Social Capital, Volunteering, and Charitable Giving

Original Paper

DOI: 10.1007/s11266-008-9055-y

Cite this article as:
Wang, L. & Graddy, E. Voluntas (2008) 19: 23. doi:10.1007/s11266-008-9055-y

Abstract

This paper explores the impact of social capital—measured by social trust and social networks—on individual charitable giving to religious and secular organizations. Using United States data from the national sample of the 2000 Social Capital Community Benchmark Survey, we find that social trust, bridging social network, and civic engagement increase the amount of giving to both religious and secular causes. In contrast, organizational activism only affects secular giving. Volunteering activity, and human and financial capital indicators positively affect both religious and secular giving. Finally, those who are happy about their lives and those who are religious give more to religious causes, but these factors do not affect secular giving. We find evidence of important differences in the determinants of religious and secular giving, suggesting the need to distinguish these two types of charitable giving in future work.

Keywords

Social capital Volunteering Religious giving Secular giving United States 

Résumé

Cet article examine l'impact du capital social - évalué en fonction du trust social et des réseaux sociaux - sur l’octroi charitable de personnes offrant leur soutien aux organisations religieuses et laïques. En utilisant les données disponibles aux États-Unis sur le sondage national, résultant de l'enquête de banc d’essai de la communauté du capital social, nous constatons que le trust social, les réseaux sociaux, et l’engagement civique accroît le montant des dons aux causes religieuses et laïques. Par opposition, l’activisme organisationnel ne profite qu’aux dons laïques. L’activité du volontariat et les indicateurs du capital humain et financier affectent positivement à la fois les dons religieux et laïques. En fin de compte, tous ceux et toutes celles qui sont heureux dans leur vie et tous ceux et toutes celles qui sont religieux offrent davantage au causes religieuses, mais ces facteurs n’affectent pas les dons aux causes laïques. Nous avons la preuve de différences importantes qui déterminent les dons religieux et laïques, ce qui permet de suggérer le besoin de distinguer deux types de contributions dans recherche à l’avenir.

Zusammenfassung

Dieser Beitrag untersucht die Auswirkung von sozialem Kapital - gemessen an sozialem Vertrauen und sozialen Netzwerken - auf die individuelle Bereitschaft zu wohltätigen Spenden an religiöse und nicht kirchliche Organisationen. Unter Bezugnahme auf U.S.-amerikanische Daten aus der landesweiten Stichprobe im Rahmen einer in 2000 von Bürgerstiftungen durchgeführten Umfrage zum Sozialkapital (2000 Social Capital Community Benchmark Survey) stellen wir fest, dass soziales Vertrauen, die Überbrückung des sozialen Netzwerks und Bürgerengagement die Spendenbereitschaft für religiöse und nicht kirchliche Zwecke erhöhen. Dagegen wirkt sich organisatorisches Engagement lediglich auf nicht kirchliche Spendenbereitschaft aus. Ehrenamtliche Tätigkeiten sowie Human- und Finanzkapitalindikatoren wirken sich sowohl auf religiöse als auch nicht kirchliche Spenden positiv aus. Und letztlich spenden glückliche und religiöse Menschen mehr für religiöse Zwecke, während diese Faktoren bei Spenden für nicht kirchliche Zwecke keine Rolle spielen. Wir sehen Anhaltspunkte für wesentliche Unterschiede zwischen den Bestimmungsaktoren religiöse und nicht kirchliche Spenden, was auf eine notwendige Unterscheidung dieser beiden Spendenarten bei zukünftigen Untersuchungen hinweist.

Resumen

Este trabajo analiza la influencia del capital social, ―medido por trusts y redes sociales― en las donaciones de caridad que hacen las personas a organizaciones religiosas y seculares. Utilizando datos sobre Estados Unidos tomados de una muestra nacional de la encuesta de referencia comunitaria de capital social 2000 (Social Capital Community Benchmark Survey), hemos descubierto que el trust social, la red social y el compromiso cívico incrementan la cantidad de donaciones a obras tanto religiosas como seculares. En contraste, el activismo de las organizaciones solo afecta a las donaciones seculares. Las actividades de voluntariado y los indicadores humanos y de capital social tienen un efecto positivo tanto en las donaciones religiosas como seculares. Por último, aquellos que están contentos con sus vidas y las personas creyentes dan más a las causas religiosas, factores éstos que no afectan a las donaciones seculares. Hemos hallado pruebas de la existencia de importantes diferencias en los factores que determinan las donaciones seculares, lo que apunta a la necesidad de distinguir entre estos dos tipos de donaciones para futuros trabajos.

