Chapter 1. Trust matters for governance
This chapter explores the theoretical and practical relevance of trust in government institutions by providing a critical review of literature and presenting the framework and methodology constituting the basis of this report. It introduces the concepts of competence and values as the main drivers of institutional trust according to the public management literature. The measurement approach presented in this chapter allows disentangling the dimensionality of trust and generating meaningful evidence that could inform and guide policy action.
Trust plays a very tangible role in the effectiveness of government institutions. Few perceptions are more palpable than that of trust or its absence. This chapter sheds light on the relevance of trust in government institutions by providing a critical review of literature and presenting the framework and methodology constituting the basis of this report. Trust is a person’s belief that another person or institution will act consistently with their expectations of positive behaviour (OECD, 2017a). While trust may be based on actual experience, in most cases trust is a subjective phenomenon, reflected in the eyes of the beholder. Trust supports the entire range of human interactions, allowing people and businesses to make decisions without having to renegotiate with their counterparts at each interaction. Two different types of trusting relationships are commonly treated in the academic literature: interpersonal or generalised trust and institutional trust (OECD, 2017a). Both of them have been recognised as influencing people’s well-being (OECD, 2017b) through different channels, of which the most relevant are specified below.
Trust reduces transaction costs and improves economic performance: where high levels of trust exist, the need for formal contracts is reduced, thereby reducing transaction costs (Fukuyama, 1995; Knack and Keefer, 1997). This in turn encourages exchanges in the economic and non-economic spheres. Similarly, common expectations reduce the amount of time that has to be spent in agreeing on the exact terms of an exchange. Government performance also benefits from higher levels of trust, which (by reducing transaction costs) can facilitate agreement, collaboration and innovation in state bureaucracies (Knack, 2000).
In turn, trust promotes efficient allocation of resources. Trust and co-operative norms enable forms of collective action which can replace or supplement market or government-based action, such as natural resource management, or respond to emergencies. Collective action strategies based on trust and co-operative norms often provide the most efficient way to manage common resources such as forests, agricultural land or fisheries stocks, thereby contributing to the maintenance of environmental quality. Mechanisms for collective action are also important to ensure the production of various public goods, on which the effective functioning of markets (and several aspects of wider well-being) depends. This applies at both the national and international level – and “global public goods”, such as the environment, security, financial stability, and respect for property rights – require international co-operation to be sustained (Kaul et al., 1999; Sachs 2008).
In addition, trust also influences quality of life and human capital formation. People living in higher-trust and more co-operative communities tend to be happier (Helliwell and Putnam, 2004; Helliwell and Wang, 2010; Hudson, 2006), healthier (Lochner et al., 2003), better educated and more civically engaged (Putnam, 2000), and less likely to be victims of crime (Sampson, Raudenbush and Earls, 1997). Although the causal mechanisms for many of these relationships remain unclear, trusting relations have been shown to influence behaviours relevant to individual well-being.
Systemic or institutional trust focuses on the interaction between government institutions and citizens and within government. In general terms, institutional trust happens when citizens appraise the government and their institutions in general and/or the individual political leaders as promise-keeping, efficient, fair and honest (Blind, 2006). But in this realm, a number of nuances and additional trust relationships can be considered, from different aspects of citizen trust in government institutions to whether government agencies trust citizens or each other.
This report is primarily concerned with institutional trust and its drivers. As a starting point an important distinction should be made between political and institutional trust. Political trust refers to an assessment of the elected leadership, while, for the purpose of this report, institutional trust has been restricted to trust in government institutions defined1 as the “central administrative branches, local governments, public enterprises, public institutions and all other kinds of social organisations”. Nevertheless, a key challenge when addressing institutional trust is that these dimensions (i.e. institutional and political trust) are not fully independent. Evidence for a handful of countries with sufficiently long time series (i.e. the United States, Norway and Sweden) has shown that a negative evaluation of government performance by citizens, which (as will be explained later) are considered a key driver of institutional trust, leads to political distrust (Miller and Listhaug, 1999). In turn, Dahlstrom and Lafuente (2017) find strong evidence that a clear separation of officials at the top of government and the creation of two independent channels of accountability (politicians who are responsive to the electorate and bureaucrats to professional peers) lower levels of corruption, increases the effectiveness in the delivery of public goods and could therefore result in higher institutional trust.
In addition to the distinction between institutional and political trust, another distinction has been made by Bouckaert (2012a) between macro-, meso- and micro-level trust, depending on whether trust is related to the functioning of the democratic system, policy making or service delivery. Others use “macro and micro” to describe trust in the political system, its institutions or in the personnel in charge of these institutions (Blind, 2006). Still others address how trust within government (among government institutions or between government and its employees) may influence efficiency in government outcomes through co-ordination or principal-agent failures. Trust can also be analysed from the point of view of how much government institutions trust people, and whether this influences, for example, the regulatory activity of the state (Yang and Holzer, 2006).
There is another conceptual distinction to make with respect to institutional trust. A common theme in the academic literature on institutional trust (e.g. Nooteboom, 2007) is the distinction between “trust in competence” (i.e. whether the functioning of institutions matches people's expectations about the competencies of those steering them) and “trust in intentions” (which captures whether institutions act in a way that is perceived by people as ethical and fair). These distinctions are extended by Bouckaert (2012b), who distinguishes between the “logic of consequences”, where trust is derived causally from outcomes, and the “logic of appropriateness”, where trust is based on the values of integrity and transparency. This distinction between the outcomes of an action and the intention that guided it forms the basis of the OECD Trust Framework, endorsed by the Organisation’s Public Governance Committee.
The importance of institutional trust is manifold. Trust in government institutions is one of the foundations upon which the legitimacy and sustainability of political systems are built. It influences individual behaviour in ways that could support desired policy outcomes. This may range from rather narrowly defined policies and programmes (such as participation in vaccination campaigns) to broader policy reforms (e.g. environmental regulation or pension reform). Trust is important because many public programmes create the opportunity for free riding and opportunistic behaviour. Trust could reduce the risk of such behaviour to the extent that people are prepared to sacrifice some immediate benefits if they have positive expectations of the longer-term outcome of public policies, either at a personal level (pensions) or by contributing to the common good (e.g. redistribution of income through taxation) (OECD, 2013).
In turn, trust in government institutions may help governments to implement structural reforms with long-term benefits. Many reforms involve sacrificing short-term satisfaction for longer-term gains and will require broader social and political consensus to be effective and sustainable. In a high-trust environment, such reforms may not only be properly enacted and implemented, but could be sustained long enough to bear fruit. This extends the time frame for policy decisions. In a low-trust climate, citizen will prioritise immediate, appropriable and partial benefits, and will induce politicians to seek short-term and opportunistic gains through free-riding and populist attitudes (Gyorffy, 2013).
