Catchall or Catch and Release? The Electoral Consequences of Social Democratic Parties’ March to the Middle in Western Europe

Abstract

Although the move to the center of many European Social Democratic parties in the 1990s was first rewarded with victories, these parties have since faced a remarkable electoral drought. What explains the seeming inability of these catchall parties to cast a wider but sustainable net for voters? Incorporating a temporal dimension helps explain when and why the broadening of party platforms fails and produces counterintuitive electoral outcomes. Our empirical study analyzes the votes of individuals in three European countries in the past three decades. The individual level allows us to track changes in parties’ voter structures, which are necessarily omitted from studies using aggregate vote shares. Our findings indicate that current analyses of the electoral effects of strategy shifts are misleading inasmuch as they fail to account for individual-level motivations for vote switching.

Keywords

political parties party positions public opinion neoliberal convergence social democracy institutions historical institutionalism

We are living through a period of right-of-centre political dominance in Europe not seen in the whole age of democratic suffrage. In Britain, Germany, France, the Netherlands, Italy and Sweden, six countries with good claims to represent the historic heartland of social democracy, there are now centre-right governments. This has not happened since the First World War.

—David Miliband (London Times, March 8, 2011)

What explains the current ailing of Social Democratic parties in Western Europe? In the late 1990s, left parties or coalitions with left participation governed 12 of the then 15 European Union (EU) countries. By 2006, Social Democracy was suffering throughout Europe; even Sweden’s powerful Social Democratic Party (SAP) lost power for the first time in 12 years. The period 2006–2011 has been even less fruitful for Social Democracy in Europe.¹ These last rounds of defeats for the Social Democrats have taken place in the midst of the most substantial economic recession in 75 years. Yet even in the face of the financial system’s breakdown—amid a weakened regulatory system under the stewardship of center–right governments (see, e.g., Schmidt, 2008)—the major left-leaning parties of Europe have been unable to take advantage of the right’s difficulties.

Combining a number of insights from recent literature on voting and party competition (Adams, Ezrow, & Somer-Topcu, 2011; Adams & Somer-Topcu, 2009; Ezrow, 2005) with a modified version of Kirchheimer’s “catchall” thesis informed by historical institutionalist scholarship (Allen, 2009; Hall & Taylor, 1996; Immergut & Anderson, 2008; Kirchheimer, 1966; Pierson, 2004), we argue in this article that the policy platform choices of Social Democratic leadership were decisive in explaining electoral outcomes. A rightward shift of the left’s mainstream political parties in the mid-1990s contributed substantially to Social Democrats’ slide over the past 10 years. Despite other exogenous and structural impediments, the article suggests it is necessary to add a temporal dimension to explain when and why the broadening of party platforms fails and produces counterintuitive electoral outcomes. We posit that the gains these parties derived from the policy shift toward the middle in the 1990s were short-lived and came at the expense of electoral success in the subsequent decade, mottling the ideological coherence of the parties as political organizations in the process.

The article is organized as follows. First, a brief review of the relevant literature suggests a considerable rift between arguments based on the spatial voting model, implying policy moderation as a beneficial strategy, and a broadly institutionalist perspective that focuses on parties as complex entities. Our main argument relies on the conjecture that time is of the essence in assessing the benefits and costs of parties’ programmatic moves: In the long run, moderating strategies cost parties voters in the political center and at their core. We suggest that this should be observable in the structure of Social Democratic voters over time as well as through measures of party attachments of different types of Social Democratic voters. In the third section, we describe changes in the Social Democratic electorate for the cases of Germany, Sweden, and Great Britain, using data on party positions and individual-level voting data for the 1980s, 1990s, and 2000s. We generally find support for our hypotheses. In the conclusion, we suggest next steps for quantitative tests of our argument and suggest a reevaluation of the popular reading of Kirchheimer’s catchall strategy.

The Promises and Pitfalls of Policy Moderation

A study by Adams and Somer-Topcu (2009) finds evidence of a “lagged policy moderation benefit” (p. 678) for political parties in 25 postwar democracies. Parties that moderated their left–right position at the previous election experience increased electoral support at the current election. Voter support increased when left-wing parties moved right or when right-wing parties moved left in the previous election; parties that shifted in a more radical direction in the previous electoral cycle lost votes in the current election.

To the contrary, a case study focusing on Germany and Sweden (Allen, 2009) suggests that the Social Democratic parties of these countries essentially moderated themselves out of government as they continued to move toward the political center throughout much of the 1990s. Building from Kirchheimer’s “catchall” party thesis (Kirchheimer, 1966), which expected that parties of the center–right and center–left would moderate their positions to cast nets for floating voters near the political center (see Safran, 2009 for a recent review), Allen (2009) argues that focusing on the center exposed the Social Democrats to attacks from smaller left-wing parties, particularly given the PR systems of both countries.

Adams and Somer-Topcu stress the electoral ramifications of moderation for the immediately subsequent electoral cycle. And Allen (2009), for his part, asserts that the moderating moves of Social Democrats in the 1990s, “while initially successful, contributed to their losing power by the early to mid-2000s” (p. 636). That the centralizing strategy was initially successful for the Social Democrats of Germany and Sweden and other European countries suggests the possibility that the benefits of moderation found by Adams and Somer-Topcu are present. That most of these political victories by the moderate democratic left proved short-lived, however, suggests that the spike in support these parties experienced as a result of moderation is not necessarily durable over an even longer time period.

This article tests the compatibility of these two arguments across three European countries with important institutional differences structuring party competition.² First we turn to a brief discussion of center parties and the spatial voting model.

Taking Time Seriously When Studying Party Positions

When Moving to the Center Increases Vote Share

The argument advocating policy moderation as a rational strategy to expand parties’ vote share is firmly grounded in the spatial voting model (Downs, 1957). Assuming that the electorate can be aligned along a single dimension and that the distribution of voters on this dimension peaks in the center, the optimal strategy of nonextremist parties is to move to the center, where most voters are located. The assumption of the electorate’s single-peaked and symmetric distribution has been supported by cross-country surveys (Ezrow, 2005). It states that parties have a consistent support base and implies that they can also fish successfully in the political center while maintaining their core base. The theoretical and formal foundations for this argument are robust (cf. Downs, 1957; Kirchheimer, 1966; Schofield & Sened, 2005). If this strategy is successful, we should observe gains in the vote share of those parties after they expand their ideological profile toward the center:

Hypothesis 1 (H1): Large catchall parties will receive more votes from centrist voters in elections where they moderated their policy positions.

