INTERNATIONAL POLICY ANALYSIS

The Voters Choosy Heart Beats Left of Centre

Satisfaction with the Government and Attitudes of Social Democratic Voters in European Comparison, 20022008
BERNHARD WESSELS May 2012
How satisfied are social democratic voters with the performance of their governments? Furthermore, do Social Democratic voters still differ at all from other voters with regard to their political objectives and orientations? These are the two overarching questions considered in this contrast analysis, based on data from the European Social Survey. Comparing the voters of social democratic and conservative parties we find that social democratic voters punish their governments more severely with expressions of dissatisfaction if they are not satisfied with policy outcomes. Moreover, they reward their governments less with expressions of satisfaction. Regardless of a governments performance the duration of a period of government is alone sufficient to bring about a slight fall in voter popularity and an increase in the number of those who abstain from voting. Interestingly, this erosion by voter fatigue is much more pronounced among social democratic voters and their governments than among conservatives. Social democratic and conservative voters differ clearly in terms of their political goals and orientations. Surprisingly, the biggest differences between social democratic and conservative voters concern tolerance and integration: in other words, attitudes about lifestyles, immigrants, migration and European integration and not traditional issues of the left like social differentiation and inequality.

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Content
1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Government Performance and Satisfaction with the Government . . . . . . . . . . . . . 3. Its the Economy, Stupid The State of Things and Satisfaction with the Government . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Voter fatigue and Voter Behaviour . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 3

6 8

5. Is There a Typical Social Democratic Voter? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 6. Summary and Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

1. Introduction
For what reasons do parties win or lose elections in Europe? The answers to this important for both election winners and election losers question are generally extremely involved and complex. However, simple answers sometimes have considerable explanatory power. One example of this is the obvious response that the losers political achievements were unconvincing. Throwing the rascals out, as they say in the United States, tends to be the result of a desire to punish a government perceived to have performed poorly. In recent years, social democratic parties have often failed to be returned to office. For example, government participation or leadership was lost in Germany in 2009 and in the United Kingdom, Slovakia, the Czech Republic and Cyprus in 2010, to name only a few. Can a negative trend be discerned here? Are social democrats losing achievement potential and profile in Europe? The question applies to parties as well as to party supporters and people voting for social democratic parties in Europe. Power and profile for political parties mean, first, effective and competent policymaking and achieving corresponding results and, second, substantive differences in comparison to the other parties. As far as supporters and voters are concerned, achievement potential can only mean that they express their own convictions in private and public debates. However, that requires that social democratic voters political profile differs, in particular from that of conservative party voters. This paper raises questions at both levels. With regard to achievement potential and performance we first ask how well social democratic parties in government do with the voters in comparison to their right-of-centre rivals, and how their performance assessment translates into electoral success. The second question concerns whether governments tend to become exhausted the longer they remain in office; or rather, whether voters tend to tire of governments the longer they remain in office and thus lose support. Finding an answer to this question is thus a matter of some significance because it provides reasons for lost votes besides poor or strong performance. Both questions relate to how social democratic governments are judged by the voters and what the consequences are. Turning to political profile, we ask whether social democratic voters differ from their conservative counterparts

in terms of their political objectives and values. There are at least two reasons why this question is worth pursuing. On the one hand, given increasing voter volatility, the diminishing influence of socio-structural characteristics when it comes to deciding who to vote for and the end of ideology in politics, doubt is increasingly being raised concerning whether distinct voter profiles still exist differentiating between the so-called national or catch-all parties. On the other hand, political parties and social democratic parties are by no means an exception are in competition for the political centre-ground, the socalled median voter. But contesting the centre-ground makes sense only if there is no longer much of a difference between social democratic voters and their major right-of-centre counterparts regarding their ideas about society and politics. The questions are posed in both dimensions the performance of social democratic governments and the political profile of social democratic voters using the data of the European Social Survey (europeansocialsurvey.org), an instrument for observing European societies created by social scientists for social science purposes. The European Social Survey commenced in 2002 and is conducted every two years. To date, there have been four waves, the latest in 2008. Here we make use of the cumulative data of all four survey waves.1 A total of 24 European countries can be examined. The European Social Survey was not conducted in all of them at all four points in time. Altogether, a total of 85 surveys can be used for analysis. Table A1 provides an overview.

2. Government Performance and Satisfaction with the Government

It is not only intuitively plausible that performing well in government is important for getting re-elected, but it has also been frequently demonstrated. V.O. Key as early as 1961 pointed out the two aspects of peoples assessment of governments that can lead to re-election or being
1. The cumulative data record brings together the following waves for all comparable issues: ESS Round 1: European Social Survey Round 1 Data (2002). Data file edition 6.2. Norwegian Social Science Data Services, Norway Data Archive and distributor of ESS data. ESS Round 2: European Social Survey Round 2 Data (2004). Data file edition 3.2. Norwegian Social Science Data Services, Norway Data Archive and distributor of ESS data. ESS Round 3: European Social Survey Round 3 Data (2006). Data file edition 3.3. Norwegian Social Science Data Services, Norway Data Archive and distributor of ESS data. ESS Round 4: European Social Survey Round 4 Data (2008). Data file edition 4.0. Norwegian Social Science Data Services, Norway Data Archive and distributor of ESS data.

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

voted out: the nature and direction of policy and policy outcomes (Key 1961: 474). The prescription throw the rascals out characterises this mechanism precisely: dissatisfaction with a governments achievements leads to a party being voted out, as Miller and Wattenberg were able to demonstrate in a long-term study. For parties in office, retrospective performance assessment plays a central role (Miller and Wattenberg 1985), often together with performance expectations, i.e. prospective evaluations. Performance and competence are the two dimensions that determine whether governments are voted out and which ones are voted in again. That can also be shown in Bundestag elections (Weels 2002). The comparative surveys of the European Social Survey cover three social domains: assessment of the state of the health care system; assessment of the state of the education system; and satisfaction with the economic situation. Two key policy areas are thus addressed which are at the heart of social democratic party policies, while the economic situation is equally important for all citizens and parties. Without a benchmark, however, it makes little sense to try to determine how satisfied voters are with social democratic parties in government on the basis of citizens assessments of the state of social domains. On the one hand, the problem arises that conditions in all areas of society, but particularly the economy, cannot be entirely determined politically, even if parties and the political sphere frequently give the impression of omnipotence and voters assume that politics is all-powerful. A grim economic situation and problems in the education and health care systems in many respects cannot be traced back to bad policy as ultimate cause. Evaluations made at different points in time are therefore difficult to compare and it is impossible to determine whether, for example, conservative governments are more capable on the basis of such judgements. Direct comparison of social democratic and other governments should therefore be eschewed and instead assessments of conditions in the areas of health care, education and the economy by social democratic and conservative voters should be contrasted. This contrastive comparison should be carried out for social democratic governments or government participation without conservative party involvement; for governments made up of social democratic and conservative parties; and for conservative

