Beyond Unions and Collective Bargaining by Leo Troy, Armonk, N.Y.: M. E. Sharpe, 1999, 224 pp., ISBN 0-7656-0469-8 (hardcover) and 0-7656-0470-1 (pbk).

Leo Troy is well-known to Canadian readers for two particularly controversial articles he has written about the North American decline in union density in which he argues that the fate of Canadian unions inevitably will follow that of our neighbours to the south. Private sector U.S. union density reached a historic peak of 36 percent in 1953 and ended the 20^th century at just over 9 percent. He has written that Canadian unions are following the same trend line, but, he asserts, this fact is being obscured deliberately by Canadians who are misreporting and misinterpreting their own data. In this latest work, perhaps mercifully he drops this controversy (except for the occasional aside), but opens another, much larger and more significant one. The book addresses the question of what is happening to the vast majority of American (and by extension Canadian) workers who are not part of the unionized workforce. Troy’s position is that there exists a system of individual representation. Troy does not follow the tendency that has become fairly widespread among industrial relations scholars to attribute much of the decline in American union density to anti-union employer behaviour. Although he readily admits that employers’ preference for a nonunion system “can be taken as a given” (p. 10), he turns his attention elsewhere for an explanation of union decline. He presents the fairly novel argument that there is a “separate and independent demand by nonunion workers for individual representation.” He arrives at this conclusion by the confluence of two presumptions. First, ninety percent of American workers are not covered by unions and hence must belong to a different system of representation. Second, every “no” vote in a union election must be an affirmative vote for independent representation. He builds a case that nonunion workers have some sort of atavistic yearning, or articulate position or preference for an individual system of representation. Troy turns industrial relations scholarship on its head by arguing that there is a “participation rate” in nonunionism (p. 16). This assertion merits serious consideration in the field of industrial relations, and I do believe it deserves sustained scholarly attention. In this book, however, principal evidence upon which this argument has been constructed is a 1984 Harris survey for the AFL-CIO. The survey found that for many American workers the union route is irrelevant to their jobs and their lives. Though the possible existence of a nonunion participation rate is an important insight, I am not satisfied that the book contains sufficient evidence to transform a rejection of unionism by both employers and workers into an endorsement of nonunionism. Further, to establish the cluster of practices that constitute a system of individual representation, Troy tends to rely on generalizing from the practices of an anonymous group of exemplar large companies (for example, which establish wage rates by systematically surveying their local labour markets) and then arguing that these rational, thoughtful practices are characteristic of the entire nonunion system. Perhaps, with evidence, the notion of an independent system might be a compelling proposition for certain sectors of the economy in which workers have high individual bargaining power and greater mobility among firms. But Troy has lumped janitors and nannies in with computer analysts and physicians. He needs a much finer and more nuanced appreciation of the labour market. I have made the argument myself that workers with high personal efficacy should be free to chose a union or an alternative institutional arrangement that suits their needs, or nothing at all—but I would never generalize from this group to all workers. It concerns me that Troy’s giant net in …