The interwar years provide a rich body of evidence concerning the causes and consequences of unemployment. Yet, despite the frequent parallels drawn between the 1930s and 1980s, the interwar experience has not received the detailed analysis it deserves. At a lunchtime meeting on 10 June, Barry Eichengreen described the results of new research using interwar data. The research is reported in a new CEPR volume, Interwar Unemployment in International Perspective, edited by Professor Eichengreen and Dr Tim Hatton. The new research confirmed the importance of analysing unemployment at the microeconomic level.

Barry Eichengreen is Professor of Economics at the University of California at Berkeley, and a CEPR Research Fellow. Interwar Unemployment in International Perspective, edited by Barry Eichengreen and T&nbspJ Hatton, is based on a conference held at Harvard University in 1987, organized by the Centre for Economic Policy Research and the Harvard Center for International Affairs, with sponsorship from the NATO Scientific Affairs Division. It is available in paperback for £24.50 from Kluwer Academic Publishers Group, Distribution Centre, PO Box 322, 3300 AH Dordrecht, The Netherlands. The opinions Eichengreen expressed at the meeting were his own and not those of CEPR, which takes no institutional policy positions.
Two contrasting views have dominated research on the interwar years. One portrays interwar unemployment as the result of economic collapse in the post-1929 slump, not as the fault of the individual. As the experience of unemployment began to undermine the employability of unemployed workers, there emerged a large pool of long-term unemployed for whom idleness was a permanent condition. The second view characterizes interwar unemployment as a voluntary practice, encouraged and subsidized by generous unemployment insurance systems.
Eichengreen noted that previous research on the interwar period had failed to exploit the exceptionally rich body of information on unemployment under a variety of institutional conditions. It has focused almost entirely on two countries, Britain and the United States. Moreover, it has been highly aggregative and so has provided little insight into who was unemployed or into the effects of unemployment. Interwar Unemployment in International Perspective attempts to broaden the discussion of interwar unemployment, through explicit international comparisons and through detailed microeconomic studies of eight industrialized countries.
According to Eichengreen, the detailed analysis of microeconomic data in the volume revealed five features of particular importance in understanding interwar unemployment. These findings not only help us to understand the experience of the 1930s, but are also relevant to the policy debates of the 1980s.
An important feature of the British labour market (like those of a number of other countries), analysed in the CEPR volume in the chapter by Mark Thomas, was its bifurcation into two segments. In one segment a large number of individuals moved smoothly back and forth between work and dole, experiencing a succession of short spells of employment and unemployment. In the other, both employment and unemployment were more lasting, and there accumulated a hard core of long-term unemployed for whom re-employment prospects were dim. The segmented structure of labour markets makes blanket statements about the effects of unemployment insurance dangerously misleading, Eichengreen argued. Benjamin and Kochin have claimed that between one-third and one-half of UK interwar unemployment (i.e. five to eight percentage points) could be attributed to the generous provision of insurance benefits. Eichengreen's own research (in CEPR Discussion Paper No. 207) and the chapter by Mark Thomas suggest that the effects of unemployment benefits were considerably smaller. It was essential to distinguish between adult male household heads, who were unlikely to remain unemployed merely because the dole replaced some fraction (e.g. a half) of their income, and `secondary workers' such as women and young men, whose labour supply was likely to be much more elastic with respect to the level of benefits.
The second feature of interwar unemployment revealed in the volume, according to Eichengreen, was structural (or `mismatch') unemployment and labour immobility. In the 1920s the UK's staple industries suffered intense foreign competition and high unemployment. Together with the costs and risks involved in long-distance migration in search of work, this led to prolonged concentrations of unemployment in depressed regions; even by 1936, unemployment still varied from 32% in Wales to 4% in the South-East. The importance of labour immobility was confirmed by the studies in the volume of the labour markets of France (by Robert Salais), Italy (Gianni Toniolo and Francesco Piva), the United States (Robert Margo), and Canada (Alan Green and Mary MacKinnon).
The volume's emphasis on the study of microeconomic data also revealed that workers moved rapidly in and out of the pool of unemployed according to Mark Thomas, at two to three times the rates of the 1980s. During 1932, for example, when the unemployment rate was 22%, as many as 53% of workers experienced at least some unemployment. Part of the explanation lay in the operation of the unemployment insurance system. In addition it was simply more common in the interwar period for workers to change employers.
Despite this rapid turnover, Eichengreen noted, the burden of unemployment was borne very unevenly: unemployment rates were almost twice as high for men aged 55-59 as for those aged 18-24, and unemployment was lowest among `prime-age' workers in their late 30s and early 40s. The duration of unemployment spells also varied widely: despite rapid turnover for some workers, at least a quarter of those entering unemployment (particularly older males) could expect to remain there for more than a year.
The final issue discussed by Eichengreen was the effects of unemployment on health. Contemporary economists and health officials doubted that unemployment had adverse effects on health, pointing to improvements in aggregate mortality and infant mortality. But their scepticism is convincingly disputed in the volume by Bernard Harris, who conducted a cross-section analysis of children's height and adult male unemployment in 11 UK towns. The analysis revealed a clear negative relationship between height and unemployment. This relationship further reduced the plausibility of theories based on over-generous unemployment benefits, Eichengreen argued.
One member of the audience asked what insights could be obtained by examining the 1930s reduction in unemployment. Although the experience of the industrialized countries varied widely in the rapidity with which unemployment fell, the multicountry analysis in the chapter by Andrew Newell and J S V Symons revealed that all 14 of the countries studied had experienced high unemployment, combined with wages above market-clearing levels. Tim Hatton noted that, in contrast to the 1980s, unemployment fell quite rapidly in the 1930s, suggesting that aggregate demand policies were of some importance. There was also some evidence, Eichengreen noted, that unemployment was reduced more effectively in economies exhibiting greater real wage flexibility. The evidence on the unemployment decline is difficult to explore, however, since the usefulness of time-series data ends abruptly in 1939