While trust takes time to be established, it can be lost quickly. It is not sufficient to discuss the impact of trust in government institution on the performance of those institutions, the economy and society; it is also necessary to describe what might happen if there is an increasing distrust in government. This might lead to less willingness on the part of citizens (and businesses) to obey the law, to make sacrifices during crises or to pay taxes. This could raise costs for government institutions – resulting in declining efficiency – or erode revenues. Declining trust might also make it more difficult to attract and retain talent to work for government institutions.
Finally, it is key to emphasise that – as will be described in further sections –, this report focuses largely on better understanding how changes in public policies could influence the trustworthiness (i.e. the notion of deserving trust or confidence) of public institutions, and how governance changes may strengthen or weaken the trust drivers, with a particular focus on South Korea and its specific context. This scope reflects the concerns of leaders and policy makers about the deterioration of institutional trust in the past few years and their interest in “actionable” policy insights to help them reverse this trend.
Approaches and challenges for measuring institutional trust
Still, trust remains an abstract concept encompassing several actors and instances. The complexity of trust relations could be illustrated by Table 1.1. This framework classifies measures of trust primarily in terms of the parties involved in the trusting relationship, and it has the advantage of capturing a very comprehensive range of situations. However, the framework also has some limitations in that it primarily focuses on distinguishing individual trust from the different elements of institutional and political trust.
Despite its limitations, the framework in Table 1.1 is useful for narrowing down the scope of this report. Some of the types of trust identified by Table 1.1 are not suitable for measurement in surveys (e.g. civic trust), which – as will be explained later – is the methodology followed by this report. Similarly, trust by institutions or organisations (as opposed to trust in institutions) is neither suitable for measurement through surveys of the general population, and hence none of the measures of trust by institutions are covered by this report. Similarly, a household survey cannot specifically target political leaders as respondents, making this group out of scope even if there were no other reason to avoid collecting data of this sort. For this reason, the scope of the report is kept relatively specific, focusing on people's trust in government institutions (i.e. institutional trust), its measurement and policy implications of the evidence.
A body of cross-country comparative survey-based measures of institutional trust is available. For many years, the main source of internationally comparable data on institutional trust has been the World Values Survey, which started collecting these data in 1981. More recently, a wider range of non-official sources have provided comparative data on trust, including the Asian Barometer, Gallup World Poll (GWP), the European Social Survey (ESS) and the European Quality of Life Survey (EQLS), while in 2013 the European Union Statistics on Income and Living Conditions (EU-SILC) provided first official estimates for European countries. The OECD has compiled a wide range of institutional trust measures in the OECD Trust Database (see Box 1.1) The size of the database and range of co-variates make it possible to identify the underlying patterns captured by survey-based measures of trust in institutions and systematically test the accuracy (i.e. reliability and validity) of these measures.
A key challenge in building a better understanding of the drivers of trust and its impact on other outcomes is the limitations of the available data. This limitation has also had an important impact on the ability to assess the validity and reliability of trust data rigorously.
The OECD Trust Database is an effort to map existing sources of (so far non-official, apart from EU-SILC) data and compile them into a single repository of information. Table 1.2 displays the different surveys included in the OECD Trust Database. Its coverage goes beyond OECD states and includes up to 124 countries, spanning the period between 2002 when the collection of trust data became more common and 2015 or the latest year available.
Different surveys have different geographical coverages and collection frequencies. In the case of the Gallup World Poll, data are collected annually for countries in all regions of the world, while for other surveys – including the WVS, ESS and the European Quality of Life Survey (EQLS) – the coverage is limited to a smaller set of countries. Data collection takes place every 2 years for the ESS, every 3 years for the EQLS and roughly every 5 years for the WVS.
Both institutional and interpersonal trust feature in the OECD Trust Database. In the case of institutional trust, questions are traditionally formulated through a common heading (e.g. “do you have confidence in your…”) followed by a list of primarily public institutions (e.g. government, congress, etc.) and, less commonly, private (e.g. major companies). Survey wording varies considerably, both in terms of the general construction of the question and in the use of the term trust or one of its various synonyms (e.g. confidence). Also, different surveys, or even different questions within the same survey, make use of different response scales. For example, while the Gallup World Poll relies primarily on a “yes/no/don’t know” response format, other surveys such as the ESS and EQLS use longer numeric scales (0-10 in ESS and 1-10 in EQLS). In the case of the WVS, questions are usually answered using a 4-point Likert scale (i.e. “a great deal”, “quite a lot”, “not very much” and “none at all”). In the OECD Trust Database, different questions have been re-scaled to a binary “yes/no” format that allows comparability across surveys.
Further observation of the available data on institutional trust reveals that questions sometimes refer to similar concepts while using quite different descriptions. For instance, some surveys refer simply to “the courts” while others ask about the “judicial system”. While in most cases the interpretation of these concepts is straightforward, in others the lack of clarity may have more significant implications. For example, although most surveys ask about trust in government, the ESS asks about trust in politicians, and EU-SILC addresses trust in the political system. In turn, the Latinobarometer has included questions about trust in government, the state and public administration; concepts traditionally related to each other but not strictly synonymous.
Generally, non-official household surveys provide less information on interpersonal trust compared to institutional trust and, where interpersonal trust is included, the focus is limited to generalised trust. The most common question (asked by four of the surveys under study), which is very similar to the version introduced by Rosenberg in 1957, is the following: “Generally speaking, would you say that most people can be trusted or you can’t be too careful in dealing with other people?”
In the case of the WVS, in addition to “most people”, the questions enquire about additional parties to be trusted (e.g. family, neighbourhood, known people, people met for the first time). Only the EQLS asks a direct question about trusting people in general.
The surveys included in the OECD Trust Database vary in terms of data quality. While all of the surveys have different strengths and weaknesses, some of them place a greater emphasis on methodological rigour and consistency than others. Both the ESS and EQLS are directly funded by the European Commission and aspire to very high standards of data quality. A great deal of attention is paid to consistency across countries, and changes between “waves” are carefully managed. Covering a much greater range of countries, the Gallup World Poll varies more in survey content from wave to wave, but retains a strong focus on methodological consistency and minimises the impact of questionnaire changes by having a fixed core questionnaire. The WVS has evolved over time, and data quality is higher in more recent waves than in earlier ones. In contrast, both the Eurobarometer and Latinobarometer put more emphasis on responsiveness to policy issues, and thus the questionnaires change more frequently and the response rates are generally lower. The OECD Trust Database is available online as an electronic annex of González and Smith (2017).
Source: Gonzalez, S. and C. Smith (2017), “The accuracy of measures of institutional trust in household surveys: Evidence from the OECD Trust Dataset”, https://doi.org/10.1787/d839bd50-en.
