of a wages; 12.1-27.8% increase Monica, Calif.: The recognized union for fire wages. Rand Corporation, Presence (absence) of a Unionism increases fire P-5924, 1977). collective bargaining fighters employment, but

agreement not police employment

David Shapiro, "Relative Males 45-59 years of AU 99-1136 1971 Hourly wage rates for Whether or not wages are Unionization of white-collar Wage Effects of Unions age blue collar and white set by collective government workers does in the Public and Private collar workers bargaining not result in significant Sectors," Industrial and gains in earings Labor Relations Review. Unionization of blue-collar 31 (January 1978), workers significantly 193-203. increases wages -

W 0,W

TABLE 1 (continued)

Government Sample Yeads) Data Were

Study Scope of Sample Function Studied Size Collected Dependent Variable(s) Independent Variable($ Major Findings

Stanley Benecki, "Muni- Cities with populations All 81-347 1969-1 972 Expenditure, revenue, Percentage employees r e p Collective bargaining is asso- cipal Expenditure Levels over 10,000 debt, and employment resented by unions or ciated with higher levels of and Collective Bargain- levels employee associations personnel expenditures ing," Industrial Rela- Institutional bargaining relative to other expendi- tions. 17 (May 1978), variables tures ;larger budgets; and 216-230. lower levels of employment

Richard C. Kearney, "The southwestern cities with Police 109 1967-1976 Police personnel expen- Percentage of police em- Police unions appear to Impacts of Police Unioni- populations between ditures as a percentage ployees represented by exert a slight impact on zation on Municipal Bud- 10,000 and 250,000 of total police an organization which budgetary outcomes getary Outcomes." Paper expenditures conducts formal or in- presented at the Annual formal negotiations Meeting of the Midwest Presence (absence) of a Political Science Associa- collective bargaining tion, Chicago, April 1978. contract

Russell L. Smith and Cities with populations Fire fighters 342-41 1 1960-1 970 Annual salary for Presence (absence) of a 12% increase in wages William Lyons, "Public over 25,000 in 1960 entering employees local union 18% reduction in hours Sector Unionization and Weekly work hours Presence (absence) of a Presence of a union was Municipal Wages: The contract more significant than exis- Case of Fire Fighters." Presence (absence) of a tence of a contract Paper presented a t the dues checkoff provision Annual Meeting of the Midwest Political Sci- ence Association, Chicago, April 1978.

Philip B. Coulter, "Or- Cities with populations Fire fighters 324 1973 Expenditures Unionization Unionization was related to ganizational Effective- over 25,000 Prevention effectiveness higher expenditures, but ness in the Public Sec- Suppression effectiveness not to suppression effec- tor: The Example of Productivity tiveness or productivity Municipal Fire Protec-

z P 4. P

tion," Administrative Science Quarterly, 24 (March 1979), 65-81.

C n

5-w m 0

0

I-

S 2 w 9 P 0->-z TABLE 1 (continued) z 0

Year($ 3 Study

James L. Perry, Harold

Scope of Sample

Public transit organiza-

Government Function Studied

Transit

Sample Size

28

Data Were Collected

1977

Dependent Variable(s)

Service efficiency (labor

Independent Variable(s)

Legal, organizational

Major Findings

Organizational policies in

P I-

8$

L. Angle, and Mark E. tions in the Western productivity, operating structure, attitudinal the collective bargaining Pittel, The Impact o f Labor-Management Re- lations on Productivity and Efficiency in Urban

United States expense ratios) Service effectiveness Employee withdrawal

(turnover, absenteeism,

and policy variables agreement significantly influenced service effi- ciency and employee withdrawal

n < ? m CA

Mass Transit (Washing- tardiness) Service effectiveness was not ton, D.C.: U.S. Depart- Organizational systematically related to ment of Transportation, adaptability the labor-management Research and Special variables Programs Administxa- tion, Office of Uni- versity Research, 1979).