Adams and Somer-Topcu (2009) have argued that this effect is visible only in the election after the policy moderation occurred. This argument is consistent with the spatial voting model and catchall party literatures. For instance, voters’ slow updating of information about parties’ political profile could cause the delay. Of importance, this does not necessarily imply that voters are inattentive or uneducated about political issues and party politics. Voters might evaluate the validity of a party’s policy shift before reacting to the shift. Thus, a lagged effect of policy moderation is a second potential implication from the general catchall argument:

Hypothesis 2 (H2): Large catchall parties will receive more votes from centrist voters one election cycle after they moderated their policy positions.

Adams and Somer-Topcu (2009) provide support for the second hypothesis at the aggregate level. Our study presents a more rigorous test for it at the individual level.

The Fuzzy Center

A basic version of the catchall argument does not address the long-term implications of programmatic moves to the center. We focus on this exact aspect. To set up the argument, we first explore the center of the electorate in more detail. The spatial voting model describes the electorate as a normally distributed, single-peaked population on the left–right dimension of political competition. Because of the shape of the distribution, a party can maximize its utility (gain more new voters) with small moves to the center, whereas moves toward the fringe will gain comparatively fewer voters.

This logic ignores any substantial qualitative differences in voters at either position on the left–right scale. The term swing voters, one of the most prominent and intensely explored concepts in electoral studies, illustrates this point well. Voters in the political center do not display a particularly solid association with any political party. Rather, they change their vote frequently over the course of subsequent elections.³

Because of the volatility at the political center, any assessment of the success of catchall strategies should take into account the long-term implications of such moves. It is entirely consistent with the story of the spatial voting model that shifts to the center result in immediate or once-lagged electoral gains. The more important question, especially with regard to the puzzle of this article, is if these gains are sustainable. In the following section, we present arguments suggesting that they cannot be sustained particularly because the political center is “fuzzy” and easily switches votes. In addition, and maybe more problematically, moves to the center also affect the core voters of a party. Together with the natural movement of swing voters, this effect can result in the electoral postmoderation drought we have observed among European Social Democratic (SD) parties in the past few years.

Moderate Now, Win Later, Lose Eventually? The Advantage of a Historical Institutionalist Perspective

The first two hypotheses suggest that moving to the center will benefit parties’ electoral success in the short term, up to one election after the shift occurred. Swing voters are the primary mechanism behind this argument: Programmatic moderation is the key strategy for large center parties to attract swing voters. With this strategy, however, come side effects that are all too often ignored by recommendations derived from the spatial model and the catchall argument. We introduce these effects step by step and conclude with an overall hypothesis.

First, as the previous section suggests, voters in the middle of the political spectrum rarely display strong affiliations with any political party. This quality is useful for catchall parties as they cast a wider net, but it can play out to their disadvantage when centrist voters switch their votes again at a future election. The “fuzzy center” of the electorate provides votes that are easy to gain, but hard to keep over time. This assertion can be tested empirically:

Hypothesis 3a (H3a): Voters from the center of the electorate display weak or no affiliations with SD parties.

Hypothesis 3b (H3b): Voters from the center of the electorate are likely to make up larger shares of Social Democrats’ voters immediately after these parties moderate their policies. Yet these voters will decrease one or two periods after moderation occurred.

Second, a party’s move to the center not only sends signals to centrist swing voters but also contains information for voters on the far from center end of the party’s ideological spectrum. These voters can be made up of different types—party activists, in the case of SD parties, for instance, union activists, but also strategic voters with more extreme preferences than the pool of competitive parties can cover. These activist voters may be willing to accept it as a means for short-term success, but a move to the center is not their true preference for the party they vote for. If, however, their party maintains this new position for longer than would be necessary for one-time electoral gains, these far-from-center voters can be alienated and cease supporting the respective party. Recent research indeed suggests that SD parties tend to adjust their position according to changes in the mean voter position rather than shifts in the preferences of the parties’ supporters (Ezrow, de Vries, Steenbergen, & Edwards, 2011), making the alienation of rank-and-file party members even more likely.

Historically, SD parties were less likely to shift ideological positions because of changes in public opinion and more apt to attempt shaping opinion rather than be shaped by it. The long interdependence of leftist parties and trade unions and socialist societies also differentiated Social Democrats from other political parties. Although these ties helped maintain class identity salience while the left parties pursued strategies aimed at expanding their electoral reach and ideological flexibility, these same ties also constrained the extent to which ideology could be strategically expanded, creating what Przeworski and Sprague (1986) refer to as the “dilemma of electoral socialism.” SD parties also differed from those of the center and right in the extent to which they enrolled dues-paying members and allowed participation of institutionalized member activists in internal decision making, the foundations of the mass party (Duverger, 1963). Despite this historical legacy, office-seeking rather than policy-seeking behavior is found more in postwar SD parties (Kitschelt, 1994) and is even more prominent within the “Third Way” movement of the late 1990s (Keman, 2011). At the same time, the ties to unions and other social groups have declined (e.g., Poguntke, 2002). The end result is a party family, historically predicated on a concrete ideological stance, with a much less distinctive ideological profile.

In the late 1990s, these parties could not ignore the threat that globalization posed to the welfare state, social spending, and the trade union working class, the primary base of Social Democrats’ support (Blyth, 2003).⁴ It served to reinforce beliefs within Social Democrats that the growth of the service sector required the party to appeal to more professional workers at the expense of the working class (Anderson, 2006). By the late 1990s Tony Blair’s New Labour in 1997, Gerhard Schröder’s Neue Mitte in 1998, and the Italian Olive Tree coalition in 1996 had all won elections pursuing a moderation strategy (Kitschelt, 1999). As others have noted, however, it may prove difficult if not impossible in the long run for these parties to reconcile the “middle way” with the domestic demands of their SD constituents (Pierson, 2002). Having moved away from their old base in the “rust” for the “postmaterial center,” the catchall parties of the left undermined their ability to return credibly to materialist arguments.

