Political Studies (1998), XLVI, 572±588

Parties, Party Systems, and Satisfaction with
Democratic Performance in The New Europe
CHRISTOPHER J. ANDERSON1
Introduction
Political parties and the party systems they form constitute the major channels of interest aggregation and citizen input in contemporary democracies. They are the vehicles through which political elites supply policy alternatives, and they constitute the major route for citizens to organize the demand for such alternatives.2 Parties also play a crucial role in the nature of democratic governance because they help legitimize the state. After all, free and fair elections in which parties compete for oce are a prime criterion for whether a system should be considered a democracy in the ®rst place. Outside of elections, political parties also have long been the most important mediating institutions between citizens and the state, in particular as parties have taken on the roles as simultaneous agents both of the state and its citizens.3
While virtually all democracies have political parties that compete for oce, political systems di€er in a number of important ways with regard to how they go about channeling inputs or providing policy alternatives, and with regard to the roles they assign parties in this process. Moreover, and perhaps most importantly, the ways in which political institutions condition the formation, functioning, and development of political parties and party systems varies as well. While there is an extensive literature linking electoral systems and the development of party systems, few researchers have investigated their link with how citizens ± the ultimate arbiters of democratic governance ± feel about the way the political process performs.
This essay focuses on two types of political performance that are expected to a€ect citizen attitudes about the political system: party and party system performance. While party performance focuses on how individual political parties perform their roles as mobilizing and organizing agents in elections, party system performance involves the aggregate constellation of parties in a political system and its electoral consequences in the form of party system fragmentation and volatility. This essay investigates whether, and how, crossnational di€erences in both party and party system performance a€ect citizens' evaluations of their country's political system.
1 Many thanks to Richard I. Ho€erbert, participants at the SUNY-Binghamton political science graduate colloquium, and the political science department at Washington University in St. Louis for thoughtful comments on earlier ideas and drafts.
2
H. D. Klingemann, R. Ho€erbert, and I. Budge, Parties, Policies, and Democracy (Boulder,
CO, Westview, 1994).
3
R. Katz, and P. Mair, `Changing models of party organization: The emergence of the cartel party', Party Politics 1,1 (1995), 5±28.

# Political Studies Association 1998. Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main
Street, Malden, MA 02148, USA.

CHRISTOPHER J. ANDERSON

573

On the basis of directly comparable survey evidence from about 20 democracies collected between 1993 and 1995, this article examines the determinants of cross-national di€erences in democratic support in both old and new democracies. Speci®cally, I focus on satisfaction with democracy in the 15 member states of the European Union as well as in several of the emerging democracies of east central Europe. I investigate whether and to what extent electoral rules and party system performance help us understand di€erences in levels of satisfaction with the political system in both old and new democracies.
The next section reviews the literature on electoral systems and party performance; subsequently, I develop a model of how these may be related to levels of democracy satisfaction in contemporary democracies. The empirical analysis then tests these relationships systematically, and a concluding section discusses the results and spells out implications for further research.
Electoral Systems, Party Systems, and Performance
What are the relationships between a country's electoral institutions and the outcomes they produce? For one, it is well known that electoral systems are the primary institutional factor that constrains the development and nature of a country's party system. Moreover, it is widely assumed that the laws governing electoral systems have politically non-neutral consequences.4 Simply put, it is believed that the nature of the electoral system in place in a country a€ects the type of party system that develops (two-party versus multi-party) by in¯uencing the strategic considerations of voters and political elites. The more proportional the electoral system, the more likely it is that the party system is fragmented.5
Consequently, much of comparative political science has sought to extend the notion that political institutions a€ect parties and party systems in democracies.6
Electoral rules, however, do not simply impose such non-neutral consequences on the participants in the governing process; they also provide them with strategic opportunities. Moreover, electoral rules are not dictated by some magic force, but structured and used by citizens and elites alike. Electoral rules can be said to have two kinds of consequences: ®rst, they delineate the probabilities that particular political parties will gain representation in parliament.
Second, they in¯uence the electoral strategies of both voters and political parties. Hence, they also constrain both the representative and policy making processes in a country and a€ect how citizens evaluate the performance of the democracy they live in.
Aside from institutional determinants of party and party system development and performance, a number of scholars have pointed to factors rooted in the socio-structural environment as important determinants of party and party
4
M. Duverger, Political Parties (New York, Wiley and Sons, 1954). D. W. Rae, The Political
Consequences of Electoral Laws (New Haven, Yale University Press, 1967). W. H. Riker, `The number of political parties: A re-examination of Duverger's law', Comparative Politics 9 (1976):
93±106. A. Lijphart, `The political consequences of electoral laws, 1945±85', The American Political
Science Review 84,9 (1990): 481±496.
5
A. Lijphart, Electoral Systems and Party Systems: A Study of Twenty-Seven Democracies, 1945±
1990 (New York, Oxford University Press, 1994).
6
A. Lijphart, `Constitutional choices for new democracies', Journal of Democracy 2,1 (1991), 72±
84. D. W. Rae, Political Consequences of Electoral Laws.

