Sociology paper, experts only

Meows369

ReligionandCharitableFinancialGivingtoReligiousandSecularCausesDoesPoliticalIdeologyMatter.pdf

Home >Social Science homework help >Sociology homework help >Sociology paper, experts only

Religion and Charitable Financial Giving to Religious and Secular Causes: Does Political Ideology Matter?

BRANDON VAIDYANATHAN Department of Sociology, University of Notre Dame

JONATHAN P. HILL Sociology Department, Calvin College

CHRISTIAN SMITH Department of Sociology, University of Notre Dame

Previous research on charitable giving has identified a significant relationship between political conservatism and greater financial giving to charitable causes. Yet that research has not adequately explored the important role of religion in that relationship, nor differences in financial giving targets (i.e., religious congregations, noncongregational religious organizations, and nonreligious organizations). Support for competing theories concerning political ideology, religious practice, and charitable financial giving is assessed using data from the Panel Study on American Ethnicity and Religion (PS-ARE). For both religious and nonreligious giving, the effect of political ideology is completely mediated by participation in religious and civic practices. These findings support recent arguments on “practice theory” in cultural sociology and suggest that it is less the effect of ideology than of active participation in religious, political, and community organizations that explains Americans’ financial giving to religious and nonreligious organizations.

Keywords: political ideology, charitable giving, congregations.

INTRODUCTION

In 2009, an estimated $227 billion were charitably donated to religion, education, human services, health, the arts, and other nonprofit organizations by individuals (Giving USA Founda- tion 2010). These individual gifts represent three-fourths of all charitable donations in the United States and make up substantially more of the GDP than in similar industrialized nations (Charities Aid Foundation 2006).1 Despite this, scholars have not reached consensus on what social factors drive this “culture of giving” (Brooks 2003; Johnstone 2005; White 1998).

Recent work has argued that political ideology plays a key role in explaining charitable giving behavior in the American context—in particular, the differences between those who are generous and those who are not (Brooks 2006; Brooks and Lewis 2001; Derin-Güre and Uler 2010; Wolpert 1995). Empirical evidence points to self-identified political conservatives’ greater financial generosity when compared with liberals (Brooks 2006; Brooks and Lewis 2001). These findings have been used to make the argument that political ideology is a central social

Acknowledgments: A previous version of this article was presented at the Annual Conference of the Society for the Scientific Study of Religion in 2009. We would like to thank Michael Emerson and Adele James, as well as Marie Cornwall and three anonymous referees, for their helpful comments. Correspondence should be addressed to Brandon Vaidyanathan, 810 Flanner Hall, University of Notre Dame, Notre Dame, IN 46556. E-mail: [email protected] 1 In the United States individual charitable gifts make up 1.67 percent of GDP. The comparable statistic for the United Kingdom is 0.73 percent and 0.72 percent for Canada. France ranks lowest with charitable donations making up 0.14 percent of GDP (Charities Aid Foundation 2006).

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 451

mechanism that helps to explain differences in charitable giving: conservative political ideology makes people more generous by inclining them to take individual charitable initiative; conversely, liberal political ideology, by supporting more taxation and redistribution, makes adherents less willing to make charitable donations, which, Brooks contends, is a case of “substituting political opinions for private donations” (2006:61).

But there are both theoretical and empirical reasons to suspect whether political ideology really does have a direct influence on charitable behavior. Recent research in cultural sociology and social psychology calls into question the extent to which beliefs and ideology explain action. These studies conclude that beliefs only correlate weakly with behavior and are often situation- dependent. Social action, according to several proponents of this approach, can often be better explained by examining the repertoires of practices, habits, and schemas accumulated over time and “activated” by environmental cues (Chaves 2010; DiMaggio 1997; Lizardo and Strand 2010; Swidler 2001). Such a view need not assume that values and ideology are irrelevant. First, as Vaisey (2009:1686–87) has recently argued, certain kinds of cultural schemas—particularly, values that are deeply ingrained, implicit, and unconscious—can be strongly motivating, even though people may not be able to articulate them. But such schemas are not the subject of rational deliberation; rather, they are largely internalized through slow accumulation in our “practical consciousness.” In other words, using Jonathan Haidt’s (2006) memorable analogy of the elephant and rider, these cultural schemas primarily motivate action by shaping the elephant, our “practical consciousness,” which is what governs our actions most of the time. The rider, our “discursive consciousness,” comes into play when deliberation and justification are demanded by situations, but this is not characteristic of the bulk of our actions. This view is developed further in the discussion of “practice theory” below.

The implication of this approach is that scholars need to be wary of explanations of charitable giving behavior based primarily on political ideology. If political ideology influences giving, it will largely be contingent on concrete institutional arrangements for the cultivation of cultural habits that, largely unconsciously, sustain generous dispositions. Claims about the purported causal relationship between political ideology and charitable giving (and the implications that conservatives therefore “care” more [Brooks 2006]) may therefore be misleading. Consistent with this critique, the little empirical work on the subject suggests that it may be religious participation, rather than political ideology, that explains this variation in giving outcomes (Gibson 2008:505; Theiss-Morse 2009:113).

The key question addressed in this article is: Does political ideology have a direct effect on charitable giving, or is its effect largely meditated by active participation in religious, civic, and political associations? In doing so, the analyses also take into account different recipients of charitable gifts—congregations, noncongregational religious causes, and secular causes—a distinction that has been ignored in most previous studies (but see Wilhelm 2010).

EXPLAINING CHARITABLE FINANCIAL GIVING

Both political liberals and conservatives express commitments to the poor and needy. While the prevailing stereotypes are of “bleeding heart” liberals who care about social welfare and tightwad conservatives concerned with little beyond individual liberty, recent research paints a rather different picture. Notably, Brooks’s (2006) study reveals that when it comes to giving either money or time, political conservatives turn out to be significantly more generous than liberals. In addition, Brooks and Lewis (2001) find that people who express less confidence in the government (i.e., presumably conservatives) are more likely to give time and money to both religious and nonreligious nonprofits than those who express a great deal of confidence. Still another study shows liberals to be less likely than conservatives to donate money but more likely to express willingness to pay taxes toward helping the needy (Kim, Jung, and Oh 2008).

452 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

A central explanatory mechanism suggested from such studies is that liberals are less generous because they believe that it is the government’s responsibility to care for the poor, whereas conservatives are more generous because they believe that caring for the poor is the responsibility of individuals and not of the government (Brooks 2006). This can be considered an ideology- based explanation for giving. Political ideology, as Tedlin (1987:65) proposes, can be defined as “an interrelated set of attitudes and values about the proper goals of society and how they should be achieved.” The relevant elements advanced in the argument here pertain to the role of government in redistributing wealth (Brooks 2006; Brooks and Lewis 2001). While there has been some debate about whether coherent political ideologies exist, recent studies show a consistency between political ideology and affiliation: in general, people who support such a role for government consistently self-identify as liberal, while those who oppose such a view consider themselves political conservatives (see Jost 2006:656–57). But, is such political ideology really a proximate cause of charitable giving?

Ideology, Practice, and Action

Recent research and theorizing at the nexus of cultural sociology, social psychology, and cognitive neuroscience cast doubt on the direct causal relationship between ideology and behavior. An ideology-driven account, or more generally, a values-driven account, was in fact the dominant approach for much of the past century in explaining the role of culture in shaping action (Parsons and Shils 1951; Weber 1930/2001). In recent years, however, cultural sociologists have begun to criticize such accounts for being overly deterministic in positing necessary connections between values and action. Such criticism seems warranted especially in light of studies showing that people can appeal to the same ideology to support radically different courses of action (e.g., Hart 1992), and that people are often unable to articulate any consistent ideological explanations for their behavior (Chaves 2010; Swidler 2003). People’s espoused values, some scholars contend, are often merely “rationales for predetermined ends” (Kaufman 2004:340).

