Does process matter more for predicting trust in government? Participation,performance,and process,in local government in Japan

Abstract

Trust in government and its antecedents and development remain leading policy and research concerns. Drawing on a broadly representative online survey of 3100 respondents in Japan, we examine measures of trust in three local government actors. We find political participation is not associated with trust in local government, contrary to our expectations. Civic participation is associated with trust in the mayor, but not councillors or administrators. Satisfaction with services provided by local government, and positive perceptions of policy process, are associated with trust, with interactions suggesting process is the stronger antecedent. To develop greater trust in local government, it is important that public sector actors exhibit respect for rights and follow procedure, laws, and regulations, as well as deliver positive outcomes.

Points for practitioners

Building trust in government remains a key concern for policy makers, as it is related to successful adoption of policies.

Trust in local government in Japan is related to perceived performance and citizen satisfaction.

Civic participation is also related to some forms of trust in government.

However, possibly more important are citizen perceptions that policy processes respect rights, procedures, and laws.

To increase trust in government, practitioners need to practice, and show that they practice, good processes in developing and delivering policy, and show that policy leads to better perceived outcomes.

Keywords

Due process citizen participation civic participation policy process political participation social capital local government performance trust

What factors predict trust in local government in Japan? Trust in government influences effective delivery and adoption of policies, instruments, directives, ‘moral suasion’, tax collection, and income redistribution; and its antecedents and development remain leading policy and research concerns (Goldfinch et al., 2021; Mahmood et al., 2020). Local government, owing to its importance in service delivery to citizens, is central to this debate (Manning and Guerrero, 2013). Moreover, there is a growing interest in, and literature on, Japan, Asia, and trust (Imada, 2010; Kim, 2010; Noda, 2019; Pekkanen, 2006), albeit with claims the Asia-Pacific and Japan have been neglected in public sector studies (Suzuki et al., 2020; Van der Wal and Demircioglu, 2020). Surveying 3100 Japanese citizens, we focus on three possible antecedents of trust in local government: civic and political participation (social capital), performance of local government measured by satisfaction, and perceptions of policy processes. We test these for three local government actors: the mayor, councillors, and administrators. We add novelty by comparing these possible antecedents of trust, and making conclusions on their relative verisimilitude. We are not aware of research that tests all three antecedents in one study.

By focussing on these three antecedents in this single article, we are also engaging with the broader literature on public administration reform. In the context of public administration reform models or ‘paradigms’: Is trust related to the way local government performs its functions, by exhibiting attention to law, procedure, and rights as ‘traditional public administration’ is claimed to promote (the process of our title)? Is it about what is achieved either through results, outcomes, or outputs of New Public Management (NPM) doctrine, perhaps measured by citizen satisfaction (performance)? Or is it whether citizens participate in local government, in accord with precepts of ‘new governance’, ‘networked governance’, or ‘Post-NPM’ where co-production and civic and political participation are key (participation) (Goldfinch and Yamamoto, 2019)?

Japan is an important field within which to study trust in government. As the third largest economy of strategic importance, it is worth study in its own right. Moreover, this article addresses calls for a greater focus on the Asia-Pacific (Van der Wal and Demircioglu, 2020). Japan, owing to its differing characteristics, is interesting in the context of the trust and participation literature. First, trust in central government is low at 17% in 2012, compared with the OECD average of 40%, and still low at 38% in 2018 compared with the OECD average of 45% (OECD, 2019). Second, there are low rates of political participation, with, for example, only a quarter of individuals in their 20s voting in the Tokyo Metropolitan elections, compared with around half of 40–50-year-olds, and two-thirds of over-60s (MIC, 2017a). Voter turnout in the 2015 elections for governors averaged 47% (MIC, 2017b). Such low levels of trust and/or participation may be of concern if they are related to the ability of a government to successfully introduce policy changes, with a consensus that Japan's sluggish economic growth of the last three decades, high debt ratio, and an aging and shrinking population, means significant changes in policy direction are needed. In particular, the Japanese National government has looked to encourage participation in social life for individuals across generations, hopefully to assist in this policy adjustment (CAO, 2017; Suzuki, 2017). In sum, building trust in government is of direct policy relevance for Japan, with generalizable interest elsewhere.

Local government also provides a useful field to investigate trust, with its activities ‘easily recognized by citizens … [who] tend to have [an] accurate perception’ (Manning and Guerrero, 2013: 94–95). Some may trust a government to which they are physically close, but distrust more distant central government (Goldfinch et al., 2009). Moreover, local government has a central role in Japan, providing around two-thirds of public services, including schooling, waste collection, fire protection, health and welfare, urban planning, and infrastructure. It accounts for around 60% of governmental expenditures and 25% of GDP. Hence, how it operates is important to the whole of Japan. Local government is also relevant to understanding public administration broadly, with it a key driver of reform elsewhere (Boyne, 1998); albeit with the relationship between public sector reform at the local level and trust found to be somewhat ambiguous (Reitan et al., 2015). Japan is important to this debate, with local governments adopting NPM-type instruments including: performance measurement, accrual based financial reporting, contracting-out, agencification, and post-NPM attributes such as co-production (Aoki, 2015; Goldfinch and Yamamoto, 2019).