# Political Studies Association, 1998

574

Democratic Performance in The New Europe

system performance.7 Usually phrased in terms of changes in party attachment or realignment and dealignment, a number of electoral researchers have pointed out that party systems undergo periodic changes as a result of the changing composition and political socialization of mass electorates.8 According to this line of research, party systems change or remain the same because of changes
(or lack thereof) in a country's social structure that produce changes in individuals' attachments to political parties.
Yet, due to di€erently structured political and electoral institutions and di€erent electoral strategies, similar changes in social structure across two countries do not necessarily have similar consequences. While some countries' electoral institutions encourage the establishment of new political parties or the reorientation of old ones, others inhibit such responsiveness. As a consequence, political, and in particular electoral, institutions play a crucial role with regard to how social changes are translated into changes at the level of political parties.9 Similarly, depending on how electoral rules condition the development of political alternatives, parties choose to target particular segments of the electoral market, thus mobilizing particular constituencies and neglecting others. Taken together, both the institutional and socio-structural approaches have shown quite convincingly that electoral rules produce di€erent constellations of parties and party systems in the context of di€erent societal and social structural constraints.10
This essay serves two tasks: ®rst, it will examine the consequences electoral systems have for party and party system performance, as well as the relationships among party and party system performance. Second, it will also examine how institutions (such as electoral systems) and the performance they produce are related to how citizens evaluate the democratic performance of their particular polity.
A Model of Party System Performance and Democracy Satisfaction
Logically, the establishment of electoral procedures and institutions precedes the measurement of support for the system; this assumes a chain of causality running from institutions, party performance (e.g., voter mobilization), and party system performance (e.g., electoral volatility) on one hand to satisfaction with democracy on the other. Electoral rules frequently have been hypothesized to a€ect the levels of fragmentation as well as volatility of party systems.11

7
S. M. Lipset, and S. Rokkan, Party Systems and Voter Alignments (New York, Free, 1967).
R. Rose, and D. Urwin, `Persistence and change in western party systems since 1945', Political
Studies, 18 (1970), 287±319.
8
R. J. Dalton, Citizen Politics: Public Opinion and Political Parties in Advanced Industrial
Democracies (Chatham, NJ, Chatham House, 1996). R. Inglehart, Modernization and Postmodernization (Princeton, Princeton University Press, 1997).
9
S. Bartolini, and P. Mair, Identity, Competition, and Electoral Availability: the Stabilisation of
European Electorates 1885±1985 (New York, Cambridge University Press, 1990).
10 M. Duverger, Political Parties. A. Lijphart, Electoral Systems and Party Systems: a Study of
Twenty-Seven Democracies, 1945±1990 (New York, Oxford University Press, 1994). R. J. Dalton,
S. Flanagan, and P. Beck, Electoral Change in Advanced Industrial Societies: Realignment or
Dealignment (Princeton, Princeton University Press, 1984).
11
M. Duverger, Political Parties. D. W. Rae, Political Consequences of Electoral Laws.
A. Lijphart, Electoral Systems and Party Systems.

# Political Studies Association, 1998

CHRISTOPHER J. ANDERSON

575

Based on this research, electoral rules thus should result in particular kinds of party and party system performance, which, in turn, should a€ect citizen satisfaction with democratic governance. The e€ect of electoral institutions on democracy satisfaction, however, should be indirect and hence mediated by political performance. We can express this model of how electoral rules, party system performance, and democracy satisfaction are related as follows:
Electoral system 3 PartyaParty system performance 3 Satisfaction with democracy

(1)

According to this model, electoral system and party system performance indicators can be taken to be independent variables that a€ect the level of citizen satisfaction with democratic performance. They a€ord citizens greater or fewer choices; preferences are translated in a very proportional or not very proportional fashion; and they de®ne the extent to which citizens can vote strategically. Taking a similar approach, Miller and Listhaug have argued, for example, that a smaller number of parties in a system (which, to some degree, may be a consequence of electoral laws) is correlated with popular dissatisfaction with democratic governance because there are fewer policy choices for citizens.12 Put di€erently, because some electoral systems inhibit the emergence of new parties to take account of new demands, system support may be lower in the long run. Thus, countries with fewer parties also should have lower levels of citizen satisfaction with the system. In contrast to the research by Miller and
Listhaug, however, Weil's cross-national study of system support in western democracies found that party system fragmentation ± that is, a larger number of parties ± was systematically associated with lower levels of democratic support.13 However, to date few scholars have sought to systematically disentangle the fragmentation-citizen support linkage.
Some scholars have pointed to the changed role parties play in western democracies as an in¯uence on system support. In terms of their representative role, established parties seem to have become less relevant and to be losing some of their key functions, while at the same time they seem more privileged than ever in public oce and in terms of their linkage to the state.14 This combination is probably one of the main factors associated with anti-party sentiment, electoral volatility, and the emergence of new and protest parties.15 Following this line of argument, a greater number of parties also should be associated with lower levels of system support. Below, I will test whether party system fragmentation and volatility ± that is, indicators of party system performance ± indeed are systematically associated with levels of performance satisfaction in contemporary democracies.
Based on the notion that electoral institutions work in distinct and variable socio-structural as well as politically relevant strategic contexts, it is hypothesized that electoral systems also are related to party performance ± that is, to parties' electoral and organizational strategies as well as to their e€ectiveness.
12
A. H. Miller, and O. Listhaug, `Political parties and con®dence in government: A comparison of Norway, Sweden and the United States', British Journal of Political Science, 29 (1990), 357±86.
13 F. Weil, `The sources and structure of legitimation in western democracies: a consolidated model tested with time-series data in six countries since World War II', American Sociological
Review, 54 (1989), 82±706.
14
Katz and Mair, `Changing models of party organizations'.
15
S. Scarrow, and T. Poguntke, `The politics of anti-party sentiment', European Journal of
Political Research, 29,3 (1996): 57±262.

