Chapter 2. Trust in government in Korea: A puzzle

This chapter focuses particularly on the Korean case by analysing evidence stemming from several existing sources. It shows that while Korea has achieved and maintained rapid economic growth and development, and performs comparatively well in several existing measures of the quality of public administration, trust in government institutions is relatively low. Based on data specifically collected for this report this chapters shows empirically what are the main drivers of institutional trust in Korea.

What the evidence says

The previous chapter signalled that levels of trust in government institutions are low in Korea compared to OECD member countries (see Box 1.3). Still, it could be argued that trust levels at specific points of time reflect the current situation, and not necessarily the structural trend of a given indicator. Based on the Gallup World Poll, Figure 2.1 presents trends over time (2006-15) of trust in a handful of institutions compared to the OECD average. Even though year-to-year changes can be seen for the set of institutions (i.e. national government, judicial system, local police, and military), trust in government institutions in Korea seems to be at comparatively low levels.

Similarly, by using survey data from Trustlab (see Box 1.2) it is possible to establish a comparison between the six countries that have participated in that initiative. Figure 2.2 displays the average trust, on a scale of 0 to 10, for a handful of public and private institutions. For all public institutions presented in the figure (i.e. police, judicial system, government and parliament) Korea reports the lowest average values.

However, self-reported trust in government based on surveys should be assessed critically as evidence suggests that they may be biased in some circumstances, and for a number of key aspects of validity there is simply no evidence one way or another (González and Smith 2017). It can indeed be argued that in today’s political climate, expressing trust in government is likely to evoke at least mild social disapproval and actually be the socially undesirable response (Intawan and Nicholson, 2017). In addition, salient negative events involving the government in media and public discourse may lead to availability bias when making evaluations of government performance. Easton (1975) has warned of the danger of conflating cynicism vis-à-vis the government with low levels of trust. This discrepancy between self-reported and implicit trust confirms that self-reports may be affected by such factors and a closer look into deeper, more ingrained implicit sentiments towards institutions could be warranted.

While it is not completely clear what is being captured by it implicit association tests (IAT)1 have been conducted as means of generating alternative measures of institutional trust. IAT results consistently show (Figure 2.2a) that people have a higher level of implicit (i.e. experimental) trust in government than what they declare in survey questions (Murtin et al 2018). In turn, among the countries that participated in Trustlab, Korea shows the highest discrepancy between self-reported and implicit trust in government. Understanding the discrepancy between survey and implicit based measures is an area where further research is required.

Beyond the fact that self-reported trust in institutions is at comparatively low levels in Korea, the analysis of a longer discontinuous series – extracted from six waves of the World Values Survey – evidences that trust in government institutions shows interesting variations since the country’s democratic transition in the late 1980s (see Figure 2.3). With the democratic transition, public trust in the military and the court declined, at least initially, a plausible explanation is that previous cover-ups and unlawful actions under the military government were exposed. In contrast, in 1990-1995, trust in civil service increased reflecting, amongst others, efforts to make public service delivery more responsive to citizens’ needs. Over the same period, trust in the National Assembly remained stagnant at low levels, a possible explanation is the lag with which citizens oriented reforms commonly reached citizens and the political nature of this institution that could shape people’s perception. Over the next period (1995-2000), which included the 1997 economic crisis, trust in government institutions generally declined. Since Korea recovered from the crisis, trust in government institutions moderately increased over the next decade (2000-2010) but never returned to 1980s levels2.

The OECD-KDI survey has also included a version of the institutional trust questions included in most household surveys. This is a general heading similar to “How much confidence do you have in…” followed by a list of institutions. Compared to non-official comparative household surveys, one of the main improvements achieved through the OECD-KDI measurement instrument is the inclusion of a larger set of institutions, both political and administrative. For example, the Gallup World Poll only includes five institutions (government, financial institutions, judicial system and courts, police and military).

