Mikael Persson, "Social
network position mediates the effect of education on active
political party membership," Party Politics, 20
(September, 2014), 724-739. [Available at
Research on participation in political parties has shown
that active party members have generally had a substantially
higher education than average citizens (e.g. Cross and
Young, 2008; Scarrow and Gezgor, 2010; Whiteley, 2009,
2011).1 However, research in this area seldom problematizes
the possible causal mechanisms linking education with active
party membership, i.e. exactly why do the high-educated
participate in political parties to a greater extent than
the low-educated? According to the conventional view,
education increases civic skills and cognitive capabilities,
factors that in turn increase the likelihood of
participation in political parties (Lewis-Beck et al., 2008;
Verba et al., 1995). One of the major puzzles in political
behaviour research is therefore why increased education at
the aggregated level in Western countries has not resulted
in a corresponding aggregate increase in levels of political
party participation (Brody, 1978). powers to punish and
reward legislators, for example through withholding
re-selection and promising promotion or financial
- Table 1. Estimates from model 1.
- Figure 1. Illustrations of the direct and indirect
effects from models 1-5.
- Table 2. Estimates from model 2.
- Table 3. Estimates from model 3.
- Table 4. Estimates from model 4.
- Table 5. Estimates from model 5.
- Table 6. Summary of results.
(Second paragraph of Conclusion) This study moves beyond
previous research by confirming the causal path proposed by
NJS. It uses the best data available on social network
position and analyses them with a more accurate technique
than has previously been employed. However, all studies have
their limitations and this is of course no exception. In
this case, the problem is that this study is limited to one
single country, Sweden, and one form of political
participation, i.e. active political party membership.
Further research is needed in order to evaluate whether the
conclusions are generalizable to other forms of
participation in other contexts. Moreover, as always in
research drawing on cross-sectional data, there is a problem
with possible reversed causality: not only education but
also political activities are likely to affect social
network position. Panel data are needed to establish the
relative strength of these paths. Hence, this study brings
some important insights to our knowledge of how the
relationship between education and participation functions,
yet the black box remains far from totally