with relatively low human capital endowment are only slightly better off by working than by
being unemployed. If these workers are forced to enter unemployment, their human capital
degrades. If this degradation is pronounced enough and if human capital investments during
unemployment are costly enough, their human capital erodes to the point where it does not pay
to work again at the time when re-employment is possible.
5. Conclusion
This paper has offered a micro-economic tool for the analysis of short- and long-term unem-
ployment. Still simple enough to be solved analytically and discussed diagrammatically, the
model has highlighted the importance of labor market institutions (generosity and duration
of unemployment benefits, welfare assistance, taxes, and retraining costs) and how they inter-
act with personal characteristics (age, retraining costs, rate of human capital degradation) and
stylized exogenous labor market fluctuations (labor market tightness) to explain unemployment
duration and retraining efforts. Our model provides a theoretical foundation for recent empirical
studies who find institutional variables at the heart of the unemployment problem.
Despite our mostly intuitive results which are in line with much empirical evidence for several
reasons we would like to caution the reader to jump to policy conclusions. First, any quantitative
policy assessment would require to embed the model in a macroeconomic context and integrate
it into a general equilibrium model of job search and matching. While this would possibly
destroy the general solvability and beauty of the model, we expect little value added. This is
because the micro-economic model already suggests that “it depends”. Therefore, the second
and perhaps more important reason for caution is that the comparative statics have shown that
the individual characteristics of the unemployed (education, skill loss, and age) determine the
impact of institutional change. This holds in all cases with respect to the magnitude of effects
and sometimes even with respect to their sign. As a consequence, the heterogeneity of the
workforce must be taken into account when analyzing the implications of policy interventions.
Nevertheless, we believe some general conclusions follow from our model. We identify skill
degradation as an important determinant of long-term unemployment as in Ljungqvist and Sar-
gent (1998). In contrast to these authors, we do not find the problem to be aggravated by
unemployment benefits. Higher unemployment benefits, while increasing short-term unemploy-
ment, can actually be favorable in the long-run because individuals may use a spell of short-term
21