1 Introduction

The goal of this paper is to use the domains-of-life approach to provide insight into the relationship between religious practice and subjective well-being in the United States. The domains-of-life approach is based on a hierarchical understanding of the explanation of life satisfaction: at a first level of analysis, life satisfaction depends on satisfaction in many domains of life; at a second level of analysis, satisfaction in each domain of life depends on many factors and life events, such as: education, income, marriage, illnesses, accidents, divorce, religious practice, and so on (Cummins, 1996; Rojas, 2007). This paper studies the relationship between religious practice and satisfaction in domains of life (second level of analysis), as well as the relationship between satisfaction in domains of life and life satisfaction (first level of analysis). The approach allows mapping out all potential domains-of-life channels connecting religious practice with life satisfaction (Rojas, 2009).

Research on the relationship between religious-related variables and life satisfaction began during the last decades of the past century (Ellison, 1991; Ellison et al., 1989; Gee & Veevers, 1990; Reed, 1991; Lee, 1992, Pargament & Hahn, 1986; Pollner, 1989), and it has grown steadily thanks to the larger availability of information. Research on religious practice provides consistent results: life satisfaction is positively associated with religious practice (Dilmaghani, 2018; Kim-Prieto & Miller, 2018; Lelkes, 2006); different explanations have been advanced for the existence of this positive relationship. Some studies emphasize the role of social networks; it has been shown that religious practice expands the quantity and quality of social networks (Eliassen et al., 2005; Lim & Putnam, 2010) and it has also been shown that social networks are important for life satisfaction; this explanation points to the importance of the friendship domain: religious practice increases satisfaction in the friendship domain, and, through this domain, life satisfaction is raised. Other studies point to the role played by having a direct and personal relationship with a divine other; these studies emphasize spiritual aspects in people’s lives (Ellison, 1991; Ellison et al., 1989), and they suggest that the relationship between religiosity and life satisfaction takes place through the spiritual domain of life. Other channels, such as the capacity of copying with health and economic difficulties, have been mentioned in the literature; they imply that the health and economic domains of life may be relevant channels in the relationship between religious practice and life satisfaction.

Research on the relationship between religious practice and life satisfaction tends to concentrate in the study of a single facet, such as: social networks (friendship domain), spiritual life, or health. By following a domains-of-life approach, this paper studies the different channels at play in the relationship between religious practice and life satisfaction. Hence, it allows knowing which domains of life play a larger role in the contribution of religious practice to life satisfaction.

The empirical study relies on a survey applied in 2018 to the adult non-Hispanic white population in the United States; the survey was implemented to gather information on subjective well-being; financial constraints intervened in the decision to focus on the non-Hispanic white population rather than on the whole population of the United States. The survey is representative at the country level for this segment of the population.

Econometric techniques are used to address the relationship between religious practice and satisfaction in domains of life, as well as to address the relationship between satisfaction in domains of life and life satisfaction. The investigation finds that religious practice is positively associated to life satisfaction, but its significance shows up only for those who do practice their religion markedly. It is shown that the impact of religious practice on life satisfaction takes place throughout many domains of life. The larger impact takes place throughout the economic domain, which is a channel that has not received much attention in the literature. The spiritual and family domains of life do also play an important role; the spiritual channel is always mentioned as a relevant one in the relationship between religious practice and life satisfaction; however, the family domain does not receive similar attention. The paper also shows that the friendship, occupational, and health satisfaction channels do also play a relatively minor but statistically significant role.

The paper is structured as follows: Sect. 2 presents the literature review regarding life satisfaction and the domains-of-life approach, it also discusses the existing literature on the relationship between religious practice and well-being. Section 3 explains the database and the main variables in the study. Section 4 deals with the relationship between religious practice and life satisfaction; it exploits the domains-of-life approach to provide a map of the different channels that connect religious practice and life satisfaction. Some final considerations are made in Sect. 5.

2 Literature Review

2.1 The Domains-of-Life Approach

2.1.1 The Approach

The domains-of-life approach introduces hierarchy in the explanation of the relationship between events and life satisfaction. The first level of analysis explains people’s satisfaction with life as emerging from their satisfaction in different domains of life. People can assess their situation with different facets in their life and can report their satisfaction in these domains, for example, their satisfaction with relations in the family, with health situation, with economic situation, with their job, and with many other domains of life. In this way, satisfaction in domains of life constitutes the substrate of information people use to make an overall assessment of life (Cummins, 1996; Headey et al., 1984; Saris & Ferligoj, 1996; Rojas, 2006). A second level of analysis explains satisfaction in domains of life as emerging from many events and factors in life, such as: income, religiosity, crime, aging, illnesses, and so on (Praag & Ferrer-i-Carbonell, 2004; Rojas, 2007, 2009). For example, illnesses may affect life satisfaction because of their impact on health satisfaction, but also because of their impact on satisfaction with family relations, with leisure activities, and with the personal economic situation (Rojas, 2009). Thus, in order to understand how illnesses end up impacting on life satisfaction, it is useful to know how they relate to satisfaction in all domains of life, as well as to know how satisfaction in these domains of life relate to life satisfaction.