In any case, the availability of data is not a guarantee of its fitness for use.2 Recent evidence on the accuracy (i.e. reliability and validity) of trust measures based mainly on non-official household surveys has found that room for substantial improvement exists in the case of institutional trust measures. While the reliability (the degree to which repeated measurements of the same thing produce the same results) of institutional trust measures is relatively good, evidence about the validity (the extent to which the measure in question is biased) is mixed. Institutional trust measures generally perform relatively well in terms of construct validity (measures behave in a way that is consistent with expectations), but the situation is less clear with respect to face validity (the degree to which a measure is intuitively plausible) while evidence to test convergent validity (i.e. it correlates well with other measures of the same construct) is insufficient (González and Smith, 2017; OECD, 2017a). All in all, existing measures of institutional trust have been considered experimental in nature, calling for further development in the context of official statistics (OECD, 2017a).
The shortcomings of existing trust measures are diverse. Some of them are technical (e.g. sampling, questionnaires) some are conceptual (e.g. the meaning of “government”). How general the question is generates substantial ambiguity on the perspective respondents may take. Some of these shortcomings are related to the scope of existing surveys. Existing cross-country comparative surveys ask specifically about government as a set of institutions, including the different branches (i.e. parliament, judiciary). Additionally, while it is clear that trust in government refers to the executive branch it does not consistently differentiate between the different layers within it, namely politicians and public servants. Of the cross-country comparative surveys included in the OECD trust dataset, only the World Values Survey includes a question about trust in the civil service.
In turn, the notion of trust itself as defined previously is associated with expected positive behaviour. Existing measures fail to specify the type of interaction or behaviour under consideration: which aspects of trust the survey respondents are being asked to think about. The analysis of existing evidence also sheds light on the fact that there is no single framework for classifying the different approaches to measuring trust in the academic literature. Nonetheless, drawing on a range of sources, it is possible to identify several distinct measurement approaches. As stressed previously, at the most basic level, a long tradition of survey questions has directly asked people about their trust in institutions (e.g. WVS). Compared to interpersonal trust, the literature and findings on institutional trust are more limited. For example, in the case of interpersonal trust Morrone, Tontoranelli and Ranuzzi (2009) make a distinction between traditional trust questions and attempts to measure trust through peoples’ expectations about others’ behaviour; specifically, a survey question on whether a lost wallet is likely to be returned. Although the use of such “expectations questions”, drawing on specific hypothetical scenarios is, so far, relatively limited, they set a distinctly different conceptual task for respondents than direct questions about trust, and thus might represent a source of additional information.
Beyond this, a large body of literature has compared actual trusting behaviour in experimental settings with survey questions on trust (Glaeser et al., 2000; Fehr et al., 2003; Gachter, Herrmann and Thoni, 2004; Lazzarini et al., 2004; Naef and Schupp, 2009; Johnson and Mislin, 2012; Algan and Cahuc, 2013; Falk et al., 2016). Although experimental approaches to measuring trust might be regarded as beyond the scope of traditional household surveys, these measures provide important insight into the validity of more conventional survey-based measures, and they have been used as a basis to develop better survey questions (Falk et al., 2015). Finally, it is important to note the existence of a wide suite of questions grounded in people’s experiences that, while not focused directly on trust, can provide information on the subject. Taken together, these different approaches to measuring trust can be organised into four broad groups: evaluations, expectations, experiences and experiments, of which the first three have been traditionally based on surveys.
Evaluations: these focus on the respondent’s response to questions on whether they have trust in an individual or institution. These questions ask the respondent to make an evaluation of their own feelings and/or beliefs at the current point in time, rather than recalling information about past experiences or speculating about the future. Most of the commonly used survey questions related to trust are evaluative. An example of an evaluative question is the standard generalised trust question by Rosenberg, used in the WVS: “I am going to name a number of organisations. For each one, could you tell me how much confidence you have in them: is it a great deal of confidence, quite a lot of confidence, not very much confidence or none at all?” Organisations include government, political parties, civil service and parliament.
Expectations: an alternative to asking people to evaluate how they feel is to ask them about their expectations of what would or will happen in a given, hypothetical situation. In further sections of this chapter, the term “situational” will be used to refer to these expectations questions. Because these questions focus on expectations of how people will behave, it is sometimes argued that they are more specific or quasi-behavioural than evaluative questions (Morrone, Tontoranelli and Ranuzzi, 2009). The lost wallet question used in some waves of the Gallup World Poll is an example of an expectations question: “If you lost a wallet or a purse that contained items of great value to you, and it was found by a stranger, do you think it would be returned with its contents, or not?” In the case of institutional trust, these types of questions have not been commonly formulated.
Experiences: an alternative to asking what people expect to happen in a given situation is to ask the respondent about past experiences. In the context of institutional trust, questions relating to people’s experiences of corruption or discrimination have been used in both research and official surveys. Experience questions have not traditionally been a major source of data on trust, but are included here as they have the potential to provide information on the degree to which respondents have experienced situations where their expectations of (positive) behaviour from others have or have not been met. An example of such questions could be extracted from the European Quality of Government Survey (EQoG): “How much do you agree with the following statement: certain people are given special advantages in the public education/health care system?”
Experiments: an alternative to collecting information on trust through surveys is to use experimental techniques to measure trusting and trustworthy behaviour by participants, either in the field or in controlled laboratory conditions. While certainly relevant, this type of exercise is beyond the scope of this report. Still, the OECD has launched the Trustlab initiative (see Box 1.2.) in Korea in partnership with the Korea Development Institute (KDI).
The OECD has launched Trustlab, a joint initiative between the OECD and a range of academic and government partners, to carry out a series of comparable experimental studies on trust across OECD countries. The OECD has conducted Trustlab studies in Korea in partnership with the Korea Development Institute (KDI).
Trustlab’s studies are based on a nationally representative sample of 1 000 individuals stratified by age, gender and income. In the current set up, respondents log onto a custom-made online platform to participate in three experimental games to provide behavioural measures of social norms and values. A number of games study interpersonal trust: the Dictator Game, which provides information on altruism; the Trust Game, which provides information on trust and trustworthiness; and the Public Good Game, which provides information on willingness to co-operate and contribute to public goods. In all three cases, respondents are matched with other respondents for the games, which are played with approximately USD 10 in actual money at stake. In some countries, Trustlab includes “conditional trust games” that assess trust in specific population groups. For institutional trust, after the games respondents complete a series of implicit association tests on attitudes towards a range of institutions (such as the government, the judicial system and the media).
Implicit association tests are a psychometric technique used to test respondent attitudes where issues of social desirability may make them unwilling to respond honestly, or in areas that are difficult to measure through explicit self-reporting due to lack of awareness (Greenwald, Nosek and Banaji, 2002). These tests have been applied successfully to measure perceptions, stereotypes and attitudes towards commonly stigmatised social groups such as black people, women and the elderly (Dasgupta and Asgari, 2004; Aberson, Shoemaker and Tomolillo, 2004). The final part of Trustlab involves respondents completing a survey questionnaire containing a battery of trust questions. These cover a range of different approaches to measuring both interpersonal and institutional trust; self-reported items on other social norms, such as altruism and reciprocity; and a range of questions on the potential policy drivers of trust, along with basic demographic and socio-economic information.