摘要

篇报告研究了社会资本——由社会信任度和社会网络衡量——对于向宗教和非宗教机构进行个人慈善捐赠的影响。利用来自于2000年全美“社区社会资本基准”调查案例的数据,我们发现社会信任度,建立社会网络和公民责任感能促进对宗教和非宗教事业的捐赠。相比之下,组织积极行动只能影响非宗教捐赠。志愿行为,及人类和财政资本指数对宗教和非宗教捐献有着积极影响。最后,那些生活快乐和信仰宗教的人们向宗教事业捐赠更多,但这些因素不影响非宗教捐赠。我们发现了宗教和非宗教捐赠取决因素的差异,这表明在未来的研究中需要区分这两种慈善捐赠。

要約

本論文では、宗教団体および非宗教団体に対する個人のボランティア活動の提供における、社会的信用および社会的なネットワークから判断される社会資本への影響について論じる。2000 Social Capital Community Benchmark Survey(社会資本共同体の標準調査)の米国のデータのサンプルから、社会的信用、社会的ネットワークへの架け橋、市民の関与により、宗教的要因および非宗教的要因への提供が増加することがわかった。一方、 組織的な行動主義は世俗的な提供にのみ影響する。ボランティア 活動と財政的な資本インディケータは明らかに宗教的かつ世俗的な提供に影響する。最終的に、自分の人生に満足し、宗教的な要因にとらわれずに宗教的な提供を行う者に対しては、これらの要素は世俗的な提供に影響しない。宗教的かつ世俗的な提供の決定要因には明らかに重要な違いがあることが明らかになったが、今後の研究においては、この2つのタイプの慈善提供を区別する必要性を示していく。

ملخص

هذا البحث يستكشف تأثير رأس المال الإجتماعي – تم قياسه بالثقه الإجتماعيه و الشبكات الإجتماعيه – من شخص خيري يعطي لمنظمات دينيه وعلمانيه . بإستخدام معلومات الولايات المتحده الأمريكيه من عينه قوميه لدراسه نموذجيه لعام 2000 للرأس المال الإجتماعي للمجتمع٬ وجدنا أن الثقه الإجتماعيه ٬ ربط الشبكه الإجتماعيه ٬ المشاركه المدنيه يزيد كميه العطاء لأسباب دينيه و علمانيه . على النقيض ٬ الفعاليه التنظيميه تؤثر فقط على المنحه العلمانيه . النشاط التطوعي٬ و مؤشرات الرأس المال البشري و المالي تؤثر إيجابيا? في العطاء الديني و العلماني . أخيرا? ٬ أولئك الذين هم سعداء في حياتهم و الذين هم متدينون يعطون أكثر لأسباب دينيه ٬ هذه العوامل لا تؤثر على العطاء العلماني . و نجد أدله على الإختلافات الهامه في محددات العطاء الديني و العلماني ٬ مما يشير إلى الحاجه للتمييز بين هذين النوعين من العطاء الخيري في العمل المقبل .

Introduction

Charitable giving is a part of civic life in American society. According to Giving USA 2006, from 70 to 80 percent of Americans contribute annually to at least one charity, and the amount of individual charitable giving in the United States reached nearly $190 billion in 2004 (American Association of Fundraising Counsel 2005). In the past 10 years, the United States has witnessed a dramatic growth in the number of charitable organizations—about 904,000 501(c)(3) public charities are listed in the Internal Revenue Service’s official roster in 2006, a 70 percent increase from 1996 (National Center for Charitable Statistics 2007). This growth in charitable organizations coupled with the economic downturn at the beginning of the twenty-first century has raised concerns about intense fundraising competition among charities. In addition, governments around the globe increasingly acknowledge the important role of civil society in dealing with social needs, such as health, poverty, and the impact of global disaster. The level of giving is one indicator of the strength of civil society. In many counties, individual donation is a main source of funding for voluntary and community organizations and therefore the level of individual giving affects the contributions these organizations can make to society (Charities Aid Foundation 2006). In this context, understanding the factors that promote an individual’s willingness to give is critical to the growth and the financial strength of charitable organizations worldwide.

Literature Review

Due to the dominance of the rational choice view of human nature, which portrays human beings as self-interested, individual charitable behavior remains a puzzle to many social scientists. Contributing to the public good is not a rational choice because the donor cannot directly consume the benefits, and one additional contribution does not make a noticeable difference for the collective outcome (Bekkers 2004). Forty years ago, Olson (1965) pointed out the discrepancy between the rational choice view of human beings and observed unselfish behavior. This problem is now known as the “collective good problem” or the “participation paradox.” Individual charitable behavior has subsequently been of interest to scholars in the fields of economics, sociology, and psychology, and they have used different lenses to explain the behavior.

Influenced by rational choice theory, economists often explain charitable behavior based on the benefits people receive through donating, such as tax incentives and the effect of “warm glow” (Andreoni 1990; Brown and Landford 1992; Clotfelter 1985, 1997; Reece and Zieschang 1985). They draw an analogue between the consumption of certain goods and charitable giving. When the price of a donation falls (i.e., when the marginal tax rate falls), people choose to donate more. In addition, charitable behavior is not considered as purely altruistic. People donate because they enjoy the pleasure that derives from the act of making a gift. This economic explanation of charitable behavior, however, has flaws. First, the tax incentive argument fails to explain why donors who do not itemize their tax returns, and thus do not enjoy tax benefits, still choose to donate. Second, the economic model of charitable behavior pays little attention to the role of friendship, propinquity, and social networks in charitable giving (Clotfelter 1997). In other words, the purely economic explanation of charitable behavior overlooks the social construction of individual values. The expected benefits an individual enjoys through charitable donations are affected by his or her social networks, prior experiences, and organizational involvement, as well as other social and psychological factors.

