The Influence of Public Corruption and Human Values on Trust in the Police: A European Cross- National Perspective The Influence of Public Corruption and Human Values on Trust in the Police: A European Cross-National Perspective

Police forces are an important part of every democratic system and depend on societal legitimacy to perform their duties. Societal legitimacy stems from compliance with binding principles, like human rights and national law, and from broad acceptance from the people. Accordingly, police forces need the public’s trust to perform their duties in today’s European democracies. This paper investigates trust in police forces based on individual and contextual factors found in European countries. To explain trust in police forces, this paper uses human values on an individual level as well as perceived public-sector corruptionon a macro level. The analysis is based on data from the European Social Survey 2016 (ESS) and the Corruption Perception Index 2015 (CPI). It combines individual and contextual factors using multilevel models to explore differing levels of trust in police forces as well as the dynamics between perceived public-sector corruption, human values, and trust in the police. The results confirm previous findings that trust in police forces erodes in the face of perceived public-sector corruption, while human values play a sub-ordered role in the explanation of trust in the police in European countries.


Introduction
In Greece, the level of trust in the police dropped dramatically between 2002 and 2011.1 Such a decline prompts a question: how can trust in police be explained? Public institutions need to be trusted to function effectively.2 For state institutions to fulfill their tasks, they need a certain amount of operational space. But this space can be maintained long-term only if citizens trust theirstate's institutions.3 When citizens trust their state's institutions, they legitimate them.4 Thus, institutional legitimacy depends on citizen trust.
Differences in citizen trust levels can be explained by differences in institutional corruption among states. Interest in corruption has increased notably in recent years, both among social scientists and the public. For example, in 2015 a leak of over 11 million financial documents known as the 'Panama papers' had an impact in Europe.5 Though offshore business is legal, the Panama papers revealed that some shell corporations were being used for illegal purposes, such as fraud or tax evasion.6 Some former and current European leaders were involved. The former Prime Minister of Iceland, for example, had links to anonymous offshore companies.7 Russian President Vladimir Putinhad links to hidden money.8 While instances of real or perceived corruption like those revealed in the Panama papers have gained considerable international attention, corruption is widespread and affects the lives of many people in many societies.9 And when corruption is systematically present in one part of society, it has a lasting impact across thatsociety. 10 While real and perceived public-sector corruptionaffects trust in public institutions and the policeon a macro-level,11human values can explain people'strust in public de/cgi-bin/dokserv?id=2774541&prov=M&dok_var=1&dok_ext=htm. institutions and the police on a micro-level.12 Strikingly, though various factors have been used to explain trust in the police, hardly any study has investigated the influence of human values. Schwartz defines 'values' "as desirable transsituational goals, varying in importance, that serve as guiding principles in the life of a person or other social entity." Each value has consequencesthat may be compatible or in conflict with other values. Schwartz's theory focuses onthe following 10 values: universalism, benevolence, tradition, conformity, security, power, achievement, hedonism, stimulation, and selfdirection. 13 The police can support or undermine values through their actions or inaction. Therefore, the differences in the importance amongthese values across a society's population can be assumed to establish different levels of trust in the police within that society. Human values, therefore, play an important role in understanding differences in trust in the police.14 Since both perceived public-sector corruptionand human values influence trust in the police, this paper studies their separateand combined effect on trust in the police. It begins by explaining the theoretical conceptsbehind trust in the police, perceived public-sector corruption, and human values. This explanation is followed by a description of the data and the statistical methods for the empirical analysis. Then, analysis's descriptive and bivariate results are presented, followed by the results of the multilevel analysis of the 23 European countries and a discussion of them. Finally, this paper draws various conclusions and identifies where more research is needed.