Source: Adapted from Murtin, F., et al. (2018), "Trust and its determinants: Evidence from the Trustlab experiment", OECD Statistics Working Papers, No. 2018/02, OECD Publishing, Paris, https://doi.org/10.1787/869ef2ec-en. .
The different approaches to trust identified by these four groups do not represent substantively different concepts of trust. Rather, they correspond to different approaches to measuring the same concept. The first three approaches are all survey-based and vary in the focus of the question asked to respondents. In particular, while evaluative questions focus on the respondent’s views at the time of the interview, expectations questions have a future orientation and relate closely to the concept of trustworthiness (e.g. the notion deserving trust or confidence), and questions on experiences focus on the past. In contrast, an experimental approach to measuring trust captures actual behaviour from respondents in a situation that is designed to elicit trusting (or non-trusting) behaviour.
Existing methodological challenges for trust measurement have not impeded the wide use of trust data. Trust in government institutions or the lack of it has received increased attention and received many headlines in the aftermath of the 2007-08 crisis. The source of trust data that is most commonly used and has the largest coverage is the Gallup World Poll (see Box 1.3). Different sets of analysis on the basis of the Gallup data reveal trends and correlations that are revealing of the state of trust in government in OECD countries and may assist policy makers in digging deeper into the subject. On the one hand, based on data levels, it has been found that trust in government is highly correlated with the approval of current leadership and the perception of government corruption (OECD, 2015). On the other hand – and using longer time series – evidence has been found on the link between trust in government and economic outcomes (e.g. gross domestic product per capita, unemployment rate) as well as overall life satisfaction (González and Smith, 2017; OECD, 2017a).
This evidence suggests that while trust in government has predictive power, it is influenced by a wide range of determinants that limit the capacity of institutions to act on existing metrics. The existing trust measures’ limited capacity to discriminate calls for better understanding of the drivers and dimensions of trust, in order to disentangle its multidimensionality. Refined measures of the drivers of institutional trust are therefore required to enable government institutions to propose and adjust actions aimed at regaining trust from their citizens. In conclusion, addressing the concerns of policy makers over declining levels of trust, and improving current measurement tools, requires not only a narrowing down of the scope of the trust relationship to be measured, but a more operational notion of trust that can be deconstructed into meaningful policy questions. The next section proposes a framework, based on the public management literature, to be used as basis for developing questions aimed at capturing the trustworthiness of public institutions.
Despite the complexity of the subject, a variety of approaches and scattered empirical testing find consistently across the literature that institutional trust is affected by at least two key aspects. First, the literature highlights two different but complementary components that matter in understanding and analysing trust: 1) competence – or what concerns operational efficiency, or ability, capacity and good judgement to actually deliver on a given mandate; and 2) values – or the underlying intentions and principles that guide actions and behaviours. Second, there is consistency in the literature regarding specific attributes that matter for trust, both in relation to the “competence” and “values” components (see Annex A).
On average, trust in OECD governments has slightly (by 3 percentage points) decreased since 2006 – a likely consequence of the global financial crisis, with significant up-and-down fluctuations that could be explained by diverse type of factors (e.g. economic conditions, long term changes in political systems, short term approval of the incumbent government events such as natural disasters or corruption cases as well as how public institutions perform their roles). According to international available data, on average, only 42% of citizens have confidence in national government in OECD countries (2016). Further, trust levels generally show a mixed picture. The larger drops in trust occurred in countries facing either a political, fiscal or economic crisis, such as Chile (a 38-percentage point decrease since 2007), Greece (-36 p.p.) Finland (-27 p.p.), United States (-26 p.p.), Slovenia (-23 p.p.) Spain (-23 p.p.). In the case of Korea, trust in government between 2016 and 2006 remained relatively stable at a comparatively low level (24% in 2016) (see Figure 1.1).
Note: Data refer to the percentage who answered “yes” to the question: ‘‘Do you have confidence in national government?’’ (data arranged in descending order according to percentage point change between 2006 and 2016). Data for Iceland are 2013 rather than 2015
* In the English language, it is recognised that there is a definition gap between trust and confidence. Some international surveys use the term trust (Eurobarometer, World Economic Forum, Edelman) while others use the term confidence (World Values Survey, Gallup, Latinbarometer). A strain of research argues that trust relates to the individual expectations of receiving effective and fair treatment from public institutions. In contrast, confidence relates to the belief that overall public institutions are effective and fair (Roberts and Hough, 2005). As a result, it may be argued that comparability between surveys could be difficult. Part of the problem however stems from the generality of available trust/confidence questions, a gap that the methodology of this report has sought to address. In addition, the OECD Statistics Directorate has addressed this question in collaboration with the Office of National Statistics of the United Kingdom. Results from split sample testing run in 2016 suggest that the use of the word “confidence” or “trust” does not have a significant impact on the results.
Source: OECD calculations based on Gallup World Poll, www.gallup.com/services/170945/world-poll.aspx
The OECD measurement and policy framework
As outlined in the previous section, several questions calling for complementary evidence remain to be answered. For example, are respondents to surveys able to distinguish between trust in government institutions and trust in political leadership? What is the relative effect of the different drivers of institutional trust, in general, and particularly for the Korean context? How can novel measurement approaches inform our understanding of trust in government institutions? What are the trusting patterns for different population groups?
Limitations to the understanding of what drives trust in government are partly due to the lack of an analytical framework that can help organise concepts, links and causality relations beyond perceptions. To answer those questions, both a theoretical framework and empirical evidence are required. Such a framework should take stock of the key findings identified in the relevant literature and test them in an integrated way, while allowing the development of actionable metrics. The OECD’s approach to trust in government institutions develops an analytical framework which offers an instrumental approach to citizens’ trust in public institutions, facilitating measurement efforts (both based on experience and expectations) and policy attempts to influence trust.
As mentioned previously, the focus of the framework is on trust in public institutions, i.e. on better understanding how trust influences the outcomes of public policies and how governance changes may strengthen or weaken trust. The goal is to unpack trust by the general public in government as a network of institutions, instead of the more short-term issue of political leadership. This scope reflects the concerns of leaders and policy makers about the deterioration of institutional trust in the last few years, and their interest on actionable policy insights to help them reverse this trend, something that existing measures of trust are not able to offer (OECD, 2017c). The framework identifies two key components of trust in institutions: competence and values (See Box 1.4). Within each component, relevant dimensions that are amenable to policy change are identified based on the common threads in the literature (Mcknight, Choudhury and Kacmar, 2002) and on the OECD update of this evidence (see Annex A).
The starting point of our analysis was the mapping of a generalised set of government mandates, i.e. the basic delegation of functions that informs the social contract between society and its elected representatives. While the list can be formulated in a number of ways, essential public governance concepts point to the following key mandates: to provide public services; anticipate change and protect citizens; use power and public resources ethically; listen, consult and explain to citizens; and improve living conditions for all. For each mandate, we then mapped the concerns, or specific expectations which, if and when unmet, could lead to a breakdown of trust (Table 1.3).
