Winners,losers and the Grand Coalition: Political satisfaction in the Federal Republic of Germany

Abstract

This article examines how electoral outcomes nurture democratic sentiment. Elections generate winners and losers, and voters alter their beliefs in ways that are congruent with electoral choices. Yet, while winning and losing shape democratic sentiment, comparatively less is known about the intermediary influences through which this relationship flows. Analysing data from the Federal Republic of Germany, this article probes for contextually driven effects using cross-level modelling of German General Social Surveys (Allgemeine Bevölkerungs Umfrage der Sozialwissenschaften). Four hypotheses are tested at the individual level, with the results confirming the following. First, winners’ democratic sentiments remain firm in the context of electoral stalemate, but are mediated through their ideological proximity to the party for which they voted. Second, time conditions winners’ democratic sentiments, but in discrepant ways in the east and west of the country. There is no evidence that the western relationship has systematically weakened over a 20-year period. The eastern relationship, meanwhile, was especially strong following unification, but stabilised thereafter. In demonstrating that winners’ attitudes are influenced by political context and time, the article presents a more refined analysis and a more comprehensive account, and develops fresh lines of inquiry into the structure of mass democratic sentiment.

Keywords

Germany grand coalition political context satisfaction with democracy winning and losing

Introduction

Few claims of 20th-century political science have proved as enduringly relevant as the notion that stable democracy depends upon a concordant wellspring of supportive citizen attitudes. Of all the comparative investigations into the mass underpinnings of democracy, however, few variables have predicted the direction and intensity of democratic sentiment with greater accuracy than that of the individual’s party choice in the most recent national election (Anderson and Guillory, 1997: 74). Electoral outcomes commonly distinguish between supporters of the winning and losing parties, and this distinction steers political evaluations in ways that are congruent with voter choices (Clarke and Kornberg, 1992: 185). Losing exacerbates antipathy towards the authorities, while winning stimulates more sanguine attitudes (Kaase and Newton, 1995: 60). Yet, despite its inherent conceptual potential, and its longevity in political science, there has been a disappointing lack of urgency to isolate the additional forces through which winners’ sentiments are mediated (Ginsberg and Weisberg, 1978: 32). It is the aim of this article to redress this.

Extensive research has been conducted into the connections between electoral outcomes and attitudes towards democracy (Almond and Verba, 1963; Norris, 1999a; Pharr and Putnam, 2000). Supporters of the winning candidates or parties are more politically satisfied compared with those who supported the losers – although the distinctions are sharper in majoritarian electoral systems (Anderson and Guillory, 1997: 76). Winners are more trusting of representative institutions (Anderson and LoTempio, 2002: 342; Norris, 1999b: 230). Winners are more likely to believe in the efficacy of political involvement (Clarke and Acock, 1989: 559). And winners are also more supportive of policy output: they are more inclined to believe that government is legislating in ways that align with their interests (Beasley and Joslyn, 2001: 536). Laudable though this research has been, there is a notable deficiency. It has generally divorced the winning phenomenon from the political context in which it is located (Wells and Krieckhaus, 2006: 569–570). In suggesting that winners’ support is largely uniform from one election to another, much of the existing research has neglected the diversity of ways in which context can harden and sharpen winners’ support. In short, extant studies have viewed the relationship as static.

This study, by contrast, suggests that it is dynamic (Nadeau and Blais, 1993: 557). It posits that winners’ support is contingent upon the political context in which it is located, and systematically tests this by analysing two types of effect in the Federal Republic of Germany. The first are short-term effects that are generated by the context of electoral stalemate. Although elections are supposed to generate winners and losers, this is not guaranteed – as the results of the 2005 German Federal Election showed (Saalfeld, 2009: 81). The Red–Green (SPD–Die Grünen (Green Party)) government failed to secure sufficient votes to continue in power, while it was numerically impossible for the opposition Christian Democratic Party–Christian Social Union (CDU–CSU) to form a government with a majority of its own – or, indeed, with its preferred coalition partner, the Free Democratic Party (FDP). A Black–Red Grand Coalition between the Union (CDU–CSU) and SPD was thus formed. But how, if at all, did the context of a Grand Coalition retrospectively alter winners’ support? Did it change supportive attitudes between the winners and losers, and in what circumstances, if any, is there variation among the winners?

The second are long-term effects that are linked to processes of democratisation and modernisation (Nadeau and Blais, 1993: 560). Germany is a particularly useful setting in which to examine these types of effect. Persistent east–west differences in political attitudes (R Campbell, 2011, 2012) enable us to test if easterners’ political satisfaction has been shaped through democratic exposure (Anderson and Tverdova, 2003: 93), and thus extend knowledge about how citizens of new democracies develop the levels of mass legitimacy that are necessary to sustain democratic arrangements (Anderson et al., 2005). But the German context also enables us to examine if long-term processes of modernisation have altered the winning phenomenon in the west of the country. On this logic, modernisation has increased the political sophistication of mass publics and made voters less reliant upon the durable ties to political parties that structure electoral competition (Dalton, 2000: 32; Inglehart, 1981: 881). An important corollary of this is that if partisan commitments are fading as political heuristics, then the supportive attitudes that flow from them may be diminishing over time.

To investigate these issues, the study systematically examines how winning influences political satisfaction, with the analysis sharply concentrated on the individual level. The article utilises data from the Allgemeine Bevölkerungs Umfrage der Sozialwissenschaften (ALLBUS; German General Social Survey). The quantitative analyses comprise multivariate models based on surveys collected between 1988 and 2008. The article is divided into six sections. The first evaluates the relationship between the principal variables and draws hypotheses to test this relationship. The second introduces the data on which the empirical testing is based, leading into a third section outlining the empirical connections between political satisfaction and winning. The fourth section introduces the quantitative procedures used in the analysis. The fifth section presents results and the sixth section draws conclusions.