Figure 2.3 shows that not all public institutions are valued equally when it comes to trust levels. In fact, respondents make clear differences in their assessment at two levels: 1) the political versus the institutional dimensions, as noted earlier; and 2) between the network of institutions, particularly public services, government agencies and the National Assembly (Figure 2.4). Consistently with findings from other surveys (KIPA, 2015) the National Assembly (4.0) emerges as the least trusted institution (see Box 2.1). In turn, most service-related institutions (i.e. public health and school system) have the highest comparative scores. The scores of 5.08 for trust in government and 5.19 for local government are very close to the average for all institutions taken together (5.09).

Box 2.1. Trust and the National Assembly

A good functioning relationship between the executive and legislative branches of government is essential for defining and delivering solid policies and long-term structural reforms.

Some authors argue that public trust in the National Assembly is positively related to citizens’ support for the majority political party, national economic situation, and approval rating for the president. The findings of the OECD-KDI Trust Survey, however, do not identify any significant differences across respondents in their perceptions of the National Assembly. In other words, there is a generalised mistrust and, thus, deeper institutional factors must be at play.

Empirically investigating the factors behind the low levels of public trust in the National Assembly, Yi and Jeong (2013) find that the most influential factors are distrust in politics and an overall public perception that the National Assembly does not represent citizens’ interests. In addition, the basic elements of the political system such as lack of policy consistency, institutional competition between political parties and lack of internal collaboration in Korea may have a role of to play.

Low levels of interpersonal trust

While the focus of this report is on systemic or institutional trust, a general finding of the literature is that trust, as well as most of its drivers, are interlinked and often self-fulfilling. On the one hand, modern theorists suggest that interpersonal trust is influenced by increasing political and social participation. Contact with other members of the community allow people to know each other better and to improve their level of trust. In turn, this positive disposition derived from their civic experience is often extended to strangers (Fukuyama, 1995; Levi, 1997).

On the other hand, “new-institutionalist” scholars maintain that a trustworthy government generates interpersonal trust (Veenstra, 2015 Blind, 2006). For example, Rothstein and Stolle (2008) show that high trust in legal institutions has a positive impact on interpersonal trust. This means that in addressing low levels of institutional trust, Korea may want to consider approaches to strengthening interpersonal trust as well, which may also be related to the overall improvement of well-being in the country.

Interpersonal trust, used primarily as a proxy for “social capital”, is highly influenced by the set of common values shared by a given society; as such, groups of countries with mutual cultural and religious backgrounds tend to have similar levels of interpersonal trust (OECD, 2009). Interpersonal trust is an important concept, as it contains values such as truth telling, promise keeping, fairness and solidarity. Because of this, questions on interpersonal trust have been used in many cross-national surveys.3 Data from the World Values Survey show that, while distrust of other people has always exceeded trust in Korea, that gap has widened from 23% in Wave 1 of the WVS to 47% in Wave 6 (see Figure 2.5).

Interpersonal trust questions have been classified as two types: “generalised trust” and “limited trust”. Generalised trust refers to trust in people who are not known to the respondent or to trust in situations where the person being trusted is not specified. Limited trust focuses on people known to the respondent, including family, friends and neighbours (OECD, 2017a). The OECD-KDI survey includes questions both on generalised trust and limited trust, with a greater emphasis on generalised trust due to its greater relevance for most policy and analytical purposes.

The OECD-KDI survey also asked questions on interpersonal trust, showing complementary results to the WVS surveys. When asked whether or not respondents thought a lost wallet or purse containing items of great value to them would be returned to them if found by a stranger, 59.7% said “no” compared to 40.3% who said “yes”. In turn, analysing trust in the different groups – consistently with the academic literature – finds that limited trust tends to be higher than generalised trust (Figure 2.6). Evidence for six countries from the Trustlab initiative (see Figure 2.2.), comparing them with available data, shows that generalised trust in Korea is the second lowest after France.

The data presented above poses a number of questions, particularly from the point of view of public performance. Indeed, in the context of Korea, we can talk about a public performance paradox. Korea rates consistently above the OECD average in many standard measures of public governance and performance (Box 2.2). Yet its levels of trust in public institutions rank, as shown above, well below the OECD average.