Trust in the Police
Trust is a complex social phenomenon.15For example, trustcan exist between individuals, between a person and an institution, andbetween institutions.16This paper focuses on trust in the police as an institution.
Institutional trust refers to the complex structure of an institution as a whole and not to the individuals within the institution. It focuses on the institution's functioning and serves to reduce the complexity of managing an institution.17 Institutions fulfill a necessary and important role in our complex world. Institutions provide"stable, valued, recurring patterns of behavior." 18 Trust is an important factor for managing and reducing complexity.19 The police are part of the political system and therefore can be regarded as a representative of the government.20 Trust in the police, however, is a special case of trust in state institutions. The police's legitimacy dependson whether the public accepts the police. The police need the trust of the population to properly perform their duty, especially in democracies like Europe's democracies.21 Public attitudes towards the police are influenced by the police's structure and operations; all facets of the police function matter.22

Perceived Public-Sector Corruption
In this paper, "corruption is the abuse of entrusted power for private gain." 23 "It is understood to mean either offering and accepting money, gifts, or other benefits. The aim is to commit someone in the business world to a breach of duty or a breach of trust." 24 Research on corruption has found that corruption has negative consequences.25 In public services, corruption means that goods and services are not distributed fairly 17 Höhne, Vertrauen. among the people.26 Corruption also undermines political processes.27 Corruption is a threat to a good government 28; it diminishes the quality of state-providedservices.29 Furthermore, corruption reduces trust in state institutions.30 This loss of trust in turn leads to a delegitimization of the political system.31 And corruption violates fundamental democratic principles because it can lead to systematic manipulation of political institutions and laws.32 The police are responsible for enforcing formalized norms impartially.33 Many studies show that generalized trust or trust ininstitutions increaseswhen all persons are treated fairly and equally.34 Corruption, on the other hand, violates impartiality.35 Corrupt governmental behavior means benefits the corrupt and at a cost to everyone else. When the police are corrupt, public trust in government erodes along with the police's loss of legitimacy.36 Consequently, perceived public-sector corruption can be assumed to have a negative impact on trust in the police. Based on these theoretical considerations, the following hypothesis can be formulated: Hc: The higher the perceived public-sector corruptionis, the lower the level of trust in the police is.

Human Values
Other important factors that explain trust in the police are human values. To understand individuals, groups, or institutions from a scientific point of view, the concept of value is often used to explain social phenomena. Human values on the individual level express abstract beliefs about desirable goals, ordered by relative importance, that guide individuals as they evaluate events, people, and actions.37 Schwartz analyzed the different value structure of individuals within cultures and providedatheoretical foundation. The Schwartz Value Theory is used instead of Inglehart Theory,38 mainly because of the availability of the data in the ESS and its better-drawn distinctionsamong values and attitudes.39 Schwartz human value theory says that human values form a motivational continuum and that the importance of each value varies between individuals. That means that each individual hierarchically orders his or her values. This hierarchical ordering of values is relatively stable across time and situations.40 In contrast, attitudes are less ordered hierarchically and less stable across time and situations. Each value has consequences, and these consequences may be compatible or in conflict with other values.41 The model in Figure 1 shows the relations between the 10 first-order values in the motivational continuum, as well as the four second-order values. Values next to each other are more similar than values on the opposite side, which often conflict.42 From each of the four second orders of values, the value with a higher theoretical explanation is included for the analyses.
The police have three main duties: firstly, to provide personal protection, secondly, to promotesocial order, and thirdly, to prevent actions that could violate rules, social norms, or citizen's rights.43 These duties are the basic motivations behind the three first-order values security, tradition, and conformity.44Studies show that these three first-order values are often not discriminatory,45 so the value of conservation, which is a second-order value, is used. The hypothesis is accordingly: Hi1: The more important the value of conservation is, the higher the level of trust in the police is.
The downside of the basic functions of the police in particular is that they restrict the scope for action and limit individual freedom.46 However, it is precisely this autonomy that is the basic motivation for the value of self-direction, which is why the assumption is that the police block this value. This leads to the following hypothesis: Hi2: The more important the value of power is, the higher the level of trust in the police is.  The police hold a position of power in society.47 There is a homosociality between people for whom dominance over people and resources is important and the police, who have this authority to enforce the law. Homosociality reduces complexity and thus increases trust.48 From this, the following hypothesis can be formulated: Hi3: The more important the value of self-direction is, the lower the level of trust in the police is.
As a government institution, the police must treat all people fairly and equally.49 This is also the basic motivation behind the universalism value. Therefore, people for whom the universalism value is important will trust the police more because the police protect the welfare of all fairly and equally, thereby satisfying that value. The universalism value is preferred to the benevolence value for analysis since it is more important to examine equality for all, which is assumed by the police, than to capture the emphasis on close relationships, which is covered by the benevolence value. The hypothesis is accordingly: Hi4: The more important the value of universalism is, the higher the level of trust in the police is.

