democracies are more satisfied with democracy. We have not found such a
relationship. Thus, proportionality seems to go together with positive perceptions of
the accountability and representativeness of democracy, but is not associated with
higher evaluations of democracy.
Our final analysis combines the micro- and macro-level explanations of
satisfaction with democracy. We confine ourselves again to observations in old,
majoratarian or proportional democracies. Table 6 reports the regression results. As
expected on the basis of the previous tables, constitutional design does not have an
impact on satisfaction with democracy, but both perceived accountability and
perceived representation do positively affect satisfaction. Their combined effect leads
to the highest levels of satisfaction.
We are now also in a position to answer the question to what extent the
relationship between type of democracy and satisfaction with democracy can be
interpreted by people’s perceptions of accountability and perception (cf. figure 1). The
answer is that it can’t. There is no direct relationship between proportional design and
satisfaction with democracy (table 5), and the introduction of the intervening variables
perceived accountability and perceived representation does not make any difference in
this respect – the nonsignificant result is basically replicated.
Table 6 Evaluations, perceptions and constitutional design (excluding new
democracies and mixed design)
Satisfaction with democracy
Proportional design –0.04 (0.02)
Perceived accountability 0.31 (0.03)
Perceived representation 1.30 (0.04)
Accountability*representation 0.13 (0.05)
n = 27,547
Pseudo R
2
= 0.06
cutting point 1 –1.64 (0.03)
cutting point 2 0.14 (0.03)
cutting point 3 3.21 (0.04)
Note: Entries are coefficients, cutting points, and associated standard errors from an ordered logistic
regression (proportional odds model); the dependent variable is the four-category “satisfaction with
democracy” measure.
Conclusion and discussion
How do the voters in different political systems perceive accountability and
representativeness, and what effects does this have for their satisfaction with
democracy? Political systems are often divided into majoritarian and consensus types,
and accountability and representativeness have a quite different importance in these
two types. Whereas the consensus type of democracy is believed to maximize the
representation function, the majoritarian type enhances the accountability function.
Satisfaction with democracy is thought to be greater in consensus type democracies,
because the representation function supposedly keeps the voters of opposition parties
relatively satisfied.
CSES Module 2 provides the first internationally comparative data needed to
put these expectations to a test. We have operationalized the accountability and
representation functions at the micro-level by two indicators from the CSES survey.
The type of democracy was, following Powell (2000) operationalized in terms of a
majoritarian, proportional or mixed constitutional design. Our findings, based on a
total of 28 elections in 27 countries during the years 2001-2005, show that people’s
satisfaction with their democracy primarily depends on their perception of the
representation function, and to a lesser degree on the accountability function.
19
Convention