Ethnic Hostility among Ethnic Majority and Minority Groups

7. Educational Attainment and Ethnic Hostility | Results
however that Dutch media usage does not affect opposition to mixed relationships for Antilleans
(of both fi rst and second generation), for second generation Moroccans and only reaches the
boundary of signifi cance for second generation Surinamese. But all in all we fi nd more corroborative
evidence that, at least for fi rst generation ethnic minority members, Dutch media usage is an
indicator of involvement with the host country and as such is negatively related to ethnic hostility
than that the Dutch media is an important source of negative feelings towards natives. Somewhat
surprisingly, Dutch media usage is also negatively related to opposition to mixed relationships
among native Dutch b=-0.08 (se=0.04; Table 7.3).
Negative contact experiences are, as predicted (hypothesis 10), positively related to
opposition to mixed relationships for fi rst generation Turks (b=0.10, se=0.05), fi rst generation
Surinamese (b=0.12, se=0.04), second generation Moroccans (b=0.19, se=0.10), and second
generation Antilleans (b=0.09, se=0.05). The parameter estimates are positive for other ethnic
categories as well (with the exception of second generation Turks) and even reach the boundary
of signifi cance for second generation Antilleans (b=0.08, se=0.05; Table 7.3).
The total effect of education is signifi cantly interpreted through the proposed mediators
for native Dutch and for all ethnic groups of fi rst generational status; the 90 per cent biased
corrected confi dence interval lies below zero for all ethnic groups (Table 7.3). Among minorities of
second generation, we observed a signifi cant interpretation of the education effect for Surinamese
and Antilleans but not for second generation Turks and Moroccans. Contact with outgroup members
in leisure time most consistently signifi cantly interprets the effect of education, but not for
second generation Turks and Moroccans. This is not because positive contact does not reduce
ethnic hostility among these latter groups, but because higher educated Turks and Moroccans do
not have (signifi cantly) more positive contact with natives in leisure time than lower educated
Turks and Moroccans (Appendix 7.1).
Dutch media usage explains in part the effect of education on opposition to ethnically
mixed relationships among ethnic minorities. There is one exception: for second generation
Moroccans, Dutch media usage does not interpret the education effect. Origin country media usage
was consistently positively related to more opposition to ethnically mixed relationships, however,
we do not observe an indirect effect of education through origin country media consumption.
Higher educated ethnic minority members do not use less media from their country of origin than
lower educated ethnic minorities (Appendix 7.1). We tentatively conclude that involvement with
the country of origin is less easily altered by policy measures directed at the structural integration
(e.g. educational programs) than involvement with the host country.
We fi nd far less signifi cant mediation effects for second generation minorities than
for their fi rst generation counterparts. The variation across generations is due to differences
in the direct effects of educational attainment on the mediator variables (Appendix 7.1). The
positive effect of educational attainment on Dutch media usage, and contact in leisure time, and
the negative effect of education on personal experienced discrimination, is weaker for second
generation Turks and Moroccans than for their fi rst generation counterparts (Appendix 7.1).
157