Article body

This is an ambitious book. Blau and Kahn summarize the differences between labour market outcomes in the USA and other OECD countries over the last twenty years, describe the key economic events and labour market institutions that have shaped labour market performance and evaluate the research that has attempted to explain it. They manage (somehow) to do it all in a readable 266 pages, without the crutch of mathematics. The result is an impressive accomplishment that deserves to be widely cited—but it is also not surprising that some important gaps remain.

The book begins with an overview of economic performance in the USA and elsewhere, concentrating on trends in employment, unemployment, working hours, wage levels and wage inequality. In the end, Blau and Kahn will conclude that “institutions matter,” and Chapter 4 is devoted to the conceptual issue of how one might be able to distinguish the influence of institutional structure from the workings of market forces. However, a prior issue is the extent of international institutional differences in collective bargaining, wage setting, minimum wages, employment protection and mandated benefits—and many readers will likely find themselves using the succinct summary of Chapter 3 as an authoritative reference for their own research.

The book then focuses on the impact of labour market institutions on unemployment, relative wages and the gender pay gap. The puzzle is that European labour markets always have been more highly regulated than those in the USA and on average had much less unemployment until the late 1970s. Since about 1980, however, high unemployment has plagued France, Germany and several other countries. The organizing framework for Blau and Kahn’s discussion is the so-called “unified theory,” which explains all this in terms of the interaction of economic shocks and labour market institutions: “The EU experience of rising unemployment, rising real wages, and comparatively stable relative-wage levels and the U.S. experience of falling unemployment, falling to steady real wages, and rapidly rising wage inequality are two sides of the same coin. The United States permitted real and relative wages to adjust, while, in other Western nations, employment took the brunt of the shocks.” (p. 5). A key idea in this approach is the presumption that U.S. labour markets are more “flexible”—although it is notable that “flexibility” is never defined explicitly or tested directly.

Blau and Kahn summarize their own research and that of others with admirable clarity and encyclopaedic thoroughness. In the end, they conclude that although the evidence for institutional impacts on the wage structure is much stronger than for employment effects, the unified theory is a useful framework. However, the demand side of labour markets is not directly examined. Micro data on households and individuals is used extensively, but firm level micro data is not examined—an issue which is important since it is the presumed hiring behaviour of firms in response to wage (in)flexibility that is assumed to be responsible for differences in employment levels. Nor is there any consideration of macro economic demand management policies—one looks, for example, in vain for any reference to the hypothesis of Akerlof, Dickens and Perry (“Near-Rational Wage and Price Setting and the Long-Run Phillips Curve,” Brookings Papers on Economic Activity 1, 2000, 1-60) that when monetary policy is focussed solely on attaining a very low inflation target (as in the EMU), a permanently higher unemployment rate is the cost.

However, for this reviewer, the book also raises the broader issue of how we should construct, and test, theories. In economics, one often encounters the idea that theory is a way of making sense out of “stylized facts,” i.e. a plausible explanation of broad empirical generalizations. The “unified theory” is just such a simplified way of thinking about the many differences between labour markets in the USA and “elsewhere” and it has the important policy implication of a trade-off in social choices about labour market interventionism or a more “free market” approach. If a lower unemployment rate (as in the USA in the late 1990s) can only be had by sacrificing the institutional protections and lower wage inequality observed “elsewhere,” then one can perhaps rationalize both standard economic theories and US policy choices with a single model.

However, in the natural sciences, theory is usually thought of in terms of potentially refutable hypotheses. In this approach, the blurry idea of “stylized facts” is avoided and it only takes one clear counter example to reject a theory. In, for example, aeronautics a century ago, because all previous attempts at flight with heavier than air craft had failed, the prevailing “unified theory” held that such flight was impossible—but the single counter example of the Wright brothers was enough to disprove that hypothesis.

Although Blau and Kahn file numerous caveats that the trends they observe are “not universal,” their general approach is to contrast outcomes in the USA with average outcomes “elsewhere.” In making broad generalizations about “stylized facts,” exceptions to the general rule are ignored or down played. In particular, the Scandinavians, Swiss and Austrians often do not fit the “unified theory” at all well, since they combine extensive labour market regulation, low inequality and high real wages for poorer workers with unemployment that is just as low (or lower) than in the USA and labour force participation rates that are equally high. It is only when these countries are pooled with other “non-USA” countries that the broad generalization or “stylized fact” appears to emerge—an aggregation process which presumes implicitly that the information content of these exceptions is nil.

In aeronautics, it was the Wright brothers’ counter-example that was seen as having the most information content—and we have air travel today because people tried to explain why they had succeeded, where others had failed, in manned flight. If Blau and Kahn had followed this approach, they would have similarly focussed on explaining the counter examples to the “unified theory” and would have asked why some countries have succeeded in generating both less wage inequality and less unemployment. This they did not do—which implies that much remains to be learned from the international comparison of labour market outcomes.