Political satisfaction and electoral outcomes

A rich diversity of survey measures are used to investigate mass attitudes towards democracy (Dalton, 2004: 24). One is obtained from the question: ‘How satisfied are you with democracy as it is practised in the Federal Republic of Germany?’ The conceptual foundations of this indicator are commonly provided by Easton’s (1965) framework of political support. Easton conceived that above the capricious flux of attitudes towards government performance lies diffuse support for the political community; a basic national attachment, transcending the actions of incumbents, which enables democracy to persist in times of crisis (Easton, 1965: 273). This is followed by regime support for the norms, values and institutions of democracy – as well as more ephemeral, performance-based support for authorities (Fuchs et al., 1995: 330–331). But, as Easton recognised, support for these objects is coterminous and reflexive, the consequence of which is that surveys are an imperfect instrument with which to delineate empirical dimensions of support (Linde and Ekman, 2003: 393). Thus, the form of support that satisfaction with democracy operationalises is the subject of sharp methodological disagreement (Anderson, 2002; Canache et al., 2001). Imprecision in question wording, or variations in respondents’ focus, raises the possibility that mass interpretations are formed from such a disparate catalogue of stimuli that any sense of an aggregate pattern is lost. One study, for example, uncovered heterogeneity of focus in sufficient measure to justify profound methodological concern about the validity and reliability of cross-national comparisons (Canache et al., 2001: 525). Yet, the small range of countries included, together with empirical analyses that were far from clear-cut, led to a reappraisal of the evidence (Anderson, 2002: 2–9), which suggested that mass responses were, in fact, more concrete. And more recent studies, meanwhile, suggest that satisfaction with democracy is formed from a creative synthesis of institutional and performance-based stimuli (Wagner et al., 2009: 32). The present study holds to this latter interpretation and uses this indicator to extend and develop the insights of pre-existing studies.

Winning and losing

These studies contend that elections are a principal avenue through which political sentiment is manipulated. Electoral outcomes act as mechanisms that structure politically relevant attitudes (A Campbell et al., 1960: 100–101; Key, 1961: 189). Underpinning this is the contention that elements of belief systems (attitudes, behaviours and values) are functionally integrated (Rokeach, 1970: 82–83). Beliefs have a diversity of forms and varying degrees of connection, with important sub-domains involving interrelated subjects – including the political (Tetlock, 1989: 135). But within sub-domains, beliefs have integrative complexity: they are cognitively linked, closely related and mutually reinforcing, with the most central and enduring values harnessing them together (Rokeach, 1970: 124). Beliefs within the political sub-domain, for example, have strong partisan coherence (A Campbell et al., 1960: 122), as partisan orientations impose order and meaning on political objects, binding together seemingly disparate political stimuli in ways that provide for varying degrees of structural consistency (Judd and Krosnick, 1989: 100). Electoral outcomes, then, act as opinion-forming cues: political attitudes are configured in ways that are congruent with partisan electoral choices – analogous to a ‘home-team’ effect (Holmberg, 1999: 117–119). Elections raise a partisan screen through which beliefs within the political sphere are refracted (Ginsberg and Weisberg, 1978: 32). The corollary of this is that elections are not just periodic reviews of government conduct; they are highly salient interventions, the outcomes of which prompt re-evaluations of political objects (Citrin, 1974: 978).

Yet, the psychological and rational properties of winning provide a deeper appreciation of why it shapes democratic sentiment. Psychologically, individuals strive to minimise dissonant or non-fitting elements of belief systems (Festinger, 1962). Inconsistency generates ‘cognitive dissonance’, a psychological discomfort in maintaining contradictory ideas (Festinger, 1962: 12–13). The intensity of discomfort is a function of the importance of the beliefs, as well as the importance of the discrepancy between them: it is greatest between highly important but inconsistent beliefs (Beasley and Joslyn, 2001: 522). Dissonance, however, is not merely confined to a cognitive awareness of – or aversion to – its presence; it generates pressure to reduce it (Beasley and Joslyn, 2001: 523). Information is screened, selected and modified on the basis of its compatibility with pre-existing elements of the belief system. There is greater receptivity to information that reinforces previous preferences and hostility to conflicting information. Dissonance, then, produces cognitive reorganisation: individuals decrease in importance, or revise altogether, one or more of the dissonant elements (Festinger, 1962: 18–19). On the basis of this, individuals’ political beliefs should align with their electoral choice. Individuals’ electoral choice is sufficiently salient in character and partisan in structure to alter political attitudes in ways that reduce post-election dissonance (Blais and Gélineau, 2007: 426–427).

Yet, winning also has rational properties. Rational choice approaches emphasise goal-directed attitudes and behaviour reflecting strategic utility maximisation (Downs, 1957: 27). Rational behaviour means choosing the best means to gain a predetermined set of ends. As voters adopt choices reflecting their underlying goals and preferences, as well as the constraints of the situation (Morrow, 1994: 17), they behave in ways that maximise utility in pursuit of purposive goals that have been reshaped by values or preferences. Voters are guided by self-interested calculations of the costs involved and the benefits derived. According to Downs, ‘a rational man always takes the [course of action] which yields him the highest utility … he acts to his own greatest benefit’ (Downs, 1957: 36–37). Post-election rationality can thus be expected to guide political evaluations because electoral outcomes impinge upon assessments of marginal utility. They offer contrasting calculations of system output, and varying opportunities to satisfy wants and needs (Sarsfield and Echegaray, 2005: 156–157). Winners should be most satisfied with the government’s programmatic agenda and have the least incentive to push for institutional change; losers, meanwhile, with little opportunity to realise self-interested calculations, should be less satisfied and have the greatest incentive to push for institutional change.

These psychological and rational properties suggest that there should be contrasting levels of political satisfaction between the winners and losers. Winners should be more politically satisfied compared with losers. Thus:

Hypothesis 1: Winners will be more politically satisfied compared with losers net of the impact of political, economic and demographic variables.

