J Gambl Stud DOI 10.1007/s10899-015-9532-3 ORIGINAL PAPER

Religious Background and Gambling Among Young Adults in the United States Jeremy E. Uecker • Charles E. Stokes

 Springer Science+Business Media New York 2015

Abstract Despite the rapid growth of the gambling industry over the last 40 years, there have been few large-scale, nationally representative longitudinal studies of gambling among young adults. We use data from the National Longitudinal Study of Adolescent to Adult Health to investigate whether and how the gambling behavior of young adults is associated with their religious beliefs and practices during adolescence. We find that young adults who grew up as conservative Protestants, mainline Protestants, Mormons, and Jehovah’s Witnesses; those were raised in a community with a higher percentage of conservative Protestants; and those who attended religious services weekly are less likely to have ever gambled. Among gamblers, young adults who attended religious services up to three times per month as adolescents are more likely to experience gambling problems than those who never attend. Notably, accounting for a young adult’s propensity for risk-taking behavior does not explain the associations between religion and gambling. Keywords gambling

Religion  Young adults  Longitudinal analysis  Gambling  Problem

Introduction The social factors associated with gambling and problem gambling have received relatively little attention compared with the psychological correlates or the economic costs

Information on how to obtain the Add Health data files is available on the Add Health website (http://www. cpc.unc.edu/addhealth). J. E. Uecker (&) Department of Sociology, Baylor University, One Bear Place #97326, Waco, TX 76798-7326, USA e-mail: [email protected] C. E. Stokes Samford University, Birmingham, AL, USA

123

J Gambl Stud

of gambling. Studies on gambling behavior among young adults, the part of the life course where risk-taking behavior is highest (Arnett 1992), are especially scarce (for exceptions, see Welte et al. 2008; 2009). Most of the published research on American young adults and gambling uses nonrandom samples of college students (e.g., LaBrie et al. 2003; Lightsey and Hulsey 2002; Browne and Brown 2001). Perhaps this lack of attention should not be surprising given the rapid rise in the legality, availability, and popularity of gambling; there has simply not been enough time for the gambling behaviors of young adults to have attracted the attention of large studies. Judging from the public discourse on gambling, religion may be a key factor in determining gambling behavior among young adults. Although for most young adults religious priorities and pursuits decline from earlier childhood levels (Uecker et al. 2007), religious influences may still be very strongly associated with the gambling behavior of young adults for three reasons. First, religion is still important for a large number of young adults. Second, religious influences during childhood and adolescence may still influence the thinking and behavior even of non-observant young adults. Third, religious organizations are among the few social organizations that present clear messages about gambling behavior, a factor which could certainly influence adherents but may also precipitate ecological effects on gambling. While multiple dimensions of religion may help to explain involvement in gambling and problem gambling, studies examining these possibilities are few and are often hampered by data limitations. In this study we use data from the first and third waves of the National Longitudinal Study of Adolescent to Adult Health (Add Health), a nationally-representative sample of young adults who were enrolled in school as adolescents, to conduct an investigation of the associations between religion and gambling behavior among young adults. Though some minors are known to gamble (Stinchfield and Winters 1998), gambling typically becomes legal at age 18. Thus, our data capture gambling among its youngest legal participants, and we are able to establish time order by measuring religious characteristics during adolescence. Moreover, young adults are a particularly salient group to study because they are the most likely to gamble and to be ‘‘at-risk’’ for problem gambling (Mok and Hraba 1991; Gerstein et al. 1999). Utilizing the richness of the Add Health data, we are able to improve upon previous studies of religion and gambling in several ways. First, we are able to look at the impact of several dimensions of young adults’ religious experience on gambling, including aspects such as religious characteristics of the community. Second, we are able to include a rich array of controls, including measures of young adults’ propensity for risky behavior. Third, we make use of longitudinal data that can measure religion prior to participation in gambling, thereby avoiding issues of reverse causation—namely, that those who begin gambling or develop gambling problems participate less in religion as a result of cognitive dissonance or competing demands for time. Thus, by using nationally representative panel data on young adults, employing innovative measures, and investigating the associations with religion, we are able in this study to more clearly answer the questions: (1) How are different religious characteristics of young adults related to the prevalence of gambling and problem gambling, and (2) To what extent can these religion-gambling associations be explained in terms of propensities for risk-taking?

Religion and Gambling Religion is a multidimensional construct, involving (among other things) personal identity, religious practices, and beliefs. Any of these can be linked theoretically to gambling

123

J Gambl Stud

behavior. Below we explain how identification with different religious traditions, religious communities, and personal religiosity may influence gambling behavior. Religious Traditions Among the major influences in the lives of young adults, religious organizations present the clearest messages regarding gambling. Major religious traditions in the United States have, to different degrees, discouraged their adherents from gambling. In many cases, the opposition to the activity is strikingly unmitigated. Despite relatively widespread opposition to gambling on an institutional level, however, important subcultural variations among religious traditions might affect the way in which religion affects gambling. Choices about all kinds of behaviors, including gambling, are not made in social vacuums. Instead, individuals are embedded in subcultures, which have their own history, beliefs, norms, and practices (Granovetter 1985). Beliefs and practices are susceptible to subcultural variation; different religious traditions have different norms and beliefs about gambling. The largest group on the anti-gambling end of the spectrum is conservative Protestants. This group uses both the social ills of gambling and biblical arguments to build their case against gambling. The Southern Baptist Convention, for example, appeals to the ‘‘pain and destruction [left by gambling] in the lives of countless people, especially the children, poor, and elderly’’ (Southern Baptist Convention 1997) in its resolution against gambling. The Assemblies of God, another conservative Protestant group, articulates four arguments against gambling based on their interpretation of the Bible, including assertions that gambling (1) promotes irresponsible stewardship of money, (2) encourages profiting at another’s expense, (3) helps develop a poor work ethic, and (4) may be habit-forming (General Council of the Assemblies of God 1983). Other biblical arguments from conservative Protestants conceive of gambling as challenging the sovereignty of God and akin to worshiping false gods (Hoffmann 2000). Conservative Protestants themselves tend to echo this opposition. Texans with conservative Protestant affiliations were far more likely to oppose a state lottery than their counterparts (Ellison and Nybroten 1999). Some crosssectional evidence also suggests that they gamble less frequently than others (Ellison and McFarland 2011). Joining conservative Protestants in their staunch opposition to gambling are Mormons (i.e., Latter-Day Saints), Jehovah’s Witnesses, and Muslims. For example, the Church of Jesus Christ of Latter-Day Saints (2015) states unequivocally on its web site that it is ‘‘opposed to gambling, including lotteries sponsored by governments.’’ Jehovah’s Witnesses are also opposed to gambling, citing among other things trust in luck instead of God (Watchtower 2011). For Muslims, gambling is explicitly denounced in the Quran: ‘‘[Gambling is] … but defilement from the work of Satan, so avoid it that you may be successful’’ (5:90). Mainline Protestant groups, who tend to be more theologically liberal, nonetheless also tend to produce anti-gambling messages. For example, the United Methodist Church— itself no bastion of religious conservatism—condemns gambling as ‘‘a menace to society, deadly to the best interests of moral, social, economic, and spiritual life, and destructive of good government’’ (United Methodist Church 2004), before advocating abstinence from gambling on the part of all Christians. These groups have even joined with their conservative Protestant cousins in efforts to stave off state lottery initiatives (Olsen et al. 2003). Yet the social attitudes and behaviors of mainline Protestants are not necessarily reflective of their leaders’ positions. Clergy from the Evangelical Lutheran Church in America and the Episcopal Church perceive themselves to be more liberal than their congregation on a

