Jump to ContentJump to Main Navigation
Critical CitizensGlobal Support for Democratic Government$

Pippa Norris

Print publication date: 1999

Print ISBN-13: 9780198295686

Published to Oxford Scholarship Online: November 2003

DOI: 10.1093/0198295685.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in OSO for personal use.  Subscriber: null; date: 17 September 2019

The Economic Performance of Governments

The Economic Performance of Governments

Chapter:
(p.188) 9 The Economic Performance of Governments
Source:
Critical Citizens
Author(s):

Ian McAllister (Contributor Webpage)

Publisher:
Oxford University Press
DOI:10.1093/0198295685.003.0009

Abstract and Keywords

A major conclusion of this book is that there are few consistent trends in popular support for the political community but there is high and perhaps even growing support for democratic values and declining support for regime institutions and political leaders. Earlier chapters have traced these patterns with respect to democracy worldwide, and, in particular, regions of the world. This one focuses specifically on the role of public policy in shaping popular support for democratic institutions among OECD countries. The data are aggregate indicators across twenty‐four of the twenty‐nine member countries of the OECD in 1997, as well as individual‐level data from the same twenty‐four countries based on the 1990–1 World Values Survey. The different sections of the chapter are: Explaining Institutional Confidence; Evidence for Trends in Institutional Confidence; The Role of Democratic Experience; Policy Outputs and Confidence: A Macro Analysis; Policy Outputs and Confidence: A Micro Analysis; and Discussion and Conclusions.

Keywords:   democracy, democratic experience, democratic institutions, democratic values, economic performance, government performance, institutional confidence, OECD countries, policy outputs, political leaders, popular support, public policy

ONE major conclusion of this book is that there are few consistent trends in popular support for the political community but there is high and perhaps even growing support for democratic values and declining support for regime institutions and political leaders. An earlier chapter by Hans‐Dieter Klingemann has traced these patterns with respect to democracy world‐wide, while Russell Dalton, William Mishler and Richard Rose have confirmed the existence of these patterns in particular regions of the world. This chapter focuses specifically on the role of public policy in shaping popular support for democratic institutions among OECD countries. The data are aggregate indicators across 24 of the 29 member countries of the OECD in 1997, as well as individual‐level data from the same 24 countries based on the 1990–1 World Values Survey.1

The role of public policy in shaping electoral outcomes—most particularly with regards to government economic performance—has long been a major preoccupation of political science. Aside from the voter attributes of party identification or social group membership, economic management has traditionally been viewed as the single most important factor that can deliver electoral success or failure. By contrast, studies focusing on the role of the economy in shaping confidence in democratic institutions are comparatively rare (but cf. Clarke, Dutt, and Kornberg, 1993; Finke, Muller, and Seligson 1987), and almost no attention has been devoted to the impact of the broader policy outputs of government. Most studies which examine confidence have focused on aggregate over‐time trends, usually concentrating on democracy as a value and tracing changes among particular social, political, and national groups (Listhaug and Wiberg 1995; Fuchs, Guidorossi, and Svensson 1995; Miller and Listhaug 1990).

Yet if economic performance has a major impact on electoral outcomes, as we know, then it should also have some influence on popular confidence in democratic (p.189) institutions. Equally, social policy outputs may have some influence in shaping confidence. The impact of public policy on institutional confidence is likely to be greatest where there is widespread economic dissatisfaction for a prolonged period of time, as was the case in many of the advanced democracies following the Arab oil embargo in 1973. The embargo initiated a prolonged period of ‘stagflation’ that most governments were unable to quell until the early 1980s, and which had significant consequences for government support (Weatherford 1984). It should also be apparent in newly democratic societies, where civil society is underdeveloped and where many citizens fail to make a clear distinction between democratic governments and the democratic institutions of the state. Many Central and Eastern European countries have experienced severe economic dislocation, just at the time when they are endeavouring to generate popular support for their embryonic democracies.