Winning and the Grand Coalition

This analysis contends, however, that winners’ political satisfaction is contingent upon political context. It begins testing this claim by probing for short-term effects driven by the formation of the Grand Coalition. This permits analysis of if and how electoral stalemate alters voters’ perceptions of winning. True, the psychological-rational properties present a compelling theoretical case for why the winning effect should hold, but this is yet to receive empirical verification. And a Grand Coalition may be a sufficiently unusual electoral result to alter the basic psychological and rational properties of winning. Perhaps the fact that the parties won but failed to get over the line without the support of their main rival alters the psychological importance of the result to winners and reduces the cognitive pressure to align individuals’ attitudes with their vote choice. And perhaps electoral stalemate – and the degree of policy concession required by the parties to form the coalition – is sufficient to alter voters’ perceptions of marginal utility in ways that deplete the dividends of political satisfaction that normally accompany winning. These are open questions that we investigate.

But the Grand Coalition also constitutes a particularly interesting context in which to scrutinise the properties of winning. In particular, the psychological properties suggest that political satisfaction should be higher among voters of the CDU–CSU and SPD compared with those who voted otherwise. The structure of political satisfaction should exhibit binary, winner–loser distinctions. But the rational properties also generate expectations that political satisfaction should vary among the winners. If, for example, utility calculations of self-interest vary by party ranking, then contrasting levels of political satisfaction should be found between voters who supported the party that narrowly came first (CDU–CSU) and those who supported the party that came second (SPD). Winners, then, may use party rankings as a partisan guide from which they assess the likelihood of realising self-interested goals, with their support adjusted on the basis of this.

A conceptually deeper proposition, however, is that winners’ political satisfaction is conditioned by their ideological distance to the party for which they voted (Holmberg, 1999: 117). Winners – no less than ordinary voters – are not all equally proximate to their party; some are closer than others. And this, in turn, might impinge upon the levels of support they confer (Citrin, 1974: 976). One way to investigate it, therefore, would be to trace over time how winners’ assessments of party policy positions influence their support, but the inferences of this type of analysis would be confined to discrete policy areas (Borre, 1995: 345). Ideological proximity, by contrast, links broader principles underlying the government’s policy agenda to supportive sentiments (Klingemann, 1979: 219). Ideology is commonly measured through the Left–Right self-placement scale, but this is not without methodological problems (Klingemann, 1979: 233). In particular, scholars have long recognised that the active use of an ideological mode of thought is confined to a minority of the population (A Campbell et al., 1960: 250). Nevertheless, the Left–Right scale has its uses and has implications across a range of policy fields (Sarsfield and Echegaray, 2005: 162). It is also a shorthand indicator of the intensity and extremity of individuals’ political beliefs. Those on the ends of the scale are the most likely to have intense political beliefs that favour far-reaching solutions to social problems, while those in the median position should be more satisfied with the status quo. The ideological distance between voters and their party may thus signify important preferences about how winners’ policy styles and governing priorities shape their support (Anderson et al., 2005: 84–85). Losers with extreme ideological views, for example, should be the least supportive group of all: they are part of the political minority, with little likelihood of having their preferences translated into policy. Ideologically proximate winners, by contrast, closely aligned with the governing party, should be the most mainstream in their political views and may be the most supportive:

Hypothesis 2: Ideologically more proximate winners should be more politically satisfied than less proximate winners.

Democratisation and winning

In addition, the analysis tests if the east–west division of Germany has provided long-term structure to winners’ support. It examines if the relationship between winning and political satisfaction was especially strong in the east immediately following unification. Research suggests that voters of old and new democracies should hold quite different political attitudes (Anderson et al., 2005: 92–93). Anderson et al. (2005: 105), for example, have shown that losing has a negative impact on political support, but that ‘the effect is more pronounced in post-communist countries than in stable democracies of western Europe’. The basis of this claim is that voters’ reactions are conditioned by democratic exposure. Thus, in circumstances in which democracy is nascent, and the esoteric intricacies of electoral processes remain elusive, voters’ reactions should exhibit a degree of volatility, with (at least) three specific reasons cited as to why this may be the case. First, losers in the new system may have been disproportionately drawn from those for whom the old system guaranteed victory. As these voters experience losing for the first time, they may be particularly aggrieved, confer especially low levels of political satisfaction and thereby generate striking winner–loser contrasts. Second, in the immediate post-transition phase, voters’ reactions should not reflect the winning party’s record in government; after all, it has none (or very little) from which to be judged. Rather, its principal contribution has been to sweep away the old system and inaugurate the new, and it may thereby receive a strong (but temporary) fund of support for doing so. In effect, winners’ support for the new system may be especially strong merely because it is not the old system. Third, citizens – and especially winners – new to democracy often have heightened expectations about what the system can deliver, which may temporarily elevate winners’ support (Anderson and Tverdova, 2001: 324). But as voters become familiar with the challenges of governing and consensus-building, so these expectations should become increasingly realistic over time, thereby resulting in less distinctive levels of support between winners and losers. Overall, then, winners’ support in new democracies should be tempered by processes of democratic learning (Nadeau and Blais, 1993: 553). But scholars are less certain about the time frame in which it develops (Anderson et al., 2005: 93). This analysis extends knowledge on this point by investigating if, compared with westerners, easterners’ political satisfaction had sharper winner–loser distinctions in the period following unification, and how quickly this effect faded. Thus:

Hypothesis 3: The individual-level relationship between winning and political satisfaction was stronger in the east compared with the west in the period following unification.

Winning and modernisation

In the west, meanwhile, writers contend that long-term forces of modernisation have altered the political context in which the winning–losing relationship operates (Dalton, 2004: 67–69). On this perspective, socio-structural change has extended the cognitive skills of contemporary citizens and reshaped their basic value priorities (Inglehart, 1997: 311). The effects of this are sufficiently far-reaching to impinge upon the basic partisan heuristics with which individuals orient themselves in the political sphere (Dalton, 2000: 19). In particular, cognitive sophistication has led to a pronounced weakening in the underlying partisan attachments of contemporary electorates. As individuals eschew party attachments, so partisanship is less likely to structure electoral competition (Dalton, 2000: 25). And if partisan orientations are increasingly less likely to structure electoral competition, then they may be less likely to structure the attitudes linked to that competition (Blais and Gélineau, 2007: 426). Gradually, this may mean that the relationship between winning and losing and politically supportive attitudes is weakening. In the long term, winning may count for increasingly less:

Hypothesis 4: The western individual-level relationship between winning and losing and political satisfaction has weakened over time.

