Public Sector Unionization and Government Contracting: A Meta-analysis of Four Decades of Empirical Evidence

• Published date: 01 January 2024
• DOI: http://doi.org/10.1177/02750740231193177
• Author: Jiahuan Lu,Yiying Chen,Wan-Ju Hung
• Date: 01 January 2024

Public Sector Unionization and

Government Contracting: A Meta-analysis

of Four Decades of Empirical Evidence

Jiahuan Lu

1

, Yiying Chen

1

and Wan-Ju Hung

1

Abstract

Within the large body of literature on government contracting, the effect of public sector unionization on contracting out is

still unsettled even after decades of research. Previous literature proposes that unionization may both inhibit and motivate

contracting out, making the net effect diff‌icult to predict. Through a meta-analysis of 232 effects drawn from 49 existing stud-

ies spanning over four decades, we f‌ind that jurisdictions with higher levels of public sector unionization generally contract out

more in public service delivery. Further metaregression analysis suggests that unionization has a weakereffect when govern-

ments engage in intergovernmental contracting but a stronger effect when governments contract out for technical services.

Unionization also has a stronger effect on how much a government contracts out than on whether a government contracts

out. Overall, unionization is a relevant, but not necessarily robust, factor in driving contracting out, and its exact effect may

vary slightly by context.

Keywords

government contracting, contracting out, public sector unions, unionization

Introduction

Over the past several decades, under the new governance

model, government contracting has become a prevalent gov-

ernment tool in public service delivery (Salamon, 2002).

When providing public services to citizens, governments

may choose not to rely on their own employees (in-house

delivery) but to hire third-party contractors (contracting

out). Although contracting out has been used widely in the

public administration landscape, its benef‌its and risks are

always under scrutiny (Bel et al., 2010; Brunjes, 2022;

Hodge, 2000; Leland et al., 2021; Peng & Lu, 2021; Savas,

2000; Sclar, 2000). Given the pros and cons of contracting

out, a lingering line of inquiry in public administration

research seeks to understand what factors lead governments

to contract out service delivery rather than providing services

in-house. The large volume of literature on government

make-or-buy decisions identif‌ies a myriad of economic, ideo-

logical, managerial, and political forces (Alonso & Andrews,

2020; Bel & Fageda, 2007; Chen et al., 2022; Fernandez

et al., 2008; Ferris & Graddy, 1986; Hefetz & Warner, 2012).

Within this body of literature, an often-debated subject

is the effect of public sector unions.

1

Unions are formed to

represent the interests of government employees, and they

play an active role in shaping various aspects of government

operations. In the context of government contracting, the

effects of unionization on government contracting out are

inconclusive. On the one hand, given that contracting out

may bring a series of negative consequences to government

employees (Lee et al., 2021; Vrangbæk et al., 2015),

unions are likely to exert political pressures to oppose con-

tracting out. This argument is supported by the empirical

f‌indings that jurisdictions with more unionized public

employees are associated with lower levels of contracting

out (Foged & Aaskoven, 2017; Jerch et al., 2017;

Lopez-de-Silanes et al., 1997). On the other hand, union

activities including collective bargaining and lobbying sub-

stantially increase the labor costs of in-house delivery

(Anzia & Moe, 2015; Chandler & Gely, 1995), motivating

public managers to contract out service delivery to achieve

cost savings. This line of argument is supported by the

studies reporting that jurisdictions with higher levels of

unionization contract out less (Lu, 2013; McGuire et al.,

1987; Warner et al., 2021). In addition, these two opposing

mechanisms may cancel each other out, resulting in unioniza-

tion having a null effect on contracting out (Ferris, 1986;

1

School of Public Affairs and Administration, Rutgers University Newark,

Newark, NJ, USA

Corresponding Author:

Jiahuan Lu, School of Public Affairs and Administration, Rutgers University

Newark, Newark, NJ, USA.

Email: jiahuan.lu@rutgers.edu

Article

American Review of Public Administration

2024, Vol. 54(1) 71–89

© The Author(s) 2023

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/02750740231193177

journals.sagepub.com/home/arp

Warner & Hefetz, 2020). These competing arguments and

empirical f‌indings, although they highlight the complex

nature of the relationship, call for an effort to synthesize

existing f‌indings.