We test four hypotheses looking at the antecedents of trust in local government, examining civic and political participation, policy process, and outcomes/performance measured by satisfaction. We test interactions between these, influenced by studies suggesting some interdependence (Goldfinch and Yamamoto, 2019). Trust in the directly-elected mayor, directly-elected local councillors, and local administrators/bureaucrats are all separately tested. We find trust in government is related to positive perceptions of policy process, and to service performance satisfaction. However, contrary to our expectations and other research (Noda, 2019), we find political participation is not associated with our measures of trust in government. Civic participation is related to trust in the mayor only. Despite a large body of research suggesting otherwise, some socio-economic and demographic factors are not related to trust. The exceptions being that females are less trusting of councillors, while older respondents also trust councillors less. Those identifying as politically conservative trust councillors more. Trust in local government, rates of participation, and positive perceptions of local government processes, are comparatively low in Japan, suggesting some cause for concern.

Encouraging political and civic participation has normative appeal and may contribute to better policy and other outcomes. It has been a recent policy focus of the Japanese Government, and a supposed link with trust finds support in other studies of Japan (Noda, 2017). But it is not a panacea for building trust in local government (Uddin et al., 2019). The performance of local government remains key, with satisfaction an important antecedent of trust. However, we reject the view that citizens care only about what is achieved, rather than how it is achieved. Indeed, perhaps our most interesting finding is that policy processes perceived to protect human rights and exhibit accordance with laws and regulations may be more important in building citizens’ trust in local government than either citizen satisfaction or civic and political participation. Positive perceptions of process may be key in developing trust in government and bringing citizens along for needed policy change.

Participation and trust

We measure both civic and political participation. Civic participation is publicly-spirited collective action with concerns above and beyond public policy. Political participation is more directly related to engagement with the machinery of government and its actors, the election and selection of government actors, voting, and the influencing of public policy (Theiss-Morse and Hibbing, 2005). We are inspired by social capital literature – including that examining collaboration between government and community in Japanese society (Fukuyama, 2001; Pekkanen, 2006) – and draw measures from Putnam's classic studies (Putnam, 2000; Putnam et al., 1993).

The relationship between civic participation, political participation, and trust in government is contested. A positive view of the state may encourage political and other participation as individuals expect their activities to lead to positive outcomes (Noda, 2019; Wilkinson et al., 2019). Perhaps those who trust government see less reason to participate – they are happy to leave their representatives to deliver what is ‘best’ (Kamin, 2010). A competing body of research finds trust in government has little relationship with political participation (Reitan et al., 2015). Alternatively, Goldfinch et al. (2009) found less trust in government was associated with greater political participation in Australia and New Zealand. Perhaps the relationship is reciprocal, where trust increases participation, and where participation increases trust (Kettl, 2017).

Other studies suggest a lack of trust in government might encourage civic engagement and civic participation, but also a disengagement from electoral and other direct forms of political participation (Magni, 2015). Alternatively, civic participation may not be related to political participation or trust in government, and perhaps serves other needs (Imada, 2010). Indeed, in Japan civic engagement may act as an alternative to formal government processes and service provision (Pekkanen, 2006; Suzuki, 2017).

This leads us to the following hypotheses:

H1: Trust in local government is stronger with higher levels of political participation in local government.

H2: Trust in local government is not associated with higher levels of civic participation.

Policy processes and trust

A body of research suggests higher trust in government is related to positive perceptions of policy processes. This can be characterized as ‘following procedure’, perceived fairness, impartiality, integrity, ethical conduct by government actors, ‘due process’, ‘procedural justice’, and perceptions of less corruption, or combinations of such (Dahlberg et al., 2015; Grimes, 2017; Grimmelikhuijsen and Knies, 2017; Suzuki and Demircioglu, 2021; Van de Walle and Migchelbrink, 2022). For example, Tyler (2006: 7) found policy processes characterized by ‘neutrality, lack of bias, honesty, efforts to be fair, politeness, and respect for citizen's rights’ to be important for legitimacy. Indeed, process may matter ‘more than outcomes of government activity’ (Van Ryzin, 2011: 758–759). Drawing from this disparate literature, we focus on respondent perceptions of policy processes as exhibiting respect for rights, and following procedures, laws, and regulations (Christensen and Lægreid, 2005). From this we derive Hypothesis 3:

H3: Positive perceptions of policy processes in local government are associated with higher levels of trust in local government.

Service performance, satisfaction, and trust

A literature suggests policy outcomes and quality of government services – and satisfaction with these – are central to explaining trust in government (Kim, 2010; Lanin and Hermanto, 2019; Oyama, 2015). Citizens’ perceptions or evaluations of political and policy processes, or political and other participation, are argued to be of lesser or mixed importance (Dahlberg et al., 2015). As Popkin (1991: 99) claims, voters ‘generally care about ends not means; they judge government by results’. Manning and Guerrero (2013) found both direct and indirect measures of trust in government were correlated with government policy outcomes in Latin America. Downe et al. (2013) found satisfaction with services was related to trust in UK local governments. Greasley’s (2013) study of local government in the UK also found a link between performance as measured by inspections and centrally determined performance indicators, and trust of citizens as established by survey.

This leads us to our fourth hypothesis:

H4: Higher levels of satisfaction with local government is associated with stronger trust in local government.

