37 minute read

Table 3 Breaking informal standards in Western Europe

quality of social services concerns many families. Immigration and threats to national identity can be important too. When asked to name the three most important problems facing the country, only 16% of Britons mention corruption and 23% of the french. Spain was exceptional because protest parties have made fghting corruption a major theme: 66% said it was a major problem there. When people cast their vote, they make a holistic judgment about the alternatives on the ballot before marking their preference for a party or candidate that they view positively or that they view as the lesser evil.

references

Advertisement

Becker, Gary. 1997. The Economics of Life. New york: McGraw-Hill. della Porta, D., and y. Meny (eds.). 1997. Democracy and Corruption in Europe.

London: Pinter. Eurobarometer, 2014. Corruption: Special Eurobarometer 397. Brussels:

European Commission. fisher, J., E. fieldhouse, M. franklin, R. Gibson, M. Cantijoch, and C. Wlezien (eds.). 2018. The Routledge Handbook of Elections, Voting Behavior and Public

Opinion. London: Routledge. Hine, D., and G. Peele. 2016. The Regulation of Standards in British Public Life:

Doing the Right Thing? Manchester: Manchester University Press. Ho, L.K. 2011. Public Policy and the Public Interest. Abingdon: Routledge. Kavanagh, D., and P. Cowley. 2010. The British General Election of 2010.

Basingstoke: Palgrave Macmillan. Rose, Richard. 1992. East Europe’s Need for a Civil Economy. In Finance and the International Economy 6, ed. Richard O’Brien, 5–16. Oxford: Oxford

University Press. Rose, Richard, and Bernhard Wessels. 2018. Money, Sex and Broken Promises:

Politicians’ Bad Behaviour Encourages Distrust. Parliamentary Affairs, in press. Schumpeter, J. 1952. Capitalism, Socialism and Democracy, 4th ed. London:

Allen & Unwin. Tänzler, K., K. Mara, and A. Giannakopoulos (eds.). 2012. The Social

Construction of Corruption in Europe. Aldershot: Ashgate. Ungoed-Thomas, J. 2015. Tony’s Here to Help—At a Price. Sunday Times,

March 8. vanHeerde-Hudson, J. (ed.). 2014. The Political Costs of the 2009 British MPs’

Expenses Scandal. London: Palgrave.

CHAPTER 7

The Impact of Corruption on Citizens

Although corruption is a problem of governance, development economists have been the leaders in research on the impact of corruption, because it is seen as having a substantial negative effect on economic growth. Corrupt politicians who control national resources and foreign aid may abuse their political power to make sure that they take what they want for themselves. At worst, corrupt rulers have been ‘stationary bandits’, enforcing an undemocratic repressive order to maintain a fow of personal riches from the exploitation of their country’s people and resources (cf. Olson 2000). Macroeconomists have used the Corruption Perceptions Index to produce quantitative estimates of the loss of the full beneft of development expenditure due to the ineffciencies and ineffectiveness that national corruption may have on economic growth. Whatever methods are used, the cost of corruption is estimated to be billions of dollars annually and cumulatively tens of billions (Rose-Ackerman and Soreide 2011).

By contrast with what has preceded, this chapter is about the impact of corrupt governance on the political behaviour and attitudes of citizens. When corruption is seen as signifcant, this is likely to discourage the political engagement of individuals (cf. Easton 1965; Dahl 1989). Instead of participating in elections and trusting political institutions, individuals can show incivisme, a french term for individual attitudes and activities that reject the idea that people ought to engage in politics.

© The Author(s) 2019 R. Rose and C. Peiffer, Bad Governance and Corruption, Political Corruption and Governance, https://doi.org/10.1007/978-3-319-92846-3_7 127

Voting is the one form of political activity in which a majority of citizens are engaged (Blais 2010). Non-voting can be caused by people being indifferent to what distant governors do or because they are alienated by politicians abandoning their election pledges once in offce. Whereas voting against the governing party or for a protest party registers dissatisfaction with representatives, not voting registers disengagement from politics or even rejection of the political system as a whole. Because the names and appeals of parties differ with national contexts, non-voting is more suited to comparative analysis than party identifcation. Elections also differ cross-nationally: Some are free and some are not. In the latter context, not voting can be a means of protesting against undemocratic regimes.

Political trust is integral to theories of good governance, because it refects confdence that public offcials will perform their duties by the book rather than by hook or crook. However, empirical studies often fnd that distrust is widespread. When offcials break formal and informal standards of good governance, it is rational for individuals to distrust their political institutions (Rose and Mishler 2011). Given that political trust is variable, a variety of reasons may explain why some people trust their political institutions while others do not (Zmerli and van der Meer 2017; Uslaner 2018: Part VIII).