123

J Gambl Stud

wide range of public policy issues. For example, while two-thirds of these clergy agreed that the federal government should do more to solve social problems, only 30 % felt a majority of their congregation agreed with that statement (Djupe and Gilbert 2003). Moreover, Wilcox (2004: 75) suggests that, unlike conservative Protestants, the family ideologies of mainline Protestants are loosely linked with those of their leaders. Thus, official mainline denominational stances on gambling may not be good predictors of gambling attitudes and behaviors among the rank-and-file. Some religious groups, however, are not so clearly against gambling. The Catholic Church, for example, has not taken the hard-line that many Protestant groups have adopted. Instead, the Catholic Church has a long tradition of sponsoring low-stakes games of chance in fundraising and social activities, and their official stance on gambling is much more moderate. According to the Catechism of the Catholic Church 1994 (paragraph 2413), ‘‘Games of chance (card games, etc.) or wagers are not in themselves contrary to justice. They become morally unacceptable when they deprive someone of what is necessary to provide for one’s needs and those of others.’’ Though the existing research on religion and gambling has used rather crude distinctions among religious traditions, Catholics tend to gamble more frequently and are more likely to be problem gamblers than Protestants (Hraba and Lee 1996; Feigelman et al. 1998; Diaz 2000; Ellison and McFarland 2011) and may even be more likely than religious nones (i.e., those with no religious affiliation) to have gambled, perhaps because their tradition has sponsored and sanctioned such activity (Diaz 2000; Eitle 2011). Similarly, gambling has a long history in the Jewish tradition (Brozan 1984) even as it is often viewed as incompatible with the goal of building character (Bell 1974). Based on the above discussion, then, we expect that young adults who were raised as conservative Protestants, mainline Protestants, Mormons, Jehovah’s Witnesses, and Muslims will be the least likely to have ever gambled while those raised Catholic or Jewish will be the most likely. Those raised with no religious affiliation will not have been exposed to either positive or negative messages about gambling from religious institutions and thus we expect them to fall in between adherents to anti-gambling religious traditions and those who have been more accepting of gambling (i.e., Catholics and Jews) in their gambling behavior. The non-religious are, in some ways, a ‘‘control group’’ against which we compare adherents to different religious groups. Religious Communities Since Durkheim’s ground breaking work on suicide (1951[1897]), social scientists have been interested in how individuals respond differently according to the religious composition of their communities. The religious composition of communities may have important effects on individual gambling behaviors in at least two ways. First, religious organizations that oppose gambling may take collective actions which impact the availability of gambling. Conservative Protestants in particular have led efforts to stop state lotteries (usually unsuccessfully) as well as taking more localized actions against casinos and video gambling machines (Ellison and Nybroten 1999; Olsen et al. 2003). Second, young adults may grow up in areas with high concentrations of religious people who oppose gambling (such as conservative Protestants). Even young adults who themselves are not particularly religious may be influenced by anti-gambling messages from religious people. Or, more simply, young adults may be influenced by normative non-gambling behaviors (i.e., peer pressure) in communities with a strong anti-gambling religious presence. Regnerus (2003), testing a form of Stark’s ‘‘moral communities’’ thesis (1996), proposed and found that

123

J Gambl Stud

youth were less likely to engage in delinquency when they lived in communities with higher concentrations of conservative Protestants. Because a high concentration of conservative Protestants may reduce the availability of gambling and/or increase the normative pressure to avoid gambling, we predict that young adults from communities with a higher percentage of conservative Protestant adherents will be less likely to gamble, regardless of their own religious affiliation or participation.1 Religiosity Religiosity has both a public and private component. Attendance at religious services is one way in which individuals publicly display their religiousness, and attendance at religious services might be an important factor in determining gambling behavior. Young adults who make repeated decisions to attend religious services may be exposed to religious messages condemning gambling more frequently. People who attend religious services more frequently may fear negative sanctions from their co-religionists, as well as feelings of guilt and embarrassment, should they violate their religious norms (Grasmick et al. 1991; Ellison and Levin 1998). Even in religious communities where gambling is more widely accepted, religious attendance may cultivate habits of self-discipline, moderation, and careful stewardship of money, thus reducing the appeal of financially risky behaviors such as gambling, which might ‘‘tempt’’ frequent attenders into excess. Indeed, the extant literature suggests that religious attendance does appear to be related to gambling behavior. Hoffmann (2000) finds that both religious service attendance and strength of faith in God are negatively related to individuals’ lifetime gambling participation, but only religious service attendance predicts a lower incidence of problem gambling. Another recent study suggests that religious service attendance is negatively correlated with all types of gambling and with the respondent’s frequency of gambling (Lam 2006). Ellison and McFarland (2011) report an inverse relationship between frequency of church attendance and frequency of gambling. Eitle (2011), however, finds no association between religious attendance and problem gambling. Despite this last finding, we expect that those young adults who attend religious services more frequently will be less likely to gamble and to exhibit gambling problems. Religiosity also has a private component. Individuals internalize their faith and use it to guide their decision-making. As mentioned above, strength of one’s faith is negatively associated with the incidence of gambling (Hoffmann 2000). Indeed, religious salience may explain why religious service attendance affects gambling behavior. That is, it may be that people who attend religious services more frequently are merely more personally religious, and it is this personal religiousness and internalization of religious teachings that has a direct influence on gambling behavior. Does Risk-Taking Orientation Explain Religious Influences? It may be that religion has no direct effect on gambling. Some scholars question whether apparent religious influences are due to unobserved endogenous factors such as inherent dispositions or other spurious associations (e.g., Cochran et al. 1994). One possibility is that religiosity is related to individual propensity for risk, specifically that persons who are 1

We also hypothesized that persons would be more likely to gamble when living in areas with high concentrations of Catholics, who occasionally sponsor low stakes games of chance (Diaz 2000). We found no evidence of this in our ancillary analysis and thus do not pursue this line of argument here.