governments or government participation without social democratic party involvement.2 The state of the health care system, the education system and the economy during the period 20022008 in the 24 countries considered here is assessed differently if the government was social democratic without conservative participation: 35 per cent more social democratic voters are satisfied with the state of the health care system than are dissatisfied; 9 per cent with regard to the education system. When it comes to the economic situation, the proportion of dissatisfied persons is 17 percentage points higher than that of satisfied persons. These values can be compared to the evaluations made by conservative voters. It is clear from this that social democratic voters are relatively more satisfied with the state of the health care system, the education system and the economy. Assessments of the situation are therefore not objective, but the result of a specific party perspective: voters are biased, in other words. Social democratic voters clearly evaluate the same situation in the three areas under a social democratic government more positively than those who voted for their major conservative rivals. Conversely, conservative voters under a conservative government without social democratic participation evaluate the situation as much better than the voters of social democratic parties (see Figure 2, second panel). However, the differences in the assessments here are less pronounced and thus the bias is smaller. But it may not be concluded from this that, by comparison, conservative governments would do better: the assessments relate to different points in time and overwhelmingly to other countries.4 It is of interest not only politically, but also in terms of electoral strategy that social democratic voters evaluate the situation in the areas of health care, education and the economy much less positively than conservative voters when social democrats and conservatives are in gov2. Depending on the time of the survey, information on the relevant government (composition) was added to the surveys of the European Social Survey. 3. There were such governments in the following countries (see Table A2 for abbreviations and time points) CY, CZ, DE, ES, GB, GR, HU, NO, PL, PT, SE, SI and SK. 4. Conservative governments in question in AT, BE, DK, EE, ES, FI, FR, GR, IE, IT, LU, NL, NO, PL, PT, SE, SI, SK. Regarding specific time points and abbreviations see Table A2.

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Figure 1: Assessment of the state of the health care system, the education system and the economy by social democratic and conservative voters under social democratic and conservative governments

Social democrats in government without conservatives

Conservative voters social democratic voters education 1% 9% 5% 17% 32% 40% 30% 20% 10% 0% 10% 20% 30% 40% 50% 5%

Health care

economy 50%

% more dissatisfied/very bad than satisfied

% more satisfied/very good than dissatisfied

Conservatives in government without social democrats

Health care Conservative voters social democratic voters 9% 15% 31% 6% 40% 30% 20% 10% 0% 10% 20% 30% 40% 50% 14% 17%

education

economy 50%

% more dissatisfied/very bad than satisfied

% more satisfied/very good than dissatisfied

Social democrats and conservatives in government together

Conservative voters social democratic voters education 1% 16% 6% 9% 40% 30% 20% 10% 0% 10% 20% 30% 40% 50% 4% 15%

Health care

economy 50%

% more dissatisfied/very bad than satisfied

% more satisfied/very good than dissatisfied

The percentages present the difference between those who make a positive assessment minus those making a negative assessment. Based on assessment scales from 0 (very bad) to 10 (very good), summarised as follows: 0, 1, 2, 3 = 1; 4, 5, 6 = 0; 7, 8, 9, 10 = +1.

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

ernment together.5It is often the case that, in the eyes of their voters, (genuinely) grand coalitions pay off better for conservative parties than for social democratic ones.

3. Its the Economy, Stupid The State of Things and Satisfaction with the Government
What is the effect of assessments of the state of the health care and education systems and the economy on satisfaction with the government of the day? Generally, it should be noted that neither social democratic nor conservative voters are particularly satisfied with their respective governments. When social democrats are in power without their biggest conservative rivals the judgement of social democratic voters is slightly negative and that of conservative voters strongly negative. When conservatives are in power without social democratic parties a small majority of voters are fairly satisfied, while social democrats are very dissatisfied. When social democrats and conservatives are in power together the majority of neither electorate is satisfied and social democrats are even significantly more dissatisfied than those who voted for the conservative coalition party (see Figure 2). Overall, there is quite a wide range of variation in satisfaction with the government (see Figure A2 in the Appendix). However, that does nothing to change the general impression that social democratic voters are more critical of their governments than conservative voters are. In the 13 countries in which social democrats govern without conservatives, among all voters there was only one response tending towards satisfaction, in Cyprus. When account is taken of the assessment of social democratic voters it was five countries: Cyprus, Spain, Norway, Sweden and Slovakia. For conservative governments the result is more positive. When they rule without social democrats which is the case in 17 countries there are four in which satisfaction outweighs dissatisfaction among all voters; among conservative voters that applies to 12 of these 17 countries. However, such comparisons must be interpreted with caution: different countries or the same countries at different time points can be responsible for significant differences with regard to problem issues and therefore for diverging satisfaction with the government.
5. Coalition governments are concerned in the following countries: AT, BE, CH, CZ, DE, EE, FI, NL, SI. See Table A2 for details concerning time points.

Looking at governments in which both social democrats and conservatives participated, government satisfaction with social democrats and conservatives is perfectly balanced. Taking into consideration the countries in which social democrats have ruled for a period without conservatives and in another period conservatives have ruled without social democrats which was the case in Spain, Norway, Portugal, Slovenia and Slovakia, as well as in Greece, Poland and Sweden social democratic governments come off better than conservative governments among voters overall in Cyprus, Spain, Norway, Sweden and Slovakia. What is principally responsible for satisfaction or dissatisfaction with the government? The phrase Its the economy, stupid comes from Bill Clintons election campaign against George Bush Senior in the US presidential election in 1992. The meaning is apparent from the context: elections are to be won primarily over economic issues at least in the United States. But in Germany, too, the state of the economy is an important factor in voter decision-making. German electoral researchers also use the phrase to provide an explanatory framework for voters behaviour (for example, Walz/Brunner 1998). As a rule, peoples assessment of the general rather than their personal economic situation is more significant here. The former is described as sociotropic, the latter as egocentric voting. Egocentric voting, also called pocket-book voting, because it is based on ones personal financial benefit, plays a less important role (Maier/ Rattinger 2004). Empirical testing of whether and to what extent the economy really has decisive significance for satisfaction with the government, can be done by determining the relative contribution of each of the three factors state of the health care system, state of the education system and state of the economy make to overall satisfaction with the government.6 The results show the following:

6. This involves the static procedure of regression analysis. Here satisfaction with the government is regressed (or traced back) in a model to the assessment of the state of affairs in the three domains. As in the case of the preceding analyses across countries and time points the data are design- and population-weighted in order to avoid distortions of the results by sampling errors and different sample sizes, as well as calculated using standard error correction, corresponding to the cluster-like data.