Box 2.2. The performance paradox in Korea

On many measures of “good public governance”, the Korean government ranks highly compared to other OECD countries. Two key examples are for public disclosure of private interests and use of digital services (Figure 2.7). In 2014, the level of asset disclosure and public availability of information for all branches of government was higher in Korea than across the Group of Seven (G7) and OECD countries.

As another example, since the 1990s the Korean government has invested heavily in digital services and has developed state-of-the-art e-government systems and infrastructure. Digital services are a key component of the government's strategy to increase competence and foster a citizen-centred approach to service delivery. In 2014, more than 70% of all Koreans reported having used the Internet at least once over the past 12 months to interact with public officials, highest in the G7 and OECD. Moreover, Korea is the OECD leader in providing the enabling conditions for the re-use of open government data inside, and outside, the public sector.

Paradoxically, despite these and many other good practices, the results from the OECD-KDI survey show that Koreans are still highly distrustful of their government institutions. Data show that citizens rate their government quite well for basic government functions (protecting citizens and delivering public services), but lower regarding measures of integrity and openness. However, these results might also reflect that asset disclosure in itself is not sufficient to ensure that that government officials avoid capture by powerful interests. Similarly, digital infrastructure does not automatically lead to meaningful engagement with citizens in policymaking.

Note: The Group of Seven consists of Canada, France, Germany, Italy, Japan, the United Kingdom and the United States – an informal bloc of industrialised democracies that meets annually to discuss issues such as global economic governance, international security and energy policy.

Source: OECD (2016), Government at a Glance: How Korea compares, https://doi.org/10.1787/9789264259003-en.

The context of trust in Korea makes this report particularly relevant; both in terms of the challenges the country faces today in consolidating itself as a high-growth, productive and inclusive society, and of closing the gap between institutional performance and perceived levels of trust.

Why are trust levels so low in Korea? A review of the literature

The issue of low trust in Korea has been addressed by numerous scholars, particularly nationally. This section summarises some of their key findings, starting with a good summary of reasons for low trust in the Korean government provided by Jung and Sung (2012) (Box 2.3).

Box 2.3. Reasons for low trust in government: Jung and Sung, 2012

According to Jung and Sung (2012), several factors have contributed to the low public trust in Korea:

1. Political culture and democratisation. Democratisation since the 1980s has raised the public’s expectations of government to the point that they are easily dissatisfied with how the government operates and the policies it adopts. While it is easy to assume that the level of trust in the government will be much lower in authoritarian regimes, in reality, trust in government institutions can be lower as democratisation proceeds.
2. Public officials’ competence and integrity. Government’s critical policy failures have damaged the Korean people’s trust in government institutions. Korea experienced a financial crisis in the late 1990s, when the government failed to adapt to the rapidly changing global economic environment. Public officials’ malpractice and corruption also contributed to low trust levels; while Korea’s Corruption Perception Index scores from Transparency International began to exceed the world average in 2000, they are still well below the average of OECD member states.
3. Harsh criticism toward the government by the mass media. Some studies claim that the media’s corrosive criticism of the government is another factor that undermines public trust in the government. Moreover, the spread of social networking services creates an environment where this criticism of government by the media and individuals can quickly proliferate. Chapter 6 of this report deals specifically with the topic of government institutions’ communication and how it relates to trust.
4. Political institutions and competition. Ideological conflict between the parties has continued to intensify since the democratic transition of 1987. Politicians must now compete all the time for power under the current five-year single-term presidential system, and this negatively impacts trust in government institutions by causing incumbent presidents to become “lame ducks” at an ever-earlier stage in their terms.

Source: Jung, Y. -D. and S. Y. Sung (2012), “The public’s declining trust in government in Korea”, http://mjpse.meiji.jp/articles/files/01-04/01-04.pdf.