Equations 1–5 formalize the domains-of-life approach:

$$LS_{i} = f\left( {DS_{1i} ,DS_{2i} , \ldots DS_{ni} } \right)\quad i = 1, \ldots p$$
(1)
$$DS_{ki} = g_{k} \left( {X_{i} } \right)\quad k = 1, \ldots n$$
(2)
$$X=({x}_{1},{x}_{2}, . . . , {x}_{m})$$
(3)
$$\frac{\partial LS}{\partial {x}_{j}}=\sum_{k=1}^{n}\frac{\partial {DS}_{k}}{\partial {x}_{j}}\frac{\partial LS}{\partial {DS}_{k}}$$
(4)
$${P}_{dj}= \frac{\frac{\partial {DS}_{d}}{\partial {x}_{j}}\frac{\partial LS}{\partial {DS}_{d}}}{\sum_{k=1}^{n}\frac{\partial {DS}_{k}}{\partial {x}_{j}}\frac{\partial LS}{\partial {DS}_{k}}}$$
(5)

where LSi: life satisfaction of person i. DSki: satisfaction in domain of life k of person i. X: vector of m events and relevant factors. Pdj: Proportion of the total impact of event j on life satisfaction that takes place through domain d.

Equation 1 indicates that satisfaction in a vector of n domains of life constitutes the substrate of information which people use to assess their life satisfaction; it defines the first level of explanation, which takes place between satisfaction in domains of life and life satisfaction. Equation 2 indicates that satisfaction in each domain of life can be understood as emerging from the realization of different events and factors, which are described by the vector X (Eq. 3). Equation 4 provides the information that allows understanding the relationship between a specific event or factor xj and life satisfaction; according to the domains-of-life approach this relationship passes throughout the n domains of life. It is possible to map out the relationship based on the information provided by Eq. 4; the proportion of the total impact of event j on life satisfaction that takes place through domain d can be computed using Eq. 5.

It is important to distinguish the domains-of-life approach from mediating-model frameworks; the domains-of-life approach states that life satisfaction directly depends on satisfaction in domains of life, and not on life events and factors, which, hierarchically, belong to a lower level of explanation. Thus, the impact of these factors and events takes place exclusively through their impact on satisfaction in the domains of life.

2.1.2 The Enumeration and Demarcation of Domains of Life

The enumeration and demarcation of the domains of life is arbitrary; it can go from a small number to an almost infinite recount of all imaginable human activities and spheres of being. Cummins (1996, p. 304) states that “The possible number of domains is large. If each term describing some aspect of the human condition is regarded as separate, then their number is very large indeed.” Rojas (2006, p. 469) argues that there are many possible ways of operationalizing the domains-of-life construct; nevertheless, “any partition must value parsimony -the number of domains must be manageable and domains should refer to clearly separable information-, meaning -the domains of life, as delimited by the researcher, must relate to the way people think about their lives-, and usefulness -the delimitation must contribute to the understanding of the subject-.”

Based on a meta-study of the literature, Cummins (1996) argues for a seven-domain partition: material well-being, health, productivity, intimacy, safety, community, and emotional well-being. Argyle (2002) mentions domains such as money, health, work and employment, social relationships, leisure, housing, and education. Day (1987) considers thirteen areas, among them: family life, working activity, social activity, recreation, personal health, consumption, ownership of durable commodities and properties, self, spiritual life, and country’s situation. Rojas (2006, 2007, 2009) used the following categories in the demarcation of domains of life: family, friends, economic, occupation, leisure, health, and community. Wills (2009) presents evidence on the marginal contribution of the spiritual domain to the explanation of life satisfaction.

Most research regarding domains of life focuses on the study of satisfaction in few or just one domain of life, for example: job satisfaction (Clark & Oswald, 1994). Some scholars work with multiple domains of life to provide an integral view, where life satisfaction is understood as the result of satisfaction in many domains of life (Easterlin & Sawangfa, 2009; Rojas, 2006; van Praag et al., 2003). Rojas (2009) uses the domains of life approach to map out the impact of health-related events on life satisfaction. Bardo (2017), Bardo and Yamashita (2013), and McAdams et al., (2012) study the evolution of the relevance of domains-of-life satisfaction throughout the life course.

2.2 Religious Practice and its Relationship to Subjective Well-being

2.2.1 Religious Practice in Well-being Studies

Religion is a central institution in society; it provides norms and beliefs that influence social and economic life (Inglehart & Norris, 2004; Putnam & Campbell, 2010; Weber, 1966); it also shapes people’s identity. Religion is associated with the inculcation of specific values and beliefs that have an impact in many aspects of life (Inglehart et al., 1992; Lenski, 1963). Thus, religion’s influence on people’s well-being takes place through many facets in people’s lives, facets which have been proven to be relevant to people’s satisfaction with life. For example, religion influences people’s life purposes, their material and non-material aspirations, the nature of their interpersonal relations, and the ways they cope with adversity and change; religion does also promote specific values and attitudes -such as the joy of giving and sharing-, and it fosters virtues -such as gratitude and caring- (Fischer et al., 2010; Jung, 2014).