Research design

A commercial company was contracted to deliver an internet-based survey of Japanese citizens. The company invited its 20,816 registered social studies panel members to complete an online survey. Sampling requirements were derived from Japanese government quota methods, on a gender, age, and residential basis, with the required number for the 20 subsamples set at 155 (Cabinet Office, 2020). Once this number was satisfied (i.e. 20 × 155 = 3100), the survey finished (i.e. the sample size was 3100). We targeted all municipalities, of which there are 1751 in Japan, with around half represented in the sample. We are not comparing municipalities with each other (although we control for urban/rural), so a broad-based citizen survey is appropriate to answer our research question. As Suzuki et al. (2020: 365; emphasis added) notes, ‘Japanese municipalities have relatively homogenous administrative structures, culture, ethnicity, and economic levels, allowing us to control for non-cross sectional variating factors.’ The data we generated were broadly representative in terms of gender, age, and residence. Income is broadly representative, at￥5.253m compared with the national average of ￥5.419m. Respondents are slightly better educated than the general population (Table 1).

Table 1.

Quota sample versus Japanese population (%).

	Sample⁵	Population
Age ¹
20–29	20	14.2
30–39	20	17.8
40–49	20	20.9
50–59	20	17.6
60+	20	29.3
Gender
Female	50	51.4
Male	50	48.6
Residence ²
Urban	50	43
Rural	50	57
Income ³
<\2 m	13.9	19.8
2–3	12.2	13.7
3–4	17.5	13.2
4–6	24.7	19.2
6–8	14.7	14
8–10	8.4	8.6
10–12	4.1	4.7
12–15	2.2	3.7
16+	2.2	3.3
Education ⁴
Junior high school	2.4	3
High school	31	38.3
Vocational/Junior college	23.6	26
University	39	28.9
Postgraduate	4	3.4

Source: Population survey by MIC. Owing to low internet use, most 60+ are 60–74; ² Tokyo, Kanagawa, Chiba and Saitama Prefectures and ordinance-designated major cities; ³Source: Basic Survey for National Life. Average income of sample is Y5.253 m; Average is Y5.454 m in wider population. At 08/12/21 Y1000 = $US8.81; ⁴Source: National Survey of Social Stratification and Social Mobility; ⁵Sample size = 3100.

We expect satisfaction and trust to interact with participation and perceptions of policy process (Mizrahi et al., 2010). Civic participation may influence a stronger association between satisfaction and trust. In addition, the greater the positive perceptions of process, the weaker the expected association between performance and trust. These hypothesized relationships are presented in Figure 1.

Figure 1.

Hypothetical framework.

Multiple regression analysis was used to test our hypotheses, with and without interaction effects.¹

Dependent variables

Trust in government

There is sometimes a relationship found between trust in government in general, and trust in particular political actors and organizations, particularly in strong presidential (or mayoral) systems (Manning and Guerrero, 2013). However, trust in different government actors, including public servants, can be conceived of, and measured, separately (Houston et al., 2016). In this case, Japan has a two-tier local government system, classified into 47 prefectures and 1719 municipalities; the latter is the focus of our study. Municipalities consist of cities, towns, and villages, with direct election of assemblies, and direct election of local government heads (Mayors). Mayors exercise executive power. Accordingly, our dependent variables² are:

Trust in the directly-elected mayor as mayoral trust.

Trust in councillors, who constitute the directly-elected local (municipal) assembly, as councillor trust.

Trust in local government administrators (i.e. municipal government bureaucrats), as administrative trust.

Each is measured by asking ‘to what extent do you trust your local government and leaders?’, on a five-point Likert scale from ‘high levels of trust’ = 5, to ‘do not trust at all’ = 1.

Independent variables

Participation, policy process, and satisfaction

Civic participation is measured by responses to the following questions:

Are you ‘involved in local organizations, clubs and volunteer activities?’

Are you ‘a member of the neighbourhood association?’

Answers are coded ‘Yes’ (=1) or ‘No’ (=0), and summed. These questions are derived from Putnam (2000) and Putnam et al. (1993).

Political participation is measured by responses to the following questions:

Have you ‘discussed local issues with politicians or bureaucrats?’

Have you ‘voted in the most recent local government elections?’

Have you ‘actively searched for information on the local assembly or the local government?’

Answers are coded ‘Yes’ (=1) or ‘No’ (=0), and summed. Measures are adapted from Putnam (2000), Putnam et al. (1993), and Goldfinch et al. (2009).

Policy processes are measured by asking whether respondents believe local government:

‘places emphasis on procedures and people's rights’.

‘does its best to comply with laws and regulations’ in administration.

Likert scales ranging from ‘strongly agree’ = 5 to ‘strongly disagree’ = 1 are used, and the two scales are combined. These measures are derived from Christensen and Lægreid (2005) and Park and Joaquin (2012).

Respondent satisfaction is measured by answers to: ‘To what extent are you satisfied with the general policy/services provided by your local government/council?’, adapted from Christensen and Lægreid (2005), and using a five-point Likert scale ranging from ‘very satisfied’ = 5 to ‘very dissatisfied’ = 1.

Control variables

Ideology

Ideology can influence political perceptions and behaviour. Politically conservative individuals can be happier with, and supportive of, existing power structures and outcomes (Napier and Jost, 2008). Marginalized groups may also be supportive as they internalize the social order (Cichocka and Jost, 2014). In this case, Japanese political conservatives are more supportive of the status quo and Liberal Democratic Party politics and policies (Ikeda, 2016).

Ideology is tested by respondents’ self-rating on the following question, derived from the World Values Survey (2016): ‘When we define our political standpoint, we often use the terminology left (liberal) or right (conservative). On a scale of 1 (more left) to 10 (more right), how do you define your political ideology?’

Socio-economic and demographic variables

Socio-economic and demographic variables can influence participation, satisfaction and process, and trust, albeit in contested ways (Goldfinch et al., 2021). We measure these as:

Gender coded 1 for male; 0 for female.