Explaining Institutional Confidence

One explanation why government policy outputs in general, and the economy in particular, has received relatively little attention in studies of institutional confidence is the view, originating in the earliest empirical studies of democracy, that popular support for democratic norms has its origins in political culture. A more recent view identifies economic performance as the main‐spring for how citizens evaluate their governments and, in turn, for how they rate their democratic institutions. Many studies have demonstrated the pivotal role of popular economic expectations in deciding the fate of incumbent governments in established democracies (Lewis‐Beck 1988; Fiorina 1981). Governments that are perceived as delivering wealth and prosperity to their voters are more likely to be re‐elected, while those that are perceived as depressing economic conditions through their policies are likely to be replaced. In general, it is collective rather than individual judgements that have most weight in the popular economic calculus, and those judgements are usually (though not exclusively) retrospective rather than prospective.2

In evaluating the influence of economic conditions on political behaviour, there is a crucial distinction between economic perception and economic reality. Considerable evidence suggests that voters believe that it is a central responsibility of government to deliver high levels of economic performance. Since economic performance is judged by collective (sociotropic) rather than individual (egocentric) criteria, popular perceptions about the economy are shaped principally by the mass media and through an assessment of national economic conditions, and less by individual economic circumstances. Since these perceptions are mainly collective, they relate directly to governments and also, to some degree, to political institutions. But the linkages remain (p.190) unclear, as do the specific circumstances in which economic performance and expectations will impact on the direction and strength of citizen beliefs about democratic institutions.

The conventional way in which citizens deal with an economic performance deficit in a democratic polity is to assign responsibility for the problem to the government of the day and to vote them out of office. The economic performance hypothesis is predicated on the assumption that citizens distinguish between the role of the government and the role of the political system. This is more likely to occur in the established democracies, where citizens make a clear distinction between the operation of the system as a whole, and the activities of particular party governments. In the newer democracies, which often do not enjoy widespread popular support, or where support is unevenly distributed across social groups, such distinctions are more likely to be blurred since the ‘reservoir of good will’ that allows citizens to accept performance deficits does not exist. However, a more generalized loss of confidence is possible in an established democracy where the economic performance deficit is more intense and sustained (Fuchs, Guidorossi, and Svensson 1995: 327). The policy hypothesis we will explore therefore identifies a performance deficit as the explanation for institutional confidence, through changes in macro and micro socio‐economic circumstances that affect popular perceptions of collective socio‐economic conditions.

There are, of course, a range of difficulties in evaluating and measuring the policy performance of governments (Bok 1997; Klingemann, Hofferbert, and Budge 1994). Public opinion may be used as the main measure evaluating policy performance, but opinions often change for reasons unrelated to policy. Moreover, popular satisfaction with government policy is mediated by expectations, which also vary according to external criteria: a major policy success when expectations are high will have less impact than a more modest success, delivered when popular expectations are low. One alternative is to evaluate the outcomes of specific policies, aggregated across several major areas, such as health, welfare, or education. This approach, too, has significant disadvantages, most notably changes in external conditions which may affect programme performance, as well as the inability to make evaluative judgements (Bok 1997). For these reasons two sources of data are used in the chapter, aggregate‐level indicators of policy performance across the OECD countries, and individual‐level public opinion data from the 1990–1 World Values Survey. These two perspectives enable a more informed judgement to be made of the impact of public policy on institutional confidence.

Evidence for Trends in Institutional Confidence

‘Support’ and ‘confidence’ in democratic institutions are used interchangeably since they convey a broad meaning concerning the links between popular (p.191) beliefs about government and representative institutions. As discussed in the Introduction, the concept of support for democratic institutions can be defined in various ways (for a review, see Fuchs, Guidorossi, and Svensson 1995; Kornberg and Clarke 1992). Narrowly defined, the object of support can mean the political system itself, encompassing parliaments and the bureaucracy. A broader definition might take in the institutions that preserve public order, namely the legal system, police, and armed forces. More broadly still, the concept of democratic institutions could encompass the core institutions of civil society, such as the educational system, the mass media, the trade unions and major companies.3 The narrower definition based on parliaments and the bureaucracy is used here for two reasons. First, it is less ambiguous, particularly where cross‐national analyses are concerned. Secondly, in newly democratic societies, the democratic institutions are associated most directly with representative government, and much less so with the general institutions of civil society which are often underdeveloped.4

Arriving at an appropriate list of democracies to include in the analysis presents several difficulties. First, many of the democracies currently in existence have very different levels of socio‐economic development. India, for example, has been a democracy since 1949, yet it ranks close to, or at the bottom of, any measure of economic development. Secondly, some of the current democracies only recently achieved that status, which raises technical difficulties in matching the survey results with the time at which the country became a competitive democracy. A third problem concerns the items included in the 1990–1 World Values Survey. Although the survey contains a wide range of questions and covers over 40 countries,5 not all of the items were asked in every country. To arrive at a suitable list of countries for analysis it was decided to take the 29 countries which were OECD members in 1997, and for whom adequate data were available in the World Values Survey; this results in the list of 24 countries shown in Table 9.1.

