1、Religion and subjective well-being among the elderly in China 1. Introduction Since Easterlins (1974) pioneering analysis of the interplay between aggregate economic growth and the average subjective well-being of a countrys citizens, economists have embraced subjective well-being as an important ec
2、onomic outcome and proxy for individual utility. One prominent line of research has shown that aggregate data on happiness may be used to inform macroeconomic policy. For example, Di Tella et al. (2001) use data from a dozen European countries to infer each countrys subjective preferences for the tr
3、ade-off between unemployment and ination. Helliwell (2006) estimates the social valuation of good and transparent governance, economic stability, and the rule of law. Alesina et al. (2005) and Gruber and Mullainathan (2002) assess the effect of labormarket regulation and cigarette taxes, respectivel
4、y, on collective well-being. At the disaggregated level, economists have long held that revealed preference more accurately represents true well-being than subjective states of mind, yet deducing changes in happiness from observed behavior is often difcult in practice. Although care must be taken in
5、 the use and interpretation of subjective data,Lelkes (2006) and Frey and Stutzer (2002a) note that measures of subjective well-being are reliable measures of “experienced utility,” and the use of subjective data on well-being has recently been embraced by economists. A popular line of empirical inq
6、uiry in the recent research on individual-level well-being has been identifying the determinants of happiness among various population groups. Large-scale surveys conducted in the United States,2 the European Union,3 and 81 countries from across the socioeconomic spectrum demonstrate a considerable
7、degree of consensus: regardless of survey location, robust indicators of subjective well-being include relative income, health status, the strength of social networks, the happiness of friends and relatives, and recent changes in income, marital status, or social networks (Frey and Stutzer,2002b). N
8、otably, men and women are equally likely to report high-levels of happiness or life satisfaction according to surveys of 170,000 adults in 16 countries (Inglehart, 1990) as well as to a meta-analysis of 146 studies (Haring et al., 1984). Because faith communities provide social support for their mem
9、bers and encourage hope in the face of vulnerability (Ellisonet al., 1989), because religiously-active individuals tend to rebound from divorce, unemployment, illness, and bereavement more quickly and more fully (Ellison, 1991), and because religion may foster higher expected utility in the afterlif
10、e (Azzi and Ehrenberg, 1975), participation in religious activities may also inuence subjective well-being. The preponderance of evidence is overwhelming. For example, Myers (2000) uses a survey of 35,000 American adults to show a monotonic positive relationship between the frequency of attendance a
11、t religious services and subjective well-being. Gruber (2005) nds that the effect on self-reported well-being of moving from never attending religious services to attending weekly is comparable to the effect of moving from the bottom income quartile to the top quartile. Swinyard et al. (2001) nd tha
12、t religious participation is among the most deterministic predictors of subjective well-being in Singapore. Indeed,Witter et al. (1985) conduct a meta-analysis of 28 previous studies to nd that religious belief and religious participation account for between 2% and 6% of the variation in adult subje
13、ctive well-being. There is nevertheless some controversy about the interplay of religion and gender in subjective well-being: although Moberg (1965) proposes that religion is a less important determinant of well-being among men than women because of its less central role in the life of men,Witter et
14、 al. (1985) nd no evidence to support this position. In as much as religion serves as a “coping mechanism” for elderly people (Cox and Hammonds, 1988) and because religious capital may accumulate across the lifetime (Iannoccone, 1990), religious participation is likely to be particularly important i
15、n subjective well-being among the aged. Again, the empirical evidence over-whelmingly supports this conjecture. For example, Blazer and Palmore (1976) and Guy (1982) use longitudinal data to demonstrate that the importance of religion in self-reported well-being increases over the life span. In Japa
16、n, religious participation among elderly populations leads to higher subjective well-being (Krause, 2003). Indeed, Okun and Stock (1987) conclude that religion is among the two most important positive inuences on subjective well-being (the other being health) in their meta-analysis of the determinan
17、ts of healthy aging. While the majority of studies report positive relationships, at least two examples of religiosity negatively inuencing subjective well-being have nevertheless been described. First, Gee and Veevers (1990) use data for 6500 Canadians to demonstrate a positive correlation between
18、religious involvement and satisfaction with life. Within the subpopulation of adult men in British Columbia, however, 48.7% of religiously “unafliated” survey respondents report being “very satised” with life, whereas only 38.3% of “” survey respondents report such high levels of satisfaction. Secon
19、d, Willits and Crider (1988) nd that religiosity is positively associated with overall life satisfaction among middle-aged Pennsylvanians. Among men, however, the frequency of church attendance negatively impacts marital satisfaction.Unfortunately, neither study controls for health, demographics, li
20、festyle, and other correlates of well-being that may bias the estimates. Finally, as we describe below, religious participation and subjective well-being may be negatively related in the presence of widespread religious persecution. This paper analyzes the inuence of religiosity on subjective well-b
21、eing among Chinese octogenarians, nonagenarians, and centenarians using a robust multivariate framework that controls for demographics, health and disabilities, living arrangements, wealth and income, lifestyle and social networks, and location. Given the evidence from previous studies, including ev
22、idence from countries that share religious traditions with China,we expected to nd a positive relationship between religious participation and satisfaction with life； however, we nd a robustly negative relationship. Moreover,we nd that religious participation has a stronger inuence on mens well-bein
23、g than on womens well-being. To our knowledge, this is the rst study to uncover such relationships for a large sample while controlling for such a large number of correlates of religiosity.We interpret this nding to be indirect evidence of vulnerability associated with religious persecution, although we cannot test for persecution directly. Section 2 provides an overview of religion and religious participation in China.