Willingness to vote and readiness to trust governors refect two complementary forms of political engagement. Voting is an input that citizens can make to give direction to governors, while trust is about how people react to the outputs of government (Rose 1989; Kumlin 2004). If these outputs are delivered free of corruption, this should encourage trust in political institutions, but if popular standards of good governance are not met, this should stimulate distrust.

To test the impact of corruption on political engagement requires survey data that not only records the experience of individuals with corruption but also their political attitudes and behaviour. As an anti-corruption agency, Transparency International’s Global Corruption Barometer questionnaire concentrates intensively on different forms of corruption and excludes other questions about political behaviour. The 2013 Eurobarometer survey that asked many questions about corruption similarly did not include questions about voting and political trust. Democracy surveys that collect a great deal of data about political attitudes tend to give little or no attention to political corruption.

To provide data for this chapter, we analyse three continental surveys that have measures of voting turnout and of political trust as well as corruption measures that are part of the Global Corruption Barometer. A total of 77 countries meet both requirements in the sub-Saharan Afrobarometer survey, the Latin American Public Opinion Project (LAPOP) and countries with a Communist legacy included in the European Bank for Reconstruction and Development (EBRD) survey. Since breaking informal standards rather than bribery is the chief form of corruption in democratic countries of Western Europe, we test the impact of breaking informal standards on political trust with data from the Politicians Behaving Badly survey analysed in Chapter 6.

1 Theories of The imPacT of corruPTion

Theories of the impact of corruption on individuals assume that its effects are negative for all forms of political activities. Hence, our generic model postulates that corruption will have the same effect on two different forms of political engagement, voting and trust in political institutions. To avoid the reductionism of simple correlations between a single measure of corruption, it employs multiple measures to test which particular aspects of corruption may be signifcant. To avoid spurious associations, the model introduces controls for the established effect of individual socio-economic resources on voting, and for national context, which consistently has an impact on corruption.

Political Engagement (f) Corruption, Resources, National Context.

The continental Barometer surveys provide four indicators suitable for testing the infuence of corruption on voting and trust in government. The payment of a bribe directly exposes an individual to bad governance; insofar as corruption is important, it should have a signifcant major effect on the two dependent variables. Among sub-Saharan Afrobarometer respondents, 17% reported that they had paid at least one bribe in the past year, and the fgure was 14% in Latin America and in formerly Communist countries. The experience of bribery increases by more than one-quarter if one controls for the minority of respondents who have had no contact with any public service in the past year. Vulnerability to bribery increases when people have contact with more than one service during the year. Among Afrobarometer respondents, 54% have contact with two or more services; 39% do so in Latin America; and 39% have multiple contacts in EBRD countries.

In the course of a few years, there is a cumulative increase in the proportion of people who have had to pay a bribe at least once. This can create a long-term effect that is refected in their perception of corruption. Among Afrobarometer respondents, 26% perceived most or all public offcials as corrupt; in formerly Communist countries 31% did so; and in Latin America the percentage perceiving most or all offcials as corrupt was high, 81%. Even if an individual is not caught up in bribery every year, a long-term perception of corruption can reduce the likelihood of voting and increase distrust of government.

By defnition, corruption is a problem in badly governed countries, but such countries have many problems competing for their attention, such as dealing with an economy that leaves many people poor, and domestic and cross-border challenges to national security that leave many citizens feeling insecure. If turnout at elections is low and distrust in political institutions high, the cause may not be corruption but the government’s ineffectiveness in dealing with other problems to which citizens give a high priority. Given competition between multiple problems, corruption is rated as one of the top problems of their country by 36% in formerly Communist countries, 12% of Africans and 8% of Latin Americans.

Another way in which perceptions may affect political engagement negatively is if people see corruption as increasing. When corruption gets worse, politicians appear unable to maintain corruption at even a steady level. The resulting uncertainty can discourage people from voting and increase distrust. At the time of the continental surveys, many African respondents and people in formerly Communist countries thought that corruption was increasing and very few thought it was decreasing.

Controlling for an individual’s resources is necessary, since theories of voting behaviour make these resources ‘the essential core’ of explanations of whether people decide to vote (Plutzer 2018: 690). Education should make voting more likely since it gives individuals a better understanding of electoral procedures and issues and helps identify the party that comes closest to their own views. However, more educated people are more likely to follow political news and thus be aware of media accounts of politicians engaging in bad behaviour, and therefore less likely to trust their political institutions. The effect of income on the likelihood of voting is problematic. The cost of taking the time to vote will be higher in absolute terms for a person with higher earnings, but for the working poor taking time off work may have a higher relative cost. The infuence of income on political trust is ambiguous. Insofar as being

better off confers a status on individuals that secure better treatment by public offcials, its effects on political engagement should be positive. However, better-off people may be exploited for bribes because they can afford to pay them, thereby reducing their trust in political institutions.