123

J Gambl Stud

more risk-averse are also more likely to be religious. This position has been most clearly articulated as an explanation for the persistent, cross-cultural, male–female gap in religiosity and typically conceives of religious persons as avoiding the risks of punishment in the afterlife (Miller and Hoffmann 1995; Miller and Stark 2002). While the idea that women (and religious persons in general) are more religious because they fear hell has been hotly refuted (Roth and Kroll 2007), it may still be the case that religious persons are more risk-averse on average. Gambling, of course, is a classic risk-taking behavior, and one would expect risk-averse individuals to avoid it. If these risk-averse persons are more likely to be religious and also less likely to gamble, then associations between religion and gambling behavior may be spurious or, at best, indirect (Lesieur 1994). If this is true, we expect that controlling for the risk-taking propensity of individuals would explain or significantly reduce any observed effects of religion on gambling behavior. Religion and Problem Gambling Thus far we have hypothesized that young adults raised in certain religious traditions, young adults raised in conservative Protestant communities, and young adults who attend religious services more frequently will be less likely, and those raised as Catholics or Jews more likely, to have ever gambled. Among young adults for whom religious influences are associated with avoiding gambling altogether, problem gambling will naturally also be avoided. It is also possible that religious young adults who do experiment with gambling will be less likely to experience gambling problems as they heed religious warnings against excess. Additionally, those young adults who might otherwise struggle with gambling problems may find coping resources in the religious community and thus feel less inclined to turn to excess gambling. This may be true for Catholics as well, even if they are more likely to experiment with gambling, as the Catholic Church clearly proscribes excessive or harmful gambling. Thus, we expect high attending youth who have experimented with gambling, regardless of religious tradition, to be less likely to experience gambling problems. On the other hand, some religious influences might have a negative impact among those who have experimented with gambling. We suggest two related reasons why this might be the case. First, because of the strong anti-gambling messages offered by many religious groups, young adults who have experimented with gambling may be less likely to discuss their gambling exploits among co-religionists, and thus be less likely to have access to religious resources which could stave off gambling problems. Second, religious institutions which take a strong abstinence position toward gambling may be unable or unwilling to offer resources to assist young adults in avoiding problem gambling. Thus, among those who have gambled, young adults who grew up in anti-gambling religious traditions, regardless of their level of participation, and young adults raised in a heavily conservative Protestant community, may be more likely to report gambling problems.

Data The data for this study are from Waves I and III of the National Longitudinal Study of Adolescent to Adult Health (Add Health). Add Health is funded by the National Institute of Child Health and Development (NICHD) and 23 other federal agencies. Wave I was conducted in 1994 and 1995, including in-depth interviews with 20,745 American

123

J Gambl Stud

adolescents in grades 7–12. Adolescents’ parents, siblings, friends, romantic partners, fellow students, and school administrators were also surveyed. Wave III was conducted in 2001 and 2002 and consisted of interviews with 15,197 of the Wave I respondents. More than 99 percent of the Wave III respondents were age 18–25, and questions in this wave focused on topics pertinent to young adults. More information about Add Health is available at www.cpc.unc.edu/addhealth. We used only cases with valid sampling weights and excluded cases with missing values on the dependent variables or cases from counties with fewer than five respondents in the sample. This yields an analytic sample of 13,980 cases. To account for missing data on the predictor variables we performed multiple imputation in Stata using the ice command (StataCorp 2013).

Measures Dependent Variables Participation in Gambling Questions on gambling were asked in Wave III of Add Health. The questions used for constructing the dependent variables were: ‘‘Have you ever bought lottery tickets, such as daily, scratch-offs, or lotto?’’ ‘‘Have you ever played casino tables or video games for money—such games as craps, blackjack, roulette, slot machines, or video poker?’’ and ‘‘Have you ever played any other games, such as cards or bingo, for money, or bet on horse races or sporting events, or taken part in any other kinds of gambling for money?’’ The response choices for the questions were ‘‘yes’’ and ‘‘no.’’ From these three questions, we construct a dichotomous dependent variable measuring whether the respondent has participated in either casino-style gambling, lottery gambling, or other gambling.2 Problem Gambling Following the questions about ever having participated in gambling, Add Health included five questions concerning gambling problems. These questions were only asked of those respondents who indicated that they had participated in some form of gambling. The first question asked, ‘‘In all the time since you first started any type of gambling, what would you say was the largest amount of money that you have ever been behind across an entire year of gambling?’’ Respondents were given eight possible responses to this question: never gambled, never lost money, lost less than $100, lost between $100 and $500, lost between $501 and $1000, lost between $1001 and $5000, lost between $5001 and $10,000, and lost more than $10,000. From this question we constructed a simple dichotomous measure of whether or not the respondent lost more than $500.3 Four other problem gambling questions, to which respondents could answer ‘‘yes’’ or ‘‘no,’’ were as follows: ‘‘Have there ever been periods lasting 2 weeks or longer when you 2

We also conducted exploratory analyses with each of the three types of gambling as a separate dependent variable. The results were substantively similar for our religion predictor variables across the models. Religious predictors were slightly more effective at explaining lottery gambling then casino or other types of gambling.

3

In exploratory analyses we also examined money lost with an ordered logistic regression (predicting the full range of the variable) and found that there was little significant variation in money loss above $500, at least as predicted by our religion measures.

123

J Gambl Stud

spent a lot of time thinking about your gambling experiences or planning out future gambling ventures or bets?’’ ‘‘Have you ever gambled to relieve uncomfortable feelings such as guilt, anxiety, helplessness, or depression?’’ ‘‘Has there ever been a period when, if you lost money gambling 1 day, you would return another day to get even?’’ and ‘‘Has your gambling ever caused serious or repeated problems in your relationships with any of your family members or friends?’’ From these four questions we constructed four dichotomous measures, each one indicating that the respondent answered ‘‘yes’’ to the question.4 Independent Variables Religious Affiliation For this analysis we use the Wave I information on the primary respondent’s religious denomination to classify respondents as conservative Protestant, mainline Protestant, Catholic, Mormon (LDS), Jewish, Jehovah’s Witness, Muslim, other religion, or no affiliation; this is a modified form of the RELTRAD coding scheme that has been used frequently in recent years (Steensland et al. 2000). The primary differences of our scheme from RELTRAD is that we move black Protestants into either the mainline or conservative Protestant category, depending on the religious beliefs and traditions of the particular black Protestant denomination,5 and—taking advantage of the large Add Health sample—we break out Mormons, Jehovah’s Witnesses, and Muslims from the catch-all ‘‘other religion’’ category. In our analysis of problem gambling, using these small categories sometimes resulted in perfect prediction; thus, for our problem gambling analysis we include Mormon, Jewish, Jehovah’s Witness, and Muslim young adults in the ‘‘other religion’’ category. Religious Community Add Health includes information on the proportion of adherents to major religious traditions by county at Wave I. For our investigation we use variables for the proportion of conservative Protestants, transformed to indicate the percentage of conservative Protestants in the respondent’s county. Unfortunately, the county-level data do not include the proportion of the population who adhere to small religious groups. Religious Attendance We use religious attendance as our first measure of religiosity. Our measures are created from the Wave I question: ‘‘In the past 12 months, how often did you attend religious services?’’ Using the four response categories to this question, we create four dummy 4

In exploratory analyses we also combined the five problem gambling measures described above into an index measure of problem gambling (alpha = .79; tetrachoric = .89). This index was similar to measures of problem gambling used in other studies (Hoffmann 2000). We experimented with a number of modeling techniques, including count-based (poisson, negative binomial) and ordered logistic regressions. In each case the results were heavily driven by the variable for money lost. Ultimately, we felt the index measure obscured important variation among the problem gambling measures.