Data
The individual-level empirical data come from the eighth wave (2016)  The country-level empirical data come from the Transparency International's CPI 2015. The CPI ranks countries by the perceived corruption that exists among public officials and politicians. The index collects its data from different sources, including adata composite from country experts, business analysts, and public pools.51 The CPI has a value from theoretical 0 to 100. Zero means the perceived corruption is the highest, while 100 stands for the lowest perceived corruption.

Variables
Trust in the police. This variable is the dependent variable in the model. Respondents were asked the following question: 'Using this card, please tell me on a score of 0-10 how much you personally trust each of the institutions I read out. Zero means you do not trust an institution at all, and 10 means you have complete trust: The police?' Values. The ESS includes a shortened version of the Portrait Values Questionnaire.52 Each value is measured with two items (with one exception: universalism is measured with three items). Respondents answered the following question: 'Now I will briefly describe some people. Please listen to each description and tell me how much each person is or is not like you. The responses are on a six-point rating scale ranging from 'very much like me' (1) to 'not like me at all' (6). The scores of the items were reversed so that higher scores indicate greater value of importance.
Control variables. Variables that had an influence on trust in the police in previous studies, as well as their availability, resulted in the selection of control variables used. Previous research show that older people have more trust in the police than younger people.53 Age is grouped into five categories (15 to 17, 18 to 29, 30 to 45, 46 to 60 (reference category), and 61 and older). A higher educated person has a higher trust in the police.54 The level of education was measured by taking the highest achieved level of education (low, middle, high). Middle education level is the reference category. Gender was used as a control variable (man as a reference). The majority of studies find that women have more trust in the police than men.55 Generalized trust is an additional 51 For further detailed information: https://www.transparency.org/cpi2015. control variable. A higher generalized trust leads to higher trust in institutions.56 A migration background exists if at least one of the parents was not born in that country (no migration background is the reference category). In general, having migrated leads to a lower level of trust in the police.57 A person with right-wing political orientation has a higher trust level.58 For the political orientation, a three-classed scale (left, middle (reference category), right) is used.

Modeling strategy
To analyze individual and contextual factors in the same model, two-level multilevel models are used to explore differing levels of trust in police forces. In this case individuals are nested within countries. It is possible to use a multilevel analysis, although the number of cases on the second level is low with 23 countries. However, distortions of the standard error are possible.59 The calculated models increase in complexity; this strategy is proposed by Hox (2002). To do so the ML win program was used.60 The first model is a so-called empty model, and it only includes the intercept. The second model includes the four human values, followed by the control variables. The third model includes the context variable CPI. Descriptive statistics are calculated with missing values, while bi-und multivariate statistics handle the missing data by using multiple imputation.61 For the imputation the package HMISC 62 in the statistic program R was used.63 All variables were standardized.  Causality analyses can only be generated with experiments 64 and/or panel data.65 Nevertheless, in this paper, the CPI is used upstream to take into account the temporal aspect to generate a clear temporal sequence and thus treat the perceived corruption as a cause for trust in the police.
Individual values are among the most abstract types of cognition,66 and the more abstract a cognition, the more resistant it is to change.67 It can be assumed that values influence the attitudes of individuals. Values are therefore upstream of attitudes.68