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Figure 2: Satisfaction of social democratic and conservative voters with the government under social democratic and conservative governments

7% Social democrats without conservative participation 54%

Social democratic voters Conservative voters

45% Conservatives without social democratic participation 11%

16% Coalition of social democrats and conservatives 0%

75%

50%

25%

0%

25%

50%

75%

% more satisfied than dissatisfied

% more dissatisfied than satisfied

The model regresses satisfaction with the government on the state of the health care system, the education system, and the economy. The model explains, with regard to the electorate as a whole, around 40 per cent of the variation in satisfaction with the government. That is a fairly decent figure for a model at the level of individual respondents and only three explanatory variables. It underlines that people in fact hold their governments responsible to a considerable extent for the state of affairs in the three domains. Among social democratic voters the model even explains one more percentage point of variance, while among conservative voters it is six percentage points less. All in all, the three assessments of how things stand explain satisfaction with the government at the individual level statistically more than satisfactorily. The question of which of the three assessments of how things stand is key to satisfaction with the government can be answered very clearly: It really is the economy. Evaluations of neither the state of the health care system nor the education system impairs or improves satisfaction with the government to the same degree as evaluations

of the economy. For every point that a persons assessment of the economic situation increases on an 11-point scale (from 0 very bad to 10 very good) satisfaction with the government improves by around 5 percentage points, while a better assessment of the health care or the education system yields only between 0.6 and 2 percentage points (see Table 1). There is an interesting difference between social democratic and conservative voters with regard to the importance of the two systems for satisfaction with the government: for social democratic voters the two are of approximately equal importance, while for conservative voters the health care system is around three times more important for satisfaction with the government than the education system. However, neither finding comes anywhere near the importance of the economy. The finding of the importance of the assessment of the economic situation can be examined a little further. If governments are examined at particular time points with regard to peoples assessments of the economic situation and satisfaction with the government the result is

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Table 1: Assessment of the state of the health care system, the education system and the economy, and satisfaction with the government
Satisfaction with the government increases by percentage points if the assessment of the state of the improves by one point on the scale: Social democratic voters, social democratic gov- Conservative voters, conservative government ernment without conservative participation without social democratic participation State of the health care system State of the education system State of the economy 1.4 1.2 5.1 2.0 0.6 4.5

Improvement in satisfaction with the government from the worst to the best assessment in percentage points State of the health care system State of the education system State of the economy 13.6 12.4 50.8 19.6 6.1 45.4

Results of a regression analysis; proportion of explained variance 41 per cent for social democratic voters and 34 per cent for conservative voters. Design- and population-weighted; all coefficients are statistically significant (clustered standard errors).

unambiguous for governments in a country at a particular point in time: the better the economic situation, the higher the higher government satisfaction (Figure 3). Once again, we find confirmation of the influence of party orientation on peoples standpoint: regardless of the fact that the general economic situation is objectively the same for both social democratic and conservative voters, social democratic voters under a social democratic government without conservative participation judge the economic situation to be better and thus are also more satisfied with their government (Figure 3A). Conversely, conservative voters under a conservative government without social democratic participation evaluate the economic situation more positively than social democratic voters at the same time point and, accordingly, conservative voters are more satisfied with a government that is really theirs (Figure 3B). The party electorate-specific difference is even more marked under conservative governments than under social democratic ones. Furthermore, social democratic voters punish their governments more with their dissatisfaction than conservatives do their own, and reward them less with satisfaction if the economic situation is satisfactory. Among social democratic voters who are dissatisfied with the economic situation the proportion of those dissatisfied with their government is 41 percentage points higher than that of those who are satisfied. Among conservative

voters and a conservative government those dissatisfied with the government exceed their counterparts by only 22 percentage points (Table 2). Conversely, those who are satisfied with the government in the case of a satisfactory economic situation exceed the dissatisfied among social democratic voters by 56 percentage points and among conservatives by 65 percentage points. The same pattern is evident with regard to assessment of the economic situation and satisfaction with the government. This result confirms that social democratic parties benefit less among their voters in a coalition than their conservative partners in government among theirs (Table 2).

4. Voter Fatigue and Voter Behaviour

Dissatisfaction with their performance is not the only reason parties in government lose votes. There is much to suggest that people grow tired of governments, regardless of their performance, simply because they have been in power so long. The notion that governments somehow grow stale is misleading. Strictly speaking, it is rather the favour of the voters that fades over time. The US electoral researcher Philip E. Converse, for example, has shown that the regional stability of voter behaviour decays over time (Converse 1969): the same phenomenon can be observed in Germany (Weels 1998). The

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Figure 3: Assessment of the economic situation and satisfaction with the government
A. Social democratic government, no coalition with conservatives
10 9 Satisfaction with the government (0: Completely dissatisfied; 10: Completely satisfied) 8 7 6 5 4 3 2 1 0 0 1 2 3 4 5 6 7 8 9 10 Satisfaction with economic development (0: Completely dissatisfied; 10: Completely satisfied) y = 0.65x + 0.91 R = 0.50 y = 0.71x + 1.65 R = 0.68 Social democratic voters Conservative voters Linear (social democratic voters) Linear (conservative voters)

B. Conservative government without a coalition with social democrats

10 9 Satisfaction with the government (0: Completely dissatisfied; 10: Completely satisfied) 8 7 6 5 4 3 2 1 0 0 1 2 3 4 5 6 7 8 9 10 Satisfaction with economic development (0: Completely dissatisfied; 10: Completely satisfied) y = 0.57x + 1.09 R = 0.61 y = 0.70x + 2.63 R = 0.41 Social democratic voters Conservative voters Linear (social democratic voters) Linear (conservative voters)

Mean values for governments at the relevant time point.

