Unemployment in an Interdependent World

a good proxy for overall openness to trade, as well. Not surprisingly, it correlates well
with conventional openness measures. It should, however, be less prone to endogeneity
concerns.

The estimates presented in column (2) imply that a one standard-deviation increase of
bit and of b*t leads to an increase in the domestic unemployment rate by 0.318 and 0.065
standard deviations, respectively.⁴⁰ Hence, on average, the domestic tax wedge is about 5
times as important quantitatively than the foreign one. This seems like a sensible result
which may, however, hide potentially large differences across countries.

Column (3) adds additional labor market institutions but drops PMR. This has an
important effect on the coefficients of both b and b*, where the former grows substantially
larger and the latter looses statistical significance. However, as shown in column (4), the
inclusion of PMR restores the picture that we have already found in column (2). One
reason for the importance of PMR may lie in the fact that the construction of the weights
ωijt relies on exogenous geographical and demographic data only; it does not reflect trade
regulations that may curb the amount of trade between nations. Including PMR mitigates
this problem.

Columns (5) and (6) add foreign labor market institutions to the regression. Not
surprisingly, adding variables for which the direct effect on home unemployment is already
dubious (union density or EPL), does not improve accuracy of estimation. The coefficients
on b* are insignificant and seem unplausibly large; not to speak of the coefficient on EPL*,
to name only the most striking case. Hence, the lack of a robust relationship between
these variables in standard equations such as (28) also impairs inference when using their
spatial lags.

Finally, columns (7) and (8) replicate our preferred specifications (2) and (4) with
the exception that they now also include the Heckscher-Ohlin relative capital-abundance
k^*/k and its interaction with the foreign wedge b^* . Theoretical considerations suggest that
a higher foreign wedge should lower the domestic rate of unemployment if the foreign
economy is on average more capital-rich than the domestic. Our results suggest that the
opposite holds: the more capital-rich the foreign country is, the stronger is the adverse
effect of foreign distortions on domestic unemployment. As in Table 1, we do not find
evidence in favor of the Heckscher-Ohlin view. Instead we find a strong negative direct
effect of k^*/k on domestic unemployment as in Table 1.

4.5 The role of real wage rigidity, country size and entry regu-
lation

Table 3 sheds additional light on the channels through which foreign institutions affect
domestic unemployment. First, we ask whether foreign labor market institutions have a
larger effect on domestic unemployment when domestic labor market institutions imply
a high degree of real wage rigidity. There is no clear consensus how to measure real wage
rigidity in a single indicator. Rather, the degree of real wage flexibility is a complicated

⁴⁰0.075 × 18.21/4.29 = 0.318 and 0.012 × 23.39/4.29 = 0.065.

39

<<   38   39   40   41   42   43   44   >>

More intriguing information