Subjective well-being (SWB) research shows that it is religious practice, rather than religious affiliation, which matters the most for people’s well-being (Kim-Prieto & Miller, 2018). For example, using a database from Canada, Dilmaghani (2018) finds that the main driver in the happiness and religion relationship is the intensity of religious belief, rather than the specific religious denomination a person professes. Researchers have found out that frequency of churchgoing, which may be an imperfect proxy for religious practice, positively correlates with well-being (Ferris, 2002 -for USA-; Greene & Yoon, 2004, Hayo, 2007, and Lelkes, 2006 -for Europe-; Kortt et al., 2015 -for Australia-; Elliott & Hayward, 2009 -worldwide-).

2.2.2 Religious Practice and its Association to Satisfaction in Domains of Life

Because life satisfaction emerges out of satisfaction in many domains of life, it is worthwhile to review the literature that associates religious practice with domains of life such as: spiritual, friendship, health, economic, family, and others.

The religious experience is closely related to the spiritual domain of life, and satisfaction in this domain may contribute to life satisfaction. James (1982) provides in-depth considerations on what the religious experience is about. Satisfaction in the spiritual domain may depend on many factors associated to the religious experience; some authors point to the role of religious-founded afterlife promises (Azzi & Ehrenberg, 1975), as well as to religion providing a framework which gives coherence, harmony, and certainty to human life (Ellison, 1991; Ellison et al., 1989). Meaning of life and transcendent values are promoted by religious practice, and they may contribute to well-being (Headey et al., 2010). The spiritual domain is also associated to the experiencing of pinnacle moments and flow states (Manglos, 2013). The benefits of having a direct and personal relationship with a divine other are underlined by some researchers (Bulman & Wortman, 1977; Ellison, 1991; Pargament & Hahn, 1986; Pollner, 1989). This personal relationship with a divine other may help in confronting difficulties in life, as well as in accepting and dealing with them; it may also help in the conveying of responsibility regarding vital decisions (Gorsuch & Smith, 1983; Spilka & Schmidt, 1983). Hence, it is reasonable to assume that the spiritual domain of life plays a central role in the relationship between religious practice and life satisfaction. Religious practice may contribute to increase spiritual satisfaction, and spiritual satisfaction contributes to life satisfaction.

Researchers have also stressed the importance that social networks and social participation play in explaining the positive relationship between religious practice and life satisfaction. Evidence on how religious practice contributes to the creation of social networks in the congregation is vast (Eliassen et al., 2005; Krause et al., 1999; Krause & Wulff, 2005; Lim & Putnam, 2010; Shor & Roelfs, 2013). Interacting with people who share similar values and beliefs may directly be a source of SWB, and it may expand social ties and social support (Ellison, 1991; Lee, 1992). This literature emphasizes the impact of religious practice in the relational community (friendship) domain of life; thus, the literature points to the central role of the friendship domain in the relationship between religious practice and life satisfaction. Religious practice contributes to greater friendship satisfaction, and greater friendship satisfaction contributes to life satisfaction.

Religious practice may also help people in coping with adversity (Headey et al., 2010), as well as in developing an optimistic attitude (Van Cappellen et al., 2016). These skills are particularly important in coping with health-related problems (Idler, 2004; Pargament & Ano, 2004). Hence, the literature suggests that some of the well-being benefits from religious practice may take place through the health domain of life; it would be expected for highly religious persons to report greater health satisfaction -after controlling by other socio-demographic and economic factors-, and, in turn, greater health satisfaction contributes to greater life satisfaction.

The relationship between religion and economic factors has attracted some attention, usually in the economic development literature. At the microeconomic level, Lelkes (2006) shows that the relationship between income and SWB is less important for the religious than for the non-religious. Economic values may be modified by religious factors (Colvin & McCracken, 2017); in fact, it is widely accepted that religion may influence the importance a person attaches to worldly possessions (Pace, 2012) and, in general, the materialistic values a person holds (LaBarbera & Gurhan, 1997; Swinyard et al., 2001). If religious practice is accompanied by a decline in materialistic values and in the importance of worldly possessions then, for a given income, greater economic satisfaction would be expected as religious practice raises. Hence, the economic domain may play a role in the association between religious practice and life satisfaction; greater religious practice contributes to greater economic satisfaction and greater economic satisfaction contributes to more life satisfaction.

The literature on religion and family relationships is vast; however, there is not much research on its impact on family satisfaction. Religion may influence the nature and quality of interpersonal relations at the family level (Mahoney, 2010; Mahoney et al., 2001). The stability of marriages, the quality of parenting practices, the attitude in coping with marriage problems, the strength of filial bonds, and many other behaviors which are important for satisfaction in the family domain of life could be associated with religious practice (Pearce & Axinn, 1998; Vermeer, 2010, 2014). Hence, even though there is not much research on religious practice and family satisfaction; it may be expected for religious practice to positively contribute to family satisfaction, while it is a well-established fact that the family domain is very relevant for life satisfaction.