Jean Baderschneider, Cities with populations AU 289 1967 and Fiscal capacity (financial Presence (absence) of Some support that collective "Collective Bargaining over 50,000 in 1975 1975 capability) collective bargaining bargaining legislation is Pressure on Municipal Fiscal effort (extent legislation covering associated with ability to Fiscal Capacity and Fiscal Effort." Paper

government utilizes its financial capability)

police and fire fighters pay and substantial support that i t is associated with

presented at the An- increased fiscal effort nual Meeting of the National Academy of Management, Atlanta, August 1979.

coincides obviously with the period during which collec- tive bargaining became an important dimension of public management. This period has been characterized as an era of substantial progress for public employee organiza- tions, in terms of both membership growth and bargain- ing effect iveness Therefore, any conclusions derived from these studies must recognize the possible con-founding influences of this historical period.

A final generalization worth noting concerns the dependent variables most frequently analyzed in this research. Wages, i.e., inputs t o local services, have received by far a greater amount of attention from scholars than any other variable. This imbalance of emphasis can probably be attributed t o the tradition of relative wage research in economics6 and t o the greater ease of conducting research on wages than on output variables. However, as the chronological progression in the dependent-variable column of Table 1 illustrates, research is increasingly being focused on variables within the output and consequence categories of the taxonomy in Figure 1.

Inputs

Wages. As we noted earlier, the question researchers have most frequently asked about the impact of public sector unionization and collective bargaining concerns employee wage gains. Have municipal employees im- proved their wages through collective bargaining and, if they have, by how much? The answer t o this question seems to be yes-depending upon the occupational group involved-but wage levels overall have risen only mar-ginally because of collective bargaining. Unionized fire fighters have reaped the greatest benefits, with estimates of their hourly wage increases ranging from 2 percent t o 28 percent.' However, the magnitude of these hourly wage increases is partly attributable t o decreases in average duty hours, since annual salaries of fire fighters have been estimated t o have increased n o more than 12 percent due t o collective bargaining8

Scholars have consistently concluded that union- ized public employees do no better than union- ized private employees. Thus, the observed differences between unionized and nonunion- ized employees reflect nothing endemic to the public sector, contrary to assertions that have received wide currency.

Other municipal employees have not fared as well as fire fighters, but transit employees are probably a close second with an estimated 9-12 percent wage differential between unionized and non-unionized organizations.' The wages of unionized police employees have actually been found t o lag behind their non-unionized counter- parts in two studies, although two more recent analyses found that police unionization had a positive influence on wages1 These contrasting results probably indicate that employees in low-wage police departments or-

PUBLIC ADMINISTRATION REVIEW

ganized principally t o increase wages, but that their efforts did not achieve instantaneous success.l

While the research results are not in complete accord, general municipal employees (e.g., highways, sewerage, sanitation, parks and recreation, and libraries) d o no t appear t o have achieved significant wage gains from unionization. Only Ehrenberg and Goldstein concluded that unionization has had an upward influence on the wages of general municipal employees." Two studies, using more recent samples, found that unionization had n o effect on the wages of general municipal em-ployees.'

Although the aggregate effects of collective bargain- ing on local employee wages would appear t o be positive, nothing sets the results for public employees apart from their private union counterparts. Scholars have consistently concluded that unionized public em- ployees do no better than unionized private em-ployees.14 Thus, the observed differences between unionized and nonunionized employees reflect nothing endemic t o the public sector, contrary t o assertions that have received wide currency.'

Fringe Benefits. The price of labor might not be the only input that has been affected by unionization and collective bargaining. Although research on employee benefits has been quite limited, there is some indication that employees, by organizing and bargaining collective- ly, are more likely t o improve their benefits and working conditions than they are wages. Four studies that have looked directly a t fringe benefits concluded that em-ployees achieved fringe benefit gains through either unionization or collective bargaining.l Kochan and Wheeler's finding that collective bargaining has contri- buted t o significant gains in overall bargaining outcomes for fire fighters indicates that significant wage increases for fire fighters have not dampened their success in achieving non-wage gains as well. There is some support, however, for the belief that unions gradually shift their relative priorities between wages and fringe benefits as they become more established. For example, in his study of transit motormen, Lurie concluded that after 1938 transit unions placed a higher priority on securing fringe benefits than they had in the preceding two decades.' '