An alternative account stresses the primacy of practices over ideas in shaping action out- comes, which reflects a new theoretical turn in sociology (Schatzki, Knorr-Cetina, and von Savigny 2001). Rather than a cohesive theoretical framework, “practice theory” denotes a diverse multiplicity of approaches that, to varying degrees, emphasize the causal centrality and priority of habits, skills, and routine activities in determining various outcomes that previous theories attributed to, for example, “rational choice” calculations or internalized value systems (Biernacki 1995; Lizardo and Strand 2010).

Perhaps the most famous example of a practice approach in American sociology is Ann Swidler’s argument that it is not through internalized values but through people’s repertoires of skills and habits that culture shapes action.2 Culture here functions as a “toolkit” composed of “symbols, stories, rituals, and world-views, which people may use in varying configurations to solve different kinds of problems” (Swidler 1986:273). People’s actions, then, are not primarily driven by the values or ideologies they espouse; rather, people tend to act in a manner consistent with their existing “cultural equipment.” Culture serves as a “style or set of skills and habits” used by individuals to devise “strategies of action” (Swidler 1986:275, 276). In Swidler’s account, “the causal claim is that people are more likely to act in ways that utilize their skills than in ways that enhance their values” (Swidler 2008:616).

2 By highlighting Swidler’s approach here, we do not intend to suggest that there is any consensus among “practice theorists” that supports her approach. There is certainly a great deal of variation in what scholars identify as constitutive elements of practices as well as their causal effects (Schatzki, Knorr-Cetina, and von Savigny 2001). Swidler’s perspective, as Lizardo and Strand (2010) argue, is a “weak” form of practice theory. Yet, we draw on it because of its ubiquity and popularity in contemporary American sociology, and because of its central emphasis on the importance of people’s repertoires of skills, habits, and routines in shaping action—a claim that most practice theorists would accept.

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 453

This, however, does not eliminate the role of values and ideology. For example, Vaisey (2008) notes that values need to be taken into consideration in order to explain why people only deploy certain elements from their cultural repertoires at any given time. Recent scholarship proposes a dual-process account of cognition that considers the relationship of conscious and unconscious processes in shaping motivation, judgment, and action (Cerulo 2010; Vaisey 2009). Such an approach can explain people’s frequent lack of articulate ideological explanations for their behavior, and points to the importance of factors such as implicit schemata—mental cate- gories that are implicit and not verbalized (DiMaggio 1997; Vaisey 2009).3 The primary content of socialization, in this view, is not propositional ideological beliefs about which people have discursive consciousness, but, rather, skills, habits, schemas, and codes, learned through expe- rience and stored in procedural memory, operating through automated rather than deliberative cognition (DiMaggio 1997; Lizardo and Strand 2010). Most versions of practice theory do not completely dismiss the possibility that explicit values and ideology can shape action. Ideology can still have a role to play, but in a more contingent manner, depending on the context and existing repertoires. In particular, it is during “unsettled times” (Swidler 1986, 2003) or contexts in which “institutional prescriptions” are weak or breaking down (Lizardo and Strand 2010) that ideologies matter, particularly for the construction of new repertoires of practices. But in much of ordinary life, practice theorists argue, their role is less important than that of repertoires of routines, skills, and habits.

Extending these theories to the realm of philanthropy, one might expect that political ide- ologies only lead to charitable giving behavior when they are coupled with the proper cultural “toolkit.” Beliefs about the role of government in public life, divorced from the cultivation of practices and skills related to giving, are unlikely to lead to philanthropic behavior. But what are the contexts in which such repertoires and toolkits are acquired?

Cultivating Charitable Giving

Diverse institutional settings can contribute to the acquisition and habituation of practices of charitable giving. Much previous research has emphasized the importance of religious congrega- tions in particular. Religious congregations provide a context in which people are made aware of the needs of others in the community, are recruited into social networks, and become involved in planning and carrying out charitable activities (Chaves 2004; Hodgkinson 1990; Wuthnow 1991).

All major world religions emphasize the importance of charitable giving to the poor and needy, and some religious traditions even institutionally mandate the practice of regular charitable giving. Many religious institutions teach people attitudes of caring and compassion for others, and provide the context to mobilize these attitudes into concrete action (Wuthnow 1991). In addition, giving in a congregational context is repetitive in nature, with giving opportunities occurring on at least a semi-regular basis. For most congregations, soliciting funds to support the institution is carried out weekly and is part of the ritual of worship itself. This is distinct from the one-time charitable gifts that are typical of many forms of nonreligious giving. Once these dispositions, skills, and habits become part of an individual’s repertoire (e.g., writing a check to the church on a weekly basis), it becomes easier to extend this repertoire into other giving domains (e.g., giving regularly to secular charities). Similar to other cultural practices of religious institutions in the United States, the practice of charitable giving in this context may be transposable to other contexts (Chaves 2004:142–43).

3 There is considerable variation in how scholars understand what constitute “schemas” or “schemata”: for example, whether they are to be understood as primarily linguistic, nonlinguistic, or embodied (see the differences between DiMaggio 1997; Lizardo and Strand 2010; and Vaisey 2009). We do not need to enter into these discussions here, but focus on common elements across accounts.

454 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

This view of congregations is consistent with past literature that demonstrates the strong link between religiosity and charitable giving (see Bekkers and Wiepking 2007:6–7 for a summary). Despite the focus of Brooks’s (2006) book, his own earlier research suggests that the relationship between conservative ideology and charitable giving might be primarily driven by religion. In a previous study (2003), after controlling for nonreligious factors—gender, education, income (and thus, tax incentives), age, race, marital status, region, and political beliefs—Brooks still finds significant differences between frequent and infrequent church attendees in their giving habits, which he admits to being “more due to religion than to politics” (2003:42). These findings support other studies finding that political affiliation is not statistically significant net of religious factors when it comes to giving money to charities that serve the poor and needy (Regnerus, Smith, and Sikkink 1998:489). If political liberals who go to church behave like conservatives who go to church when it comes to their financial giving, then the explanation for differences in charitable giving seems to be primarily about the effects of going to church, rather than of the influence of political ideology.

More importantly, as Brooks (2003:46) argues, “charitable behavior is habitual.” In the absence of a context in which to learn and develop such generous habits, people are less likely to adopt and sustain the practice of charitable giving. People who regularly attend religious institutions might thus be more likely to develop the habitual practice of financial giving. Since such religious people are considerably more likely to consider themselves politically conservative rather than liberal (Brooks 2003:42), this might help explain the higher charitable giving of conservatives.

Brooks, however, claims that because religious people are more likely to be politically conservative, political liberals lack a key source of financial generosity. Secularism, he argues, is antithetical to charitable giving because religion is a key source of social capital, which he defines, following Putnam, Fukuyama, and others, as “the stock of trust and social cohesiveness that promotes giving, volunteering, and participation in civil society” (Brooks 2003:46–47). But even if this tautological conceptualization of social capital is accepted (see Fishman 2009 and Portes 2010:30–31 for a critique of problems with such an approach), it is not clear that religion is the only source of such community bonds and sentiments. Regular involvement in various kinds of political and civic activities—whether through one’s congregation (Chaves 2004:196; Regnerus, Smith, and Sikkink 1998) or through political or community organizations—may also foster a commitment to supporting charitable giving. Participation in community groups, school boards, and local civic and political groups can give people a sense of ownership in these groups, improve their awareness of pressing local needs that need financial support, and can thus generate a commitment to financial giving.