Although the spatial voting literature has no immediate explanation for this flaw of the strategy, a historical institutionalist perspective, a theoretical approach seldom applied within the voting literature, provides important insights on the matter. Most of the research in the spatial voting framework discussed above considers parties as black boxes and rational unitary actors striving to maximize vote share. Scholars in the historical institutionalist tradition see parties as much more complicated entities and have shown more interest in parties as evolving organizations and battlegrounds of different member groups with different interests (see, e.g., Berman, 1998). This perspective allows for a more fine-grained look at the development of party platforms. For instance, Hirschman’s (1970) classic study of exit, voice, and loyalty as potential responses to organizational decline pertains to political parties as much as to the firms and various other institutional entities he discusses. The shifts within SD parties at the end of the 20th century and beginning of the 21st increased the importance of party elite at the expense of rank-and-file party members, altering the incentive structure for the possible responses to these parties’ decline throughout the past decade. With less space within the party for membership to express voice, loyalty decreased and exit became a more attractive option.

A key aspect of Streeck and Thelen’s (2005) historical institutionalist approach is that “much transformative institutional change takes place gradually, so we need to analyze the ways in which these processes unfold” (Immergut & Anderson, 2008, p. 356). This allows scholars to focus on aspects of political parties that the spatial voting framework often misses: namely, the organic nature of political parties as membership organizations. Consequently, this approach suggests that parties as durable institutions cannot move along the policy space without generating “lagging effects” among their members and voters. That is, the relationship between individual ideology (or left–right self-placement) and vote choices is not an immediate one with regard to time.

Voters on the far side of a party’s spectrum usually constitute the more active core of the organization. They are the ones who have invested time, or developed ties with the party they have voted for, over a long period of time. Although they may tolerate a strategic shift to the center, they are not indefinitely tied to the party. This assertion can also be tested empirically by tracing the movement of this type of voter over time:

Hypothesis 4 (H4): When a major party moves to the center, its voters from the noncenter side of the party’s left–right profile will keep voting for the party for one or two electoral cycles. But the party will lose them two or three elections after the move to the center.

H3 and H4 combine to provide a consistent explanation for Social Democrats’ electoral woes after a successful period in the late 1990s. Pointing to the importance of temporal dynamics in voter behavior, we argue that catchall strategies are not an effective tool to increase a party’s vote share and maintain the increase over the course of several electoral cycles.

Evidence: Changes in Social Democratic Party Constituencies Over the Past Three Decades

To test these hypotheses, we present survey evidence, illustrating that a closer look at the temporal dimension of programmatic shifts may help explain surprising cases like the recent failure of SD parties to exploit otherwise favorable circumstances. We start by inspecting possible trends in the overall electorate to see if changes in the economic structure of these countries correlated with substantial changes in the ideology of the electorate. Subsequently, we examine (a) the composition of the SD electorate and (b) what options “leftist” voters pick among their choice sets of SD parties, other parties, and abstentions.

Cases

We focus on SD parties in three countries: Germany’s Sozialdemokratische Partei (SPD), Sweden’s Socialdemokratiska Arbetareparti (SAP), and Great Britain’s Labour Party. All three parties were able to make significant electoral gains in the 1990s but have suffered since at least the middle of the first decade of the 21st century. None of these parties is currently in power, and each of them faced considerable difficulties to establish coherent narratives about how their centrist policies in the 1990s and early 2000s contributed to a distinctly SD position toward the economic crisis of the past three years. We chose these three cases because they offer interesting variation of the political landscape with regard to the political Left. In Sweden, several feasible alternative choices for leftist voters exist with the Green and Left parties, both of which regularly exceed the 4% threshold to achieve representation in the Swedish Riksdag. In Germany, the Greens have been an established political party for at least two decades, but the Left Party is still frequently stigmatized as the formal successor of the East German SED/PDS, thus limiting the potential for cooperation among the Left. And in Great Britain, the institutionalized inability of any party other than Labour, the Tories, and the Liberal Democrats to achieve a significant caucus in Parliament exercises significant constraints on left voters to switch their vote. This variation allows us to control, at least casually, for the mediating impact of electoral institutions and party landscapes on the effect of Social Democrats’ programmatic shifts.

Operationalization

Party positions

First, to explore programmatic shifts of the SD parties, we rely on two common sources. The Comparative Manifesto Project (CMP; Budge, Klingemann, Volkens, Bara, & Tanenbaum, 2001; Klingemann, Volkens, Bara, Budge, & McDonald, 2006) data place parties on a left–right spectrum based on how their manifestos rank on a number of issues, such as economic and social policies. But party manifestos as sources of party positions may suffer from a gap between “words and deeds,” where parties’ programs may differ from the policies they actually pursue in government or opposition. For this reason, we also rely on the Chapel Hill Expert Survey (CHES; Bakker et al., 2012; Hooghe et al., 2010; Steenbergen & Marks, 2007) as an alternative measure. These surveys compile party positions based on the evaluations of experts, mostly political scientists specializing in the respective country. We map both CMP and CHES data on a left–right scale from 0 (far left) to 10 (far right).⁵

Individual-level survey data

We then use several national and European election surveys from each country to test our hypotheses and explore information on voters’ characteristics.⁶ The left–right dimension of political competition marks the core of this article.⁷ We use survey respondents’ political self-placement on a scale from far left to far right to examine the structure of parties’ voters, but collapse the continuous scales into three elements, left of center, center, and right of center. This ensures the comparability across different surveys where voters were given different scales for placing themselves.

Voters of a specific party are treated as those respondents who indicated they had voted for the party of interest (or abstained) in the last general/national election before the survey was taken. For the arguments involving party support and party attachment, we use survey items that asked respondents if they supported or favored a particular political party and consecutively asked about the strength of respondents’ association to that party. Finally, we approximate the concept of likely vote choice or party supporters—both terms appear interchangeably below—through survey items that ask respondents about their political identity or affiliation. Thus, all respondents who listed the respective SD party as their “political home” enter our analysis as likely SD voters or supporters.⁸

Findings and Discussion

Although anecdotal evidence such as the Blair–Schröder (2000) article suggests that SD parties moved toward the center in the 1990s, systematic data present a clearer picture of parties’ shifts. Figure 1 shows how the German SPD, the Swedish SAP, and the British Labour party have moved across the left–right continuum in the past three decades. The following sections discuss these movements and their consequences for voters’ political behavior in more detail with respect to each case.⁹

Figure 1.

Programmatic changes of social democratic parties.