The two items used in the survey to measure institutional confidence relate to confidence in parliament and the civil service. Table 9.1 shows considerable variation in levels of confidence across the 24 countries, with Poland registering a mean value of 6.6 for confidence in parliament (on a 0–10 scale) to 3.7 for Mexico. In most countries parliament is rated at a similar level to the civil service. The main exceptions are countries where there is stronger confidence in the civil service than in the parliament: this occurs in the United States, South Korea, and Hungary. The average correlation between the two items is 0.53, ranging from a low of 0.39 in Korea to a high of 0.80 in Japan. There is, (p.192) then, a wide range of variation in popular support for democratic institutions across the 24 countries under examination.

Much of the literature on confidence in democratic institutions assumes that there has been a decline; the assumption is that the challenges to representative democracy have been sufficiently severe both in their frequency and intensity to erode popular institutional confidence. Fortunately, we can provide a partial test of this hypothesis since the first round of the World Values Survey, conducted among a more limited range of countries in 1981, also asked about confidence in institutions. Although there are only two time‐points, 1980 and 1991, the results are clear. Trends across the 19 OECD countries for which over‐time data exist show that, overall, popular confidence in democratic institutions has registered a modest decline, from a mean of 4.8 in 1981, to 4.5 in 1990–1. This provides further confirmation of the analysis presented earlier by Dalton (Chapter 3).

Table 9.1. Confidence in Political Institutions

Rank

Country

Confidence (0–10)

Parliament

Civil Service

(r)

(N)

1.

Poland

6.6

6.7

(.73)

(938)

2.

Turkey

5.5

5.2

(.52)

(1,030)

3.

Norway

5.3

4.6

(.44)

(1,239)

4.

Iceland

5.1

4.8

(.43)

(702)

5.

Ireland

5.1

5.4

(.53)

(180)

6.

Germany

5.0

4.4

(.51)

(3,075)

7.

Netherlands

4.9

4.7

(.45)

(522)

8.

Britain

4.8

4.8

(.44)

(2,805)

9.

Sweden

4.8

4.6

(.64)

(1,047)

10.

United States

4.7

5.6

(.46)

(2.010)

11.

Denmark

4.6

5.0

(.47)

(155)

12.

France

4.5

4.6

(.43)

(2,806)

13.

Austria

4.5

4.6

(.62)

(1,460)

14.

Czechoslovakia

4.5

4.0

(.45)

(1,396)

15.

Belgium

4.4

4.4

(.46)

(307)

16.

Canada

4.4

4.9

(.48)

(1,729)

17.

Hungary

4.2

4.8

(.57)

(999)

18.

Finland

4.1

4.2

(.53)

(588)

19.

Spain

4.1

4.0

(.54)

(2,224)

20.

Japan

4.0

4.2

(.80)

(2,193)

21.

South Korea

3.9

5.5

(.39)

(1,251)

22.

Portugal

3.9

4.0

(.63)

(478)

23.

Italy

3.8

3.3

(.60)

(2,307)

24.

Mexico

3.7

3.3

(.62)

(1,531)

MEAN

4.5

4.6

(32,847)

Note: The question was: ‘Please look at this card and tell me, for each item listed, how much confidence you have in them, is it a great deal, quite a lot, not very much or none at all?’ Scaled from 0 (low) to 10 (high).

Source: 1990–1 World Values Survey.

(p.193) In the G7 countries, however, Figure 9.1 shows that there has been a change in only two countries—the USA and France—while institutional confidence in the remaining five countries has remained stable. The results for the G7 countries tend to emphasize the Fuchs, Guidorossi, and Svensson (1995) conclusion that confidence has remained remarkably unchanged in the established democracies, despite a variety of challenges.6

The decline in the US is confirmed by other trend studies which have shown a consistent weakening in confidence in Congress throughout the 1970s and early 1980s (Lipset and Schneider 1987). Some suggest that the mainsprings of this decline were the widespread opposition to the Vietnam War and the Watergate crisis in the late 1960s and early 1970s, both of which had major effects on political trust (Abramson 1983; Nye et al. 1997). Nevertheless, the results demonstrate the wide variation in institutional confidence that exists across the G7 countries. Although there has been a decline in the USA, it is still almost 50 per cent higher than trust in Italy, the lowest among the G7 countries. The Italian results underline the consequences of widespread political corruption and the accompanying loss of confidence in the political class. These survey results date from 1990; the 1994 Italian elections caused a major upheaval in the party system, reflecting the low standing of democratic institutions.

