From Social Trust and Happiness to Government Trust: The Moderating Role of Political Systems and Governance in the Philippines

There are frequent calls to enhance citizens’ trust in government to pave the way towards a new paradigm of participatory governance and strong citizen support for government. In various realms, citizens may directly or indirectly engage with the government through various available mediums, even though, despite the availability of various policies and services provided by the government, citizens are generally passive and adamant in trusting the public sector. While many studies have explored a set of determinants that influence citizens’ trust in government (i.e., central government, local government, parliament, and the legal system), few studies have ascertained the relationship and the role of social trust, happiness, governance, and political systems. These are critical factors that may influence trust in government. To address this gap, this study draws on the theoretical lens of social capital theory, proposing that cognitive social trust and citizen happiness—environment and performance—are the most likely predictors of citizen trust in government. This study assumes that citizens’ perceptions of governance and political systems will moderate the effect of social trust and happiness on trust in government. Using data from the Asia Barometer Survey 2007 and focusing on data collected from the Philippines, this study tests a latent model employing the structural equation modelling technique. It finds that happiness negatively predicts trust in the central government and the legal system, while all other predictors do not have a significant effect. The findings also show that the political system moderates the impact of social trust and happiness on trust in government. Finally, this article points out its theoretical, empirical, and practical implications and provides directions for future research.


Introduction
There are frequent calls to enhance citizen trust in government and thereby pave the way towards a new paradigm of participatory 1 governance and strong citizen support for government. In various realms, citizens may directly or indirectly engage with the government through various available mediums. However, despite the availability of various policies and services provided by the government, citizens tend to remain passive and adamant in trusting the public sector. Scholars have recognised that the causes and effects of trust in government are complex (Kim, 2010). There is broad evidence that the public is more likely to evaluate government trustworthiness based on recent government actions.
Trust has been acknowledged as a key factor in the relationship between citizens and government, and the lack or failure of trust can hinder the creation of public value (Kelly, Mulgan, & Muers, 2002). According to Ba and Pavlou (2002), trust is defined "as the subjective assessment of one party that another party will perform a particular transaction according to his or her confident expectations, in an environment characterized by uncertainty" (p. 245). While there is a growing body of literature on trust in government, there are few theoretical and empirical studies that detail the role of individual values or well-being-social trust (social capital) and happiness (subjective well-being)-on citizens' trust in their governments in an Asian context. Thus, this study analyses how social trust and happiness are associated with government trust, and then explores the interaction of the government's performance and the political system.
Generally, this study examines how social trust and happiness are associated with trust in government-be it the central government, local government, legal system, or legislature. Do the political system and government's governance performance moderate the impact of social trust and happiness on trust in government? Accordingly, this study asks four questions: (1) How does social trust influence citizens' happiness and trust in government? (2) Does happiness mediate the relationship of social trust and trust in government? (3) What is the relationship of happiness and trust in government? (4) Do governance and the political system moderate the impact of these predictors on trust in government? To answer these questions, first, we provide a literature review using social capital theory (Putnam, 2008;1993) as our primary theoretical lens for explaining the relationship of the variables (i.e., social trust, happiness, and trust in government). Second, we discuss the research methods and measures used to denote the variables. Third, we present the research findings and results of validity and reliability tests, i.e., exploratory factor analysis (EFA), regression analysis, moderation, and mediation test. Lastly, following discussion of the results, we explain the findings; discuss their theoretical and practical implications; and present the limitations and paths for future research. Figure

Trust in Government
Trust in government is generally a reflection of a government's performance (Keele, 2007), which may serve as the foundation for successful public policies (OECD, 2013). It refers to the degree of trust and confidence the public has towards the executive, legislative, and judicial branches of its government (Gallup, n.d.). Scholars have recognised that the causes and effects of trust in government are complex (Kim, 2010). There has been wide evidence that the public is more likely to evaluate government trustworthiness based on recent government actions, that is, by looking at the recent performance of politicians-public executives, legislators, and the legal system (Keele, 2007;Citrin, 1974).
Citizen trust is important for governments to make "binding decisions, commit resources to attain the societal goals" (Gamson, 1968;Chanley, Rudolph, & Rahn, 2000, p. 240), and ascertain citizens' acquiescence and/or agreement to public policies (Chanley et al., 2000;Levi, 1998;Scholz & Lubell, 1998;OECD, 2013). Interpersonal trust is learned through interactions and civic activities (Brehm & Rahn, 1997) which can lay the foundations for trust in government (Putnam, 2000;Brehm & Rahn, 1997). Citizens' involvement in civic activities connotes trust in the political process and in the government's efforts to bring about social change (Keele, 2007).

Consequences of Social Trust: Happiness and Trust in Government
Familiarity is expected to breed trust (Dinesen & Sønderskov, 2018). People trust others through interaction. The more experience they gain from interaction, the more trust they gain. As indicated by Putnam (1993Putnam ( , 2000, trust derives from reciprocity, which can be learned only in cooperation with others. Generalised or Social Trust (generalised) refers to the positive feeling when members of society, in general, are perceived trustworthy (Uslaner, 2002), and disposed to trust unknown others by default (Dinesen & Sønderskov, 2018). Such a feeling also implies that people can rely on others and will provide help in various ways (Hardin 2006). Previous studies concerning social trust have focused on the effect of ethnic diversity (Putnam, 2007;Abascal and Baldassarri, 2015). For example, Putnam (2007) found that residential ethnic diversity in the United States has negatively impacted social trust. Putnam (1993) is the scholar behind the conceptualisation of civic engagement. He points out the importance of "social capital", which refers to the social connections, networks, and interpersonal trust that occur in communities (Putnam, 2000). It also indicates the quantity and quality of social interactions in a community (Petrou & Kupek, 2008). Social capital has two aspects: (a) civic engagement in a community, state, or nation, and (b) interpersonal trust, or willingness to ascribe benign intentions to others (Keele, 2007, pp. 243-244). Helliwell and Putnam (2004) argue that social capital is an essential predictor of happiness, as it-social trust-can enhance happiness through health and well-being (Tokuda, Fujii, & Inoguchi, 2010). As such, the more people interact, quantitatively and qualitatively, the higher the social capital (social trust) developed.
Happiness becomes inevitable when an entire community develops trust within itself, and become fixated with the idea that social capital is embedded in their social interactions. These theoretical backdrops enable us to propose the following hypothesis: Hypothesis 1: Social trust is positively related with happiness.
In the same manner, interpersonal trust is learned through interactions and civic activities (Brehm & Rahn, 1997), which can lay the foundation for trust in government (Putnam, 2000;Brehm & Rahn, 1997). People who join associations are naturally trusting, because trust has a moral dimension and virtue that is learned since childhood (Uslaner, 2002). Citizens' involvement in civic activities connotes a trust in the political process and in the government's efforts to bring about social change (Keele, 2007). Thus, we propose: Hypothesis 2: Social trust is positively related with trust in government

Relationship of Happiness and Trust in Government
Contemporary public policy and governance have started to notice the importance of subjective wellbeing (SWB), a factor going beyond economic growth. Indeed, happy people live longer, and happiness is what people ultimately pursue (Diener & Seligman, 2004). To capture SWB empirically, scholars use both cognitive (i.e., life satisfaction or happiness) and emotional (i.e., positive and negative affection) measures (Arthaud-Day, Rode, Mooney, & Near, 2005;Simsek, 2011).
Studies on happiness in western countries such as the United States have lent no support to the proposition that the general public puts its trust in the government's efforts to maximise happiness (Duncan, 2013). This is surprising, given the range of services delivered by governments that are intend to enhance the wellbeing of citizens. According to the OECD, trust in government is also anchored in "citizens' experiences when receiving public services" which may enhance happiness. In this study, we argue that happiness will foster public trust in government. This theoretical backdrop enables us to propose the following hypothesis: Hypothesis 3: Happiness is positively related with trust in government.
In addition to the direct relationship discussed above, the relationship between social trust and trust in government could be mediated by happiness. The mediation process can be observed in the relationship established between social trust and happiness, and in the discussion explicating the relationship between happiness and trust in government. Thus, we propose: Hypothesis 4: Happiness mediates the relationship between social trust and trust in government.

The Moderating Role of Governance and the Political System
Governance, as distinguished from government, refers to "the processes of interaction and decision-making among the actors involved in a collective problem that lead to the creation, reinforcement, or reproduction of social norms and institutions" (Hufty, 2011). We argue that it is in the process (i.e., creation, reinforcement, and reproduction) through which social trust and happiness are most likely to be enhanced. It may provide a basis for the government to be trusted. As such, we propose: Hypothesis 5: Governance moderates the relationship of (a) social trust and (b) happiness and trust in government.
As a broad and encompassing system of politics and government, the political system-the set of formal legal institutions that constitutes a "government" or "state" (Heslop, 2014)-has an influential role on social trust, happiness, and trust in government. On the contrary, trust does not exist independently of government. People lose trust in governmental institutions when faced with corruption, which may lead them to assume that others commit misfeasance and malfeasance in their official functions, i.e. by committing bribery. Although such notion still confirms trust and good governance are related (Rothstein, 2003), Rothstein's hypothesis still maintains that good governance is what people expect in a political system where the core standard for public service is public trust. Easton (1953;1957;1965) writes that political systems are closer to trust in government-i.e., have trust as the central dimension. As such, in democratic institutions, the more political systems are open, as shown through the redress of grievances and economic opportunities, the more they can moderate the dynamics of social trust, happiness, and trust in government. Thus, we propose: Hypothesis 6: The political system moderates the relationship of (a) social trust and (b) happiness and trust in government.

DATA AND METHOD
The causal relationship of latent variables in the hypothesised model was determined by employing multiple regression analysis on SPSS (Version 21). Prior to causal analysis, we employed factor analysis testing on the distinctiveness and convergent validity of the measuring constructs. This estimation is based on a weighted least square (WLS) parameter (Hox, Maas, & Brinkhuis, 2010) and principal component analysis (PCA). Expectation Maximisation (EM), compared with multivariate analysis, has the ability to measure the relationships of multiple variables in a model (Byrne, 2001) and define a model explaining an entire set of relationships (Kline, 2005). Finally, the multiple linear regression technique was employed to assess the relationship between the variables and the moderating role of governance and the political system.

Data and Instrumentation
The data employed in the study come from the Asia Barometer Survey 2007 (Asia Barometer, 2007). The proponent of the survey, jointly conducted by the Research and Information Center for Asian Studies, Institute of Oriental Culture, University of Tokyo, described it as: (a) representing the largest ever comparative survey in Asia, covering East, Southeast, South, and Central Asia; (b) focusing on daily lives of ordinary people (bumi putra) and their relationships with their family, neighbourhood, workplace, social, and political institutions and marketplaces; and (c) involving country-wide faceto-face surveys using standardised instruments designed around a common research framework. The survey was administered in seven Asian countries, namely: (1) Cambodia, (2) Indonesia, (3) Laos, (4) Malaysia, (5), Myanmar, (6) the Philippines, and (7) Thailand, with 7,020 total useful samples. The current study focuses only on data collected from the Philippines, with a total of 1,000 respondents.
Exploratory factor analysis (EFA), using a principal component analysis technique, was used to bring intercorrelated variables together under one general and underlying variable. We computed factor extraction and found composite factor scores for each identified construct in the research model. Results show that all latent variables were clearly defined by each underlying variable.

Missing Data Adjustments
We attempted alternative techniques for dealing with missing data, such as mean substitution; however, these did not have any significant change on the overall mean (see Cohen, Cohen, West, & Aiken, 2013). While listwise or pairwise deletion techniques were considered, these dropped a significant amount of data in the analysis (Roth, 1994). In lieu of mean substitution and listwise deletion techniques, we employed the Expectation Maximisation (EM) method to deal with missing values. This approach is an iterative method that is appropriate for imputing single values, using other variables to impute a value and accordingly verify that the imputation is the most likely value of the variable through the iteration of the E (expectation) and M (maximisation) steps (Dempster, Laird, & Rubin, 1977;Gold & Bentler, 2000).

Measures
The Asia Barometer Survey 2007 was designed to gather citizens' perceptions on government efforts to implement various democratic policies; the fulfilment of rights enshrined in the constitution such as the right to vote and demonstrate; trust in government institutions (central government, local governments, congress, legal system, educational system, etc.); political predispositions; and engagement in politics. Table 1 shows the exploratory factor analysis employing the principal component analysis rotation technique and Varimax extraction method. The factor loading ranges from .436 to .825.
Trust in Government. The items used to measure trust in government were based on the responses of participants to the survey questionnaire, "Please indicate to what extent you trust the following institutions to operate in the best interests of society. If you don't know the reply or have no particular opinion, please say so (SA for each institution): the central government; your local government; the legal system; and the parliament, congress." The selected items depict trust in government institutions that represent the executive, legislative, and judicial branches. The responses are classified as "trust a lot", "trust to a degree", "don't really trust", "don't trust at all", "haven't thought about it", and "don't know". The Cronbach's α value (four items) is equal to 803.
Social Capital. The measures of social capital were derived from the survey item that asked respondents to indicate their responses to several questions: "Generally, do you think people can be trusted or do you think that you can't be too careful in dealing with people (that it pays to be wary of people)," with the responses being, "most people can be trusted", "can't be too careful in dealing with people", and "don't know." Participants were also asked, "Do you think that people generally try to be helpful or do you think that they mostly look out for themselves", with the responses being, "people generally try to be helpful", "people mostly look out for themselves", and "don't know." Lastly, this item includes responses to the question, "if you saw somebody on the street looking lost, would you stop to help?", with the responses being, "I would always stop to help", "I would help if nobody else did", "it is highly likely that I wouldn't stop to help", and "don't know." The Cronbach's α value (four items) is equal to .273.
Happiness. The measures for happiness asked respondents how they perceived the happiness they experience based on the following survey items: "All things considered, would you say that you are happy these days?", with the responses being, "very happy", "quite happy", "neither happy nor unhappy", "not too happy", "very unhappy", and "don't know." Also, "how often do you feel you are really enjoying life these days?", with the responses being, "often", "sometimes", "rarely", "never", and "don't know." Lastly, participants were asked, "how much do you feel you are accomplishing what you want out of your life?", with responses being "a great deal", "some", "very little", "none", and "don't know." The Cronbach's α value (three items) is equal to .773.
Political System. The measures adapted for the construct political system were based on whether participants perceived the following systems: (a) governance by a powerful leader without the restriction of parliament or elections, (b) a system whereby decisions affecting the country are made by experts (such as bureaucrats with expertise in a particular field) according to what they think is best for the country; (c) military government; and (d) a democratic political system, as "very good", "fairly good", "bad", or "don't know". The Cronbach's α value (three items) is equal to .626.
Governance. The measures used to describe governance were taken from responses to the question, "how well do you think the [YOUR COUNTRY] government is dealing with the following issues?", with responses being, "very well", "fairly well", "not so well", "not well at all", and "don't know". These fields include:  a. Rotation converged in 5 iterations.

Results
Testing the causal relationship provided in the hypothesised model, we employed regression analysis to determine the direct impact of the independent variables (i.e., social trust and happiness) on trust in government (i.e., central government, local government, legal system, and parliament/congress) as well as the moderating role of the political system and governance. Bootstrapping analysis was used to check the mediating effect of happiness on the relationship of social trust and trust in government.
First, as the results show (see Table 4), social trust has no significant effect on happiness (β=.042 ns ), thus rendering Hypothesis 1 not supported. The mediation analysis also indicates that happiness does not play a mediating role in the relationship between social trust and trust in government; as such, Hypothesis 2 is not supported. Although the results do not show any concrete evidence of the role of social trust in happiness or the subsequent intervening role of the latter on social trust's impact on trust in government, it may imply that the social capital of citizens may not be strong enough to trigger individual happiness. Second, the results reveal that social trust (see Table 5) does not have any significant effect on trust in government. However, happiness is significantly and negatively associated with trust in the central government (β=-.555, p < .01) and legal system (β=-.466, p <. 05); while not significant associated with trust in local government and parliament/congress. These results show that Hypothesis 3 is partially supported. The results may imply that happiness may influence the direction of trust in government. For example, happiness may negatively affect trust in the central government for several reasons, such as government failure to provide expected public services. Similarly, happiness is associated with the judiciary when it fails to satisfy the public's need for justice and fair dealings.
Lastly, examination of the moderating role of governance reveals no significant effect on the impact of social trust and happiness on trust in government, thus rendering Hypotheses 4 and 5 not supported. Meanwhile, the political system dampens the positive impact of social trust on trust in the central government (β=-.490, p < .05), thus partially supporting Hypothesis 6. Also, the political system shows a significant moderating effect on happiness and its impact on trust in the central government (β=.517, p < .05), local government (β=.693, p < .001), legal system (β=.508, p < .05), and parliament/congress (β=-.615, p < .05). This reveals that the political system plays an influential role in the process of enhancing happiness as well as trust in government. As it clearly suggests some critical notes on the identification of an effective and applicable political system, there is a need to carefully identify this political aspect. Some specific political systems may apply in a given context. For example, the Philippines, which has been a democracy for almost three decades, may require a stronger political system that may embody the public's demand for total happiness and trust in government. Table 5 summarises the results of the multiple regression and the moderation analyses.

Moderation Test of Political System
The regression model has an R-squared value of .175, meaning that 17.5% of variance in trust in the central government is explained by the predictors in the hypothesised model. The F-value is equivalent to 17.29 at p<.001 level, signifying that predictors of trust in the central government have significant influence. From the results, we can see that the political system has a significant moderating effect on happiness and its impact on trust in the central government. From plot analysis (see Figure 2), we can observe a significant change in the slope based on the level of the political system. For trust in local government, the findings show that the political system moderates the impact of happiness on trust in government. The regression model has an R-squared value of .098, meaning that 9.8% of variance in trust in local government is explained by the predictors in the hypothesised model. The F-value is equivalent to 8.80 at p<.001 level, signifying that predictors of trust in local government have significant influence. From the results, we can see a significant effect on the impact of trust in local government. The plot analysis (see Figure 3-A) shows a significant change in the slope based on the level of the political system. For trust in the legal system, the model that includes the interaction effect reveals an R-squared value of .118, which means that 11.8% of variance in trust in the legal system is explained by the predictor variables. The F-value is equivalent to 10.89 at p<.001 level. The moderation plot (see Figure 3-B) displays a significant change in the slope depending on the level of political system. Lastly, for trust in parliament/congress, the regression model has an R-squared value of .060, meaning that 6.0% of variance in trust in parliament/congress is explained by the predictors in the model. The F-value is equivalent to 5.19 at p<.001 level, showing that the predictors have a significant effect on trust in parliament/ congress. The slope in the plot analysis (see Figure 3-C) shows a significant change as political change is added to the interaction model.

Discussion and Conclusion
Given seemingly borderless public governance, the public sector now faces the tremendous challenge of advancing a more participatory, collaborative, and open government. These concepts, which define the principles of democracy, require effective and efficient government that is founded on trust, transparency, and accountability. While the government must engage citizens in promoting full participation in government affairs, the issues of credibility and daunting corruption have undermined public trust, demoralised citizens, and eroded support for the government. This study has identified and assessed the predictors of trust in government-trust in central government, local governments, legal systems, and parliament/congress. Among the important C determinants that may be considered are social trust, citizen's happiness, the impact of governance, and the political system. This study has analysed the relationship of social trust, happiness, and trust in government. Also, it has investigated the mediating role of happiness in the relationship between social trust and the outcome variable-trust in government (i.e., trust in the central government, local governments, legal system, and parliament/congress). The research model and causal relations were tested employing the Asia Barometer Survey 2007, focusing on data collected from the Philippines. The results of the EFA, reliability analysis, and CFA confirmed that the proposed causal model is a good fit for the data.
First, this study found that social trust does not have any significant impact on happiness and trust in the government, which means that Hypotheses 1 and 2 are not supported. Although the results did not show any evidence connecting these variables, we argue that social trust has a contributory impact on happiness and trust in government (see, for example, Keele, 2007;Tokuda, Fujii, & Takashi, 2010).
Second, the results of the study revealed that happiness influences trust in the central government and the legal system, thus partially supporting Hypothesis 3. Conversely, the findings of the study indicate that happiness does not mediate the relationship between social trust and trust in government; therefore, Hypothesis 4 is not supported. The results suggest that happiness does not necessarily affect trust in government. This finding supports Duncan's (2013) apprehension that the general public's trust in government maximises happiness.
Lastly, examination of the moderating role of governance and the political system provides interesting findings. Governance does not show any significant moderating impact on social trust and happiness; therefore Hypothesis 5 is not supported. Meanwhile, the political system has a significant moderating role on the interactions between social trust and trust in the central government, as well as the interactions between happiness and trust; therefore, Hypothesis 6 is supported.

Theoretical and Practical Implications
This study shows that social trust is found to be lower among Filipinos. From the time the survey was conducted until 2007, the President was allegedly involved in election fraud, dubbed the "Hello Garci" controversy, and her administration was shattered by anomalous government contracts, e.g. Fertilizer Fund scam, NBN-ZTE deal, Northrail Project, and the like. These issues have eroded the trust of people in the government, and invited Filipinos to conduct massive protest (Ugaddan & Calata, 2018). The political system at that time may imply a need to enhance government intervention and implement important policy strategies that promote active and favourable interactions between citizens. There is a need, therefore, for the government to not just be transparent in entering contracts where large amounts of money are involved, but also in the creation of policies and programmes that help elevate the lives of the people in general. This also suggests that the government must improve fundamental competencies that can enhance citizens' trust (Kim, 2010). For example, "improving the economic development and the quality of public services, and reducing corruption are necessary components of leadership development for enhancing public trust in government" (Kim, 2010, p. 808). The government-the central government, local governments, legal systems, and congressmust encourage the adoption of a political system that enables citizens to enjoy free and fair interactions, enhance their happiness through various projects and activities that may directly ask their involvement, and provide services geared towards real governance.
The presidential election of 2016, which elected the current President, involved some 40 million voters in the electoral process (Palatino, 2016). This means that at least 81 per cent of eligible voters participated in the election. Such voter turnout is higher than that of the 2010 election. This indicates that there was high hope that a new government would create a productive environment for the people, as also shown by the electoral mantra that dominated the election: "change is coming!" At the cornerstone of the new administration has been the "Build, Build, Build" (BBB) programme, which seeks to accelerate government spending on infrastructure projects (P8 to 9 trillion from 2017 to 2022). BBB is focused on industries that will yield robust growth, create jobs, and improve the lives of Filipinos (Mawis, 2018). With the BBB programme, the government seems to have again opened its windows to the political milieu of the previous decade, when it frequently entered into allegedly dubious and anomalous contracts. This scenario could test governmental transparency and put it at stake because of the involvement of significant amounts of the people's money, as well as foreign loans (mostly from China). At present, while BBB projects are underway, the government is also facing problems and issues. Since the current President came into power, from July 1, 2016, up to June 11, 2018 -the police have recorded 23,518 Homicide Cases Under Investigation (HCUI), equivalent to an average of 33 people killed a day, excluding 4,279 suspects killed in anti-illegal drug operations, with an average of 6 a day (Talabong, 2018). The alleged "extrajudicial killings" were perpetuated against drug users and pushers. The President has promised to end corruption in 3-6 months but failed. Except for the police and military personnel that have already been benefited by salary increases, other government workers, like the teachers, have yet to see the increases promised to them by the President at the beginning of the term. The territorial disputes with China is also a hot issue as the President has been staging "friendly" relations with China even if the latter has encroach upon Philippine sovereignty by occupying the islands in the West Philippine Sea. With these issues and other economic problems such as high inflation rates, rice shortages (Romero, 2018), and high prices of goods and commodities, thus making the President's trust rating fluctuate ("Public trust in Duterte," 2018). It is up to the government to either make sense of these issues to regain the trust of the people or continue building infrastructure that may lock the country in a "debt trap" and controversies happened decade ago.

Limitations and Future Research
This study has some limitations. Relying solely on survey data may affect the reliability and generalisability of the findings. Although the survey collected data from a large pool of respondents, "common method bias" and social desirability might not be fully eliminated in this research. For this study, respondents were randomly approached in their houses. This may pose a bias challenge. Finally, this study obtained data from only one time period; therefore, the findings should be generalised with some level of caution.

Construction of Indices
Predictors of Trust in Government Social Trust (3 items) • Generally, do you think people can be trusted or do you think that you can't be too careful in dealing with people (that it pays to be wary of people)? • Do you think that people generally try to be helpful or do you think that they mostly look out for themselves? • If you saw somebody on the street looking lost, would you stop to help?