permanently. Furthermore, the effect is strong if taxes are already high, if welfare assistance is
generous, and if skills degrade rapidly during STU.
Yet, lower net returns of retraining also apply to workers of intermediate productivity who
previously considered an STU spell and now prefer to exit unemployment immediately. While
higher labor taxation increases LTU, it reduces STU for two reasons: an outflow into employment
of workers from the higher productivity segment and an outflow into LTU of workers from the
lower productivity segment who are permanently discouraged by higher taxes.
We now discuss how personal characteristics of the unemployed determine retraining and the
duration of unemployment. As explained below, some variables which are interpreted as reflect-
ing personal characteristics might also be interpreted as due to policy choices. For example, the
effect of the variable remaining work-life might be interpreted as a workers age or as (mandatory)
retirement age.
3.5. Human Capital Degradation. The rate at which a workers human capital erodes can
depend on his personal characteristics, i.e. on his skill level, or be a consequence of exogenous
factors like the rate at which technological innovations are made. Rapid degradation of human
capital increases LTU because retraining is less attractive for the STU, and it reduces STU
because the direct cost of degradation while unemployed are too high for high-skill workers. In
Figure 1, higher human capital degradation during unemployment moves hA to the right and
hC to the left.
∂hA = [s + λτ(1 - θ)] - e [(1 - θ)δτ - c] = hA 0 ∂hc = _ hC 0
∂d δτ (1 - θ)(1 - d)2 (1 - d) > , ∂d δτ ∩ - b)+ d<
δτ '` '
Hence, rapid human capital degradation reduces STU at both ends. Some workers from the
upper end of the education spectrum who previously thought about using a spell of STU to
upgrade their human capital, now decide against it because the opportunity costs of skill loss
while not working are too high. At the lower end of the educational spectrum some workers who
previously found a short spell of unemployment attractive, now see their human capital (and
potential future wages) erode so fast that staying permanently out of work becomes attractive.
It is worthwhile to point out some similarities and differences of effects of human capital
degradation in our model and the model by Ljungqvist and Sargent’s (1998). The main goal
of their analysis is to explain the puzzle how labor market institutions can account for high
13