The sample mean of isb is .988.Wefind that the figure is slightly
higher for employers than the others (1.057 for employ =1and .978 for the
rest). However, there is little difference between the means for those who
employed only themselves and the rest (1.024 and .985, respectively).17 We
find that those who employed at least one person other than themselves were
on average in sectors with a high proportion of foreign workers (1.107 for
those who employed at least one person other than themselves and .982 for
the rest).
Other explanatory variables include purely exogenous variables such as a
respondent’s gender (female) and approximate age in years at the time of
interview. We also have an indicator of whether or not at least one parent
of a respondent was born abroad (fparent); the respondent belonged to an
ethnic minority in the country where the interview took place (ethnic); and
he or she was a citizen of that country (citizen).
A respondent’s closeness to immigrants is approximated by the number
of immigrant friends she or he had (friend dummies). In addition, we use
a measure of a respondent’s exposure to the media on current affairs and
politics (media) in hours per weekday.18 The media is often thought to in-
fluence one’s view, and we want to control for such potential influence. Since
the status of being unemployed has been found an important determinant
17 The 95 percent confidence intervals for these figures overlap each other.
18We created this variable by using the responses to the 3 separate questions: On an
average weekday, how much of your time is spent watching television (ESS Question A2)
/ listening to the radio (A4) / reading newspapers (A6) about politics and current affairs?
The responses to these questions were given on the same scale that has an equal interval
in hours between categories. This enables us to easily aggregate the responses at the
individual level.
13