# Political Studies Association, 1998

576

Democratic Performance in The New Europe

Speci®cally, in systems that have a more proportional translation of seats into votes ± that is, systems in which the electoral market is more likely to be segmented ± parties should rely more on mobilizing voters along traditional lines of social group ties and party membership. Such mobilization strategies are relatively e€ective and inexpensive because they are based on existing networks and social milieus as well as enduring formal and informal group ties. The extent to which parties in particular systems are able to do so can be considered an indicator of party performance.
Finally, party and party system performance can be expected to be related to each other as well. Party system fragmentation and volatility are expected to be lower in systems with parties that have strong social-group ties and high rates of organizational membership. Taken together, Model 1 can be modi®ed to include a link between party and party system performance as well:
"

"

"

"

Electoral system

performance
"

" Party

Party system performance

Democracy satisfaction

Whether citizens evaluate the functioning of their democracy based on the recent performance of their electoral institutions and organization or based on the more lasting institutions, such as electoral laws, holds importance for the broader scholarly debate about the design of democratic institutions. Speci®cally, political theorists have debated whether it is more important that the institutions produce superior outcomes or that the institutions are designed in a way that produces maximum process fairness to all participants.16 Translated for the present study, this would mean asking whether party/party system performance or the type of electoral institutions have a stronger e€ect on democracy satisfaction.
Below, I test these models of the relationships among electoral systems, party system performance, and democracy satisfaction in greater detail. Based on systematic data collected in 19 European democracies, the analysis examines the following hypotheses:
Hypothesis 1: The more proportional the electoral system, the more fragmented the party system and the higher the level of electoral volatility.
Hypothesis 2: The more proportional the electoral system, the stronger the linkages between parties and voters.
Hypothesis 3: The stronger the linkages between parties and voters, the lower the level of party system fragmentation and volatility.
Hypothesis 4a: A greater number of parties in a system is correlated with higher levels of system support (Miller and Listhaug).
Hypothesis 4b: A greater number of parties in a system is correlated with lower levels of system support (Weil).
Hypothesis 5: Higher levels of electoral volatility are correlated with lower levels of system support.
Hypothesis 6: Closer links between citizens and political parties are correlated with higher levels of system support.
16
A. Lijphart, Democracies (New Haven, Yale University Press, 1984). G. B. Powell, Contemporary Democracies: Participation, Stability and Violence (Cambridge MA, Harvard University
Press, 1982).

# Political Studies Association, 1998

CHRISTOPHER J. ANDERSON

577

The countries included in this study are 14 of the member states of the
European Union (Austria, Belgium, Denmark, Finland, France, Germany,
Greece, Ireland, Luxembourg, the Netherlands, Portugal, Spain, Sweden, and the United Kingdom),17 as well as ®ve new democracies of east central Europe
(Bulgaria, Czech Republic, Hungary, Poland, and Slovakia). Because the rise of democratic systems among the countries of the former Soviet bloc is such a recent phenomenon, much of the theorizing and most of the empirical studies about electoral systems, party systems, and system support has been based on the experience of established democracies. The countries included here were chosen because they allow for the testing of theories of democracy both in established and newly established democratic systems; the new democracies included here also ful®ll the criterion of having the most stable democratic systems of east central Europe.
Electoral Systems and Party System Performance
The model proposed above posits that the nature of a country's electoral system is related to party system performance. To compare electoral systems across the variety of democracies included in this study, I rely on an index of proportionality (or disproportionality) originally developed by Loosemore and
Hanby.18 The disproportionality index represents the total deviation from proportionality (% votes minus % seats). To calculate the index, the absolute values of all vote-share di€erences are added and divided by 2:19
D ˆ 1 jvi À si j
2
One can easily transform the index of disproportionality into an index of proportionality by subtracting its values from 100 (P ˆ 100 7 D); in this case,
100 would be equal to perfect proportionality.20 This kind of index has the advantage of providing a representation of any electoral system, regardless of whether it is based on proportional representation, single-transferrable vote counts, or plurality rules.
To assess party system performance, I rely on indicators of fragmentation and volatility. For the fragmentation measure, I utilize the Laakso and Taagepera index measuring the e€ective number or parliamentary parties in a system.21
The e€ective number of parties is calculated as follows:22 n ˆ
N ˆ 1a pi iˆ1

17 Italy was left out of the analysis because it experienced signi®cant electoral volatility, elections under a new electoral law, as well as the fundamental reformation of its party system during the period investigated here. Interested readers may obtain the empirical results with the Italian case included by contacting the author.
18
J. Loosemore and V. Hanby, `The theoretical limits of maximum distortion: some analytic expressions for electoral systems', British Journal of Political Science, 1 (1971), 467±77.
19 A. Lijphart, Electoral Systems and Party Systems, p. 60.
20 T. Mackie and R. Rose, The International Almanac of Electoral History. rev. 3rd ed. (Washington DC, Congressional Quarterly Press, 1991).
21
M. Laakso and R. Taagepera, `E€ective number of parties: a measure with application to west
Europe', Comparative Political Studies, 12 (1979), 3±27.
22
Instead of using the more widely known Rae/Taylor index of party system fractionalization, I measure party system change by changes in the e€ective number of parliamentary and electoral