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

passage of time therefore makes a difference. Something similar can be observed and read about in the public debate, in the media and in election campaigns. People say that its time for a change or talk about a sea change (in Germany) Kanzlermdigkeit (chancellor fatigue) and so on, all of which refer to the time aspect, besides other issues that might be raised. Even in a very parsimonious forecast model of the outcome of parliamentary elections how long a government has been in power plays a role, as Gschwend and Norpoth have shown (Gschwend and Norpoth 2005). Does voter support fade, therefore, the longer a party is in government? In order to answer this question, for each survey time point in each country the length of time the government had been in power at the last election was determined. The date of the current survey therefore was not at issue because the question was posed retrospectively in relation to the previous election. In a first step, whether satisfaction with the government and its performance declines more sharply the longer it remains in office was examined. In a second step, voter behaviour was investigated depending on the length of a governments period of office. This resulted in clear and statistically significant effects. They can therefore be

related to the mere passage of time because a correlation with government performance could not be established. It was also examined whether there are party voter- and government party-specific effects. The calculated model determines the extent to which casting a vote for party can be explained by the length of time this party has been in government. In a further step, whether abstention can be explained by the length of time a party has been in office was examined. In every model for which the results are presented in what follows the effects of length of time in government are statistically significant. The results show that both social democratic and conservative parties slowly but surely lose electoral support the longer they are in office. Although the decline in voter support from one year to the next is very modest and, given the usual duration of a parliamentary term around four years appears insignificant, small margins often determine whether a government is able to remain in office. For example, while in the case of social democrats a 5 percentage point loss in voter support takes only 12 years, conservative parties do not register such a figure even after 20 years in power. But even a loss of 2 percentage points is enough to banish a party from the government to the opposition benches: and in the case

Table 2: Economic situation and satisfaction with the government in various government constellations
Satisfaction with the economic situation Dissatisfied So so Satisfied

Proportion of those satisfied with the government in percentage point Social democratic voters Social democrats in government without conservatives Coalition of social democrats and conservatives Conservative government without social democrats Conservative voters Social democrats in government without conservatives Coalition of social democrats and conservatives Conservative government without social democrats Social democratic voters Difference social democratic/conservative government Conservative voters Difference conservative/social democratic government 46.9 64.0 26.4 52.3 68.6 41.5 21.7 32.7 3.5 14.7 0.5 47.8 64.5 41.1 54.1 67.5 1.1 18.0 27.4 55.5 35.2 3.2

Scale of satisfaction with the economy: 0 completely dissatisfied; 10 completely satisfied. Values summarised as follows: 0, 1, 2, 3 = 1; 4 5 6 = 0; 7, 8, 9, 10 = +1. The values in the table show the extent to which dissatisfaction outweighs satisfaction (negative values) or vice versa (positive values).

10

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

of social democrats governing without conservatives this happens after only around six years in office. What is particularly significant about these results, however, is that social democratic governments without conservative participation lose their shine much more rapidly than conservative governments without social democratic participation (see Figure 4). When social democrats and conservatives are in government together the difference between the two in terms of voter fatigue disappears almost completely. The progression of vote loss for conservatives is in that case similar to that of social democrats in government without conservative participation. With regard to coalitions of this kind, what happens is similar to what happens to social democratic governments in Figure 4. To some extent, these negative developments arising from fatigue can be mitigated by means of relative stability in party ranks. Various models have investigated the extent to which overly rapid turnover in that respect accelerates the decline in support. This shows that continuity in government can mitigate the progression of voter fatigue: the effect of a long period in office can consider-

ably compensate the effect of voter fatigue among social democrats, while among conservatives continuity in party ranks in government can even overcompensate, when the effect is measured in terms of annual change. However, these results are not particularly stable and robust in comparison to assessments of fatigue effects. The results indicate that a combination of a long period in government and prolonged continuity in terms of personnel produces less fatigue than a long period in government with frequent personnel changes. What happens to voters who become estranged from their parties due to fatigue? Length of term of office appears to cause people to abstain from voting. If abstention is regressed onto length of period in government, equally strong effects result for social democratic and conservative government participation (Figure 5). Although it cannot be concluded from this with certainty that voters gravitate towards the abstention camp as the period in government lengthens, for various reasons it is not implausible. One reason, for example, is that a long term of office can be attributed to asymmetry and dominance of a particular political party in a political system and thus an important motivation for voting is lacking: if

Figure 4: Duration of continuous government and development of share of the vote in the case of social democratic and conservative parties
0%
.4% 0

.1%

.2%

.7%

0 .7

.9%

1% Fall in the share of the vote (regression) 2% 3% 4% 5% 6% 7% 8% 9% 10% 0.01 0.25 0.5

.5%

.2%

.5%

.4%

.6%

.9%

.3%

2 .1

.0%

3 .7

.5%

.2%

.0%

.7%

Social democrat

Conservative

10

12

14

16

18

.5%

20

Successive years of government participation

Estimated values from regression analysis. The differences in the fall in the share of the vote between social democratic and conservative governments are statistically significant.

11

.4%

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

there is a tight race people feel that their votes actually count for something, at least psychologically (Feld and Kirchgssner 1998). That would also be consistent with the result that this phenomenon occurs fairly evenly in the case of both social democratic and conservative government participation. What are the ramifications of the results? On the one hand, they show that voters may drift away from a party regardless of its performance in government, probably, but not exclusively into the abstainers camp. In a medium-term perspective, this means a heightening of the danger of government participation and in principle points to the cyclical development of voter support for parties. Parties can counter this only by making a special effort to mobilise support but with no guarantee of success. For social democratic parties the trend towards fading support is much stronger than for their conservative rivals. It is hard to say why that should be the case. But this finding is consistent with findings with regard to satisfaction or dissatisfaction with the economic situation in the country and how this converts into satisfaction or dissatisfaction with the government: social democratic voters appear to punish their parties in government more and to reward them less than conservative voters do.