Research on the role of religion in the workplace is limited (Benefiel et al., 2014); and, to the best of the authors’ knowledge, there is no research on the association of religious practice to job satisfaction.

2.2.3 An Integrated View

The study of the relationship between religious practice and SWB has commonly focused on specific domains, with few studies taking into consideration many domains of life (Gee & Veevers, 1990; Reed, 1991). The domains-of-life approach can contribute to the literature by providing an integrated view on how religious practice translates into greater life satisfaction throughout the different domains of life.

Some causality issues arise when studying the relationship between religious practice and SWB. Religious practice may lead to greater SWB, but it is also possible for happier people to practice their religion with greater intensity. For example, healthy people may have greater opportunities to practice their religion; stronger families may be more inclined to incorporate religion into their daily life, and non-materialistic people may find some reassurance in the religious message. Thus, a directional causality from religious practice to SWB is reasonable, but a bidirectional relationship cannot be rule out.

2.3 Main Research Question

Based on the literature review, religious practice is expected to be positively associated to life satisfaction; this relationship is expected to take place because: First, religious practice is positively associated to satisfaction in many domains of life and, second, satisfaction in these domains of life is positively associated to life satisfaction. However, the specific association between religious practice and satisfaction in domains of life is not clear, while the importance of the different domains of life in explaining life satisfaction does also require further study. Hence, this paper addresses the following two questions: First, what is the association between religious practice and satisfaction in an integrated set of domains of life? Following Rojas (2006, 2007), this paper takes into consideration the following seven domains of life: Family, Friends, Economic, Free time, Health, Occupation, and Spiritual. The literature has emphasized the spiritual and the social-network (friendship) channels; but other channels should not be ruled out. Second, what domains of life are more important in explaining life satisfaction? By addressing these questions, it becomes possible to use a domains-of-life approach to map out the relationship between religious practice and life satisfaction.

3 Information

3.1 The Survey

A survey was implemented in the United States (USA) during the first semester of 2018; its objective was to gather subjective well-being information for a representative and relatively homogeneous segment of the adult USA population. A well-established polling company took care of the survey, applying all monitoring procedures to ensure the quality of the information and its representativeness for the population segment of interest. Due to financial constraints, as well as to the desire of having a representative survey from a culturally homogeneous segment of the adult USA population, the population of the study was defined as the adult non-Hispanic white population in the USA. In total, 1091 questionnaires were applied.

3.2 Main Variables

3.2.1 Religious Practice

A general self-reported question on religious practice is asked: “How do you consider yourself about your religion?”; the response options where: Very practicing, Practicing, Little practicing, Non-practicing. A Non-religious category is also incorporated based on information from a previous question in the survey regarding religious affiliation. This variable is ordinal; it is unclear whether all ‘steps’ are equally important; this is, whether a change from non-practicing to little practicing is similar to a change from practicing to very practicing. It is also unclear where the non-religious category should be positioned, at the same level than the non-practicing category? Below the non-practicing category? Hence, to add flexibility and to allow for ‘steps’ being of different magnitude, the religious-practice variable is handled as a categorical variable, and a vector of dichotomous variables is generated; these variables have a value of 1 if the person reports a practicing category and 0 otherwise. When dealing with religious practice it is frequent to gather information on concrete practices, such as: church attendance and time spent in parish activities; however, as Dargent (2019) shows, these specific practices tend to miss the whole perspective on people’s religious involvement and commitment: thus, Dargent recommends working with self-reported information on religious practicing, which provides useful data on people’s religiosity. This approach is also followed by Eliassen et al. (2005), and questions such as ‘How important your religion is to you?’ are not rare in the literature (Dilmaghani, 2018).

3.2.2 Life Satisfaction

Based on the following question: “All things considered, how satisfied are you with your life as a whole in the present?” The response scale is assumed to be categorical; response categories are: Extremely dissatisfied, Very dissatisfied, Somewhat dissatisfied, Neither satisfied nor dissatisfied, Somewhat satisfied, Very satisfied, Extremely satisfied. The life-satisfaction variable is a dependent one; thus, for simplicity of analyses, the response scale is treated as cardinal in a 1 (Extremely dissatisfied) to 7 (Extremely satisfied) range. A well-known finding in the subjective well-being literature states that the handling of happiness variables as cardinal, instead of ordinal, does not imply important changes in the main results (Ferrer-i-Carbonell & Frijters, 2004).Footnote 1