Great Britain: The Rise of New Labour and Decline of Core Labour Support

The trend toward the center is most pronounced for the Labour party, which moved further across the political spectrum than any other European party in the past 30 years. Notoriously weak despite frequent leadership changes, the party had made the biggest leap toward (in fact, beyond) the political center in its history around its landslide election victory under Tony Blair in 1997.¹⁰ As Blair’s formerly rising star began to decline (Quinn, 2006), one can also see that Labour moved slightly back toward the left in 2005–2006. The data point in 2010—indicating the party’s position before the elections—is almost exactly at the political center.¹¹

Change and continuity in the British electorate

Moving the party to the political center would be a successful strategy for party leaders if one could observe a trend that voters’ affinity to SD values had declined. As a proxy indicator for the predisposition toward these values in the electorate, we examined the proportion of voters who considered themselves “left of center.” Figure 2 tracks the share of these voters, compared to “centrist” and “right of center” voters, among the British electorate in the past three decades.¹²

Figure 2.

Composition of the British electorate.

Contrary to arguments that SD parties’ moves to the center were a natural adjustment to fundamental changes in the electorate, the 1990s actually witnessed an increase of left of center voters among British voters. The group of centrist voters grew moderately, whereas right-of-center voters declined until they increased again in 2005. Only in the early 2000s, several years after New Labour, do we observe centrist voters as the majority of the electorate, whereas the share of left voters was still higher than in the 1980s. For the purposes of this analysis, this suggests two major points. First, Labour’s move to the center, albeit successful in the short term, did not follow a trend of a “vanishing Left” among the electorate; an increase in centrist voters occurred only after Labour’s shift. Second, it raises the question of how a significant part (just short of one third) of the electorate would cast their votes if the only feasible party in Britain’s first-past-the-post system moved beyond the political center and thus significantly further away from this group’s political identity. The next two paragraphs address each of these points separately.

Changes in the composition of Labour voters: Political identity and party attachment

The plan behind Blair’s catchall policy moderation strategy was to tap swing voters, without whom a dominating election victory would not have been possible. Above, we suggested that although catchall strategies may attract such swing voters in the short run, they change the composition of a party’s electorate inasmuch as they drive out voters who are further away from the political center, but also more attached to the party and thus more loyal long-term supporters. Figure 3 evaluates this proposition for the British context.

Figure 3.

Labour voters and supporters.

Panel a shows that electoral gains in 2001, the second win following the 1997 landslide, depended much more on nontraditional SD voters compared to Labour’s drought in the early 1990s. That is, voters from the political center and from the right made up for larger shares of Labour’s electorate, whereas the importance of left voters decreased. This supports H2 (delayed gains) more than H1 (immediate gains) but also illustrates the changes in the structure of the voters that come with higher vote shares after policy moderation. In 2010, almost no Labour voters came from the right or center, while the overall votes decreased. Meanwhile, the “new” voters of New Labour were also less attached to the party, as Panel b illustrates. The share of strong Labour supporters among its voters has been on the decline since 1987; New Labour was not particularly able to mobilize these voters. Rather, its success rested on attracting voters with few or no ties to the party. In this vein, Panel c supports H3a on the “fickle center.” Across time, centrist Labour voters display significantly less attachment to the party than voters further to the left of the political spectrum. As we explored above, this strategy can backfire if voters with little attachment—the fickle center—choose to switch votes again at the next election, all the while Labour’s core constituency continues to flock away from the party. This is well represented in the 2010 elections, where less than 10% of Labour voters had little or no attachment to the party. In other words, those voters had left Labour again, as we expected in H3b.

Tracking likely Labour voters

This last statement about the trend of traditional Labour voters switching their votes is corroborated by an examination of the voting behavior of self-identified Labour supporters.

From the 1997 elections onward, more individuals with ties to the Labour party than ever before chose to either abstain or vote for another party. The high rates of abstentions, more than 20%,¹³ are not surprising in an electoral system that offers few feasible alternatives with realistic chances of securing seats in the House of Commons. Conversely, it is notable that even under those conditions more than 10% of Labour supporters cast their vote for a different party.

In sum, the British case offers consistent support for H4, illustrating a time-dependent argument about the effects of SD parties’ catchall strategies on the composition of their electorate and thus their long-term success. Aggregate findings of a positive effect of policy moderation on vote share necessarily overlook the trends that are clearly visible in the British case: Labour’s policy moderation resulted in a shrinking share of core Labour supporters and a rise of noncommitted centrist voters in the party’s electorate.

Germany: Neue Mitte Loses Out to Neue Linke

Under Helmut Schmidt, Germany’s SPD had assumed a position as a cornerstone of the German model of a Democratic Capitalism or Soziale Marktwirtschaft. After the Free Democrats left the coalition in 1982 and the SPD assumed the opposition, moderately left programs and chancellor candidates failed to lead the party back into power. Gerhard Schröder’s course of portraying the SPD as Neue Mitte, clearly reflected in the party’s position change between 1996 and 1998, indeed resulted in the party’s return to government offices. But the ability of the Christian Democrats in opposition to plausibly lay claim to supporting most social spending programs the SPD might intend to cut made this attempt at a programmatic change difficult for the SPD (Zohlnhöfer, 2007). The narrow defeat in 2005 and the ensuing 4 years in a grand coalition allowed the party to fill a slightly more left role, but also constrained it to carry policies that stood in sharp contrast to more traditional SD values.¹⁴ In 2009, the SPD lost with a historically low vote share, running on a fairly central platform under Schröder’s former foreign minister Frank-Walter Steinmeier. Since this defeat, the SPD has started a new process of soul searching, elected new party leaders, and begun to readjust its program.

Change and continuity in the German electorate

The yearly Politbarometer surveys on German politics display a slight dispersion of the three voter groups in Germany in the mid-1990s (see Figure 4). After 1996–1997, centrist voters rose to the largest group among the electorate, whereas fewer voters identified themselves as right or left than before—although the decline is more pronounced for right-of-center voters. Throughout the early 2000s, the share of left-of-center voters remained at essentially the same level as it had been since the 1980s, just more than 30%.

Figure 4.

Composition of the German electorate.

Although the increase in centrist voters is remarkable, it also appears that the ratio of centrist voters has been in decline for a few years now. Meanwhile, the consistency of the one third of the electorate located left of center again suggests that moving the SPD toward the center is not a necessary consequence of an erosion of the leftist voter base alone.

Changes in the composition of SPD voters: Political identity and party attachment

Under Gerhard Schröder in the run-up to the 1998 elections, the SPD pursued the center of the German electorate, marketing itself as the moderate and established part of the red–green project. This catchall strategy paid off, as the data reveal.

