Journal of Socio-Economics |
Vol. 31 (2002), 431–432 |
Book Review Truman Bewley, Harvard University Press, 1999, 527 pages, ISBN 0-674-95241-3 Truman Bewley's Why Wages Don't Fall During a Recession reports the detailed results of interviews of more than 300 businesspeople, labor leaders, counselors of the unemployed, and business consultants in the northeastern United States during the recession of the early1990s. It is made up of 22 chapters that explore aspects of why wages don't fall. The focus is on nominal wages, but since there was little inflation during the period when he conducted the interviews, there was little difference between nominal and real. This is both a usual and an unusual book. It is usual in that it is discussion of a natural first step in how a reasonable person would start a consideration of why wages don't fall-go out and ask people. It is unusual in that it is written by a theoretical economists and asking people is not the way theoretical economists typically go about finding information. Instead, theoretical economists develop formal models, based on standard assumptions of rationality, which they integrate into larger models and empirically test. Why did Bewley not follow this standard approach? He states simply, "I undertook this one because I could think of no other way to answer my questions." The book begins with three introductory chapters where Bewley discusses the methods he used and the methodological arguments about surveys. The middle chapters present the survey results organized by issues such as turnover, hiring, and the giving of raises. Most of these chapters have a brief introduction explaining the issue being discussed, a quantitative overview of the interview evidence, discussion of special issues that relate to the topic, a discussion of empirical evidence, a summary of the arguments in the chapter, and an appendix on related evidence. These chapters quote extensively from the responses Bewley's interviewees made during their conversations, which is reasonable since this is primarily a qualitative survey, interviewees filled out no questionnaires, although Bewley quantified some of his results into tables. He does not report statistical significance because of the non-random aspect of tile survey, but the results are strong enough so that they are very persuasive. In terms of his major question, the results are quite convincing. It is concern over morale that leads to wage rigidity. He concludes "Most employers do not view pay cuts as a feasible alternative to layoffs, mainly because the cuts would have little impact on their need for labor." (p. 434). He argues that in economics it is normally assumed that people must be either coerced or bribed into performing tasks. Managers don't see it that way; they see quite different motivators — motivators such as maintaining spirit and morale. Bewley argues that most theories that economists put forward to explain wage rigidity fail because of the lack of realism of their basic assumptions and inconsistency with the survey evidence, although they may be correct in specific instances. For example, he reports that the shirking theory did not apply in 87% of the firms, he studied. He also provides insight into a number of smaller questions. For example, he found differences between primary and secondary job sectors. In the primary sector, wages were less flexible downward, and there was a general shunning of overqualified job applications. In the secondary sector, the wages of new hires was more flexible downward and employers were more willing to hire overqualified workers. Bewley concludes the book by building his results into a more general theory, which is an elaboration of previous morale theories of Solow and Akerlof. Like Asking About Price, the earlier study on price flexibility by Blinder et at. (Russell Sage Foundation, 1998), this book is refreshing in its concept and realization. It would be nice to think that here we seethe beginning of a new realism in economics, but I suspect not. The institutional pressure on economists to develop models independent of simple observation is just too strong. David Colander |