AbstractBACKGROUNDGermany, like many OECD countries, faces a shift in the age composition of its population, and will face an even more drastic demographic change in the years ahead. From a theoretical point of view, decreasing cohort sizes may on the one hand reduce unemployment due to inverse cohort crowding or, on the other hand, increase unemployment if companies reduce jobs disproportionately. Consequently, the actual effect of cohort shrinkage on employment and unemployment is an empirical question.OBJECTIVEWe quantitatively assess the relationship between (un)employment and cohort sizes.METHODSWe analyze a long panel of population and labor-market data for Western German labor market regions. We isolate the direct, statistical effect of aging in a decomposition approach and estimate the overall effect by regression. In this context, we account for the likely endogeneity of cohort size due to migration of the young workforce across regions using lagged births as instrument.RESULTSThe direct effect of the age composition of the labor force on unemployment is negligible. In contrast, the elasticities of unemployment and employment with regard to the labor-market entry cohort's size are significantly positive or negative, respectively. The causal effect indicates an over-elastic reaction by unemployment.CONCLUSIONSOur results provide good news for the Western German labor market: small entry cohorts are indeed likely to decrease the overall unemployment rate and thus to improve the situation of job seekers. Accordingly, we find the employment rate is positively affected by a decrease in the youth proportion.(ProQuest: ... denotes formulae omitted.)1 IntroductionIn this paper we analyze the consequences of labor market entry cohort size on employ- ment and unemployment in Western Germany. Germany experienced a sharp decline in birth rates in the second half of the 1960s when the so-called boom was followed by a bust. As a result, the youth proportion, i.e. the population aged 15-24, di- vided by the population aged 15-64, has dropped significantly 15 years later. In the late 1970s, this youth share was around 23% in Germany but 26% in the U.S. and even 32% in Canada (OECD average: 24%). Due to an increasing birth deficit and the aging of the baby boomer, Germany is today - together with Japan - among those industrialized coun- tries that display the lowest youth proportion among OECD member states (Germany: 17%, OECD average: 19%). Paralleling demographic development, unemployment rates have been decreasing, whereas participation rates have been increasing over the last few years in Germany.Based on these stylized facts, we analyze whether or not the size of the labor mar- ket entry cohort has affected unemployment and employment rates in Western Germany. From a theoretical perspective, the effects of cohort size changes on the labor market are unclear, as the juxtaposition of the cohort crowding hypothesis and the hypothesis of Shimer (2001) shows. With regard to the cohort crowding hypothesis, Easterlin (1961) argues that workers do not perform as well in the labor market if they belong to bigger co- horts. The reason is that labor demand is supposed to react less flexibly than labor supply. In contrast to this rationale, Shimer (2001) points out that companies have an incentive to create more jobs in regions with large youth entry cohorts, which eventually overcompen- sates the increase in the labor supply. In addition to these theoretical discrepancies, even empirical investigations provide ambiguous results (e.g. Korenman and Neumark 2000; Skans 2005; Foote 2007). It is at this point that our analysis adds to the existing literature. We study the impact of shrinking cohort sizes on (un)employment over a long time span, i.e. from 1978 to 2009, consider functional regional labor markets as an appropriate unit of analysis and apply various econometric robustness checks. …