5. Is There a Typical Social Democratic Voter?

The results on differences in the assessment of the same objective situation in the health care system, the education system and the economic situation among social democratic and conservative voters, depending on government participation, point to a political coloration of peoples judgement. Known as partisan bias in electoral research, this phenomenon has already been observed by the authors of American Voters, a classic study in this field, and has since been established for different time points, countries and political systems (Campbell et al. 1960). This also applies and in particular to economic perceptions and evaluations (Fiorina 1981; Bartels 2002). Different formulations have been worked out for what this means in substantive terms, each with different implications. One perspective is that attachment to a party filters perceptions, enabling people to see only what fits in with their political orientation and that they only accept a reality in line with their party orientation. The explanation here is based entirely on the assumption of cognitive psychology that people have a need to avoid inconsistency or cognitive dissonance. According to another perspective, differences in perception derive from differences in interests, values and general political orientation. On this understanding, bias is not a dis-

Figure 5: Duration of continuous rule and development of the proportion of abstainers in the case of social democratic and conservative government participation
0%
0 .2% 0 .3% .6% 0 0 % 0 .8

1% 2% Fall in turnout (regression) 3% 4% 5% 6% 7% 8% 9% 10% 0.01

.4%

.2% 1

.3%

.7%

.7%

.3%

.1%

.9%

.5%

.9%

.5%

3 .3

.1%

.8%

.6%

.2%

4 .2

Social democratic government participation

Conservative government participation

0.25

0.5

10

12

14

16

18

20

Successive years of government participation

12

.8%

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

tortion of reality to establish cognitive consonance but rather a filter through which reality is viewed. In that case, the filter is virtually a yardstick against which reality is measured. If this is to explain the established differences with regard to assessment of the same situation between voters of different parties these voters have to be differentiated in terms of their values and goals. Is there, therefore, a substantive social democratic voter profile that can be distinguished from that of conservative voters? The European Social Survey measures a whole series of social orientations and values, yielding clear implications with regard to policy goals. Overall, there are around 40 indicators, which can be categorised in terms of the following dimensions: tolerance and integration, domestic security, social conservatism, individual autonomy, hedonistic orientations, environmental protection and equality and social differentiation. In order to ascertain whether and in what way social democratic voters are distinguished from conservative voters the indicators were all considered in terms of whether they systematically and consistently produce a relevant difference between electorates. Two results were obtained by means of this first step. First, there are systematic differences, but second, not in all areas. The following results take into account only the dimensions that indicate systematic and significant differences between party electorates. There are differences in the dimensions equality and social differentiation, tolerance and integration and domestic security. There are no systematic and relevant differences in questions on hedonistic lifestyle (for example, it is important to have a good time), individual autonomy (for example, it is important to have new ideas and to be creative), social conservatism7 (for example, it is important to follow rules) and environmental protection. The dimensions that generate the sharpest differences (equality and differentiation, tolerance and integration) are at the core of social democratic politics. The results on domestic security can be rapidly summarised. In contrast to social democratic voters conservatives incline towards
7. A more than 5 percentage point greater difference can be observed only on the question of whether it is important to maintain traditions (17 to 23 per cent). However, even among conservative voters only 23 per cent consider that important.

tough law and order measures to a statistically significant degree. Concerning the infringements of the rule of law within the framework of the fight against terrorism asked about here, however, there is no majority among either social democrats or conservatives. Two sets of questions were available on the dimension of tolerance and integration. The results are presented in Figures 6 and 7. Social democratic voters are much more in favour of gays and lesbians being able to pursue their own lifestyles than conservative voters. Although in Germany this is probably not a significant issue the same cannot be said for certain Central and Eastern European countries. Furthermore, significantly more social democrats than conservatives, regardless of ethnic-cultural background, would allow many to come here and live (Figure 6). It is therefore not surprising that, in comparison to the conservative electorate, a much greater preponderance of social democratic voters operate on the assumption that immigrants enrich cultural life; a clear majority take the view that immigrants are good for the economy; and a smaller proportion but still a majority consider that immigrants make Germany a better place to live in. Consistent with this is the fact that a clear majority of social democratic voters would like to see European unification proceed further, while among conservatives around half favour it, while the other half are against it (Figure 7). What is the attitude of social democrats to social differentiation and inequality? Given the fact that this is one of, if not the core areas of social democratic politics it is not surprising that social democratic voters consider factors that generate social differentiation it is important to be rich; it is important to be successful less important. However, the differences in relation to conservative voters are quite small. More marked are the differences concerning approval of the importance of equal treatment and equal opportunities, as well as the role of government in reducing income inequalities (Figure 8). However, this analysis conceals differences that sometimes are more pronounced than the results suggest. It also conceals different patterns with regard to the differences between social democratic and conservative voters in the countries under consideration. In order to investigate the question of patterns of differences and country-specific social democratic profiles,

13

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Figure 6: Differences with regard to questions concerning tolerance and integration between social democratic and conservative voters
40% 35% 30% 25% 20% 15% 10% 5% 0%
Gays and lesbians should be able to live how they choose Immigrants of the same ethnicity should be allowed Immigrants of different ethnicities should be allowed Immigrants from poor countries should be allowed

35%

Social democratic voters Conservative voters

23% 19% 16% 13% 8% 12% 7%

Figure 7: Differences with regard to questions concerning immigration and European integration between social democratic and conservative voters
(Percentage point difference) More in favour than against

30% 25% 20% 15%

27%

Social democratic voters Conservative voters

10% 10%

10% 6%

5% 0% 5% 10% 15%

1%

1%

More against than in favour

4%

13% Immigrants enrich cultural life European unification should be taken further Immigrants are good for the economy Immigrants make Germany a better place to live in

Answers were given on an 11-point scale, with low values indicating rejection and high values approval. Summarised, the values were 0, 1, 2, 3 = 1; 4 5 6 = 0; 7, 8, 9, 10 = +1. A negative mean over these summarised values indicates the proportion by which negative opinions outweigh positive opinions and vice versa.

14

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Figure 8: Differences with regard to questions concerning social differentiation and inequality between social democratic and conservative voters
40% Democratic voters 35% 30% 25% 20% 15% 10% 5% 0% 10% 19% Conservative voters 30% 34% 30% 29%

37%

11%

It is important to be rich

It is important to be successful

Governments should reduce Equal treatment and income inequalities equal opportunities important

Proportion of those who strongly agree among the responses: Strongly agree, Agree, Neither agree nor disagree, Disagree, Strongly disagree.