Panel a of Figure 5 shows how the share of centrist voters among all those who cast their ballot (Zweitstimme) for the SPD jumped up from around 30% in 1994 to almost 50% in the 1998 elections. This is in line with the catchall argument based on the implications of the simple spatial model of voting, and supports H1. However, as our argument suggests, centrist voters are fickle and will not stay with one catchall party over longer periods. The SPD experienced exactly this scenario, contrary to H2, as the share of centrist voters declined again subsequently to 44% in 2002 and 36% in 2005.

Figure 5.

SPD voters and supporters.

In Panel b of Figure 5, we see analogously that the share of noncommitted SPD voters increased dramatically at the 1998 elections, where these voters for the first time in several decades exceeded strong and moderate SPD supporters. Again, in the following election in 2005 that yielded significant losses for the SPD, this type of voter turned away from the party, and their share among its voters fell far under 40%.¹⁵ Exactly as in the British case, we also find strong evidence that centrist voters are indeed fickle; Panel c again illustrates the strong relationship between centrist self-placement and low or no party attachment. These data reflect our expectations from H3a and H3b relatively well.

Tracking likely SPD voters

More convincing evidence for this scenario can be found in the voting behavior of self-identified SPD supporters who turned away from the party. Panel d illustrates how historically low abstention rates and vote switches of SPD supporters carried the 1998 election win. In 2002 and even more in 2005, more supporters than before abstained or chose to cast their ballot for another party. This observation matches our expectation based on H4. Unlike the British case, the German proportional system offers at least two feasible alternatives for disappointed left of center voters, the Greens and the former PDS and current Linkspartei.

The comparison with the British case yields an interesting difference. After 1998, typically twice as many Labour supporters abstained than switched to a different party. For the self-identified SPD supporters, this relationship is reversed. In all three elections in which the SPD ran under the catchall Neue Mitte platform, two or three times as many of the party’s supporters voted for a different party. Most of these were likely strategic votes for the Green party to secure the option of a red–green coalition, an outcome that would be highly unlikely in a majoritarian system. However, the share of switchers increased significantly in 2002 and 2005, whereas abstentions doubled between 1998 and 2005. We interpret this as further evidence for our argument that the catchall strategy has the potential to alienate the core constituency of SD parties in the medium term, especially if alternative and feasible options are available.

Sweden: Social Democracy Slips in a Former Stronghold

The Swedish Social Democrats (SAP) changed their position more cautiously than both Labour and the SPD. Failure to achieve core political projects and the loss of power in 1976 resulted in a significant reexamination of the party’s future programmatic strategy. Although the party was able to regain power in 1982, the soul searching and the creeping separation from the unions continued and were later met by the challenge of new political realities and constraints through the deepening of the European Community and EU. By the mid-1990s, the party had joined Europe’s other SD parties in moving to the center. Comparing the CMP and CHES data suggests that this move was stronger in name, that is, in the party’s manifestos, than in overall policy. Yet from both perspectives it is clear that the Swedish Social Democrats significantly changed course toward the center. As election results in the 1990s did not reward this move, the SAP readjusted its program again, but was not immediately rewarded and lost power in 2006.¹⁶ In 2010, again with a more centrist program, the party failed to defeat the incumbents, so that for the first time in Swedish history the Moderaterna were able to maintain governing power.

Change and continuity in the Swedish electorate

Unlike their British and German counterparts, centrist voters are by far the smallest part of the Swedish electorate in most years. Throughout the 1990s, when the SAP had moved the closest toward the political center, it is notable that the electorate had become less centrist, whereas both left-of-center and right-of-center self-placements became more prominent. Only the first years of the 21st century saw a significant increase in centrist voters, a trend that may be reversed again in the 2010 elections. In any case, the Swedish case also fails to offer substantial evidence for propositions of a dramatically eroding leftist voter base.

Changes in the composition of SAP voters: Political identity and party attachment

The Swedish case differs from the developments in both Germany and Great Britain. As indicated in the previous paragraph, the SAP’s moves on the left–right scale were much less pronounced than Labour’s and the SPD’s programmatic readjustments (see Figure 6). Concurrently, SAP voters appear to be decidedly more stable in their left–right positions and party attachments.

Figure 6.

Composition of the Swedish electorate.

Both Panel a and Panel b of Figure 7 display more continuity than the same indicators did for Labour and the SPD, although similar trends are also present. In 1998, the party’s manifesto was rated slightly to the right of center, further right than ever before (see Figure 1). In the same year, the party attracted more voters with moderate or no attachment to the SAP than it had in previous elections, although leftist voters were still most pronounced among its voters. H1 and H2 are mostly supported in the this case when we consider the party attachment of SAP voters, with higher levels of moderates and independents making up the party’s electorate in 1998 and 2002. As in the other two cases, we also find strong evidence for H3a in Panel c for a correlation between voters’ centrist positions and their lack of party attachment. In 2006, the decrease of voters with moderate attachment and increase of independent voters in the SAP’s electorate is ambiguous with regard to H3b. We would expect that new data on the 2010 Swedish elections could yield a more conclusive evaluation in this case.

Figure 7.

SAP voters and supporters.

Where did they go? Tracking likely SAP voters

The (comparatively) strongest trends among SAP supporters are not related to the composition of the party’s voters, but appear in the voting behavior of SAP supporters. In 1994, the most successful election for the SAP in the past two decades, only 0.5% of self-identified SAP supporters abstained, whereas 5% cast a ballot for another party. Again, in Sweden’s PR system, this behavior is expectable as some SAP supporters might push for a coalition government or strengthen possible supporting parties of a minority government. As the SAP continued to pursue its catchall strategy, however, its supporters increased abstentions significantly (close to 3% in 1998, 2% in 2002), whereas the overall election returns for the SAP declined. And in 2006, the SAP’s worst electoral performance since 1921, 14% of SAP supporters reported to have abstained, again in line with the expectation expressed in H4.

Although the surprisingly low numbers on abstentions in the Swedish case before 2006 must be treated with a limited degree of caution—we refer to our previous comment about the underreporting of abstentions, although this only biases our inferences against our favor—they offer a similar account of the effects of programmatic moderation and catchall strategies on the SAP’s voters and supporters. The historically low returns of the SAP in 2006 were partially accompanied by both a failure to retain centrist and nonattached voters that drove in previous election successes (H3b) as well as the turn of SAP supporters to alternative votes or empty ballots (H4).