Age ranked from ‘under 20’ (=1) to ‘60 and over’ (=6).

Education ranked from ‘junior high school’ (=1) to ‘postgraduate’ (=5).

Income per year ranked from ‘less than ￥1350’ (=1) to ‘more than ￥15,000’ (=10) (as at 08/12/21, ￥1000 = $US8.81).

Place of residence is divided into urban (=1) and others or rural (=0); and length of residence ranked from ‘less than 1 year’ (=1) to ‘more than 20 years’ (=5).

Results

Trust in Japanese local government is strikingly low, with ‘high trust’ or ‘trust’ in mayors, councillors, and administrators, at 20.3%, 10.5%, and 15.9%, respectively. Civic and political participation is also low, at 36.7% and 23.7%, respectively. The number of respondents satisfied by public services is also low at 26%, while only 18% positively assessed policy processes (Appendix 1).

Moving to the hypotheses, and starting with mayoral trust (Table 2), we find political participation is not associated with mayoral trust. As such, Hypothesis 1 is not supported. Notably, civic participation is significantly associated with mayoral trust both directly, and in interactions. Hence, Hypothesis 2 is partly falsified. Both process and satisfaction are significantly and positively associated with mayoral trust, providing support for Hypotheses 3 and 4. We found no significant relationship between ideology and socio-demographics, and mayoral trust. The exception is length of residence in the local community, with longer residence associated with lower mayoral trust. Neither participation nor process show significant associations with the linkage between satisfaction and mayoral trust in interactions. As such, civic participation, process, and performance directly affect mayoral trust.

Table 2.

Regression results for mayoral trust.

Independent variable	Model 1 without interaction		Model 2 with interaction
		VIF		VIF
Participation
Civic	0.059(0.022)**	1.364	0.059(0.022)**	1.367
Political	0.004(0.017)	1.203	0.007(0.017)	1.221
Satisfaction	0.287(0.019)***	1.260	0.282(0.02)***	1.306
Process	0.274(0.01)***	1.284	0.272(0.01)***	1.300
Ideology	0.006(0.006)	1.019	0.006(0.006)	1.019
Socio-demographics
Gender	0.006(0.024)	1.064	0.007(0.024)	1.065
Age	0.002(0.01)	1.364	0.001(0.01)	1.369
Education	−0.019(0.013)	1.143	−0.018(0.013)	1.144
Income	0.001(0.006)	1.103	0.001(0.006)	1.105
Residence period	−0.019(0.01)+	1.238	−0.019(0.01)+	1.241
Living place	−0.027(0.024)	1.031	−0.028(0.024)	1.032
Participation × Satisfaction
Civic × Satisfaction			0.014(0.012)	1.158
Political × Satisfaction			0.004(0.011)	1.184
Process × Satisfaction			−0.013(0.009)	1.072
Const.	0.470(0.089)***		0.497(0.09)***
N	3093		3093
F	158.340***		124.704***
Adjusted R²	0.359		0.359

Standard errors in parentheses. The difference between the number of respondents and data is caused by ineffective responses *p < 0.05, **p < 0.01, ***p < 0.001, +p < 0.10.

Councillor trust is not significantly associated with civic and political participation (Table 3). However, as in mayoral trust, there are significant positive relationships between trust, policy process, and satisfaction. Males are less trusting than females, both with and without interactions. Residence period is also negatively associated with trust. Older people exhibit greater distrust in local councillors, perhaps partly because of longer residence. However, perceptions of process negatively affect the linkage between satisfaction and political trust. Other measures are not significant.

Table 3.

Regression results for councillor trust.

Independent variable	Model 3 without interaction		Model 4 with interaction
		VIF		VIF
Participation
Civic	0.021(0.021)	1.364	0.022(0.021)	1.367
Political	−0.009(0.016)	1.203	−0.006(0.017)	1.221
Satisfaction	0.25(0.019)***	1.260	0.247(0.019)***	1.306
Process	0.256(0.01)***	1.284	0.254(0.01)***	1.300
Ideology	0.006(0.006)	1.019	0.006(0.006)	1.019
Socio-demographics
Gender	−0.069(0.024)**	1.064	−0.067(0.024)**	1.065
Age	−0.016(0.009)+	1.364	−0.017(0.009)+	1.369
Education	−0.006(0.013)	1.143	−0.005(0.013)	1.144
Income	−0.009(0.006)	1.103	−0.009(0.006)	1.105
Residence period	−0.017(0.01)+	1.238	−0.016(0.01)+	1.241
Living place	−0.001(0.023)	1.031	−0.002(0.023)	1.032
Participation × Satisfaction
Civic × Satisfaction			0.016(0.012)	1.158
Politic × Satisfaction			−0.008(0.011)	1.184
Process × Satisfaction			−0.018(0.009)*	1.072
Const.	0.593(0.088)***		0.614(0.089)***
N	3093		3093
F	133.516***		105.500***
Adjusted R²	0.320		0.321

*p < 0.05; **p < 0.01; ***p < 0.001.

Moving to administrative trust, we find political and civic participation are not associated with trust. Significant positive relationships between trust, process, and satisfaction are discovered, however (Table 4). Notably, ideology is associated with administrative trust – with and without interactions with satisfaction – with conservative respondents having higher levels of trust in local government administrators. Socio-demographic variables, including residence period, are not related to administrative trust. In contrast to mayoral and councillor trust, interactions between satisfaction and political participation show a positive and significant impact on administrative trust; that is, political participation strengthens the performance-trust link. In addition, poor perceptions of process negatively affect the relationship between satisfaction and administrative trust.