Our emphasis on the contexts in which individuals develop habits and practices of financial giving results in specific predictions about the relationship between political identification, reli- giosity, and charitable giving. In statistical models that assess the effects of political ideology on various forms of generosity, active involvement in religious, political, and civic practices will be a substantial predictor of all forms of financial generosity, while political ideology will have no independent influence.

METHOD

Data

Data from the first wave of the Panel Study on American Ethnicity and Religion (PS-ARE) are used to assess whether religious, political, and civic practices explain the relationship between political ideology and charitable giving, PS-ARE is a multilevel panel study focused on religion and ethnic and racial diversity in the United States (Emerson and Sikkink 2006). The PS-ARE

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 455

covers the civilian, noninstitutionalized household population in the continental United States who were 18 years of age or older at the time the survey was conducted, and speak either English or Spanish. The sample design and interviews were conducted by RTI International, which estimates that the combined sampling frame accounts for over 98 percent of the occupied housing units in the United States. The study used a four-stage sampling procedure. Census data were used to construct a nationally representative sampling frame of three-digit Zip Code Tabulation Areas. Two five-digit zip codes were then selected from each area. One hundred addresses, on average, were selected from each selected zip code, and digital maps were generated for a subsample of selected addresses. RTI then randomly selected one individual per selected housing unit for interview. Face-to-face interviews were conducted with 2,610 respondents. The PS-ARE yielded a contact rate of 83 percent, a screening completion rate of 86 percent, a cooperation rate of 82 percent, and an overall response rate of 58 percent. For details confirming the national representativeness of the PS-ARE sample, see Emerson, Sikkink, and James (2010:164–66). Multiple imputation methods available in Stata are used to impute missing data. The final N for our models ranges from 2,389 to 2,466 depending on the outcome measure.4

Measures

Dependent Variables Three dependent variables are examined. The first variable is the natural log of the total

donations that the respondent made to his or her local religious congregation in 2005. The second variable is the natural log of the total donations made to religious organizations in 2005. The third variable is the natural log of the total donations made to nonreligious charitable organizations in 2005. Respondents were asked the following questions to measure these outcomes: (1) For congregational donations: “During the year 2005, what was the total dollar value of all donations, if any, that you and your spouse/partner made to your local congregation. Donations include any gifts of money, assets, or property.” (2) For noncongregational donations: “During the year 2005, what was the total dollar value of all donations you and your spouse/partner made to other charitable organizations not including your congregation? Charitable organizations include religious or nonprofit organizations that help those in need or that serve public interests. Please do not include political contributions.” (3) To assess noncongregational religious vs. secular giving, a further question was asked: “Of the amount you just reported shown below, what percentage went to religious organizations or groups outside of your congregation?”

Independent Variables Political ideology is measured using a four-category variable indicating whether respondents

thought of themselves as politically liberal, moderate, conservative, or unaffiliated.5 Categories are entered as separate dummy variables, with liberal omitted as the reference group. Religious service attendance is an eight-category ordinal variable measuring how often respondents reported attending worship services, with values ranging from “never” to “three times a week or more.” Religious ideology is measured using a seven-category ordinal variable indicating whether respon- dents identify themselves as liberal, moderate, or conservative on religious matters (ranging from

4 It is standard practice in such situations to not impute data for the dependent variable (Allison 2002; Cohen et al. 2003). Differences in nonresponse on the giving outcomes explain the divergent Ns in our models. 5 Respondents were asked whether, when it came to politics, they thought of themselves as “very liberal,” “somewhat liberal,” “middle of the road,” “somewhat conservative,” “very conservative,” or “haven’t thought much about this.” Be- cause only a small minority of respondents identified themselves with the extreme cases (9 percent as “very conservative” and 7 percent as “very liberal”), the extreme categories of liberal and conservative were collapsed. Those who consider themselves “middle of the road” as “moderate” and “haven’t thought much” are labeled as “unafilliated.”

456 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

1 = very conservative, . . . 4 = moderate, . . .7 = very liberal). Importance of religious faith is an indicator of how important the respondent personally considers religious faith to be (ranging from 1 = not at all important to 5 = most important). The models with the outcome of noncongrega- tional giving also include the logged congregational giving variable as an independent predictor.

The effects of other practices and activities that might affect financial giving outcomes are also assessed. These variables can also to some degree control for a “joiner effect” (see Regnerus and Smith 2005), whereby some people might simply have a tendency to get involved in various associations and activities, which could explain both their religious participation and their charitable giving. A measure of the number of times in the past year the respondent has attended meetings about local government, school, or community affairs is included along with a measure of the number of times in the past year the respondent has volunteered for a local community project (both variables ranging from 0 = never to 7 = more than once a week). Political activity is measured by counting the number of the following political activities the respondent participated in during the past year: signing a petition, working for a candidate or party, contacting a public official, trying to influence someone’s vote, attending a political meeting or rally,6 or participating in a demonstration or protest. This measure can also help assess whether any mediation of political ideology by religious factors is in fact attributable to active political engagement. The following controls are included: gender, race, educational attainment, region, age, work status (full-time vs. part-time), marital status (married vs. unmarried), household size, and self-reported health. In addition, respondent income is measured using the natural log of the midpoints of closed-ended response categories. Table 1 presents descriptive statistics of the variables in our analyses.

Analysis Since the dependent variables measure total-dollar amounts donated, they are left-censored

at zero7—many people did not donate anything during the year, and it is not possible to donate a negative number. Ordinary least squares parameter estimates in such cases would be biased and inconsistent, and so it is standard practice in these cases to use the Tobit model (e.g., Wilhelm 2010), which takes the form:

y∗i =Xi β+εi , where εi ∼ N(0; σ 2), and y∗ is a latent variable representing the underlying desire or propensity to donate that is observed for values greater than 0. The observed y is therefore defined by the following measurement equation: yi= y∗ if y∗ > 0; yi= 0 if y∗ ≤ 0. All models include the marginal effect of the independent variable on the expected value of the outcome for uncensored observations in brackets. These marginal effects are intuitively easier to interpret than the standard Tobit coefficients. With all other variables held at their mean, the marginal effect included in our analyses can be interpreted as the proportional increase in giving caused by a one unit change in the independent variable given that the dependent variable is not censored.

Four models are estimated for each outcome.8 All models include control variables (see Table 2). Following the approach of Wilhelm (2010), for each of the four models, we estimate (a)

6 An index of political activity was created that excluded attendance at a political meeting or rally because of concern over similarity to our measure of attendance at local educational or civic meetings. However, findings were robust and very similar across both measures in the final models. For this reason, the more complete measure of political activity is included, recognizing that the same event may be counted in two different variables for a minority of respondents. 7 The number of zeros in the first multivariate table is 1,217, the number of cases greater than zero is 921. The number of zeros in the second table is 595 (1,475 greater than 0). The number in the third table is 1,220 (865 greater than 0). 8 A technical appendix that includes a list of variable names, a correlation matrix, and the Stata output for Models 4a and 4b in Tables 3–5 is posted online at wileyonlinelibrary.com.

