19
compare the late 1980s with the late 1990s. This number should be taken with a due grain of salt,
however, since it is bound to be sensitive to how one measures the GDP gap.
A further perspective on the adjustment of employment is obtained by examining the
simulated flows. We focus here on flows involving employment and unemployment. Figure 4.2
reveals that permanent and temporary workers contribute by roughly similar (absolute) numbers
to the flow into unemployment. Panel B shows the effects of the recession on the outflows from
unemployment to employment. The flow from unemployment into both types of employment
rises during the recession, with the rise into temporary work being particularly strong.
The marked positive correlations between unemployment inflows and outflows illustrated in
Figure 4.2 are similar to what have been found in data for other countries; see Blanchard and
Diamond (1990) for the US, and Burda and Wyplosz (1994) for some European countries. As
revealed by ocular inspections of transition rates (Figure 3.5) and the regressions in Table 4.1, a
recession causes a decline in transition rates from unemployment to employment and an increase
in transition rates from employment to unemployment. However, as the pool of unemployed
builds up, more workers become available to match with job vacancies. This “matching effect”
dominates during most of the recession and the level of the flows from unemployment to
employment is therefore higher in a recession than in a boom.
This concludes our exposition of how adverse business cycle conditions impact on temporary
employment. It is time to take stock of alternative explanations of why Sweden has experienced
a remarkable rise in temporary work during the 1990s.