Employment. An enduring controversy in economics involves the disemployment effects of uni0nisrn.l The controversy involves the extent t o which bargained increases in wages and other terms of employment will result in reduced levels of employment. This controversy is particularly meaningful t o the current predicament of municipal governments. Given the "essential" nature of many municipal services, i t seems highly unlikely that increases in the price of labor could be offset by reductions in employment. This does appear t o be the case for uniformed services, where the employment levels of police have remained constant and fire fighter employment levels have actually increased despite in- creases in the price of 1abor.l However, the inelasticity of demand for these occupational groups is not charac- teristic of all municipal services. Benecki found that

367 COLLECTIVE BARGAINING FOR LOCAL SERVICES

unionization significantly reduced overall employment levels, particularly in large cities.' O These reductions in employment, of course, might result from a variety of sources, among them decisions by the local legislature t o decrease the quantity of municipal services in the face of higher labor costs, substitution of new technologies for relatively more expensive labor, o r increases in the productivity of existing employees.

Environmental Moderators. As we noted in our discussion of the taxonomy, environmental factors affect other variables within each component of the model. Environmental influences are particularly impor- tant in assessing the effects of unionization and collec- tive bargaining on local government inputs. Two specific environmental influences, the city government structure and the degree of monopsony power held by the city, seem especially likely t o moderate the influences of unionization and collective bargaining on local govern- ment inputs.

Three types of city government structure have been studied in assessing the impacts of collective bargaining: city manager, commission, and mayor-council. It is generally believed that a manager or commission type government, because it is more professional, will be able t o hold the line against unions more effectively than a mayor-council form of government. However, manager and commission structures d o not appear t o moderate union influences o n wages and fringe benefitsS1 In fact, there is some indication that the wages of union members might actually be higher in the city manager cities, bu t these wage differentials are probably offset by higher employee p r o d u ~ t i v i t y . ~

Another factor that may blunt the power of a union is the monopsony power held by the municipal em-ployer. Monopsony refers t o the demand side of the demand-supply equation. It is analogous t o a monopoly, but, instead of one supplier, there is only one purchaser of a good or service. Thus, the degree t o which a city is the only buyer of particular labor skills in a given geographic area, it exercises monopsony power. With only one exception, research indicates that city monop- sony power does dampen union i n f l ~ e n c e . ~ However, Hall and Vanderporten suggest that monopsony power might moderate union influence only under limited c i r c ~ m s t a n c e s . ~ Hall and Vanderporten found that the ability of geographically isolated cities t o wield monop- sony power is not sustained when employees have achieved formal bargaining rights. In such cases, bilateral monopoly settlements yielded salaries similar t o those reached in cities located in more competitive urban areas. Thus, a strong union might be able t o overcome the monopsonistic power of a city.

Other research, while not testing the monopsony issue directly, has looked into differentials in union power associated with city or metropolitan area size. Both Ashenfelter and Ehrenberg found that the union had greater impact upon wages in smaller cities than in larger cities.' Gerhart, in assessing aggregate contract outcomes, discovered that unions operating in small

SMSAs fared better than unions in jurisdictions outside of SMSAs or than unions in large SMSAS. '~ Surprising- ly, contract index scores were lowers in large SMSAs. Among his explanations for the relative success of governmental jurisdictions in large SMSAs was that management might be more sophisticated and that the process may be more political, and hence informal, in the largest cities. Both of these would have a tendency t o decrease union power relative t o management. Thus, it appears that employer monopsony can have a dampen- ing effect on union power, but only in limited circum- stances.

Activities

While a great deal of heat has been generated in the last several years about collective bargaining and em-ployee productivity, work rules, and policy encroach- ment, the research t o date has shed little light upon these controversies.' ' Although his research was in-tended t o be exploratory rather than conclusive, Stan- ley's multiple-case study remains the most comprehen- sive analysis of collective bargaining and municipal a c t i v i t i e ~ . ~ His general conclusion was that unions had won victories over working conditions in a narrow sense, but that management maintained firm control over policy determination and operations management.

