Transitional Winners and Losers: Attitudes toward EU Membership in Post-Communist
Countries
Author(s): Joshua A. Tucker, Alexander C. Pacek and Adam J. Berinsky
Source: American Journal of Political Science, Vol. 46, No. 3 (Jul., 2002), pp. 557-571
Published by: Midwest Political Science Association
Stable URL: https://www.jstor.org/stable/3088399
Accessed: 08-07-2019 14:00 UTC
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
Midwest Political Science Association is collaborating with JSTOR to digitize, preserve and
extend access to American Journal of Political Science
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
Transitional Winners and Losers:
Attitudes Toward EU Membership
in Post-Communist Countries
Joshua A. TilCker Princeton University
Alexander C. Pacek Texas A&M University
Adam J. Berinsky Princeton University
We present a model of citizen sup?
port for EU membership designed
expllcitly for post-communist coun?
tries. We posit that membership in the
EU can function as an implicit guarantee
that the economic reforms
undertaken since the end of communism
will not be reversed. On this
basis, we predict that "winners" who
have benefited from the transition are
more likely to support EU member?
ship for their country than "bsers"
who have been hurt by the transition.
We also predict that supporters of
the free market will be more likely to
support EU membership than those
who oppose the free market, We test
these propositions using survey data
from ten post-communist countries
that have applied for membership in
the EU and find strong support for
our hypotheses. The article concludes
by speculating about the role
attitudes towards EU membership
may play in the development of parti?
san preferences.
557
What accounts for post-communist citizens' support for mem?
bership in the European Union? Students of both the Euro?
pean Union (EU) and post-communist politics have examined
which factors are likely to influence public opinion on the issue of EU enlargement
(Tverdova and Anderson 2000; Cichowski 2000). These initial
studies are based largely on the wealth of theories and findings from decades
of research on West European citizen attitudes. While the studies parallel
each other in many ways, their findings are varied.
We present a model of support for EU membership designed explicitly
for post-communist countries and test it using survey data from ten
East-European applicant nations. We suggest that, for post-communist
citizens, membership in the EU can function as an implicit guarantee that
the economic reforms undertaken since the end of communism will not
be reversed. We then argue that the unprecedented economic transition
across the former communist states has divided societies into "winners"
who have benefited from the transition and "losers" who have been hurt
by the transition. We predict that winners are more likely to support EU
membership, while losers are more likely to oppose it. As the framework
suggests that EU membership is the ultimate guarantee of a free market
economy in post-communist countries, we also expect that that those who
support the free market will be more likely to support EU membership,
Joshua A. Tucker is Assistant Professor of Politics and International AfFairs, Princeton
University, 322 Bendheim Hall, Princeton, NJ 08544 (jtucker@princeton.edu).
Alexander C. Pacek is Associate Professor of Political Science, Texas A&M University,
College Station, Texas 77843 (e339ap@polisci.tamu.edu). Adam J. Berinsky is Assistant
Professor of Politics, Princeton University, 41 Corwin Hall, Princeton, NJ 08544
(berinsky@princeton.edu).
We thank Guy Whitten, Randy Stevenson, Christopher Anderson, Jan Leighley, and
Benjamin Radcliff for useful comments and support and Sara De Master and James
McGhee for research assistance. We also wish to thank the anonymous reviewers for
their helpful comments and suggestions. All data analysis was conducted using Stata 6.0
and Clarify 1.3. A previous version of this article was presented at the 2001 Annual Conference
ofthe Midwest Political Science Association. Authors' names are listed in reverse
alphabetical order.
American Journal of Political Science, Vol. 46, No. 3, July 2002, Pp. 557-571
?2002 by the Midwest Political Science Association ISSN 0092-5853
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
558 JOSHUA A. TUCKBR, ALBXANDER C. PACEK, AND ADAM J. BERINSKY
while those who oppose the free market will oppose EU
membership.1
We begin the article with a brief review of work on
determinants of EU support in both advanced industrial
democracies and transitional states, followed by a systematic
explication of our model, including our assumptions
and hypotheses. We then present empirical tests of
our hypotheses utilizing survey data from ten East European
countries with methods of analysis that are appropriate
to the data. We conclude with implications of our
findings for future research.
Determinants of EU Support
The previous decade has produced a burgeoning field of
research on the relationship between societal characteris?
tics, mass attitudes, and European Union integration in
the advanced industrial world (for a range of reviews, see
Gabel 2000; Gabel and Whitten 1997; Anderson and
Kaltenthaler 1996; van der Eijk and Franklin 1996; Gabel
and Palmer 1995; Eichenberg and Dalton 1993). One line
of reasoning emphasizes the salience of utilitarian fac?
tors. Simply put, citizens assess the various costs and
benefits associated with membership in the European
Union. This argument is grounded largely in economic
terms; EU membership represents a specific international
economic policy?the liberalizing of the movement
of goods, labor, and services among EU member
states (Gabel 1998a, 937). A wealth of studies demon?
strates that perceptions of favorable national economic
conditions and personal financial conditions are positively
associated with EU support (Palmer and Whitten
1999; Gabel 1998a; Gabel and Whitten 1997; Anderson
and Kaltenthaler 1996; Anderson and Reichart 1996;
Gabel and Palmer 1995; Eichenberg and Dalton 1993).