Age is a resource that consistently infuences voting in Western nations because it increases the experience that people have of governance. Older people are more likely to vote because they tend to have more time to do so, they have learned about elections from experience, and repeated exposure is likely to make voting a habit. Insofar as contemporary political systems in Africa, Latin America and ex-Communist countries are viewed as improvements on past regimes, older people ought to be more trusting than young people, who may take an improved governance environment for granted. However, insofar as youthful socialization into an undemocratic regime creates a long-term predisposition to distrust governors, then older people will be less trustful of their contemporary institutions.

Whereas citizens have the chance to vote only once in every few years, people have continuing opportunities to engage with the outputs of political institutions. Independent of age and education, frequency of contact with the state builds up experience of how public services are delivered. Most people contact more than one public service during a year and, however many services are contacted, a majority do not pay any bribe (fig. 4 in Chapter 4). As long as service outputs are delivered by the book, people should be readier to engage with the input side of governance, voting. However, contact with a greater number of services also increases the risk of having to pay a bribe, particularly in countries where corruption is widespread (see Table 1 in Chapter 5). In such circumstances, more frequent contacts could increase distrust in political institutions.

Involvement in networks of informal face-to-face relationships gives people social capital, that is, a resource that people can use when they want to get things done. Most people with social capital are in networks that are not directly engaged in politics and Putnam (2000) has theorized that being socially engaged ought to spill over into political engagement. When an election is held, people with social capital should be more likely to go along to vote as others in their network do. People with social capital should also be more likely to have in their network at least one person who can help in dealing with public services to get what they want and therefore be more inclined to trust political institutions.

Among many features of national context, the most important for voting is whether an election is free and fair or rigged to make the outcome a foregone conclusion. Individuals should be more likely to vote in competitive elections where their collective choices could make a difference. freedom House evaluates how free and fair elections are on a scale ranging from 0 to 12. Two-thirds of Latin American countries are at the top end of the scale with a rating of 10 or higher. fewer than half of formerly Communist countries and one-third of sub-Saharan African countries are rated as holding free elections, and the median country in these continents holds partly free elections. If an election is not free and the outcome appears predetermined, it is rational for an individual not to bother casting a vote that has no value. Abstention from voting can be a silent but visible way of registering distrust in governors.

Given differences between voting, a form of political behaviour, and trust, a subjective attitude, corruption may affect citizens differently. for example, the perception of government as corrupt may affect trust but not voting, while the experience of seeing corruption as increasing may infuence whether an individual votes. Therefore, we test the effects of corruption separately for voting and for trust.

2 corruPTion has LiTTLe imPacT on voTing

Theories that are based on the assumption that corruption has an important effect on political behaviour imply:

H1. The more that individuals are affected by corruption, the less likely they are to vote.

Whether democratic or not, all the countries analysed here hold elections that give their citizens the opportunity to engage with their political system. At the national election preceding the surveys, offcially reported turnout was 62% in Afrobarometer countries, 67% in Latin American countries and 63% in formerly Communist EBRD countries. These fgures are as high as turnout in British elections and signifcantly higher than in American elections. In Western democracies there is a tendency for some non-voters falsely to report that they voted, making survey results for turnout 10–20 percentage points higher than the offcial record. The three Barometer surveys showed less tendency

to exaggerate turnout. The self-reported turnout of Afrobarometer respondents was eight percentage points above offcial fgures; in LAPOP countries, six points, and in the EBRD countries surveyed, self-reported turnout was 3% higher than in the offcial record.

While all three continental surveys collect comparable indicators, the classifcation of replies is not identical. This makes it unsuitable to pool the surveys in a single fle for analysis. Instead, we apply the same statistical techniques in three parallel multilevel logit regressions testing the effect of hypothesized infuences on voting. When a given variable is signifcant in two or three independently conducted surveys on different continents, this provides robust support for conclusions. The strongest form of statistical signifcance is a p-value lower than 0.001, and a p-value of 0.01 or lower is also strong. Since signifcant infuences vary greatly in their impact, Table 1 reports shifts in predicted probabilities too. They estimate the size of the impact of standardized independent variables on the likelihood on an individual voting.

