what the main factors of party choice were in the visegrad countries
Project description
Based on uploaded article, please discuss in the reflection sheet WHAT THE MAIN FACTORS OF PARTY CHOICE WERE IN THE VISEGRAD COUNTRIES.
1
Gábor Tóka and Andrija Henjak:
Party Systems and Voting Behaviour in the Vise
grad Countries 15 Years After the Transition
Published in:
Visegrad Votes: Parliamentary Elections 2005-2006
, ed. by Pavel Ã…Â aradÃÂn and Eva
Bradová. Olomouc: Palacky University Press, pp. 210-244.
It is probably fair to say that in the course of the 1990s the emphasis in the international scholarly
literature on party systems and electoral alignments in Eastern and Central Europe in general, and
the Visegrad countries in particular, shifted far away from the initial emphasis on the inherent
instability of post-authoritarian contexts repres
ented by works like à gh (1998), Cotta (1994), Rose
and Mishler (1995). Instead, the main theme in the
literature of the late 1990s was the remarkable
ability of some or even most of these party systems to offer intelligible programmatic alternatives
and party-voter linkages, and the need to explain a surprisingly large cross-national variation across
the post-communist landscape in any aspect of the political and policy changes that the observers
looked at (Grzymala-Busse 2002; Kitschelt
et al.
1999; Mateju and Reháková 1996; Mateju and
Vlachova 1997; Miller
et al.
1998; Tóka 1996b, 1997b; Tworzecki 2002). Yet this shift was
probably as much a reflection of intellectual fashion as that of actual political developments in the
region.
1
It is probably not so surprising, then, that in recent years the pendulum is again swinging in
the evaluative remarks about the outcome of post
-communist political transitions. This time, the
new trend appears to be an accentuation of how unde
rdeveloped is the soft, social infrastructure of
democracy basically everywhere east of Germany and Austria, in spite of the apparent consolidation
of democracy and market economy west of the Be
lorussian-Ukrainian border. A new generation of
assessments is now pointing out that strongly pa
rtitocratic regimes and weak civil societies,
generating relative little
de facto
citizen participation and popular control of the policy making
process emerged in the Eastern annex of the European Union (Adam
et al.
2005; Biezen 2003;
Howard 2003; Markowski 2005; Rose-Ackermann 2005).
In the present paper, we aim at a fresh assessment of party systems and electoral alignments
in the Visegrad countries, and at giving justice to all three assessments by specifying in which
respect each of them holds true. First we sha
ll look at some basic parameters of party system
development, then we turn our attention to voting behaviour, and finally to cleavage structures.
Limits of space prevent us to reiterate basic fact
s about elections, parties, and political events in
individual countries. It is not our choice but the (un)
availability of cross-nationally comparable data
1
In fact, the passing of time probably obfuscated rather than clarified programmatic alternatives in some of
these party systems once the initial big issues of post-communist transition like the regaining of national independence,
2
that make us pay a great deal of attention to somewhat dated evidence from the last European
Values Study in 1999/2000 and the Comparative Study of Electoral Systems surveys in the 1996-
2004 period, and that Slovakia is not as extensivel
y covered in our tables as the Czech Republic,
Hungary and Poland.
2
Partly to compensate for these hiatuses, we tried to make the paper
empirically richer by including a wide range of comparative referents from other East European
countries and occasionally even from outside of Europe.
Party systems
Citizens’ voting behaviour cannot stop reflecting or even mirroring the choice set that they
face, i.e. the supply side of the electoral market. Not surprisingly, then, since the transition to
democracy electoral politics in the countries of East Central Europe has been in a state of permanent
flux (Kostelecky 1999). Parties were emerging and vanishing, merging and splitting – though
remarkably less frequently in the South than in the North. Hungary (and beyond that Slovenia,
Croatia, Romania, Bulgaria, Serbia, Bosnia-Herzegovina, Albania and Macedonia) witnessed far
fewer changes in the identity of the major parties than Poland (and beyond that the Baltic States). In
both North and South, however, the electoral alignments of the mass public were inevitably shifting
with the changes in the party system, the crystallization of political alternatives and the ideological
shifts of parties. Obviously, after half a century
of communist dictatorship, the translation of social
divisions into political alignments was bound to take
some time – conceivably much more than after
the comparatively short-lived and socially less transformative fascist regimes in post-war Italy or
Germany (Cotta 1994). Consolidation was further delaye
d by rapid changes in the issues of political
contestation, which can also be related to the te
ndency of communist rule – and consequently of
post-communist transitions – to impact deeply on every sphere of life.
Following previous studies of Southern Europe and Latin America, we can benchmark the
stabilization of post-democratization East Central European party systems by looking at the
fragmentation of the party system, the inter-election changes (volatility) in individual parties’ share
of the vote, and – as a logical corollary of the latter – the percentage of partisan identifiers in the
electorate (Mainwaring and Scully 1995; Morlino 1995). In Table 1, party system fractionalization
is measured with the effective number of electoral parties (
ENEP
) in two subsequent elections
around 2000. If
n
is the number of parties winning any votes, and
v
i
is the fraction of votes obtained
by the
i
th
party, then
2
1
1
n
i
i
E
NEP v
=
=
?
. Thus,
ENEP
conveniently equals
x
when there are
x
parties
competing in an election and each wins an equal share of the vote, but falls far below
x
if at least
large scale privatization, and market opening were settled for good.