In the present study, we employ meta-analysis to synthe-

size previous f‌indings of the unionization–contracting rela-

tionship, with a focus on two questions: (1) what is the

generalized impact of unionization on government contract-

ing out, and (2) under what conditions does the impact

vary? Through a meta-analysis of 232 effects drawn from

49 existing studies spanning over four decades, we f‌ind

that jurisdictions with higher levels of unionization generally

contract out more in the delivery of public services. Further

metaregression analysis suggests that the relationship is

weaker when governments engage in intergovernmental con-

tracting (as compared to contracting with for-prof‌it and non-

prof‌it providers), but it is stronger when governments

contract out for technical services (as compared to social ser-

vices). Unionization also has a stronger effect on how much

governments contract out than on whether governments con-

tract out. Overall, unionization seems to be a catalyst of,

rather than a barrier to, contracting out, and its exact effect

may vary slightly by context. This study offers a systematic

synthesis of previous f‌indings and adds new knowledge to

the government contracting literature.

Literature Review

Public sector unions represent the job interests of public

employees and are an important force in democratic gover-

nance and public policy (Goldf‌ield & Bromsen, 2013;

Kearney & Mareschal, 2014). In the United States, after

decades of explosive growth in membership since the

1960s, public sector unions have been “formidable players

in electoral campaigns, as well as in every major sphere of

governmental decision making”(Moe, 2009, p. 156).

According to the U.S. Bureau of Labor Statistics (2022),

7.0 million employees in the public sector belonged to

unions in 2021, equivalent to a union membership rate of

33.9%, a rate more than f‌ive times higher than that of

private sector workers (6.1%). In public administration schol-

arship, a growing body of literature examines the impacts of

public sector unions on various aspects of government oper-

ations and performance (Bolton, 2021; Davis, 2011;

Nicholson-Crotty et al., 2012; Oberf‌ield, 2021; Riccucci,

2011). However, the effects of unionization on government

make-or-buy decisions are still inconclusive.

In the government contracting literature, there is much

evidence suggesting that contracting out would adversely

affect public employees on a wide range of parameters,

including job security, wages and benef‌its, working condi-

tions, and work-related attitudes. For example, Fernandez

et al. (2007) found that contracting out leads to declines in

full-time employment and increases in part-time employment

in local governments, with the net employment effect being

negative. Lee et al. (2021) reported that increases in contract-

ing activities among federal agencies reduce employees’job

satisfaction and increase their turnover intention. Hebdon

(2006) also noted that contracting out has a deunionizing

effect, causing previously unionized jobs to lose union

status after the work has been contracted out. Vrangbæk

et al. (2015, p. 17) systematically reviewed this body of liter-

ature and concluded that “the negative effects of contracting

out would outweigh positive effects for the employees.”

All these negative consequences inevitably lead unions to

oppose contracting out and to seek maintenance of in-house

delivery. Union strategies for opposing contracting out can

take different forms (AFSCME, 2013; Chandler & Gely,

1995; Flavin & Hartney, 2015; Jalette & Hebdon, 2012).

For example, unions can mobilize resources to exert pressure

on governments through collective bargaining, striking, and

lawsuits. They can also indirectly inf‌luence policy making

through public campaigns, legislative lobbying, and electoral

mobilization. Cases of unions successfully blocking contract-

ing out and keeping service delivery in-house are reported in

the news and literature (AFSCME Council 31, 2021; Isitt &

Moroz, 2007). However, although unions generally are not in

favor of contracting out and attempt to challenge it, their

opposition efforts are “neither predictable nor consistently

effective”(Naff, 1991, p. 23), which makes the effect of

unionization on government make-or-buy decisions less

clear. In particular, the relationship may proceed in three pos-

sible directions.

First, through various political and legal measures, unions

restrict the government’s consideration and adoption of con-

tracting out. This line of argument is supported by empirical

studies demonstrating a negative association between union-

ization and contracting out. For example, Chandler and

Feuille (1994) found that the existence of a municipal sanita-

tion union reduces the likelihood that a municipality consid-

ers or adopts sanitation service contracting. Foged and

Aaskoven (2017) found that higher union membership rates

among public eldercare workers are associated with lower

levels of privatization of elderly home care services among

Danish municipalities. Lopez-de-Silanes et al. (1997) noted

that even if governments initially contract out service deliv-

ery, union opposition could eventually push local govern-

ments to bring contracted services back in-house.

Second, unionization and its associated activities (e.g.,

bargaining, striking, and lobbying) increase the labor costs

of in-house delivery. Some studies suggest that unionized

employees have higher wages and better benef‌its than their

nonunionized counterparts do, which increases in-house

delivery costs (Anzia & Moe, 2015; Chandler & Gely,

1995; Jerch et al., 2017). Given cost saving as a primary

reason for contracting out, increased labor costs would moti-

vate public managers to consider contracting out to cut

service costs. This line of argument is supported by empirical

f‌indings of a positive association between unionization and

contracting out. For example, using a panel of U.S. local

72 American Review of Public Administration 54(1)