Considering Alternative Arguments: Dealignment, Lower Turnout Trends, and Incumbency Effects

These findings suggest that although the conventional wisdom on the benefits of catchall strategies is empirically supported in the short run (H1 and H2), it warrants a closer look when it comes to behavior of the undecided middle of the electorate (H3a and H3b) and especially the core supporters of SD parties (H4). The three cases examined in this article correspond significantly to our expectations about these two voter groups. Before concluding with implications and suggestions for further strategies to build time into spatial approaches to party strategies, we discuss alternative explanations of the trends described in the previous section.

Dealignment

The observation of increases in SD voters with low or no attachment to the party could equally be caused by a general trend of partisan dealignment. To test this argument, we examined the voter structure of the three major parties on the right, the Tories, the CDU/CSU, and the Moderaterna; our exploration did not yield results that supported this argument. In the United Kingdom, we find a slight increase of Conservative voters with no party attachment (at the cost of strong Conservatives), but the structure of Conservative voters is visibly more stable than in the case of Labour. The other two cases show a much stronger difference between the voter structures of SD and Conservative parties. In Germany, the size of these groups virtually does not change at all over time for CDU/CSU voters (strong ID: ~55%, medium ID: ~40%, no ID: ~5%), as is the case in Sweden for Moderaterna voters (strong ID: ~50%, medium ID: ~30%, no ID: ~20%).

Trends toward lower turnout

We also considered possibility that the perceived changes in SD supporters’ voting behavior might be rooted in a secular trend toward lower turnout in Western Europe. Tracking the votes of self-identified Conservatives (supporters of the Tories/CDU/CSU/Moderaterna), we find no convincing evidence of such a pattern. In the United Kingdom, self-identified Conservatives’ abstention rates (based on their survey responses) move between 10% and 18%, but with no clear trend toward more abstentions (in 2010, only 11% of Conservatives abstained). In Germany, CDU/CSU supporters never abstained at a rate higher than 7%, also with no clear upward trend. In Sweden, the same applies to voters sympathetic to the Moderaterna. Tracking alternative votes similarly fails to reveal a clear pattern of partisan dealignment or increased vote switching in the 1990s or 2000s for Conservatives. Although we do not consider our additional analyses to be the last word on these major arguments and hope for future studies in this direction, we see them as evidence that supports our argument about idiosyncratic effects of Social Democrats’ programmatic shifts on their voter base.¹⁷

Incumbency effects

A third objection to our interpretation of the evidence is rooted in reverse incumbency effects, where the “cost of governing,” rather than policy moderation, materializes in voters deserting incumbent SD parties.¹⁸ When Labour was voted out of power after three terms, a historically long time in office for that party, its loss was mainly attributed to the party’s inability to distance itself from unpopular policies. Similarly, the SPD faced its most dramatic loss in 2009 after two terms of a red–green coalition, and one subsequent term of a Grand coalition with the Christian Democrats. Such incumbency effects are inadvertently linked to our argument. Policy moderation and governance often correlate, as parties tend to face constraints in realizing more ambitious (or radical) policies once in office. This makes it intrinsically difficult to tease out the effects of moderation alone from incumbency. In our view, though, several points strengthen the argument about the effects of moderation against incumbency costs.

The cost of governing argument is linked to the clarity of responsibility, such that for the electorate to punish incumbents, it must be clear that incumbent parties were responsible for undesired policies (Powell & Whitten, 1993). Our cases offer interesting variation in this regard. Labour was clearly the only party that can be made responsible for the Iraq War or unpopular fiscal policies. But Powell and Whitten (1993) suggest, among other factors, coalition governments as one scenario where the responsibility for policies is diluted. The SPD’s participation in the Grand coalition thus makes it less likely that voters punished the SPD more than the CDU for undesirable policies. We believe our data rather indicate that the incumbency punishment was likely stronger for the Social Democrats than for the Conservatives because the SPD could not sustain centrist voters and it was losing its supporters to other parties or abstentions. The German case can thus be read as supporting the moderation argument over a negative incumbency effect.

Second, the Swedish SAP was in power only until 2006, when our analysis ends. However, the SAP’s unsuccessful run in the 2010 elections (with 4.3% less than in 2006, a historically low result; see Table 1) also corresponds to our argument more than a punishment for the cost of governing. With the SAP in the opposition for 4 years, and a new leader with Mona Sahlin, one would expect that voters attribute only a limited degree of responsibility for pre-2006 SAP government in the 2010 elections. However, the party maintained a centrist profile under Sahlin (see Figure 1). With incumbency not present as a factor in 2010, but the party running on a centrist platform, we consider our argument more likely than negative incumbency effects.

Table 1.

Election results of Social Democratic Parties, 1979–2010.

Party	Year	% of the popular vote
Labour	1979	36.9
	1983	27.6
	1987	30.8
	1992	34.4
	1997	43.2
	2001	40.7
	2005	35.2
	2010	29.0
SPD	1983	38.2
	1987	37.0
	1990	33.5
	1994	36.4
	1998	40.9
	2002	38.5
	2005	34.2
	2009	23.0
SAP	1991	37.6
	1994	45.3
	1998	36.4
	2002	39.9
	2006	35.0
	2010	30.7

Vote shares for each election year were retrieved from the national election committees: Great Britain: http://www.electoralcommission.org.uk/elections/results/general_elections and http://news.bbc.co.uk/2/hi/uk_news/politics/election_2010/8654338.stm; Germany: http://www.bundeswahlleiter.de/de/bundestagswahlen/; Sweden: http://www.scb.se/Pages/TableAndChart____32065.aspx and http://www.val.se/in_english/index.html (all retrieved January 18, 2011).

Finally, further examples similar to the SAP in out-of-sample cases not discussed in this article might render the incumbency argument less convincing. The French socialists, running on a moderate–centrist platform, experienced a similar scenario as the SAP, being voted out of power in 2002, but losing again unambiguously in 2007. In the Netherlands, the PvdA has not led the government since Wim Kok left in 2002; incumbency effects are unlikely here as well. Overall, we acknowledge the importance of voters of both types, core supporters and centrist swing voters, punishing incumbent SD governments for different reasons. However, our discussion mostly suggests that it is not incumbency per se but the moderate policy platforms of these parties that cause these voters to switch their votes.