# Political Studies Association, 1998

578

Democratic Performance in The New Europe

where pi is the proportion of parliamentary seats for the I-th party.23 This formula contains information about the number and relative size of the parties in the system. It thus helps to di€erentiate not only between two- and multiparty systems, but it is also a more subtle measure than simply counting the number of parties that gain representation or receive votes. The proposed measure takes the relative strength and parliamentary viability of parties into account. Or, as Taagepera and Shugart have put it:
The advantage of using the e€ective, rather than the actual, number of parties is that it establishes a nonarbitrary way to distinguish `signi®cant' parties from less signi®cant ones. The construction of the index is such that each party weights itself by being squared. Tiny parties contribute little to the index, while large parties contribute relatively more.24

Finally, to gauge electoral volatility, I utilize the volatility index devised by
Pedersen,25 measuring the net electoral shifts among political parties in a system. Volatility estimates the total change in electoral outcomes, irrespective of the direction of change. Thus, changes across parties, can be measured by the total net change from one election to the next:
€
V ˆ 1 jPiYt j
2
where Pi,t is the percentage of the vote obtained by party I at election t.
Table 1 shows the distribution of the countries along the proportionality index, as well as the level of party system fragmentation and electoral volatility.
There is signi®cant variation among the country cases included here on all three dimensions. Table 2 reports the Pearson correlation coecients for the association between the electoral system and party system performance indicators. Two correlation analyses were conducted: To see whether the inclusion of the new democracies of east central Europe a€ected the results, one included just the older democracies of Western Europe and one included all cases.
Con®rming earlier work by Lijphart and others, the results indicate that the proportionality of the electoral system indeed is positively associated with party system fragmentation, here measured by the e€ective number of parliamentary parties. The association is of roughly equal strength regardless of whether we include the new democracies of East Central Europe, although it does not reach conventional levels of statistical signi®cance: The more proportional the

parties in a system. Note, however, that the formula for the e€ective number of parties carries the same information as Rae's index of party system fractionalization, only expressed in a di€erent metric. The reader can obtain Rae's fractionalization index (F) by substituting the value for the e€ective number of parties (N) in the following formula: F ˆ 1 7 (1/N).
23
A. Lijphart, Democracies, p. 120.
24 R. Taagepera, and M. Shugart, Seats and Votes (New Haven, Yale University Press, 1993),
p. 455.
25 M. Pedersen, `The dynamics of European party systems: changing patterns of electoral volatility', European Journal of Political Research, 7 (1979), 7±26. M. Pedersen, `Changing patterns of electoral volatility in European party systems, 1948±1977: explorations in explanation', in
H. Daalder and P. Mair (eds), West European Party Systems: Continuity and Change (London,
Sage, 1983), pp.29±66.
# Political Studies Association, 1998

CHRISTOPHER J. ANDERSON

579

TABLE 1. Electoral Proportionality, Fragmentation, and Volatility in 19 Democracies
Country

Proportionalitya

Austria
Belgium
Bulgaria
Czech Republic
Denmark
Finland
France
Germany
Greece
Hungary
Ireland
Luxembourg
Netherlands
Poland
Portugal
Slovakia
Spain
Sweden
United Kingdom

E€. no. of partiesb

97.83
94.02
92.14
95.90
96.66
94.94
79.11
96.69
81.04
85.00
94.78
95.22
97.04
79.85
93.11
86.14
84.89
97.45
87.09

3.40
7.95
2.72
4.85
4.70
4.88
2.96
2.78
2.17
2.89
3.48
3.90
5.38
3.85
2.55
4.41
2.67
3.51
2.26

Volatilityc

Electionsa,c

7.45
7.10
20.00
19.90
8.90
9.20
19.15
6.90
10.75
25.00
15.05
5.10
19.55
27.60
10.10
25.90
6.30
11.25
5.10

15
16
1
1
20
14
15
13
8
2
15
11
15
2
8
2
6
15
14

Sources: a A. Lijphart, Electoral Systems and Party Systems: a Study of Twenty-Seven Democracies
(New York, Oxford University Press, 1994), Appendix B; own calculations on the basis of data reported in European Journal of Political Research (various issues). b L. LeDuc, R. Niemi, and P. Norris, Comparing Democracies: Elections and Voting in
Global Perspective (London, Sage, 1996), Table 1.4. Most recent election only. c P. Mair, Party System Change (Oxford, Clarendon Press, 1997), Table 8.1; own calculations on the basis of data reported in European Journal of Political Research
(various issues).

TABLE 2. Pearson Correlations of Electoral System, Party System Fragmentation, and
Electoral Volatility
Western Europe
Proportionality of electoral system

All cases
Proportionality of electoral system

E€ective no. of parl. parties
Volatility

0.45
À0.12

0.39
À0.39*

N:

14

19

*p 5 0.10.
# Political Studies Association, 1998

580

Democratic Performance in The New Europe

electoral system, the greater the fragmentation of the party system (Pearson correlation coecient 0.4).26
Levels of electoral volatility, however, are largely unrelated to the proportionality of the electoral system when we consider the mature democracies alone.
However, when all cases are included, electoral volatility appears to be negatively associated with the nature of the electoral system: more proportional systems are associated with lower levels of volatility.27 Given that it also is the case that more proportional systems produce a greater number of parties, it appears that more proportional electoral institutions lead to stronger attachments to political parties; or, at the very least, it provides an aggregate indication of producing fewer incentives to switch parties from one election to the next. Electoral Systems and Party Performance
While the relationship between electoral system type and party system performance appears to point in the direction that more proportional systems are associated with higher levels of party system fragmentation, and possibly lower levels of volatility, the evidence presented so far is not overwhelming when viewed from the perspective of total variance explained. To investigate whether the linkage between electoral system and party performance is any stronger than that between electoral system and performance of the party system, we require indicators of the linkage between voters and individual political parties. For this, I rely on two variables: One is the extent to which class voting takes place; that is, the strength of the bivariate association between indicators of social class and party choice in a system.28 The other is political party membership in a country as a percentage of the electorate.29 Table 3 displays these variables by country. We again estimated correlation coecients between electoral system
(proportionality) and these party performance measures.
The results, shown in Table 4, are particularly strong with regard to the relationship of electoral system and party membership: countries with more proportional electoral systems also have signi®cantly higher levels of party membership. Regardless of whether the east central European cases are included in the analysis, the relationship holds and displays statistically signi®cant coecients (Pearson's r ˆ 0.5). The relationship between strength of class voting and electoral system type also points to stronger class voting in more proportional systems. (As an aside, it is noteworthy that class voting and party membership ®gures are positively correlated.30) Thus, the results suggest that electoral system type is strongly associated with party performance. Where there
26
The somewhat lower levels of signi®cance reported in some of the analyses may be due to the small number of cases analysed here. Thus, our interpretive strategy examines the direction and size of the correlation coecient ®rst, and its statistical signi®cance second.
27
The relationship between party system fragmentation and electoral volatility is virtually zero.
Thus, it would be dicult to conclude that systems with a greater e€ective number of parties also display higher levels of electoral volatility.
28 For details, see R. Dalton, Citizen Politics: Public Opinion and Political Parties in Advanced
Industrial Democracies. 2nd ed. (Chatham, NJ, Chatham House, 1996).
29
These data are taken from R. Katz, and P. Mair, Party Organisations: a Data Handbook on
Party Organisations in Western Democracies, 1960±1990 (London, Sage, 1992).
30 The Pearson correlation coecient is 0.33 (Western Europe) and 0.46 (all cases; p 5 0.1).

# Political Studies Association, 1998

CHRISTOPHER J. ANDERSON

581

TABLE 3. Strength of Class Voting and Party Membership in 19 Democracies
Country

Class votingd

Party membershipe

0.20
0.16
NA
0.05
0.21
0.16
0.15
0.13
NA
0.03
0.14
NA
0.18
0.12
0.11
0.04
0.15
0.16
0.18

21.80
9.20
NA
1.80
6.50
12.90
1.70
4.20
NA
2.50
5.30
NA
2.90
1.50
4.90
3.10
4.00
21.20
3.30

Austria
Belgium
Bulgaria
Czech Republic
Denmark
Finland
France
Germany
Greece
Hungary
Ireland
Luxembourg
Netherlands
Poland
Portugal
Slovakia
Spain
Sweden
United Kingdom

Sources: d P. Mair, Party System Change (Oxford, Clarendon Press, 1997), Table 8.2. e R. Dalton, `Political Cleavages, Issues, and Electoral Change', in L. LeDuc,
R. Niemi and P. Norris (eds), Comparing Democracies: Elections and Voting in
Â
Global Perspective (London, Sage, 1996), Table 13.1; and G. Toka, `Parties and
Electoral Choices in East-Central Europe', in G. Pridham and P. Lewis (eds),
Stabilising Fragile Democracies: Comparing New Party Systems in Southern and
Eastern Europe (London, Routledge, 1995), Table 4.4 and own calculations.
TABLE 4. Pearson Correlations of Electoral System, Party Membership, and Class Voting
Western Europe
Proportionality of electoral system
Party membership
Class voting
N:

0.53*
0.22
12

All cases
Proportionality of electoral system
0.55**
0.35
16

*p 5 0.10; **p 5 0.05.

is a more proportional translation of votes into seats, I also ®nd a closer link between parties and their voters. These results are consistent with those found for the relationship between electoral system and party system performance, indicating that proportional systems produce a greater number of parties, less electoral volatility, and stronger linkages between citizens and parties. What is not clear, however, is whether this is the case because parties in proportional
# Political Studies Association, 1998

582

Democratic Performance in The New Europe

systems are better able to target well-de®ned segments of the electoral market or because voters are more likely to ®nd parties that are close to them ideologically
(or both). In all likelihood, it is a combination of both, although these data do not allow us to address this question.
Party and Party System Performance: Is there a Link?
The theoretical models speci®ed above hypothesized that there should also be a link between individual party performance and party system performance.
Speci®cally, it postulated that party system volatility should be lower in systems that have strong organizational and mobilization linkages between individual parties and voters, whereas it is unclear what the relationship between fragmentation and party performance should be. Table 5 shows the associations between these conditions.
TABLE 5. Pearson Correlations of Party System Fragmentation, Electoral Volatility,
Class Voting, and Party Membership
Western Europe

All cases

Class voting
Party membership

E€ective no. of parl. parties
0.29
0.14

E€ective no. of parl. parties
0.10
0.10

Class voting
Party membership

Volatility
À0.37
À0.30

N:

12

Volatility
À0.71***
À0.48*
16

*p 5 0.10; **p 5 0.05; ***p 5 0.01.