Table 4.

Regression results for administrative trust.

Independent variable	Model 5 without interaction		Model 6 with interaction
		VIF		VIF
Participation
Civic	0.019(0.019)	1.364	0.019(0.019)	1.367
Political	0.019(0.015)	1.203	0.024(0.015)	1.221
Satisfaction	0.29(0.017)***	1.260	0.28(0.018)***	1.306
Process	0.276(0.009)***	1.284	0.273(0.009)***	1.300
Ideology	0.016(0.006)**	1.019	0.016(0.006)**	1.019
Socio-demographics
Gender	−0.003(0.022)	1.064	−0.002(0.022)	1.065
Age	0.009(0.009)	1.364	0.009(0.009)	1.369
Education	0.014(0.012)	1.143	0.014(0.012)	1.144
Income	0.000(0.005)	1.103	−0.001(0.005)	1.105
Residence period	−0.012(0.009)	1.238	−0.012(0.009)	1.241
Living place	−0.019(0.021)	1.031	−0.018(0.021)	1.032
Participation × Satisfaction
Civic × Satisfaction			−0.001(0.011)	1.158
Politic × Satisfaction			0.025(0.01)*	1.184
Process × Satisfaction			−0.014(0.008)+	1.072
Const.	0.220(0.080)**		0.266(0.081)**
N	3093		3093
F	198.770***		157.137***
Adjusted R²	0.413		0.414

*p < 0.05; **p < 0.01; ***p < 0.001.

To sum, political participation is not directly associated with our three measures of trust in government. Hypothesis 1 is not supported. Civic participation is significantly and positively associated with mayoral trust. It is not statistically significant in councillor or administrative trust. Consequently Hypothesis 2 is only partly supported. Process and satisfaction are significantly associated with all types of trust, whether directly or through interactions, providing support for Hypotheses 3 and 4. Notably, standardized coefficients for process range from 0.430–0.476; in contrast to those of satisfaction which range from 0.216–0.257. As such, process might be a stronger antecedent of trust than satisfaction.

Interactions can perhaps be most readily demonstrated graphically. The marginal effects of civic participation, political participation, and process, at varying levels of satisfaction, are shown in Figures 2–4. The impacts of civic participation and process are greater at lower levels of satisfaction in administrative and councillor trust, respectively. The impact of political participation is smaller, however, at lower levels of satisfaction, when testing for councillor trust.

Figure 2.

Effect of civic participation at varying levels of satisfaction.

Figure 3.

Effect of political participation at varying levels of satisfaction.

Figure 4.

Effect of process at varying levels of satisfaction.

Moving to our control variables in more detail, we find some are not significant, or significant in unexpected directions. Conservative ideology is positively and significantly associated with administrative trust, but is not associated with mayoral and councillor trust. Income and education have no significant relationship with trust. Older residents exhibit less councillor trust, in contrast to studies suggesting age is positively related to trust (Christensen and Lӕgreid, 2005). Gender is significantly related to councillor trust, with Japanese females exhibiting lower councillor trust than males, which is an unexpected finding (contra Christensen and Lægreid, 2005). Those with longer periods of residence exhibit lower mayoral and councillor trust. We suspect this is partly caused by a reduced ability amongst Japanese respondents to ‘exit’ owing to greater moving costs (see Tiebout, 1956). Elsewhere, citizens with lower trust in local government may have moved to other jurisdictions removing them from the sample. Residence period is correlated to age (0.404, p < 0.01) and civic participation (0.242, p < 0.01) (see Appendix 2). Just over 45% of respondents have a residency period of over 20 years. Place of residence (urban or rural) is not related to any type of trust, but is significantly corelated to income (0.47, p < 0.01) and education (0.80, p < 0.01) (see Appendix 2).

Discussion and conclusion

Local government in Japan provides an important field in which to examine the antecedents of trust in government, with our study showing service delivery and process matter for trust in local government. A body of research finds higher levels of trust in government facilitates policy engagement, compliance, and change, while promoting greater levels of social cohesion and reducing political polarization (Newton et al., 2018). However, Japan faces both seemingly intractable policy problems, and a relatively (by OECD standards) politically disengaged and distrustful population. If increasing trust in government does indeed facilitate policy change and other positive results, then understanding antecedents of trust in Japan has direct and important policy relevance; as well as informing debates elsewhere. Moreover, focusing on trust at local government level is of policy relevance for Japan, with local government providing around two-thirds of public services, and accounting for around 60% of governmental expenditures, and 25% of GDP.

To further this understanding, we draw on a considerable body of research suggesting three key avenues of inquiry: first, political and civic participation (sometimes termed social capital) as potentially being associated with trust in government. Second, perceptions of policy process as an antecedent of trust. Third, perceptions of local government performance measured by satisfaction, as contributing to greater trust. We test four hypotheses investigating these, drawing on a broadly representative online survey of 3100 citizens across Japan sampled by quota, and separately measuring trust in three different local government actors. Interaction effects of satisfaction, participation, and process are tested; as are socio-economic, demographic, and ideological factors.