In contrast to the economic perspective, the psychological explanation of charitable behavior links it to individual personalities and the perception of nonprofit organizations. The decision to contribute to collective goods is viewed as the result of pro-social or altruistic personality characteristics (i.e., agreeableness, extraversion, and emotional stability), the perceived efficacy of contributions, and feelings of empathy (Bekkers 2004; Schervish 1997). To explain individual philanthropic behavior, psychologists often use experimental methods. The environments of the experiments are carefully controlled so that scholars can focus on one or a few key factors that might affect charitable behavior. The experimental basis of these findings of psychological impacts on pro-social behavior, however, may limit their generalizabilitiy.

Another significant contribution to our understanding of charitable behavior comes from the sociological perspective, which emphasizes the importance of the social environment, norms, and social networks in promoting charity. Exposure to requests to donate, an individual’s organizational involvement, and community size are factors shown to be associated with the decision to contribute (Schervish and Havens 1997). The findings associated with the impacts of those factors on charitable behavior, however, are not consistent. For example, Schervish and Havens (1997) found that informal helping behavior and participation in organizations that serve as channels for giving and volunteering, especially religious organizations, are strongly related to giving behavior, but general levels of social participation are not. O’Neill and Silverman (2002), however, found that, for Californians, religious affiliation makes no difference in either the rate or level of giving, especially to secular agencies.

In view of the different theoretical explanations of charitable behavior and the discrepancies in the empirical results on charitable giving, further studies on this subject and a synthetic view of an individual’s decision to donate are necessary. Recent developments in the concept of social capital provide a new opportunity for us to examine the impact of social forces on charitable giving behavior.

Social capital refers to the “networks, norms, and social trust that facilitate coordination and cooperation for mutual benefits” (Putnam 1995, p. 67). Like its counterparts of financial and human capital, social capital has been increasingly applied as an analytical tool to explain variances in the performance of individuals and organizations (Oztas 2004). Empirical studies have demonstrated that social capital enhances the chances of individuals getting better jobs (Lin 2001), promotes economic development (Fukuyama 1995; Putnam 1993), helps government reform efforts (Putnam et al. 1993), improves school performance (Helliwell and Putnam 1999) and improves perceived health (Kawachi et al. 1999a). Halpern (2005) has further shown that the impact of social capital on economic performance, health and well-being, crime, education, and government effectiveness are evident at the individual, community, and national levels across the globe.

Only a handful of studies, however, have examined the relationship between social capital and individual philanthropic behavior—the voluntary contribution of time and/or money to collective goods. In addition, these studies have different interpretations of how social capital, volunteering and giving are related. For example, Narayan and Cassidy (2001) used volunteering and/or giving as one of the indices to measure social capital. However, Putnam argued: “Doing good for other people… is not part of the definition of social capital” (Putnam 2000, p. 117), and therefore we should separate philanthropic behavior from social capital. Goss (1999) treated volunteering as a function of several types of social capital and found that most increase volunteering. Brown and Ferris (2007) found that individuals’ associational networks and their trust in others and in their community were important determinants of giving and volunteering. Brooks (2005) argued that testing the impact of social capital on volunteering may really be “measuring the associations between various types of social capital.” Therefore, to establish a link between social capital and charity, he focused only on monetary contributions and found that different social capital types—measured in terms of civic group involvement, social and racial trust, and political engagement—have differing levels of impact on giving. Schervish and Havens (1997) found that people who volunteered to one or more philanthropic organizations and who participated in religious organizations gave more of their household income.

With a view to these discrepancies, we re-examine here the relationship among social capital, volunteering, and giving. We adopt the view that social capital is distinct from individual charitable behavior (Brooks 2005; Brown and Ferris 2007; Goss 1999; Putnam 2000) and posit that both social capital and volunteering behavior promote charitable giving, controlling for individual demographics, human and financial capital, religiosity and psychological inclination to give.

Moreover, we consider two separate but related aspects of giving—religious and secular giving. Previous research on charitable giving has demonstrated some differences in the determinants of religious and secular giving, including the effects of education, marital status, citizenship, and race (Brooks 2005; Brown and Ferris 2007). The findings, however, have not been consistent. In this paper, we use the same set of predictors for both secular and religious giving, and test whether the determinants have different impacts on secular and religious giving.

We begin by developing a theoretical model of charitable giving (to both secular and religious organizations), which incorporate social capital indices and volunteering activity. Then, we estimate the model using data from the 2000 Social Capital Community Benchmark Survey. We conclude with a discussion of our findings and their implications.

