The Impact of Collective Bargaining Legislation on Strike Activity and Wage Settlements

• Author: Michele Campolieti,Benjamin Dachis,Robert Hebdon
• Date: 01 July 2014
• DOI: http://doi.org/10.1111/irel.12063
• Published date: 01 July 2014

The Impact of Collective Bargaining Legislation

on Strike Activity and Wage Settlements

*

MICHELE CAMPOLIETI, ROBERT HEBDON, and BENJAMIN

DACHIS

We examine the effects of collective bargaining legislation, such as (among oth-

ers) bans on replacement workers and reinstatement rights, on private sector strike

activity and wage settlements using Canadian data from 1978 to 2008. Our esti-

mates indicate that this legislation does not have a statistically significant effect

on the incidence of strikes. However, we do find that some of the policy variables

have a statistically significant effect on strike duration and wage settlements.

Introduction

ONE OF THE KEY FUNCTIONS THAT UNIONS PROVIDE FOR THEIR MEMBERS IS THE

negotiation of collective agreements. While unions and employers may come

to an agreement after bargaining, sometimes there is a breakdown in negotia-

tions, resulting in an impasse that cannot be resolved. The consequence of this

impasse or breakdown in negotiations after exhausting third-party procedures

may be a strike (or a lockout, which are much less frequent). Since strikes can

be costly, from both an economic and social perspective, labor relations legis-

lation is often implemented to increase the efficiency of negotiations and

perhaps also reduce the consequences of strikes. For example, collective

bargaining legislation might include policy instruments that try to get the

parties to restart bargaining (e.g., some form of conciliation or cooling-off

period), or to ensure that a union’s leadership reflects the desires and wishes

of the rank-and-file members (mandatory or employer initiated strike votes) or

*The authors0affiliations are, respectively, Department of Management, University of Toronto Scarborough

and Centre for Industrial Relations and Human Resources, University of Toronto, Toronto, Ontario, Canada

Email: campolie@chass.utoronto.ca; Desautels Faculty of Management, McGill University, Montreal,

Quebec, Canada Email: robert.hebdon@mcgill.ca; C.D. Howe Institute, Toronto, Ontario, Canada Email:

ben@cdhowe.org.

Campolieti gratefully acknowledges the financial support of SSHRC. The authors would like to thank

Genevi

eve Laurence, Charles Philippe Rochon, Catharine Emberly, Sylvie Gratton, Philip Malfara, and

Victoria Hanga of Human Resources and Skills Development Canada for kindly providing information on

legislation and the data used in this analysis.

INDUSTRIAL RELATIONS, Vol. 53, No. 3 (July 2014). ©2014 Regents of the University of California

Published by Wiley Periodicals, Inc., 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington

Road, Oxford, OX4 2DQ, UK.

394

reduce picket-line violence (bans on replacement workers). While most of

these initiatives are intended to be neutral in terms of the relationship between

the bargaining parties, there can be unintended consequences of some legisla-

tion that could alter the relative bargaining power of either unions or firms.

Although there are a number of studies looking at the effects of collective

bargaining legislation on strikes, most of these analyses only contain data up

until the early 1990s (e.g., Gunderson, Kervin, and Reid 1986, 1989; Gunder-

son and Melino 1990; Budd 1996; Cramton, Gunderson, and Tracy 1999).

However, there have been some dramatic changes in the economic landscape

since the end of the sample periods in these earlier studies. The effects of glob-

alization and outsourcing on manufacturing, service industries, and industrial

relations systems in North America are now more apparent than they were in

the early 1990s (Autor 2003; Bartel, Lach, and Sicherman 2005; Gunderson

and Verma 2003; Lipsig-Mumme 2009; Peters 2002). The North American

Free Trade Agreement (NAFTA) also had an impact on the manufacturing sec-

tors as well as industrial relations systems in these countries (Lipsig-Mumme

2009; Trefler 2004).