Results are not always as expected, and our study generates some novel and important findings. Trust in local government is strikingly low, with only 20% trusting the mayor, 10.5% trusting councillors, and 16% trusting local government administrators. Political participation is not associated with greater trust in local government, contrary to Hypothesis 1 and a considerable, albeit somewhat contradictory, body of research including other studies of Japan (Noda, 2019). However, interactions show political participation strengthens the relationship between satisfaction and trust in administrators. Civic participation (Hypothesis 2) is not significantly associated with trust as expected; with an exception being a positive association found between civic participation and trust in the mayor; the latter contrary to our expectations. As such, there are some grounds at least to suggest some civic participation may be important in building trust.

Significant relationships exist between perceptions of policy process and trust in government, as predicted (Hypothesis 3). Greater satisfaction with public services is associated with greater trust in government; again, as predicted (Hypothesis 4). That both process and performance outcomes are associated with trust is supported by other studies (Van de Walle and Migchelbrink, 2022). Our interactions do show a direct relationship between process and trust. Moreover, the balance of evidence suggests perceived process is a stronger antecedent of trust in government than perceived performance/satisfaction. However, our findings do suggest if NPM-type reforms improve service performance – one of their stated aims – they can enhance trust (Van de Walle, 2011). Our interactions show interdependence amongst some aspects of participation, process and satisfaction, as previous studies suggest (Kettl, 2017).

Socio-economic and demographic factors had mixed significance for our trust measures, and again our study generates some interesting and somewhat novel findings. Income and education were not associated with any type of trust (contra Goldfinch et al., 2021). Gender did have a relationship – but a somewhat unexpected one – with females exhibiting lower trust in councillors. This might reflect Japan's high levels of female inequality, with the World Economic Forum (2018) ranking it 110 out of 149 countries on equality measures. Longer residence was found to be associated with lower levels of trust in mayors and councillors; place of residence (i.e. rural or urban), however, was not associated with trust. Older respondents also exhibited lower levels of trust in councillors; again, an unexpected finding, but perhaps a longer experience with local government encourages a degree of scepticism towards it (Yamagishi and Yamagishi, 1994). Conservative ideology was associated with higher levels of trust in councillors, but statistically insignificant otherwise.

Our findings sit somewhat in tension with a body of work suggesting high levels of social capital in Japan, which is often assumed to be associated with higher trust in government. The importance of social capital in Japan for recovery from large earthquakes, and where citizen engagement and collaboration with state actors facilitates provision of social care, especially care of the elderly, has been noted by Pekkanen (2006). Further study is needed to examine this relationship, but we suggest that social capital – and particularly the measure of civic engagement we use – may capture an alternative to state provision, and only be loosely related to trust in government. For example, civil participation in the neighbourhood organizations known as chonaikai (Imada, 2010), and more recent evidence of co-production (Dollery et al., 2020), may act as alternatives to government service provision. Indeed, Suzuki (2017) found cuts in government spending led to increases in voluntary activities of citizens in Japanese municipalities.

Civil participation perhaps also serves other demands, including sociability. However, our study does not preclude the possibility that building greater trust in government (or at least the mayor) could lead to greater civic engagement, and vice-versa. In any event, an aging and shrinking population has seen a decline in some measures of social capital, with membership of community chonaikai falling from 61% in 2003 to 54% in 2013. Such demographic shifts may in the future require the state to provide social services currently supplied by civil society.

This study has limitations. The use of an online survey method may exclude part of the population marginalized by a ‘digital divide’. Moreover, teasing out other methods of engagement and trust relationships – including a greater and explicit focus on the use of alternative digital, social network and e-government means, and the role of transparency and open government – might have improved our results (Mahmood et al., 2020; Zhang et al., 2021). We have not controlled for respondents’ predispositions towards or expectations of government, while our satisfaction measure may not be sufficiently finely-grained, which may have affected the reliability of our study (Bouckaert and van De Walle, 2003). Our snap-shot study may not capture changing relationships over time. Some of our unexpected findings – particularly on the lack or unexpected association found between trust and socio-economic factors – need to be better understood; as do causal relationships. Our use of five-point Likert scales could generate clumping around the mid-point, particularly with neutral answers, perhaps owing to an unwillingness to commit or provide critical responses (Chyung et al., 2017). Even-numbered or four-point scales might generate more reliable results. There may be some endogeneity and reverse causality issues with participation, trust, and other variables, not accounted for in our cross-sectional analysis, but worth future analysis with more sophisticated techniques, such as structural equations and panel data (Kettl, 2017). Future studies using panel data, alternative methods of sample selection including non-digital ones, expanded antecedents, qualitative interviews, and comparative studies, might strengthen findings, including providing understanding of causal relationships.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the Japan Society for the Promotion of Science (JSPS) KAKENHI (grant number 25245025).

ORCID iD

Shaun Goldfinch

Notes

Shaun Goldfinch is ANZSOG and WA Govt. Chair in Public Administration and Policy, Curtin University, Western Australia. He has published widely on comparative public administration and policy, and consulted for governments around the world.

Kiyoshi Yamamoto is Professor of Management at Kamakura Women's University and Professor Emeritus, the University of Tokyo. He has published many articles and book chapters on public sector management and higher education policy from an international comparative perspective.

Saizo Aoyagi is a lecturer at the Faculty of Global Media Studies, Komazawa University, Japan. His research interests are mainly focused on area and tourism informatics, and human computer interaction.

Appendix

Appendix 2.

Correlation matrix.