However, the costs associated with membership in
the EU are not borne equally among groups in particu?
lar societies. In the short run, both workers and owners
in specific industries bear the burden of adjustment to
lNote the important distinction between this approach and the
more traditional "utilitarian" approach used in established democ?
racies. We claim that citizens choose to support EU membership
not on the basis of how joining the EU will change their economic
position in the future as the utilitarian approach posits, but rather
because of how their economic position has changed since the
transition from communism. We maintain that the utilitarian ap?
proach makes unrealistic assumptions about both the stability of
post-communist societies and citizens* understanding of the realities
of EU membership. Instead, we posit merely that citizens have
a sense of whether they are doing well in this new world and
whether they would like the new free market economy to continue
in the future.
capital mobility during integration. Opportunities exist
for those skilled enough to compete with other workers
in their occupation throughout the market, while unskilled
workers will be at a comparative disadvantage
(Gabel 1998a, 1998b). Not surprisingly, cognitive skills,
education, and occupation type have all been found to
affect support for EU integration (Gabel 1998a; Ander?
son and Reichart 1996; Gabel and Palmer 1995; Janssen
1991).
A second line of argument focuses on the political
values of citizens and their impact on EU support. The
issue of EU membership as a means of strengthening
democratic institutions was raised during the 2nd Enlargement
in Greece, Spain, and Portugal (Wallace 1990;
Duchene 1982; Tsoukalis 1981). Scholars have further argued
that political parties might affect attitudes toward
European integration (Anderson 1998; Taggert 1998), or
perceptions of the incumbent governments (Hug and
Sciarini 2000). In short, citizens rely on attitudes toward
familiar political objects to make sense of a complex
transnational issue (Anderson 1998).
Both these arguments, developed in the context of
Western European countries, have driven contemporary
inquiries into determinants of EU support in the transi?
tional post-communist world. Initial studies find some
support for both arguments, but findings are mixed
across studies and indicators utilized. Tverdova and
Anderson's study of six post-communist countries (Bulgaria,
Czech Republic, Estonia, Hungary, Latvia, and
Slovakia) found strong relationships between a range of
utilitarian indicators and EU support, with the exception
of measures of education (2000, 16). By contrast,
their results were much less clear for indicators of gov?
ernment support and satisfaction with democracy
(2000,17). Cichowski reported somewhat different find?
ings in her study of five transitional countries (Czech
Republic, Estonia, Hungary, Poland, and Slovenia). Specifically,
she finds that those citizens who are satisfied
with democracy, favor the free market, and take cues
from political parties are more likely to favor EU mem?
bership (2000, 1270). Economic perceptions appear to
have little impact (Cichowski 2000,1267).
What are we to make of this? While the small num?
ber of initial studies precludes generalization, the picture
of what exactly determines EU support among transi?
tional citizens is far from clear. While the arguments
based on the Western experience with integration are
valuable, they are but one piece of the puzzle. In the fol?
lowing section, we outline a new model of the determi?
nants of support for EU membership in transition coun?
tries that is specifically designed to account for the
realities of the post-communist world.
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
TRANSITIONAL WINNERS AND LOSERS 559
Support for EU Membership
in Post-Communist Countries
We argue that the conditions produced by the ec
transition in post-Communist Europe in large
termine attitudes toward EU support among
We begin with three assumptions. First, EU mem
might carry with it quite different connotations
zens of post-communist countries than it does fo
zens of western democracies. Cichowski notes that the
peculiarities of the economic transition mean that in
principle, East Europeans interpret EU cues through the
reality of how the transition has affected them (2000,
1248). Joining the EU represents not only a set of specific
economic guarantees, but a guarantee that the economic
reforms will continue as well. As Cichowski notes, "Euro?
pean integration stands as a further institutionalization
of free-market reforms, a prospect not necessarily welcomed
by these individuals" (2000,1248). Such concerns
about the impact of EU membership on the reform pro?
cess have been observed across post-communist societies
(Kirka 2000; Tuszynski 1997).
Our second assumption concerns the division pf
post-communist societies into winners and losers. While
there is a rough scholarly consensus that this division
exists, there is disagreement and differing emphasis on
precisely what it constitutes (Bell and Smeltz 2000).
In the wake of free market reforms and economic dislocation,
scholars have argued that these divisions are
based on skill levels (Mackie 1995), cognitive abilities,
and overall living standard levels (Plasser, Ulram, and
Waldrauch 1998). Attempts to identify winners and los?
ers have often involved ascribing status as a winner or
loser to certain demographic groups: elderly pensioners,
state workers, and women are often categorized as los?
ers, while winners tend to include the young, the better
educated, and those employed in the private sector.
However, in the absence of information clearly denoting
the citizens who have benefited or lost from the transi?
tion experience, ascribing winner status on the basis of
demographic characteristics will remain at best one step
removed from individual perceptions and at worst an
imprecise and inaccurate system of extrapolation. To
avoid these problems, we rely directly upon individuals'
self-assessments of their economic progress during the
transition. Therefore we assume that the population can
be placed along a continuum of winners and losers and
that this continuum can best be measured by individual
self-assessment.
Our third assumption is that attitudes towards EU
membership are not a result of one's political party preference
in transitional democracies. While a number of
studies treat party support as a measure of EU attitudes
in the West European context (Anderson 1998; Taggart
1998), we question the appropriateness of such a technique
in the post-communist context. Recent studies
have produced evidence of considerable distance be?
tween party elite and supporter positions on a host of reform-oriented
issues, including European integration
(Kitschelt et al. 1999). Moreover, we find it theoretically
implausible that on such a fundamental concern as
membership in the EU, voters would be likely to take
their cue from such new political parties. This dynamic is
due both to the presence of so many new parties and the
constant fluctuations between parties being in power and
being marginalized.2
Our predictions can therefore be concisely stated.