The results indicate that there is a positive relationship between party system fragmentation (i.e., e€ective number of parties) and organizational linkages, and a negative one between electoral volatility and the organizational linkage variables. Thus, in systems that have higher levels of class voting and party membership, the party system also tends to have a greater number of parties.
Conversely, electoral volatility tends to be lower in systems with high levels of class voting and party membership.
Overall, these relationships are very consistent. In particular the link between class voting and volatility (both indicators of the strength of the connection between voters and parties) turns out to be extremely strong, whereas the association between party membership and fragmentation is weaker, and possibly also less obvious from a theoretical perspective. Countries with parties that are able to mobilize core electorates along the traditional cleavage-lines thus experience less volatility.
Satisfaction with Democratic Governance: Measures and Data
To assess current levels of system support, I rely on comparative survey data from the Eurobarometer averaged over the period of 1993±95 in the member
# Political Studies Association, 1998

CHRISTOPHER J. ANDERSON

583

states of the European Union.31 Moreover, for reasons of comparability I make use of the data collected as part of the Central and Eastern Eurobarometer
(CEEB) in a number of the emerging democracies of east central Europe during the same time period. For 1993 and 1994 the (western) Eurobarometer includes random national samples for Belgium, Denmark, France, (West) Germany,
Greece, Ireland, Luxembourg, the Netherlands, Portugal, Spain, and the
United Kingdom. For 1995, it also includes the new member states Austria,
Finland, and Sweden. The CEEB surveys include data from Bulgaria, the Czech
Republic, Hungary, Poland, and Slovakia.
System support is de®ned in a straightforward manner as citizens' satisfaction with democracy. Appropriately, the question wording for the system support measures di€ers slightly in Western and east central Europe in order to account for di€erent political realities in these environments. Thus, citizens in the member states of the European Union were asked: `On the whole, are you very satis®ed, fairly satis®ed, not very satis®ed or not at all satis®ed with the way democracy works in (our country)?' In central and eastern Europe, respondents were asked: `On the whole, are you very satis®ed, fairly satis®ed, not very satis®ed or not at all satis®ed with the way democracy is developing in (our country)?' System support is measured by the percentage of respondents in a country who indicate that they are very or fairly satis®ed with democracy.32
Satisfaction with democracy measures system support at a low level of abstraction. It does not refer to democracy as a set of norms, but to the functioning of the democratic political system.33 Thus, it gauges people's responses to the process of democratic governance34; that is, a country's `constitution in operation'35 or its `constitutional reality'.36, 37
31
I use averages here to smooth out any temporary ¯uctuations that may exist in the data. Note, however, that the substantive conclusions are unchanged when I employ year-by-year data.
32
Studies that have employed the democracy satisfaction measure as a dependent variable include C. J. Anderson, and C. A. Guillory, `Political institutions and satisfaction with democracy: a cross-national analysis of consensus and majoritarian systems', American Political Science Review,
91,1 (1997), 66±81. H. D. Clarke, N. Dutt, and A. Kornberg, `The political economy of attitudes toward polity and society in western European countries', Journal of Politics, 55 (1993), 998±1021.
R. Harmel, and J. Robertson, `Government stability and regime support: a cross-national analysis',
Journal of Politics, 48 (1986),1029±40.
33 For analyses of support for democratic norms in Eastern Europe and the former Soviet Union, see J. L. Gibson, `A mile wide but an inch deep (?): The structure of democratic commitments in the former USSR', American Journal of Political Science, 40,2 (1996), 396±420. G. Evans, and
S. White®eld, `The politics and economics of democratic commitment: support for democracy in transition societies', British Journal of Political Science, 25,4 (1995), 485±514.
34
See F. Weil, `The sources and structure of legitimation in western democracies'.
35
J.-E. Lane, and S. Ersson, Politics and Society in Western Europe, 2nd ed. (London, Sage,
1991) p. 194.
36 D. Fuchs, G. Guidorossi, and P. Svensson, `Support for the Democratic System', In
H.-D. Klingemann and Dieter Fuchs (eds) Citizens and the State (New York, Oxford University
Press, 1995). p. 328.
37
H. D. Clarke, and A. Kornberg, `Support for the Canadian Federal Progressive Conservative
Party since 1988: the impact of economic evaluations and economic issues', Canadian Journal of
Political Science, 25 (1992), p. 47, n. 24; and A. Kornberg, and H. D. Clarke, `Beliefs about democracy and satisfaction with democratic government: the Canadian case', Political Research
Quarterly, 47 (1994), 537±63, report on a variety of tests designed to establish construct validity of the satisfaction with democracy question as an indicator of system support. They ®nd that satisfaction with democracy is clearly an indicator of actual system support and not coterminous with support for the incumbent government. Similarly, Fuchs (D. Fuchs, `Trends of political support', in D. Berg-Schlosser and R. Rytlewski (eds), Political Culture in Germany (New York,