	MT	CT	AT	IT	SP	PP	ST	PJ	ID	GE	AG	ED	IN	RP
MT	1
CT	0.680**	1
AT	0.673**	0.731**	1
CP	0.107**	0.55**	0.92**	0.095**	1
PP	0.064**	0025	0.078**	0.062**	0.354**	1
ST	0.440**	0.411**	0.473**	0.494**	0.28	0.026	1
PJ	0.559**	0.527**	0.598**	0.629**	0.130**	0.100**	0.444**	1
ID	0.061**	0.063**	0.086**	0.078**	−0.46*	−0.081**	0.063*:	0.075*	1
GE	−0.019	−0.066**	−0.016	−0.038**	−0.35	0.111**	−0.009	−0.032	−0.040*	1
AG	0.042*	−0.007	0.050**	0.031	0.398**	0.213**	−0.030	0.086**	−0.028	0.000	1
ED	0.021	0.019	0.063**	0.038*	0.006	0.130**	0.064**	0.059**	−0.042*	0.200**	−0.045*	1
IN	0.055**	0.022	0.065**	0.053**	0.137**	0.142**	0.078**	0.082**	−0.008	0.038*	0.079**	0.239**.	1
RP	−0.011	−0.037*	−0.005	−0.020	0.242**	0.154**	−0.042*	0.030	−0.001	0.038*	0.404**	−0.83**	0.101**	1

MT = Mayoral Trust, CT = Councillor trust, AT = Administrative trust, CP = Civic participation, PP = Political participation, ST = Satisfaction, PJ = Process, ID = Ideology, GE = Gender, AG = Age, ED = Education, IN = Income, RP = Residence period.

*p < 0.05, **p < 0.0. Sample size = 3100.

References

Aoki

(2015) Let’s get public administration right, but in what sequence? Lessons from Japan and Singapore. Public Administration and Development 35(3): 206–218.

Bouckaert

van de Walle

(2003) Comparing measures of citizen trust and user satisfaction as indicators of ‘good governance’: Difficulties in linking trust and satisfaction indicators. International Review of Administrative Sciences 69(3): 329–343.

Boyne

(1998) Public Choice Theory and Local Government: A Comparison of the UK and the USA. Houndmills: Macmillan.

CAO (2017) Designing 100-Year Life Society. Tokyo: CAO.

Cabinet Office (2020) Public Opinion Survey on Social Awareness (in Japanese). Tokyo: Cabinet Office.

Christensen

Lægreid

(2005) Trust in government: The relative importance of service satisfaction, political factors, and demography. Public Performance & Management Review 28(4): 487–511.

Chyung

Roberts

Swanson

, et al. (2017) Evidence-based survey design: The use of a midpoint on the Likert Scale. Performance Improvement 56(10): 15–23.

Cichocka

Jost

(2014) Stripped of illusions? Exploring system justification processes in capitalist and post-communist societies. International Journal of Psychology 49(1): 6–29.

Dahlberg

Linde

Holmberg

(2015) Democratic discontent in old and new democracies: Assessing the importance of democratic input and governmental output. Political Studies 63(S1): 18–37.

10.

Dollery

Kinoshita

Yamazaki

(2020) Humanitarian co-production in local government: The case of natural disaster volunteering in Japan. Local Government Studies 46(6): 959–978.

11.

Downe

Cowell

Chen

, et al. (2013) The determinants of public trust in English local government: How important is the ethical behaviour of elected councillors? International Review of Administrative Sciences 79(4): 597–617.

12.

Fukuyama

(2001) Social capital, civil society and development. Third World Quarterly 22(1): 7–20.

13.

Goldfinch

Yamamoto

(2019) Citizen perceptions of public management: Hybridisation and post-new public management in Japan and New Zealand. Australian Journal of Public Administration 78(1): 79–94.

14.

Goldfinch

Gauld

Herbison

(2009) The participation divide? Political participation, trust in government, and E-government in Australia and New Zealand. Australian Journal of Public Administration 68(3): 333–350.

15.

Goldfinch

Taplin

Gauld

(2021) Trust in government increased during the COVID-19 pandemic in Australia and New Zealand. Australian Journal of Public Administration 80(1): 3–11.

16.

Greasley

(2013) Trust in government, performance information and democracy. In: Llewellyn

Brookes

Mahon

(eds) Trust and Confidence in Government and Public Services. New York: Routledge, pp. 65–79.

17.

Grimes

(2017) Procedural fairness and political trust. In: Zmerli

Meer

(eds) Handbook of Political Trust. Cheltenham: Edward Elgar, pp.256–269.

18.

Grimmelikhuijsen

Knies

(2017) Validating a scale for citizen trust in government organizations. International Review of Administrative Sciences 83(3): 583–601.

19.

Houston

Aitalieva

Morelock

, et al. (2016) Citizen trust in civil servants: A cross-national examination. International Journal of Public Administration 39(14): 1203–1214.

20.

Ikeda

(2016) The Japanese Way of Thinking and World People’s Thinking: From the World Values Survey (in Japanese). Tokyo: Keiso Shobo.

21.

Imada

(2010) Civil society in Japan: Democracy, voluntary action, and philanthropy. In: Vinken

Nishimura

White

, et al. (eds) Civil Engagement in Contemporary Japan. New York: Springer, pp. 21–40.

22.

Kamin

(2010) Of sunlight and sausages. Kennedy School Review 10(10): 45–49.

23.

Kettl

(2017) Can Governments Earn Our Trust? New York: Polity.

24.

Kim

(2010) Public trust in government in Japan and South Korea: Does the rise of critical citizens matter? Public Administration Review 70(5): 801–810.

25.

Lanin

Hermanto

(2019) The effect of service quality toward public satisfaction and public trust on local government in Indonesia. International Journal of Social Economics 46(3): 377–392.