                      The Economic Performance of Governments

Fig. 9.1. Trends in institutional confidence in the G7 countries, 1981–1990.

Sources: 1981 and 1990–1 World Values Surveys.

(p.194) The overall downward trend in institutional confidence across the 19 countries included in the over‐time analysis is largely explained by significant declines in five countries: confidence declined in Hungary from 7.2 in 1980 to 4.5 in 1990; South Korea, from 6.9 to 4.7; Finland, from 5.3 to 4.1; Norway, from 5.9 to 5.0; and Spain, from 4.7 to 4.0. Several countries showed a much more modest increase in institutional confidence, although the only one of significance is Iceland, where confidence increased from 4.4 to 4.9 over the period. In most of these countries special factors may help explain the decline. In Hungary, the transition to democracy has caused particular strains, especially in regard to economic conditions. In South Korea, uncertainty over the intentions of the communist North, widespread corruption within the political elite, and martial law between 1979 and 1981, all served to undermine popular confidence in democratic institutions (Rose, Shin, and Munro, Chapter 7 this volume).

The Role of Democratic Experience

Do broad patterns exist in support for democratic institutions across the 24 OECD countries? While it is obviously possible to construct a simple classification based on the level of institutional confidence that exists, this would group countries with high levels of confidence, such as Poland and Turkey, together with countries such as Norway and Iceland, which have clearly different experiences with democratic institutions. A more fruitful approach is to relate the level of confidence in democratic institutions to the length of time that these institutions have remained in existence. Studies of democratic stability suggest that the longer a democracy persists, the greater the probability that it will attract widespread popular support and, as a consequence, be seen as legitimate by its citizens and act more effectively in coping with crises (Powell 1982; Dahl 1989). If we relate the period of time during which each of the 24 countries has been a representative democracy to levels of popular confidence in democratic institutions, four distinct groups of countries can be identified.

The largest group of countries is those in which levels of confidence are high, democracy is well established, and there have been few, if any, threats to the operation of the democratic system. These countries include Britain, the United States, and France, as well as Ireland and the Scandinavian countries. The second largest group is those countries at the other end of the spectrum, where democracy is a more recent innovation and confidence is lower. Some of these countries are the new democracies of Central and Eastern Europe, such as Hungary; countries where democracy emerged in the later post‐war years out of dictatorship, such as Portugal and Spain; and countries where there have been periodic challenges to democracy, such as Mexico and South Korea. (p.195)

                      The Economic Performance of Governments

Fig. 9.2. Institutional confidence and democratic experience, 1990.

Note: A year of continuous democracy is the period that has elapsed since the introduction of universal manhood suffrage, disregarding the periods of wartime occupation by a foreign power.

Source: 1990–1 World Values Survey.

The remaining two groups of countries are comparatively small. Divided countries have well‐established democracies, but their citizens exhibit low popular confidence; the two countries in this category, Belgium and Canada, have both experienced major regional and ethnic challenges to their legitimacy, and this explains the relative lack of popular confidence in their democratic institutions. The remaining three countries, Poland, Turkey, and Germany,7 have high popular confidence in democratic institutions, but are less well‐established democracies. The first two countries have strong religious elements in their societies, which have had periodic political involvement, and this would seem to have bolstered confidence in political institutions. In the case of Germany, where there is modest confidence coupled with a comparatively newly established democracy, national pride has been associated with the institutions of civil society, rather than with political institutions (Conradt 1980).

(p.196) Policy Outputs and Confidence: A Macro Analysis

Studies of the political economy of vote choice suggest that macroeconomic conditions have a major impact on election outcomes, particularly if those conditions are channelled through the mass media so that they influence collective (sociotropic) rather than individual (egocentric) perceptions of economic performance (Kinder and Kiewiet 1981; Lewis‐Beck 1988). However, there is also a demonstrated link between egocentric economic evaluations and emotional reactions towards the government that is viewed as responsible for a performance deficit (Conover, Feldman, and Knight 1986). We would therefore expect that a range of macroeconomic factors might influence confidence in democratic institutions.