First, winners who have done well by the transition are
likely to support EU membership as it ensures a continuation
ofthe transition process from which they have benefited;
losers who have been hurt by the transition will
oppose this step. Second, respondents who support the
free market will be more likely to support EU member?
ship. Third, these effects will be present even after controlling
for demographic effects; in other words, the relationships
we predict are not mere proxies for underlying
demographic effects. In addition, we consider the possibility
that citizens' orientation towards the EU may have
an effect on their political party preferences.3
Description of Data and Variables
For our analysis, we rely upon data from the 1996 Cen?
tral and Eastern European Eurobarometer (CEEB) sur?
vey (Cunningham 1996).4 The CEEB data set is not without
its drawbacks?most notably it only asks a small
number of questions?but it remains attractive due to its
cross-national coverage. The 1996 CEEB survey included
2While it would be desirable to test this assumption quantitatively,
the data available for this project were not sufficiently rich to allow
for the necessary models. Therefore, the direction of causality be?
tween political party preference and preference over EU member?
ship remains a theoretical question (see below for further discussion).
Moreover, we are suspicious of any attempt to use the same
survey data to specify the positions of political parties (i.e., hy?
pothesis building) and assess the effects of those positions (i.e., hy?
pothesis testing).
3We also considered testing the effect of citizens* proclivity towards
democracy as a potential explanatory variable. Unfortunately, the
CEEB survey did not contain appropriate measures.
4We chose the 1996 survey because it dealt most explicitly with EU
issues for post-communist countries (Cunningham 1996). It remains
to be seen how time-serial data will affect our findings.
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
560 IOSHUA A. TUCKER, ALEXANDER C. PACEK, AND ADAM /. BERINSKY
approximately 1000 respondents each from ten postcommunist
countries considering membership in the
EU: Bulgaria, the Czech Republic, Slovakia, Estonia,
Hungary, Latvia, Lithuania, Poland, Romania, and
Slovenia.5
For our dependent variable in this analysis, we use
the answer to the following question: "If there were to
be a referendum tomorrow on the question of (OUR
COUNTRY'S) membership, would you personally vote
for or against membership?"6 We grouped respondents
into four categories: support EU membership, oppose
EU membership, would not vote in such a referendum,
and undecided/don't know/no answer.7 There is a similarly
straightforward variable for respondents' attitudes
towards the market economy. Respondents were asked
"Do you personally feel that the creation of a free mar?
ket economy, that is one largely free from state con?
trol, is right or wrong for (OUR COUNTRY'S) future?"
Those who answered yes were coded as a 1; those who
said no were coded as a 0.8
As noted above, we chose not to ascribe winner or
loser status to respondents on the basis of their socioeconomic
characteristics, relying instead on self-assessment.
To do so, we used traditional measures of eco?
nomic self-assessment long used in the economic voting
literature. Voters in the survey were asked to place their
evaluation of their own financial situation over the past
twelve months on a five-point scale ranging from "got a
lot better" to "got a lot worse." They were also asked to
answer this question in terms of their anticipated finan?
cial situation over the next twelve months along the same
scale. While the two measures were closely correlated (r
5The process of being officially considered for membership in the
EU began for Hungary and Poland in 1994, Bulgaria, Estonia,
Latvia, Lithuania, Romania, and Slovakia in 1995, and the Czech
Republic and Slovenia in 1996 (Cichowski 2000).
6As a robustness test, we also conducted our analyses using an al?
ternative survey question as the dependent variable that asked respondents
"As you might know, fifteen states of 'Western' Europe
form together the 'European Union'. Would you say that your impressions
of the aims and activities of the European Union are
generally positive, neutral, or negative?" The results presented below
were robust to using the alternative measure of attitudes to?
wards the EU.
7The question actually had five coded responses, with undecided
and don't know/no answer as separate categories. Since the two
categories are not very distinct substantively, we collapsed them,
but the results are robust to rescaling the variable in its five-category
format.
8Slightly fewer than 18 percent of the respondents were coded as
"don't know." We reran our analyses including these respondents
using a series of dummy variables for the three responses. The re?
sults were robust to this respecification. The only additional insight
gained is, not surprisingly, that the respondents that failed to an?
swer this question also tended towards the "don't know/no answer"
and the "would not vote" categories of the dependent variable.
=.50), they were not identical. Rather than rely on either
measure, we combined them in a single measure. Our intuition
was that the strongest "winners" were those who
both thought their financial situation had improved
most over the past twelve months and who thought that
it would continue to improve over the next twelve
months. Likewise, the biggest losers were those who
thought things had gotten a lot worse and would con?
tinue to get a lot worse.9 Technically, the measure was
calculated by taking the mean of each voter's score across
both categories.10 For ease of interpretation, we rescaled
the variable along a 0-1 continuum (as we did with all of
our independent variables). The distribution of this vari?
able is listed in Table 1.
For demographic control variables, we include: gender,
education, age, residency, profession, and income.
The latter three are included as a series of dummy vari?
ables (with one category excluded) so as to not assume
any ordinal or linear relationships.11
Empirical Findings
As the dependent variable in the analysis is categorical
and nonordered, an appropriate model for analysis is
Multinomial Logit (MNL). We focus here on substantive
quantities of interest, namely the effect that our primary
independent variables have on an individual's likelihood
of supporting EU membership.12 To reiterate our predic-
9Results are robust to respecifying the variable as either retrospective
or prospective evaluation. When we use either alone we find
the same results, albeit with slightly smaller effects. When we in?
clude both measures in the analysis, we again find the same results,
although it appears that the prospective variable is doing more of
the work; but even in this case we still find the retrospective vari?
able helpful.
l0Respondents were assigned a value on this scale as long as they
answered at least one of the two questions. If they answered both
questions (92 percent of respondents), the "winner" value was the
mean of the two answers. If they answered only one of the two (7
percent), the "winner" value was the value of the one nonmissing
response; only 1 percent refused to answer both questions.
nThe excluded variables are residence in a village/rural area,
"other" occupation, and income in the third quartile. In two of the
single country analyses (Bulgaria and Poland), some ofthe demo?
graphic variables predicted categories perfectly, resulting in nonsensical
coefficients and standard errors. For these countries, we
dropped the variables in question. The key results of interest were
robust to these changes.