Economic performance and trust

Over the past six decades, the Korean population has benefited from remarkable improvements in standards of living. Since the 1960s, Korea achieved fast economic growth based on export-oriented industrialization and human capital development. Korea did suffer a financial crisis in 1997 and was also affected by the global financial crisis in 2008 – but the country overcame those crises within relatively short periods of time, showing high levels of institutional resilience. Post-crisis recovery boosted public trust in government (Figure 2.1 and Figure 2.3). In 2016, Korea’s per capita income reached more than 85% of the average in OECD countries, compared to slightly more than 20% in the early 1980s. Considering the steady catch up in standards of living with other OECD countries, it is again surprising that trust levels are so low; especially given that the interrelation between economic performance and trust, both institutional and interpersonal, has been found in the academic literature (Algan et al., 2017).

An analysis of the different phases of Korean development could shed light on the comparatively low levels of trust. A study on trust in the Korean government during 1980s argues that mistrust in government institutions increased as people were dissatisfied with government achievements from a distributional fairness perspective, or the extent to which all groups of the population were receiving the benefits, as well as how different groups were treated. In addition, during the same period, while the government’s policy achievements regarding economic growth and national security did not have much impact on trust in government institutions, those relevant to politics and societal issues were found to have significant impact (Park, 1991).

In the meantime, democratisation since the late 1980s has raised people’s expectations of the Korean government to a level that can easily lead to dissatisfaction if not properly met. Moreover, Korean citizens expect the public sector to be at the forefront in guaranteeing their rights and enabling them to realise their full potential in life, taking limited personal responsibility for their own well-being. The gap between people’s renewed expectations and government institutions’ capacity to meet them is one of the reasons why the media and public tend to criticise the public administration, which is expected to rapidly adapt to keep providing high-quality services.

In parallel with the Korean population’s rising expectations, other complex challenges for Korean government institutions to address include widening household income disparity and the productivity gap between industries, companies and the labour market. Based on data from the European Social Survey for 22 countries, recent evidence has found that declining levels of trust are especially evident among the most vulnerable groups of the population (Ruelens and Nicaise, 2016). As in many other OECD countries, income inequality in Korea has widened over the past two decades and relative poverty has increased, especially among older people, while the share of young people unemployed or not in employment or training is now above the OECD average (OECD, 2016). As the next chapters of this report will show, some of the groups (e.g. the youth) tend to be particularly distrustful of public institutions in Korea. Addressing the needs of especially vulnerable groups is essential to avoid increasing the trust gap in Korean society.

Additionally, Lee (2014) suggests that public satisfaction with social policies, rather than economic policies, could have an effect on trust in government in Korea. Overall, there is significant evidence that people’s trust in public institutions is affected by more than just economic performance. As the OECD Trust Framework suggests, trust is affected by institutional performance in terms of core competences and values. While it is certainly true that institutions could create the conditions for fostering economic activity this is only one aspect of institutional performance.

Understanding better institutional trust and its drivers

Beyond economic factors, the objective of this report is to analyse some of the key drivers of trust in government institutions. Based on the policy and measurement framework presented in the previous chapter, the objective is to test the policy drivers that the public management literature has identified as influencing trust in government institutions. A first step in that direction is to identify any latent factors that are captured by the data through a factor analysis (see Box 2.4).

Box 2.4. Factor analysis

Factor analysis is a methodology commonly used for data reduction purposes. In other words, several questions (variables) could be capturing similar phenomena, causing redundancies and difficulties in making a comprehensive analysis of the issue under study. Based on the results of the factor analysis, it is possible to construct a synthetic measure that reflects the structure of the data and allows a large number of variables to be combined into a few easily interpretable factors. It is a methodology commonly used to measure complex concepts, which may apply in the case of trust in public institutions.

The idea of the methodology is to retain only factors where the data signal a common underlying concept. Only factors with an “eigenvalue” higher than 1 capture a latent or underlying phenomenon (eigenvalues measure the amount of variation in the total sample accounted for by each factor). “Factor loadings” are scores assigned to each variable and can range from -1 to 1. Loadings close to -1 or 1 indicate that the variable strongly affects the factor. Loadings close to zero indicate that the variable has a weak effect on the factor. Commonly, variables with a loading above 0.5 are retained within a single factor.