Table 1 Corruption has little impact on going to vote

Afro LAPOP EBRD

(Maximum shift probability of voting)

Corruption Paid bribe Not sig Not sig Not sig Perception corruption −0.09** Not sig Not sig Important problem Not sig Not sig 0.05*** Increasing or decreasing Not sig Not avail Not sig Resources Age 0.56*** 0.46*** 0.32*** Education Not sig 0.15*** 0.17*** Income Not sig Not sig 0.12** Social capital 0.07*** 0.04** 0.08** Number services contacted 0.09** 0.18*** 0.23** Context free, fair elections Not sig Not sig 0.28** Pseudo R squared 7.3% 10.5% 5.6%

***Signifcant p-value < 0.001. **p-value < 0.01. Source Multi-level logit regression. Shift: Change in the predicted probability of voting if an independent variable moves from its minimum to maximum value. F test for each model: p-value 0.000.The EBRD regression included a variable for whether a country was a Soviet successor state; it was not signifcant

Corruption consistently has little impact on an individual casting a vote. Even though the payment of a bribe may be expected to make people want to disengage from government, in all three surveys it is not signifcantly associated with whether a person votes. Similarly, the perception of political institutions as corrupt has no signifcant effect for whether people vote, except in Africa, where it is associated with a nine percentage point lower likelihood of voting.

An alternative interpretation, favoured by campaigning organizations such as Transparency International, is that people who think corruption is an important problem should become politically engaged in order to rid their government of corruption. However, in Africa and Latin America the small minority who see corruption as a big problem are not signifcantly more likely to vote than those who see other problems as most important. Only in EBRD countries does regarding corruption as a big problem give a signifcant albeit limited boost of fve percentage points to the likelihood of voting. The slight effect of corruption on the readiness of people to vote is confrmed by other studies. A host of competing infuences, such as party loyalty, ideology and party system characteristics, restrict the effect of corruption on whether citizens vote (Charron and Bagenholm 2016; Dahlberg and Solevid 2016).

An individual’s resources have by far the biggest impact on voting (Table 1). Age is the most important resource. Older Africans are 56% more likely to vote than young people; the impact is almost as great in Latin America, 46%. Older citizens who were socialized under Communism are 32% more likely to vote than young EBRD respondents who did not have this experience. The importance of age refects learning during the life cycle. While young people have little knowledge or experience of voting, by middle age people have been exposed to up to half a dozen elections, and casting a ballot has become habitual. Even though men are much more likely than women to be successful candidates, since statistical analysis found no signifcant substantial gender effect on voting or trust, this variable is not included in the tables in this chapter.

Education gives stronger encouragement to voting than does income. It boosts the likelihood of voting by 17% in EBRD countries, fve percentage points more than the effect of income. In Latin America education increases the likelihood of voting by 15%, while income has no statistical signifcance. Africa appears to be an outlier: Voting is not signifcantly increased if people have more education or a higher income. This may be due to African politicians making more effort to mobilize

the poor, uneducated voters, who are numerous enough to decide their political fate. The mobilization of the uneducated is supported by the infuence of age, since older Africans are the most likely to have had little or no basic education.

Notwithstanding the extent to which bribery is present in these continents, the more frequent use of public services gives a signifcant boost to turnout. People with the most contacts with public services rather than no contact are 23 percentage points more likely to vote in EBRD countries, 18% more likely in Latin America and nine points more likely to vote in Africa. This association suggests that going to vote and making use of public services may both be part of a syndrome of activities involving engagement in society. This interpretation is consistent with participation in a social capital network also having a signifcant effect on voting. Having social capital gives a boost of eight percentage points to voting in EBRD countries, 7% in Africa and four points in Latin America. Unlike most corruption indicators, the effect of social capital is consistently signifcant, but the impact is much less than age, which relies on individual experience.

When elections are not free and fair, this should discourage people from casting a ballot because their vote can have no effect on the outcome and may be fraudulently counted. Moreover, a low level of participation challenges the claim of an undemocratic government to represent all the people. Notwithstanding this theoretical logic, the quality of an election has no signifcant effect on whether Africans or Latin Americans vote, perhaps because past traditions have made people accept their national practices as normal. By contrast, respondents in Central and East European countries can compare free elections since the fall of the Berlin Wall with Communist-held elections that decided nothing. Thus, in EBRD countries the likelihood of people voting is 28 percentage points higher where free elections are held than where they are not.

Paradoxically, including multiple measures of corruption in the cross-continental analysis confrms the major importance of individual resources as the chief determinant of why people vote in developing countries as well as in established Western democracies (Table 1). All fve indicators of resources are signifcant in formerly Communist countries, four are signifcant in Latin America and three in Africa. By contrast, neither paying a bribe nor viewing corruption as increasing is signifcant in any of the three surveys. In two, seeing most politicians as corrupt and viewing corruption as important likewise fails to achieve signifcance.

3 corruPTion gives a Big BoosT To DisTrusT

Although corruption is not present in the majority of contacts that individuals have with public services, when the proportion of people paying a bribe rises above the trivial level and the proportion seeing most offcials corrupt is higher, this creates a ripple effect as more and more people realize that they are vulnerable to being asked to pay a bribe to get a service to which they are entitled. When vulnerable citizens are not confdent that public offcials will perform their duties as they ought, the resulting uncertainty should encourage distrust.