12Predicted probabilities were calculated using stochastic simulation.
Demographic variables were held at their mean, and the explanatory
variables were set at the various levels noted in the
figures. Reported probabilities are the mean ofthe 1000 simulated
predicted probabilities generated by Clarify 1.3 (Tomz,
Wittenberg, and King 2000) used in conjunction with Stata 6.0.
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
TRANSITIONAL WINNERS AND LOSERS $6l
Table 1 Distribution of "Winners" Across the
tions, we expect that individuals who are winners w
more likely to support EU membership than tho
are losers. In addition, we predict that individual
favor a free market economy will be more supporti
EU membership than those who do not.
As a first cut, we consider all respondents simu
neously in a single pooled dataset, with the cavea
these are general effects and may be masking cou
by-country variation (which we address below). C
cients and standard errors for the pooled analysis c
found in the first column of Table 2; predicted
abilities are plotted in Figure 1. Figure 1 clearly dem
strates support for both of the primary hypotheses.
controlling for demographic effects, moving from
a loser to a winner increases one's likelihood of supp
ing EU membership (denoted by the line wit
square markers). Likewise, support for a free m
economy has a large effect on the probability of
porting EU membership, as we can see by the acros
board higher probabilities of supporting membe
and lower probabilities of opposing membershi
angle markers) in the bottom panel, regardless o
ner status. And taken together, the effects are dra
an extreme "loser" who opposes the free market on
slightly more than a 30 percent chance of supportin
membership, while an extreme "winner" who sup
the free market has an almost 70 percent chance of
porting EU membership. Conversely, an extrem
who opposed the free market had an almost 20 p
chance of opposing EU membership, while an ex
winner who supported the free market had less tha
percent chance of opposing membership. Moreo
quick glance at the coefficients and standard err
Table 2 shows that not only are these effects su
tively meaningful, they are also statistically signifi
Thus the central point of our analysis is clear: w
and supporters of the free market are more likely t
port EU membership.
Figure 1 also gives us an o
effects of winner/loser statu
ket on the probability of a r
did not plan to vote (star m
have an opinion on the mat
estingly, we see largely the sa
the "oppose EU membership
probability of giving this a
from loser to winner status
posing the free market to su
it appears that as the proba
bership goes up, the increas
from the ranks of those w
but also from those who m
have an opinion. These resu
MNL analysis and should
against merely modeling eit
intermediary category (e.g.
Although Figure 1 conta
demographic effects, Table
of observations. As one m
less likely to support EU m
residents of rural areas an
than their more urban cou
ship (although, quite surpri
difference between reside
villages). Men seem to be so
membership than women
after controlling for loser/
the free market, more educ
oppose EU membership th
voting, or even not have an
to note that almost none
variables seem to have im
coefficients and relatively h
the demographic picture i
clear patterns, standing in m
planatory variables of the st
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
562 JOSHUA A. TUCKER, ALEXANDER C. PACEK, AND ADAM J. BERINSKY
Table 2 Country-by-Country MNL Analysis of Preference for EU Membership. Pro-EU Membership
Coefficients with Standard Errors in Parentheses
Note: "Oppose Membership" is the base category
***p < .001, **p<.05, *.p<.01
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
TRANSITIONAL WINNERS AND LOSERS 563
Table 2 (continued)
Don't Know / Undecided Coefficients with Standard Errors in Parentheses
Note: "Oppose Membership" is the base category
***p < .001, **p<.05, *.p<.01
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
564 IOSHUA A. TUCKERj ALEXANDER C. PACEK, AND ADAM J. BERINSKY
Table 2 (continued)
Would Not Vote Coefficients with Standard Errors in Parentheses
Note: "Oppose Membership" is the base category
*"p<.001,**p<.05,*.p<.01
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
TRANSITIONAL WINNERS AND LOSERS
565
Figure 1 Pooled Analysis of Attitudes Toward EU Membership
Oppose Free Market Economy
0.800
0.700 ]
cd 0.600
C/)
c
?- 0.500CD
CC
^ 0.400 -I
= 0.300
2 0.200
CL
0.100
0.000
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Winner Score
0.800
0.700
? 0.600
C/)
c
8" 0.500
CD
CC
\ 0.400
j? 0.300
o
Cu
0.200
0.100
0:000
Support Free Market Economy
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Winner Score
-*- Support -Oppose -Don't Know -NoVote
Next, we analyzed the data for each country separately.
We first ran ten MNL analyses?one for each
country?and then computed the predicted probabilities
for each country separately. The coefficients and
standard errors are found in Table 2. The substantive re?
sults of these analyses are summarized in Figure 2. In an
effort to keep the figure readable, we plot only the pre?
dicted probabilities of supporting EU membership.13
13Readers interested in the predicted probabilities for the other
categories of the dependent variable can find additional figures at
http://www.wws.princeton.edu/~jtucker/pubs.html.
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
566 JOSHUA A. TUCKER, ALEXANDER C. PACEK, AND ADAM J. BERINSKY
Figure 2 Probability of Supporting EU Membership by Country
Oppose Free Market Economy
1.000
0.900 ?