The Determinants of Charitable Giving

Past research on charitable behavior generally finds that financial resources, human resources, religiosity, psychological inclinations, and demographics affect one’s decision to donate and the amount to donate (Clotfelter 1997; Hodgkinson and Weitzman 1996; Mesch et al. 2006; Schervish 1997; Wolpert 1997). In this study, we add two dimensions to the model—individual social capital and volunteering—and posit that individual charitable giving is an outcome of the combined forces of social capital, volunteering, psychological propensity, human and financial capital, and demographic characteristics.

Our implicit model is that one needs both the inclination and the capacity to give. Inclination is determined by one’s connection to charitable organizations (measured by social capital and volunteering), and one’s psychological propensity to give (as measured by religiosity and happiness). Capacity is captured primarily by human and financial capital variables. We also include a set of demographic characteristics, which may affect either the inclination or the capacity to give, as control variables. These are not the focus of this paper, and so their role is not developed in detail. Each set of expected determinants is discussed in detail below.

Social Capital

Social capital is a multifaceted concept. Indeed the literature reveals a variety of definitions, with different scholars sometimes using “social capital” to mean different things. Nevertheless, the concept is generally regarded as including aspects of social networks and/or social trust. Here, we use social capital to refer to both an individual’s social networks of friends, families, and organizations, and his or her social trust of others and authority. More specifically, we consider five social capital indices: four indices of networks—bridging social networks, informal social networks, civic engagement, and organized group activism, and one index of social trust. These are discussed in turn.

Social Networks

Overall, social networks are expected to have a positive impact on charitable giving. As Putnam argued in Bowling Alone, “social networks provide the channels through which we recruit one another for good deeds, and social networks foster norms of reciprocity that encourage attention to others’ welfare” (Putnam 2000, p. 117). With our four indices of social network, we address whether the different kinds of social networks—bonding versus bridging, formal versus informal, membership versus participation—have similar impacts on both religious and secular charitable giving.

Social networks generally include bonding and bridging networks (Putnam 2000). The former refers to networks among homogenous groups of people, while the latter refers to networks among heterogeneous groups. Bridging social networks represent the extent and diversity of social relations in which people engage. Individuals who have friends from different social backgrounds are expected to be more open and respectful of others. They are thus more likely to support a variety of charitable causes or issues that might affect their diverse social networks. Therefore, we expect that individuals with more bridging social networks will donate more.

Informal social networks refer to interactions among family, friends, and others. These types of interaction foster a sense of reciprocity and caring among people, which are deemed to be virtues that lead to philanthropy (Martin 1994). Therefore, we expect that individuals with more informal social networks will donate more.

Civic engagement refers to formal group involvement. Normally, individuals become involved in formal groups because they believe in the mission or causes of the group. Membership in organizations can involve people in a wide variety of giving (Radley and Kennedy 1995). Through group involvement, individuals build a sense of connection. This “sense of being connected with another or categorizing another as a member of one’s own group,” is a main determinant of helping (Jackson et al. 1995, p. 74) because the combination of personal beliefs and associational ties brings the needs of others into one’s purview (Schervish 1997). Indeed, the 1996 Giving and Volunteering survey found that group involvement were highly correlated with the likelihood that a household makes a charitable contribution (Hodgkinson and Weitzman 1996). Therefore, we expect that individuals involved in more groups will donate more.

Organized group activism refers to the extent of civic participation. Being a member of an organization is different from actually attending local community events, club meetings, and public meetings. If being a member shows the breadth of an individual’s social network, actual participation in organized group activities indicates the depth of the network. Face-to-face interaction with others strengthens social networks and connections. Connections promote charitable giving, and thus we expect that individuals who participate more frequently in organized group activities will donate more.

Social Trust

Giving is a matter of trust. Studies show that individual charitable decisions and the amount of contribution are highly influenced by the individual’s level of trust in institutions and in other people, which is also called general social trust. In his study of charitable giving in The Netherlands, Bekkers (2003) found that general social trust increased the amount people gave to charitable causes. He argued that the general social trust determines to an important extent the level of trust in the philanthropic sector in a given society. Therefore, we expect that individuals with higher levels of social trust will donate more.

The two aspects of social capital are expected to affect charitable behavior differently. In general, individual charitable behavior is affected by both exposure to giving opportunities and the willingness to give. Different types of social networks increase one’s chances of being asked to donate, enhance one’s sense of connectedness to certain organizations, foster reciprocity, and therefore increase charitable giving. Social trust, on the other hand, affects the psychological aspect of giving. It is particularly important in explaining the charitable behavior of those who are not well connected to others or involved in organizations. Therefore, we explore the impact of these two social capital dimensions separately.

Volunteering

Several studies have shown that volunteering behavior is a reliable predictor of charitable giving. The Independent Sector’s series of studies on giving and volunteering in the United States showed a clear trend of more giving by volunteers than non-volunteers (1999). In the six national surveys on giving and volunteering, contributing households with a volunteer gave more than twice the percentage of household income than the households that contributed but did not volunteer. This relationship held even in periods of uncertain economic condition as in 1991 and 1993. Schervish (1997) examined whether the number of hours and the number of different organizations people volunteered in had any impact on their charitable giving. He found that the number of hours volunteered significantly increased the amount of giving. Brooks (2005) also found that volunteering for various types of organizations increased people’s charitable giving.