questions about the dimensions addressed here were all examined at the level of percentage point differences with regard to approval or rejection between social democrats and conservatives and the three questions were chosen that produce the biggest difference in the respective country. The following areas of attitudes are the ones that produce the biggest difference between social democratic and conservative voter in one of the 24 countries: Social differentiation and inequality, Tolerance and integration or Social conservatism. Attitudes within the framework of social conservatism, however, produce one of the three most important differences in only three countries, Belgium, Hungary and Poland. Attitudes concerning tolerance and integration are among those that indicate the biggest difference between social democratic and conservative voters in 22 out of the 24 countries, and attitudes on social differentiation and inequality in 15 out of 24. In Table 3 the dimensions are presented with different colours: grey for social differentiation and inequality, light grey for tolerance and integration and dark grey for social conservatism. The results indicate that it is not only the traditional core of social democratic goals that distinguishes social

democratic voters from conservatives. Peoples orientation with regard to questions concerning tolerance and integration, in particular in the context of increasing immigration, also point to specific social democratic profiles. Among traditional left-wing aims, the expectation that the government should reduce income inequalities is prominent. This classic distribution-related orientation plays a role in distinguishing conservative voters only in Denmark. In Switzerland, Germany, Estonia, Spain, the United Kingdom, Ireland and Slovenia only the dimension of tolerance and integration supplies the three questions that differentiate most sharply between social democrats and conservatives. In the remaining countries it is always a combination of social differentiation and inequality and of tolerance and integration, except for Belgium, Hungary and Portugal, where social conservatism plays a role. These results make clear that the orientations and thus assessment criteria of social democratic voters are clearly distinguishable. Their view of the world differs from that of conservatives and thus generates different judgements concerning the state of society. The fact that newer questions concerning inequality, as they arise in the context of immigration and cultural differences, play a prom-

15

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Table 3: Country-specific patterns of differences in the political orientations of social democratic and conservative voters contrast analysis
AT Tolerance with regard to gays and lesbians Reduce income inequality Continue European integration 17% 14% 14% DK Reduce income inequality Equal treatment/ equal opportunities Success and achievement 12% 10% 7% GR Immigrants good for the economy Immigrants make the country a better place to live in Reduce income inequality 12% 10% 9% NO Reduce income inequality Continue European integration Immigrants good for the economy 17% 17% 9% BE Equal treatment/ equal opportunities Reduce income inequality It is important not to attract attention 18% 14% 8% EE Immigrants make the country a better place to live in Immigrants provide cultural enrichment Immigrants good for the economy 16% 16% 15% HU Continue European integration Immigrants of the same ethnicity Important to maintain traditions 15% 12% 7% PL Immigrants good for the economy Reduce income inequality Immigrants make the country a better place to live in 17% 16% 13% CH Continue European integration Immigrants provide cultural enrichment Immigrants good for the economy 83% 73% 56% ES Immigrants provide cultural enrichment Immigrants good for the economy Immigrants make the country a better place to live in 27% 20% 18% IE Immigrants provide cultural enrichment Immigrants make the country a better place to live in Tolerance with regard to gays and lesbians 26% 16% 15% PT Important not to attract attention Immigrants provide cultural enrichment Immigrants good for the economy 6% 5% 4% CY Immigrants good for the economy Reduce income inequality Continue European integration 8% 8% 7% fI Reduce income inequality Immigrants good for the economy Success and achievement 22% 18% 10% IT Reduce income inequality Immigrants provide cultural enrichment Immigrants good for the economy 22% 21% 17% SE Reduce income inequality Immigrants good for the economy Tolerance with regard to gays and lesbians 12% 7% 5% CZ Immigrants good for the economy Immigrants provide cultural enrichment Reduce income inequality 27% 25% 23% fR Immigrants provide cultural enrichment Immigrants make the country a better place to live in Reduce income inequality 30% 27% 22% LU Tolerance with regard to gays and lesbians Immigrants good for the economy Reduce income inequality 19% -12% 10% SI Immigrants make the country a better place to live in Immigrants of the same ethnicity Immigrants of different ethnicities 11% 8% 7% DE Immigrants provide cultural enrichment Continue European integration Tolerance with regard to gays and lesbians 16% 14% 14% GB Continue European integration Immigrants provide cultural enrichment Immigrants make the country a better place to live in 35% 22% 16% NL Tolerance with regard to gays and lesbians Immigrants provide cultural enrichment Reduce income inequality 19% 16% 12% SK Immigrants provide cultural enrichment Immigrants good for the economy Reduce income inequality 22% 20% 14%

For country abbreviations, see Table A1. The given percentages indicate the percentage-point difference and the direction of the difference between social democratic and conservative voters. Colours of the dimensions: social differentiation and inequality grey; tolerance and integration light grey; social conservatism dark grey

16

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

inent role in the orientation profiles of social democratic voters in terms of differentiating them from other political camps is an indication that the problem agenda is adjusted to changing problem situations, without abandoning the general orientation towards problems of inequality and difference. However, the picture is not uniform. In nine out of the 38 cases in which the immigration issue in its various aspects (good for the economy, culture and the country) makes an important difference between social democratic and conservative voters, the judgements of social democratic voters about the consequences of immigration are more negative than those of conservative voters. This becomes clear in five of the 24 countries considered here (Czech Republic, Finland, Luxembourg, Poland and Slovakia) in answers to the question of whether immigration is good for the economy, and in the Czech Republic and Slovakia also in answers to the question of whether immigrants provide cultural enrichment and in Poland whether immigrants make the country a better place to live in. Furthermore, in Austria and Norway European integration is assessed more critically by social democratic voters than by conservative voters. Overall, however, in the great majority of countries a more positive assessment prevails among social democratic voters on the question of the cultural and economic dimension of immigration than among conservative voters. Thus social democratic voters are more liberal and more open on issues of immigration and cultural difference than conservative voters. How sharply a persons highly generalised orientation on the left/right dimension of politics still sets off social democratic voters from conservatives is shown by self-categorisation on this dimension. The heart of social democratic voters still beats left and very much in marked contrast to conservative voters (Figure 9). Here, too, there is considerable variation in the strength of the difference between social democrats and conservatives. The key finding, however, is that social democratic voters regardless of the extent of their separation are not found to the right of centre in any of the 24 countries under consideration. As far as social democratic voters are concerned, their centre is left of centre.