-3 0.800 ?
LU
v> 0.700?.
0.600
a
co 0.500
JD
CO
-Q
O
0.400
0.300
0.200 -I
0.100
0.000
0.1 0.2 0.3 0.4 0.5 0.6
Winner Score
0.7 0.8 0.9
1.000 n
0.900
Gj 0.800 ]
O)
% 0.700
o
Q.
Q.
w 0.500
0.600
^ 0.400 '"?
0.300
(? 0.200
0.100
0.000
Support Free Market Economy
0.1 0.2 0.3 0.4 0.5 0.6
Winner Score
0.7 0.8 0.9
- Bulgaria
- Estonia
- Poland
- Hungary
- Romania
? Latvia
- Czech Republic
- Lithuania
- Slovakia
- Slovenia
Figure 2 provides strong additional support for both
of our central hypotheses. Consider first winner status.
Regardless of whether or not one supports the free mar?
ket, moving from being a loser to a winner has a substantively
meaningful effect on the likelihood of supporting
the EU in eight out of the ten countries; Table 2 illustrates
that these effects are statistically significant as well.
Indeed, across these eight countries, the average increase
in the likelihood of supporting EU membership as one
moves from being a loser to a winner is 29.5 percent for
those who support the free market and 31.8 percent for
those who oppose the free market. In some cases, the in?
crease is significantly larger: Lithuanian winners who
support the free market are 47.3 percent more likely to
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
TRANSITIONAL WINNERS AND LOSERS 567
support EU membership than Lithuanian losers
support the free market.
Only two countries provide exceptions to the
tern. In Slovakia, the effects are positive but are n
substantively interesting nor statistically significan
Poland, the effect is in the incorrect direction, but
not even approach statistical significance. Why the
countries do not follow the same patterns as the
eight is an interesting question for future research
for the time being we leave this as an interesting p
and evidence that there is nothing inherent in our
sis that prevents results that falsify our hypotheses.
Turning to support for the free market, the evide
is even stronger. Across all ten countries, and
single possible value for loser winner/status in
countries, the probability of supporting EU me
ship is always higher when the respondent supp
free market economy. This effect ranges from a f
insignificant 3.3 percent increase in the likeliho
supporting EU membership for an extreme Rom
winner (winner = 1) to a 24.9 percent increase for a
treme Polish loser (winner = 0).14 The average va
the effect across all possible combinations was
percent increase in the likelihood of supportin
membership. While the magnitude of this effect is
large as for loser/winner status, it is striking that t
fect is in the correctly predicted direction in every
While Figure 2 again does not tell us anything ab
the effects of demographic variables, a quick gla
Table 2 reveals a fairly inconsistent pattern. An
whelming majority of the coefficients are nowhere
to being statistically meaningful. In addition, the s
coefficients switch frequently from country to co
indicating a lack of cross-national consistency in th
fects of these variables. Moreover, the findings fro
pooled results regarding demographic variables l
disappear. For example, we only have confidence in
out of ten countries (Lithuania, Poland, Czech Re
that men are more likely to oppose than prefer EU
bership, two countries (Bulgaria and Hungary) for
people, and two countries (Lithuania and Sloveni
more educated people (see Table 2). In the other
tries, not only are the effects not statistically signi
they are also sometimes in the opposite direction
stands in marked contrast to the two variables used to
test our hypotheses, which are in the correct direction in
the first panel (comparing pro-EU to oppose EU) of
Table 2 in nineteen out of twenty cases and statistically
significant (p<.05) in fifteen of those cases.
14The low value for the effect in Romania is no doubt due in part
to the fact that Romanians had the highest overall support for EU
membership, with over 82 percent supporting membership.
In short, we find extremely strong support for our
hypotheses' predictions, even controlling for demo?
graphic variables. Furthermore, the effects are much
more consistently significant and important than the de?
mographic variables.
Speculation on the Effects
of EU Membership
The hypotheses tested above were based on an assump?
tion that attitudes towards EU membership are not a
function of one's political party preference. Instead, we
suggested that attitudes towards joining the EU could
have an effect upon the types of parties that voters are
likely to prefer. In this section, we develop this line of reasoning
further and present some initial speculative evi?
dence in this regard.
If attitudes towards the EU are a function of a fundamental
belief on the part of citizens about the ultimate
outcome of the transition to a market economy,
then it is likely that such attitudes could help structure
other political preferences. One possibility is that atti?
tudes concerning EU membership serve to direct citi?
zens toward particular policies and the parties that mar?
ket them. Like citizens, political parties are faced with
the host of issues, benefits, and consequences that EU
membership provides. Political parties that direct programmatic
appeals to either transitional winners or los?
ers certainly weigh the benefits and costs of EU mem?
bership to their respective constituents (Andor 2001;
Szczerbiak and Taggart 2000). Both the economic and
national identity issues that shape attitudes toward the
EU are often reflected in the positions of parties on EU
membership. Szczerbiak and Taggart note that some
parties are predisposed toward "policy Euroskepticism"
in large part over economic issues, while "national iden?
tity Euroskepticism" involves standing up for the per?
ceived national interest (2000, 3).
To this end, we would expect that voters who sup?
port EU membership would gravitate towards pro-EU
political parties, while those who oppose EU member?
ship would favor "Euroskeptic" parties. As an initial exploration
of this contention, we attempted to see if such
patterns could be found in two countries: Slovakia and
Hungary. We chose these two countries because both had
elections two years prior to the survey (1994) and both
would go on to have elections two years after the survey
(1998). In this manner, we hoped to hold constant the sa?
lience of political competition in the minds of the electorate.