1

In addition, union density rates in both Canada and the

United States have continued to decline (Farber and Krueger 1993; Riddell

1993; Riddell and Riddell 2004), although the private sector decline in the Uni-

ted States has been much more dramatic (Farber 2005). These changes may

have had an effect on the environment in which firms and unions operate, and

on the strategic bargaining power of unions or firms when negotiating collective

agreements. Consequently, understanding the effects of labor relations legisla-

tion on strike activity and wage settlements in this more globally integrated

economy, reflected in the more recent data, is important.

In this paper, we look at the effects of collective bargaining legislation on pri-

vate sector strike activity (i.e., the incidence of strikes as well as their length

when they occur) and wage settlements using Canadian data from 1978 to 2008.

Some (e.g., among others, Budd 1996; Cramton, Gunderson, and Tracy 1999)

have noted that Canadian data provide some advantages over U.S. data in terms

of reliability because there are mandatory reporting requirements in Canada and

the U.S. strike data since 1984 is restricted to bargaining units with 1000 or more

workers. In addition, the U.S. industrial relations system and its counterpart in

Canada are very similar in structure, since Canada adopted, with some modifica-

tions, the Wagner Act model of collective bargaining in the United States (Budd

1996; Thompson and Taras 2009). This suggests that the Canadian estimates will

likely be instructive for the United States. Our data include bargaining units from

1

NAFTA began as the Canada–U.S. Free Trade Agreement in 1988, with Mexico joining in 1994. Tre-

fler (2004) found evidence of outsourcing as well as declines in the employment in industries that experi-

enced tariff cuts in his examination of the effects of the Canada–U.S. Free Trade Agreement.

Strike Activity and Wage Settlements / 395

all ten provinces as well as the federal jurisdiction.

2

Another advantage of the

Canadian data is that labor relations legislation is set at a disaggregated level (ten

provinces as well as the federal jurisdiction), not the national level. Conse-

quently, there is cross-sectional and time-series variation in many of the legisla-

tive variables during our study period. In addition, our sample period includes

data from four decades, so it includes data that overlap with earlier studies as

well as more contemporary data. Our empirical work also makes use of more

appropriate procedures to compute standard errors and avoids the potential biases

in standard errors in earlier studies (Moulton 1990). More specifically, we use

cluster-robust standard errors that account for the fact that our legislative vari-

ables only vary at the jurisdiction and year level. Consequently, the clustered

standard errors we compute in our duration, incidence, and wage analyses are

much more conservative than those in the earlier literature, which have not incor-

porated the clustering in the computation of the standard errors.

Our results for strike incidence without clustered standard errors are similar to

earlier findings. However, using clustered standard errors we find that the legisla-

tive variables do not have a statistically significant effect on strike incidence.

Moreover, we find that we cannot reject the null hypothesis that the policy vari-

ables are jointly equal to zero. From a policy perspective our results suggest that

collective bargaining legislation is essentially neutral on strike incidence. The

findings from our analysis of strike duration are less sensitive to clustered stan-

dard errors and are similar to earlier results in many respects. We find that some

of the policy variables are associated with strong increases in the duration of

strikes (i.e., decreases in the settlement hazard). In our analysis of wages, our

preferred fixed-effect estimates indicate that bans on replacement workers and

reinstatement rights are associated with decreases in wage settlements. Overall,

our wage results suggest that the bargaining power of firms may be stronger than

in earlier studies. We present some plausible explanations for these differences

in the Empirical Results section. In addition, we also test for time variation in the

effects of the policy variables, i.e., the effects of the policy variables are different

after 1992, for all of the outcome measures we consider. We selected 1992 as the

base year because the earlier studies considering the effects of collective bargain-

ing legislation on strike activity and wage settlements have not included data

after 1992. We present tests for all the policy variables as well as for reinstate-

ment rights and bans on replacement workers (both jointly and individually).

2

According to Human Resources and Skills Development Canada, about 12,000 businesses and 820,000

of their employees, which account for about 6 percent of Canadian workers, are federally regulated by the

Canada Labour Code, i.e., they fall under federal jurisdiction. The federal jurisdiction applies to all firms

that operate across provincial borders and covers sectors such as banks as well as other financial institutions,

interprovincial and international shipping, telecommunications, and broadcasting.

396 / CAMPOLIETI,HEBDON,AND DACHIS