Data

To investigate these issues requires data. The empirical testing is conducted by quantitative analyses of ALLBUS, a standardised, biennial survey consisting of face-to-face interviews of the German population collected using clustered random sampling of adults aged 18 years or older. ALLBUS is a cross-sectional survey designed to track trends in mass opinion, demography and socio-economic change in Germany. It has provided a representative sample of individual-level data on West Germany since 1980 and East Germany since 1991. Periodically, questions have been asked about the respondents’ political satisfaction and their party choice in the most recent Federal elections, as well as political, economic and demographic variables. ALLBUS, however, cycles the survey content, meaning that not all surveys in the series contain the same questions.

Winning, losing and political satisfaction

The empirical analysis commences by examining winners’ and losers’ political satisfaction in Germany, beginning in the west. Two clear points emerge from Figure 1. First, winners are more politically satisfied compared with losers, including those who voted for the Black–Red Grand Coalition. But the consistency with which this finding is confirmed is striking when one considers that the first three time points all relate to the same government – a government moreover that held power for 16 years. Persistent dividends of support across so tumultuous a period are remarkably impressive evidence of the importance and durability of winning. But it would be a misreading of the data to infer that winning increments of supportive attitudes are inevitable, for a second point from the figure concerns the only instance in which political satisfaction was indistinguishable between winners and losers. This was in 2002 – towards the end of the first term of Gerhard Schröder’s Red–Green government. The unpopularity of the Red–Green government at this time is well-known, but the sheer scale of its mass unpopularity is underlined by the fact that unlike any government for which there are comparable data, and merely four years after its inauguration, it no longer carried additional dividends of political satisfaction from those who had voted for it. Our first glimpse at the data reveals contrasts between winners’ support: on the one hand, there is evidence of its tenacity and stability; on the other hand, there is evidence of the rapidity with which it can drain away.

In the east, meanwhile, the findings differ in two critical respects. First, at every time point, winners are more satisfied than losers. This holds for the Black–Red Grand Coalition, but it also holds for the Red–Green Coalition, suggesting that eastern winners’ support for the leftist Red–Green Coalition was slightly more tenacious than that of western winners. Second, the starting base of easterners’ political satisfaction is noticeably low. At every time point, for every government, western losers are more politically satisfied compared with eastern winners. Between 1998 and 2008 eastern winners’ political satisfaction fluctuated between 2.73 and 2.76 (on a 0–5 scale), levels between the ‘somewhat dissatisfied’ and ‘somewhat satisfied’ points on the scale. Losers’ political satisfaction, meanwhile, varied between 2.52 and 2.31 – between the same points on the scale. If winning elevates political satisfaction, the degree to which it does so is scarcely sufficient to conceal pervasive levels of political discontent in the territory of the former East Germany.

But this is not the whole story. Separate analyses revealed that political satisfaction varied among the winners of the Grand Coalition. Although not shown in Figures 1 or 2, in 2008, CDU–CSU voters (mean 3.32) had higher levels of political satisfaction compared with SPD voters (mean 3.2). These slight differences (.14) reveal that while winning matters, so does party ranking.

Figures 1 and 2, however, illustrate that winners’ political satisfaction has the potential to vary. Table 1 develops this point by displaying the amount by which political satisfaction differs between winners and losers over time. In the west, the differences diminish. They were sizeable in 1988 (.51), increasing in 1991 (.56), but declining thereafter – by 1998, they were .39; by 2002, they had vanished altogether. Furthermore, although they reappeared in 2008 (.19), they did so modestly and tentatively. Over a 20-year period, then, winner–loser distinctions in political satisfaction have faded incrementally; winning is delivering increasingly less with the passage of time.

Table 1.

East–west differences in political satisfaction between winners and losers, 1988–2008.

	1988Black–Yellow	1991Black–Yellow	1998Black–Yellow	2002Red–Green	2008Black–Red
East	n.a.	.68	.26	.23	.45
West	.51	.56	.39	0	.19

Notes: German nationals only. Black–Yellow denotes coalition government between the CDU–CSU and the FDP; Red–Green denotes coalition government between SPD and Die Grünen; Black–Red denotes ‘Grand Coalition’ between CDU–CSU and SPD. Design weights applied.

Source: ALLBUS (1988, 1991, 1998, 2002, 2008).

The eastern pattern, meanwhile, also provides cautious support to predictions about political context and time. Winner–loser distinctions in political satisfaction were sharpest in the immediate aftermath of unification, but recede as time elapses. The sizeable differences found in 1991 (.68) were a fleeting aberration; they reduced to .26 in 1998, and dropped further to .23 in 2002. Overall, then, between 1991 and 2002, the eastern differences between winners’ and losers’ political satisfaction reduced by almost two-thirds. But the data caution against sweeping conclusions of much-reduced differences since 1991, for by 2008, the differences had recovered to .45. The trajectory in the east, then, is more dynamic than one of simple, monotonous decline.

Methods

The quantitative models develop the analysis. In every model, the dependent variable is political satisfaction. The main independent variable is winning and losing, measured by the ‘zweitestimme’ question, which asks for the party for which respondents voted in the most recent Federal election.¹ More details on the survey questions and recoding procedures are provided in the Appendix.² The analyses are conducted in three stages. In the first stage, regression models establish the effect of winning on political satisfaction, with different models produced over the periods for which there are data. This permits analysis of if and how voters’ response to the Grand Coalition changed compared with that of previous governments. Separate regressions are run for eastern and western Germany. The second stage focuses exclusively on the Grand Coalition. A series of multivariate regression models investigate whether: (1) political satisfaction differs between CDU–CSU and SPD voters; and (2) whether political satisfaction varies by ideological proximity. Ideological proximity was calculated as the difference between the individual’s self-reported placement on the Left–Right scale and their Left–Right placement of the party for which they voted.