H2. The more that individuals are affected by corruption, the less likely they are to trust political institutions

Because trust is a generic social science concept, its radius depends on the objects of trust (Zmerli and van der Meer 2017; Uslaner 2018). Psychologists and social psychologists focus on interpersonal trust involving face-to-face relations with family, friends and work mates. Political psychologists go further; they postulate that if individuals have positive trust in face-to-face relations, this trust will be projected to people they have never met. National policymakers are remote from the frst-hand experience of individuals. Exposure to politicians on television is a oneway process; it may infuence perceptions of politicians but not engage with the masses of citizens personally.

Corruption and trust in developing countries. Governance is about how institutions relate to citizens; in developing countries the extent to which institutions are fully bureaucratized is problematic. Barometer surveys normally ask about trust in institutions or about categories of people in charge of institutions, such as politicians or religious leaders. The Afrobarometer is distinctive; in keeping with late bureaucratization in the continent, it asks about people such as Members of Parliament rather than about Parliament. Since trust can vary between institutions, the three surveys ask for evaluations of a variety of institutions on a scale ranging from the most to the least trusted. Political institutions tend to be among the least trusted. factor analysis was used to identify four institutions that refect a common attitude towards political institutions: trust in parliament, the government executive, local government and political parties. We have combined evaluations of these institutions in a standardized scale in which 1.00 stands for maximum trust and 0.00 for maximum distrust.

Because theories of political trust are developed in countries where good governance is the norm, it is conventional to give a positive label to a scale. However, survey results show that it could just as accurately be described as distrust. The mean is near the scale’s mid-point. In the Afrobarometer, the mean is 0.53; among EBRD respondents, 0.54; and in Latin American countries it is 0.44. The mean position is thus almost equally far from complete trust or distrust; it shows scepticism about whether political institutions are or are not to be trusted (cf. Mishler and Rose 1997).

The second hypothesis is confrmed: Corruption in all its forms signifcantly reduces the trust of citizens in their political institutions. Moreover, corruption indicators are much more likely to be signifcant than indicators of resources. They thus contribute most to the high level of variance in individual attitudes accounted for in the regression analysis: almost 27% in the Afrobarometer and 18% in the EBRD study and 17% in Latin America (Table 2).

Table 2 Corruption increases distrust in government

Afro LAPOP EBRD

Round 6 2014 2016

(Maximum shift probability of trust)

Corruption Perception corruption Increasing or decreasing Paid bribe Important problem Resources Age Social capital Education Income Number services contacted Context free, fair elections OLS adjusted R2 (%) −0.43*** −0.11*** −0.29*** −0.16*** Not avail −0.23*** −0.05*** −0.04** Not sig Not sig Not sig −0.02***

0.05*** Not sig Not sig 0.02** 0.02* Not sig −0.11*** Not sig Not sig −0.10*** Not sig 0.08** 0.06** Not sig Not sig

Not sig Not sig −0.16** 26.7 17.1 17.6

*** Signifcant p-value < 0.001. ** p-value < 0.01. *p-value < 0.05 Source See Table 1. Estimated change in trust (0–1 scale) associated with a minimum to maximum shift in the independent variable. The LAPOP regression included an additional variable: perceptions of the courts’ fairness; it was signifcant, (Trust shift: 0.29, p-value: 0.000). F tests for all models: p-value<0.000. The EBRD regression included a variable for whether a country was a Soviet successor state; it was not signifcant

The perception of corruption in political institutions has the strongest and most consistent effect. In Africa, people who see institutions as very corrupt are less trusting than people who see institutions as having integrity; the effect is large, lowering trust by 43 percentage points. The perception of corruption lowers trust by 29 percentage points in formerly Communist countries and by 11% in Latin America. The effect of perceptions is underscored by the fact that the actual experience of paying a bribe has no signifcant effect on trust in formerly Communist countries and comparatively limited effect among Afrobarometer and Latin American respondents.

The dynamics of corruption also has a substantial impact. If people see corruption increasing, this boosts distrust. Among Africans, those who see corruption as increasing a lot are 16% less likely to trust their political institutions than those who see corruption as decreasing a lot. In EBRD countries, those who see a big increase in corruption are 23 percentage points less likely to trust government. By contrast, regarding corruption as an important problem had no signifcant effect on trust among Afrobarometer and LAPOP respondents. In formerly Communist countries, where 36% see corruption as one of the biggest problems, it lowers trust by only two percentage points. This suggests that where corruption is relatively widespread people no longer see it as a problem that must be solved but accept it as a chronic condition that they must put up with. However, if a regime is not even able to prevent corruption growing, a loss of trust will result.