3.2.3 Satisfaction in Domains of Life

Based on the following question: “How satisfied are you with the following aspects of your life?”, the following facets are considered: Relationship with partner, Relationship with children, Relationship with Parents, Relationship with other members of the family, Relationship with friends, Economic situation in your home, Availability of free time, Your health, Your job or the activity you perform, and Your spiritual life. The response scale is similar to the one used for the life-satisfaction question. The high correlation among satisfaction in the relationship with partner, children, parents, and other family members, as well as the existence of many missing values in some of these subdomains, makes it convenient to construct a Family domain computed as a simple average of the available information for the four questions: partner, children, parents, and other members of the family. Thus, this paper works with the following seven domains of life: Family, Friends, Economic, Free time, Health, Occupation, and Spiritual.Footnote 2

3.2.4 Socio-economic and Demographic Variables

Satisfaction in domains of life may depend on many other variables besides the religious one (Frey, 2018; Leyva et al., 2016); thus, the survey also gathered information regarding the following variables, which can be used as control ones: gender, age, marital status, education level, self-reported health, and household per capita income.

4 A Domains-of-Life Approach to the Relationship Between Religious Practice and Well-being

4.1 The Well-Being Variables

Table 1 presents bivariate correlation coefficients for the main well-being variables. The study is interested in how religious practice is associated to life satisfaction through the seven domains of life; thus, it is required to have information for all variables under consideration; this requirement is satisfied for 767 observations in the sample.

Table 1 Life satisfaction and satisfaction in domains of life bivariate correlations
Full size table

It is observed that life satisfaction is positively correlated with satisfaction in all domains of life, with correlation coefficients being relatively high for the economic domain (0.60) and relatively low for the free time domain (0.36). The correlation coefficients for satisfaction between different domains of life are also positive, ranging from 0.54 for the correlation between family satisfaction and friends satisfaction to 0.30 for the correlation between friends satisfaction and health satisfaction.

Table 2 provides descriptive information regarding satisfaction in domains of life by religious practice; this information is useful in grasping the subjective well-being situation in the population under study.

Table 2 Life satisfaction and satisfaction in domains of life mean values by religious practice
Full size table

The information presented in Table 2 allows for the following comments:

First, life satisfaction is clearly higher for the more religious; mean life satisfaction is 5.04 for the ‘non-practicing’ and 4.97 for the ‘non-religious, and it is 5.74 for the ‘very practicing’.

Second, regarding the whole population, high satisfaction values are observed in the Family (5.61) and Friends (5.55) domains of life. They are followed by the Spiritual (5.37) and Free Time (5.21) domains. Low satisfaction values are observed in the Economic (4.65) and Health (5.04) domains.

Third, about three quarters of the population report some religious affiliation; they are distributed as follows: 18 percent report being ‘very practicing’, 25 percent ‘practicing’, 25 percent ‘little practicing’, and 9 percent are ‘non-practicing’. 23 percent of the population report being ‘non-religious’.

Fourth, satisfaction in domains of life is higher for the more religious people; this effect is general, but not symmetrical, across all domains of life. It is observed that the sensibility is greater for the spiritual domain, with an increase in mean values from 4.81 (for the non-practicing) to 6.23 (for the very practicing).

Fifth, as expected, the ‘very practicing’ report outstandingly high satisfaction levels in their Spiritual domain; their satisfaction is also very high in the Family and Friends domains of life.

Sixth, the non-religious persons have a domains-of-life satisfaction pattern which resembles that of non-practicing religious persons. It is observed that they report, on average, greater Spiritual satisfaction than that of the ‘non-practicing’ and the ‘little practicing’.

It is important to remark that the association between religious practice and satisfaction in domains of life does not necessarily imply direct causality. In some cases, it is possible to imagine reverse causality; for example, people with poor health could report lower health satisfaction and could also face more difficulties to practice their religion.

4.2 Religious Practice and Satisfaction in Domains of Life

Equation 6 is used to study the relationship between religious practice and satisfaction in the seven domains of life under consideration; religious practice is treated as a vector of dummy variables -rather than as an ordered variable- to introduce flexibility in the relationship. The quantitative exercise controls by a group of socio-demographic and economic variables, which have recurrently been shown to be relevant in the explanation of satisfaction in domains of life (Frey, 2018; Leyva et al., 2016). Ordinary least square (OLS) regressions are run to estimate the parameters for each equation j.

$$DS_{ji} = \beta_{j0} + \sum\nolimits_{r = 1}^{3} {\beta_{jr} RP_{ri} + } \beta_{j4} NR_{i} + \gamma_{j} Z_{i} + \varepsilon_{ji} \quad j = 1, \ldots 7$$
(6)

where DSji: Satisfaction in domain j reported by person i; j = 1,.., 7 (Family, Friends, Economic, Free time, Health, Occupation, Spiritual). RPri: Vector of three dichotomous variables for religious practice category reported by person i; r = 1, 2, 3 corresponds to the ‘very practicing’, ‘practicing’, and ‘little practicing’ categories; the reference category is ‘non-practicing’. NRi: Dichotomous variable with value of 1 for non-religious person, 0 otherwise. Zi: Vector of control variables; Gender, age, age squared, reported health, marital status (single, married, stable partner, divorced/separated, and widowed), education level, and logarithm of household per capita income. εji: error term.