# Political Studies Association, 1998

584

Democratic Performance in The New Europe

Figure 1. Democracy Satisfaction in Europe
This dependent variable is also useful for reasons of research design with regard to the number of cases that can be analysed. The Eurobarometer data used here are among the few available sources that permit an analysis of mass political support across a meaningful number of contemporary democracies.
Moreover, they uniquely allow for comparative studies of attitudes that combine both established and emerging democracies. Figure 1 shows the distribution of responses across the countries included in this study, averaged over the
1993±95 period.
The graph shows clearly that democracy satisfaction in east central Europe is systematically lower than in the mature democracies of western and southern
Europe. This holds true at both ends of the distribution. The highest levels of democracy satisfaction in western Europe can be found in countries such as
Denmark, Luxembourg, and the Netherlands; they are about twice the level of the highest levels in central/eastern European states such as the Czech Republic and Poland. Conversely, the lowest levels of satisfaction in eastern Europe (e.g.,
Bulgaria) are roughly half that of the lowest levels in the west (e.g., Greece).
Support ranges from 70±80% in the highest western nations to about 10±15% in the lowest eastern ones.
Attitudes toward democracy should di€er somewhat across older and newer democratic systems because the emerging democratic systems have existed for only a fairly short period of time relative to the older democracies of the West.
Given that they have not been able to develop deep reservoirs of di€use system support, it should not be surprising that levels of satisfaction with democratic
St. Martin's), p. 242) examines the validity of the satisfaction with democracy indicator and ®nds that the results constitute `a successful validation of the indicator as a measuring instrument for a generalized attitude towards the political system on the legitimacy dimension'. Weil (`Structure and
Sources of Legitimation', pp.692±3) provides some indirect construct validity for the satisfaction with democracy indicator as well by reporting high positive correlations with political trust.
# Political Studies Association, 1998

CHRISTOPHER J. ANDERSON

585

performance are consistently lower in central and eastern Europe than in the
West.38
Electoral Institutions, Political Performance, and Democracy Satisfaction
Are electoral institutions, party performance, and party system performance related to levels of system support? That is, do di€erent kinds of electoral institutions and political performance lead to systematically di€erent levels of satisfaction with democratic performance, as hypothesized in Model 1?
Moreover, do institutions ± in the form of electoral laws ± have a stronger impact than recent performance, or do they contribute equally to variation in democracy satisfaction across the democracies studied here?
These questions are of obvious relevance to our understanding of how democracies function. Most immediately, if there is only a weak link between how electoral institutions and organizations perform and how citizens evaluate the system as a whole, the central role of elections and parties in advanced democracies may be called into question. Conversely, if there is a link, policy makers and those who write the rules of the electoral game may want to know what the direction of these relationships are. After all, low levels of citizen support can pose serious problems for democratic systems because both the functioning and the maintenance of democratic polities are intimately linked with what and how citizens think about democratic governance.39
This is particularly the case for systems that are undergoing the transition to a more democratic political system. In contrast to the more stable democracies of western Europe, questions of popular support for the regime are particularly important for emerging democracies in Latin America and central and eastern
Europe. In these systems citizen support for the political system is of practical and immediate relevance for the design and continued stability of democratic institutions.40 Moreover, the relationships among electoral institutions, party system performance, and democracy satisfaction also will shed light on the question concerning the e€ects of institutions versus performance on citizen attitudes toward the political system. Put simply, can we explain cross-national di€erences in system support with the help of institutions, performance, or both?
To answer the question posed above, I again conducted correlation analyses between the institutional and performance variables on one hand, and levels of democracy satisfaction on the other. Table 6 shows the results.
The results indicate that the strongest correlations exist between proportionality (electoral system) and democracy satisfaction: the more proportional the electoral system, the higher the level of satisfaction. I also ®nd that class voting is signi®cantly and positively associated with democracy satisfaction: the higher the level of class voting, the higher the level of democracy satisfaction. This relationship is most signi®cant when central/east European countries are
38
Â
G. Toka, `Political support in East-Central Europe', H.-D. Klingemann and D. Fuchs (eds),
Citizens and the State (New York, Oxford University Press, 1995). For an overview, see A. H. Miller,
V. L. Hesli, and W. M. Reisinger, `Conceptions of democracy among mass and elite in post-soviet societies', British Journal of Political Science, 27 (1997), 157±90.
39 S. M. Lipset, Political Man: the Social Bases of Politics (Garden City NY, Doubleday, 1959).
G. B. Powell, Contemporary Democracies. G. B. Powell, `Extremist parties and political turmoil: two puzzles', American Journal of Political Science, 30 (1986), 357±78.
40
W. Mishler and R. Rose, `Trust, distrust and skepticism: popular evaluations of civil and political institutions in post-communist societies', Journal of Politics, 59 (1997), 418±51.

# Political Studies Association, 1998

586

Democratic Performance in The New Europe

TABLE 6. Pearson Correlations of Democracy Satisfaction with Electoral System and
Party/Party System Performance
Western Europe
Electoral system

0.74**
[N ˆ 14]
0.38
[N ˆ 14]
À0.04
[N ˆ 14]
0.11
[N ˆ 12]
0.51*
[N ˆ 12]

E€ective no. of parties
Volatility
Party membership
Class voting

All cases
0.61***
[N ˆ 19]
0.31
[N ˆ 19]
À0.56**
[N ˆ 19]
0.34
[N ˆ 16]
0.79***
[N ˆ 16]

*p 5 0.10; **p 5 0.05; ***p 5 0.01.

included in the analysis. Thus, both institutions and party performance appear to have strong e€ects on democracy satisfaction across the countries included in this study.
Regarding the associations of the other conditions with democracy satisfaction, the results suggest that levels of electoral volatility and democracy satisfaction are negatively related. However, this is only the case when the new democracies of east central Europe are included in the analysis. Levels of democracy satisfaction are thus lower when levels of electoral instability are particularly high; conversely, in those countries where election results are fairly stable over time, levels of satisfaction tend to be higher. However, these correlations are not nearly as strong as those found for electoral system and the party performance variables and should be treated with some caution.
Party system fragmentation (that is, e€ective number of parties) is positively related to democracy satisfaction. This suggests that levels of democracy satisfaction are marginally higher in systems that have parties which target segments of the electoral market. Thus, the results found here support Miller and Listhaug's view (and argue against Weil's) that a greater number of parties is, in fact, positively correlated with measures of system support. Thus, a greater probability for any one voter to ®nd a political party that is fairly close to her/ his true preferences is associated with higher levels of system support.
Finally, party membership levels are positively associated with levels of democracy satisfaction; a higher percentage of citizens is active in political parties in countries that display higher levels of democracy satisfaction. Despite recent disenchantment with political parties across a number of European democracies, these results indicate that closer linkages between citizens and individual political parties ± the prime organizational vehicle for political representation in modern democracies ± are associated with positive results in terms of citizen satisfaction with political performance.
Conclusions and Implications
This essay examines the interrelationships among electoral institutions, party and party system performance, and citizens' support for democratic governance
# Political Studies Association, 1998