Building upon an extensive literature review, we grouped these attributes into two broad components of trust: competence, or the ability of governments to deliver to citizens the services they need, at the quality level they expect; and values, or the drivers and principles that inform and guide government action. Each of them was then further deconstructed into key dimensions, or drivers of trust in public institutions (Table 1.4).
This approach led to the formulation of a unique framework for understanding trust in public institutions, strongly focused on actionable policy insights.
To start with, the framework deconstructs trust into two key components – competence and values – that closely follow the broad distinction reflected in the literature between the actual outcome of an action and the intention that guided it. Second, for each component, the framework identifies relevant dimensions that could make it amenable to policy change: responsiveness, reliability, integrity, openness and fairness. The framework is summarised in Table 1.5 and Box 1.4. A detailed description of these dimensions is presented in Annex B of this chapter.
This framework provides a basis to operationalise alternative measures of trust in public institutions and better link the policy discussion on trust to an actionable reform agenda. By focusing on the role of public institutions, we partly build on the finding of Helliwell et al (2018) that service delivery quality and democratic quality play an important role in supporting better lives although the relative importance varies according to the development level in different countries.
As with any analytical framework, it is not possible to capture every possible related element. Governments perform many functions and are subject to seemingly conflicting mandates (e.g. security versus respect for privacy or transparency). Citizens are influenced in their assessment of trust by exogenous factors, such as culture and attitudes towards authority. Policy relevance may also be extracted by means of better understanding the potential causality between different dimensions of trust (e.g. political vs. institutional, interpersonal vs. institutional) and the influence of context specific factors (culture, system of government). In addition, there are dimensions (e.g. political institutions; see Box 1.5), which are not aligned with the goal of facilitating a more operational approach to the understanding of trust in public institutions by addressing dimensions that are both critical and amenable to policy change.
There are certainly factors that fall outside the OECD Trust framework, which focuses on core competences and values of public institutions on the performance side. Trust in public institutions is clearly influenced by the demographic and other characteristics of respondents on the perception side. One of them is political views. For instance, Park and Lee (2012) posit that public perceptions of public policies are much more influential in explaining the level of trust in government than individual traits (sex, income or political preferences) or civic culture (participation, social trust) in Korea.
Focusing on the fact that a democratic government maintains public legitimacy and trust through elections, Cho (2013) examined how voters’ electoral experiences influence their trust in government. By analysing the Korean data culled from the Asian Barometer Project in 2011, this study demonstrates that government trust was high among those voters who evaluated the last presidential election as being fairer and freer. Based on these findings, the author argues that elections should be considered as an important variable to examine in studying government trust, and they should be made more fair and free.
In some contexts positive relationships between institutional trust and educational status and, although less strong, income has been found (Helliwell and Wang, 2010; Christensen and Laegreid 2005). However, in the Korean context evidence of the importance and type of these relationships is not conclusive (Cho 2013 and Lee 2014)
Kum and Baek (2010)’s empirical analysis shows that the effects of political trust (i.e. trust in the incumbent conservative government at the time) are significantly more pronounced among conservatives than among liberals, which means the effect of ideology on support for government spending is moderated by political trust. This finding suggests that political trust has a normative importance as well as practical consequences in understanding the policy process. Park (2006) also finds that supporters of the ruling party have a higher trust in the government than the supporters of the opposition party. Wong, Po-san and Hsin-Huang (2011) found that institutional factors, particularly the economic and political performance of government, are powerful determinants of political trust, whereas the effects of such cultural factors as post-materialism, traditionalism, and authoritarianism are either insignificant or weak.
This approach is consistent with the fact that citizens view and assess government not only from the perspective of service delivery, but also according to the efficacy and fairness of government policy and government ethics (Yang and Holzer, 2006). It brings renewed attention to process in addition to output. Further, it is consistent with the general finding that institutional trust happens when citizens appraise public institutions and/or the government and individual political leaders as promise-keeping, efficient, fair and honest (Blind, 2006). Lastly, it can provide better guidance to the measurement of trust, monitoring trust over time and the analysis of which factors may drive it in the future, opening the door to an alternative set of data to that currently available.
In addition, it is proposed to use this framework to assess “trustworthiness” of institutions rather than trust in institutions per se. The literature on the measurement of generalised trust has made a distinction between trust and trustworthiness. While trust has been commonly captured by measuring attitudes, trustworthiness is traditionally associated with expectations on future behaviours (Ben-Ner and Halldorsson, 2010). In the past, surveys of people’s attitudes have been weak predictors of people’s actual behaviours (Ajzen and Fishbein, 1977). In consequence, recent developments suggest that trustworthiness might be a better predictor of actual trusting actions (Johnson and Mislin, 2011). Building “trustworthy” institutions is something that governments and policies can do. Further, with the right statistical measures, the impact of policy efforts to build “trustworthy” institutions could possibly be assessed.
The next section builds upon the proposed analytical framework to identify avenues for potential improvement of existing measures of trust by citizens in public institutions. For this purpose, measurement could consider the dimensions of trust as actionable policy drivers and design, for each of them, questions that link more directly with individual experiences, expectations, attitudes and behaviour. Questions along these lines could be included in household surveys without running the risk of entering into political opinion surveying. In addition to monitoring trust better, this would allow countries to contrast perception with government performance, creating a clearer gap analysis that could guide policy and communication decisions.
The OECD-KDI trust measurement instrument
To address the challenges specified in the previous section of this chapter, the OECD and the Korean Development Institute (KDI) designed a pioneering survey instrument to be applied in Korea. The survey was fielded in early 2016 to a representative sample of the Korean population (see Box 1.6, and the survey instrument and methodological details in Annex B). In addition to a series of questions on standard socio-economic characteristics of the population (e.g. age, education, income, gender, religion) and interpersonal trust, the survey focused on the drivers of institutional trust as outlined in the previous section, where the policy and measurement framework was presented.
The joint OECD-KDI survey was conducted through face-to-face interviews from 20 January to 22 February 2016. The main focus of the survey was on administrative institutions of government. As such, respondents were requested not to consider subjective feelings about political organisations or political agents to the extent possible, even though it is acknowledged that this distinction could be difficult to make as the heads of administrative institutions are either elected officials or political appointees. A full description of the survey methodology can be viewed in Annex B. The essentials of the survey design are below.
Subjects and time period
-
population: all general citizens over 20 years old
-
sampling size (number of respondents): 3,000 people
-
time period: 20 January – 22 February 2016.
Sampling design and sampling
-
sampling frame: Korea Census 2010
-
sampling method: stratified random sampling
-
stratification criteria: region, gender, age
-
data collection method: face-to-face interview using a structured questionnaire.