2
Regarding our data sources and other technicalities about our computations the reader is referred to the notes
adjoining the tables.
3
one of these parties wins a much larger share of the vote than the others. Aggregate electoral
volatility (
AV
) measures the stability of the party system with a similarly straightforward formula,
namely
1
1
2
n
it it
i
vv
AV
–
=
–
=
?
, where
v
it
and
1
it
v
–
are the percentage of votes obtained by the
i
th
party at an
election held at time
t
and in the previous election, respectively. The formula sums up the party by
party differences between two subsequent election
outcomes and divides the result by two because
the percentage gains of the vote-winning (or new)
parties necessarily equal the combined losses of
all other parties together. While Table 2 provides more detailed data on both indicators for the
Visegrad countries across the entire period since 1990, Table 1 focuses only on the average of the
two elections closest in time to the dates when the survey data summarized in the column about
partisanship in the electorate were collected.
A further technical note is due before the analysis. Following previous findings in the
comparative literature, one would expect that party system fractionalization increases volatility and
reduces partisanship in the electorate, while the
latter are strongly and negatively correlated with
each other (Bartolini and Mair 1990; Schmitt and Holmberg 1995). These expectations are only met
in these data if the suspiciously high percentage of party identifiers in the Russian and Ukrainian
survey data – in fact, and quite counter-intuitively
, higher than what turnout in legislative elections
was in these countries at the time – are discarded in the analysis.
3
Therefore this is exactly what we
will actually do below in the belief that translati
on problems or other factors made the survey data
from these two countries incomparable to those
obtained elsewhere with ostensibly the same
question. In other words, Russia and the Ukraine are excluded from the calculation of the East
European regional average reported at the bottom of the last column.
Table 1 about here
Since the table is otherwise unwieldy, some regional averages are presented in the bottom
rows. One interesting finding is that around the turn
of the millennium the four Visegrad countries
were not really different from the
West European average in terms of the number of parties and the
partisanship of the electorate, but even then, at least Poland and Slovakia were dramatically
different from Western Europe in terms of volatilit
y. Yet, as the regional averages and previous
studies make it plainly clear, by then Western Eu
rope was significantly influenced by a new wave
3
Of the mentioned correlations only the one between
ENEP
and
AV
(R=.475, p=.007) is significant across all
data points shown in Table 1. The correlations between the first two indicators one the one hand and the incidence
partisanship on the other substantially increase (from -.03 to -.35 between
ENEP
and partisanship, and from -.31 to -.69
between
AV
and partisanship) and reach statistical significance when the Russian and Ukrainian survey data are
ignored.
4
of relatively volatile election results, increased
party system fragmentation and a strong erosion of
partisan attachments and traditional party alignments in the electorate (cf. Franklin
et al.
1992; Mair
2002; Schmitt and Holmberg 1995). This new upheaval can be attributed both to recent socio-
economic changes in affluent postindustrial societies and to an aging of their party systems. Hence
a comparison with Southern and Eastern Europe may be more interesting, since these regions are
still largely unaffected by these newer developments. In this comparison, the Visegrad countries
turn out to be far more similar to the new democracies of Eastern than to the not-so-old democracies
of Southern Europe. The latter are, in fact, ma
rked by higher levels of partisanship and lower
volatility and party system fragmentation than th
e older and more affluent democracies of North-
Western Europe and Italy, and thus have far more
consolidated party systems than the Visegrad
countries. It is indeed striking how much higher electoral volatility still is in the latter than it was in
Greece, Italy, Portugal and Spain –
or post-fascist Germany for that ma
tter – after the first four-five
rounds of democratic elections in the pos
twar period (cf. Mair 1996; Morlino 1995).
Table 2 about here
Table 2 shows a more nuanced picture regarding party system development in the Visegrad
countries. Here we further distinguish between ‘raw
’ and ‘adjusted’ electoral volatility. The first
shows a naïve calculus that counts as relevant electoral change even those instances where a party
changed its name or formal organizational structure (thus turning, for example, from OH in 1992 to
SD-LSNS in 1996 and then to CSNS in 1998), to recall a familiar example from recent Czech
history. Adjusted volatility, in its turn, disregards
every numerical change in vote distributions that,
mathematically speaking, could be explained away as a mere artifact due to changes in party names,
electoral alliances, party mergers and party splits. For instance, the main right-wing party in
Hungary, commonly referred to as Fidesz, run a joint list with the Hungarian Democratic Forum
(MDF) in 2002, but then changed its official na
me (from Fidesz-Hungarian Civic Party to Fidesz-
Hungarian Civic Movement) and run a joint list with the KDNP – which was electorally non-
existent in 2002 -, rather than the MDF in the 20
06 election. The ‘raw’ estimate treats the 2006 vote
(47 percent of the total) for the Fidesz-KDNP joint list and the MDF as votes for two entirely new
parties, while the ‘adjusted’ count treats them as the vote for the same party that, under the guise of
the Fidesz-MDF joint list, received 41 percent of the vote in 2002. In practice, judgments about
party continuity inevitably involve drawing a more
or less arbitrary threshold between ‘continuity’
and ‘discontinuity’, and hence neither of the two figures is perfect.
4
That is why we present here
4
For instance, in the present analysis we treated the Freedom Union (US) as one successor – alongside with
ODS itself – to what the ODS vote represented in the 1996 Czech elections. However, we decided to consider the ODA