5

We removed the black Protestant category from RELTRAD to prevent potential collinearity issues with the race controls and because we think that preserving race as an analytical category is more important to this analysis than any advantages which would be gained by using the black Protestant category.

123

J Gambl Stud

variables for levels of attendance: weekly (or more) attendance, one to three times per month attendance, less than once per month attendance, and no attendance. We use a set of dichotomous variables for attendance for two reasons. First, dichotomous measures better represent groups of people and more accurately reflect the ordered, categorical nature of the underlying variable. Second, in preliminary analyses, we discovered strong evidence of a non-linear relationship between religious attendance and the dependent variables, rendering an interval-level variable inappropriate. Other studies have demonstrated value in including multiple measures of religiosity (especially attendance and salience) and including them in the same models despite the high degree of overlap (e.g., Regnerus and Elder 2003). We include a measure of religious salience, coded as four dichotomous variables. We include categories for religion is very important, somewhat important, somewhat unimportant, and not important at all. In addition to salience, we examined a few other religious variables available in the Add Health, such as respondent’s frequency of prayer and beliefs about the Bible. None of these significantly predicted gambling behavior once the other religious variables were included in the model. For parsimony we include religious attendance and religious salience as the only measures of religiosity in our analysis. In Wave I, Add Health respondents who indicated no religious affiliation were skipped out of subsequent religion questions, including the questions on religious attendance and religious salience. Following other Add Health users, we impute the lowest values on attendance and salience for those who indicated no religious affiliation (e.g., Regnerus and Elder 2003). Note that all of our religion measures are taken from Wave I, indicating the religious characteristics of the respondent during adolescence. We employ this analytic strategy for three reasons. First, because the dependent variables are based on young adults’ reports of ever having engaged in various gambling behaviors, using Wave I measures is our best attempt at modeling religious characteristics before gambling behavior occurred. While this strategy certainly does not give us a pass to make causal claims, it at least allows our empirical models to better fit our theory, that religious characteristics have some influence on young adults gambling behaviors. Second, adolescence and young adulthood are turbulent and dynamic, with religious change (usually decline) often occurring during this part of the life course (Uecker et al. 2007). Because of the typical religious declines during young adulthood using Wave III religion measures may overlook important effects of religion on gambling that stem from behavior patterns rooted in adolescent and childhood religious characteristics. Finally, we did explore using Wave III measures and found no major substantive differences from those reported (using Wave I measures) in this study. Thus, we report only Wave I religion measures here. Risk-propensity To control for the possibility that both religious responses and gambling behaviors might be due to underlying risk-propensity, we use an index of sensation/thrill-seeking tendencies (Miller and Hoffmann 1995; Miller and Stark 2002; Lightsey and Hulsey 2002). The thrillseeking index consists of nine items (standardized alpha = .88) from Wave III,6 each of 6

Unfortunately, the items we use in our thrill-seeking index are not included in Wave I of Add Health; they only appear in Wave III. Because we use this measure to tap risk-propensity as an underlying personality characteristic we believe our use of this Wave III variable is reasonable. Wave III also includes a shorter list of items to measure risk-propensity but this set of items was asked only of a small subset of Wave III participants.

123

J Gambl Stud

which asks respondents to indicate how true for them a particular statement of thrillseeking behavior is. The nine statements are: ‘‘I often try new things just for fun or thrills, even if most people think they are a waste of time.’’ ‘‘When nothing new is happening, I usually start looking for something exciting.’’ ‘‘I can usually get people to believe me, even when what I’m saying isn’t quite true.’’ ‘‘I often do things based on how I feel at the moment.’’ ‘‘I sometimes get so excited that I lose control of myself.’’ ‘‘I like it when people can do whatever they want, without strict rules and regulations.’’ ‘‘I often follow my instincts, without thinking through all the details.’’ ‘‘I can do a good job of ‘‘stretching the truth’’ when I’m talking to people.’’ ‘‘I change my interest a lot, because my attention often shifts to something else.’’ The five response categories range from ‘‘not true’’ to ‘‘very true.’’ Controls We include control variables for gender (coded one for female), age, and race. We also include a dichotomous variable for never having married, leaving all previously and currently married respondents as the reference category. For education level of the respondent, we create dummy variables indicating that the respondent did not complete high school and that the respondent had been awarded a college degree. This leaves a reference category of all those who completed high school but who have not yet completed a college degree. Similarly, to control for SES and family background, we include a dummy variable indicating at least one parent had earned a college degree. We also include a measure of personal earnings at Wave III. Finally, as one study (Hoffmann 2000) suggests that time usage might be an important factor in choices about gambling, we also include controls for weekly hours worked (continuous). Table 1 lists descriptive statistics for our analytic variables.

Analytic Approach We begin our analysis with a brief examination of descriptive statistics to establish the magnitude of gambling and problem gambling among young adults. In multivariate analyses we first examine the association of various aspects of religion with any reported type of gambling and then move to an investigation of religious effects on reports of gambling problems. Because gambling problems are relevant only for those who have ever gambled, we conduct our analysis of problem gambling only on the subsample of respondents who report having ever gambled. For our analysis of whether young adults report any type of gambling, we use nested models, adding religion measures and the thrill-seeking measure one at a time. Starting with religious affiliation in our first model, we subsequently include measures for the percentage of conservative Protestants in the county, religious attendance, religious salience, and finally thrill-seeking. In this way we can observe whether subsequent religion measures explain any observed effects from previous models. While other modeling orders can be justified, we believe our strategy best fits our theoretical framework. Namely, differences in gambling behavior by religious affiliation may be explained by availability to gambling (religious community). Similarly, differences in gambling behavior by religious affiliation which are not accounted for by religious community may be best explained by religious participation. Effects of religious participation may in turn be explained by one’s internal commitment to religion. Finally, observed religious differences

123

J Gambl Stud Table 1 Descriptive statistics (Weighted)—Add Health (N = 13,980) Mean Gambling behaviors (W3) Ever gambled Lost more than $500 (N = 10,249) Gambled to get even (N = 10,247) Relational problems from gambling (N = 10,246) Planned gambling in advance (N = 10,248) Gambled for a feeling (N = 10,247) Religious affiliation (W1) Conservative Protestant Mainline Protestant Catholic Mormon (LDS) Jewish Jehovah’s Witness Muslim Other religion No affiliation Religious community (W1) % Conservative Protestant in county (2.300–77.600) Religious attendance (W1) Never Less than once per month One to three times per month Weekly or more Religious salience (W1) Not important at all Somewhat unimportant Somewhat important Very important Thrill-seeking (W3) (5–45) Controls Female (W1) Age (18–28)(W3) Hispanic (W1) African–American (W1) Other race (W1) White (W1) No HS diploma or GED (W3) HS graduation and some college (W3) Graduated 4-year college (W3) At least one parent 4-year college graduate (W1) Average hours worked weekly (0–90)(W3) Single–never married(W3) Personal earnings (1–8)(W3)

SD

.736 .050 .025 .004 .014 .007 .318 .233 .246 .011 .008 .012 .003 .043 .123 22.851