26.

Magni

(2015) It’s not fair! The impact of perceived injustice on political participation. In: Proceedings of European Political Science Association Annual Conference, Vienna, Austria, 25–27 June 2015. EPSA.

27.

Mahmood

Weerakkody

Chen

(2020) The role of information and communications technology in the transformation of government and citizen trust. International Review of Administrative Sciences 86(4): 708–728.

28.

Manning

Guerrero

(2013) Why do politicians reform? Do they know something about the drivers of trust in government? In: Llewellyn

Brookes

Mahon

(eds) Trust and Confidence in Government and Public Services. New York: Routledge, pp. 80–120.

29.

MIC (2017a) Trends of Voting Rate by Age. Tokyo: MIC.

30.

MIC (2017b) Survey on Results of the 18th Local Elections. Tokyo: MIC.

31.

Mizrahi

Vigoda-Gadot

Cohen

(2010) Trust, participation and performance. Public Management Review 12(1): 99–126.

32.

Napier

Jost

(2008) Why are conservatives happier than liberals? Psychological Science 19(6): 565–572.

33.

Newton

Stolle

Zmerli

(2018) Social and political trust. In: Uslander

(ed.) The Oxford Handbook of Social and Political Trust. New York: Oxford University Press, pp.37–56.

34.

Noda

(2017) Trust in the leadership of governors and participatory governance in Tokyo Metropolitan Government. Local Government Studies 43(5): 776–797.

35.

Noda

(2019) Citizen expectations and satisfaction of service performance: Lessons from subnational governments in Japan. Asia Pacific Journal of Public Administration 41(3): 142–156.

36.

OECD (2019) Government at a Glance 2019. Paris: OECD.

37.

Oyama K (2015) Which effects trust in the civil service, NPM or Post-NPM? Outcomes and process from comparative perspective. Journal of Law, Politics, and Sociology 88(9): 1–24.

38.

Park

Joaquin

(2012) Of alternating waves and shifting shores: The configuration of reform values in the US federal bureaucracy. International Review of Administrative Sciences 78(3): 514–536.

39.

Pekkanen

(2006) Japan’s Dual Civil Society: Members without Advocates. Redwood, CA: Stanford University Press.

40.

Popkin

(1991) The Reasoning Voter: Communication and Persuasion in Presidential Elections. Chicago: University of Chicago Press.

41.

Putnam

(2000) Bowling Alone: The Collapse and Revival of American Community. New York: Simon & Schuster.

42.

Putnam

Leonardi

Nanetti

(1993) Making Democracy Work: Civic Traditions in Modern Italy. Princeton, NJ: Princeton University Press.

43.

Reitan

Gustafsson

Blekesaune

(2015) Do local government reforms result in higher levels of trust in local politicians? Local Government Studies 41(1): 156–179.

44.

Suzuki

(2017) Government expenditure cuts and voluntary activities of citizens: The experience of Japanese municipalities. Asia Pacific Journal of Public Administration 39(4): 258–275.

45.

Suzuki

Demircioglu

(2021) Is impartiality enough? Government impartiality and citizens’ perceptions of public service quality. Governance 34(3): 727–764.

46.

Suzuki

Avellaneda

(2020) Direct and non-linear innovation effects of demographic shifts. Australian Journal of Public Administration 79(3): 351–369.

47.

Tiebout

(1956) A pure theory of local expenditures. Journal of Political Economy 64: 416–424.

48.

Theiss-Morse

Hibbing

(2005) Citizenship and civic engagement. Annual Review of Political Science 8(1): 227–249.

49.

Tyler

(2006) Why People Obey the Law. Princeton, NJ: Princeton University Press.

50.

Uddin

Mori

Adhikari

(2019) Participatory budgeting in a local government in a vertical society: A Japanese story. International Review of Administrative Sciences 85(3): 490–505.

51.

Van der Wal

Demircioglu

(2020) Public sector innovation in The Asia-Pacific trends, challenges, and opportunities. Australian Journal of Public Administration 79(3): 271–278.

52.

Van de Walle

(2011) New public management: Restoring the public trust through creating distrust? In: Christensen

Lægreid

(eds) Ashgate Research Companion to New Public Management. Aldershot: Ashgate, pp.309–320.

53.

Van de Walle

Migchelbrink

(2022) Institutional quality, corruption, and impartiality: The role of process and outcome for citizen trust in public administration in 173 European regions. Journal of Economic Policy Reform, 25(1): 9–27.

54.

Van Ryzin

(2011) Outcomes, process, and trust of civil servants. Journal of Public Administration Research and Theory 21(4): 745–760.

55.

Wilkinson

Briggs

Salt

, et al. (2019) In participatory budgeting we trust? Fairness, tactics and (in)accessibility in participatory governance. Local Government Studies 45(6): 1001–1020.

56.

World Economic Forum (2018) The Global Gender Gap 2018. Geneva: World Economic Forum.

57.

World Values Survey (2016) World Values Survey. Available at: http://www.worldvaluessurvey.org/WVSDocumentationWV6.jsp (accessed February 25).

58.

Yamagishi

(1994) Trust and commitment in the United States and Japan. Motivation and Emotion 18(2): 129–166.

59.

Zhang

Zhao

X-y

Xue

Y-x

, et al. (2021) A meta-analysis of how the culture and technical development level influence citizens’ adoption of m-government. International Review of Administrative Sciences. DOI: 10.1177/00208523211057358.