CHRISTOPHER J. ANDERSON

587

from a cross-national perspective. Starting with the assumption that electoral institutions condition the development and dynamics of party systems as well as parties themselves, and the notion that party and party system performance a€ect how citizens evaluate democratic performance, the analysis tested these relationships with recent data collected in 19 democracies of western, southern, and east central Europe.
On the basis of directly comparable survey evidence and data on party system performance collected between 1993 and 1995, this article examined the determinants of cross-national di€erences in democratic support in both old and new democracies; that is, the 15 member states of the European Union as well as several of the emerging democracies of east central Europe. I investigated whether and to what extent electoral rules and party system performance help us understand di€erences in levels of satisfaction with the political system in both old and new democracies.
Regarding the relationship of electoral institutions and party/party system performance, the empirical analysis indicated that more proportional electoral systems are associated with higher levels of party system fragmentation and a closer link between parties and their voters. Furthermore, systems with a greater number of parties have higher levels of class voting and party membership, the party system also tends to have a greater number of parties. Conversely, in systems with high levels of class voting and party membership, electoral volatility tends to be lower.
The empirical analysis also demonstrated that electoral institutions a€ect democracy satisfaction more than either party or party system performance.
Thus, institutional factors ± such as electoral system types ± are, in fact, better predictors of whether citizens are supportive of democratic governance than is current political performance of the party system. The results also showed that, among the performance measures, party performance (class voting) was the best indicator of di€erences in democracy satisfaction across contemporary democracies. Coupled with the negative relationship of electoral volatility on democracy satisfaction, we can thus conclude that systems where electoral dealignment has taken place or is taking place have lower levels of democracy satisfaction than those where voters' ties with political parties are relatively stable. Overall, these results show that di€erent attributes of democratic governance ± electoral systems along with party and party system performance ± move together in systematic and predictable ways. Moreover, these syndromes of democratic politics systematically a€ect how citizens perceive the functioning of democratic institutions. Perhaps most importantly, these relationships hold in both mature and emerging democracies, lending greater validity to our theories of how democracies function.
Finally, this essay extends an existing and developing research agenda that examines the relationships among political institutions and political/democratic performance on one hand, and citizens' attitudes about the political system on the other.41 Building on work in political economy and the comparative study of
41
C. J. Anderson, and C. A. Guillory, `Political institutions and satisfaction with democracy'.
P. Norris, `Designing democracies: Institutional arrangements and system support', Paper presented at the JFK School of Government Workshop on Con®dence in Democratic Institutions: America in
Comparative Perspective, 25±27 August 1997, Washington D.C.

# Political Studies Association, 1998

588

Democratic Performance in The New Europe

political institutions, this line of research has only recently begun to systematically investigate the interrelations of institutions, performance ± both political and economic ± and system support. This analysis has shown that electoral institutions in particular are an important factor to consider in this context.
While the relationship between party system performance and democracy satisfaction was more ambiguous, this analysis should be taken to be more suggestive than de®nitive, and further research into this web of interrelations is clearly warranted in order to gain a more complete understanding of the functioning of modern democracies.
Appendix: Descriptive Statistics of the Variables Used in the Analysis
Minimum
systema

Proportionality of electoral
E€ective no. of parl. partiesb
Volatilityc
Party membership
(as % of electorate)d
Class votinge
Democracy satisfactionf

Maximum

Mean

Std. Dev.

N

79.11
2.17
5.10
1.50

97.83
7.95
27.60
21.80

91.00
3.75
14.23
6.68

6.47
1.40
7.60
6.51

19
19
19
16

0.03
12.33

0.21
80.40

0.14
48.57

0.05
18.32

16
19

Sources:
A. Lijphart, Electoral Systems and Party Systems: a Study of Twenty-Seven Democracies
(New York, Oxford University Press, 1994), Appendix B; own calculations on the basis of data reported in European Journal of Political Research (various issues). b L. LeDuc, R. Niemi, and P. Norris, Comparing Democracies: Elections and Voting in
Global Perspective (London, Sage, 1996), Table 1.4. Most recent election only. c P. Mair, Party System Change (Oxford, Clarendon, 1997), Table 8.1; own calculations on the basis of data reported in European Journal of Political Research (various issues). d P. Mair, Party System Change (Oxford, Clarendon, 1997), Table 8.2. e Â
Dalton (1996, Table 13.1) and G. Toka, `Parties and Electoral Choices in East-Central
Europe', in G. Pridham and P. Lewis (eds), Stabilising Fragile Democracies: Comparing
New Party Systems in Southern and Eastern Europe (London, Routledge, 1995), Table 4.4 and own calculations. fEurobarometers (39±44) and Central and Eastern Eurobarometers (4±6). a # Political Studies Association, 1998