The third stage tests for long-term effects. In the west, the analyses examine if there has been long-term change in the strength of the relationship between winning and losing and political satisfaction. This was investigated by creating a time variable for the period covering 1988–2008. A multiplicative interaction term was then computed for winning multiplied by time. In the east, meanwhile, a dummy variable was used to discriminate eastern from western respondents (west coded 0; east coded 1). A multiplicative variable was then built for winning × east and was regressed on the available surveys to examine if there is evidence of cross-regional variation in the effects of winning on political satisfaction.

Results

Stage one: Winning and losing and political satisfaction

Table 2 presents the results of the first stage in the analysis.

Table 2.

OLS multivariate analyses: winner–loser, political satisfaction.

	East				West
	1991	1998	2002	2008	1988	1991	1998	2002	2008
	b (beta)	b (beta)	b (beta)	b (beta)	b (beta)	b (beta)	b (beta)	b (beta)	b (beta)
Constant	−.021	.537^c	.135	.370	2.848^a	1.266^a	1.373^a	1.278^a	1.890^a
Winner–loser (winner = 1)	.521^a (.203)	.264^c (.101)	.196^c (.079)	.277^c (.112)	.437^a (.187)	.236^c (.109)	.206^c (.080)	.120 (.050)	.294^a (.128)
Controls
Current national economy	.240^a (.146)	.434^a (.291)	.295^a (.193)	.331^a (.218)	n.a.	.174^b (.115)	.265^a (.189)	.223^a (.151)	.207^a (.145)
Future national economy	.216^a (.135)	.342^a (.204)	.247^a (.163)	.082 (.047)	n.a.	.070 (.053)	.135^b (.089)	.242^a (.172)	.212^a (.136)
Current personal economy	.314^a (.215)	.075 (.053)	.261^a (.183)	.218^a (.159)	n.a.	.275^a (.187)	.195^a (.131)	.318^a (.205)	.154^a (.118)
Future personal economy	.212^a (.143)	.067 (.034)	.113 (.058)	.345^a (.207)	n.a.	.226^b (.130)	.175^b (.102)	.020 (.011)	−.037(−.023)
Age	.007^c (.083)	.002 (.031)	.004 (.058)	.003 (.041)	.010^a (.174)	.009^a (.148)	.003 (.002)	−.003 (−.040)	.002 (.038)
Sex (male = 1)	.157^c (.064)	.022 (.009)	.091 (.038)	−.048 (−.020)	.135 (.062)	.096 (.044)	.024 (.010)	.032 (.013)	.052 (.023)
Education (degree = 1)	−.259^c (−.065)	.125 (.030)	−.185(−.052)	.037 (.009)	.175 (.035)	−.039 (−.009)	−.132 (−.026)	.142 (.037)	.128 (.038)
Income	−.114 (−.054)	.020 (.014)	.138^b (.120)	−.017 (−.013)	.069 (.060)	−.053 (−.054)	.006 (.006)	.059 (.063)	−.001 (−.002)
Adj. R²	.23	.20	.20	.20	.09	.13	.13	.15	.10
N	936	546	660	773	1006	619	950	1206	1394

Notes: German nationals only. Design weights applied. Dependent variable = political satisfaction. ^aSig. at .0005 level; ^bsig. at .005 level; ^csig. at .05 level.

Source: ALLBUS (1988, 1991, 1998, 2002, 2008).

The table can be neatly summarised as follows: winning increases political satisfaction. In the west, the 2002 results notwithstanding, winners are more politically satisfied compared with losers, including winners of the Grand Coalition. Moreover, the relationship appears to be reasonably strong. Winning increased political satisfaction by .437 in 1988, by .236 in 1991, by .206 in 1998, and by .294 in 2008, holding all other variables in the models constant. Further, the standardised beta coefficients confirm that winning is one of the principal influences on political satisfaction. Indeed, in one out of the five analyses (1988, beta = .187) winning exerts the greatest relative effect on political satisfaction of all the variables included in the models – although fewer control variables were available for this model.

In the east, the picture is not quite identical. Overall, however, winning increases political satisfaction, including for the Grand Coalition (b = .277). But the eastern findings differ from the western results in two key respects. First, the winning effect held for the 2002 analysis of the Red–Green government (b = .196). On this evidence, eastern winners were slightly more supportive of the leftist Red–Green government compared with westerners. Second, the beta coefficients for winning are smaller compared with those for economic perceptions, indicating that eastern political satisfaction is not especially politicised; it is more heavily steeped in economic perceptions. Yet, while the beta coefficients cannot be compared across the models, one cannot but notice that the relatively high ranking winning obtained in 1991 (beta = .203) was not replicated in any of the subsequent models. This, together with the sharp winner–loser distinctions uncovered in the east in 1991 (Table 1), may be cautiously indicative of a changing relationship.

Overall, winner–loser distinctions in political satisfaction obtain to most governments, including the Grand Coalition. But we construct further models to probe more deeply into voters’ response to the Grand Coalition, the results of which follow.

Stage two: Grand Coalition and ideological proximity

Table 3 analyses winners’ political satisfaction under the Grand Coalition.

Table 3.

Multivariate analyses, grand coalition.

	East		West
	Model 1	Model 2	Model 3	Model 4
	b (beta)	b (beta)	b (beta)	b (beta)
Constant	5.344^a	5.312^a	4.725^a	4.674^a
Main effects
Voted Union (CDU–CSU)	.340^b (.128)	.429^c (.162)	.236^b (.108)	.258^b (.118)
Voted SPD	.249^c (.098)	.401^b (.157)	.168^c (.075)	.293^a (.131)
Distance from Union	−.012 (−.023)	−.015 (−.028)	.002 (.004)	.002 (.005)
Distance from SPD	.068^c (.114)	.063^c (.105)	.006 (.012)	.005 (.01)
Interaction effects
Voted Union × distance from Union	–	−.060 (−.044)	–	−.019 (−.015)
Voted SPD × distance from SPD	–	−.131^c (−.095)	–	−.104^b (−.096)
N	579	579	1161	1161
Adj. R²	.18	.19	.09	.10

Notes: German nationals only. Dependent variable = political satisfaction. The models controlled for: age; sex; income; current national economic evaluations; future national economic evaluations; current personal economic evaluations; and future personal economic evaluations. ^aSig. at .0005 level; ^bsig. at .005 level; ^csig. at .05 level.