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 457

Table 1: Descriptive statistics of variables in analysis (see Table 2 for additional controls)

N Mean Std Dev Min Max

Total donations to congregation 2,466 2.63 3.29 0 11.29 Total donations to noncongregational

religious charities (log) 2,420 1.75 2.51 0 10.77

Total donations to noncongregational nonreligious charities (log)

2,383 3.34 2.69 0 11.51

Income (log) 2,610 10.59 .98 7.31 13.43

Political Ideology Liberal 2,610 .21 .41 0 1 Moderate 2,610 .24 .43 0 1 Conservative 2,610 .28 .45 0 1 “No thought” 2,610 .26 .44 0 1

Church attendance 2,610 3.59 2.24 1 8

Religious Tradition Evangelical Protestant 2,610 .27 .44 0 1 Black Protestant 2,610 .06 .23 0 1 Mainline Protestant 2,610 .14 .35 0 1 Catholic 2,610 .27 .44 0 1 Jewish 2,610 .02 .12 0 1 Other religion 2,610 .06 .24 0 1 Unaffiliated 2,610 .16 .36 0 1 Other Protestant 2,610 .04 .20 0 1

Religious ideology 2,610 3.76 1.99 1 7 Importance of religious faith 2,610 3.00 1.22 1 5 Meeting attendance 2,610 1.19 1.53 0 7 Volunteer activity 2,610 1.45 1.90 0 7 Political activity 2,610 .90 1.20 0 6

the marginal effects of independent variables on the probability of giving, and (b) the marginal effects on the log amount given, conditional on household giving being greater than zero. In each of the tables, we assess whether measures of religious, political, and civic practice explain the effect of political ideology. For each of the giving outcomes, we also include in the model measures of the other two giving outcomes; this examines the extent to which the practice of giving in one context is associated with giving in other contexts.

RESULTS

Table 3 shows the results from a Tobit regression of the natural log of total donations to a religious congregation on political ideology and religious service attendance. Models 1a and 1b estimate the effect of political ideology on giving to congregations, controlling for income along with other standard controls. Model 1a estimates marginal effects on the probability of giving, and Model 1b estimates marginal effects on log amount given, conditional on giving being greater than zero. We find here that those who identify as politically conservative give significantly more to religious congregations than those who consider themselves politically liberal, while other groups (moderates and unaffiliated) are not statistically significantly different from liberals. Among those who give to religious congregations, political conservatives give an estimated 123 percent more than political liberals.

458 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

Table 2: Descriptive statistics of additional control variables

N Mean Std Dev Min Max

Gender = Male 2,610 .48 .50 0 1 Married 2,610 .57 .49 0 1

Race White 2,610 .69 .46 0 1 Black 2,610 .11 .31 0 1 Hispanic 2,610 .12 .33 0 1 Asian 2,610 .04 .20 0 1 Pacific Islander 2,610 .00 .06 0 1 Native American 2,610 .00 .07 0 1 Mixed race 2,610 .02 .12 0 1 Other 2,610 .01 .11 0 1

Education 2,610 4.27 2.48 1 12

Region Northeast 2,610 .17 .37 0 1 Midwest 2,610 .24 .43 0 1 South 2,610 .34 .47 0 1 West 2,610 .26 .44 0 1

Age 2,610 45.19 16.48 18 80 Full-time work 2,610 .48 .50 0 1 Household size 2,610 3.07 1.55 1 10 Overall health 2,610 2.55 1.12 1 5

Models 2a and 2b introduce our measure of religious attendance. Upon doing so, the statistical significance of political conservative disappears, and the effect sizes of all the political ideology coefficients drop substantially. Religious attendance has a strong and statistically significant effect on congregational donations. For each unit increase in our eight-category measure of religious service attendance, we estimate a 73 percent increase in congregational giving. We also see that there is no significant direct effect of political ideology on congregational giving once we control for attendance.

In Models 3a and 3b, we introduce measures of political activity and frequency of volun- teering and participation in meetings in the local community. We find that meeting attendance and political activities have statistically significant effects on congregational giving outcomes, but the influence on congregational giving is small in comparison to attendance. A comparison of standardized coefficients suggests that the effect of religious service attendance is approximately eight times greater than the effect of political activity and nearly 11 times greater than the effect of frequently attending local civic or educational meetings.

In Models 4a and 4b, we add the effect of various religious controls—religious tradition, religious ideology, and the importance of religious faith. We also add the effect of giving to noncongregational religious and secular charities. We find that, net of all other factors, only those who are religiously unaffiliated are significantly less likely to donate to congregations than evangelical Protestants. Religious ideology has no independent effect, but importance of religious faith is a significant predictor of congregational giving. Giving to secular—but not to religious— noncongregational charities is also significantly associated with congregational giving. Overall, controlling for these religious factors as well as participation in political and community activities does not substantially alter the effect of religious attendance on congregational giving.

Table 4 examines donations to noncongregational religious charities. Models 1a and 1b estimate the effect of political ideology controlling for income, and find that people who identify

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 459

T ab

le 3:

W ei

gh te

d T

ob it

m od

el s

of do

na ti

on s

to co

ng re

ga ti

on (n

at ur

al lo

g) ,

t- va

lu e

in pa

re nt

he se

s: co

lu m

n (a

) es

ti m

at es

th e

m ar

gi na

l ef

fe ct

on th

e pr

ob ab

il it

y of

gi vi

ng ;

co lu

m n

(b )

es ti

m at

es th

e m

ar gi

na l

ef fe

ct on

th e

lo g

am ou

nt gi

ve n,

co nd

it io

na l

on th

at am

ou nt

be in

g gr

ea te

r th

an ze

(1 a)

(1 b)

(2 a)

(2 b)

(3 a)

(3 b)

(4 a)

(4 b)

In co

m e

(l og

) .0

73 ∗∗

∗ .4

33 ∗∗

∗ .0

68 ∗∗

∗ .4

56 ∗∗

∗ .0

62 ∗∗

∗ .4

16 ∗∗

∗ .0

54 ∗∗

∗ .3

69 ∗∗

∗

(4 .8

6) (4

.8 2)

(5 .5

4) (5

.5 5)

(5 .2

0) (5

.2 1)

(4 .5

0) (4

.5 1)

P o li

ti ca

l id

eo lo

g ya

M od

er at

e .0

38 .1

97 −.

00 8

−. 05

3 −.

00 0

−. 00

3 −.

01 4

−. 09

2 (1

.0 7)

(1 .0

7) (−

.3 1)

(− .3

1) (−

.0 2)

(− .0

2) (−

.5 1)

(− .5

1) C

on se

rv at

iv e

.2 04

∗∗ ∗

1. 23

∗∗ ∗

.0 26

.1 75

.0 34

.2 23

−. 00

2 −.

01 6

(6 .5

4) (6

.7 6)

(1 .0

1) (1

.0 3)

(1 .3

3) (1

.3 5)

(− .0

8) (−

.0 8)

U na

ffi li

at ed

.0 49

.2 59

−. 02

0 −.

12 7

.0 10

.0 65

−. 00

4 −.

02 5

(1 .4

5) (1

.4 5)

(− .6

9) (−

.6 9)

(. 35

) (.

35 )

(− .1

3) (−

.1 3)

R el

ig io

us at

te nd

an ce

.1 10

∗∗ ∗

.7 30

∗∗ ∗

.1 07

∗∗ ∗

.7 14

∗∗ ∗

.0 91

∗∗ ∗

.6 15

∗∗ ∗

(4 5.

05 )

(3 0.

18 )

(3 9.

38 )

(2 7.

78 )

(2 0.

20 )

(1 8.

09 )

P ol

it ic

al ac

ti vi

ty .0

24 ∗∗

∗ .1

61 ∗∗

∗ .0

22 ∗∗

∗ .1

52 ∗∗

∗

(3 .6

0) (3

.6 1)

(3 .4

4) (3

.4 6)

M ee

ti ng

at te

nd an

ce .0

15 ∗∗

.0 98

∗∗ .0

12 ∗

.0 81

∗

(2 .8

4) (2

.8 3)

(2 .1

7) (2

.1 7)

V ol

un te

er in

g .0

01 .0

09 .0

01 .0

06 (.