6. Summary and Conclusions

How satisfied are social democratic voters with their parties? Are there still social democratic voters in the sense of people who distinguish themselves as such from other voters in their political aims and orientations? These are the two general questions explored here. In a comparison of voters of the biggest conservative rivals and social democratic voters the results with regard to both questions point towards quite considerable differences. As far as satisfaction with policy outcomes is concerned to the extent that policy can be made entirely responsible it turns out that social democratic voters are more satisfied with their governments in particular when they are not in coalition with conservative parties. Conversely, conservative voters are relatively more satisfied with their governments. Relatively more satisfied means not unconditionally and not always satisfied. Among social democratic voters there is also a mechanism for asserting satisfaction and dissatisfaction with policy outcomes with regard to the government that differs from conservative voters not in tendency but in level and can constitute a particular challenge for social democratic parties. First, social democratic voters punish their governments more harshly with dissatisfaction if they are not satisfied with policy outcomes than conservative voters do theirs. Second, social democratic voters reward their governments less with satisfaction if they are satisfied with policy outcomes than conservative voters do theirs. It therefore appears to be more difficult for social democratic parties to generate support for their activities in government in the sense of the satisfaction of their voters. It is also evident that coalitions with conservative parties do not produce the same satisfaction as is the case among conservative voters. When the two parties are together in government, dissatisfaction with policy outcomes among social democratic voters manifests itself to a much greater extent in dissatisfaction with the government than among conservative voters. Conversely, satisfaction with policy outcomes among conservative voters expresses itself much more strongly as satisfaction with the government than among social democratic voters. In this sense, conservative parties profit much more from a coalition with social democrats than vice versa. Basically, however, such grand coalitions always produce less satisfaction for both social democrats and conservatives than when the parties govern without their biggest rival.

17

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Figure 9: Difference with regard to left-right self-placement among social democratic and conservative voters
SK SI SE PT PL NO NL LU IT IE HU GR GB FR FI ES EE DK DE CZ CY CH BE AT 5 4 1.24 0.94 1.20 0.95 1.49 0.83 1.05 1.19 2.19 0.74 1.56 0.33 0.47 1.75 0.50 1.66 0.03 0.46 1.03 0.84 0.06 1.54 1.21 0.89 1.56 0.48 0.59 4 5 0.72 2.92 3.57 0.37 1.96 1.45 1.62 1.29 1.06 1.93 1.13 2.10 1.09 0.91 1.85 0.76 2.12 2.57 1.30 1.94 2.34

3 2 1 0 1 2 3 Left <------------------------------------------|---------------------------------------> Right Conservative voters Social democratic voters

Another result that is more disadvantageous for social democratic parties than conservative parties is that governments lose their appeal, regardless of performance, or rather their popularity among the voters declines the longer they are in government. This voter-fatigue effect is much stronger among social democratic voters and their governments than among conservative voters and their governments. Although the effects of voter fatigue lead to medium-range vote losses only after a relatively long period of government, it should not be overlooked that often a mere two or three percentage points decide whether a party participates in government or not. Two percentage points can be lost in terms of voter popularity, according to the model presented here, among social democrats after only five or six years, while among conservatives it is only after 16 years. Another effect is shifting into the abstention camp, although that affects social democrats and conservatives equally. However, even that can be a disadvantage for social democratic parties. Al-

though it is disputed whether this is generalisable it has often been observed that an increase in abstentions occurs to the detriment of social democratic parties, and therefore that there is party-specific selectivity with regard to abstention. This finding, too, may be connected to the party voterspecific bias observed with regard to the assessment of performance in government and satisfaction with the government. If social democratic voters apply stricter standards to their governments than conservative voters do, it may also have the result that social democratic governments lose support among their voters more quickly. Basically, these results show that social democratic and conservative voters obviously judge the same social and political circumstances and situations differently. That suggests that they apply different benchmarks, which in turn implies that they differ with regard to their political

18

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

goals and orientations. Concerning this second general question about the political profiles of social democratic voters in contrast to those of conservative voters there is a whole series of differences that absolutely enable us to talk of different profiles. One dimension, naturally, is the traditional domain of social democratic politics: social differentiation and inequality. But it would be wrong to assume that this is the dimension that marks social democrats off from conservatives across all countries. Rather, in many European countries it is clear that for social democratic voters the new aspects of inequality and difference now play a central or equally important role: tolerance and integration, in other words, attitudes about lifestyles, immigrants, immigration and European integration in many countries generate bigger differences in relation to conservative voters than the classic issue of the Left. The fact that typical distinctions arise in this dimension suggests that social democratic voters can also categorise newer problem situations from a social democratic perspective. The differences of orientation between social democratic and conservative voters point to a classic contrast, that between Left and Right. This is also confirmed by the fact that the difference between social democratic and conservative voters on the left/right dimension is not only clear, but in each of the 24 countries social democratic voters classify themselves as left of centre, while conservative voters classify themselves as right of centre. The heart of social democratic voters still beats left and the centre that has to be found for them lies left of centre. It may be that it is precisely these standards and orientations of social democratic voters that make it more difficult for social democratic parties to engender satisfaction and to keep social democratic voters on board. The results suggest that social democratic voters tend to see relatively more reason for change and improvement. Someone who wants those things is less easy to satisfy than someone who is happy with the status quo.

19

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Table A1: Countries and time points in the cumulative European Social Survey
Abbreviation AT BE CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU NL NO PL PT SE SI SK Country Austria Belgium Switzerland Cyprus CzechRepublic Germany Denmark Estonia Spain Finland France GreatBritain Greece Hungary Ireland Italy Luxembourg Netherlands Norway Poland Portugal Sweden Slovenia Slovakia X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X 2002 X X 2004 X X X 2006 X X X X X X X X X X X X X X X X X X 2008

20

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Table A2: Governments in the countries and time points of the cumulative European Social Survey
Abbreviation AT BE CH CY CZ DE DK EE ES FI FR GB GR HU IE IT LU NL NO PL PT SE SI SK Country Austria Belgium Switzerland Cyprus CzechRepublic Germany Denmark Estonia Spain Finland France GreatBritain Greece Hungary Ireland Italy Luxembourg Netherlands Norway Poland Portugal Sweden Slovenia Slovakia Conservative Conservative Social democratic Conservative Social democratic Social democratic & conservative Conservative Social democratic & conservative Conservative Social democratic Social democratic Social democratic Conservative Conservative Conservative Conservative Conservative Social democratic Conservative Social democratic Conservative Conservative Conservative Social democratic Conservative Social democratic Conservative Conservative Social democratic Social democratic & conservative Social democratic Conservative Social democratic Conservative Social democratic Social democratic Social democratic Social democratic Conservative Social democratic Social democratic Conservative Conservative Social democratic Social democratic & conservative Conservative Social democratic Conservative Social democratic Conservative Social democratic Conservative Social democratic & conservative Conservative Social democratic Social democratic & conservative Conservative Conservative Social democratic Conservative Conservative Social democratic Social democratic Social democratic Conservative 2002 Conservative Social democratic & conservative 2004 Conservative Social democratic & conservative Social democratic & conservative 2006 Social democratic & conservative Social democratic & conservative Social democratic & conservative Social democratic Conservative Social democratic & conservative Social democratic Social democratic & conservative Social democratic & conservative Conservative 2008

