difference stems from the gap between those who employed at least one person
other than themselves and the rest. While we find little difference between
the means for those who employed only themselves and the others (1.050 and
1.077, respectively),11 the figure for those who employed at least one person
other than themselves is much higher than the figure for the rest (1.534 and
1.050, respectively).12
We speculate that immigrant labor is more employable in some sectors
than others due to the nature of production. If immigrant labor were more
employable, employers would benefit from an increase in labor supply due to
immigration. In order to differentiate observations by the sectoral employ-
ability of immigrants, we create a measure that approximates the relative
extent of the sectoral employment of foreign workers in each country. By
using Eurostat Census 2001, we compute the share of foreigners in each sec-
tor’s total employment in a country, divided by the share of foreigners in that
country’s total employment. Table 2 presents the obtained figures.13 A fig-
ure greater than 1 indicates that the share of foreigners in the corresponding
sector’s employment is relatively high in the country. We assume that this
indicates a relatively high employability of immigrant labor in the sector.14
Unfortunately, the corresponding Belgian data are missing in the source.
11The 95 percent confidence intervals for these figures overlap each other.
12The figures are computed using the sample after listwise deletion, i.e., the observations
actually used for regression.
13 OECD has produced a similar table in its annual publication, International Migration
Outlook (formerly Trends in International Migration), but it does not give the figures as
we do in our table.
14 However, this assumption may not be valid in some countries. Even in a sector where
foreign workers are highly employable, we may observe a low value of isb because of, for
instance, strict national policy regarding the issuance of work permits. We discuss this
possibility further in the conclusion.
11