In the decade since Stanley's research, n o one has undertaken t o reassess systematically his conclusions. However, two recent studies have looked a t various aspects of collective bargaining and work management. Coulter examined the influence of unions on the productivity of fire departments (as measured by the total cost of fires, i.e., expenditures plus property loss, per ~ a p i t a ) . ~ Using a multivariate procedure called discriminant analysis, Coulter concluded that unionism neither decreased nor increased productivity.

Clues t o the reasons why unionization or collective bargaining might have n o net effect on productivity are provided in a recent study of public mass transit o rgan i~a t ions .~ Collectively-bargained work rules, gen- erally believed t o reduce productivity, were found t o have both positive and negative effects on labor produc- tivity and operating costs. This finding confirms Stan- ley's earlier assertion that "some forces tend t o offset each ~ t h e r . " ~ Overall, however, the transit study' concluded that productivity and efficiency could be improved if labor would agree t o the relaxation of certain scheduling and guaranteed-minimum work rules, and if both labor and management cooperated in developing work attendance incentives and improving grievance procedures.

Outputs

Although the research is far from voluminous, we know much more about the impact of collective bargaining on municipal outputs than about its impacts on municipal activities. For instance, widespread consen- sus surrounds the conclusion that collective bargaining has driven municipal expenditures upward.j In the

most comprehensive study of the influences of collective bargaining on expenditure patterns, Benecki found that unionization was associated with higher personnel ex-penditures and higher overall expenditures. These re-sults, however, were not uniform for large and small cities. In the largest cities, unionization was associated with lower levels of governmental activity because of a highly significant and negative relationship between unionization and employment. On the other hand, unionization in the smallest cities was associated with higher levels of governmental activity related t o signifi- cantly higher personnel expenditures.

A possible reason for this disparity between large and small cities is that larger cities tend t o offer a greater quantity and variety of services while smaller cities might offer an irreducible core. The greater quantity and variety offers a source of slack t o the larger cities. Benecki speculates that large cities can eliminate peri- pheral programs and still maintain their core while smaller cities ~ a n n o t . ~ Thus, large cities are more likely t o cut back on output and incur disemployment. Smaller cities are more likely t o incur higher levels of expendi- tures, especially in the personnel sector, and t o imple- ment some revenue stopgap, e.g., incurring short-term debt, in order t o cope with the aftermaths of unioniza- tion and collective bargaining.

The differences between large and small cities might also be accounted for by differences in the development of unionization in the two size classification^.^ Larger cities exhibit greater unionization, which might reflect longer-standing bargaining relationships. If this is true, then the adjustment mechanism used by smaller cities, that of expenditure adjustment, may represent the short-run impact of bargaining. In the long run, however, employment might be expected t o decline in smaller cities just as it has in large cities.

In the fact of expenditure increases, revenues and taxes have risen correspondingly. Schmenner found tax rates positively correlated with police-fire wage settle- m e n t ~ . ~However, when the tax rate was lagged back s

two years, it was negatively correlated with police-fire wages. Perhaps the most plausible explanation for these temporal differences is that, in the short run, politicians seldom react t o negotiated wage increases by increasing taxes. However, local decision makers must ultimately raise taxes as a result of negotiated wage increases. In a similar vein, Baderschneider found that fiscal effort was greater in jurisdictions that were covered by collective bargaining legislation for police and fire fighter^.^

Consequences

Like the activities category of the taxonomy, very little research has been directed at discovering how and in what ways unionization and collective bargaining might affect local service effectiveness. Two studies, focusing on fire departments and public transit agencies, have explored simple associations between collective bargaining variables and measures of service effective- ness. Coulter examined the relationship between union-

PUBLIC ADMINISTRATION REVIEW

ism and fire suppression effectiveness (dollars of proper- ty loss per 1000 p o p ~ l a t i o n ) . ~ Perry, et al . , looked at " the associations of a variety of legal, organizational, and policy variables with two measures of transit service effectiveness (revenue passengers per service-area-popu- lation and revenue-passengers per revenue vehicle hour).3 Neither study found any significant relation- ships between the effectiveness measures and the inde- pendent variables.