17.435

.237 .182 .197 .381 .154 .066 .366 .412 23.257 .491 21.811 .118 .155 .073 .653 .109 .788 .103 .316 33.474 .815 2.191

8.291

1.867

15.493 1.505

123

J Gambl Stud

in gambling behavior may be explained by an underlying propensity for risk. For the multivariate analysis of problem gambling we show only the full models. For our multivariate analysis predicting whether respondents have ever gambled (Table 2), we employ hierarchical linear models. Since we do not find (in ancillary analyses) any county level religious affiliation effects on problem gambling, we estimate models predicting problem gambling (Table 3) as simple individual-level models using the svy command in Stata (StataCorp 2013) that account for clustering within schools and the individual probability of selection into the sample. All models are estimated using the mim command to analyze our multiply imputed data (Carlin et al. 2008). We might reasonably expect some statistical interactions among our various religion measures. For instance, perhaps conservative Protestants are only less likely to have gambled if they also attended religious services frequently and were thus exposed to the anti-gambling messages of the religious tradition. Or possibly Catholic young adults who grew up in heavily conservative Protestant communities might be less likely to gamble compared with their Catholic peers who grew up elsewhere. Cross-level interactions, in particular, are expected by the ‘‘moral communities’’ theory, though they are somewhat rare finds empirically (see Regnerus 2003). In exploratory analyses, we examined a number of (individual and cross-level) interactions among our various religion measures but none of these reached statistical significance. In this study, for parsimony, we have limited both our theorizing and analysis to main effects.

Results Table 1 presents descriptive statistics for our analytic sample. About 74 % of young adults have participated in some form of gambling. Far fewer have experienced gambling problems. Only about 5 % have lost more than $500 while gambling. About 3 % have gambled to get even for a previous day’s losses. About 1 % of young adults report planning their gambling ahead of time. Fewer than 1 % say they have experienced relational problems because of gambling or have gambled because of the feelings they experience. The gambling behaviors reported seem reasonable compared with estimates based on the full US adult population. We might expect less gambling and fewer gambling problems among young adults simply because they have had fewer opportunities. The plurality of our sample reported a conservative Protestant affiliation during adolescence, and 38 % report having attended religious services weekly during adolescence. While these figures might strike some as a bit high, they are on par with other studies of religion among adolescents (e.g., Regnerus and Smith 2005). In other respects, our sample is largely white (65 %), 10 % already have completed a 4-year college degree, 32 % report having at least one college educated parent, and 82 % are never-married. Table 2 reports odds ratios from hierarchical linear models predicting the odds of having ever gambled. In Model 1, even while controlling for a number of key predictors, there are a number of differences in gambling across religious affiliations. Conservative Protestants, mainline Protestants, Mormons, Jehovah’s Witnesses, and adherents to other religions are all statistically less likely than the religiously unaffiliated to report having ever gambled. Conservative Protestants have about 39 % lower odds, mainline Protestants have 21 % lower odds, Mormons have 61 % lower odds, Jehovah’s Witnesses have 77 % lower odds, and those from other religions have 30 % lower odds. Catholics, Jewish young adults, and Muslims do not differ significantly from those with no religious affiliation. Most of these findings fit well with previous studies and resonate with our theoretical

123

.614*** .790** 1.146 .389*** 1.361 .231*** .667 .704**

.048 .063 .096 .078 .395 .041 .235 .085

.049 .063 .096 .078 .391 .041 .236 .085 .004

.625*** .795** 1.144 .389*** 1.348 .232*** .669 .708** .988**

SE(B)

.004

.107 .088 .062

1.241* 1.053 .802**

.066 .082 .122 .094 .385 .050 .240 .096

SE(B)

.989**

.655*** .820* 1.171 .443*** 1.291 .263*** .672 .720*

OR

Model 3

.108 .093 .070

.176 .172 .147

1.065 1.181 .994

.004

.095 .118 .169 .101 .377 .055 .242 .119

SE(B)

1.226* 1.057 .832*

.990**

.606** .752 1.064 .413*** 1.178 .245*** .640 .662*

OR

Model 4

1.093 1.200 1.023 1.031***

1.220* 1.057 .843*

.990**

.598** .739 1.035 .405*** 1.165 .242*** .576 .651*

OR

Model 5

.182 .176 .153 .003

.108 .094 .071

.004

.095 .117 .165 .099 .375 .055 .220 .118

SE(B)

* p \ .05; ** p \ .01; *** p \ .001

Reference groups in parentheses. All models control for, but do not display results for, gender, age, race-ethnicity, education, parent’s education, hours worked per week, marital status, and personal earnings. N = 13,980

Religious affiliation (W1) (No affiliation) Conservative Protestant Mainline Protestant Catholic Mormon (LDS) Jewish Jehovah’s Witness Muslim Other religion Religious community (W1–Level 2) % Conservative Prot. in county Religious attendance (W1) (Never) Less than one time per month One to three times per month Weekly or more Religious salience (W1) (Not important at all) Somewhat unimportant Somewhat important Very important Thrill-seeking (W3)

OR

OR

SE(B)

Model 2

Model 1

Table 2 Odds ratios from hierarchical linear models predicting ever having gambled—Add Health

J Gambl Stud

123

123

.731

1.739

2.064*

1.051

One to three times per month

Weekly or more

1.502

1.062***

.315*

.421

.615

1.733

3.663**

3.034**

1.318

1.509

1.592

.802

.012

.165

.225

.368

.793

1.587

1.197

.780

.769

.851

1.076**

.180

.498

.383

2.017

5.701

2.849

2.097

1.627

1.338

1.044

OR

.026

.219

.539

.450

1.793

5.072

2.369

2.436

1.698

1.562

1.426

SE(B)

Relational problems

1.062***

.401

.562

.868

1.976

3.823*

3.506*

2.702

2.288

2.689

2.169

OR

.015

.335

.461

.709

1.154

2.044

1.703

2.391

1.822

2.295

1.821

SE(B)

Planned in advance

1.064***

1.602

1.637

2.377

1.983

2.109

3.726*

1.506

.815

.636

1.106

OR

.016

1.803

1.806

2.773

1.284

1.331

2.366

1.922

1.051

.774

1.384

SE(B)

Gambled for a feeling

* p \ .05 ** p \ .01 *** p \ .001

Reference groups in parentheses. All models control for, but do not display results for, gender, age, race-ethnicity, education, parent’s education, hours worked per week, marital status, and personal earnings. N for each outcome is 10,249 (lost [ $500), 10,247 (gambled to get even), 10,246 (relational problems), 10,248 (planned gambling in advance), and 10,247 (gambled for a feeling)

.008

.221

1.047***

.558

Thrill-seeking (W3)

Very important

.287

.980

.754

Somewhat important

.334

.333

.637

.492

Somewhat unimportant

Religious salience (W1) (Not important at all)

1.692

Less than one time per month

Religious attendance (W1) (Never)

Other religion

.479

1.094

Catholic

.547

Mainline Protestant

.543

1.328

1.287

Conservative Protestant

Religious affiliation (W1) (No affiliation)

