No abstract

No abstract

This study analyzes the relationship between strike activity and output among disaggregated manufacturing industries. A major finding is that in many manufacturing industries, strikes have no discernible effect on industry output. Even when strikes are found to have a statistically significant effect on output, the net loss of output appears to be small. Overall, the evidence suggests that the ability of nonstruck firms to increase their output, and of struck firms to draw on inventories, makes it highly unlikely that strikes in manufacturing will cause a national emergency. STRIKES have long commanded the attention of economists and other social scientists precisely because they are such striking events. Like wars, duels, and violent encounters of all kinds, the occurrence of strikes would seem to reflect either miscalculation or love of combat. When the parties to a bargain have adequate information about each other's capabilities and incentives, the carrying out of threats would appear unnecessary and, indeed, irrational. And usually such events do not happen: strikes occur only infrequently. Yet they do occur, and not only among inexperienced bargaining pairs. Moreover, when aggregated over firms or industries, strikes exhibit statistical regularities that affect economic decision making.

A s of the late 1980s, American unionism as a whole has been in decline for about thirty years. This decline has been concentrated in the unions serving workers in the private sector and contrasts sharply with the growth of public sector unions during much of the same period. This paper attempts to explain these divergent trends, but takes their existence as given. By decline (or rise) of unionism I refer to changes in the percentage (relative number) of union adherents among a specified group of workers. Though I believe that movements in the relative number of union adherents have run parallel to broader, but more vague, movements of union "influence," the reader must draw such inferences for himself as best he can.A union adherent is assumed to be both a dues payer and covered by a collective bargaining contract; possible differences between these two concepts of union affiliation are unimportant for this paper. Individuals are assumed always to act as if they had compared the costs and benefits of union adherence and chosen adherence if the benefits were greater, and otherwise if the opposite applied. As used here, costs and benefits are limited to hourly compensation (wages and fringes), employment opportunities and working conditions. While ideological preferences undoubtedly influence a small minority of workers, it is assumed that such individuals are too few to affect the trends analysed. The above assumptions obscure important interdependencies among the decisions of individual workers and those of an employee and his employer. For example, if an 1 I do not deny either that changes in attitudes toward the relative desirability of working under union conditions may have occurred from time to time, or that such changes might be of interest. However, I assume that the dominant factors affecting movements in the quantity of union adherents are changes in the benefits and costs of unionization, particularly hourly compensation and employment opportunities.

No abstract