This is a study of how trade unions and firms interact during periods of large-scale job loss. It uses some simple game-theoretic models to delineate the conditions under which these interactions will or will not engender industrial conflict. It explores the utility of these models with empirical analyses of strikes over workforce reductions in two industries – automobile production and coal mining – and in three countries: Britain, Italy, and Japan. Since the same models prove useful in explaining why workforce reductions typically fail to generate industrial conflict in many other countries, I also look at mass job loss in the United States. In the United States, fluctuations in employment are chronic, but rarely do they engender organized union resistance.

The main claim of this study is that unions resist job loss only when the latter presents a severe threat to the union organization itself. This argument is quite general and could in principle be extended across the advanced capitalist countries. I have limited the empirical analysis to four countries in order to examine specific industrial disputes in detail, since the usual constraints of time and access to relevant secondary materials limited my efforts. At the same time, the four countries investigated in this study exhibit substantial historical, cultural, and political differences, as well as marked variations in their industrial and economic structures. Despite this, I show that unions' responses to job loss can be analyzed within the same theoretical framework for all four cases.