Descriptive overview
In Figure 2 trust in the police in European countries is presented. The scale is an 11-pointscale (0 means 'no trust at all,'10 means 'complete trust'). Trust in the police in the European countries vary, and it is high especially in Scandinavia, being thehighest in Finland (8.22), followed by Iceland (7.84), and Norway (7.44). Only in Switzerland (7.23), Austria (7.04), and Germany (7.01) does the level of trust in the police also lie between 7 and 8.
On the other side of the range Russia, has the lowest level of trust in the police (4.43), followed by Israel (5.13), Poland (5.74) and Czech Republic (5.78). Russia and Israel have 11% and 10%, respectively, of answers in the lowest category: 'no trust at all' . The average of all 23 countries lies around 6.47.
The CPI 2015 displays the varieties of the perceived public-sector corruption in the European countries in Figure 3. On this 0-100 scale, a higher score means that the perceived public-sector corruptionis lower, while the lower values mean that the perceived public-sector corruptionin the respective country is higher.
The following picture shows the rating of the CPI 2015 for the relevant countries for this paper. Finland leads this ranking with a score of 90, followed by Sweden (CPI: 89), Norway (CPI: 88), Switzerland (CPI: 86), the Netherlands (CPI: 84), as well as Germany and United Kingdom (each CPI: 81). The lowest score lies with 29 in Russia, followed by Italy (CPI: 44), and Hungary (CPI: 51).    IT  HU  CZ  ES  LT  SI  IL  PL  PT  EE  FR  IE  AT  BE  IS  DE  GB  NL  CH  NO  SE  FI   10

Human Values
In the following Figure 4, the mean scores of all human values of each country is presented. In this 1-6 scale, a higher score means the value is more important. While the value conservation in Sweden with 3.9 is on the lower end of the scale, in Poland and Slovenia the value conservation is on the other end of the scale with a mean of 4.7. The importance of the power value is more diverse within European countries. In France the power value is the most unimportant (2.6) and in Hungary (4.0) the most important. The power value for each country is the most unimportant one. The self-direction and universalism values are more important in European countries than conservation and power. The range for the self-direction valueextendsfrom 4.1 (LI) to 5.0 (CH) and the universalism value has a range from 4.3 (LI, CY) to 5.1 (ES).

Correlations
The highest two correlations are between the universalism and conservation values (0.40), followed by the correlation of the universalism and self-direction values (0.39). The correlation in between generalized trust and the trust in the police is with 0.27 moderate.69 The correlation between trust in the police on the aggregated country level and the CPI 2015 is 0.79. These results must be treated with due caution because of the low case number of 23 countries. In general, in countries with a lowerlevel ofperceived public-sector corruptionthe level of trust in the police is higher. In countries with a higher perceived public-sector corruption the level of trust in the police is lower.