The results of the OECD-KDI trust survey suggest that it is indeed possible to identify latent factors. The factor analysis conducted over the entire sample shows the dimensions of analysis loading clearly into two differentiated clusters, signalling common underlying concepts (Figure 2.8): Factor 1 on institutional trust and Factor 2 on trust in political systems and politicians. Interestingly, some of the most commonly asked questions of trust in institutions, such as trust in government and trust in local government, have no discriminatory power and could load into any of the factors. Further discussion can be found in the methodological note included as Annex B to this report.

Having such composite measures allows institutional and political trust to be regressed (e.g. estimating the relationships among variables through econometric processes) against a series of explanatory variables, based on situational questions and other relevant covariates with possible explanatory power (e.g. the capacity that a hypothesis or theory possess to effectively explain its subject matter). The institutional trust composite is regressed on the dependent variables through a stepwise regression.4

Figure 2.9 shows exactly which variables in the multivariate analysis represent the most important determinants of trust in government institutions in our cross-sectional study. The most important explanatory factor of trust in government institutions is having a recent experience with those institutions. In turn, both elements of competence (e.g. expected ability to innovate, expected stability of conditions for starting a business, expected effectiveness of disaster management plans) and values (expected availability of information on administrative procedures, expectation that actions would be taken in case of discrimination, expected prosecution of a corrupt high-level official) have an effect on trust in public institutions.

Turning to societal determinants of trust in government institutions, feeling part of a community is associated with higher levels of trust in government institutions. When it comes to economic determinants it is found that the regional economic growth rate has a positive, albeit relatively small, effect on trust.

As discussed previously, institutional and political trust are not fully independent concepts. Despite the fact that no measurement framework has been specifically designed in this case study for the measurement of political trust and therefore no econometrical models have been specified for this purpose, testing whether or not some of the drivers of institutional trust have explanatory power over political trust could deepen the understanding of how trust in Korea is constructed. Information on political trust could shed light on specific and common determinants of both dimensions (i.e. institutional and political) and the relative size of their effect. In turn, this enhanced understanding can help tailor specific policy recommendations to restore trust in public institutions in Korea.

Based on the political trust factor resulting from the factor analysis described in Figure 2.8, it is possible to conduct regression analysis using a unified measure of political trust as a dependent variable and a similar specification of the model (e.g. same explanatory variables) as for institutional trust (see Annex B for details). Figure 2.10 presents the expected changes on both institutional and political trust following a one standard deviation increase in each of the factors that turned out to be statistically significant. As can be seen in Figure 2.10, differences in components of competence and values positively affect levels of institutional and political trust, although the drivers (or variables) with explanatory power and – when common across the two equations – the relative size of the effect vary across the two factors. While increases in most explanatory variables have positive effects on both institutional and political trust, changes in political orientation – defined as the level of conservativeness – negatively influence political trust. As the level of conservativeness increases, it is expected that trust in political institutions will decrease.

Figure 2.11 shows elements of competence and values, as well as other explanatory variables that have a statistically significant effect on institutional and political trust, including the relative size of the effect when a determinant influences both (see Annex B for the full output of the regressions). For example, while the expected prosecution of a corrupt high-level official influences both institutional and political trust, it has a higher relative effect on political trust in our cross-sectional study. In turn, the expected ability of civil servants to innovate has a higher relative effect on institutional than on political trust. Some variables influence one type of trust exclusively, such as consultation processes on contentious issues, which are associated with political rather than with institutional trust. In turn, variables on the effectiveness of disaster management plans, availability of information on administrative procedures, responses to citizens’ complaints and fair treatment in the provision of services influence institutional trust exclusively.