Table 3 presents the estimated coefficients from the econometric exercises. A parallel exercise with the religious variable as the only independent one is run, and the estimated R squared coefficients are also presented in Table 3.

Table 3 Religious practice and satisfaction in domains of life OLS technique
Full size table

The following comments are pertinent based on these findings:

First, ‘very practicing’ religion is positively associated to satisfaction in all domains of life under consideration; thus, its impact is not restrained to the Spiritual domain nor to the friendship domain. ‘Very practicing’ religion is associated to a higher level of Spiritual satisfaction (about 1.4 higher)—with respect to the ‘non-practicing’ category-. ‘Very practicing’ is also associated to higher Economic satisfaction (about 0.7 higher). In addition, ‘very practicing’ religion is also associated to higher Occupational satisfaction, Family satisfaction, Friends satisfaction and, slightly but not statistically significant, to Health satisfaction. Hence, it seems that there are many connecting routes between religious practice and well-being; these routes go beyond the frequently emphasized channels of social networks and spiritual aspects.

Second, those who report ‘practicing’ their religion do also tend to report greater satisfaction in the Spiritual and Occupational domains -with respect to the ‘non-practicing’ category-. The higher Spiritual satisfaction was expected, while the higher satisfaction in the Occupation domain points towards the need of studying the role religiosity plays in people’s working activities.

Third, there are no statistically significant differences in domains-of-life satisfaction between the ‘little practicing’ and the ‘non-practicing’. This corroborates a general finding in the literature that shows that, from a well-being perspective, it is not a matter of having a religious affiliation, but of seriously practicing your religion. It is also observed that, with the surprising finding for the Spiritual domain, the non-religious do tend to report domain-of-life satisfactions which are more or less similar to those reported by the non-practicing religious persons.

Fourth, Table 3 also presents the explanatory power of a regression exercise where satisfaction in domains of life is explained by the religious practice variable alone. It is observed that the explanatory power (R squared coefficient) of the religious practice variable is generally low, except for the Family domain, where the religious variable explains about 6 percent of the variability in family satisfaction, and, as expected, the Spiritual domain, where the variable explains about 14 percent of the variability in spiritual satisfaction.

Fifth, the exercises presented in Table 3 control by a group of socio-demographic and economic variables, including the logarithm of income. The estimated coefficient for the logarithm of household per capita income can be used as reference to assess the relative magnitude of the estimated religious practice coefficients. It is observed that income is not significant in explaining satisfaction in the following domains: Free time, Health, Occupation, and Spiritual. Income is relevant in explaining satisfaction in the Family, Friends and Economic domains; in all cases the estimated coefficient for the income variable is smaller than that for the ‘very practicing’ variable. This means that the higher satisfaction levels of those very practicing their religious can be considered as substantial, at least in comparison to what a 100 percent increase in income would contribute to.

4.3 Satisfaction in Domains of Life and Life Satisfaction

Not all domains of life are equally important in explaining life satisfaction; a regression analysis is used to estimate the importance of each domain of life in explaining life satisfaction. However, before running the quantitative exercise, it is important to recognize that a regression between life satisfaction and satisfaction in domains of life exclusively involves subjective measures as independent and dependent variables; this exposes the quantitative analysis to a potential common-method bias (Podsakoff et al., 2003). There could be a person-level factor -such as a personality trait- which, being common to all variables in the analysis, biases the estimated parameters in cross-section analyses. Hence, to address this potential bias a ‘trait’ exercise is run (van Praag & Ferrer-i-Carbonell, 2004); the exercise relies on those so-called objective variables in the vector of control variables used in Eq. 6 to observe the existence of a systematic pattern in responding to subjective questions. Equation system (7) is run:

$$DS_{ji} = \beta^{\prime}_{j0} + \gamma^{\prime}_{j} Obj_{i} + \varepsilon^{\prime}_{ji} \quad j = 1, \ldots 7$$
(7)

where Obj refers to the vector of objective variables: gender, age, age squared, marital status, educational level, and logarithm of income. The vector of seven residuals obtained for each person i is used to observe the existence of a systematic pattern in responding to subjective questions, and it may be considered as person i’s trait. Principal component analysis is used to construct a variable based on the seven residuals for each person i; this variable provides a trait value for each person, and it can be used as a control variable in the study of the relationship between life satisfaction and satisfaction in domains of life.Footnote 3

In consequence, the study of the relationship between life satisfaction and satisfaction in domains of life relies on Eq. 8, which, besides the life satisfaction variables, does also include the Trait variable:

$${LS}_{i}={\propto }_{0}+\sum_{j=1}^{7}{\propto }_{j}{DS}_{ji}+\rho {Trait}_{i}+{\mu }_{i}$$
(8)

where LSi: Life satisfaction reported by person i. Traiti: variable portraying person i’s trait, constructed from exercise presented in Eq. 7. µi: error term. All other variables already defined.