21

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Figure A1: Assessment of the state of the health care system, the education system and the economy under social democratic, conservative and social democratic/conservative governments
Social democrats in government without conservatives: ratio satisfied/dissatisfied

85% 75% 65% 55% 45% 35% 25% 15% 5% 5% 15% 25% 35% 45% 55% 65% 75% 85% CY CZ DE

% more dissatisfied

% more satisfied

Economy ES GB GR HU NO PL PT

Education SE SI

Health care SK

85% 75% 65% 55% 45% 35% 25% 15% 5% 5% 15% 25% 35% 45% 55% 65% 75% 85% AT BE DK EE

Conservatives in government without social democrats: ratio satisfied/dissatisfied

% more dissatisfied

% more satisfied

Economy ES FI FR GR IE IT LU NL NO PL

Education PT SE

Health care SI SK

85% 75% 65% 55% 45% 35% 25% 15% 5% 5% 15% 25% 35% 45% 55% 65% 75% 85% AT BE

Social democrats and conservatives in government together: ratio satisfied/dissatisfied

% more dissatisfied

% more satisfied

Economy CH CZ DE EE FI

Education NL

Health care SI

22

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Figure A2: Satisfaction with social democratic, conservative and social democratic/conservative governments
Satisfaction with social democratic governments

Very satisfied

5.0 4.0 3.0 2.0 1.0 0.0 1.0

Governments without conservative participation

Governments with conservative participation

Very dissatisfied

2.0 3.0 4.0 5.0 CY CZ DE ES GB GR HU NO PL All citizens PT SE SI SK AT BE Social democratic voters CH CZ DE EE FI NL SI

Very satisfied

5.0 4.0 3.0 2.0 1.0 0.0 1.0

Satisfaction with conservative governments

Governments without social democratic participation

Governments with social democratic participation

Very dissatisfied

2.0 3.0 4.0 5.0 AT BE DK EE ES FI FR GR IE IT LU NL NO PL PT SE SI SK AT BE CH CZ DE EE All citizens Conservative voters FI NL SI

23

BERNHARD WESSELS | THE VOTERS CHOOSY HEART BEATS LEfT Of CENTRE

Bibliography
Bartels, Larry M. (2002): Beyond the Running Tally. Partisan Bias in Political Perceptions.Political Behavior24 (2): 117 150. Campbell, Angus; Converse, Phillip E.; Miller, Warren; et al. (1960): The American Voter. New York: Wiley. Converse, Philip E. (1969): Of Time and Partisan Stability. Comparative Political Studies 2 (2): 139171. feld, Lars P.; Kirchgssner, Gebhardt (1998): Erwartete Knappheit und die Hhe der Wahlbeteiligung: Empirische Ergebnisse fr die neunziger Jahre. In: Klingemann, Hans-Dieter; Kaase, Max (eds): Wahlen und Whler Analysen ausAnlass der Bundestagswahl 1998. Wiesbaden: WestdeutscherVerlag: 415442. fiorina, Morris P. (1981): Retrospective Voting in American National Elections. New Haven: Yale University Press. Gschwend, Thomas; Norpoth, Helmut (2005): Prognosemodell auf dem Prfstand: Die Bundestagswahl 2005. PolitischeVierteljahresschrift46 (4): 682688. Key, V. O. (1961): Public Opinion and American Democracy. New York: Knopf. Maier, Jrgen; Rattinger, Hans (2004): Economic Conditions and Voting Behaviour in German Federal Elections, 19942002. German Politics 13 (2): 201217. Miller, Arthur H.; Wattenberg, Martin P. (1985): Throwing the Rascals Out: Policy and Performance Evaluations of Presidential Candidates, 19521980. The American Political Science Review 79 (2): 359372. Walz, Dieter; Brunner, Wolfram (1998): Its the Economy, Stupid! Revisited. In: Pickel, Gert; Walz, Dieter (eds): Politische Einheit kultureller Zwiespalt? Die Erklrung politischer und demokratischer Einstellungen in Ostdeutschland vor der Bundestagswahl 1998. Frankfurt/M.: Peter Lang: 113126. Weels, Bernhard (1998): Wahlprferenzen in den Regionen: Stabilitt und Vernderung im Wahljahr 1994 oder: Die Heimkehr der CDU/CSU-Whler von 1990. In: Kaase, Max; Klingemann, Hans-Dieter (eds): Wahlen und Whler. Analysen aus Anla des Bundestagswahl 1994. Opladen/Wiesbaden: Westdeutscher Verlag: 259284. Weels, Bernhard (2002): Ideologische Atomisierung oder Whlerwandel? Wahlverhalten in Deutschland 19761998. In: Brettschneider, Frank; Deth, Jan van; Roller, Edeltraud (eds): Politik heute: Das Ende der politisierten Sozialstruktur? Opladen: Leske + Budrich: 4980. About the Author Dr Bernhard Wessels is a senior researcher at the Wissenschaftszentrum Berlin fr Sozialforschung (WZB) and teaches political science at Humboldt University, Berlin. His main research areas are political sociology, international comparative interest intermediation and political representation, and electoral and attitudinal research.

24

About the author Dr Bernhard Wessels is a senior researcher at the Wissenschaftszentrum Berlin fr Sozialforschung (WZB) and teaches political science at Humboldt University, Berlin. His main research areas are political sociology, international comparative interest intermediation and political representation, and electoral and attitudinal research.

Imprint Friedrich-Ebert-Stiftung International Policy Analysis Hiroshimastrae 28 | 10785 Berlin | Germany Responsible: Dr. Gero Maa, Head, International Policy Analysis Tel.: ++49-30-269-35-7745 | Fax: ++49-30-269-35-9248 www.fes.de/ipa To order publications: info.ipa@fes.de

International Policy Analysis (IPA) is the analytical unit of the Friedrich-Ebert-Stiftungs department of International Dialogue. In our publications and studies we address key issues of European and international politics, economics and society. Our aim is to develop recommendations for policy action and scenarios from a Social Democratic perspective. This publication appears within the framework of the working line International Monitoring of Social Democracy, editor: Jan Niklas Engels, Jan.Engels@fes.de; editorial assistant: Nora Neye, Nora.Neye@fes.de

The views expressed in this publication are not necessarily those of the Friedrich-Ebert-Stiftung or of the organization for which the author works. This publication is printed on paper from sustainable forestry.

ISBN: 978-3-86498-129-6