Whereas resources have the most infuence on whether people vote, they have virtually no impact on trust in government in Latin America and in formerly Communist countries. In both continents, neither age nor education, the two major infuences on voting, has a statistically signifcant infuence on trust. The lack of any effect of education suggests that, in countries where corruption is high, efforts to inculcate a positive view of government in school has less effect than what people learn in the streets. Having contact with more public services likewise has no signifcant effect on trust. Governors may hope to gain electoral advantage by providing more people social benefts, but they do not gain more trust if their institutions are perceived as corrupt. The one resource signifcant in post-Communist countries, having a higher income, boosts trust in political institutions by eight percentage points. In Latin America having social capital has a signifcant association with trust, but the size of the boost is slight, two percentage points.

In Africa, by contrast, where differences in resources are particularly great, all fve indicators have a signifcant effect on trust. Education reduces trust in political institutions by 11 percentage points. This implies that more educated Africans have tended to accept universalistic standards of good governance, standards by which their governors fall short. The negative effect of income on trust may be due to better-off people being more anxious to protect the advantages they have against bad governance that may threaten their wealth and well-being. Older Africans are fve percentage points more likely to trust their political institutions than younger people. This implies that older Africans are more likely to have a traditional faith in whoever holds political power rather than question powerholders as younger and more educated Africans may do. Africans who contact more services are six percentage points more likely to trust their political institutions. This positive response may refect the relative recency of widespread social service provision in Africa.

In the abstract, free elections ought to encourage political trust because they enable citizens to hold their governors accountable and encourage governors to deliver what their voters want. Among Latin Americans and Africans, free and fair elections do not have any significant direct effect on political trust. In other words, fair procedures in choosing governors do not begin to offset the negative effect on trust resulting from corrupt practices by those who win control of political institutions. The views of citizens in formerly Communist countries are signifcantly affected by free elections, but the effect is negative: a reduction in trust of 16 percentage points. This may be due to the tendency of politicians as a class being infuenced by a legacy of manipulative skills that politicians used in Communist times to gain and hold on to power.

Breaking informal standards in Western Europe. Although corruption in the form of bribery is low in Western Europe by comparison with developing countries (fig. 3 in Chapter 4), there is much less difference between continents in the perception of corruption and in distrust of political institutions. Among West Europeans interviewed in the Politicians Behaving Badly (PBB) survey reported in Chapter 6, a total of 15% of Britons and 23% of french named corruption as one of the country’s three most important problems. In Spain, where leading government ministers have been convicted for taking large bribes, 66% see corruption as a big problem. This level of concern is similar to that in developing countries. Moreover, distrust in political parties is as low

or lower in Western Europe than in Africa, Latin America or Eastern Europe and post-Soviet countries. On the PBB scale in which 10 represents complete trust and 0 no trust, the mean score is closer to the distrust end of the scale: It is 4.2 in Britain, 3.7 in france and 3.0 in Spain.

The combination of low corruption by legal standards and a high level of distrust suggest that actions by politicians that break informal standards can have a substantial negative impact on trust. for clarity in presentation, we present a regression analysis for the three-nation pooled data set, and to provide a fne-grained test for cross-national similarities in the causes of distrust separate regression analyses for Britain, france and Spain (Table 3). Given sample sizes of one thousand, we use p-value of less than 0.05 for statistical signifcance in the three national analyses and p-value of less than 0.01 for statistical signifcance in the larger pooled database. We calculate shifts in predicted probabilities for the effect on trust of standardized independent variables (for full details, see Rose and Wessels 2018).

Breaking informal standards of how politicians should behave has a strong impact on political trust. In each of the three regression analyses, the three indicators of the informal violation of standards increase distrust in political parties. Altogether, our parsimonious model accounts for more than 25% of high levels of distrust in political parties and 27.9% in the pooled three-country analysis. Informal corruption has just as much infuence on trust in Western European countries as conventional measures of corruption used to test the effect of corruption on trust in our three-continent analysis (cf. Tables 2 and 3).

Notwithstanding many differences between Britain, france and Spain, all three ways of breaking informal standards consistently reduce trust by similar amounts in all three countries. The behaviour that generates the most distrust in political parties is central to the theory of representative government: elected representatives misleading voters by failing to keep promises that they make when seeking votes. In the pooled data analysis there is a fall of 2.6 points on the 11-point trust scale among people who see politicians saying one thing to win votes and doing the opposite once in offce. The effect is biggest in Spain; dishonesty in political promises lowers trust by 3.6 points. The effect of politicians being seen as taking money for favours is second highest in impact; it lowers trust by 1.5 points in the pooled data analysis and by just over two points in Britain. If citizens think most of their politicians overindulge in their private lives, this has a signifcant effect on trust too, but it is substantially less than that of misleading voters and taking money for favours.