32 )

(. 32

) (.

23 )

(. 23

) R

el ig

io u s

tr a d it

io n

B la

ck P

ro te

st an

t −.

02 3

−. 15

2 (−

.6 1)

(− .6

2) M

ai nl

in e

P ro

te st

an t

−. 00

3 −.

01 8

(− .1

0) (−

.1 0)

C at

ho li

c −.

01 5

−. 09

8 (−

.6 9)

(− .6

9) Je

w is

h −.

03 9

−. 25

7 (−

.6 3)

(− .6

(C o

n ti

n u

ed )

460 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

T ab

le 3

(C o n ti

n u ed

)

(1 a)

(1 b)

(2 a)

(2 b)

(3 a)

(3 b)

(4 a)

(4 b)

O th

er re

li gi

on −.

08 1

−. 51

1 (−

1. 87

) (−

1. 96

) U

na ffi

li at

ed −.

13 1∗

∗ −.

79 9∗

∗∗

(− 3.

12 )

(− 3.

42 )

O th

er P

ro te

st an

t −.

07 3

−. 46

3 (−

1. 40

) (−

1. 47

) R

el ig

io us

id eo

lo gy

(c on

se rv

at iv

e– li

be ra

l) −.

00 4

−. 03

0 (−

.8 3)

(− .8

3) Im

po rt

an ce

of re

li gi

ou s

fa it

h .0

25 ∗∗

.1 71

∗∗

(2 .9

9) (2

.9 9)

N on

co ng

re ga

ti on

al re

li gi

ou s

gi vi

ng (l

og )

.0 01

.0 06

(. 27

) (.

27 )

S ec

ul ar

gi vi

ng (l

og )

.0 15

0∗ ∗∗

.1 02

∗∗ ∗

(3 .7

0) (3

.6 9)

N 24

66 24

66 P

se u d o

R 2

.0 5

.1 8

.1 9

.2 0

∗ p <

0. 05

,∗ ∗ p

< 0.

01 ,∗

∗∗ p

< 0.

00 1.

a R

ef er

en ce

ca te

go ry

is li

be ra

l. b R

ef er

en ce

is ev

an ge

li ca

l P

ro te

st an

t. N

o te

: A

ll m

od el

s co

nt ro

l fo

r ge

nd er

,r ac

e, ed

uc at

io na

l at

ta in

m en

t, re

gi on

,a ge

,w or

k st

at us

,m ar

it al

st at

us ,

ho us

eh ol

d si

ze ,a

nd se

lf -r

at ed

he al

th (n

ot sh

ow n,

bu t

av ai

la bl

e up

on re

qu es

t) .

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 461

T ab

le 4:

W ei

gh te

d T

ob it

m od

el s

of do

na ti

on s

to no

nc on

gr eg

at io

na l

re li

gi ou

s ch

ar it

ie s

(n at

ur al

lo g)

, t-

va lu

e in

pa re

nt he

se s:

co lu

m n

(a )

es ti

m at

es th

e m

ar gi

na l

ef fe

ct on

th e

pr ob

ab il

it y

of gi

vi ng

; co

lu m

n (b

) es

ti m

at es

th e

m ar

gi na

l ef

fe ct

on th

e lo

g am

ou nt

gi ve

n, co

nd it

io na

l on

th at

am ou

nt be

in g

gr ea

te r

th an

ze ro

(1 a)

(1 b)

(2 a)

(2 b)

(3 a)

(3 b)

(4 a)

(4 b)

In co

m e

(l og

) .0

55 ∗∗

∗ .2

37 ∗∗

∗ .0

54 3∗

∗∗ .2

37 ∗∗

∗ .0

47 ∗∗

∗ .2

06 ∗∗

∗ .0

12 .0

52 (3

.7 5)

(3 .7

1) (3

.8 9)

(3 .8

4) (3

.3 5)

(3 .3

1) (.

87 )

(. 87

)

P o li

ti ca

l id

eo lo

g ya

M od

er at

e .0

41 .1

71 .0

24 .1

05 .0

27 .1

15 .0

09 .0

42 (1

.2 0)

(1 .2

1) (.

73 )

(. 73

) (.

81 )

(. 81

) (.

30 )

(. 30

) C

on se

rv at

iv e

.1 18

∗∗ ∗

.5 13

∗∗ ∗

.0 48

.2 12

.0 49

.2 13

.0 12

.0 52

(3 .6

6) (3

.7 1)

(1 .4

8) (1

.4 9)

(1 .5

2) (1

.5 3)

(. 34

) (.

34 )

U na

ffi li

at ed

.0 12

.0 47

−. 01

4 −.

05 9

.0 12

.0 50

.0 03

.0 14

(. 34

) (.

34 )

(− .4

1) (−

.4 1)

(. 34

) (.

34 )

(. 10

) (.

10 )

R el

ig io

us at

te nd

an ce

.0 49

∗∗ ∗

.2 12

∗∗ ∗

.0 44

∗∗ ∗

.1 91

∗∗ ∗

.0 25

∗∗ ∗

.1 13

∗∗ ∗

(1 0.

10 )

(9 .6

5) (8

.7 6)

(8 .4

4) (3

.8 2)

(3 .7

8) P

ol it

ic al

ac ti

vi ty

.0 11

.0 49

.0 03

.0 11

(1 .1

1) (1

.1 1)

(. 27

) (.

27 )

M ee

ti ng

at te

nd an

ce .0

28 ∗∗

∗ .1

24 ∗∗

∗ .0

22 ∗∗

.0 98

∗∗

(3 .8

7) (3

.8 5)

(3 .2

4) (3

.2 4)

V ol

un te

er in

g .0

10 .0

46 .0

06 .0

29 (1

.7 3)

(1 .7

3) (1

.1 3)

(1 .1

R el

ig io

u s

tr a d it

io n

B la

ck P

ro te

st an

t −.

04 3

−. 18

9 (−

1. 00

) (−

1. 02

) M

ai nl

in e

P ro

te st

an t

−. 00

9 −.

04 2

(− .2

6) (−

.2 6)

C at

ho li

c .0

14 .0

66 (.

52 )

(. 52

)

(C o

n ti

n u

ed )

462 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

T ab

le 4

(C o n ti

n u ed

)

(1 a)

(1 b)

(2 a)

(2 b)

(3 a)

(3 b)

(4 a)

(4 b)

Je w

is h

.0 22

.1 00

(. 29

) (.

29 )

O th

er re

li gi

on −.

06 6

−. 28

6 (−

1. 15

) (−

1. 18

) U

na ffi

li at

ed −.

09 6∗

−. 40

8∗

(− 2.

38 )

(− 2.

45 )

O th

er P

ro te

st an

t .0

21 .0

97 (.

37 )

(. 37

) R

el ig

io us

id eo

lo gy

(c on

se rv

at iv

e– li

be ra

l) −.

00 7

−. 03

2 (−

1. 16

) (−

1. 16

) Im

po rt

an ce

of re

li gi

ou s

fa it

h .0

31 ∗∗

.1 41

∗∗

(2 .8

5) (2

.8 4)

C on

gr eg

at io

na l

gi vi

ng (l

og )

.0 00

.0 02

(. 09

) (.

09 )

S ec

ul ar

gi vi

ng (l

og )

.0 54

∗∗ ∗

.2 44

∗∗ ∗

(1 2.

53 )

(1 1.