Volunteer work may promote charitable giving in two ways. First, volunteers tend to have pro-social personalities (e.g., empathy, caring) that make them more likely to make monetary contributions to public goods. Second, volunteering increases the awareness of public needs, helps people establish networks and relationships with charitable organizations, strengthens beliefs in the organizations’ missions, and enhances their understanding of the significance of the charitable work. Pro-social characteristics and increased exposure and interaction with charitable organizations should foster charitable giving. Therefore, we expect individuals who volunteer more to give more.

Happiness

An individual’s decision to donate may be affected by his or her psychological perspective. Studies show that an agreeable, extraverted, emotionally stable individual is more likely to give (Bekkers 2004). In this study, we use the individual’s self-rated happiness as an indicator of psychological inclination to donate. Compared to those who are unhappy about their lives, those who are generally happy are more emotionally capable to help others. In addition, happy people normally have an optimistic personality, which resonates with other characteristics that foster charitable giving behavior.

Religiosity

Religiosity has proven to be a significant contributor to charitable giving in American society (Hodgkinson and Weitzman 1996; Schervish 1997). Religion provides a cognitive framework that fosters caring and benevolent behavior. The impact of religiosity on religious donations is explicit. But, what impact if any should religiosity have on secular giving? Some studies have shown that people with religious beliefs tend to be generous and have greater concern for disadvantaged groups (Brown and Ferris 2007). Therefore, we hypothesize that people with more religiosity will contribute more to both secular and religious causes.

Human and Financial Capital

The human and financial capital of individuals reflects their capacity to give. Human capital, measured as educational attainment, is expected to increase charitable giving. Education moves people upwards in a social hierarchy, expanding a person’s information sources, fostering civic awareness and empathy to those who need help, and making people more willing to undertake actions for the public good (Brown and Ferris 2007). Previous studies have consistently found that education has a positive impact on giving, even after controlling for income (Andreoni et al. 2003; Brown and Ferris 2007; Reece and Zieschang 1985). Therefore, we expect individuals with higher education levels will give more.

We consider two measures of financial capital—income and homeownership. Annual household income determines the discretionary financial resources available for charitable donations. Of course the amount of discretionary resources is subjective as there is considerable diversity in one family’s necessity versus another family’s luxury (Schervish 1997). Nevertheless, we assume that families with higher annual incomes will find it easier to spend money for charitable purposes compared to those with less income. Therefore, we expect individuals with higher annual household incomes to donate more.1

Homeownership represents a stable and healthy financial status, and thus homeowners may be in a better financial position than renters, and thus have greater capacity for charitable giving. Moreover, homeowners may be more connected to their communities which may also increase their propensity to donate—at least to local causes. A national survey conducted by the Independent Sector in 2001 found that homeowners in the Western region of the United States gave 174 percent more than non-homeowners within the region (Independent Sector 2004). Therefore, we expect that homeowners are more likely to donate.

Demographics

Finally, we include demographic characteristics that have been found to influence charitable behavior as control variables. Age, gender, race/ethnicity, citizenship, marital status, number of children, and the length of residence in the community (Brown and Ferris 2007; Schervish 1997) have been found to affect the decision to make a donation and the amount donated.

Age is expected to positively affect both the capacity and inclination to give. Older people may have more assets, as well as more experience and exposure to philanthropy. Most large donations come as bequests or after donors have retired. Most studies have found that gender affects charitable behavior. Although some studies in experimental economics found no significant gender difference in benevolent behavior (Bolton and Katok 1995), empirical studies using survey data have found that women are more likely to donate than men (Andreoni et al. 2003). Moreover, there are significant gender differences in attitudes and beliefs about caring and empathy (Andreoni and Vesterlund 2001; Hoffman 1977). Studies on race/ethnicity differences in charitable giving have consistently shown that being white is positively associated with the likelihood to donate or to donate higher amounts compared to other minority groups. The 1998 Giving and Volunteering Survey conducted by the Independent Sector revealed that African Americans and Hispanics have lower rates of household giving and gave less amounts compared to whites. Brown and Ferris (2007) also found that compared to whites, Hispanics donate significantly less to religious purposes, and both African Americans and Hispanics donate less to secular purposes. Citizenship enhances the sense of belonging and connectedness in a country and therefore is expected to increase the likelihood and amount of charitable giving. Literature on immigrant charitable giving shows that recent immigrants—those who migrated in the past 10 years to the United States—have a significantly lower likelihood of giving compared to those who have been in the United States for more than 30 years and who most likely already have gained citizenship (Osili and Du 2005). Marital status has been found to have a positive impact; married persons are more likely to give and to give more than single persons (Mesch et al. 2006). The impact of the number of children on charitable giving is likely to be negative. More children in a family increase financial constraints, reducing the capacity for giving. Nevertheless, children increase social network activities, so the net impact on giving may be small. Finally, length of residence in a community reflects social connections and the sense of belonging in the community. So, those who have lived longer in the community may contribute more. In summary, our model of the determinants of charitable giving is presented in Fig. 1.
Fig. 1

Model of the determinants of charitable giving

Data and Methodology

To test the above hypotheses, we use data derived from the national sample of the 2000 Social Capital Community Benchmark Survey (SCCB). The survey, averaging 26 min, was the largest scientific investigation of social capital and civic engagement ever conducted in the United States. It was conducted nationally by telephone using random-digit-dialing during July to November 2000. Roughly 260 variables were generated from the questionnaire.2 The sample included a total of 3,003 interviewees. After missing values are excluded, 1,946 respondents are included in the analysis.