Source: ALLBUS (2008).

Table 3 provides two additional pieces of evidence about the properties of winners’ political satisfaction under the Grand Coalition. First, party ranking influences winners’ support. In the west (Model 3), voters who supported the Union (CDU–CSU) were more politically satisfied (beta = .108) compared with SPD voters (beta = .075). Interestingly, the same ranking order is found in the east (Model 1). Although, then, both parties formed the government, there were dissimilar responses from their supporters. Union voters were more satisfied compared with those of the SPD. A second point, however, is that winning appears to be influenced by ideological proximity. Again, the effect is found in both east and west (Models 2 and 4). But, intriguingly, it is limited to SPD voters. As the negative coefficients denote more proximate SPD voters, they provide us with a cautious basis from which to posit that more proximate SPD voters are, in fact, more politically satisfied compared with less proximate SPD voters. Accordingly, Hypothesis 2 is accepted.

Stage three: Winning and modernisation

The analyses now present results for the long-term effects, beginning in the west. Overall, Table 4 does not provide strong evidence that time is systematically weakening the influence of winning on political satisfaction in western Germany. Turning to the first model, this confirms the effects of winning and losing and time on political satisfaction. Winning boosts political satisfaction (b = .422). However, the effect of time on political satisfaction is negative: as time increases, political satisfaction diminishes (b = −.099). Model 2, however, does not suggest that the relationship between winning and political satisfaction has weakened. Our interaction effect has failed to attain significance (b = .023). As a check of this finding, separate models analysed whether there were ‘fixed effects’ in each of the years. The coefficients altered in size, but the substantive results held. There was no evidence of a long-term, systematic weakening of the relationship between winning and losing and political satisfaction. Accordingly, Hypothesis 3 is rejected.

Table 4.

Regression analyses, winning and losing, time, and political satisfaction (West Germany).

	Model 1		Model 2
	b (s.e.)	Beta	b (s.e.)	Beta
Constant	3.508^a (.022)		3.478^a (.027)
Main effects
Winner–loser (winner = 1)	.422^a (.021)	.176	.500^a (.045)	.208
Time (1988–1998)	−.099^a (.006)	−.152	−.090a (.008)	−.137
Interaction effects
Winner–loser × time (1988–2008)	–	–	−.023 (.012)	−.041
N	12722		12740
Adj. R²	.05		.05

Notes: German nationals only. Design weights applied. Dependent variable = political satisfaction. ^aSig. at .0005 level; ^bsig. at .005 level; ^csig. at .05 level.

Source: ALLBUS (1998, 1991, 1998, 2002, 2008).

However, as is shown in Table 5, the eastern results confirm that democratisation has mediated winners’ political satisfaction. There was an initially strong relationship between winning and losing on political satisfaction in the period following unification. In 1991, for example, winning (b = .503) and east–west (b = −.835) had sizeable effects on political satisfaction. Winning boosted political satisfaction by .5, but easterners were less satisfied compared with westerners. Notice, however, that the interaction term (winning–loser × east–west) also surpassed the significance threshold (b = .224). This reports the additional amount by which being an eastern winner altered political satisfaction compared with a western winner in 1991. As the coefficient is positive, eastern winners’ political satisfaction increased by an additional .224 in 1991 compared with that of western winners. Thus, the relationship between winning and political satisfaction was, in fact, stronger in the east compared with the west in the period following unification. As shown in the remaining analyses, however, this effect faded quickly. Indeed, by 1992, the relationship between winning and political satisfaction was of indistinguishable strength between east and west Germans. Accordingly, Hypothesis 4 is confirmed.

Table 5.

Winner–loser, political satisfaction, east–west.

	1991		1992		1998		2002		2008
	b (s.e.)	Beta	b (s.e.)	Beta	b (s.e.)	Beta	b (s.e.)	Beta	b (s.e.)	Beta
Constant	3.353^a (.037)		2.950^a (.031)		3.077^a (.031)		3.023^a (.036)		3.149^a (.038)
Winner–loser (winner = 1)	.503^a (.060)	.197	.448^a (.055)	.123	.373^a (.055)	.140	.167^b (.058)	.065	.334^a (.049)	.137
East–west (east = 1)	−.835^a (.052)	−.341	−.791^a (.054)	−.283	−.729^a (.054)	−.279	−.632^a (.057)	−.240	−.840^a (.061)	−.325
Winner–loser × east–west	.206^c (.087)	.062	.112 (.098)	.026	.008 (.099)	.002	.188 (.105)	.046	.120 (.086)	.036
Adj. R²	.16		.11		.10		.06		.12
N	2893		3372		3027		2678		3204

Notes: Entries in table are unstandardised coefficients; standard errors in parentheses. Design and population weights applied. German nationals only. Dependent variable = political satisfaction. ^aSig. at .0005 level; ^bsig. at .005 level; ^csig. at .05 level.

Source: ALLBUS (1991, 1992, 1998, 2002, 2008).

The evidence, then, can be summarised as follows. Winning increases political satisfaction with remarkable consistency across a range of governments examined and over a time period spanning most of the unified Federal Republic. But it does not do so in every instance; it is contingent upon political context. It is shaped by party fortunes and by ideological proximity. And the evidence suggests that although the relationship was noticeably strong in the east in the immediate aftermath of unification, it quickly stabilised as easterners moved into alignment with westerners. These findings have implications for German governance and for future research on democratic sentiment.

