countries in Europe.
[Figure 1 about here]
ESS allows us to distinguish between employers by the number of em-
ployees. Figure 2 shows the distributions of respondents who employed only
themselves and those who did not employ themselves, respectively, over the
4 ordered categories on which our dependent variable is based. It suggests
that those who employed only themselves were even more likely to prefer im-
migration restriction than the rest, compared to Figure 1. However, Figure
3 indicates that there was not much difference between those who employed
at least one person other than themselves and the rest.
[Figures 2 and 3 about here]
These pictures motivate regression analysis to control factors affecting
individual attitudes other than the status of being an employer. For ex-
ample, Table 1 shows that more of those with employ =1 are distributed
to low levels of education, compared to the distribution for the whole sam-
ple. The same table also indicates that those who employed only themselves
are responsible for this observation. On the other hand, less of those who
employed at least one person other than themselves seem to be distributed
to low levels of education, compared to the distribution for the whole sam-
ple. Hence we create 4 dummy variables (ed0 to ed3) to control for each