Legislation prohibiting the temporary hiring of replacement workers during strikes has been one of the most controversial aspects of the Quebec model since 1978. British Columbia, the only other jurisdiction in North America with that type of law, passed its law in 1993. A similar law in Ontario lasted only from 1993 to 1995.
One common justification is that allowing the use of replacement workers leads to prolonged and tougher labour conflicts and to more strikes and lockouts. Premier Bernard Landry stated some days ago with regard to the strike by Vidéotron employees (who come under federal labour law) that if Quebec law applied, maybe the conflict would have been avoided, or at least shortened. »
However, a study by U.S. researchers affiliated with the National Bureau of Economic Research, in collaboration with a Canadian university professor (Peter Cramton, Morley Gunderson and Joseph S. Tracy: « Impacts of Strike Replacement Bans in Canada, » Labour Law Journal, Vol. 50, 1999), does not seem to confirm such predictions. The study examines 5,819 contracts negotiated at large private-sector companies in Canada between January 1967 and March 1993. (A larger sample that includes some smaller companies, for which fewer data are available, provides similar results.)
The results, strongly influenced by the Quebec experience, show that the prohibition of use of replacement workers is associated with more frequent and longer strikes. Speaking in quantitative terms, these laws increase the length of strikes by 32 days, a figure statistically very significant. In other words, the average length of a strike is 86 days if there are restrictions on the hiring of replacement workers and 54 days in the absence of such laws. (The average length of strikes during the period was 59 days.) The prohibition on replacement workers is the most important of several other policy variables that were considered (such as the use of a conciliation officer or conciliation council, the mandatory strike vote and the compulsory dues check-off.)
Another result relates to the effect of labour legislation on the frequency of strikes. The authors find that the probability of a strike occurring during the negotiation of a contract is 27 per cent when there is a replacement-ban law and 15 per cent when no such law exists, a 12 percentage-point increase. These results suggest that legislation forbidding the replacement of striking workers is not necessarily effective from an economic standpoint. In fact, it increases the probability that a strike will occur and that it will last longer, contradicting the prevailing perception in Quebec.
Another important effect of the legislation is that on real wages. The study finds that the ban on replacement workers leads to an increase in real wages of 2 per cent per year on average, relative to a situation where firms can hire replacement workers. The authors estimate that the greater wage gains exceed the losses in terms of additional strike days and greater strike incidence.
This could help explain the results of longer and more frequent strikes and justify the attraction of strike threats for unions in the presence of this legislation. The unions, whose power is enhanced by replacement bans, can endure longer strikes durations, hoping that the outcome will be higher wages. These wage increases in the long run more than compensate the short-run losses. The employer also prefers to wait before settling, because his losses in the long run in terms of higher wages to be paid exceed his short-run costs, in terms of foregone production. Higher wages, when not associated with higher productivity, affect the firms’ profitability and eventually its survival, especially in the context of increased competition and globalization. The firm’s willingness to endure a costly strike or holdout is one means of convincing the union to accept lower wages.
Other arguments are advanced for or against replacement-ban legislations, such as the balance of negotiating power between employers and unions and the reduced risk of picket-line violence during strikes. These assertions deserve a separate analysis. It must be admitted, however, that the usual argument which suggests that preventing the use of replacement workers helps reduce the frequency and length of strikes does not seem to be confirmed by the data currently available.
Norma Kozhaya est économiste à l’IEDM.