Table 3 Breaking informal standards in Western Europe

Trust in Parties (Maximum effect on 0–10 scale)

Britain France Spain All

Types of corruption Mislead voters −2.149** −2.179** −3.635** −2.622**

Takes money −2.065** −1.035** −1.48** −1.479**

Over-indulgence −0.632 −0.797** −0.988** −0.677**

Seriousness of violation Strict punishment −0.653* 0.071 −0.184 −0.377* Corruption important problem −0.430 −0.099 −0.107 −0.317** Partisanship Supports governing party 1.570** 1.721** 1.083** 1.533** Supports offcial opposition 0.796** 0.914** 0.692** 0.826** Resources Higher education 0.254 0.129 −0.017 0.182 Low education 0.198 0.360* −0.041 0.156 Gender female 0.184 −0.414** 0.208 0.001 OLS adjusted R2(%) 27.4 26.2 25.3 27.9

*Signifcant p-value < 0.05. **Signifcant p < 0.01 Source Politicians Behaving Badly survey organised by the Centre for the Study of Public Policy and the Wissenschaftszentrum Berlin. Nationwide sample surveys in which Effcience 3 interviewed 1004 people in Britain, 1003 in france and 1000 in Spain between 11 December 2015 and 22 January 2016. Note Regression coeffcients show the maximum effect on the 11-point scale for trust of an independent variable moving from its lowest to its highest value. F test for each model p-value 0.000. All independent variables have been recoded to a range of 0–1

The seriousness of the standard broken is of limited importance. West Europeans who see corruption as important are not signifcantly more likely to distrust politicians in any of the three countries, and the effect in the pooled data set is the least of any signifcant infuence. People who think politicians who break informal standards do not need to lose offce or go to jail are just as likely to distrust political parties in france and Spain as those who endorse strict punishments. The effect in the pooled data set is the second smallest among signifcant infuences (Table 3).

Whether people support a party that is currently in government or a party that is currently in opposition but has alternated in and out of government has a signifcant but secondary infuence on political trust. Supporting the party currently in government increases trust in parties generally. At the time of the survey, the parties in government differed in their orientation to the left or right: the Conservatives in Britain, the

Socialists in france and the Popular Party in Spain. However, the impact of partisan support is far less than the impact of the violation of informal standards (Table 3).

Supporting a party that offers to form an alternative government reduces trust in the party system compared to supporting the party in government, but the effect is limited (Table 3). An important corollary is that non-voters and those who vote for protest parties are specially high in distrusting the party system. Collectively, the outsider winning parties are winning as many votes as are established parties in france and Spain. As is the case in the analysis of political trust in developing countries, demographic resources have little or no infuence on trust in parties.

Implications. The judgment of the American political scientist V. O. Key (1966) about his fellow citizens—ordinary people are not fools— appears applicable across continents. In the very different contexts of Africa, Latin America, and Eastern Europe and post-Soviet Eurasia, people make logical judgments. If they see their political system as corrupt and getting worse, they are likely to distrust their political institutions. If they perceive their institutions of governance as high in integrity and successfully keeping corruption down, they will be trusting (van der Meer and Dekker 2011).

Analysing the consequences of corruption globally and in a West European context reaffrms how misleading it is to confne the analysis of the impact of corruption to inferences drawn from global indexes of national corruption. Doing so implies that all forms of corruption are absent in countries that the Corruption Perceptions Index rates as virtually free of capital-intensive bribery and the Global Corruption Barometer shows have very little retail-level bribery too. Since these countries tend to be democracies and to have a high standard of living, this misleading focus also encourages claims that if a country has a political system in which parties can compete in free and fair elections and a high level of Gross Domestic Product then it should be free of corruption in all its forms. The election of Donald Trump as the president of the United States has shown that this is not the case.

It is much more accurate to say that corruption takes different forms in different types of countries. In countries with a high level of Gross Domestic Product, corruption can take the form of breaking informal standards about how public offceholders ought to behave. This can be done by behaving in private in ways that would be shameful if publicly revealed, enjoying a lavish lifestyle by receiving favours in kind or cash,

and by winning free and fair elections by saying one thing to get votes and once in offce making a U-turn that disavows a commitment. While there are similarities between this behaviour and that of autocratic leaders of undemocratic developing countries, the differences are critical. The money spent on politics in Western democracies has usually been acquired legally, whereas in many poorer countries the money can come from the public purse, foreign aid or bribes for capital-intensive contracts. In Western countries, the result of a lavishly funded competitive election is that one big spender loses, whereas in undemocratic countries those in power are immune from defeat.

Paradoxically, free competitive elections may encourage informal corruption. In pursuit of votes, politicians have the incentive to promise the voters what they want to hear, and money buys high-tech research to fnd out what to market in the best way to convince voters. The weakening of party ideologies gives party leaders more freedom of manoeuvre in campaigning. There is an incentive to nominate candidates with little or no experience of national offce so that they can appeal for votes on the grounds that their lack of experience makes them more trustworthy than politicians who have dealt with the constraints of offce. However, this asset turns negative when an inexperienced politician wins offce and fnds that getting rid of corruption is easier said than done. If the dissociation of campaigning from governing is repeatedly demonstrated by whichever party wins offce, this can create a low-level equilibrium trap of distrust in governance.

