Making charitable donations to nonprofit organizations is a wide-spread and frequently recurring form of prosocial behavior in many Western countries. This study describes and explains cross-national differences in charitable giving, using data from 22 countries included in wave I of the European Social Survey.
Why people donate money to charitable causes is an intriguing question that has been studied in psychology, sociology, and economics since the late 1960s. In a comprehensive review of the empirical literature on philanthropy, Bekkers & Wiepking (2007) bring together dozens of experimental studies showing that individual philanthropy is strongly sensitive to social influence. From a large number of studies relying on survey data we know that individual religion and education are positively related to philanthropy (Wuthnow, 1991; Brown & Ferris, 2007; Bekkers & Schuyt, 2008). The types of arguments that are given to interpret these findings usually refer to the social context in which religious and more highly educated individuals reside. These arguments are in line with findings from the experimental literature. However, the arguments on context effects have rarely been tested empirically using survey data on individual giving nested in national contexts. In this paper we explore the role of national context in the explanation of differences in philanthropy across Europe.
From both the experimental as well as the survey literature we expect that individuals in more devout (especially more Protestant) countries and individuals in more highly educated societies are more likely to donate because they are more likely to experience stronger social norms encouraging donations. We explore whether individual religiosity and education matter more or less in more devout and more highly educated societies.
First, we focus on the differences in giving between the 22 countries. The descriptive results displayed in figure 1 show large differences in giving between countries, with Hungary being the least and the Netherlands being the most generous. Generally, the highest percentage of donating households can be found in the Nordic countries (except for Finland), whereas the Southern European and former Eastern European countries have lower percentages of donating households.
In order to determine whether individuals in more devout and more highly educated societies are more likely to donate to charitable organizations, we performed multi-level analyses on the first wave of the ESS (Jowell, 2003), complemented with data covering the United States (CID, Howard, Gibson, & Stolle, 2005). Table 1 displays the results for the individual and country level effects of religious denomination on giving.
The results in table 1 show that the effects of individual religious denomination are largely in line with previous country specific findings of the effect of religious denomination on incidence of giving (e.g., Hoge & Yang, 1994; Bekkers & Schuyt, 2008). People belonging to Protestant and smaller Christian religious denominations have the highest probability of making donations, followed by those with a Roman Protestant religious denomination and people with a non Christian religious denomination. People who indicate not belonging to a religious denomination have the lowest probability of making donations.
When we consider the effects of religious context, we find that people living in countries with a higher percentage of Protestants have a higher probability of making charitable donations. People living in countries with a higher percentage of Roman Catholics have a lower probability of making donations. Thus, not only individual religious denomination affects charitable giving, but also the religious context.
That context does matter becomes even more clear when we consider the interactions between individual and country level religious denomination. The results support the hypothesis that members of religious groups give more when the group is surrounded by members of other religious groups. People with a Roman Catholic denomination have the highest probability of making charitable donations in more Protestant countries. Conversely, in Roman Catholic countries, people with the highest probability of making donations are those who belong to a Protestant denomination. Both people with a Roman Catholic denomination living in a ‘Roman Catholic country’ and people with a Protestant denomination living in a ‘Protestant country’ have lower probabilities of making donations. Conclusively, religious context matters for the effect of individual religiosity on giving.
Table 2 displays the results for the individual and country level effects of educational level on giving. People with a higher level of education are more likely to donate. Again, this is in line with results of giving research conducted in separate countries (Brown & Ferris, 2007; Wiepking & Maas, forthcoming). At the country level, we find a positive effect of educational expansion. People living in countries with a higher proportion of the population holding a tertiary education diploma are more likely to donate. However, once we include interactions between individual and country level education, we find no significant effect for both educational expansion and the multi-level interactions. We conclude that the positive effect of educational level on charitable donations is not affected by the mean level of education of a country. There is no contextual influence of educational expansion on the effect of individual educational level on giving. This result is not in line with the study by Gesthuizen et al. (2008), who use Eurobarometer data and find that individual education is less predictive of charitable giving in societies with stronger educational expansion.
Our results show that, in the case of donations made by people belonging to a religious denomination, individual philanthropy is strongly sensitive to social influence. How this social influence exactly affects giving is one of the next intriguing questions that needs to be examined in future research.