Implications and Conclusions

Our article combined a number of insights from recent literature on voting and party competition with a modified version of Kirchheimer’s “catchall” thesis, informed by historical institutionalism, to explain that the strategy of policy moderation pursued by SD parties in the mid- to late 1990s yielded short-term electoral gains, but undermined the continued viability of the party as a distinct, ideologically coherent political organization. Our analysis of voters and SD parties in Germany, Sweden, and Great Britain indicates that moving toward the political center is not a durable answer to the problems the parties face nor a successful strategy in achieving their long-term goals. Rather, moving to the center captures voters that are no “safe bet” for SD parties for future elections and simultaneously has the potential to drive more attached “core” voters to other parties on the left side of SD parties. Based on our analysis, continued moderation and embrace of the catchall policies that Kirchheimer lamented will likely weaken these parties further and undermine their role as functional political organizations.

Despite their relative stability, parties, like all institutions, are not incapable of change over time. Although strategically shrewd from the perspective of the leaders of the various parties and temporarily beneficial at the polls, Social Democrats’ shift toward the middle caused substantial damage to the parties as enduring institutions in the long run. Kirchheimer (1966) himself put the point best:

The party’s transformation from an organization combining the defense of social position, the quality of spiritual shelter, and the vision of things to come into that of a vehicle for short-range and interstitial political choice exposes the party to the hazards of all purveyors of nondurable consumer goods: competition with a more attractively packaged brand of nearly identical merchandise. (p. 195)

Future research in the direction suggested by this article should focus on developing an empirical model that enables us to take into account abstentions, which have proven crucial to the recent losses of the catchall parties. A more sophisticated analysis should also probe the relationships explored within this article under the inclusion of party-level measures of left–right movement, addressing the hierarchical structure of the argument. Finally, expanding the set of countries in the study and looking at the dynamics of parties on the center–right should prove productive as well.¹⁹ The consequences of catchall strategy moves to the center are beginning to show for some of Europe’s conservative parties as well. The emerging debate about the “conservative soul” of the German CDU/CSU in the past few months, for instance, indicates that not just Social Democrats may be suffering from unsustainable majorities as a consequence of centrist catchall strategies in the past two decades.

Footnotes

Acknowledgements

We thank the Svensk Nationell Datatjänst, the Zentralarchiv für Empirische Sozialforschung, and the British Election Studies research team for making their survey data available. This study does not reflect the views of any of these institutions. We are grateful to Ryan Bakker, James Caporaso, Catherine de Vries, Ray Duch, Susan Scarrow, and two anonymous reviewers for their valuable comments on our work.

Authors’ Note

A mentor, colleague, and friend to the other authors of this article, Chris Allen contributed substantially to the present article up to the point of his untimely death in February 2011, as well as to the personal intellectual development of his coauthors. Ideas first articulated within Chris’s research form the theoretical foundations of the present study, and his thoughts and writings make up the core of what merits there are in this article. We dedicate this manuscript to his memory, and accept any errors as ours alone. Previous versions of this article were presented at the annual meetings of the American Political Science Association in Washington, D.C., 2010, and Midwest Political Science Association in Chicago, Illinois, 2011. Karreth and Polk contributed equally to this article.

Replication data and –commands are located at <>; additional empirical material referenced in the text can be obtained by contacting the authors.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Johannes Karreth is a PhD candidate in political science at the University of Colorado at Boulder, focusing on the intersections of international cooperation and domestic politics. He has previously published in the Journal of Politics and International Studies Perspectives.

Jonathan T. Polk is a postdoctoral fellow at the Centre for European Research and Department of Political Science at the University of Gothenburg. His research focuses on party positioning in European democracies and its consequences for electoral behavior. His research has been published in European Union Politics and is forthcoming in Party Politics.

Christopher S. Allen († 2011) was the Josiah Meigs Distinguished Teaching Professor in the Department of International Affairs at the University of Georgia. He was coauthor of European Politics in Transition, now in its sixth edition. He published articles on comparative political economy and on democratic representation in numerous journals, including Harvard Business Review, Comparative Politics, Journal of Public Policy, Policy Studies Review, Publius, West European Politics, German Politics, Economic and Industrial Democracy, and Journal of Policy Analysis and Management.

References

Adams

Clark

Ezrow

Glasgow

(2006). Are niche parties fundamentally different from mainstream parties? The causes and the electoral consequences of Western European parties’ policy shifts, 1976–1998. American Journal of Political Science, 50(3), 513-529.

Adams

Ezrow

Somer-Topcu

(2011). Is anybody listening? Evidence that voters do not respond to European parties’ policy programmes. American Journal of Political Science, 55(2), 370-382.

Adams

Green

Milazzo

(2012). Has the British public depolarized along with political elites? An American perspective on British public opinion. Comparative Political Studies, 45(4), 507-530.

Adams

Haupt

Stoll

(2009). What moves parties? The role of public opinion and global economic conditions on in Western Europe. Comparative Political Studies, 42(5), 611-639.

Adams

Merrill

Grofman

(2005). A unified theory of party competition. New York, NY: Cambridge University Press.

Adams

Somer-Topcu

(2009). Moderate now, win votes later: The electoral consequences of parties’ policy shifts in 25 postwar democracies. Journal of Politics, 71(2), 678-692.

Allen

C. S.

(2009). Empty nets: Social democracy and the “catch-all party thesis” in Germany and Sweden. Party Politics, 15(5), 635-653.

Alvarez

Nagler

Bowler

(2000). Issues, economics, and dynamics of multiparty elections: The 1987 British general election. American Political Science Review, 94(1), 131-149.

Alvarez

Nagler

Willette

(2000). Measuring the relative impact of issues and the economy in democratic elections. Electoral Studies, 19(2), 237-253.

10.

Anderson

(2006, August–September). Labor movement visions and welfare state restructuring in Sweden, the Netherlands and Germany. Paper presented at the annual meeting of the American Political Science Association, Philadelphia, PA.

11.

Aylott

Bolin

(2007). Towards a two-party system? The Swedish parliamentary election of September 2006. West European Politics, 30(3), 621-633.

12.

Bakker

de Vries

Edwards

Hooghe

Jolly

Marks

Polk

Rovny

Steenbergen

Vachudova

(forthcoming). Measuring party positions in Europe: The Chapel Hill Expert Survey trend file, 1999-2010. Party Politics.

13.

Berman

(1998). The social democratic moment: Ideas and politics in the making of interwar Europe. Cambridge, MA: Harvard University Press.

14.

Blair

Schröder

(2000). The third way/die neue mitte. Dissent, 2, 51-65.

15.

Blyth

(2003). Globalization and the limits of democratic choice: Social democracy and the rise of political cartelization. Internationale Politik und Gesellschaft, 3, 60-82.

16.

