Country 1, Unemployment
20
15
10
5
0.8
0
45
30
τ (%)
Figure 8: Country 1 labor market regulation and unemployment in countries 1 and 2
(=3), when ν = 0 and firms are homogeneous. [Rate of unemployment on the vertical
axis.]
Country 2, Unemployment
9.2
9
8.8
8.6
8.4
8.2
8
0.8
60
0.4 0
τ (%)
unemployment. Hence, one part of the competitiveness channel survives, even in the
homogeneous firms model without external economies of scale.
3.6 The role of real wage rigidity
So far, our numerical exercise suggests that the effect of domestic institutional change on
the domestic rate of unemployment is by about two orders of magnitudes larger than the
effect on the foreign rate of unemployment. The calibration exercise is extremely stylized
and allows countries to differ only across a small set of parameters. Yet, the quantitative
findings are very robust to different parameterizations of the model. The small magnitued
of spill-over effects clearly questions their political relevance. Moreover, in section 4, we
will measure the magnitude of spill-over effects in an empirical panel analysis of 20 rich
OECD countries. The regressions suggest a larger and more important role for spill-over
effects.
To address these concerns, this subsection shows that the lack of strong spill-over ef-
fects in our theoretical benchmark model is related to the Shimer-puzzle (Shimer, 2005),
which posits that the conventional search-and-matching model of the (closed) US econ-
omy cannot reproduce the elasticity of labor market tightness with respect to produc-
tivity shocks along the business cycle. Our model concentrates on spatial dependence
across countries rather than on time-series evidence. However, the problem is similar.
The change in country 1’s labor market institutions affects the other countries’ rate of
unemployment through aggregate productivity in those countries. If the link between
productivity and unemployment is generally weak, then spill-overs have to be small.
Therefore, we follow Shimer (2004) and contrast the analysis of earlier subsections,
29