Unfortunately, our decision to examine only a
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
568 JOSHUA A. TUCKER, ALEXANDER C. PACEK, AND ADAM J. BERINSKY
subsection of countries was also dictated by the fact that
the data on party preferences in some of the other coun?
tries were corrupted in the version of the survey that we
received from the ICPSR.15 Therefore, we advise readers
to take the results presented below as merely suggestive
of interesting grounds for future research.
To test this contention, we classified parties in our
sample according to whether they support EU member?
ship, or express various concerns or objections on the par?
ties' official programs or election manifestos based on the
programs made available on each party's website.16 If the
party in question appears to give more or less unqualified
support for EU membership, we classified that party as
pro-EU. If the particular party expresses outright opposition
to or serious reservations about EU membership,
then we classified the party as Euroskeptic.17
Our dependent variable in this analysis was straightforward;
we relied on the following survey question: "If
there were a general election tomorrow, which party or
block would you vote for, or might you be inclined to vote
for?" Because our hypothesis predicts that citizens who
support EU membership will gravitate towards EU parties
but says nothing about which of the pro-EU parties they
are likely to support, we coded the dependent variable as
a dichotomous variable: either citizens prefer a pro-EU
party (dependent variable = 1) or a Euroskeptic party (de-
15After careful investigation, we are confident that the data for the
Hungarian and Slovak cases are correct. The only exception is that
there was some confusion on the survey concerning the coding of
supporters of the Alliance of Free Democrats in Hungary. For this
reason, the model was run both including and excluding support?
ers of the party. The results were practically the same, and we re?
port the version including the party. We found no evidence of data
corruption in other areas of the survey.
16We are indebted to Leonard Pacek and Suzanne Gyzsely for their
assistance in translating party programs. The programs, platforms,
and manifestos cover the period from the mid to late 1990s. This
raises the question of shifts in party positions on EU membership
from the time ofthe 1996 CEEB surveys used in our analysis. The
contention that party programs change over time is not disputed
here (see Budge, Robertson, and Hearl 1987). However, we assume
that parties are not likely to change dramatically in such a short
period of time on such salient issues.
17Our analysis is of course restricted to those parties included in
the survey. For Hungary, pro-EU parties include Alliance of Young
Democrats, Christian Democratic People's Party, Hungarian
Democratic Forum, Hungarian Socialist Party, and Alliance of Free
Democrats; parties coded as Euroskeptics are Agrarian Alliance,
Independent Smallholders Party, Party of Hungarian Justice and
Life, and the Workers Party. For Slovakia, Party of the Democratic
Left, Christian Democratic Movement, Coexistence, Slovak Social
Democratic Party, Hungarian Christian Democratic Movement,
Democratic Party and the Democratic Union are coded as Pro-EU,
while the Movement for a Democratic Slovakia, Slovak National
Party, Association of Workers of Slovakia, and the Green Party
of Slovakia are coded as Euroskeptics. We exclude those respon?
dents who did not express a party preference in the hypothetical
election.
pendent variable = 0). As independent variables we included
dummy variables for each of the four options for
the respondent's position on EU membership (omitting
the necessary one category, here opposing EU member?
ship). We also included all of the demographic control
variables used in the previous section. As the dependent
variable is dichotomous, we used binomial logit analysis;
the results are presented in Table 3.
Table 3 suggests that an individual's attitude towards
EU membership can have explanatory power for assessing
whether an individual is likely to support a pro-EU
or Euroskeptic political party. In both cases, the coeffi?
cient for the support EU membership dummy variable is
positive and statistically significant, meaning we are confident
that those who support the EU are more likely to
support pro-EU parties than those who oppose EU
membership, even after controlling for a range of demo?
graphic effects. While it is of course possible that the introduction
of other attitudinal variables could reduce
these effects, our results support the hypothesis that gen?
eral feelings about the transition drive the desirability of
joining the EU, which in turn drives support for certain
types of political parties.
As further evidence in this regard, consider the recently
completed 2001 Polish parliamentary election. In
this election, two previously insignificant parties, the
Self-Defense Party and the Polish Family League, re?
ceived approximately 17 percent of the vote between
them, compared to less than 0.1 percent for Self-Defense
in the previous election; the Polish Family League did
not even compete in the previous election. While it is
too early to present quantitative analysis of survey data
from this election, there is a strong consensus in the
press that the anti-EU (or Euroskeptic) views of these
parties played in an immeasurable role in their electoral
success. The Self-Defense party leader, Andrzej Lepper,
was best known "for organizing road blockades and violent
protests against Poland's pro-European policies"
(Maksymiuk 2001), while the Polish Family League
"damned the EU as a 'civilisation of death' for what it
takes to be the EU's views on abortion and euthanasia"
(Economist 2001). As part of their campaign strategy,
the League blamed EU accession talks for much of
Poland's unemployment; in its platform it promised that
"We will abolish the association agreement with the Eu?
ropean Union because this agreement causes the Polish
economy losses of 20 billion dollars (21.7 billion euros)
a year." (Agence France Press 2001). Both parties were
routinely labeled?using the very term applied here?as
Euroskeptics. While there are undoubtedly many reasons
that voters turned to these parties, the fact that
their pre-election identity was so closely tied to their
concern over Poland's path towards EU membership
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
TRANSITIONAL WINNERS AND LOSERS 569
Table 3 Logit Analysis of Support for Pro-E
Parties in Slovakia and Hungary
Coefficients with Standard Errors in
Parentheses
*p < .001, **p<.05, *.p<.01
suggests that this concern must have played a nontrivial
role in attracting supporters; it would be extremely sur?
prising if future survey analysis did not bare out this
conclusion. If so, then it provides clear anecdotal evi?
dence of a case where party support was in fact driven
by voters' attitudes towards the EU.
