The Role of Shocks and Institutions in the Rise of European Unemployment: The Aggregate Evidence

Blanchard and Wolfers argue that neither adverse macroeconomic shocks alone nor adverse labor market institutions alone can account for the two central facts about European unemployment: its rise from 1.7 percent in the early 1960s to 11.0 percent in the mid-1990s, and the wide and persistent dispersion of unemployment rates across European countries (from 4 percent in Switzerland to over 20 percent in Spain by the mid-1990s). Shocks can explain the common upward trend but, the authors show, do not differ enough across countries to explain the cross-country dispersion; institutions can explain the cross-country differences but mostly predate the rise in unemployment and in many cases moved toward being more employment-friendly after the 1970s, the opposite direction from what a pure institutions story would require. Using a panel of 20 OECD countries since 1960, with two separate empirical approaches – one treating shocks as common but unobserved across countries (captured by interacted time and institution effects) and one constructing observable country-specific shock series – they find consistent evidence that the interaction between shocks and institutions is what matters: a given adverse shock produces a substantially larger and more persistent rise in unemployment in countries with poorer labor market institutions (such as high benefit replacement rates and long benefit duration, strong employment protection, and a high tax wedge) than in countries with better institutions. They interpret this as institutions determining how relevant the unemployed remain to wage-setting after a shock hits, which in turn governs whether equilibrium unemployment returns to its earlier level or stays persistently elevated.

What We Know and Do Not Know About the Natural Rate of Unemployment

Blanchard and Katz assess three decades of research on the determinants of the natural rate of unemployment and reach two main conclusions: that there has been considerable theoretical progress, organized around a framework in which the labor market’s constant “traffic” of flows generates frictional unemployment while wage-setting that departs from pure competition pins down an additional, non-frictional component; and that empirical knowledge “sadly lags behind,” with economists lacking a good quantitative understanding of the natural rate’s determinants either across time or across countries. They show the standard macroeconometric wage equation used to estimate the natural rate – in which wage inflation depends on lagged price inflation and the unemployment rate, with no separate role for the level of productivity – is in some tension with the supply-wage theory the paper develops, since on U.S. aggregate time-series data the estimated long-run adjustment coefficient toward a productivity-and-unemployment-determined wage level is actually negative, while regional panel evidence (replicating and reassessing the “wage curve” literature of Blanchflower and Oswald) suggests a long-run relation does exist, but with much slower adjustment than Blanchflower and Oswald estimated. On the rise in European unemployment since 1970, from roughly 3 percent to over 10 percent with no corresponding acceleration in inflation, they survey the leading explanations – oil and productivity shocks, taxes, labor market rigidities, skill-biased technological change, and hysteresis – and conclude that while “many suspects have been identified… none has been convicted,” closing with a call for closer integration between the empirical work of labor economists and the theoretical work of macroeconomists.

J60 | Macro Paper Warehouse

The Role of Shocks and Institutions in the Rise of European Unemployment: The Aggregate Evidence

What We Know and Do Not Know About the Natural Rate of Unemployment