The questionnaire includes several types of question on institutional trust based on evaluations (i.e. standard trust questions), experiences and expectations/situations related to the different drivers of the Trust Framework. For example, in addition to a formulation based on the standard trust question (i.e. How much confidence do you have in the government to act in the best interest of society?), other questions based on evaluations of public institutions attributes are of the type described in Table 1.6.
A number of lessons from the evolution in the measurement of interpersonal trust were applied for the questionnaire design. Traditionally, the appraisal of interpersonal trust has been done through perception questions, the most commonly asked one dating back to 1956 (the “Rosenberg question”). Recently, however, others have proposed a different approach, which moves away from perception and instead focused on specific situations. These questions are commonly referred to as “wallet questions” (see Table 1.7). In previous sections of this chapter these questions have been defined as quasi-behavioural or expectations based; from now on we will refer to them as “situational questions”, knowing that they all refer to the same concept.
The situational questions developed for our survey are not stereotypical behavioural questions: they don’t focus on the individual behaviour but rather on the expected conduct from a third party, as in the wallet question with neighbours, police officers or strangers. As such, they rather provide measurement on the trustworthiness of a given institution, community or society. Differently than attitudes (passive response) and behaviours (active response), trustworthiness is based on expectations. No attempts have been made to measure institutional trust with situational-type questions. There is a missing bridge, and thus an opportunity, between the measurement of interpersonal and institutional trust. In general terms, a situational approach to measuring institutional trust would require designing and asking the following type of questions: “How do (citizens) think their government would behave under a given specific circumstance(s)?” or “If Y happens, do you trust public institutions to do X?”
In line with OECD recommendations (OECD, 2015, 2017a), a numerical 0-10 scale with verbal scale anchors is recommended to measure trust items3. Such a scale allows for a high degree of variance in responses, increases overall data quality and facilitates translatability across languages (OECD, 2017a). In turn, as suggested, the verbal description of the scale anchors represent absolute responses (i.e. no confidence at all and full confidence) to minimise acquiescence bias and socially desirable responding to allow for the full spectrum of possible responses.
One of the main challenges in designing situational questions is to identify relevant situations that capture the underlying concept being researched, but which are still amenable to an informed judgment by survey respondents. For the questionnaire design, a number of high-visibility events in Korea were consulted that influenced trust in public institutions and which occurred before the survey was fielded. Among the key ones are the 1987 June Democracy Uprising, which started the democratic process and generated high expectations of government. In 1997, the Asian financial crisis dramatically affected the Korean economy. More recently, the sinking in 2014 of the Sewol ferry and the outbreak of Middle East Respiratory Syndrome (MERS) affected the level of trust in government institutions (see Box 1.7. for a summary).
Pre-1987 legacy: During its colonisation by Japan (1910-45), Korea’s Confucian tradition of trust and community were damaged by the Japanese government’s oppressive colonial policy. Soon after liberation in 1945 and the foundation of government in 1948, Koreans had to endure the 1950-53 Korean war. The military coup of 1961 and the military rule that followed saw a continuation of a hierarchical society, shored up with acts of violent oppression. According to some scholars, all these events have contributed to pervasive low levels of public trust in the Korean government (Kim, 2008).
1987 June Democratic Uprising and related expectations: The June Democratic Uprising was a nationwide democracy movement that generated mass protests in June 1987. The demonstrations forced the ruling government to hold elections and institute democratic reforms, which led to establishing the present-day government structure of South Korea. At the same time, democratisation has brought about challenges to trust in the government – such as very high expectations; a free media; and political opposition. While holding free and fair elections might not necessarily lead to greater public confidence in government it is a basic element of voice and accountability core elements of the openness driver of the trust framework.
1997 Asian financial crisis: Triggered in Southeast Asia, the financial crisis quickly spread to Korea, due to the bankruptcy of family-based business groups (known as the chaebol); a high level of short-term foreign debt relative to foreign exchange reserves; and the refusal of foreign creditors to roll over loans to Korean financial institutions. The ensuing IMF bailout was a shock to many Koreans. The crisis had a significant negative impact on trust in government and the chaebol, including concerns over economic policy and prudential supervision of public finances.
2011 outbreak of foot-and-mouth disease: Beginning in November 2010, the virus spread to vast regions of the country until May 2011. It led to a mass cull of over 3.5 million domestic animals. Many people pointed out the government’s failure to contain the epidemic in the early stages and blamed it for the budget loss caused by the need to compensate stockbreeders for their loss. In addition, the burial process, carried out with limited time and manpower, led to some burials leaking leachate which affected drinking water. The minister of Agriculture, Food and Rural Affairs resigned in June 2011 after the outbreak ended.
2014 sinking of Sewol ferry: In April 2014, the South Korean ship MV Sewol sank, claiming the lives of 304 passengers, mostly high school students. It transpired that the accident was a manmade disaster caused by over-loading, crating failure and, poor initial responses by the crew, coast guard, and control centre. Park (2014) notes that when government is perceived as failing to handle critical problems effectively, citizens come to lose faith in government institutions.
2015 outbreak of Middle East Respiratory Syndrome (MERS): Between May and July 2015 an outbreak of Middle East Respiratory Syndrome occurred in South Korea, affecting 186 people, and causing 36 deaths and 16 752 suspected cases. The government, in particular the Ministry of Health (MoH), was heavily criticised for not disclosing relevant information to hospitals and citizens. Initially the MoH did not disclose the names of medical institutions hospitalising the MERS patients to avoid unnecessary anxiety to the users, according to the MoH’s official statement. In turn, the MoH omitted informing some local governments (e.g. Incheon) about the transfer of some patients to its local medical institution. Medical institutions were also criticised; for example in Seoul, 35 infected patients were discharged without knowing that they were infected and allowed to move freely through the city.
Sources: Kihwan, K. (2006), “The 1997-98 Korean financial crisis: Causes, policy response, and lessons”, https://www.imf.org/external/np/seminars/eng/2006/cpem/pdf/kihwan.pdf; Dostal, J. M., and A. Ringstad (2015), “A historical-institutionalist analysis of the MV Sewol and MS Estonia Tragedies: Policy lessons from Sweden for South Korea”, The Korean Journal of Policy Studies, Vol. 30/1, pp. 35-71; Kim, Y. (2008), “Revisiting Putnam and Fukuyama: Trust and Korean society”, World Regional Studies, Vol. 26/1, pp. 5-29; Kim, K. H. et al. (2017), “Middle East Respiratory Syndrome coronavirus (MERS-CoV) outbreak in South Korea, 2015: epidemiology, characteristics and public health implications”, https://www.ncbi.nlm.nih.gov/pubmed/28153558.
The novelty and relevance of this approach stems from the possibility of providing more refined evidence that could translate into more targeted policies aimed at restoring trustworthiness. Table 1.8 presents a sample set of questions following this approach. In it, we first break down further the policy dimensions of trust into actionable policy components. For each component, we propose a sample of situational questions that aim to capture government trustworthiness, by asking citizens whether they trust their government to do X if Y happens.