Discussion and Implications
How citizens form opinions regarding potential EU
membership is a question of interest to both policy makers
and scholars alike. The vast majority of the empirical
work on this topic has examined evidence from Western
Europe. Forays into the post-communist context are lim?
ited, and have largely consisted of applying a series of hy?
potheses generated from the West European studies to a
subsample of the post-communist cases with mixed findings
(Tverdova and Anderson 2000; Cichowski 2000).
Our contribution to this literature, therefore, is threefold.
First, we have presented a theoretical model designed explicitly
for the post-communist context. Second, we have
expanded the analysis of this model to include ten postcommunist
applicant nations. Finally, we have reported a
series of remarkably consistent empirical findings. Regardless
of demographic characteristics, post-communist
citizens that support the free market economy and are
transitional "winners" are more likely to support mem?
bership in the EU. Demographic characteristics, in turn,
prove to have weak and inconsistent effects cross-nationally.
In addition, we move the debate forward by considering
not only the causes of attitudes towards EU mem?
bership, but by speculating on the effects of attitudes
towards EU membership on partisan preferences as well.
The results of our analysis provide strong evidence
that the extent of support for membership in the Euro?
pean Union among post-communist citizens is in large
part a function of the effects of the economic transition
during the previous decade. Those who have benefited?
as well as those who have lost?form opinions in a manner
that is consistent with our theoretical proposition
that citizens in transition countries view EU membership
as a continuation of free market reforms, and thus sup?
port or oppose membership accordingly. Moreover, atti?
tudes about EU membership may help citizens make
sense of political information during elections in an environment
of volatility and uncertainty. These findings
have critical implications for an understanding of how
post-communist citizens evaluate information, political
economy, and the future of post-communist countries in
an integrated Europe.
First, our findings enable us to understand the extent
to which processes in East Central and Southeast Europe
mirror those in the West. While similarities abound in
the general factors that shape mass attitudes in both West
and East Europe, crucial differences exist as well. There is
no West European equivalent to the enormous social and
economic shifts underway across the former communist
world, and thus no equivalent to their effects on citizens
as well. In forming opinions on critical issues such as
EU membership, post-communist citizens face greater
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
$J0 JOSHUA A. TUCKER, ALEXANDER C. PACEK, AND ADAM J. BERINSKY
uncertainty than their western counterparts. While it
would be foolish to suggest that we ignore the rich theoretical
contribution of studies focused on the advanced
industrial democracies, it behooves us to consider the
particulars of the post-communist experience as well.
Second, our focus on the societal division of winners
and losers links our study to several other topics of inquiry
in the field of post-communist politics. The impact
of this division has been observed with respect to support
for parties (Tucker 2001; Fidrmuc 2000; Gibson and
Cielecka 1995; Evans and Whitefield 1993; Kitschelt
1992), electoral participation (Bohrer, Pacek, and Radcliff
2000), and support for incumbent post-communist governments
(Powers and Cox 1997). Our study complements
these previous works by demonstrating that one's
status in the wake of the transition is a powerful determinant
of support for a critical issue.
Third, our findings highlight the effect that attitudes
toward the EU have in different areas. Previous studies
have argued that post-communist voters use their party
attachment as a proxy in answering questions about EU
membership (Cichowski 2000; Anderson 1998). We sug?
gest an alternate theoretical approach; post-communist
citizens use their opinions about EU membership to di?
rect their support toward particular political parties. Posi?
tions on such critical issues can help citizens sift through
uncertain information in choosing new governments.
Finally, our analysis sheds light on the extent to
which domestic and international factors are closely intertwined
in the post-communist context. For citizens of
East Central and Southeast Europe, membership in the
EU is more than simply an international issue. Rather, it
closely parallels the ongoing reform process and its consequences
throughout the region. This in turn implies
that shifts in perceptions about the reform process have
concomitant effects on support for the EU. Whether or
not the EU plays a critical role in promulgating the re?
forms underway in East Central and Southeast Europe, it
will certainly be affected by the consequences of those re?
forms across the region.
References
Agence France Press. 2001. "EU-skeptics Take Quarter of Vote in
Polish Legislative Poll.,, September 24,2001.
Anderson, Christopher. 1998. "When in Doubt, Use Proxies:
Attitudes Toward Domestic Politics and Support for Euro?
pean Integration." Comparative Political Studies 31:569-601.
Anderson, Christopher, and Karl Kaltenthaler. 1996. "The Dynamics
of Public Opinion Toward European Integration,
1973-1999." European Journal of International Relations
2:175-199.
Anderson, Christopher, and Shawn Reichert. 1996. "Economic
Benefits and Support for Membership in The EU: A CrossNational
Analysis." Journal of Public Policy 15:231-249.
Andor, Laszlo. 2001. "The Hungarian Left and the 'Euro-Atlantic
Integration'." Eszmelet (eszmelet.tripod.com/angoll/
andorangl.html).
Bell, Janice and Dina Smeltz. 2000. "Winners and Losers: Who
They Are and What They Think About Transition: Findings
from Surveys in the Czech Republic, Hungary, and Poland
1990-99." Prepared for presentation at the annual meeting
of the American Association for the Advancement of Slavic
Studies, Denver, Colo.
Bohrer, Robert E. II, Alexander C. Pacek, and Benjamin
Radcliff. 2000. "Electoral Participation, Ideology, and Party
Politics in Post-Communist Europe." Journal of Politics
62:1161-1172.