SE(B)

OR

OR

SE(B)

Gambled to get even

Lost [ $500

Table 3 Odds ratios from logistic regression models predicting problem gambling symptoms among respondents who report having ever gambled—Add Health

J Gambl Stud

J Gambl Stud

framework. Conservative Protestants consistently and (usually) unequivocally exhort members to abstain from gambling and at least some young adults raised as conservative Protestants appear to have gotten the message. The Mormon church and Jehovah’s Witnesses also provide clear anti-gambling messages that appear to be heeded by members. Mainline Protestant churches like the United Methodist Church are also opposed to gambling—mostly for its social justice implications. Catholics, on the other hand, deride excessive gambling but otherwise prescribe moderation, and while the effectiveness of the moderation message remains to be seen in Table 3, young adults with a Catholic background clearly feel free to experiment with gambling, as do Jewish young adults. Muslim young adults do not differ significantly from the religiously unaffiliated despite institutional opposition to the practice, though the odds ratio for these respondents is less than one. Before moving to Model 2, we note a few interesting features from the control variables (results not shown; available upon request). First, at least among young adults, women have lower odds of having ever gambled. This contradicts some earlier studies of adults (Gerstein et al. 1999). Not surprisingly, older young adults have greater odds of reporting having ever gambled. African Americans have lower odds of trying gambling, as do Hispanics. Interestingly, young adults with a college degree are more likely to have gambled, while young adults from families where at least one parent has a college degree are less likely to have gambled. Finally, young adults who work more also have greater odds of reporting having ever gambled, as do those with higher personal earnings. Model 2 of Table 2 shows that young adults who grew up in counties with a higher percentage of conservative Protestants also have lower odds of having ever gambled. As we theorized, the association between percentage of conservative Protestants in the county and lower likelihood of gambling may be due to a lower availability of gambling in these areas, or it may be because young adults from these areas are more likely to have been exposed to anti-gambling messages. The inclusion of county-level religious affiliation does not alter the odds ratios for religious affiliation seen in Model 1 in any meaningful way.7 Model 3 of Table 2 introduces religious attendance into our models. With never attending as the reference category we observe that young adults who rarely attended as adolescents have 24 % greater odds of having ever gambled, but those who attended weekly as adolescents have 20 % lower odds of having ever gambled. Including religious attendance does little to change the odds ratios or statistical significance of the affiliation measures or the religious community measure. It appears, then, that affiliation, religious communities, and religious service attendance have largely independent effects on gambling among young adults. Model 4 of Table 2 adds the measure of religious salience. Net of the other religion variables in the model, religious salience does not have a significant influence of the odds of ever gambling among young adults. Nor does its inclusion substantially alter the odds 7

The percentage of conservative Protestants in county only predicts a small amount of the (considerable) county level variation in the dependent variable (results not shown; available upon request). This large amount of variation is likely due to differential availability of gambling from one county to another. Unfortunately, Add Health provides only very crude measures of region (South, West, Northeast, Midwest) so we were unable to incorporate regional controls, even at the state level. In ancillary analyses using the available measures of region as level 2 predictors, we found no significant effect of region. We also experimented with county-level measures of socio-economic status and these too proved poor predictors. We expect that much of the county-level variation could be explained with measures tapping the availability of legal gambling in the area (such as the proximity to casino measure used by Hoffmann (2000) or a measure of proximity to a state with a lottery). Unfortunately, such measures were not available to us in Add Health.

123

J Gambl Stud

ratios of the other religion variables, though the odds ratio for mainline Protestants is no longer statistically significant in Model 4 (with a p value of .067). Model 5 of Table 2 introduces our measure of thrill-seeking, and it is strongly predictive of gambling behavior. Accounting for thrill-seeking, however, does almost nothing to attenuate the associations between our various religion measures and gambling behavior. If both gambling and religion were due to some underlying propensity for risk-taking we would expect thrill-seeking to largely explain any observed effects of religion. We see no evidence of that here. Table 3 presents odds ratios from logistic regressions predicting the various gambling problems. Following our analysis plan, these models are only for those roughly 74 % of respondents who reported having ever gambled. As we noted in our theoretical framework, excluding the non-gamblers may ‘‘explain’’ any religious effects we might otherwise expect for problem gambling. This would be true if religion lowers the odds of problem gambling simply by encouraging respondents to abstain from gambling. The most frequently reported gambling problem was losing more than $500 while gambling. Religious effects for this variable, however, are not as numerous as they were for the ever gambled outcome. Still, young adults who attended church one to three times per month as adolescents have more than twice the odds of losing large amounts of money while gambling compare to those who never attend. Those who attend weekly are statistically indistinguishable from non-attenders, as are those who only attend less than once per month. Other religion measures appear to have no significant effect on reports of losing over $500 while gambling. We see a similar pattern with gambling to get even and planning gambling behavior, but here those who attended less than once a month as adolescents are also more likely to exhibit symptoms of problem gambling. Additionally, those who attended less than once per month in adolescence are also much more likely to gamble to relieve negative feelings. The effects of religious predictors on relational problems are not statistically significant, though many of the odds ratios are large. In preliminary analyses we found no evidence of religious community effects on any of the gambling problems, thus, as noted earlier in our analytic plan, we exclude these variables from these individual-level, fixed-effects models. Before moving to our conclusions we note a few interesting findings from among our control variables in Table 3 (results not shown; available upon request). An inclination for thrill-seeking strongly predicts each kind of gambling problem. Women are much less likely to report gambling problems, with around 75 % lower odds for every gambling problem we measured except for gambling to relieve uncomfortable feelings, where they were not statistically different from men. Young adults with higher incomes had greater odds of reporting gambling problems on three of the five problems we studied. Finally, respondents in the other race category had greater odds of reporting gambling problems on four of the five outcomes. We conjecture that the ‘‘other race effect’’ is driven partially by Native American casino gambling as well as gambling problems among Asian Americans (Glionna 2006).

Discussion and Conclusions Despite the rapid proliferation of gambling outlets and the tremendous financial impact of the gaming industry, little is known about gambling during young adulthood, the most riskprone part of the life course and when gambling first becomes legal. In this study, we