Conclusion

This study has important implications. At a theoretical level, it suggests that mass democratic sentiment is politicised, but not one-dimensional. Although consistent winning dividends were found over a near 20-year period, winning and losing are not merely binary categories into which voters fit. On the evidence presented here, the intensity of their sentiment has been: coloured by the legacy of the east–west division; shaded by ideological proximity; and given form and substance by time. Winners’ support, then, is neither a singular condition nor a linear process. It involves diverse domains of activity, which intensify, sharpen and soften support in ways that ensure that it rarely reaches a steady state of equilibrium. While, therefore, comparative studies of mass democratic sentiment have made fruitful insights by discriminating voters’ political allegiances, this research suggests that these studies can yield an ever-richer appreciation by sensitising the research to the contextual and temporal circumstances in which these allegiances are located.

True, the findings must be carefully qualified. Any approach depending exclusively upon cross-sectional data, even if it employs multiple surveys, cannot be definitive. The availability of more surveys will present an opportunity to extend and develop the insights of this study in a wider comparative perspective. And although the models are premised upon a connection between voter choice and political satisfaction, the causal order could well be in the reverse direction: political satisfaction could condition voter choice. Or the two phenomena could be analytically entwined in ways that inhibit disentangling the direction of causality with clear-cut precision. Despite these caveats, the analyses have shown how political context is a critical tributary through which winners’ supportive sentiments flow. The analysis has not only provided a sweeping panorama of the period spanning unified Germany, but focused in on the Grand Coalition, showing that supportive sentiments were extended to the coalition, but varied by party ranking and ideological proximity.

This has implications for German governance. Ideological proximity is an important link between the governing agenda of the Grand Coalition and the supportive sentiments of those who voted for the coalition. Insofar as ideological proximity is revised on the basis of policy agenda, the range and scope of the coalition output will feed back to deplete or augment its support. Overall, then, winners’ support is no different from that of ordinary voters: governments have to earn it. They cannot take it for granted; it is revised, reinforced or removed in conformity with voters’ perceptions of performance. But as key themes of the coalition agenda were welfare retrenchment and budget austerity, and as the effect is confined to SPD voters, it was likely to develop policy solutions that not only failed to command popular support, but alienated those who voted for it. Mass sentiment, then, is at the heart of the deep and reinforcing challenges of contemporary German governance: far-reaching domestic welfare reform depends upon policy solutions that mobilise mass consensus; yet attempts to enact change of the breadth and depth required lead mass support to drain away. And this, in turn, creates circumstances increasingly less conducive to reform: it makes it ever harder to reach a binding consensus and leads decision-making to become more protracted. Mass sentiment, domestic reform and governance, then, have the potential to become closely interlocked components in a descending spiral that reinforces policy gridlock.

Yet, the most fundamental implications concern democratic learning. As easterners’ political satisfaction was deeply conditioned by the context of unification, but then swiftly fell into alignment with that of the west, it suggests that easterners did not have to experience winning and losing over the course of multiple elections in order for the volatility to dissipate. Once was enough. This raises an important point about democratic learning in post-transitional countries: while it may have developed at an uneven pace throughout the eastern bloc, the evidence here shows the rapidity with which some of the politicised aspects of it developed in eastern Germany (Rohrschneider, 1999). But this also has broader theoretical implications about governing styles in post-transitional democracies, for it suggests that in post-transition circumstances, there is an initially sizeable fund of goodwill upon which governments can draw. Government, however, must be cautious about what it chooses to do on the basis of this support, for it is double-edged. On the one hand, it confers sufficient leeway to enable policymakers to enact the far-reaching structural and institutional reforms that may be required. On the other hand, however, it can rapidly drain away in highly charged political circumstances. In other words, winners’ democratic sentiments are especially sensitive to context, points that have been reinforced by the research for this article.

Footnotes

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

Notes

Author biography

Ross Campbell is Senior Lecturer in Political Science at Glasgow Caledonian University, UK. Having obtained a PhD from the University of Strathclyde, he was post-doctoral fellow at the WZB (Social Science Center, Berlin), where he researched the nature, structure and trajectory of mass attitudes towards democracy. More recently, his research interests have broadened to include the changing repertoire of citizens’ political participation and the structure of political inequality. His recent publications have appeared in European Journal of Political Research, West European Politics and German Politics.

References

Almond

Gabriel

Verba

Sidney

(1963) The Civic Culture: Political Attitudes and Democracy in Five Nations. Boston, MA: Little, Brown and Company.

Anderson

Christopher J

(2002) Good questions, dubious inferences, and bad solutions: Some further thoughts on satisfaction with democracy. Research paper 116, Binghamton Center for Democratic Performance, State University of New York.

Anderson

Christopher J

Guillory

Christine

(1997) Political institutions and satisfaction with democracy. American Political Science Review 91(2): 66–81.

Anderson

Christopher J

LoTempio

Andrew J

(2002) Winning, losing and political trust in America. British Journal of Political Science 32(2): 335–351.

Anderson

Christopher J

Tverdova

Yuliya V

(2001) Winners, losers, and attitudes about government in contemporary democracies. International Political Science Review 22(4): 321–338.

Anderson

Christopher J

Tverdova

Yuliya V

(2003) Corruption, political allegiances, and attitudes toward government in contemporary democracies. American Journal of Political Science 47(1): 91–109.

Anderson

Christopher J

Blais

Andrè

Bowler

Shaun

. (2005) Losers’ Consent: Elections and Democratic Legitimacy. Oxford: Oxford University Press.

Beasley

R.K

Joslyn

M.R

(2001) Cognitive dissonance and post-decision attitude change in six presidential elections. Political Psychology 22(3): 521–540.

Blais

Andrè

Gélineau

(2007) Winning, losing and satisfaction with democracy. Political Studies 55(2): 425–441.

10.

Borre

Ole

(1995) Scope-of-government beliefs and political support. In: Borre

Ole

Scarbrough

Elinor

(eds) The Scope of Government. Oxford: Oxford University Press, pp. 343–366.

11.

Campbell

Angus

Converse

Philip E

Miller

Warren E

. (1960) The American Voter. London: The University of Chicago Press Ltd.

12.