Among the macroeconomic factors that may be of importance, the amount of wealth produced by the society could influence institutional confidence, with greater wealth generating reduced confidence support since governments will be unable to satisfy the rising expectations that emerge. Unemployment and inflation have both been identified as factors in weakening support for incumbent governments (Monroe 1984). We might expect, then, that high levels of unemployment and inflation would undermine support for the democratic system as a whole. This link could come about directly, by eroding the spending power and savings of citizens, or indirectly, by presenting a visible measure of economic underperformance which could harm the system's credibility in the eyes of the electorate.8 Unfortunately, the strong relationship between GDP and inflation among our 24 OECD countries precludes us testing the inflation hypothesis, but we can test the impact of unemployment.9

In addition to economic conditions, policy outputs also include social programmes, two of the most prominent of which are health and education. The ability of advanced societies to sustain high levels of health care for their populations is often seen as a primary indicator of social and economic well‐being. Equally, the ability of governments to deliver increasingly well‐educated entrants to the labour force is viewed as crucial to economic growth, as well as being central to more diffuse goals such as the successful operation of the political system via a sophisticated and knowledgeable electorate. We would hypothesize, therefore, that increased life expectancy and higher levels of human capital within a society would result in stronger support for democratic confidence. Finally, the accessibility of mass communications should reduce democratic confidence, because the watchdog role of journalists and predominant news values focus on the shortcomings not successes of public policy.10

(p.197) These hypotheses are tested in Table 9.2 which predicts confidence in democratic institutions from a range of social and economic variables across the 24 democracies. In addition, the model controls for the period of time that the country has been a democracy. The results suggest that macroeconomic conditions have a major impact, mainly via GDP (which is measured as a ratio).11 Higher levels of GDP result in lower levels of confidence, net of other things. This tends to support the hypothesis that more affluent countries may have populations which have higher expectations of their democratic institutions than countries that have a lower GDP. Increased levels of unemployment undermine popular confidence, although not significantly.

Table 9.2. Policy Outputs, Mass Media Consumption, and Institutional Confidence

Mean

b

Beta

(t value)

Health:

Life expectancy (yrs)

74.6

−.01

−.06

−0.27, p=.79

Human capital:

Schooling (yrs)

9.4

.18

.67

2.24, p=.04

Tertiary graduates (%)

13.6

.02

.21

0.81, p=.43

Economic conditions:

Adjusted GDP (ratio)

100.0

−.17

−.95

−4.90, p=.00

Unemployment (%)

7.9

−.01

−.06

−0.40, p=.70

Mass media:

Daily newspapers (copies per 100 people)

28.5

−.01

−.22

−1.15, p=.27

Televisions (per 100 people)

42.0

−.01

−.18

−.73, p=.48

Democratic experience:

Continuous democracy (yrs)

63.8

.01

.17

0.64, p=.53

Constant

20.9

Adj. R‐squared

.60

(N)

(24)

Note: OLS regression results shows partial and standardized regression coefficient predicting support for democratic institutions among 24 OECD countries. The dependent variable is scored from 0–10 and was extracted from the 1990–1 World Values Survey.

Among the social policy outputs, life expectancy has no significant effect on confidence. The increasing availability of human capital in society does increase confidence, but only through primary and secondary schooling, not via tertiary education. This supports other findings which show that democratic values and support are enhanced more by increased civic education during the teenage years, and not by university study: while tertiary education does produce a more politically sophisticated and aware electorate, it has few consequences with regard to support for the system as a whole (Neumann 1986). As we hypothesized, popular access to mass communications reduces democratic confidence, more so via daily newspapers than television, although the former just fails to reach statistical significance.

(p.198) We can conclude, then, that there is a major role for macroeconomic conditions in shaping confidence in democratic institutions, and a lesser role for social conditions. But these economic conditions influence institutional confidence indirectly, through the aggregate wealth that the society accumulates, rather than more directly, through the effects of unemployment. This endorses the finding of Clarke, Dutt, and Kornberg (1993) who concluded that while unemployment was a significant predictor of satisfaction with democracy in Western Europe, its net impact was comparatively modest. The results are, of course, based on a strictly limited sample of 24 countries; moreover, while the model controls for the period of time that the country has been democratic, it does not take into account the wide variety of other factors which may influence democratic confidence. For that reason, the next section conducts a parallel individual level analysis, using the 1990–1 World Values Survey.