. . . productivity and efficiency could be improved if labor would agree to the relaxation of certain scheduling and guaran teed-minimum work rules, and if both labor and management cooperated in developing work attendance in- centives and improving grievance procedures.

These results are obviously not sufficient t o warrant sweeping conclusions about collective bargaining and the consequences of local services, but they clearly call in to question the presumption that bargaining influences significantly the consequences of local s e r ~ i c e s . ~ What might explain the lack of association? One explanation might simply be that the only two services that have been analyzed thus far-fire and transit functions-are not representative of the relationships that exist for most local government functions; analysis of other, non-uniformed services might possibly produce different results. Another explanation is purely methodological. Considering the probably large number of determinants of the effectiveness of any local government service, the influences of collective bargaining might well be indete- ctable, especially given the crudeness of most measure-

TABLE 2 Summary of Findings About the Impacts of

Collective Bargaining on Local Services

INPUTS: Some occupational groups (e.g., fire fighters and transit operators) have achieved signifi- cant and substantial gains, other groups (police) have experienced marginal irn-provements, and some occupational groups have achieved little or no demonstrable im- provement in wages or fringes because of collective bargaining. Reductions in em- ployment appear to be mast severe in very large cities.

ACTIVITIES: Little of a general nature is known, except that collective bargaining continues to stimulate both positive and negative influ- ences on productivity and work management.

OUTPUTS: Collective bargaining has driven municipal expenditures and fiscal effort upward.

CONSEQUENCES: Limited evidence indicates that collective bargaining has had no impact on the effec- tiveness of local senices.

369 COLLECTIVE BARGAINING FOR LOCAL SERVICES

ment methods. Of course, the absence of an association between bargaining and consequences could also indicate that the consequences dimension of the taxonomy is much more complex than the variables that have been studied. For example, if collective bargaining has contri- buted t o redistributions of local government outputs, then any attempt t o measure collective bargaining's impacts should probably focus on citywide rather than service-related measures of effectiveness.

Conclusions

This paper has reviewed empirical research on the impacts of unionization and collective bargaining on local government services. Our conclusions about these impacts are summarized in Table 2. In retrospect, this collection of research findings represents significant strides beyond our knowledge of a few years ago. However, important questions about impacts within each category in the taxonomy remain t o be investigated and answered.

The various studies have shown that unions probably d o contribute positively t o employee wages. The precise magnitude of this influence is difficult t o judge. While more work needs t o be done in determining the effect, greater emphasis should be given t o non-uniformed employees. More specifically, studies which look at several categories of employees should be conducted. Furthermore, total compensation should be included more frequently in this research. Unionization and collective bargaining probably influence total compen-

sation, but the magnitude is difficult t o judge given our existing knowledge.

There is also a need t o determine if collective bargaining encourages management t o substitute capital for labor and, if so, how much. Furthermore, changes in the occupational composition of the public workforce also need t o be assessed. This type of research might shed light on the role unions have played in local government innovation. Better methods for measuring least-cost combinations of capital and labor might first have t o be developed.

As public employee unions expand and mature, bargaining emphasis appears t o change. Hours, fringe benefits, and work rules may take on greater impor- tance. If this is so, productivity bargaining may become a dominant form of bargaining in the future. It will be crucial t o better understand this type of bargaining in order t o avoid the pitfalls that Horton uncovered in New York City.4 O Methodologies must also be developed for assessing the productivity-related implications of changes in work rules.

It is with regard t o local government outputs and consequences that the most glaring research needs arise. Is it city size along that is associated with the differential union effect on expenditures o r some other variables? Virtually n o research has explored unionization's effects on consequences. Only through a more complete under- standing of these and other impacts will we be able t o sweep away the myths surrounding local government bargaining and replace them with more informed judg- ments.