23 )

N 24

15 24

15 R

2 .0

4 .0

5 .0

6 .0

6 .1

0 .1

0 ∗ p

< 0.

05 ,∗

∗ p <

0. 01

,∗ ∗∗

p <

0. 00

1. a R

ef er

en ce

is li

be ra

l. b R

ef er

en ce

is ev

an ge

li ca

l P

ro te

st an

t. N

o te

: S

ee fu

ll no

te to

T ab

le 3.

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 463

as politically conservative donate significantly more money to religious charities than political liberals. Introducing the effect of religious service attendance in Models 2a and 2b, however, we find that the statistical significance of being conservative disappears. Religious service attendance seems to explain the effect of political ideology on noncongregational religious giving outcomes. With each unit increase in our measure of religious participation, individuals are predicted to give approximately 21 percent more to noncongregational religious causes.

In Models 3a and 3b, we introduce the effects of political and community involvement. We find that frequency of attending meetings in the community has a positive and highly significant effect on noncongregational religious giving, but the effect of political activity and volunteering is close to zero. Adding these variables produces little change to the coefficient of religious attendance, whose effect size (standardized coefficients not shown) is more than twice that of meeting attendance.

In Models 4a and 4b, we introduce controls for other religious factors as well as charitable giving to other causes (congregational and secular). These controls reduce the effect size of religious attendance by approximately half, but do not change its significance level. Separate analyses reveal that this decline in effect size is largely due to the inclusion of self-rated importance of religious faith. One of the ways that worship service attendance is associated with higher noncongregational giving is through an increase in the importance of religion in the day-to-day lives of attendees. Similar to the results of the regressions on congregational donations above, only the religiously unaffiliated are significantly less likely to donate to religious charities in comparison to evangelicals, and the importance of religious faith has a positive and significant effect. Donation to religious congregations, net of other factors, does not appear to be associated with giving to noncongregational religious causes. However, giving to secular causes has a statistically significant and sizable positive effect on this outcome (a 10 percent increase in secular giving translates into a 2.4 percent increase in noncongregational religious giving), although we think caution should be extended in interpreting these effects. We elaborate on this issue further in the discussion section.

In Table 5 we examine financial donations to noncongregational, nonreligious charities. Models 1a and 1b again estimate the effect of political ideology while controlling for income and standard controls. For this outcome, we see that it is only the politically unaffiliated—people who claim they “haven’t thought much about” their political affiliation—who are statistically significantly different from political liberals, and they report giving considerably less money to secular charitable organizations.

Introducing religious service attendance in Models 2a and 2b slightly increases the statistical significance and negative coefficient of the political unaffiliated. Attendance itself has a highly significant effect, with each unit increase in our measure estimating an 8 percent increase in giving to noncongregational nonreligious organizations. However, entering the effects of political and community involvement in Models 3a and 3b reduces the coefficient size of attendance and considerably lowers the statistical significance. Frequent volunteering in the community and involvement in a number of political activities are significantly associated with charitable giving to nonreligious organizations. Notably, with each new political activity out of a possible six activities we consider here, individuals give 21 percent more to these organizations. An analysis of the standardized coefficients (not shown) suggests that the effect of political activity is more than twice that of religious service attendance. Introducing these factors also diminishes the effect size of being politically unaffiliated by nearly half and eliminates its statistical significance, suggesting that active participation in community and political activities explains why political liberals give considerably more money to secular charities than people who have not thought much about politics.

Introducing controls for other religious factors as well as giving to congregational and religious causes in Models 4a and 4b reveals that when it comes to donating to secular charities, people who are Jewish are significantly more likely than evangelical Protestants to give and to

464 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION T

ab le

5: W

ei gh

te d

T ob

it m

od el

s of

do na

ti on

s to

no nc

on gr

eg at

io na

l, no

nr el

ig io

us ch

ar it

ie s

(n at

ur al

lo g)

,t -v

al ue

in pa

re nt

he se

s: co

lu m

n (a

) es

ti m

at es

th e

m ar

gi na

l ef

fe ct

on th

e pr

ob ab

il it

y of

gi vi

ng ;

co lu

m n

(b )

es ti

m at

es th

e m

ar gi

na l

ef fe

ct on

th e

lo g

am ou

nt gi

ve n,

co nd

it io

na l

on th

at am

ou nt

be in

g gr

ea te

r th

an ze

(1 a)

(1 b)

(2 a)

(2 b)

(3 a)

(3 b)

(4 a)

(4 b)

In co

m e

(l og

) .0

71 ∗∗

∗ .5

43 ∗∗

∗ .0

72 ∗∗

∗ .5

46 ∗∗

∗ .0

66 ∗∗

∗ .5

00 ∗∗

∗ .0

50 ∗∗

∗ .3

77 ∗∗

∗

(6 .4

9) (6

.3 7)

(6 .5

6) (6

.4 3)

(6 .1

5) (6

.0 3)

(4 .7

6) (4

.6 7)

P o li

ti ca

l id

eo lo

g ya

M od

er at

e −.

00 8

−. 06

8 −.

01 2

−. 10

0 −.

00 2

−. 01

7 .0

03 .0

20 (−

.4 4)

(− .4

4) (−

.6 4)

(− .6

4) (−

.1 1)

(− .1

1) (.

14 )

(. 14

) C

on se

rv at

iv e

−. 00

8 −.

06 5

−. 02

3 −.

18 1

−. 01

3 −.

10 3

−. 01

0 −.

07 8

(− .4

5) (−

.4 5)

(− 1.

24 )

(− 1.

23 )

(− .7

2) (−

.7 2)

(− .4

9) (−

.4 9)

U na

ffi li

at ed

−. 05

0∗ −.

36 3∗

−. 05

5∗ ∗

−. 40

7∗ ∗

−. 02

5 −.

18 9

−. 02

0 −.

14 9

(− 2.

36 )

(− 2.

36 )

(− 2.

60 )

(− 2.

60 )

(− 1.

22 )

(− 1.

21 )

(− .9

8) (−

.9 8)

R el

ig io

us at

te nd

an ce

.0 11

∗∗ ∗

.0 83

∗∗ ∗

.0 07

∗ .0

51 ∗

−. 00

7 −.

05 4

(3 .6

2) (3

.6 8)

(2 .1

5) (2

.1 6)

(− 1.

72 )

(− 1.

72 )

P ol

it ic

al ac

ti vi

ty .0

27 6∗

∗∗ .2

10 ∗∗

∗ .0

23 ∗∗

∗ .1

70 ∗∗

∗

(4 .9

2) (5

.1 0)

(4 .0

8) (4

.1 6)

M ee

ti ng

at te

nd an

ce .0

07 .0

54 .0

02 .0

14 (1

.4 6)

(1 .4

6) (.

39 )

(. 39

) V

ol un

te er

in g

.0 12

∗∗ .0

88 ∗∗

.0 09

∗ .0

65 ∗

(2 .9

6) (2

.9 8)

(2 .2

6) (2

.2 6)

R el

ig io

u s

tr a d it

io n

B la

ck P

ro te

st an

t .0

19 .1

40 (.

60 )

(. 58

) M

ai nl

in e

P ro

te st

an t

.0 27

.2 07

(1 .2

9) (1

.2 7)

C at

ho li

c .0

25 .1

87 (1

.3 2)

(1 .3

3) Je

w is

h .0

73 ∗

.6 16

∗

(C o

n ti

n u

ed )

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 465

T ab

le 5

(C o n ti

n u ed

)

(1 a)

(1 b)

(2 a)

(2 b)

(3 a)

(3 b)

(4 a)

(4 b)

(2 .2

4) (1

.9 7)

O th

er re

li gi

on .0

20 .1

48 (.