Policy Outputs and Confidence: A Micro Analysis

The previous section showed the importance of macroeconomic conditions in shaping confidence in democratic institutions. In this section, the analyses move to the individual level in order to examine the micro social and economic factors that may prove to be important in this equation. In particular, we need to know the relative importance of objective economic factors, such as the material wealth that the person has, measured against subjective economic factors, such as their own feelings of economic satisfaction. The relative effects of these economic factors must also be estimated once other things have been taken into account, notably the human capital that they possess and their country of residence.

It is obviously impractical to analyse each country separately—and, in any event, we are more interested in identifying any underlying commonalities that may exist between the countries rather than country‐specific effects. Two separate analyses are conducted, one for the group of ten countries that were earlier categorized as established, and one for the group of nine countries that were categorized as unstable.12 The independent variables form four groups. First, there are three variables reflecting economic position and satisfaction, the major focus of the analysis. These are composed of three measures of objective economic position—whether the respondent is a labour‐force participant or unemployed, and total household income before tax—and one measure of economic satisfaction.13 Popular expectations of government are (p.199)

Table 9.3. Micro Social and Economic Factors and Institutional Confidence

Established democracies

Newer democracies

b

Beta

b

Beta

Economic position:

Labour‐force participant

−.36

−.08

−.31

−.06

Household income

−.08ns

−.02ns

−.02ns

−.01ns

Unemployed

−.08ns

−.01ns

.26ns

.02ns

Economic satisfaction

.08

.10

.12

.13

Economic collectivist

.14

.08

.10

.05

Health:

Self‐assessed health

.04ns

.02ns

.04ns

.02ns

Human capital:

Schooling

.02ns

.02ns

−.07

−.09

University educated

.12ns

.02ns

−.20ns

−.03ns

Country (France/Italy):

Norway/Portugal

.49

.07

.52

.05

Ireland/Japan

.75

.04

1.1

.18

Netherlands/Spain

.08ns

.01ns

.67

.10

Britain/Mexico

.36

.07

.23ns

.03ns

Sweden/Finland

.26

.03

.91

.09

United States/Hungary

.60

.10

1.4

.18

Denmark/Czechoslovakia

.15ns

.01ns

1.1

.18

Austria

.11ns

.02ns

Constant

3.5

2.9

Adj R‐squared

.04

.05

(N)

(9,343)

(8,202)

(ns) =not statistically significant at p〈.01, two‐tailed.

Note: OLS regression results shown partial and standardized regression coefficients predicting confidence in democratic institutions, which is scored from a low of 0 to a high of 10. See text for details of variables and scoring.

Source: 1990–1 World Values Survey.

based on a question about whether the respondent preferred an individualist or a collectivist approach to economic management.14

The remaining three groups of independent variables in Table 9.3 reflect the influence of social factors. Health is measured by self‐assessed personal health, a highly subjective indicator but nevertheless some measure of the state of the respondent's health. As in the aggregate analysis, human capital is measured by years of primary and secondary schooling, and by whether or not the person possessed tertiary education. Finally, the countries are measured by dummy variables, with France (in the case of the established countries) and Italy (in the case of the newer countries) forming the reference category.

(p.200) By any standards, the roots of democratic confidence in the social structure of the two groups of OECD countries are weak. Table 9.3 shows that for both groups, the models predict only 4 and 5 per cent of the total variance, respectively, even when inter‐country variances are taken into account. Disregarding for the moment these inter‐country differences, the strongest and most consistent predictor of confidence is economic satisfaction, with greater satisfaction producing stronger support for democratic institutions. However, it is notable that economic satisfaction has a greater impact on confidence in the newer democracies (with a partial coefficient of .12) than it does in the established ones (coefficient .08). In contrast to subjective feelings concerning economic conditions, objective economic position has relatively little effect on confidence. In both countries those outside the labour force are likely to exhibit higher levels of confidence compared to those who are in the labour force, while the total household income has no significant effect for either group. In line with studies of economic voting, the political consequences of egocentric economic evaluations are weaker than their sociotropic counterparts. However, since the World Values Survey did not include any economic evaluations other than the economic satisfaction question used here, it proved impossible to examine this further.

