Strike Incidence and Strike Duration: Some New Evidence from Ontario

The authors use a unique longitudinal data set from Ontario, covering the years 1984–92, to estimate the determinants of strike incidence and duration. Unlike most empirical analyses of strikes, the data set for this study contains both small and large bargaining units. The authors find strong evidence that the likelihood of a future strike was lower among bargaining units that had struck before than among those that had not (the “teetotaler” effect); the longer a strike lasted, the greater was the probability of settling (positive duration dependence); and smaller bargaining units were less likely to strike than were larger bargaining units, but had longer strikes when they did strike.

The Determinants of Strike Activity: An Interindustry Analysis

The paper presents an adaptation of a Chamberlain-Kuhn costs of agreement-costs of disagreement model explaining strike incidence and duration as functions of economic factors. The emphasis is on testing this model, rather than «explaining» strike activity. The results indicate the economic model fails to explain strike incidence, but does better at explaining strike duration, given incidence.

The Empirical Relations between Employers' Striker Replacement Strategies and Strike Duration

This analysis of General Accounting Office survey data on strikes in 1985 and 1989 shows that employers' permanent replacement of striking workers was associated with longer strikes than would otherwise have occurred. Furthermore, the authors find that simply announcing the intent to replace strikers was also associated with longer strikes. Finally, when employers permanently replaced strikers, the probability of settlement fell approximately 30%. These findings are consistent with models focusing on the joint costs of a strike and on private information in the bargaining process.