Dependent Variables

The dependent variables are religious giving and secular giving. Respondents were asked the total dollar amount the household contributed to church or religious causes and non-religious causes in the past 12 months. The charitable giving data were originally coded categorically. We estimated the amount of charitable giving based on the category’s midpoint. On average, the respondents gave $1,141 to religious causes and $566 to secular causes in 1999.

Independent Variables

Social Capital

The five social capital variables discussed earlier are measured as indices (details in Appendix). The social trust index is the mean of the standardized responses (based on national norms) to six questions: whether the respondents thought most people could be trusted, and how much they trust neighbors, co-workers, fellow congregants, store employees where they shop, and local police. The bridging social network index is a count of the different kinds of personal friends the respondent has from the 11 types provided. On average, the respondents have six different kinds of personal friends. The informal social network index is the mean of the standardized responses to five questions: the number of times the respondent played cards with others, visited relatives, invited friends home, socialized with co-workers outside of work, and spent time with friends in public places during the past 12 months. Civic engagement is a count of the different types of formal groups in which the respondent is involved from the 18 types provided. On average, the respondents were involved in three different kinds of formal groups. Organized group activism is the mean of the standardized responses to three questions: the number of times the respondent attended community events, club meetings, and public meetings during the past 12 months.

Volunteering

Volunteering activity, denoted times volunteered, is measured as how frequently people volunteered during the past 12 months. On average, the respondents volunteered six times.

Happiness

The variable “happiness” is the respondent’s answer to the question: “All things considered, would you say you are very happy, happy, not very happy, or not happy at all?” “Not happy at all” is coded as zero, and “very happy” is coded as three. It is considered an ordinal-scale variable in the analysis. Most respondents (55%) are happy about their lives, one-third of the respondents are very happy, and only a few are not happy at all.

Religiosity

Religiosity is measured as whether the respondents agree with the statement that religion is important in their lives.3 About 83% of respondents either strongly agree or agree with the statement.

Human and Financial Capital

Educational attainment, denoted education, is measured in three levels: high school or less, some college, and 4-year college and above. The educational attainment is evenly distributed among the samples (37% with high school and less education, 31% with some college, and 31% with 4-year college and above degrees). We used the high school and less category as the base category. Two education dummy variables—some college and 4-year college and above—are included in the model. To make the measurement of education level more meaningful, the study only focuses on respondents of age 25 and older. Annual household income in 1999, denoted income, is also measured in three levels: $30,000 and below, $30,000 to $75,000, and $75,000 and above. About 32% of the respondents are low income, 46% middle income, and 21% high income. The low-income category is used as the base category, and two dummy variables—middle income ($30,000 to $75,000) and high income ($75,000 and above)—are included in the analysis. Homeownership is a dummy variable denoting whether the respondent owned or rented their home. Seventy-two percent owned their home.

Demographics

More than half of the respondents are female (60%), married (54%), and citizens (72%). On average, the respondents included in the analysis are 48 years old, and have one child 17 years old or younger in the household. The length of residence, denoted residence, is a variable with five categories—less than one year, one to five years, six to ten years, 11–20 years, and more than 20 years. Most of the respondents have lived in the community for either more than 20 years (36%) or from one to five years (24%).

Table 1 provides the descriptive variables for all the variables. Given the controversy noted earlier about the conceptual distinctiveness of social capital and volunteering, we examined the correlation of these variables in our data set and found a significant, but weak association. Thus we are able to empirically differentiate these concepts in our data. We also examined correlations among all the independent variables, and no significant multicollinearity was found.
Table 1

Descriptive statistics (sample N=1,946)

Variable

Mean (s.d.)/Frequency

Max.

Min.

Dependent variables

Religious Giving

$1,141a (1,853)

0

7,500

Secular Giving

$ 566b (1,312)

0

7,500

Independent variables

Social capital

    Social trust

0 (0.7)

−2.5

1.0

    Bridging network

6.1 (2.7)

0

11

    Informal network

−0.1 (0.65)

−0.9

2.2

    Civic engagement

3.2 (2.8)

0

18

    Activism

0 (0.7)

−0.5

5.7

    Times volunteered

6 (2.5)

1

9

    Happiness—Current state of happiness

Not at all 1%; Not very 5%; Happy 55%; Very Happy 38%*

    Religiosity—Importance of religion in life

Strongly disagree: 8%; Somewhat disagree: 8%; Neutral: 1%; Somewhat agree: 20%; Strongly agree: 63%

Human and financial capital

    Education

High school and below: 37%; Some college 31%; College and above: 31%*

    Income

Low: 32%; Middle 46%; High 21%*

    Homeownership

72%

0

1

Demographics

    Age

47.5 (15.3)

25

92

    Gender

Female: 60%; Male 40%

    Citizenship

Citizen: 94%; Non-citizen: 6%

    Marital status

Married: 57%; Other: 43%

    Kids

0.9 (1.3)

0

13

    Race/Ethnicity

Asian: 1.5%; Black: 16%; Hispanic: 14%; Other 3%; White: 65.5%.