64 )

(. 63

) U

na ffi

li at

ed .0

07 .0

48 (.

26 )

(. 26

) O

th er

P ro

te st

an t

−. 08

6 −.

52 3∗

(− 1.

91 )

(− 2.

12 )

R el

ig io

us id

eo lo

gy (c

on se

rv at

iv e

– li

be ra

l) .0

02 .0

18 (.

59 )

(. 59

) Im

po rt

an ce

of re

li gi

ou s

fa it

h −.

01 8∗

−. 13

4∗

(− 2.

51 )

(− 2.

50 )

C on

gr eg

at io

na l

gi vi

ng (l

og )

.0 11

∗∗ ∗

.0 80

∗∗ ∗

(3 .7

6) (3

.7 6)

N on

co ng

re ga

ti on

al re

li gi

ou s

gi vi

ng (l

og )

.0 33

∗∗ ∗

.2 45

∗∗ ∗

(1 1.

23 )

(1 3.

84 )

N 23

89 23

89 R

2 .0

7 .0

8 .0

8 .1

1 .1

1 ∗ p

< 0.

05 ,∗

∗ p <

0. 01

,∗ ∗∗

p <

0. 00

1. a R

ef er

en ce

is li

be ra

l. b R

ef er

en ce

is ev

an ge

li ca

l P

ro te

st an

t. N

o te

: S

ee fu

ll no

te to

T ab

le 3.

466 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

donate higher amounts, while people who identify as “other Protestant” give significantly less money. In addition, importance of faith has a negative net effect on donation to secular charities. We also find that donating to religious congregations as well as noncongregational religious charities are both significantly associated with donating to secular organizations, net of all other factors. The implications of these findings are discussed below.

DISCUSSION AND CONCLUSION

What explains the apparent association between political ideology and charitable giving? We have demonstrated, consistent with research that questions the direct relationship between ideology and social action, that intervening factors—particularly, religious, civic, and political participation and congregational giving—play an important role in explaining differences between adherents of different political ideologies. In pursuing this question, three separate philanthropic outcomes are examined: donation to congregations, donations to noncongregational religious organizations, and donations to noncongregational secular organizations.

The differences in giving between self-identified liberals and conservatives exist only for religious giving (both congregational and noncongregational). In other words, net of income and demographic controls, conservatives give more to religious causes. However, the inclusion of a single additional variable, religious service attendance, sufficiently explains the statistically significant difference between liberals and conservatives. In the case of congregational giving, we expect that the church-attendance effect is mostly due to frequent solicitation of funds, coupled with the ritualized nature of giving within most religious services. On the other hand, religious service attendance is related to noncongregational religious giving as well. This may be because information about such causes circulates during announcements or in the social interaction after church, and regular attendees may be likely to donate to these causes. To understand these findings better, further research will be needed on the nature of the various noncongregational religious charities that people support. Some of these charities may be locally based, whereas others may be national para-church organizations, or religious charities with strong political components (see, e.g., Regnerus, Smith, and Sikkink 1998). There may also be a selection effect here, especially with locally based organizations, driven by people who like to be actively involved in different types of religious groups in the community, and who see the need to financially support religious causes in addition to the congregation.

While it may not be surprising that religious factors account for the association between political ideology and donations to religious causes, we also find that involvement in political and community activities is also an important factor associated with religious giving outcomes. We believe that such a finding helps to further underscore the importance of our argument that it is the cultivation of habitual practices rather than simply the espousal of beliefs that sustains charitable giving.

When it comes to giving to secular charities, we find no statistically significant differences between political liberals and conservatives. Rather, it is the absence of political reflection and affiliation that has a statistically significant, though negative, effect. People who “haven’t thought much about” their political commitments donate significantly less money to secular charities than all other categories. Our findings suggest that part of the reason may be that such individuals are less involved in political and civic activities in the community, and also might be less committed to habitually donating money to congregations, in comparison to political liberals. Previous research suggested that it is political moderates who donate the least (Brooks 2006:192–93). But here we show that it may be the case that some of these previous studies had not allowed respondents to distinguish between being a political moderate versus being politically indifferent, and this distinction matters. Future research should pay greater attention to this difference in political categories.

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 467

We also find that religious, political, and civic practices are associated with secular giving. In our final model, the effect of church attendance becomes insignificant. Separate analyses suggest that this is largely due to the inclusion of measures of religious giving. In this way, giving to religious causes may serve as a type of practice that is transposable to other contexts of giving more generally. Although we think caution is warranted in interpreting causal effects in cross-sectional data, this interpretation of the relationship between religious and secular giving is consistent with the findings of Hill and Vaidyanathan (2011) using longitudinal data.

Thus, for all three categories of giving considered here (i.e., to congregational, noncongre- gational religious, and noncongregational nonreligious organizations), we find the importance of involvement in religious congregations. Echoing Hart (1992), we argue that the effect of religion should not be equated automatically with conservative political ideology. Our study suggests that, despite their being fewer in number, more attention should be paid to religious liberals.

We thus contest the argument that it is something about conservative or liberal ideology in itself that drives people to be generous or stingy (Brooks 2006). By examining different kinds of financial giving outcomes, we demonstrate that the relationship between political conservatism and financial giving is not the same across all kinds of giving. Our analyses confirm that religion is a vital component explaining the relationship between political affiliation and charitable giving. In addition, the findings strongly support recent sociological theorizing that stresses a practice- driven account of how culture shapes action, rather than an ideology-driven account (Lizardo and Strand 2010; Swidler 2001; Vaisey 2008). Thus, it is inaccurate to argue that conservative beliefs necessarily make people more generous or that that liberal beliefs make people less generous; rather, among conservatives and liberals alike, it is the regular involvement in religious, political, and civic practices that explains differences in generosity.

While previous research on congregations helps sheds light on the mechanisms through which involvement in religious communities fosters giving to both religious and secular causes, further research is needed in order to better understand how exactly participation in political activities and local community groups facilitates financial giving. Some of these groups and activities may be facilitated through religious congregations. But other organizations such as schools and various neighborhood and community associations might play an important role in helping foster an awareness of financial needs and sustaining people’s commitments to meeting them. It also may be the case that ideological commitment is a factor that shapes why people become involved in civic and political groups to begin with, since such involvement may be more likely to occur later in life, as opposed to religious congregations in which people usually become involved from an early age. Ideology may thus matter in a more distal and indirect way, by contributing to people’s self-selection into some of the institutional contexts in which they develop habits of charitable giving.

As with all analyses that use cross-sectional data, it is nearly impossible to establish causal order. One could argue that, for example, donating to congregations might be a cause of, rather than an effect of, religious participation. However, from what is known about religious participation generally, this is unlikely. It is more plausible that people learn giving behaviors in religious institutions first—which for many tends to be a center of childhood moral socialization—and then transfer them over to charitable giving activities (e.g., Chaves 2004; Wuthnow 1991), rather than the other way round. Such an account is also supported by qualitative studies on charitable givers who often explain their commitment as resulting from practices cultivated in families and churches from an early age (e.g., Vaidyanathan and Snell 2011). Echoing Brooks (2003), charitable giving is best understood as a habit that is learned through practice, in specific institutional settings.

While the data used in this study are only from one cross-sectional survey, the PS-ARE is intended as a panel study; therefore, research in future years can examine longitudinal effects building directly on our findings from the first wave. Future research can also assess the role of contextual factors such as social networks in mediating the relationship between political

468 JOURNAL FOR THE SCIENTIFIC STUDY OF RELIGION

affiliation, religion, and financial giving. Further qualitative research is certainly needed on how charitable giving is learned and sustained, both in religious congregations as well as in other settings, such as community groups and associations.