After acknowledging that overcoming the trust crisis has become a policy priority, the difficulty of identifying credible measures of trust that could lead to actionable policy recommendations has been a practical challenge. Based on the results presented above and additional complementary evidence from the OECD-KDI survey, the next two chapters of this report will focus specifically on analysing the drivers of trust in institutions (i.e. competence and values). In light of those results, the report will propose policy recommendations informed by international experiences and aimed at influencing the drivers – and ultimately, institutional trust in Korea.

References

Algan, Y. et al. (2017), “The European trust crisis and the rise of populism”, Centre for Economic Policy Research, London, www.cepr.org/active/publications/discussion_papers/dp.php?dpno=12444.

Blind, P. (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.

Ruelens, A and I. Nicaise (2016), “Economic crisis, social disinvestment and trust in political institutions” , HIVA, Leuven.

Fukuyama, F. (1995), Trust: The Social Virtues and the Creation of Prosperity, Free Press Paperbacks, New York.

Helliwell, J., H Huang and S Wang (2016), “The Distribution of World Happiness”, World Happiness Report, Update (Vol I), Heliwell, J., R. Layard and J. Sachs (eds), Sustainable Development Solutions Network, New York

Park, C. (1991). The Impact of Governmental Performance on Trust in Government. Korean Public ,정부학연구, Vol. 17, No. 2, pp. 117-142.

Jung, Y.-D. and Y. Sea (2012), “The public’s declining trust in government in Korea”, Meiji Journal of Political Science and Economics, Meji University, Tokyo, http://mjpse.meiji.jp/articles/files/01-04/01-04.pdf.

KIPA (2015), Korea Social Integration Survey 2015, Korea Institute of Public Administration, Seoul.

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, pp. 25-58.

Levi, M. (1997), Consent, Dissent and Patriotism, Cambridge University Press, New York.

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.

OECD (2017a), OECD Guidelines on Measuring Trust, OECD Publishing Paris, https://doi.org/10.1787/9789264278219-en

OECD (2016), Government at a Glance: How Korea compares, OECD Publishing, Paris, https://doi.org/10.1787/9789264259003-en.

OECD (2009), Focus on Citizens: Public Engagement for Better Policy and Service, OECD Publishing, Paris, https://doi.org/10.1787/20774036.

Rothstein, B and D. Stolle (2008), “The State and Social Capital: An Institutional Theory of Generalized Trust”, Comparative Politics, Vol. 40, No 4, pp 441-459

Yi, K-S and J, Han-Wool (2013) “Analysis on Factors Affecting Public Trust in the National Assembly: Based on Public Attitude toward the National Assembly. (국회신뢰의 영향요인 분석: 국민인식조사를 중심으로).” Korean Public Administration Quarterly (한국행정논집), Vol. 25, No. 2, pp. 359-380. Veenstra, G. ( 2015) “Explicating Social Capital: Trust and Participation in the Civil Space” Canadian Journal of Sociology, 27(4), pp 547-572.

Notes

← 1. IAT is 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.

← 2. In 1980 Korea was under martial law. Several pro-democracy demonstrations were severely repressed including through the use of airborne troops in May 1980 resulting in several demonstrators killed. The political context in 1980 could have influenced people willingness to report freely on government through surveys.

← 3. The World Values Survey was chosen for analysis as it has asked the interpersonal trust questions in all six waves and, most recently, to a larger proportion of OECD countries than the other cross-national surveys. Gallup uses a slightly different approach to the others. Rather than asking directly about interpersonal trust, Gallup asks the “wallet question” (would a lost wallet be returned by a stranger; see Table 1.7), permitting both an analysis of the different directions of trust (i.e. relative rankings of trust between police and neighbours), but also direct comparisons between expected trust and actual trustworthiness (Helliwell et al., 2016). However, the wallet question has only been included in one Gallup Poll to date.

← 4. In statistics, stepwise regression is a method of fitting regression models in which the choice of predictive variables is carried out by an automatic procedure. In each step, a variable is considered for addition to or subtraction from the set of explanatory variables. Only variables adding to the explanatory power of the model are kept.

End of the section – Back to iLibrary publication page