    Residence

Less than 1 year: 6%; 1–5 years: 24%; 6–10 years: 16%; 11–20 years: 18%; More than 20 years: 36%

aThe median of religious giving is $300

bThe median of secular giving is $50

* Due to rounding error, the categories do not add up to 100%

Estimation Methodology

We estimate our model as a Tobit regression. Donated dollars are left-censored at zero, so ordinary least squares (OLS) regression, which assumes a symmetrical distribution of the dependent variable, may yield biased parameter estimates. The Tobit regression method corrects for the censored dependent variable (Tobin 1958). The analysis is conducted using SAS 9.0 software. To ensure that the results of the analysis are generalizable to the overall U.S. population, the analysis is weighted to generate a nationally representative sample based on age, gender, education, and race/ethnicity.

Estimation Results

Table 2 presents the results of the Tobit regression on religious giving and secular giving. The results provide strong support for our model, and reveal that social capital, volunteering, happiness, religiosity, human and financial capital, and demographics all play important roles in determining charitable giving.
Table 2

Tobit regression on religious and secular giving

Variable

Religious giving parameter estimates

Secular giving parameter estimates

Social capital

Social trust

204.85**

102.71*

Bridging network

35.45*

33.47**

Civic engagement

111.28****

112.24****

Informal network

−40.30

−4.29

Activism

−74.47

183.03***

Times volunteered

19.90****

7.01***

Happiness

186.02**

46.91

Religiosity

858.88****

−17.57

Human and financial capital

Some college

433.08***

160.73*

College

667.99****

466.13****

Middle income

342.25***

304.43***

High income

1,209.46****

1,049.76****

Homeownership

272.83**

150.17

Demographics

Age

7.40*

9.76***

Female

−359.18***

−126.76*

Citizenship

−225.57

−119.40

Married

304.80***

103.03

Kids

94.95**

−31.69

Asian

264.90

−446.20*

Black

253.39

−204.99*

Hispanic

−347.14*

−427.66***

Other

88.99

−239.20

Residence

94.55**

−56.82*

Log likelihood

−12,706.75

−11,683.45

P<0.10; ** P<0.05; *** P<0.01; **** P<0.0001

Social capital has a positive impact on both religious and secular giving. It, however, affects the two types of charitable giving in slightly different ways. As expected, social trust, bridging social networks, and civic engagement (or formal group involvement) increase the amount of charitable giving to both religious and secular causes. Associational membership and diversity of social networks increase personal and organizational connections, and presumably this exposure increases both one’s knowledge of the causes and the likelihood that one will be asked to donate. Social trust presumably facilitates the willingness to contribute to collective goods—whether provided by secular or religious organizations.

In addition, a fourth social capital measure, organizational activism, is found to increase the amount of secular giving. Attending community events, club meetings, and public meetings presumably raises awareness of secular issues and therefore increases the propensity to give. Secular group involvement, however, was not found to increase religious giving in this sample. The fifth measure of social capital, informal networks, had no significant impact on either type of charitable giving.

The results also support a relationship between volunteering and giving. Individuals who volunteer more often donate more to both religious and secular organizations. This finding is consistent with previous general findings about giving (Hodgkinson and Weitzman 1996; Schervish 1997). Earlier studies, however, did not distinguish between secular and religious giving, so this work advances our understanding of this relationship.

A feeling of happiness was found to affect religious giving, but not secular giving. This is an interesting finding. We had expected a positive psychological perspective to increase the inclination to donate generally. Individuals may be more inclined to link personal happiness with religious causes, rather than secular ones. It is also possible that religious giving gives people a sense of happiness, which makes the relationship between charitable giving and happiness reciprocal. Because the data used for this research is cross-sectional, we cannot control for this potential reciprocity.

Not surprisingly, those who think religion is important in their lives donate more to religious causes. In addition, we find that religiosity does not have a significant impact on secular giving. The latter finding is interesting because the empirical literature has been unclear on this relationship. Hodgkinson and Weitzman (1996) found that religious involvement and regular attendance of religious services increased the level of giving to religion and to other charities. In contrast, Brooks (2005) found that respondents who attended worship services weekly or more often give more to religious purposes, but less to secular purpose. Our findings are consistent with Brooks and do not show an impact of religiosity on secular giving. Those who think religion is not important to their lives are just as likely to donate to secular causes as those who think religion is important.