Efforts to improve trust measurement should pay off in terms of actionable policy insights for Korea. The proposed approach could guide the analysis for each of the dimensions identified as drivers of trust in government institutions. Tailored policy-driven recommendations could be designed as means of attaining improvement on each of the studied dimensions.
Furthermore, the purpose of this approach is to offer a tailored assessment of trust in Korea. The focus is less on comparative data across countries, and more on context-specific understanding of institutional trust and the performance of its drivers, which is consistent with the importance of cultural and historic factors in shaping trust. The next chapters of this report will present the findings of the OECD-KDI survey in detail and recommend policy action that could help increase levels of trust in Korean public institutions.
References
Aberson, C.L., C. Shoemaker and C. Tomolillo (2004), “Implicit bias and contact: The role of interethnic friendships”, The Journal of Social Psychology, Vol. 144/3, Routledge, Abingdon, UK, pp. 335-347.
Ajzen, I. and M. Fishbein (1977), “Attitude-behavior relations: A theoretical analysis and review of empirical research”, Psychological Bulletin, Vol. 84/5, APA, Washington DC, pp. 888-918, https://doi.org/10.1037/0033-2909.84.5.888.
Algan, Y. and P. Cahuc (2013), “Trust, growth and well-being: New evidence and policy implications”, IZA Discussion Paper, No. 7464, IZA, Bonn, http://ftp.iza.org/dp7464.pdf.
Ben-Ner, A. and F. Halldorsson (2010), “Trusting and trustworthiness: What are they, how to measure them, and what affects them”, Journal of Economic Psychology, Vol. 31/1, Elsevier, Amsterdam, pp. 64-79, https://EconPapers.repec.org/RePEc:eee:joepsy:v:31:y:2010:i:1:p:64-79.
Blind, P. K. (2006), “Building trust in government in the twenty-first century: Review of literature and emerging issues”, 7th Global Forum on Reinventing Government, Vienna, 26-29 June 2007, http://unpan1.un.org/intradoc/groups/public/documents/UN/UNPAN025062.pdf.
Bouckaert, G. (2012a), “Trust and public administration”, Public Administration, Vol. 60/1, pp. 91-115. Wiley-Blackwell, Hoboken, NJ
Bouckaert, G. (2012b), “Reforming for performance and trust: Some reflections”, The NISPAcee Journal of Public Administration and Policy, NISPAcee, Bratislava, Vol. 5/1, pp. 9-20, https://doi.org/10.2478/v10110-012-0001-4.
Cho, Y. (2013), 정부신뢰의 선거적 기원: 이명박 집권 중반 시기를 중심으로 (Electoral origin of government trust during the mid-term of the Lee Myung-bak administration), Korea and International Politics, Vol. 29/3, pp. 121-153.
Christensen, T. and P. Laegreid (2005), “Trust in government: The relative importance of service satisfaction, political factors, and demography”. Public Performance and Management Review, 28(4), 487-511, Newark, NJ
Dahlstrom, C and V. Lapuente (2017), Organizing Leviathan, Politicians, Bureaucrats, and the Making of Good Government, Cambridge University Press
Dasgupta, N. and S. Asgari (2004), “Seeing is believing: Exposure to counterstereotypic women leaders and its effect on the malleability of automatic gender stereotyping”, Journal of Experimental Social Psychology, Vol. 40/5, Elsevier, Amsterdam, pp. 642-658.
Dostal, J. M., and A. Ringstad (2015), “A historical-institutionalist analysis of the MV Sewol and MS Estonia tragedies: Policy lessons from Sweden for South Korea”, The Korean Journal of Policy Studies, Vol. 30/1, ACAD, Seoul, pp. 35-71.
Falk,A. et al (2016), “The preference survey module: A validated instrument for measuring risk, time, and social preferences”, Working paper, University of Bonn.
Falk, A. et al. (2015), “The nature and predictive power of preferences: Global evidence”, IZA Discussion Paper, No. 9504, IZA, Bonn.
Fehr, E. et al. (2003), “A nation-wide laboratory: Examining trust and trustworthiness by integrating behavioral experiments into representative surveys”, CESifo Working Paper Series, No. 866, CESifo, Munich; IZA Discussion Paper, No. 715, IZA, Bonn.
Fukuyama, F. (1995), Trust: The Social Virtues and the Creation of Prosperity, Free Press Paperbacks, New York.
Gachter, S., B. Herrmann and C. Thoni (2004), “Trust, voluntary cooperation, and socioeconomic background: Survey and experimental evidence”, Journal of Economic Behavior and Organisation, Vol. 55/4, Elsevier, Amsterdam, pp. 505–531, https://doi.org/10.1016/j.jebo.2003.11.006.
Glaeser, E. et al. (2000), “Measuring trust”, The Quarterly Journal of Economics, Vol. 115/3, Oxford University Press, Oxford, pp. 811-846.
González, S. and C. Smith (2017), “The accuracy of measures of institutional trust in household surveys: Evidence from the OECD Trust Dataset”, OECD Statistics Working Papers, No. 2017/11, OECD Publishing, Paris, https://doi.org/10.1787/d839bd50-en.
Gyorffy, D. (2013), Institutional Trust and Economic Policy, Central European University Press, Budapest.
Helliwell, J. and R. Putnam (2004), “The social context of well-being”, Philosophical Transactions of the Royal Society B: Biological Sciences, Vol. 359/1449, pp. 1435-1446, https://doi.org/10.1098/rstb.2004.1522.
Helliwell, J. and S. Wang (2010), “Trust and well-being”, NBER Working Paper Series, No. 15911, National Bureau of Economic Research, Cambridge, MA, www.nber.org/papers/w15911.
Helliwell, J., Haifang Huang, Shawn Grover, and Shun Wang (2018), “Empirical linkages between good governance and national wellbeing”, Journal of Comparative Economics,
Hudson, J. (2006), “Institutional Trust and Subjective Well-Being across the EU”, Kyklos, Vol. 59, No. 1,pp. 43-62.
Jang, S (2002), “Social Capital and Public Administration Government; Which Factors Determine the Interpersonal Trust Level of a Society: From a Comparative Political Rerspective”, Journal of Governmental Studies, Vol.8 No.1
Johnson, N. and A. Mislin (2012), “How much should we trust the WVS trust question?” Economics Letters, Vol. 116, Elsevier, Amsterdam, pp. 210-212.
Johnson, N. and A. Mislin (2011), “Trust games: A meta-analysis”, Journal of Economic Psychology, Vol. 32/5, Elsevier, Amsterdam, pp. 865-889, https://doi.org/10.1016/j.joep.2011.05.007.
Kaul, I., I. Grunberg and M.A. Stern (1999), Global Public Goods: International Cooperation in the 21st Century,United Nations Development Programme (UNDP), NewYork.
