Labor Law and Reaching a First Collective Agreement: Evidence From a Quasi‐Experimental Set of Reforms in Ontario

• Date: 01 July 2013
• Author: Chris Riddell
• Published date: 01 July 2013
• DOI: http://doi.org/10.1111/irel.12030

Labor Law and Reaching a First Collective

Agreement: Evidence From a Quasi-Experimental

Set of Reforms in Ontario

CHRIS RIDDELL*

This paper examines the effects of a major labor law reform package in the Cana-

dian province of Ontario on the negotiation of first agreements for newly certified

bargaining units using a quasi-experimental research design. The findings indicate

that first contract success rates were 8–14 percentage points higher under the

regime that included automatic first contract arbitration relative to the comparison

regime of “no-fault”first contract arbitration, despite imposition rates being only

2–4 percentage points higher. Further, in the more hostile-to-labor regime, only

38 percent of petitions ultimately reached a first agreement despite the presence

of quick-votes, and “no-fault”first contract arbitration.

Introduction

INDUSTRIAL RELATIONS SCHOLARS HAVE DEVOTED CONSIDERABLE ATTENTION TO

EXAMINING the role of legislative reform in affecting the changing fortunes of

unions. Studies from the United Kingdom and Canada in particular demonstrate

a strong link between changing labor laws and declining unionization rates while

recent literature from the United States focuses on American unions’efforts to

bypass the statutory component of union organizing altogether. These literatures

have generally focused on the union organizing—or certification—stage. We

know much less about the role of labor law reform in the next stage in the

“unionization”process, the first contract stage. As demonstrated recently in a

key paper for the United States by Ferguson (2008), the first contract stage may

be as important a constraint for unions as the certification stage. In particular,

Ferguson found that among those few bargaining units to successfully certify

only around half actually negotiate a collective agreement.

In Canada, most jurisdictions allow for some form of first contract arbitration

(“FCA,”also referred to as first agreement arbitration) where, for example, if

The author’saffiliation is ILR School, Cornell University, Ithaca, NY, USA.. Email: Cr443@cornell.edu.

INDUSTRIAL RELATIONS, Vol. 52, No. 3 (July 2013). ©2013 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.

702

negotiations break down either party can apply to the provincial Labour Rela-

tions Board to have the Board impose a settlement. In general, from an empiri-

cal standpoint, little is known about FCA in Canada, particularly in regard to its

possible role in reaching a first agreement. In fact, it has been difficult to study

the first contract negotiation process at all in Canada due to data constraints.

While interest in the Employee Free Choice Act (“EFCA”) in the United

States appears over for the time being, one key reform of that proposed legis-

lation is first contract arbitration; in particular, automatic FCA—one form of

FCA featured in our study. Some analysts believed nonautomatic FCA would

have to be a compromise policy for EFCA to have any chance of passing.

First contract arbitration has also featured prominently in labor law reform

in other countries. In Australia, first contract arbitration—called workplace

determinations—was a key component of the 2009 Fair Work Act (Brody

2010). In the U.K., first contract arbitration also entered the debate in the 1999

Employment Relations Bill, which brought in a statutory union recognition

procedure, although unlike Canada and Australia, the U.K. reforms ultimately

did not provide for a true arbitration component (Wood and Godard 1999).

The main purpose in having FCA—at least from the standpoint of some

analysts and scholars—is to acknowledge that a bargaining unit that has suc-

cessfully certified has a reasonable expectation that such support for a union

should be matched by a bargaining relationship with the employer. However,

achieving such a relationship is far from guaranteed in some jurisdictions. This

leads to the question of what role FCA, but also the labor law regime gener-

ally, plays in securing a first agreement.

While it may seem obvious that FCA will lead to higher rates of securing a

first agreement, this has not been empirically established. Moreover, first con-

tract arbitration can take different forms. Only a handful of U.S. studies has

had the data to empirically analyze first contract success, and given the lack of

variation in first agreement dispute resolution, those studies have not been able

to analyze the effects of an interest arbitration procedure specifictofirst agree-

ments. In Canada, there has been almost no empirical examination of first con-

tract success, likely due to the lack of a reliable, systematic first agreements

database over time in any province except for Ontario—Canada’s largest prov-

ince. Furthermore, similar to the United States (Ferguson 2008), because col-

lective bargaining in Ontario is under the jurisdiction of the Ministry of

Labour, but union certification applications are under the jurisdiction of the

Ontario Labour Relations Board, there is no system in place for matching the

two databases making the creation of a merged database a daunting task.

This article develops a framework, compiles, and matches the two databases

to empirically examine the effects of two different FCA models in Ontario over

the 1991–1998 period. In 1991, Ontario had a system where FCA was available,

Labor Law and Reaching a First Collective Agreement / 703

but had to be applied for and the criteria, resources, and time involved in success-

fully obtaining FCA were, arguably, considerable. Beginning in 1993, Ontario

switched to a system where arbitration for first agreements was automatically

available as long as sufficient time since certification had elapsed and concilia-

tion exhausted. And then, beginning November 1995, Ontario switched back to

the identical system in place during the 1991–1992 period. Using this variation

in FCA systems, we estimate the effects of automatic arbitration relative to non-

automatic on the likelihood of reaching a first agreement.

As a robustness check on our estimates, we then compile a database of union

certifications matched with their first agreement for another Canadian province,

Nova Scotia, over the same time period. Our research of provincial labor rela-

tions boards (union certification data) and ministries/departments of labor (col-

lective bargaining data) finds that these are the only two provinces for which

reliable data on first contract success during the 1990s (or earlier) can be created.

We use differences-in-differences estimators to examine the effect of changing

labor laws in Ontario using Nova Scotia—which did not experience any changes

in labor law—as a control group. The results are consistent with our analysis of

Ontario using the before-and-after, and then back to before research design.

We find that first contract success rates in Ontario under nonautomatic first

contract arbitration are lower than typically believed at around 60–70 percent

and thus have declined from the 80 percent rate computed in the 1970s. The

results suggest that automatic first contract arbitration is associated with an

increase of around 8–14 percentage points relative to nonautomatic arbitration

depending on the specification with 10 percentage points a reasonable take-

away. The fairly substantial effect of automatic FCA on first contract success

comes despite the fact that the imposition rate of FCA contracts was only 2–4

percentage points higher under automatic FCA relative to nonautomatic.

The main caveat to the analysis is that we are unable to fully isolate the

effects of the first contract arbitration system from two other potentially impor-

tant reforms that were made at the same time: (1) a ban on temporary strike

replacement workers (permanents were banned throughout the entire time per-

iod) and (2) expanded powers of the Board to combine bargaining units. If the

latter procedures affect first contract success, the estimated impacts in the

paper should be viewed as the sum of the reforms.

Finally, we find it noteworthy that for the Ontario regime within our sample

period that is most comparable to the United States, only about 38 percent of

petitions resulted in first agreements. This compares to 20 percent in the United

States as recently found by Ferguson (2008). While certainly higher than the

United States, the Ontario success rate of surviving the total unionization

sequence comes from a regime that includes “quick votes,”limited first contract

arbitration, and a ban on permanent replacement workers—all components of

704 / CHRIS RIDDELL