Among the other factors included in the models, support for economic collectivism has a consistent influence in predicting confidence across the two groups of countries, providing some support for the government expectations hypothesis. Health is unimportant, as is human capital, with the partial expectation of schooling, among the newer democracies. Finally, there are important variations between the countries in each of the two models, but also in the patterns between the two models themselves. Judged against the excluded country (France for the first group, Italy for the second), there are significant variations among the established countries. However, these never exceed three‐quarters of a point on the 0–10 scale used to measure democratic confidence. By contrast, the inter‐country variations in the second group are nearly twice as large: for example, the mean level of democratic confidence in Hungary is 1.4 points higher than Italy, even after all of the variables in the model are taken into account.15

Although the models in Table 9.3 have comparatively weak predictive power, as has been shown in other analyses (Listhaug and Wiberg 1995), the results do permit us to make some broad conclusions about the importance of economic position and expectations in shaping democratic institutional confidence in the established and newer groups of democracies. First, subjective evaluations of the person's economic position are consistently important, more so in the newer democracies than in the established ones. This undoubtedly reflects the economic instability that has occurred in the newer countries, and their comparative inexperience with democracy. Second, there is little to (p.201) suggest that the person's economic position in the society—judged by wealth or employment status—has any impact. This is particularly notable in the case of unemployment, which, like inflation, directly threatens a person's social and economic well‐being. Whatever its potential to cause personal economic disruption, it has little impact on institutional confidence, as other studies have shown (Clarke, Nutt, and Kornberg 1993).

Discussion and Conclusions

Popular confidence in democratic institutions is at the heart of representative government. Widespread confidence reduces the potential for radical change to the system, but it also encourages a constructive desire for social reform. Perhaps most importantly of all, it builds a reservoir of support which can be expended during periods of crisis when democratic institutions are placed under stress. To date, most studies have examined the origins of democratic confidence in the political culture of the country concerned, or in the norms and values of mass opinion, slowly accumulated over several generations. By contrast, this chapter has examined the political economy of popular confidence in democratic institutions, by focusing on the macro and micro social and economic conditions of the country.

The results show that institutional confidence is strongly related to the period of time that democratic institutions have been in existence. This supports the argument that confidence is formed cumulatively within the mass electorate; but it also supports the argument, often found in democratic theory, that democratic confidence is predicated on the frequency of free, competitive, national elections. In terms of the political economy theories, the macro analysis showed that national GDP was negatively related to institutional confidence, providing support for the economic expectation hypothesis, which predicted that the electorates of the established democracies had higher expectations of their systems than those in new democracies. The survey analysis found a modest yet consistent effect for economic satisfaction, but little or none for objective economic position.

In general, then, the conclusion is that economic conditions are more important than social ones, but together their impact is negligible when compared to such factors as political culture or historical circumstance. This confirms the conclusion of Clarke, Nutt, and Kornberg (1993: 1015; see also Kornberg and Clarke 1992) that ‘the political economy of public orientations toward polity and society in contemporary Western democracies is real but limited’. It also supports the argument that deep‐rooted cultural values shape popular views of political and social arrangements although, as others have pointed out (see, for example, Miller and Borrelli 1991; Miller and Listhaug 1990), there are dangers in using country variables as proxies for broadly‐based cultural values. (p.202)

                      The Economic Performance of Governments

Fig. 9.3. Economic satisfaction, political trust, and institutional confidence, Czechoslovakia and the United States.

Notes: Figures are partial regression coefficients and (in parentheses) standardized regression coefficients. The models also control for (but do not show) the full range of independent variables in Table 9.3.

Source: 1990–1 World Values Survey.

Why do policy outputs in general, and economic performance and expectations in particular, play such a minor role in shaping confidence in democratic institutions, particularly in view of the important impact that they enjoy in shaping electoral outcomes? Although there is a modest role for economic satisfaction in predicting confidence, the cultural values of the country redirect feelings about personal and collective economic conditions towards the incumbent government, rather than towards the system as a whole. This is, in effect, the reservoir of support that the established democracies can rely on to provide them with a buffer between the political system and citizen demands during times of economic stress. Such a reservoir of support should be greater within the established democracies than in the newer ones, since they have had more time and opportunity to accumulate popular goodwill. More specifically, we would expect the link between economic beliefs and democratic confidence to be weaker in the established democracies than in the newer ones, and for the impact to be mediated via beliefs about the government.