123

J Gambl Stud

investigated the associations between young adults’ religious characteristics (while they were adolescents) and their current reports of gambling. We theorized that religion was one of the few social institutions where young adults would encounter clear messages about gambling. As religion is multi-dimensional we hypothesized that religious affiliation, religious communities, and personal religiosity— both religious service attendance and religious salience—could all impact gambling behavior via different mechanisms. We also noted a small but important body of literature that suggests that any observed associations between religion and gambling may be due to an underlying propensity for risk. Finally, we predicted that religious effects on problem gambling would be largely through encouraging abstinence from any gambling, though young adults who were highly religious as adolescents may be less likely to report problem gambling, even if they had experimented with gambling. Our findings largely supported our hypotheses, though with a few unexpected twists. We found that most American young adults have experimented with gambling. Very few of them report symptoms of problem gambling. Religion is among the most important social forces predicting gambling behavior among young adults. Specifically, we observed that young adults who grew up conservative Protestant, mainline Protestant, Mormon, and Jehovah’s Witness; those raised in a county with a high percentage of conservative Protestants; and those who frequently attended religious services as an adolescent are less likely to have ever gambled. Among those who have experimented with gambling, non-attenders are less likely than those who attend less than once a month up to three times per month to report a number of gambling problems. Contrary to our expectations, Catholic and Jewish young adults—who we hypothesized would be more likely to gamble—and Muslim young adults—who we thought would be less likely to gamble—did not differ from those with no religious affiliation. Moreover, religious salience does not mediate the effect of the other religious characteristics on gambling, nor does it exert an independent effect on gambling net of the other religious characteristics and control variables. We were somewhat surprised at the independence of our three measures of religion that do influence gambling behavior. Affiliation, religious community, and religious service attendance all predict gambling behavior, largely independently of one another. Notably, religious context—growing up in counties with more conservative Protestants—affects the gambling behavior of young adults even net of their own religious characteristics. In other words, religious institutions influence the social life (specifically, gambling) even of those who are not members of their community. These findings on individual-level and contextual religious effects lend further evidence to a growing body of theory suggesting that the multidimensionality of religion requires multiple measures and levels to adequately unpack religious influences. Another unexpected set of findings was that non-attenders and frequent attenders are somewhat similar in gambling behaviors—though those who attend weekly are modestly less likely to have ever gambled than those who never attend. But it is those who attended sporadically or only semi-regularly as adolescents who exhibit the highest odds of gambling and problem gambling. We admit that this finding is somewhat puzzling as it seems that some exposure to religious messages that are clearly anti-gambling would have more impact than no exposure to anti-gambling messages. We offer a few conjectures about these findings in the hopes of inspiring future research that will more decisively elucidate them. First, it may be that non-attenders (especially to the degree that they overlap with the non-affiliated) are a unique group. Scientists have recently taken an increased interest in this growing proportion of Americans as a distinct social group (see Baker et al. 2009). It is

123

J Gambl Stud

possible that the non-religious may comprise a significant portion of those who avoid gambling for reasons of thrift or social justice, though this is only speculation. Second, it may be that there is something unique about the ‘‘middling’’ attenders. Some studies (e.g., Ellison and Anderson 2001) have found behavioral patterns suggesting that a little religion may be a ‘‘bad’’ thing. It is certainly debatable whether or not gambling is a ‘‘bad’’ thing, but the general idea is that a little exposure to religious messages and culture may be associated with negative outcomes. Finally, a third possibility, related to the first two, is that an underlying propensity for commitment or decisiveness is at work. Both the frequent attenders and non-attenders are committed in some sense to their respective positions while infrequent attenders exhibit less commitment and/or decisiveness. Decisiveness may be related to gambling behavior in that those who are less decisive tend to experiment if the opportunity arises. We await further studies to see if any of these conjectures prove accurate. One pattern of findings from our study which does not require speculation is the clear evidence that risk propensity does not render spurious (or mediate) religious effects on gambling. Whether our measure of risk propensity (thrill-seeking index) is the ‘‘right’’ one to identify religiously-related risk propensities is debatable; empirically, however, it was the strongest and most consistent predictor of gambling behavior and gambling problems across all our models. Yet, despite its robustness, it did very little, if anything, to explain the observed effects of religion. Our study of religion and gambling will not end the debate over spurious religious effects. This study does, however, deal a blow to the theory that religious differences (by gender or anything else) are due to underlying propensities for risk. Risk-aversion may yet prove to explain the pervasive religiosity gap between men and women; it does not explain why religion affects gambling behavior. This study has a number of limitations which prevent us from making causal claims and which invite further research to address these shortcomings. First, our outcome variables are inherently limited. Ideally, we would have had measures of both recent and lifetime gambling behavior. It would have been very helpful to have had information about the frequency of gambling and gambling problems. Additionally, though we understand why the questions weren’t asked (i.e., respondents were too young to legally gamble), our study would have benefitted from gambling measures at Wave I. A second limitation in our study was our inability to directly control for the availability of gambling. While we argue that we account for the primary mechanism by which we would expect gambling availability and religion to be related (religious communities), the direct measures would have been more conclusive. Third, for the reasons above and others, our study is beset by the usual vulnerabilities to selection. There may be other, unmeasured, variables underlying the observed associations between religion and gambling and/or causal processes that work in the opposite direction from those we hypothesized (i.e., gamblers reduce their religious participation). In addition to those we have already mentioned above, this study suggests a number of possibilities for future study. Our results indicate clear differences in gambling behavior by sex, race, and (to a lesser degree) socio-economic background. Whether these differences hold only for young adults or are generalizable to the entire adult population remains to be examined using high-quality nationally representative data. There is room for much more work on the impact of gambling in communities. Our findings lend evidence that community religious composition makes a difference in gambling behaviors. It is likely that other community characteristics matter for gambling behaviors as well. Finally, are there other social institutions which influence gambling behavior among young adults? While there are few clear messages about gambling outside of religious institutions, other arenas,

123

J Gambl Stud

like schools, for instance, may provide to adolescents and young adults broad messages (i.e., about experimentation and moderation) which can influence gambling. To many our primary findings may come as no surprise. Religious institutions are among the few institutions in which young adults encounter clearly articulated messages about gambling, and these institutions, via multiple mechanisms, appear to have an impact on the gambling behaviors of young adults. Perhaps more unexpectedly, religious effects on gambling are not explained by an underlying propensity for risk. For most young Americans, their gambling decisions will primarily impact their wallets, but for the few who experience gambling problems, their decisions may have more dire consequences. This study sheds light on how one area of the social world of young adults structures their decisions about gambling. Acknowledgments This research uses data from Add Health, a program project directed by Kathleen Mullan Harris and designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill, and funded by grant P01-HD31921 from the Eunice Kennedy Shriver National Institute of Child Health and Human Development, with cooperative funding from 23 other federal agencies and foundations. Special acknowledgment is due Ronald R. Rindfuss and Barbara Entwisle for assistance in the original design.