Table 4 presents the estimated coefficients for different models based on Eq. 8. Model 1 presents the standard domains-of-life analysis which corresponds to Eq. 1; Model 2 recognizes that all variables in the analysis are subjective and, in consequence, controls by a person-level trait variable. Models 3 and 4 incorporate the religious variable, with and without the trait variable, into the analysis.

Table 4 Life satisfaction and satisfaction in domains of life OLS technique, different models
Full size table

The Trait variable happens to be statistically significant; it corrects for any potential common-source bias arising from the fact that both the dependent and independent variables are reported by the same person; thus, Model 2 is preferred than Model 1. The following comments are based on the estimated coefficients for Model 2. It is important to state that the statistically significant correlations among satisfaction in domains of life do introduce multicollinearity in our analyses, with the consequence of increasing the standard errors of the estimated coefficients. Thus, the fact that most estimated coefficients for the domains of life are statistically significant, even when the standard errors may be inflated, show that the association between satisfaction in domains of life and life satisfaction is solid.

It is observed that the Economic domain has the largest estimated coefficient (0.30), which means that economic issues are very relevant for the life satisfaction of the population under study; the Family and the Health domains do also have relatively large coefficients (of 0.24 and 0.22, respectively). The estimated parameter for the Spiritual domain is statistically significant, and its magnitude is similar than that for the Friends and Occupation domains of life; hence, spiritual life matters for life satisfaction, and it matters as much as those other domains. The Free-time domain is not relevant in explaining life satisfaction for the population under study.

The goodness of fit obtained in Model 2 can be considered as relatively high, with the domains of life under consideration, plus the trait variable, explaining about 52 percent of the variability in life satisfaction. As it was stated in the literature review, the enumeration and demarcation of the domains of life is not a closed issue; based on the literature, it would be possible to think about other domains of life -such as satisfaction with sexual life, and satisfaction with neighborhood and place of living- which could increase the explanatory power of the model.

It is observed that the direct incorporation of the religious variable into the analysis adds little explanatory power to our understanding of life satisfaction (R-squared of 0.497 in Model 3 vs. 0.492 in Model 1; and R-squared of 0.526 in Model 4 vs. 0.518 in Model 2). This finding is consistent with the domains-of-life approach (as presented in Eqs. 1–5), which establishes two hierarchical levels: first, the relationship between life satisfaction and satisfaction in domains of life (as portrait in Models 1 and 2); second, the relationship between events and factors (such as religious practice, income, and others) and satisfaction in domains of life (as presented in Table 3). As Models 3 and 4 show, these events and factors play a role in the lower hierarchical level, which explains satisfaction in domains of life, but not in the higher level, which relates life satisfaction with satisfaction in domains of life alone.

4.4 Mapping out the Religious-Practice and Life Satisfaction Association

Equation (5) allows mapping out the computed impact of religious practice category r on life satisfaction; this computation is based on the estimated parameters from Eqs. 6 and 8; thus:

$$\frac{\delta LS}{\delta {RP}_{r}}=\sum_{j=1}^{7}{\alpha }_{j}*{\beta }_{jr}$$
(9)

Table 5 relies on the estimated coefficients presented in Table 3 (Model 1) and Table 4 (Model 2) to compute the product \({\propto }_{j}*{\beta }_{jr}\) for each domain of life j and for r corresponding to the ‘very practicing’ category. For example, Table 3 shows that ‘very practicing’ religion is associated to a Family satisfaction which is 0.55 higher -in comparison to ‘non-practicing’ religion-, while Table 4 shows that a one-step raise in Family satisfaction increases life satisfaction in 0.24. By combining this information, it is possible to know the magnitude of association between ‘very practicing’ religion and life satisfaction which takes place through the family domain: 0.55 times 0.24, which is equal to 0.13. By implementing Eq. 9, it is possible to have the estimated role of ‘very practicing’ religion on life satisfaction as well as to map out it in terms of domains of life. Table 5 presents the computed values for all the domains of life under consideration when religious practicing corresponds to ‘very practicing’ rather than to ‘non-practicing’.

Table 5 Influence of religious practice on life satisfactiona apportionment by domains of life
Full size table

It is observed in Table 5 that the estimated life satisfaction which is associated to ‘very practicing’ religion is 0.76 higher than that for ‘non-practicing’ religion. This higher value takes place through almost all domains of life, apart from the Free time domain, which is not relevant in explaining life satisfaction in the population of study (see Table 4).

The Economic domain constitutes the main route associating religious practice and life satisfaction; this happens because the Economic domain is very important in explaining the life satisfaction of this population (0.303, see Table 4-Model 2), and because the estimated coefficient for the very religious in the Economic domain is very large (0.698, see Table 3-Model 1).

The Spiritual and the Family domains of life do also constitute major routes in the association between religious practice and life satisfaction. The Spiritual domain has a relatively minor role in explaining life satisfaction, but it is strongly associated to religious practice. On the other hand, the Family domain has a relatively important role in explaining life satisfaction, and religious practice happens to be positively associated to Family satisfaction.