Campbell

Ross

(2011) Socialist values and political participation: A barrier to ‘inner unity’? West European Politics 34(2): 362–383.

13.

Campbell

Ross

(2012) Values, trust and democracy in Germany: Still in search of ‘inner unity’. European Journal of Political Research 51(3): 646–670.

14.

Canache

Damarys

Mondak

Jeffery J

Seligson

Mitchell A

(2001) Meaning and measurement in cross-national research on satisfaction with democracy. Public Opinion Quarterly 65(4): 506–528.

15.

Citrin

Jack

(1974) The political relevance of trust in government. American Political Science Review 68(3): 973–988.

16.

Clarke

Harold D

Acock

Alan C

(1989) National elections and political attitudes: The case of political efficacy. British Journal of Political Science 19(4): 551–562.

17.

Clarke

Harold D

Allan

Kornberg

(1992) Do national elections affect perceptions of MP responsiveness? A note on the Canadian case. Legislative Studies Quarterly 17(2): 183–204.

18.

Dalton

Russell J

(2000) The decline of party identifications. In: Dalton

Russell J.

Wattenberg

Martin

(eds) Parties without Partisans: Political Change in Advanced Industrial Democracies. Oxford: Oxford University Press, pp. 19–36.

19.

Dalton

Russell J

(2004) Democratic Challenges, Democratic Choices: The Erosion of Political Support in Advanced Industrial Democracies. Oxford: Oxford University Press.

20.

Downs

Anthony

(1957) An Economic Theory of Democracy. Boston, MA: Addison-Wesley Publishing Inc.

21.

Easton

David

(1965) A Systems Analysis of Political Life. New York, NY: John Wiley and Sons.

22.

Festinger

Leon

(1962) A Theory of Cognitive Dissonance. Stanford, CA: Stanford University Press.

23.

Fuchs

Dieter

Guidorossi

Giovanna

Svensson

Palle

(1995) Support for the democratic system. In: Klingemann

Hans Dieter

Fuchs

Dieter

(eds) Citizens and the State. Oxford: Oxford University Press, pp. 323–353.

24.

Ginsberg

Benjamin

Weisberg

Robert

(1978) Elections and the mobilization of popular support. American Journal of Political Science 22(1): 31–55.

25.

Holmberg

Sören

(1999) Down and down we go: Political trust in Sweden. In: Norris

Pippa

(ed.) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press, pp. 103–122.

26.

Inglehart

Ronald

(1981) Post-materialism in an environment of insecurity. American Political Science Review 75(4): 880–900.

27.

Inglehart

Ronald

(1997) Modernization and Postmodernization. Princeton, NJ: Princeton University Press.

28.

Judd

Charles M

Krosnick

Jon A

(1989) The structural bases of consistency among political attitudes: Effects of political expertise and attitude importance. In: Pratkanis

Anthony R

Breckler

Steven J

Greenwald

Anthony G

(eds) Attitude Structure and Function. Hillsdale, NJ: Lawrence Erlbaum Publishers Inc., pp. 99–128.

29.

Kaase

Max

Newton

Kenneth

(1995) Beliefs in Government. Oxford: Oxford University Press.

30.

Key

Valdimer O

(1961) Public Opinion and American Democracy. New York, NY: Alfred A Knopf Publishers Inc.

31.

Klingemann

Hans-Dieter

(1979) Measuring ideological conceptualisation. In: Barnes

Samuel

Kaase

Max

(eds) Political Action: Mass Participation in Five Western Democracies. London: Sage, pp. 215–254.

32.

Linde

Jonas

Ekman

Joakim

(2003) Satisfaction with democracy: A note on a frequently used indicator in comparative politics. European Journal of Political Research 42(3): 391–408.

33.

Morrow

James D

(1994) Game Theory for Political Scientists. Princeton, NJ: Princeton University Press.

34.

Nadeau

Richard

Blais

Andrè

(1993) Accepting the election outcome: The effect of participation on losers’ consent. British Journal of Political Science 23(4): 553–563.

35.

Norris

Pippa

(1999a) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press.

36.

Norris

Pippa

(1999b) Institutional explanations for political support. In: Norris

(ed.) Critical Citizens: Global Support for Democratic Governance. Oxford: Oxford University Press, pp. 217–235.

37.

Pharr

Susan J

Putnam

Robert D

(2000) Disaffected Democracies: What’s Troubling the Trilateral Countries? Princeton, NJ: Princeton University Press.

38.

Rohrschneider

Robert

(1999) Learning Democracy: Democratic and Economic Values in Unified Germany. Oxford: Oxford University Press.

39.

Rokeach

Milton

(1970) Beliefs, Attitudes and Values. San Francisco, CA: Jossey-Bass Inc. Publishers.

40.

Saalfeld

Thomas

(2009) The German party system since 1998: Cooperation and competition under growing uncertainty. In: Miskimmon

Alister

Paterson

William E

Sloam

James

(eds) Germany’s Gathering Crisis: The 2005 Federal Election and the Grand Coalition. Basingstoke: Palgrave Macmillan, pp. 80–105.

41.

Sarsfield

Rodolfo

Echegaray

Fabiàn

(2005) Opening the black box: How satisfaction with democracy and its perceived efficacy affect regime preference in Latin America. International Journal of Public Opinion Research 18(2): 153–173.

42.

Tetlock

Philip E

(1989) Political belief systems. In: Pratkanis

Anthony R

Breckler

Steven J

Greenwald

Anthony G

(eds) Attitude Structure and Function. Hillsdale, NJ: Lawrence Erlbaum Publishers Inc., pp. 129–151.

43.

Wagner

Alexander F

Schneider

Friedrich

Martin

Halla

(2009) The quality of institutions and satisfaction with democracy in Western Europe – a panel analysis. European Journal of Political Economy 25(1): 30–41.

44.

Wells

Jason M

Krieckhaus

Jonathan

(2006) Does national context influence democratic satisfaction? A multi-level analysis. Political Research Quarterly 59(4): 569–578.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.07 MB

0.01 MB