The contrast between the strong impact of corruption on trust and its slight effect on voting is bad news for anti-corruption advocacy groups that would like to mobilize people who see government as corrupt to use their right to vote to register a protest against bad governance. However, this is not the only way in which political activists and advocacy groups may affect governance—provided that there are the means to make governance transparent.

references

Blais, Andre. 2010. To Vote or Not to Vote: The Merits and Limits of Rational

Choice Theory. Pittsburgh: University of Pittsburgh Press. Charron, Nicholas, and Andreas Bågenholm. 2016. Ideology, Party Systems and

Corruption Voting in European Democracies. Electoral Studies 41: 35–49. Dahl, Robert A. 1989. Democracy and Its Critics. New Haven: yale University

Press.

Dahlberg, Stefan, and Maria Solevid. 2016. Does Corruption Suppress Voter

Turnout? Journal of Elections, Public Opinion and Parties 26 (4): 489–510. Easton, David. 1965. A Systems Analysis of Political Life. New york: Wiley. Key, V.O. 1966. The Responsible Electorate. Cambridge, MA: Harvard University

Press. Kumlin, Staffan. 2004. The Personal and the Political: How National Welfare

State Experiences Affect Political Trust and Ideology. Basingstoke: Palgrave

Macmillan. Mishler, William, and Richard Rose. 1997. Trust, Distrust and Skepticism:

Popular Evaluations of Civil and Political Institutions in Post-Communist

Societies. Journal of Politics 39 (2): 418–451. Olson, Mancur. 2000. Power and Prosperity. New york: Basic Books. Plutzer, Eric. 2018. Demographics and the Social Bases of Voter Turnout. In The

Routledge Handbook of Elections, Voting Behaviour and Public Opinion, ed.

Justin fisher et al., 69–82. Abingdon: Routledge. Putnam, Robert D. 2000. Bowling Alone: The Collapse and Revival of American

Community. New york: Simon and Schuster. Rose, Richard. 1989. Ordinary People in Public Policy. London: Sage. Rose, Richard, and William Mishler. 2011. Political Trust and Distrust in Postauthoritarian Contexts. In Political Trust: Why Context Matters, ed. Marc

Hooghe and Sonja Zmerli. Colchester: ECPR Press. Rose, Richard, and Bernhard Wessels. 2018. Money, Sex and Broken Promises:

Politicians’ Bad Behaviour Encourages Distrust. Parliamentary Affairs, in press. Rose-Ackerman, S., and T. Soreide (eds.). 2011. International Handbook on the

Economics of Corruption, vol. 2. Cheltenham: Edward Elgar. Uslaner, Eric. 2018. The Oxford Handbook of Social and Political Trust. Oxford:

Oxford University Press. van der Meer, Tom, and Paul, Dekker. 2011. Trustworthy States, Trustworthy

Citizens? In Political Trust: Why Context Matters, ed. S. Zmerli and M.

Hooghe, 95–116. Colchester: ECPR Press. Zmerli, Sonja, and Tom van der Meer (eds.). 2017. Handbook of Political Trust.

Cheltenham: Edward Elgar.

CHAPTER 8

Making Government Transparent

Transparency is intended to open up the process of governance to public scrutiny and accountability. It replaces the historical practice of political elites conducting public affairs in private. Doing so restricts information about what happens within the black box of government to a limited number of insiders. In such circumstances, outsiders lack the information needed to intervene effectively in what government does. Ordinary people, receiving public services at the grass roots, ‘know from their own experience how they are treated by public offcials but they do not know how public offcials make decisions affecting them when decisions are taken behind closed doors’ (Worthy and McLean 2015: 347). Open government gives ordinary citizens and civil society groups an opportunity to scrutinize how those in power are using their authority and to hold them accountable if they appear to be abusing it. By opening up the process of governance it reduces the inequality of political information that governors may use to the disadvantage of the governed.

Transparency International describes opening up governance as the surest way of guarding against corruption and increasing trust in political institutions. Transparency is expected to achieve the complementary goals of exposing corrupt practices and, by the threat of exposure, preventing bad behaviour in the frst place. Awareness of public scrutiny should encourage politicians to behave in ways consistent with public expectations of how they ought to behave and administer policies in accord with formal bureaucratic procedures rather than abusing their

© The Author(s) 2019 R. Rose and C. Peiffer, Bad Governance and Corruption, Political Corruption and Governance, https://doi.org/10.1007/978-3-319-92846-3_8 145

This article is from: