How Does Religious Belief Affect Happiness of Older Adults: The Participation Effect and Support Effect

Abstract

Religion can influence the realization of happiness in older adults. However, the relationship between religious belief and happiness of older adults and its mechanisms are not very clear. Using 5177 individuals aged ≥ 60 years in China for empirical analysis, the results show that there is a significant positive relationship between religious belief and older adults’ happiness. The frequency of religious participation has a significantly moderating effect on the relationship between institutional religious affiliation and happiness; participation has an effect on happiness for those involved in institutional religion, but not an affect for those with diffused beliefs, which reflects the participation effect. Social support plays a mediating role between religious belief and older adults’ happiness. People with religious beliefs are more likely to form social relationships and gain social support from religious members and groups to improve their happiness, which reflects the support effect. In addition, the influence of religious belief on happiness with low socioeconomic status is more significant, and the influence in rural areas is also more significant, showing the heterogeneity of individual socioeconomic status and registered residence. The freedom of older adult-religious believers to participate in the legal activities of religious organizations should be fully guaranteed. The religious faith of older adults in the process of active aging should be given attention.

1. Introduction

Over the past decade, aging populations have gradually become a global phenomenon. In many countries, the proportion of people aged 65 years or above has exceeded 20% of the total population, creating a super-aged society (Liao et al. 2000). In 2000, China became an aging society, with the proportion of the older adults (aged 65 and above) exceeding 7% of total population for the first time (Wu et al. 2019). There were 253 million people in China who were 60 or older in 2019, representing 18.1% of the overall population, and 12.6% of the total population was 65 or older (Yang and Yin 2022). Therefore, in the context of aging populations, improving the mental health of older adults and ensuring they can live peaceful and contented lives as they age has become a crucial concern in the realms of psychology and sociology, among others (Batterham et al. 2012; Liu et al. 2016; Tian 2016).

Studies have shown that happiness is a significant indication of mental health and successful aging in older adults (Hsu et al. 2021; Peterson et al. 2014). Many studies have examined the elements that have an impact on older adults’ happiness, including education, disposable income, health status, social participation, and family support (Bodogai et al. 2020; Lai et al. 2020; Minarro et al. 2021; Nguyen et al. 2016; Zhang and Zhang 2015). In addition, religious belief is an important manifestation of spirituality. As a result, we cannot dismiss the impact of religious belief on individual happiness. Numerous pieces of evidence indicate that religious beliefs or religious conducts are associated with happiness (Campante and Yanagizawa-Drott 2015; Snoep 2008). Using a three-dimensional model of religious engagement including organizational religiosity, nonorganizational religiosity, and subjective religiosity, and a six-dimensional model of mental health including self-acceptance, environmental mastery, purpose in life, and positive relations with others, some have tested the relationship between religious engagement and mental health in a sample of older African Americans and concluded that different dimensions of religious engagement will have significant positive effects on mental health (Frazier et al. 2005). Research shows that older adults who are religious have higher levels of well-being, and that religious belief is one of its strongest determinants (Okun and Stock 1987). Moreover, the relationship between religious belief and happiness increased with age (Witter et al. 1985). Some studied the relationship between religious practices and happiness in a group of older adults in central and eastern Europe, using life satisfaction as a measure of happiness and Orthodox, Protestant, and Roman Catholic as a measure of religious affiliation, and concluded that religious practices did not have a sufficient influence on their happiness (Bodogai et al. 2020). The above research conclusions are drawn by different researchers in the Western religious background by selecting different well-being measurement indicators. For the measurement of subjective well-being, some studies use happiness and some use life satisfaction as well-being. Yet, in fact, the two concepts, though related, are different. Well-being should also include measures of mental health. This study focuses on the influence of religious belief on older adults’ happiness. We define happiness as a feeling that everyone desires, as a psychological outcome, and as an individual’s emotional evaluation of positive emotions (Rizvi and Hossain 2017). Therefore, our measure of happiness uniformly adopts happiness as defined by internal emotional experience.

Nonetheless, these studies are typically centered on Western cultural backgrounds, with the majority of respondents being white or Christian. There are less studies on the relationship between religion and happiness in the context of Chinese culture (Sun and Zhang 2019). The researchers analyzed the relationship between religion and happiness among Chinese urban and rural residents using data from a 2007 national survey of the Spiritual Life of Chinese residents. The results showed that there was a positive correlation between religious affiliation and happiness among religious people in China, and this positive correlation was stronger in the sample of rural respondents (Zhang et al. 2019). In the study of a sample of Chinese seniors over 80 years old, the researchers found that there was a significant negative correlation between religious participation and subjective well-being after controlling for variables such as demographic characteristics, health status, living status, and wealth status (Brown and Tierney 2009). In the current literature, the relationship between religion and happiness in the context of Chinese culture is disputed. Whether this association is positive or negative requires additional discussion and testing with additional samples.

In addition, different religious affiliations have developed their own distinct cultures and religions as a result of multicultural influences. In practice, there are significant differences in the religious activities of different religious affiliations. Religious activities are the collective activities of worship, prayer, and other activities of religious organizations and believers. Participating in religious activities can help individuals obtain social assistance from religious groups and bring believers psychologically closer to the objects of their belief. The current classification of religious types is primarily based on basic philosophical concepts of religion and theological categories of religion, which include Buddhism, Christianity, and Islam (Lu et al. 2008). Aside from the three world religions mentioned above, people in China, India, Japan, and other Asian countries have a wide range of folk beliefs. In addition, the Chinese government implemented specialized supervision and management of diverse religious concerns at an extremely early stage (Yang 2006). Unlike in several nations, religious markets are not controlled. Religion has played a significant impact in the development of Chinese society. In recent years, public uneasiness has increased (Chen and Williams 2016), and many have turned to religions for solace and to deflect anger. Consequently, the number of religious adherents and religious activities has increased dramatically in China (Yang 2010).

In conclusion, we aim to fill in the gaps of existing studies by taking mainland China as an example to study the relationship between religious belief and happiness, and the mechanism of the effects of different elements related to religious belief on happiness in the context of Chinese culture. First, we used the data of 5177 older adults in China to verify the influence of religious belief on happiness through robustness testing methods, so as to ensure that the conclusions drawn were representative. Then, the econometric model was used to investigate whether religious participation played a moderating role in the relationship between different religious affiliations and happiness. In addition, we investigated the mediating influence mechanism of social support variables on the relationship between religious belief and happiness through the mediating effect. In the end, we examined whether the effects of religious belief on happiness are heterogeneous with personal socioeconomic status and registered residence.

2. Literature Review and Hypotheses

Religious beliefs make people feel better about themselves, and are widely used to cope with uncertain situations, such as serious illness, the death of a relative, or even a threat to their own life (Hogg et al. 2010). Greater religious involvement is associated with improved physical and mental health, as well as increased subjective well-being throughout life (Hoverd and Sibley 2013). Religion becomes increasingly important in the adaptation of older adults as they age (Blazer and Palmore 1976). People who become more religious experience long-term increases in life satisfaction (Headey et al. 2010).

For example, using data from 5675 older adults in the United States, researchers used structural equation models to examine the effects of religious belief on health status and mental health in a 1998 study. According to the findings, religious belief had a statistically significant positive effect on health status and mental health (Levin and Chatters 1998). Ardelt discovered in a 2003 study, using a multiple regression analysis, that religious belief based on life purpose was positively correlated with the subjective well-being of the elderly, while it was negatively correlated with fear of death and avoidance of death (Ardelt 2003). Fry discovered, through a hierarchical regression analysis, that personal meaning, the importance of religion, participation in formal religion, participation in spiritual practice, self-awareness of inner peace, and access to religious resources are all important predictors of improving subjective well-being in American national samples (Fry 2000). In addition, the researchers used 215 rural older adults as research participants and found that religious belief and spirituality were positively correlated with well-being through hierarchical regression (Yoon and Lee 2006). Another researcher worked with 346 community residents who were over 65 years old in rural Japan, analyzed the connection between several aspects of religious belief content and well-being, and found that belief in life after death and religious beliefs were positively linked with well-being (Imamura et al. 2015).

On the basis of the above conclusions, we put out the following hypotheses:

Hypothesis 1.

Religious belief has a significantly positive affect on older adults’ happiness.

As was already mentioned, various religious organizations engage in various religious activities. Some researchers have observed the development of numerous religions under the unique historical and cultural background of China by scouring Chinese religious literature. Yang Jingkun classified various religions in Chinese society according to their religious convictions, ritual symbols, and organizational characteristics and divided them into institutionalized religions and disseminating religions through in-depth observation and research of various religions in China. Catholicism, Christianity, and Islam are all recognized as institutionalized religions. They regularly participated in religious activities and had their own, largely independent institutions, rituals, and organizations. The second are the disseminating religions. Its theology, rituals, and structure were closely woven into social orderly elements such as the idea and development of secular institutions; the disseminating religions mainly referred to various folk beliefs and ancestor worship (Yang 1961).

The characteristics of religious believers who participate in religious activities are different due to the differences in the organization forms and frequency of different religious beliefs, and as a result, the differences in participation in religious activities will affect the relationship between the organization’s forms of religious affiliation and happiness.

Among religious groups, multi-form participation in religious activities including online religious training, offline group activities, and regular church services also has an impact on happiness (Okun and Nimrod 2020; Yohannes et al. 2008). Forms and contents of religious participation are associated with well-being (Keyes and Reitzes 2007; Koenig et al. 2004). This connection is particularly evident in the religious participation of older adults (Fiori et al. 2006).

For example, in an online questionnaire survey of 103 Nigerians aged 65 and older, the researchers found that the older Nigerians were, the more they engaged in religious activities, and that their level of religious participation had a significant correlation with their well-being, particularly with the overall well-being of those aged 60–70 (Ede et al. 2023). In a study of 377 urban and suburban Protestant residents in Southern Alberta of Canada, researchers found that religious attendance and comfort from religion during times of stress and crisis were the most powerful predictors of well-being (Fry 2000). Using data from the Indian Longitudinal Study of Aging, the researchers analyzed a sample of 31,464 people over the age of 60 and concluded that the association between depressed symptoms and cognitive impairment was considerably reduced by religious activity and belief. Attending religious services may help older adults with depressive symptoms reduce their chances of developing cognitive impairment. People who attend religious services are less likely to have cognitive impairments than older adults who do not attend (Muhammad 2022). Other studies have also concluded that religious participation can promote or enhance individual subjective well-being (Maselko and Kubzansky 2006; Waite and Lehrer 2003).

Therefore, participation in religious activities may play a moderating role in the relationship between different religious affiliations and happiness. When examining the relationship between religious affiliation and happiness, it is necessary to include religious activity as a variable and pay attention to whether and how it affects happiness.

On the basis of the above conclusions, we put out the following hypothesis:

Hypothesis 2.

The relationship between different religious affiliations and older adults’ happiness is moderated by religious activity participation; that is to say, there is a participation effect between different religious affiliations and older adults’ happiness.

The social support theory offers an analytical theoretical framework for examining the relationship between religion and well-being (Krause et al. 2001). Individuals can obtain social support from various dimensions such as emotion, information, and material through interaction with relatives, friends, neighbors, and other social relations, according to this theory (Lin et al. 1979). One of the key reasons religion correlates with happiness, according to this theory, is that believers who share religion and practice have more in common and achieve a high level of social integration. This can be used to form social relationships and gain social support from religious members and groups (Chatters 2000; Sosis 2000). This will help them and make them happier. Religious practice entails the construction of a platform on which religious people can interact and receive emotional or tangible support, which is particularly beneficial for improving well-being as believers are encouraged to maintain hope in the face of difficulties because of the strong social bonds they establish during interaction (Ellison et al. 1989).

For example, the researchers analyzed the relationship between religious belief, social support, and life satisfaction in a sample of 233 older religious Indians in the Americas using a hierarchical regression method. They found that those with higher levels of religious belief and social support reported higher levels of life satisfaction. Further research found that social support played a mediating role between religious belief and life satisfaction, providing a promising way for religion to improve life satisfaction (Roh et al. 2015). In addition, researchers examined the potential mediating role of social support in the relationship between religion and mental health using data from 10,828 adults and Midwestern college students. The results showed that religion-based social support, including social interaction, instrumental support, and emotional support, played an intermediary role in the relationship between religious belief and mental health in the above groups (Hovey et al. 2014; Yeary et al. 2012).

On the basis of the above conclusions, we put out the following hypotheses.

Hypothesis 3.

The association between religious belief and older adults’ happiness is mediated by social support; specifically, the support effect is reflected in how religious belief affects older adults’ happiness.

Religion is a special cultural phenomenon, which is influenced by economic and regional factors in the process of its influence on the ideological consciousness, values, and attitudes of church members. Religious belief in the supernatural and afterlife can bring comfort to people in disadvantaged social classes, improve their subjective socioeconomic status, and affect their psychological security and happiness. Using data from the China General Social Survey, the researchers analyzed a sample of 3258 people and concluded that religion has different effects on the happiness of groups in different subjective economic statuses, and that the religious happiness gap can be explained by the higher self-rated status of believers (Chen and Williams 2016). In addition, using 2018 data from the Taiwan Social Change Survey (TSCS), the researchers analyzed the relationship between religious belief and happiness and found that urban and rural areas play an important moderating role, with respondents living in urban areas more likely to gain happiness from participating in religious activities (Ding et al. 2022).

On the basis of the above conclusions, we put out the following hypotheses.

Hypothesis 4.

The influence of religious belief on older adults’ happiness is heterogeneous, which is caused by different perceptions of subjective economic status and regional factors in urban and rural areas.

3. Methodologies

3.1. Data and Study Sample

The data used in this study are from the China General Social Survey (CGSS), as the counterpart to the U.S. General Social Survey. The CGSS was carried out by several famous Chinese universities in 2003. CGSS adopts multi-stage stratified PPS random sampling technology, and samples are collected according to four levels: district and county, street and town, neighborhood committee and village committee, and individual. The database, which includes information on 28 provinces (including municipalities and autonomous territories), reflects the sociodemographic characteristics, way of life, and social opinions of Chinese citizens. It not only includes the description of the objective characteristics of the interviewees, but also includes feedback on their subjective perceptions. This dataset has a large sample size and a wide distribution area, providing important evidence for exploring the relationship between religious belief and happiness in the Chinese context. In this study, three years of mixed cross-section data of CGSS, 2015, 2017 and 2018, were used for analysis. In this study, the older adult group was the research subject in this study. In order to obtain the population with an age greater than or equal to 60 years, i.e., the older adult population, the method for screening research objects in this study is to first identify the birth year of the individuals in the questionnaire, subtract the birth year of individuals from the corresponding year of data to obtain the age of individuals, and then screen out the population. After removing the missing values of dependent, independent, and other important variables, our working sample consisted of 5177 elderly Chinese individuals.

3.2. Measures

3.2.1. Dependent Variable

The dependent variable in this study is the happiness of older adults. The dependent variable was measured with the question “Overall, do you feel happy with your life?”. The answers to the questions were “1 = very unhappy, 2 = relatively unhappy, 3 = not happy, 4 = relatively happy, and 5 = very happy.” The higher the number, the greater the happiness of the older adults.

3.2.2. Independent Variable

The independent variable in this study is religious belief. The question used to measure the independent variable was “What is your religion?”. The options were “1 = do not believe in any religion”, “11 = Buddhism,” “12 = Taoism”, “13 = folk belief,” “14 = Muslim/Islam,” “15 = Catholicism,” and “16 = Christianity”. In this study, “1 = no religion” was coded as “0 = no religion,” and options 11 to 16 are combined as “1 = religion,” so that religious belief becomes a binary variable. In addition, when classifying the types of religious affiliation, we named “11 = Buddhism,” “12 = Taoism,” “14 = Muslim/Islam,” “15 = Catholicism,” and “16 = Christianity” as institutional religions and “13 = folk belief” as diffused religions according to the views of Yang (Yang 1961).

3.2.3. Moderating and Mediating Variables

The moderating variable is the frequency of participation in religious activities.

Moderators were measured using questions such as "How often do you attend religious services?". The options are “1 = never,” “2 = less than once a year,” “3 = about once a year,” “4 = about once or twice a year,” “5 = about once a month,” “6 = two or three times a month,” “7 = almost every week,” “8 = every week,” and “9 = several times a week.” They are recoded from 0 to 8: “0 = never,” “1 = less than once a year,” “2 = about once a year,” “3 = about once or twice a year,” “4 = about once a month,” “5 = two or three times a month,” “6 = almost every week,” “7 = every week,” and “8 = several times a week.” The higher the value, the more often older adults participate in religious services.

The mediating variable in this study is social support. The mediating variable was measured using the question “Think about the people at your religious place of worship (not including your family) that you contact most frequently. How often do you contact them?”. The options were reverse coded as “0 = I don’t go to any religious places,” “1 = very few,” “2 = once a month,” “3 = 2 to 3 times a month,” “4 = about once a week,” “5 = a few times a week,” and “6 = every day.”

3.2.4. Control Variables

Based on the characteristics of older persons gathered through CGSS surveys, we chose control variables. Age is a continuous variable. Gender is the category variable: male is 1; female is 0. Highest level of education is the category variable: “0 = below primary school,” “1 = primary school,” “2 = middle school,” and “3 = university and above.” Personal income, after removing the singular value, was treated with a logarithm. Marriage was coded as “0 = separation or divorce widowed,” “1 = unmarried,” and “2 = cohabitation or marriage.” Personal socioeconomic status, which ranges from 1 to 10, is transformed into a category variable. Values 1–3 were coded as “1 = low level,” values 4–7 were coded as “2 = middle level,” and values 8–10 were coded as “3 = high level.” The rural account is 0 and the urban account is 1. The number of children is ascertained with the question "How many children do you have?”. The physical size of housing area is determined based on "What is the area of the house you are living in" in the questionnaire, and the housing area is a continuous numerical variable. The status of participating in endowment insurance is 0 if not participating and 1 if participating; the status of participating in health insurance for not participating is 0, for participating is 1.

3.3. Statistical Analysis

We analyzed the effect of religious belief on older adults’ happiness using Stata 16.0. The data distribution of the independent variables, dependent variables, moderating variables, mediating variables, and other control variables was first described using descriptive statistical analysis. Then, the ordered probit model and propensity score matching method were utilized to answer the endogenous problem of the link between religious belief and happiness among older adults. Using an interaction term and stepwise regression mediation, the moderating effect of religious participation and the mediating influence of social support were assessed, respectively.

4. Results

4.1. Descriptive Analysis

Table 1. Descriptive statistics of ordinal/categorical variables.

Variable	Variable Type	Sort Percent
Happiness	Ordinal variable	1/very unhappy 1.22%; 2/relatively unhappy 6.57%; 3/not happy 12.52%; 4/relatively happy 58.32%; 5/very happy 21.38%
Religious belief	Categorical variable	0/unbelief 78.60%; 1/belief 21.40%
Institutional religion	Categorical variable	0/unbelief 58.72%; 1/belief 41.28%
Diffused religion	Categorical variable	0/unbelief 41.28%; 1/belief 58.72%
Frequency of participation in religious activities	Ordinal variable	0/never 79.18%; 1/less than once a year 2.76%; 2/about once or twice a year 4.96%; 3/about once a year 4.54%; 4/about once a month 1.08%; 5/two or three times a month 2.28%; 6/almost every week 1.35%; 7/every week 2.55%; 8/several times a week 1.29%
Social support	Ordinal variable	0/I don’t go to any religious places 80%; 1/very few 11.9%; 2/once a month 4%; 3/2 to 3 times a month 1.6%; 4/about once a week 1%; 5/ a few times a week 0.3%; 6/everyday1.2%
Gender	Categorical variable	0/female 54.03%; 1/male 45.97%
Highest level of education	Ordinal variable	0/below primary school 26.79%; 1/primary school 33.28%; 2/middle school 33.98; 3/university and above 5.95%
Marriage	Categorical variable	0/separation or divorce widowed 26.04%; 1/unmarried 1.14%; 2/cohabitation or marriage 72.82%
Personal socioeconomic status	Categorical variable	1/lower level 39.97%; 2/middle level 56.44%; 3/higher level 3.59%
Registered residence	Categorical variable	0/rural 52.87%; 1/urban 47.13%
Participation in endowment insurance	Categorical variable	0/not participating 19.08%; 1/participating 80.92%
Participation in health insurance	Categorical variable	0/not participating 6.68%; 1/participating 93.32%

and

Table 2. Descriptive statistics of Continuous variables.

Continuous Variable	Mean Value	Standard Deviation	Min	Max
Age	69.427	7.420	60	69
Personal income	7.937	3.570	0	16.013
Number of children	2.543	1.477	0	8
The physical size of housing area	101.632	58.415	5	300

show the descriptive statistics of various variables. Among the happiness variable of the elderly, 1.22% are very unhappy, 6.57% are relatively unhappy, 12.52% are not happy, 58.32% are relatively happy, and 21.38% are very happy. Among the variables of religious belief of the elderly, those who believe in religion account for 21.40% and those who do not account for 78.60%. Among them, 41.28% believed in institutional religion and 58.72% believed in diffused religion. In the variable of the frequency of participation in religious activities, 4.96% are about once or twice a year, and 4.54% are about once a year. In the variable of social support for the elderly, 11.9% are very few, and 4% are once a month. In the variables of individual socioeconomic status of the elderly, 39.97% of them are in the lower level, 56.44% of them are in the middle level, and 3.59% of them are in the higher level. In addition, 45.97% of participants were male, most had middle school education, 72.82% were married, 52.87% were rural, the average number of children in a family was 2.543, and the average housing area was 101.632 square meters. Among the older adults, 80.92% participated in endowment insurance and 93.32% in health insurance.

4.2. Impact of Religious Belief on Happiness

Happiness is the dependent variable of this study, which is measured on a sequential scale from 1 to 5. Therefore, we chose an ordered probit model to explain the relationship between religious belief and happiness. The regression results are presented in

Table 3. Ordered probit model of the influence of religious belief on happiness.

	Happiness
Variables	Model
Religious belief	0.127 ***
	(0.048)
Age	−0.000
	(0.000)
Gender (The control group was female)
Male	−0.052 ***
	(0.019)
Highest level of Education (The control group was below primary school)
Primary school	−0.350
	(0.345)
Middle school	0.132
	(0.111)
University and above	0.045
	(0.157)
Personal income	0.228 ***
	(0.08)
Marriage (The control group was separation or divorce widowed)
Unmarried	−0.075
	(0.427)
Cohabitation or marriage	0.332 **
	(0.166)
Registered residence (The control group was rural)
Urban	0.119
	(0.143)
Number of children	0.132 **
	(0.066)
The physical size of housing area	0.034 ***
	(0.012)
	0.000
Pension (The control group was not participating)
Participating	0.094
	(0.163)
Participating	−0.149
	(0.257)
N	5177
R2	0.193

Note: Unstandardized regression coefficients are displayed, with standard errors within parentheses.** p < 0.01, *** p < 0.001.

.

Table 3 shows the regression results regarding the relationship between religious belief and older adults’ happiness. The regression results indicated that older adults who were religious rated themselves as happier. Specifically, religious believers were 0.127 times more likely to feel very happy than non-religious believers at a significance level of 0.1%, after controlling for all control variables. Gender negatively affected older adults’ happiness at a significance level of 0.1%, and women’s happiness was higher than that of men. Personal income positively affected happiness at a significance level of 0.1%. Those who cohabited or married were happier than those who were divorced or widowed at a significance level of 1%. The greater the number of children, the greater the happiness of older adults was. The larger the physical size of housing area, the greater the happiness of older adults was.

In order to further ensure the scientific results, we carried out a post analysis to test the margins effect.

Table 4. Marginal effect analysis results of variables.

	Marginal Effect of Variables (The Influence of Variables on Happiness Choice Probability)
Happiness	1	2	3	4	5
Religious	−0.0168 ***	−0.0039 ***	0.0133 ***	0.0231 ***	0.0437 ***
	(0.0003)	(0.0009)	(0.0051)	(0.0088)	(0.0111)
Control variables	All control variables	All control variables	All control variables	All control variables	All control variables
N	5177	5177	5177	5177	5177

Note: Unstandardized regression coefficients are displayed, with standard errors within parentheses. *** p < 0.001.

shows the marginal effect analysis results of the influence of religious belief on older adults’ happiness. Religious belief decreases the probability of individual happiness in categories 1–2 and increases the probability of individual happiness in categories 3–5. Specifically, for every 1% increase in religious belief, the probability of choosing category 1 “very unhappy” decreases by 1.68%, that of choosing category 2 “relatively unhappy” decreases by 0.39%, that of choosing category 3 “not happy” increases by 1.33%, that of choosing category 4 “relatively happy” increases by 2.31%, and that of choosing category 5 “very happy” increases by 4.37%.

4.3. Robustness Test

Due to confounding factors influencing the connection between religious belief and happiness in older adults, our results were affected by a sample selection bias. Consequently, propensity score matching was employed to rectify the results of the regression. First, given that the independent variable used for propensity score matching was a dummy variable, the sample was separated into two groups: those with religious belief and those without religious belief. Second, we matched the samples using the techniques outlined below: (1) we transformed the control variables into an indicator and determined the probability that each sample belonged to the experimental group. (2) To match the experimental and control groups, we used five matching methods: 1–1 nearest neighbor, 1–4 nearest neighbor, radius matching, kernel matching, and local linear regression matching, and divided individuals with similar values into the same group to ensure that group members had similar characteristics. (3) We determined the average treatment effect (ATT) of the participants, which measures the difference in happiness between the experimental and control groups.

Table 5. Propensity score matching results (ATT).

Dependent Variable: Happiness
		ATT	SE	t
Nearest Neighbor Matching	1–1 Matching	3.49 ***	0.17	4.30
	1–4 Matching	3.23 ***	0.15	4.25
	Radius Matching	3.37 ***	0.14	4.18
The whole Matching	Kernel Matching	3.43 ***	0.16	4.43
	Local Linear Regression Matching	3.28 ***	0.16	4.32

Note: *** p < 0.001.

displays the ATT calculated using various matching strategies. In all models, the ATT value is more than zero, which is statistically significant, demonstrating that religious belief has a statistically significant beneficial effect on older adults’ happiness.

4.4. Participation Effect: Testing the Moderating Effects of Participation in Religious Activities

This section tested the moderating effect of religious participation frequency on the influence of different religious affiliations on happiness. We selected the sample of religious belief, and we named the religious affiliation variable “institutional religion” as 1, and “diffused religion” as 0. As can be seen from the regression results in Model 2 of

Table 6. Testing the moderating effects of participation in religious activities.

	Happiness	Happiness
Variables	Model 1	Model 2
Frequency of participation in religious activities	0.071	0.325
	(0.124)	(0.234)
Religious affiliation (The control group was diffused religion)
Institutional religion	0.108 ***	0.271 ***
	(0.041)	(0.102)
Interactive items		0.166 ***
		(0.062)
Control variables	All control	All control
	variables	variables
Provinces	Control	Control
Years	Control	Control
Constant	2.976 ***	2.998 ***
	(0.577)	(0.879)
N	1107	1107
R2	0.206	0.315

Note: Unstandardized regression coefficients are displayed, with standard errors within parentheses., *** p < 0.00.

, the interactive item coefficient between religious affiliations and frequency of participation in religious activities is 0.166, which is significant at the level of 0.1%, indicating that the frequency of participation in religious activities had a moderating effect on the relationship between religious affiliations and older adults’ happiness.

Then, we divided the institutional religion affiliation and the diffused religion affiliation into two sub-samples to investigate the participation impact. According to the regression results in

Table 7. Sample test.

Variables	Happiness
	Institutional Religion	Diffused Religion
Frequency of participation in religious activities	0.086 ***	0.078
	(0.032)	(0.272)
Control variables	All control variables	All control variables
N	457	650
R2	0.345	0.047

Note: Unstandardized regression coefficients are displayed, with standard errors within parentheses., *** p < 0.001.

, the moderating effect of participation in religious activities only occurred in institutional religious samples, with a regression coefficient of 0.086, which is significant at the level of 0.1%. That is to say, in institutional religion, the influence of the frequency of participation in religious activities on their happiness shows a corresponding participation effect.

4.5. Support Effect: Testing the Mediating Effects of Social Support

This section adopted the stepwise method (Baron and Kenny 1986) to test the mediating effect of social support, and the results are shown in

Table 8. Testing the mediating effects of social support.

	Model 1	Model 2	Model 3
	Happiness	Social Support	Happiness
Religious belief	0.178 **	0.125 **	0.146 **
	(0.089)	(0.061)	(0.071)
Social support			0.219 **
			(0.098)
Age	−0.000	0.000	−0.000
	(0.000)	(0.000)	(0.000)
Gender (The control group was female)
Male	0.009	0.005	0.008
	(0.097)	(0.061)	(0.094)
The highest level of education (The control group was below primary school)
Primary school	−0.251	−0.157	−0.133
	(0.245)	(0.197)	(0.143)
Middle school	0.066	−0.014	0.136
	(0.071)	(0.124)	(0.142)
University and above	0.143	0.188	0.213
	(0.257)	(0.164)	(0.152)
Personal income	0.032	0.007	0.030
	(0.019)	(0.011)	(0.019)
Marital (The control group was separated, divorced or widowed)
Unmarried	−0.095	−0.251	−0.039
	(0.427)	(0.238)	(0.415)
Cohabitation or marriage	0.134 *	0.178 *	0.146 *
	(0.079)	(0.095)	(0.086)
Registered residence (The control group was rural)
Urban	0.108	0.163*	0.074
	(0.143)	(0.093)	(0.143)
Number of children	0.028	−0.241 **	0.042
	(0.051)	(0.121)	(0.050)
The physical size of housing area	0.002 **	0.001 **	0.002 **
	(0.001)	(0.000)	(0.001)
Pension (The control group was not participating)
Participating	0.094	0.190 **	0.052
	(0.163)	(0.083)	(0.161)
Medical insurance (The control group was not participating)
Participating	−0.149	−0.089	−0.129
	(0.257)	(0.115)	(0.252)
Constant	2.488 ***	1.278 ***	2.201 ***
	(0.577)	(0.331)	(0.583)
N	5177	5177	5177
R2	0.156	0.315	0.323

Note: Unstandardized regression coefficients are displayed, with standard errors within parentheses.* p < 0.05, ** p < 0.01, *** p < 0.001.

. In Model 1, religious belief positively affected older adults’ happiness at a significance level of 1%, and religious belief could improve their level of happiness. In Model 2, religious belief positively affected their social support at a significance level of 1%, indicating that religious belief improved the social support of the older adults. In Model 3, religious belief positively affected their happiness at a significance level of 1%; at the 1% level, social support positively affected their happiness significantly. The influence of religious belief on their happiness changed from 17.8% in Model 1 to 14.6% in Model 3, with a significant decrease, indicating that social support was one of the mechanisms by which religious belief affected older adults’ happiness, and that religious belief can improve older adults’ happiness by improving their social support.

To further test the mediating role of social support on the influence of religious belief on older adults’ happiness, this study used the KHB method, and the results are shown in

Table 9. Mediating effect test results.

Mediating Variables	Effect of Type	Coefficient	Standard Error	Z	p
Social support	Total effect	0.137 **	0.053	2.18	0.035
	Direct effect	0.082 *	0.052	1.68	0.096
	Indirect effect	0.037 **	0.013	2.53	0.017

Note: * p < 0.05, ** p < 0.01.

. In Table 9, the total effect of religious belief on the happiness was 0.137, passing the significance test of 1%. The direct effect was 0.082, passing the 5% significance test. The indirect effect of religious belief on older adults’ happiness through social support was 0.037, which passed the significance test of 1%, demonstrating that social support had a mediating influence. In summary, the stepwise method and KHB test both found that social support played an intermediary role in the relationship between religious belief and older adults’ happiness, and that it was also the mechanism of the relationship between religious belief and happiness in older adults. This finding also verified the supporting effect of religious belief on happiness, as stated in Hypothesis 3.

4.6. Heterogeneity Analysis

In this section, this study first tested whether the interaction term between religious belief and personal socioeconomic status significantly affected older adults’ happiness; then, the samples were grouped according to the individual socioeconomic status and divided into three sub-samples of lower socioeconomic status, middle socioeconomic status, and higher socioeconomic status to investigate the impact of religious belief on happiness. The regression results from

Table 10. Mediating effect test results.

Variables	Happiness
	Model 1	Model 2
Religious belief	0.184 *	0.196 *
	(0.112)	(0.114)
Personal socioeconomic status	0.165	0.098 *
	(0.198)	(0.059)
Interactive items		0.234 **
		(0.119)
Control variables	All control variables	All control variables
N	5177	5177
R2	0.356	0.356

Note: Unstandardized regression coefficients are displayed, with standard errors within parentheses. * p < 0.05, ** p < 0.01.

show that the regression coefficient of the interaction term between religious belief and personal socioeconomic status is 0.234, which is significant at the 1% level.

According to the regression results in

Table 11. Sample test.

Variables	Happiness
	Lower Level SES	Middle Level SES	Higher Level SES
Religious belief	0.765 ***	0.387	0.794
	(0.323)	(0.256)	(0.957)
Control variables	All control variables	All control variables	All control variables
N	2069	2922	186
R2	0.025	0.034	0.009

Note: Unstandardized regression coefficients are displayed, with standard errors within parentheses. *** p < 0.001.

, the influence of religious belief on happiness is only significant in the subsample of low socioeconomic status, with a regression coefficient of 0.765 and a significance at the level of 0.1%.

In addition, older adults’ religious beliefs were influenced by where they lived, and there was a correlation between happiness and their lives. This raises the question of whether there are differences in the relationship between religious belief and happiness among older adults who live in different places. Therefore, this study first tested whether the interaction term between residence and religious belief significantly affected older adults’ happiness, and then divided the rural and urban areas into two sub-samples to investigate the impact of religious belief on happiness. The regression results from

Table 12. Mediating effect test results.

Variables	Happiness
	Model 1	Model 2
Religious belief	0.139 ***	0.187 *
	(0.053)	(0.112)
Registered residence	0.134 ***	0.165 *
	(0.051)	(0.098)
Interactive items		0.166 **
		(0.084)
Control variables	All control variables	All control variables
N	5177	5177
R2	0.363	0.356

Note: Unstandardized regression coefficients are displayed, with standard errors within parentheses. * p < 0.05, ** p < 0.01, *** p < 0.001.

show that the regression coefficient of the interaction term between religious belief and registered residence is 0.166, which is significant at the 1% level.

Table 13. Sample test.

Variables	Happiness
	Rural	Urban
Religious belief	0.156 **	0.077
	(0.074)	(0.091)
Control variables	All control variables	All control variables
N	2737	2440
R2	0.275	0.053

Note: Unstandardized regression coefficients are displayed, with standard errors within parentheses. ** p < 0.01.

shows that religious belief has a positive and significant impact on the rural elderly, and religious belief improves the happiness of the rural elderly. Religious belief has no significant effect on the older adults’ happiness in urban areas.

5. Discussion

As the largest country with religious freedom in the world, China is governed by an atheist party, which is distinct from other countries such as the United States. As a result, studying Chinese religion is critical for comprehending the diversity of world religions. In China, however, empirical research on the impact of religion on happiness is very sparse. Furthermore, there are few studies on the difference of the influence of religion on happiness based on specific religious affiliation, religious participation, and group support in China, as well as specific cultural and economic backgrounds (i.e., registered residence, different socioeconomic status, etc.). In light of this, we explore the influence of religious beliefs on the happiness of the elderly in China, as well as the moderating and mediating mechanisms of religion’s influence on happiness.

5.1. Religious Belief Has a Significantly Positive Affect on Older Adults’ Happiness in China

In this study, we used data from 5177 Chinese samples to examine the association between religious belief and older adults’ happiness. We used the ordered probit model to test this hypothesis and drew a conclusion that Religious Belief has a significantly positive affect on older adults’ happiness in China. This conclusion is completely consistent with the results of a large number of empirical studies in the West (Dehejia et al. 2007; Ellison 1991; Lelkes 2006).

5.2. Religious Participation Plays a Moderating Role in the Relationship between Different Religious Affiliations and Older Adults’ Happiness; That Is, Religious Affiliations Have a Participation Effect on Happiness

In this study, we found that the frequency of participation in religious activities has significantly different moderating effects on the relationship between different religious affiliations and happiness. The frequency of participation in religious activities has a more significant moderating relationship between institutional religion and happiness, but has no significant moderating relationship between diffused religion and happiness. Frequent religious participation allows religious believers to enjoy a wider and denser social network, from which they might derive spiritual solace and psychological security. Participation in religion became a key coping tool for people to deal with stress and danger. Therefore, individuals are more likely to acquire greater amounts of happiness from religious participation (Azzi and Ehrenberg 1975; Myers 2000). For institutional religion, religious activities occur at times that are relatively constant. For example, Islam and Christianity have set times for religious activity. In addition, institutional religions have fixed religious rites. Institutional religion’s organization and control of religious activities are more standardized and stable. Regular involvement in religious activities at established sites may bring psychological solace and a sense of belonging among believers. As more and more time, energy, and money are invested, a believer’s sense of dependence on religion grows, which can alter their happiness.

5.3. Social Support Plays a Mediating Role on the Influence of Religious Belief on Older Adults’ Happiness. The Influence of Religious Belief on Older Adults’ Happiness Reflected the Supporting Effect

We found that social support mediated the relationship between religious belief and happiness; that is, religious belief can improve older adults’ happiness by improving their social support. Therefore, religion, as a special cultural element, can bring psychological support and security to individuals by increasing social interaction with a religious community and gaining more support, thus improving their happiness. This finding was consistent with previous research in the West (Krause et al. 2001; McFadden 1995; Rojpaisarnkit 2016). According to the social support theory, people in a society have access to crucial information resources and it allow for greater integration into the community through the ties they create in religious groups, with positive consequences on both individual and social well-being (Chatters 2000). Through their research, Lever and Martínez (2007) demonstrate that religious societies and belief in God are major sources of social support. In many circumstances, religious societies provide a source of comfort for those who are in difficult situations. People who attend religious communities experience a greater sense of social support, and they believe that being good to others can bring them happiness. Their church offers more assistance than friends and neighbors in the town (Lever and Martínez 2007).

5.4. The Positive Effect of Religious Belief on Older Adults’ Happiness with Lower Socioeconomic Status and in Rural Areas Was More Significant

The influence of religious belief on older adults’ happiness in different groups showed heterogeneity. Among them, religious belief has a more significant impact on the happiness of the elderly with a lower socioeconomic status than those with a middle and higher socioeconomic status. This may be because vulnerable groups generally lack a sense of security and social security, and often feel pressure and risks from various aspects such as income fluctuations and health. At the same time, their daily cultural life and recreational activities are relatively poor. There is a lack of public space and infrastructure to meet the entertainment needs of all vulnerable groups (Trawalter et al. 2021). Religious beliefs provide a platform for communication and the exchange of goods, information, and services (Jadidi et al. 2022). As a result, they are more likely to report higher happiness from religious beliefs. In addition, compared with the urban elderly group, the rural elderly group has fewer channels for social entertainment and more leisure time. When they participate in religious activities, they are more likely to obtain happiness and satisfaction. The findings are consistent with previous studies of older adults in Japan and India (Imamura et al. 2015; Singh et al. 2020).

5.5. Limitations

This study had some limitations. First, this paper focuses on the relationship between religious belief and happiness in the Chinese context. First of all, religious belief actually covers many types and dimensions, which are not the only ones mentioned in this paper. Secondly, the conceptual and measured differences between happiness and subjective well-being also have obvious and larger differences. However, due to limited data and questionnaire questions, this study did not study the impact of more detailed religious belief types on happiness. In addition, more detailed measurements of happiness will be further improved in the future. Second, due to the limits of public survey data, this article cannot explore the psychological process behind their interaction at a more micro-cognitive level, although revealing the association between religious belief and the well-being of older adults at the individual level. In future studies, the relationship between the two can be studied from different dimensions by combining secondary data research supplemented by quantitative methods, such as primary data surveys and experiments.

6. Conclusions

The study results indicated that religious beliefs were significantly and positively correlated with older adults’ happiness, and that religious participation had a moderating role in the participation effect between religious belief and happiness. Additionally, social support played a mediating role in the relationship between religious belief and happiness. Accordingly, we propose the following suggestions: first, the freedom of older adult religious believers to participate in the legal activities of religious organizations should be fully guaranteed. Existing studies have pointed out that religious participation can reduce cognitive impairment in older adults, improve their happiness, and contribute to healthy aging (Hill 2008). In churches and other religious organizations, older adult believers can participate in activities to enhance their interpersonal networks and obtain more resources and information. This social capital can be applied to their secular life, contributing to the improvement of their happiness. Second, in the process of realizing active aging, attention should be paid to the faith of religious older adults. An important reason for older adults to believe in religion may be that it can bring psychological comfort and security to their old age, and the research results also showed that religious belief can improve happiness by improving their social and economic status. Therefore, we should pay attention to respect for the faith of religious older adults and to the difference between religious belief and feudal superstition.