Multilevel analysis
This section describes the results of the multilevel analysis. For the analysis, the model is stepwise expanded with groups of variables. The results are shown in Table 1.
Model 0 contains only the intercepts. It shows that differences lie on the microlevel as well as on the macro-level. This means that both kind of factors are necessary to explain individual differences and differences between countries in their trust in the police. Both variances are significant and the Intra Class Correlation (ICC) is 0.111. Eleven percent of the variance is found on the country level, while 89% of the variance is on the individual level.
Model 1 includes the human values to explain trust in the police. The conservation and universalism values have a positive, significant effect on the level of trust, while the self-direction value has a negative, significant effect. Power has no effect on the level of trust in the police. In this model, the conservation value has the strongest effect on the level of trust in the police. This meansthat peoplefor whom the conservation value is  important have more trust in the police. In contrast, people forwhom the self-direction value is important have a lower level of trust in the police. Model 2 is an expanded version of Model 1. In Model 2 control variables are included, specifically age, education level, generalized trust, gender, migration background, and political orientation. Taking these control variables into account, the universalism value is losing its power of explanation. The conservation and selfdirection valuesremain stable. Both the youngest people, those 15 to17 years old, and the next youngest people, those 18 to 30 years old, have more trust in the police than the people in the reference age group, those 46 to 60 years old. Those who are 31 to 45 years old do not significantly differ from the reference category. The oldest people in the sample, those 61 years old and older, also have higher trust in the police when compared to thosein the 46 to 60 years-old-age group. The level of education plays a role in explaining trust in the police. People with a lower education have less trust in the police than people with a higher level of education.
Generalized trust also is an important factor in explaining trust in the police. People who tend to trust other people in general alsotend to have higher trust in the police. This factor of generalized trust is the most important one in this model.
In terms of gender, women trust the police significantly more than men do. People who have not migrated trust the police significantly more than people who have migrated. People on the political left trust the police less than people in the political center; while people on the right trust the police more.
Model 3 takes the context variable of CPI 2015into account. The effect of this macro variable has a powerful explanatory force. A lower level of perceived publicsector corruptionleads to a higher level of trust in the police. The explained variance in this model compared to model 0 is reduced by about 63%. When compared with model 2 the reduction is 47%, this reduction gets only with one new variable in it: the macro variable perceived public-sector corruption. The ICC lies at 0.047.
The 23 countries have different aggregated levels of trust in the police. The Scandinavian countries have the highest level of trust, while Russia has the lowest.
The study shows, in an empirical way, that a high perceived public-sector corruption findinginfluences the level of trust in the police in a powerfully negative way; therefore, my hypotheses on corruption have empirical support. This result supports earlier studiesanalyzed the link between corruption and trust in state institutions.70 In addition to the perceived public-sector corruption, human values on the individuallevel explain some of the differences in trust in the police. People for whom the conservation value is important have a higher amount of trust in the police. This can be explained by a fundamental police function: to protect the security of every citizen, which, in turns, promotes stability. People who consider the universalism value as important also have greater trust in the police. On the other hand, someone who values self-direction might have less trust in the police because the police can restrict a person's independent actions and liberty. These results show that three of four hypotheses on the individual level (Hi1, Hi3 and Hi4) are supported, while the hypotheses on power is not supported.
Furthermore, the control variables also influence the level of trust in the police. People under 45 years old and people older than 60 years old trust the police more than people 45 to 60 years old. People who have a higher trust in other people in general also have a higher level of trust in the police. A higher educated, female, who has never migrated and who identifies with the political right would trust the police the most.

Discussion and Conclusion
This study's purpose was to discover how human values as well as perceived publicsector corruption can explain the differences in citizens' trust in their police. Although perceived public-sector corruption in Europe is lower than it is globally, it nevertheless influences trust in the police and thus the legitimacy of the police as an institution. A state must first get corruption under control to increase trust in the police. In addition to this relationship at the country level, there are also explanations at the individual level. Human values have an influence on how the police are trusted. However, individual values play a subordinate role. But the police can activate different values by addressing the people differently and thus also build a stronger, more positive relationship with the population. This paper has some theoretical and empirical limitations. From a theoretical perspective, only one main variable on the country-level is used to explain trust in the police. While the explained variance in the trust in the police among the 23 countries decreasessubstantially whenthe perceived public-sector corruptionis considered, further explanatory variables can also play a role in the variance. For example, the police differ between these 23 countries. Police forces are a national institution, they depend on the rules and legitimation oftheir country, they are structuresaccording to their history and needs, and they can differ in their rights and duties.71 Another limiting factor could be the political system in which the police forces are embedded. Understanding post-communist societies (JHU Press, 1998 This paper also faced methodological limitations. The CPI is a composite index based on country-experts and other sources. While the CPI is a good indicator to compare countries, it is not possible to cover the heterogeneity of perceived publicsector corruptionwithin a country. Furthermore, a perception of corruption does not necessarily reveal the real level of corruption.74 Different measurements should be developed and used by further research in the topic of corruption to address corruption's clandestine nature.
More research is necessary to understand corruption, trust in the police, human values, and the connection between these social phenomena.