Although it is impossible to test this hypothesis across the range of OECD countries examined here, the World Values Survey did include a question about trust in government which was asked in just five countries; of those five, we can compare the impact of economic satisfaction and political trust on democratic confidence in two: the United States, and one of the new democracies, Czechoslovakia (Figure 9.3). The results show that the direct effect of economic satisfaction in shaping confidence is significantly higher in Czechoslovakia than it is in the United States, as we would expect. In the (p.203) absence of a fully developed civil society and without a reservoir of goodwill towards democratic institutions, citizens are more likely to blame economic difficulties on the institutions of the state than they are in an established democracy such as the United States. It is also notable than in Czechoslovakia there is a much stronger link between trust in the government and democratic confidence than in the United States.

The political economy of confidence of democratic institutions is therefore strictly limited: political circumstances are much more important. This conclusion underlies the gradual transformation that has taken place in the established democracies, where the frequency of national elections has slowly generated a reservoir of popular support for democratic institutions, with citizens drawing a clear distinction between the institutions of the state on the one hand, and the party and leaders elected to conduct public policy on the other. In parallel with this accumulation of democratic skill, political oppositions gradually come to be viewed as legitimate, and operate within clearly limited parameters (Weil 1989). The challenge in the twenty‐first century will be to see whether the political economy of confidence in democratic institutions has the same limited role in the newer democracies of Central and Eastern Europe as it has in the established ones.

Notes:

(1) The 1990–1 World Values Survey was co‐ordinated by Ronald Inglehart and the data supplied by the ESRC Archive at the University of Essex. The survey is used in preference to other surveys (such as the International Social Survey Program) because of its wide coverage of the OECD countries.

(2) In both cases, however, there are significant cross‐national variations in how electorates weigh these various factors in making their vote decision (Lewis‐Beck 1988).

(3) All of these items were included in the 1990–1 World Values Survey. An additional item was included which asked about trust in the political system, but this item was included in only a small proportion of the countries; this is used in the final section of the paper.

(4) Listhaug and Wiberg (1995: 306) show that across most Western European countries there is a clear distinction in how citizens view representative state institutions (parliament, bureaucracy) and how they view coercive state institutions (police, armed forces).

(5) In total the survey included 43 separately identified countries, although East and West Germany were sampled separately, as was Moscow and Russia as a whole, and Northern Ireland and mainland Britain.

(6) Fuchs, Guidorossi, and Svensson (1995: 351) show that eight countries have stable levels of confidence (Denmark, Norway, Germany, France, Britain, Ireland, Spain, and Belgium), while five have increasing confidence (Luxemburg, Portugal, Netherlands, Northern Ireland, and Italy) and one declining confidence (Greece).

(7) Because the 1990–1 survey was conducted just after reunification, ‘Germany’ relates to West Germany, rather than the reunified nation.

(8) The other major indicator of economic performance which could influence democratic confidence is unemployment; this is examined in the next section using individual‐level survey data.

(9) The correlation between inflation over the previous decade and GDP across the 24 countries is r = .84.

(10) The data come from the United Nations World Development Program, Human Development Report (New York: Oxford University Press, various years), and the World Development Report (New York: Oxford University Press, various years).

(11) The ratio sets the mean GDP across the 24 nations in $US to 100, and calculates the GDP of each country accordingly.

(12) For reasons of space, analyses of the two countries in the hybrid category and the three in the divided category are not shown here.

(13) Labour‐force participant and unemployed are both dummy variables. Household income was measured on a 1–10 scale, and has been standardized here to a ratio, where the mean household income in the country in question is taken as 1 and individual incomes are scored accordingly. Economic satisfaction was based on the question: ‘How satisfied are you with the financial situation of your household?’ and scored from a low of 1 to a high of 10.

(14) The question was: ‘We are more likely to have a healthy economy if the government allows more freedom for individuals to do as they wish’, where the respondents were asked whether they agreed completely, agreed somewhat, neither agreed nor disagreed, disagreed somewhat or disagreed completely with the statement.

(15) This is emphasized by the variances explained by the models with and without the country dummies. In the first model, the addition of the country dummies increases the variance explained from 2.7 to 3.6 per cent; in the second model, from 1.4 to 4.6 per cent.