References Arnett, J. J. (1992). Reckless behavior in adolescence: A developmental perspective. Developmental Review, 12, 339–373. Baker, J. O., & Smith, B. G. (2009). The nones: Social characteristics of the religiously unaffiliated. Social Forces, 87(3), 1251–1263. Bell, R. C. (1974). Moral views on gambling promulgated by major American religious bodies. Commission on the Review of the National Policy Toward Gambling. Browne, B., & Brown, D. J. (2001). Predictors of lottery gambling among American college students. The Journal of Social Psychology, 134, 339–347. Brozan, N. (1984). A panel explores gambling among Jews. The New York Times. Retrieved 12 February 2015 from http://www.nytimes.com/1984/12/24/style/a-panel-explores-gambling-among-jews.html. Carlin, J. B., Galati, J. C., & Royston, P. (2008). A new framework for managing and analyzing multiply imputed data in Stata. Stata Journal, 8, 49–67. Catholic Church. (1994). Catechism of the Catholic Church. United States Catholic Conference of Bishops. Church of Jesus Christ of Latter-Day Saints. (2015). Gambling. Retrieved 12 February 2015, from www.lds. org/topics/gambling. Cochran, J. K., Wood, P. B., & Arneklev, B. J. (1994). Is the religiosity-delinquency relationship spurious? A test of arousal and social control theories. Journal of Research in Crime and Delinquency, 31, 92–123. Diaz, J. D. (2000). Religion and gambling in Sin-City: A statistical analysis of the relationship between religion and gambling patterns in Las Vegas residents. The Social Science Journal, 37, 453–458. Djupe, P. A., & Gilbert, C. P. (2003). The prophetic pulpit: Clergy, churches, and communities in American politics. Lanham: Rowman & Littlefield. Durkheim, E. (1951[1897]). Suicide: A study in sociology. Glencoe, IL: Free Press. Eitle, D. (2011). Religion and gambling among young adults in the United States: Moral communities and the deterrence hypothesis. Journal for the Scientific Study of Religion, 50, 61–81. Ellison, C. G., & Anderson, K. L. (2001). Religious involvement and domestic violence among US couples. Journal for the Scientific Study of Religion, 40, 269–286. Ellison, C. G., & Levin, J. S. (1998). The religion-health connection: Evidence, theory, and future directions. Health Education and Behavior, 25, 700–720. Ellison, C. G., & McFarland, M. J. (2011). Religion and gambling among US adults: Exploring the role of traditions, beliefs, practices, and networks. Journal for the Scientific Study of Religion, 50, 82–102. Ellison, C. G., & Nybroten, K. A. (1999). Conservative Protestantism and opposition to state-sponsored lotteries: Evidence from the 1997 Texas Poll. Social Science Quarterly, 80, 356–369.

123

J Gambl Stud Feigelman, W., Wallisch, L. S., & Lesieur, H. R. (1998). Problem gamblers, problem substance users, and dual-problem individuals: An epidemiological study. American Journal of Public Health, 88, 467–470. General Council of the Assemblies of God. (1983). A biblical perspective on gambling. Retrieved 21 January 2008 from http://www.ag.org/top/Beliefs/Position_Papers/pp_downloads/pp_4186_gambling. pdf. Gerstein, D., Murphy, S., & Toce, M. (1999). Gambling impact and behavior study: Final report to the National Gambling Impact Study Commission. Washington, DC and Chicago: National Opinion Research Center (available at www.norc.uchicago.edu). Glionna, J. M. (2006). Gambling seen as no-win situation for some Asians. Los Angeles Times. http:// articles.latimes.com/2006/jan/16/local/me-gamble16. Accessed 13 Apr 2009. Granovetter, M. (1985). Economic action and social structure: The problem of embeddedness. American Journal of Sociology, 99, 481–510. Grasmick, H. G., Bursik, R. J, Jr, & Cochran, J. K. (1991). ‘‘Render unto Caesar what is Caesar’s’’: Religiosity and taxpayers’ inclinations to cheat. The Sociological Quarterly, 32, 251–266. Hoffmann, J. (2000). Religion and problem gambling in the U.S. Review of Religious Research, 41, 488–509. Hraba, J., & Lee, G. (1996). Gender, gambling, and problem gambling. Journal of Gambling Studies, 12, 83–101. LaBrie, R., Shaffer, H., LaPlante, D., & Wechsler, H. (2003). Correlates of college student gambling in the United States. Journal of American College Health, 52, 53–62. Lam, D. (2006). The influence of religiosity on gambling participation. Journal of Gambling Studies, 22, 305–320. Lesieur, H. R. (1994). Epidemiological surveys of pathological gambling: Critique and suggestions for modification. Journal of Gambling Studies, 10, 385–398. Lightsey, O. R., & Hulsey, C. D. (2002). Impulsivity, coping, stress, and problem gambling among university students. Journal of Counseling Psychology, 49, 202–211. Miller, A. S., & Hoffmann, J. P. (1995). Risk and religion: An explanation of gender differences in religiosity. Journal for the Scientific Study of Religion, 34, 63–75. Miller, A. S., & Stark, R. (2002). Gender and religiousness: Can socialization explanations be saved? American Journal of Sociology, 107, 1399–1423. Mok, W. P., & Hraba, J. (1991). Age and gambling behavior: A declining and shifting pattern of participation. Journal of Gambling Studies, 7, 313–335. Olsen, L. R., Guth, K. V., & Guth, J. L. (2003). The lotto and the Lord: Religious influences on the adoption of a lottery in South Carolina. Sociology of Religion, 64, 87–110. Regnerus, M. D. (2003). Moral communities and adolescent delinquency: Religious contexts and community social control. The Sociological Quarterly, 44, 523–554. Regnerus, M. D., & Elder, G. H. (2003). Staying on track in school: Religious influences in high- and lowrisk settings. Journal for the Scientific Study of Religion, 42, 633–649. Regnerus, M. D., & Smith, C. (2005). Selection effects in studies of religious influence. Review of Religious Research, 47, 23–50. Roth, L. M., & Kroll, J. C. (2007). Risky business: Assessing risk preference explanations for gender differences in religiosity. American Sociological Review, 72, 205–220. Southern Baptist Convention. (1997). Resolution opposing gambling and its advertisement. Retrieved 21 January 2008 from http://www.sbc.net/resolutions/amResolution.asp?ID=569. Stark, R. (1996). Religion as context: Hellfire and delinquency one more time. Sociology of Religion, 57, 163–173. StataCorp. (2013). Stata 13.1. College Station, TX. Steensland, B., Park, J. Z., Regnerus, M. D., Robinson, L., Wilcox, W. B., & Woodberry, R. D. (2000). The measure of American religion: Toward improving the state of the art. Social Forces, 79, 291–318. Stinchfield, R., & Winters, K. C. (1998). Gambling and problem gambling among youths. Annals of the America Academy of Political and Social Science, 556, 172–185. Uecker, J. E., Regnerus, M. D., & Vaaler, M. L. (2007). Losing my religion: The social sources of religious decline in early adulthood. Social Forces, 85, 1667–1692. United Methodist Church. (2004). The book of discipline of The United Methodist Church—2004. The United Methodist Publishing House. Watchtower. (2011). ‘‘Does the Bible Condemn Gambling?’’ March 1. pp. 12–14. Welte, J. W., Barnes, G. M., Tidwell, M. C. O., & Hoffman, J. H. (2008). The prevalence of problem gambling among US adolescents and young adults: Results from a national survey. Journal of Gambling Studies, 24, 119–133.

123

J Gambl Stud Welte, J. W., Barnes, G. M., Tidwell, M. C. O., & Hoffman, J. H. (2009). Association between problem gambling and conduct disorder in a national survey of adolescents and young adults in the United States. Journal of Adolescent Health, 45, 396–401. Wilcox, W. B. (2004). Soft patriarchs, new men. Chicago: University of Chicago Press.

123

Religious Background and Gambling Among Young Adults in the United States.

Despite the rapid growth of the gambling industry over the last 40 years, there have been few large-scale, nationally representative longitudinal stud...
255KB Sizes 0 Downloads 6 Views