Budge

Klingemann

H.-D.

Volkens

Bara

Tanenbaum

(2001). Mapping policy preferences: Estimates for parties, electors, and governments 1945–1998. Oxford, UK: Oxford University Press.

17.

Downs

(1957). An economic theory of democracy. New York, NY: Addison-Wesley.

18.

Duverger

(1963). Political parties: Their organization and activity in the modern state. New York, NY: John Wiley.

19.

European Parliament Election Study. (2009). Voter study: Advance release, 22/06/2011. Retrieved from http://www.piredeu.eu

20.

Ezrow

(2005). Are moderate parties rewarded in multiparty systems? A pooled analysis of Western European elections, 1984–1998. European Journal of Political Research, 44(6), 1-19.

21.

Ezrow

de Vries

Steenbergen

Edwards

(2011). Mean voter representation and partisan constituency representation: Do parties respond to the mean voter position or to their supporters? Party Politics, 17(3), 275-301.

22.

Gelman

Pasarica

Dodhia

(2002). Let’s practice what we preach: Turning tables into graphs. American Statistician, 56(2), 121-130.

23.

Green

(2011). A test of core vote theories: The British conservatives, 1997–2005. British Journal of Political Science, 41(4), 735-764.

24.

Hall

P. A.

Taylor

R. C. R.

(1996). Political science and the three new institutionalisms. Political Studies, 44(5), 936-957.

25.

Hirschman

A. O.

(1970). Exit, voice, and loyalty: Responses to decline in firms, organizations, and states. Cambridge, MA: Harvard University Press.

26.

Holmberg

Näsman

Wännström

(2010). VALU 1991–2010—SVT exit poll survey parliamentary election 1991–2010. Gothenburg, Sweden: University of Gothenburg, Department of Political Science, Royal Institute of Technology, Division of Safety Research, and SVT–Sveriges Television.

27.

Hooghe

Bakker

Brigevich

de Vries

Edwards

Marks

Steenbergen

(2010). Reliability and validity of measuring party positions: The Chapel Hill Expert Surveys of 2002 and 2006. European Journal of Political Research, 49(5), 687-803.

28.

Immergut

Anderson

(2008). Historical institutionalism and West European politics. West European Politics, 31(1), 345-369.

29.

Kastellec

J. P.

Leoni

E. L.

(2007). Using graphs instead of tables in political science. Perspectives on Politics, 5(4), 755-771.

30.

Keman

(2011). Third ways and social democracy: The right way to go? British Journal of Political Science, 41(3), 671-680.

31.

Kirchheimer

(1966). The transformation of Western European party systems. In Lapalombara

Weiner

(Eds.), Political parties and political development (pp. 177-200). Princeton, NJ: Princeton University Press.

32.

Kitschelt

(1994). The transformation of European social democracy. New York, NY: Cambridge University Press.

33.

Kitschelt

(1999). European social democracy: Between political economy and electoral competition. Cambridge, UK: Cambridge University Press.

34.

Klingemann

H.-D.

Volkens

Bara

Budge

McDonald

(2006). Mapping policy preferences II: Estimates for parties, electors, and governments in Eastern Europe, the European Union and the OECD, 1990–2003. Oxford, UK: Oxford University Press.

35.

Mikhaylov

(2008). European Parliament Election Study trend file, voter study. Version 0.96 03/05/2008. Retrieved from http://www.tcd.ie/Political_Science/staff/michael_marsh/ees_trend_file.php

36.

Nilsson

Holmberg

Weibull

(2010). National SOM 1986–2008. Gothenburg, Sweden: University of Gothenburg, SOM Institute.

37.

Pierson

(2002). Coping with permanent austerity: Welfare state restructuring in affluent democracies. Revue Francaise de Sociologie, 43(2), 369-406.

38.

Pierson

(2004). Politics in time: History, institutions, and social analysis. Princeton, NJ: Princeton University Press.

39.

Poguntke

(2002). Party organization linkage: Parties without firm social roots? In Luther

K. R.

Müller-Rommel

(Eds.), Political parties in the new Europe (pp. 43-62). New York, NY: Oxford University Press.

40.

Powell

G. B.

Whitten

G. D.

(1993). A cross-national analysis of economic voting: Taking account of the political context. American Journal of Political Science, 37(2), 391-414.

41.

Proksch

S.-O.

Slapin

J. B.

(2006). Institutions and coalition formation: The German election of 2005. West European Politics, 29(3), 540-559.

42.

Przeworski

Sprague

(1986). Paper stones: A history of electoral socialism. Chicago: University of Chicago Press.

43.

Quinn

(2006). Choosing the least-worst government: The British General Election of 2005. West European Politics, 29(1), 169-178.

44.

Safran

(2009). The catch-all party revisited: Reflections of a Kirchheimer student. Party Politics, 15(5), 543-554.

45.

Schmidt

V. A.

(2008). European political economy: Labour out, state back in, firm to the fore. West European Politics, 31(1-2), 302-320.

46.

Schmitt

Scholz

Leim

Moschner

(2010). The Mannheim Eurobarometer trend file 1970–2002. Data set edition 2.01. Mannheim, Germany: Mannheimer Zentrum für Europäische Sozialforschung and Zentrum für Umfragen, Methoden, und Analysen, Universität Mannheim.

47.

Schofield

Sened

(2005). Modeling the interaction of parties, activists, and voters: Why is the political center so empty? European Journal of Political Research, 44(3), 355-390.

48.

Sigelman

(1982). The nonvoting voter in voting research. American Journal of Political Science, 26(1), 47-56.

49.

Steenbergen

Marks

(2007). Evaluating expert surveys. European Journal of Political Research, 46(3), 347-366.

50.

Streeck

Thelen

(2005). Beyond continuity: Institutional change in advanced political economies. Oxford, UK: Oxford University Press.

51.

Tavits

(2007). Principle versus pragmatism: Policy shifts and political competition. American Journal of Political Science, 51(1), 151-165.

52.

Various Principal Investigators. (2010). British election studies. Years 1979–2010. Colchester, UK: University of Essex.

53.

Zentralarchiv für Empirische Sozialforschung, U. Z. K. (2010). Politbarometer 1997–2007 (Partielle Kumulation). Version 1.0.0, 13.04.2010. doi:10.4232/1.2391.

54.

Zohlnhöfer

(2007). The politics of budget consolidation in Britain and Germany: The impact of blame avoidance opportunities. West European Politics, 30(5), 1120-1138.