Kihwan, K. (2006), “The 1997-98 Korean financial crisis: Causes, policy response, and lessons”, presented at the High-Level Seminar on Crisis Prevention in Emerging Markets, Singapore, 10-11 July 2006, https://www.imf.org/external/np/seminars/eng/2006/cpem/pdf/kihwan.pdf.
Kim, K. H. et al. (2017) Middle East Respiratory Syndrome coronavirus (MERS-CoV) outbreak in South Korea, 2015: Epidemiology, characteristics and public health implications, Journal of Hospital Infection, Elsevier, Amsterdam, Vol. 95/2, pp. 207-213, https://www.ncbi.nlm.nih.gov/pubmed/28153558.
Kim, Y. (2008), “Revisiting Putnam and Fukuyama: Trust and Korean society”, World Regional Studies, Shanghai, Vol.26/1, pp. 5-29.
Knack, S. (2000), “Social capital and the quality of government: Evidence from the United States”, American Journal of Political Science, Vol. 46/4, MPSA, Bloomington, IN, pp. 772-785, www.jstor.org/stable/3088433.
Knack, S. and P. Keefer, (1997), “Does Social Capital have an Economic Payoff? A Cross-Country Investigation”, Quarterly Journal of Economics, Vol. 112,/4, pp. 1251-88.
Kum, H. and S. Baek, (2010), “정치적 이념, 정부신뢰 그리고 정책지지” (Ideology, political trust and policy support), Korean Journal of Public Administration, Seoul , Vol. 48/4, pp. 201-228.
Lazzarini, S.G. et al. (2004), “Measuring trust: An experiment in Brazil”, Insper Working Papers, No. 049/2004, Insper, São Paulo, https://core.ac.uk/download/pdf/6228649.pdf.
Lee, Y. (2014), “정부기관 및 사회기관 신뢰에 대한 국민과 공무원의 인식 차이에 관한 비교 연구” (A comparative study of trust in governmental and social institutions: Differences between perceptions of people and public servants), Social Science Review, Vol. 45/1, IRP Journals, Dhakar, Bangladesh, pp. 25-58.
Lochner, K., I. Kawachi and R. Brennan and S. Luka (2003), “Social Capital and Neighbourhood MortalityRates in Chicago”, Social Science and Medicine, Vol. 56, pp. 1797-1805.
Murtin, F., et al. (2018), "Trust and its determinants: Evidence from the Trustlab experiment", OECD Statistics Working Papers, No. 2018/02, OECD Publishing, Paris, https://doi.org/10.1787/869ef2ec-en.
McKnight, D. H., V. Choudhury and C. Kacmar (2002), “The impact of initial consumer trust on intentions to transact with a web site: A trust-building model”, Journal of Strategic Information Systems, Vol. 11/3, Elsevier, Amsterdam, pp. 297-323, https://doi.org/10.1016/S0963-8687(02)00020-3.
Miller, A. H. and O. Listhaug (1999), “Political performance and institutional trust”, in P. Norris (ed.), Critical Citizens: Global Support for Democratic Governance, Oxford University Press, Oxford.
Morrone, A., N. Tontoranelli and G. Ranuzzi (2009), “How good is trust? Measuring trust and its role for the progress of societies”, OECD Statistics Working Papers, No. 2009/03, OECD Publishing, Paris, https://doi.org/10.1787/220633873086.
Naef, M. and J. Schupp (2009), “Measuring trust: Experiments and surveys in contrast and combination”, IZA Discussion Paper, No. 4087, IZA, Bonn, ftp://ftp.iza.org/dps/dp4087.pdf.
Nooteboom, B. (2007), “Social capital, institutions and trust”, Review of Social Economy, Vol. 65, No. 1, pp. 29-53
OECD (2017a), OECD Guidelines on Measuring Trust, OECD Publishing Paris, https://doi.org/10.1787/9789264278219-en.
OECD (2017b), How’s Life? 2017: Measuring Well-being, OECD Publishing, Paris. https://doi.org/10.1787/how_life-2017-en.
OECD (2017c), Trust and Public Policy: How Better Governance Can Help Rebuild Public Trust, OECD Publishing, Paris, https://doi.org/10.1787/9789264268920-en.
OECD (2015), Government at a Glance 2015, OECD Publishing, Paris, https://doi.org/10.1787/22214399.
OECD (2013), Government at a Glance 2013, OECD Publishing, Paris. https://doi.org/10.1787/gov_glance-2013-en.
OECD (2009), Focus on Citizens: Public Engagement for Better Policy and Service, OECD Publishing, Paris, https://doi.org/10.1787/20774036.
Park, C (2016), Political Trust in the Asia Pacific Region, in S. Zmerli and T. Van der Meer (eds), Handbook of Political Trust, Edward Elgan Publishing, Chelton, U.K
Park, H and L Young-Ran (2012), “시민문화와 정부정책이 정부신뢰에 미치는 영향.” 한국행정논집, Vol. 24, No. 4, pp. 913-936.
Park, Chong Min and Jung Hyun Bae (2011) Sources of Trust in Government: Policy Outcomes, Processes and Outputs, Journal of Government Studies, Vol. 17. No. 2 pp117-143
Park, S. (2006) “Perceptions toward government and government trust: Focusing on undergraduate students in Korea”, Korean Public Administration Review Vol. 40/2, pp. 73-97.
Putnam, R. (2000), Bowling Alone: The Collapse and Revival of America’s Community, Simon and Schuster, New York.
Roberts, J. and M. Hough (2005), Understanding public attitudes to criminal justice. Maidenhead: Open University Press.
Sachs, J. (2008), CommonWealth: Economics for a Crowded Planet, Allen Lane, London.
Sampson, R., S. Raudenbush and F. Earls (1997), “Neighborhoods and violent crime: A multilevel study of collective efficacy”, Science, Vol. 277, No. 5328, AAAS, Washington DC, pp. 978-924, http://science.sciencemag.org/content/277/5328/918.full.
Wong T. K., P. Wan and H. M. Hsiao (2011), “The bases of political trust in six Asian societies: Institutional and cultural explanations compared”, International Political Science Review, Sage Publications, Thousand Oaks, CA, https://doi.org/10.1177/0192512110378657.
Yang, K and M Holzer (2006), “The performance-trust link: Implications for performance measurement”, Public Administration Review, Vol 66, No.1, pp. 114-26.
Notes
← 1. This working definition was presented to survey respondents when the questionnaire, used to inform this research, was fielded.
← 2. The ultimate benchmark of the quality of statistics is whether they meet the needs of the user in terms of providing useful information.
← 3. Different response options lead to different and not necessarily interchangeable measures. Therefore, a standardised approach to response format to ensure the consistency of measurement, especially in an international context, is highly advised. The available evidence in terms of general studies and specific information from trust measures suggest that a numerical 11-point scale with verbal scale anchors is preferable over the alternatives, as it follows for a greater degree of variance in responses and increase the overall data quality as well as the translatability across languages.