The capacity to give, as expected, was found to be important. Human and financial capital indicators significantly increased both religious and secular giving. Those with some college and those with at least a college level of education give more to both religious and secular organizations than those with high-school or lower education. Similarly, middle and high income individuals give more than low income respondents to both religious and secular causes. Homeownership has a differential impact. It positively impacts religious giving, but does not affect secular giving. Thus, homeowners are not significantly different from renters in secular charitable giving, after controlling for other determinants.

In terms of demographics, age, gender, race/ethnicity, and the years lived in a community significantly affect religious and secular giving. Older people contribute more than younger people, and women contribute significantly less than men to both religious and secular causes. Minorities donate less to charitable organizations than whites, but there is variability by ethnic group. Hispanics contribute less than whites to both religious and secular organizations; Asians and Blacks differ from whites only in their secular giving.4 The results also show that those who have lived in a community longer tend to donate more to religious causes, but less to secular ones, than those who have lived in the community for a shorter time. This may suggest that those who have lived in a community for more than 10 years develop stronger ties to local religious organizations than secular ones. Finally, those who are married and those who have more children contribute more to religious causes, but they are not significantly different in secular giving from those who are unmarried persons and those without children.

Conclusion

In an era of growing competition for charitable dollars, understanding individual giving behavior is critical to capacity building and the growth/fundraising effort of charitable organizations, especially small organizations. By connecting religious and secular giving to an individual’s level of social trust, social networks, volunteering behavior, happiness, religiosity, human and financial capital, and demographic characteristics, this study draws a new diagram of the personal and social forces that affect charitable giving.

The findings provide strong support for our main hypothesis that social capital fosters charitable giving. Individuals who trust others more give more to religious and secular causes. Individuals with a greater diversity of social networks and broader civic engagement give more to both religious and secular organizations. Finally, individuals actively involved in more formal groups donate more to secular causes. These results underscore the importance of trust, connections to other individuals, and connections to organizations in determining charitable behavior. Interestingly, the results indicate that the role of social capital is quite similar for both religious and secular giving. Only active participation in civic affairs distinguishes the relationship. The results thus provide evidence of a strong and consistent relationship between the different aspects of social capital and charitable giving. This study, however, does not establish a causal relationship. Indeed, there may be some cross-causality involved. Those who contribute more may then strengthen their connections to other types of people and organizations. Future research that can access time-series or panel data will be needed to confirm the direction of the relationship. Nevertheless, there are good theoretical reasons to believe the causality is as modeled here.

Another interesting and important finding is the observed differences in the determinants of religious and secular giving. Happiness, religiosity, homeownership, marital status, and number of children affect religious giving, but not for secular giving. Moreover, length of residency increases religious giving but decreases secular giving. These considerable differences in the determinants of these two types of charitable giving reveal that different forces are at play in shaping charitable giving toward religious and secular causes. This suggests that future surveys and empirical studies of charitable giving should analyze the two types of giving behavior separately whenever possible. It also suggests that faith-based nonprofit organizations and secular nonprofit organizations may want to develop different fundraising strategies.

In addition, as expected, people who volunteer more also donate more. This suggests that to attract more donations, charitable organizations—both religious and secular nonprofit organizations—may want to provide volunteering and participation opportunities that build a sense of connection to the organization. Promoting bonding and bridging networks with volunteers should enhance their trust and appreciation of the organizations’ mission, which should yield more monetary contributions.

Finally, these results are based on a national sample of United States residents. Given the strong cultural aspect of philanthropic behavior, it is unlikely that these findings can be generalized to other countries. They may however be suggestive for efforts to understand the role of social connections on individual charitable behavior more generally.

Footnotes
1

Since we are concerned here with the absolute capacity to give, we consider the relationship between charitable giving and the level of income. There are data that suggest that those with lower levels of income may give a higher percentage of their income than those with higher levels, so we are not suggesting that the poor are less generous—only that they have less capacity.

 
2

Additional details on the survey are available at http://www.cfsv.org/communitysurvey.

 
3

There are numerous ways to measure religiosity (Hill and Hood 1999). A commonly used indicator is how often people attend religious services, and attendance is likely to be associated with more religious giving as some religious organizations solicit donations during the religious services. Attendance is not, however, likely to be as useful a measure for the connection between religious belief and secular giving. We want to explore whether a general religious belief influences charitable behavior overall (the belief–behavior relationship). Thus, we use the importance religious belief has in one’s life as the measurement of religiosity.

 
4

These results, however, should not be interpreted as minorities are less generous. Studies show that minorities often give informally to their extended family (in the United States and in their country of origin), neighbors, friends, and fellow migrants from the same region (Smith et al. 1999). Moreover, they usually describe their philanthropic activities as “sharing” and helping” rather than “giving” or “volunteering.” Therefore, national surveys tend to underestimate the giving of minority groups.

 

Copyright information

© International Society for Third Sector Research and The Johns Hopkins University 2008

Authors and Affiliations

  1. 1.Taubman Center for Public Policy and American InstitutionsBrown UniversityProvidenceUSA
  2. 2.School of Policy, Planning, and DevelopmentUniversity of Southern CaliforniaLos AngelesUSA

Personalised recommendations