REFERENCES

Allison, Paul D. 2002. Missing data. Thousand Oaks, CA: Sage. Bekkers, René and Pamala Wiepking. 2007. Generosity and philanthropy: A literature review. Philanthropic Studies

Working Paper Series. Amsterdam, The Netherlands: VU University. Biernacki, Richard. 1995. The fabrication of labor: Germany and Britain, 1640–1941. Berkeley, CA: University of

California Press. Brooks, Arthur C. 2003. Religious faith and charitable giving. Policy Review 121(October):39–50. ——. 2006. Who really cares: The surprising truth about compassionate conservatism. New York: Basic Books. Brooks, Arthur C. and Gregory B. Lewis. 2001. Giving, volunteering, and mistrusting government. Journal of Policy

Analysis and Management 20(4):765–69. Cerulo, Karen A. 2010. Mining the intersections of cognitive sociology and neuroscience. Poetics 38(2):115–32. Charities Aid Foundation. 2006. Charity trends 2006. Available at http://www.cafonline.org/Default.aspx?page = 12187,

accessed October 1, 2010. Chaves, Mark. 2004. Congregations in America. Cambridge, MA: Harvard University Press. ——. 2010. Rain dances in the dry season: Overcoming the religious congruence fallacy. Journal for the Scientific Study

of Religion 49(1):1–14. Cohen, Jacob, Patricia Cohen, Stephen G. West, and Leona S. Aiken. 2003. Applied multiple regression/correlation

analysis for the behavioral sciences, 3rd edition. Mahwah, NJ: Erlbaum. Derin-Güre, Pinar and Neslihan Uler. 2010. Charitable giving under inequality aversion. Economics Letters 107(2):

208–10. DiMaggio, Paul. 1997. Culture and cognition. Annual Review of Sociology 23(1):263–88. Emerson, Michael O. and David Sikkink. 2006. Panel Study of American Religion and Ethnicity, 1st wave. Available at

http://www.ps-are.org/about/overview.asp Emerson, Michael O., David Sikkink, and Adele D. James. 2010. The Panel Study of American Religion and Ethnicity:

Background, methods, and selected results. Journal for the Scientific Study of Religion 49(1):162–71. Fishman, Robert M. 2009. On the costs of conceptualizing social ties as social capital. In Social capital: Multidisciplinary

perspectives, edited by Viva Bartkus and Jim Davis, pp. 66–83. Northampton, MA: Edward Elgar Press. Gibson, Troy. 2008. Religion and civic engagement among America’s youth. Social Science Journal 45(3):

504–14. Giving USA Foundation. 2010. Giving USA 2010. Indianapolis, IN: Giving USA Foundation: Center on Philanthropy at

Indiana University. Haidt, Jonathan. 2006. The happiness hypothesis: Finding modern truth in ancient wisdom. New York: Basic Books. Hart, Stephen. 1992. What does the Lord require? How American Christians think about economic justice. New Brunswick,

NJ: Rutgers University Press. Hill, Jonathan P. and Brandon Vaidyanathan. 2011. Substitution or symbiosis? Assessing the relationship between religious

and secular giving. Social Forces 90(1). Hodgkinson, Virginia Ann. 1990. The future of individual giving and volunteering: The inseparable link between

religious community and individual generosity. In Faith and philanthropy in America: Exploring the role of religion in America’s voluntary sector, edited by Robert Wuthnow, Virginia Hodgkinson, and associates, pp. 284–312. San Francisco, CA: Jossey Bass.

Johnstone, D. Bruce. 2005. A political culture of giving and the philanthropic support of public higher education in international perspective. International Journal of Educational Advancement 5(3):256–64.

Jost, John T. 2006. The end of the end of ideology. American Psychologist 61(7):651–70. Kaufman, Jason. 2004. Endogenous explanation in the sociology of culture. Annual Review of Sociology 30(1):

335–57. Kim, Jungbu, Taehyun Jung, and Seong Soo Oh. 2008. Funding for disaster recovery: Increased taxes or charitable

donations to nonprofits? Working Paper 08–01. Atlanta, GA: State University, Andrew Young School of Policy Studies Georgia.

Lizardo, Omar and Michael Strand. 2010. Skills, toolkits, contexts and institutions: Clarifying the relationship between different approaches to cognition in cultural sociology. Poetics 38(2):204–27.

Parsons, Talcott and Edward A. Shils. 1951. Toward a general theory of action. Cambridge, MA: Harvard University Press.

Portes, Alejandro. 2010. Economic sociology: A systematic inquiry. Princeton, NJ: Princeton University Press. Regnerus, Mark D. and Christian Smith. 2005. Selection effects in studies of religious influence. Review of Religious

Research 47(1):23–50.

RELIGION, CHARITABLE GIVING, AND POLITICAL IDEOLOGY 469

Regnerus, Mark D., Christian Smith, and David Sikkink. 1998. Who gives to the poor? The role of religious tradition and political location on the personal generosity of Americans toward the poor. Journal for the Scientific Study of Religion 37(3):481–93.

Schatzki, Theodore R., Karin Knorr-Cetina, and Eike von Savigny (eds.) 2001. The practice turn in contemporary theory. New York: Routledge.

Swidler, Ann. 1986. Culture in action: Symbols and strategies. American Sociological Review 51(2):273–86. ——. 2001. What anchors cultural practices. In The practice turn in contemporary theory, edited by Theodore R. Schatzki,

Karin Knorr-Cetina, and Eike von Savigny, pp. 83–102. New York: Routledge. ——. 2003. Talk of love: How culture matters. Chicago, IL: University of Chicago Press. ——. 2008. Comment on Stephen Vaisey’s “Socrates, Skinner, and Aristotle: Three Ways of Thinking about Culture in

Action.” Sociological Forum 23(3):614–18. Tedlin, Kent L. 1987. Political ideology and the vote. In Research in micropolitics, Vol. 2: Voting Behavior, edited by

Samuel Long, pp. 63–94. Greenwich, CT: JAI Press. Theiss-Morse, Elizabeth. 2009. Who counts as an American?: The boundaries of national identity. New York: Cambridge

University Press. Vaidyanathan, Brandon and Patricia Snell. 2011. Motivations for and obstacles to religious financial giving. Sociology of

Religion 72(2):189–214. Vaisey, Stephen. 2008. Reply to Ann Swidler. Sociological Forum 23(3):619–22. ——. 2009. Motivation and justification: Toward a dual-process theory of culture in action. American Journal of Sociology

114(6):1675–1715. Weber, Max. [1930] 2001. The Protestant ethic and the spirit of capitalism. Translated by Talcott Parsons. New York:

Routledge. White, Douglas E. 1998. The art of planned giving: Understanding donors and the culture of giving. New York: Wiley. Wilhelm, Mark O. 2010. Giving to organizations that help people in need: Differences across denominational identities.

Journal for the Scientific Study of Religion 49(3):389–412. Wolpert, Julian. 1995. Giving and region: Generous and stingy communities. New Directions for Philanthropic Fundrais-

ing 1995(7):11–30. Wuthnow, Robert. 1991. Acts of compassion: Caring for others and helping ourselves. Princeton, NJ: Princeton University

Press.

SUPPORTING INFORMATION

The following supporting information is available for this article:

Appendix S1. Stata output for Models 4a and 4b in Tables 3–5.

Supporting Information may be found in the online version of this article at wileyonlinelibrary. com.

Please note: Wiley-Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.

not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written

permission. However, users may print, download, or email articles for individual use.