Budge, Ian, David Robertson, and David Hearl. 1987. Ideology
Strategy, and Party Change: Spatial Analysis of Post-War
Election Programmes in 19 Democracies. Cambridge: Cam?
bridge University Press.
Cichowski, Rachel A. 2000. "Western Dreams, Eastern Realities:
Support for the European Union in Central and Eastern Eu?
rope." Comparative Political Studies 33:1243-1278.
Cunningham, George. 1996. Central and Eastern Eurobarometer
7: Status of the European Union, October-November
1996, ICPSR Study No. 2296.
Duchene, Francoise. 1982. "Community Attitudes." In The 2nd
Enlargement of the European Economic Community, ed.
Dudley Seers and Constantine Vaitsos. New York: St. Martins
Press.
Economist. 2001 "The left is back?in the centre," September
29,2001.
Eichenberg, Richard, and Russell Dalton. 1993. "Europeans and
the European Communisty: The Dynamics of Public Sup?
port for European Integration." International Organization
47:507-534.
Evans, Geoffrey and Stephen Whitefield. 1993. "Identifying the
Bases of Party Competition in Eastern Europe." British Jour?
nal of Political Science 23:521-548.
Fidrmuc, Jan. 2000. "Economies of Voting in Post-Communist
Countries." Electoral Studies 19: 199-217.
Gabel, Matthew J. 2000. "European Integration, Voters, and Na?
tional Politics." West European Politics 23:52-72.
Gabel, Matthew J. 1998a. "Economic Integration and Mass
Politics: Market Liberalization and Public Attitudes in the
European Union." American Journal of Political Science
42:936-953.
Gabel, Matthew J. 1998b. "Public Support for European Inte?
gration: An Empirical Test of Five Theories." Journal of Poli?
tics 60:333-364.
Gabel, Matthew J. and Guy D. Whitten. 1997. "Economic Con?
ditions, Economic Perceptions, and Public Support for Eu?
ropean Integration." Political Behavior 19:81-96.
Gabel, Matthew J. and Harvey Palmer. 1995. "Understanding
Variation in Public Support for European Integration," Eu?
ropean Journal of Political Research 27:3-19.
Gibson, John, and Anna Cielecka. 1995. "Economic Influences
on the Political Support for Market Reforms in PostCommunist
Transitions: Some Evidence from the 1993 Polish
Parliamentary Elections." Europe-Asia Studies 47:765-
785.
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms
TRANSITIONAL WINNERS AND LOSERS 571
Hug, Simon, and Pascal Sciarini. 2000. "Refere
pean Integration: Do Institutions Matter in t
cision?" Comparative Political Studies 33:3-36
Janssen, Joseph. 1991. "Postmaterialism, Cog
tion and Support for European Integration." B
of Political Science 21:443.
Kirka, Danica. 2000. "Central Europeans An
Membership." Associated Press. July 9,2000.
Kitschelt, Herbert, Zdenka Mansfledova, Rado
and Gabor Toka. 1999. Post-Communist Pa
Competition, Representation, and Inter-Part
Cambridge: Cambridge University Press.
Kitschelt, Herbert. 1992. "The Formation of P
East Central Europe." Politics and Society 20:
Mackie, Thomas. 1995. "Parties and Election
the New Europe, ed. Jack Hayward and Ed
Durham: Duke University Press.
Maksymiuk, Jan. 2001. "Left Alliance Takes
RFE/RL Newsline 5:184.
Palmer, Harvey, and Guy Whitten. 1999. "The
of Unexpected Inflation and Economic Gro
Journal of Political Science 29:623-639.
Plasser, Fritz, Peter Ulram, and Harald Wa
Democratic Consolidation in East Central Euro
St. Martins Press.
Powers, Denise, and James Cox. 1997. "Echoe
Relationship Between Satisfaction with Econo
and Voting Behavior in Poland." American Po
Revew 91:617-634.
Szczerbiak, Aleks, and Paul Taggart. 2000. "Opposing Europe:
Party Systems and Opposition to the Union, the Euro, and
Europeanization." (www.susx.ac.uk/Users/ssfj3/oppeuro.
html).
Taggart, Paul. 1998. "A Touchstone of Dissent: Euroskepticism
in Contemporary Western European Party Systems." Euro?
pean Journal of Political Research 33:363-388.
Tomz, Michael, Jason Wittenberg and Gary King. 2000.
CLARIFY: Softwarefor Interpreting and Presenting Statistical
Results. Version 1.3. Cambridge, Massachusetts, http://
gking.harvard.edu/.
Tsoukalis, Loukas. 1981. The European Community and its
Mediterranean Enlargement. London: George Allen and
Unwin.
Tucker, Joshua A. 2001. "Economic Conditions and the Vote for
Incumbent Parties in Russia, Poland, Hungary, Slovakia,
and the Czech Republic from 1990 to 1996." Post-Soviet Affairs
17:309-331.
Tuszynski, Jack. 1997. "NATO, the European Union, and Polish
Agriculture." Pigulki 23:1-6.
Tverdova, Yuliya V., and Christopher J. Anderson. 2000.
"Choosing the West: Referendum Choices on EU Member?
ship in East-Central Europe." Presented at the 2000 Annual
American Political Science Association meeting, Washing?
ton, D.C.
van der Eijk, Cees, and Mark Franklin. 1996. Choosing Europe?
Ann Arbor: University of Michigan Press.
Wallace, William. 1990. The Transformation of Western Europe.
London: Pinter.
This content downloaded from 80.77.242.154 on Mon, 08 Jul 2019 14:00:13 UTC
All use subject to https://about.jstor.org/terms