The Economic, Spiritual and Family domains account for about 71 percent of the association between religious practice and life satisfaction. The Friendship domain accounts for about 14 percent, while the Health and Occupation domains account for 8 and 7 percent, respectively.

5 Discussion

5.1 Conclusion

Life satisfaction is positively associated to religious practice; the impact is large and statistically significant for the very religious -those who report ‘very practicing’ their religion -in comparison to those who do not practice their religion and to those who are not religious-. This paper provided insight into this relationship by studying the different domains-of-life routes that intervene in this association. The paper followed a domains-of-life approach to provide a broader view of how religious practice ends up relating with people’s life satisfaction.

The previous literature has emphasized two facets: First, the spiritual one associated to the benefits of having a direct and personal relationship with a divine other (Bulman & Wortman, 1977; Ellison, 1991; Pargament & Hahn, 1986; Pollner, 1989) as well as the related benefits associated to dealing with the afterlife (Azzi & Ehrenberg, 1975) and having a transcendent and meaningful life (Headey et al., 2010). The findings presented in this paper confirm that the Spiritual domain is a relevant one; the domain is not highly important in explaining life satisfaction, but it is strongly associated to religious practice.

The literature has also pointed to the role of social networks as an important channel linking religious practice with life satisfaction. The religious congregation may constitute a friendship network that provides support and gratifying ties (Eliassen et al., 2005; Lim & Putnam, 2010; Ellison, 1991; among many). The findings from this paper do not fully support this argument; the Friendship domain is important, but it is not one of the three most important domains in the relationship. It is true that religious practice is strongly associated to friendship satisfaction, but the relevance of the Friends domain in explaining life satisfaction is not high.

The main findings from this research point to the importance of the economic and family domains of life, besides the spiritual one. When the relationship between ‘very practicing’ religion and life satisfaction is mapped out, it becomes evident that the economic domain plays the larger role. Those who ‘very practice’ their religion do report substantially higher economic satisfaction -after controlling by income and other socio-demographic variables-; in addition, the economic domain happens to be highly relevant in explaining the life satisfaction of the population under study. Hence, the results from this paper suggest that greater attention must be placed to the role that religious practice plays in economic values (Colvin & McCracken, 2017; Lelkes, 2006; Pace, 2012, and others), as well as in the exploration of other economic-related factors, such as: consumption patterns, savings attitudes, and social comparisons. The importance of the family domain is also highlighted by this paper’s findings; the very practicing do report substantially higher family satisfaction -with respect to the non-practicing and the non-religious-, and the family domain is highly relevant in explaining life satisfaction. The role of religion in influencing the quality of intra-family relations has been mentioned by some researchers (Mahoney, 2010; Mahoney et al., 2001; Vermeer, 2010, 2014, among others); but there may be many family-related aspects that require further research, such as the relationship between religious practice and nurturing practices, as well as between religious practice and some behaviors that threat family satisfaction, such as drug addiction, infidelity, and others.

The paper has also shown the value of following a domains-of-life approach to have an integrated view of the relationship between life satisfaction and life events and factors.

5.2 Final Considerations

This research project has faced some weaknesses which constitute opportunities for future research:

First, it is important to recognize that causality concerns can always be raised, and that cross-section data is limited to address issues of antecedence, which not necessarily imply causality. This paper has assumed that satisfaction in domains of life explains life satisfaction, and, in fact, there is evidence supporting this view (Rojas, 2009; Rojas & Elizondo-Lara, 2012); however, not all scholars agree with a bottom-up understanding of the relationship between domains-of-life satisfaction and life satisfaction. Headey et al. (1991) defend a top-down understanding of the relationship. Further study of causality issues may be necessary. Unfortunately, the study of causality is always a difficult issue in social science; the data shows association and antecedence, and, in the end, it is up to scholars to rely on theoretical and conceptual models to attribute causation.

Second, financial constraints limited this research project to the study of religious practice and life satisfaction in the adult non-Hispanic white population of the United States. Hence, findings apply only to this segment of the USA population, and it is unclear whether they sustain when extrapolated to other population segments. The comparison of findings across different population segments, as well as across countries may provide further insight on the role of religiosity.

Third, the measurement of satisfaction in domains of life has relied on single and general questions, which provide useful information but little texture to go deeper into the understanding of the relationship between religious practice and satisfaction in domains of life. Further studies could gather information at the sub-domain level, which would be useful to fully grasp the role that each domain of life plays in the relationship between religious practice and life satisfaction.

Fourth, this study relied on an overall subjective variable to measure religious practice. There may be some debate about the information that this variable provides. The SWB tradition has stressed the importance of using subjective variables; it states that lists of so-called objective variables cannot fully grasp well-being as it is experienced by people. Some traditions in the study of religious practice do rely on lists of concrete religious habits -such as frequency in attending services and in praying; it is unclear whether these lists can fully comprehend the religious habits people care about, as well as the meanings people assigns to them. The study of the relationship between concrete religious habits and the subjective religious-practice assessment may constitute a relevant area for future research.