Notes

1. For a sampling of opinion and evidence about the extent to which collective bargaining has affected the fiscal health of local governments, see Harry H. Wellington and Ralph K. Winter, The Unions and the Cities (Washington, D.C.: The Brookings Institution, 1971); Arvid Anderson, "Local Gov- ernment-Bargaining and the Fiscal Crisis: Money, Unions, Politics, and the Public Interest," Labor Law Journal, 27 (August 1976), 512-520; Edward M. Gramlich, "The New York City Fiscal Crisis: What Happened and What Is To Be Done?" American Economic Review, 66 (May 1976), 415-529; and Robert B. P e n n e n d and Jogindar S. Uppal, Can Cities Survive? The Fiscal Plight of American Cities (New York: St. Martin's Press, 1974).

2. See D. F. Bradford, R. A. Malt, and W. E. Oates, "The Rising Cost of Local Public Services: Some Evidence and Reflections," National Tax Journal, 22 (June 1969), 185-202 and Jesse Burkhead and Patrick J. Hennigan, "Productivity Analysis: A Search for Definition and Order," Public Administration Review, 38 (January/February 1978), 34-40.

3. These criteria have led to the exclusion of research on teachers as well as single- and multiple-city case studies. A moderate amount of research has focused upon the impacts of collective bargaining in public education. See, for

example, W. Clayton Hall and Norman E. Carroll, "The Effect of Teachers' Organization on Salaries and Class Size," Industrial and Labor Relations Review, 26 (January 1973), 843-841, and Gary A. Moore, "The Effect of Collective Bargaining on Internal Salary Structures in the Public Schools," Industrial and Labor Relations Review, 29 (April 1976), 352-362. The primary reason for excluding this research is the difference between the institutional contexts of public education and municipal governments. Differences in the composition of their workforces and their legal status threaten the validity of any comparisons.

Case studies tend to be more useful for generating than for testing hypotheses. Although case studies are not included in our review, both single- and multiplecity case studies have been useful for assessing collective bargainings' impacts. See, among others, David Stanley, Managing Local Government Under Union Pressure (Washington, D.C. : The Brookings Institution, 1972), and Raymond D. Horton, "Productivity and Productivity Bargaining in Government: A Critical Analysis," Public Administration Review, 36 (July/August 1976), 4 0 7 4 14.

4 . It should be noted that unionization and collective bargain- ing are not one and the same. The term unionization is usually used to denote the proportion of a work force that

COLLECTIVE BARGAINING FOR LOCAL SERVICES

Wheeler, op. cir., Gerhart, op. cit. and Perry and Levine, op. cir.

28. Stanley, op. cit. 29. Philip B. Coulter, "Organizational Effectiveness in the

Public Sector: The Example of Municipal Fire Protection," Administrative Science Quarterly, 24 (March 1979), 65-81.

30. James L. Perry, Harold L. Angle and Mark E. Pittel, The Impact of Labor-Management Relations on Productivity and Efficiency in Urban Mass Transit (Washington, D.C.: U.S. Department of Transportation, Research and Special Pro- grams Administration, Office of University Research, 1979).

31. Stanley, op. cit., p. 139. 32. See Benecki, op. cit.; Richard C. Kearney, "The Impacts of

Police Unionization on Municipal Budgetary Outcomes." Paper presented at the Annual Meeting of the Midwest

Political Science Association, Chicago, April 1978; and Coulter, op. cit.

33. Benecki, op. cit. 34. Zbid. 35. Schmenner, op. cit. 36. Jean Baderschneider, "Collective Bargaining Pressure on

Municipal Fiscal Capacity and Fiscal Effort." Paper pre-sented at the Annual Meeting of the National Academy of Management, Atlanta, August 1979.

37. Coulter, op. cit. 38. Perry, Angle, and Pittel, op. cit. 39. See, for example, James L. Perry and Carder W. Hunt,

"Evaluating the Union-Management Relationship in Govern- ment," Public Administration Review, 38 (September/Octo-ber 1978), 431436.

40. Horton, op. cit.