Religious giving by individuals: a cross denominational study

American Journal of Economics and Sociology, The,  Jan, 1997  by Kevin F. Forbes,  Ernest M. Zampelli

• E-mail
• Print
• Link

I

Introduction

Religious organizations received a large share of the over $60 billion contributed by Americans to charitable institutions in 1992, according to Independent Sector's 1992 Survey of Giving and Volunteering. Although religious giving is substantial, the isolating of the factors significant in motivating it remains high on the research agenda. Much of the research to date has analyzed the determinants of church collections at the congregation level. Although this research has provided important insights into religious giving patterns, it ignores the significant reality that the congregational collection is only one of the many channels through which individuals contribute to religiously affiliated organizations. One purpose of this paper is to focus on total religious giving at the individual contributor level and to determine if the results are consistent with those from the congregation specific literature. Different results may be indicative of differences in the motivations underlying the other types of religious giving.

A second purpose of this paper is to enhance the understanding of differences in the religious giving patterns of Catholics and Protestant denominations. Past descriptive analyses of religious contributions have noted consistently that Catholics contribute less than Protestants both in amount and as a share of household income. This observation has been corroborated by a number of the econometric/regression based studies whose results indicate that marginal propensity to give out of income is lower for Catholics than it is for Protestants. This paper hopes to shed light on whether this result is maintained when total religious giving by individuals of different denominations is examined.

We note at this point that one study has analyzed religious giving by individuals. Forbes and Zampelli (1993) estimate an econometric model of religious giving by individuals using standard least squares regression techniques and find that income, attendance at church services, as well as some demographic and attitudinal characteristics are statistically significant factors in explaining religious giving patterns across individuals. In addition, the results are consistent with previous evidence that the marginal propensity to give out of income is lower for Catholics than for Protestants.

These results should be interpreted with caution. The model was estimated for givers only, i.e., the dependent variable, religious giving, was restricted to be positive. A sizable portion of the entire sample, however, was comprised of non-givers. As is well known, conventional regression methods applied in circumstances where the dependent variable is zero for a significant fraction of the observations do not account for the qualitative difference between the limit (zero) observations and the non-limit (continuous) observations. A third purpose of this paper is to correct this deficiency by employing the censored regression model or what is commonly referred to as the Tobit model.(1)

The remainder of the paper is organized as follows. Section II presents an econometric model of religious giving. The data is discussed in section III. Estimation and results are provided in section IV. Section V offers a brief summary and some tentative conclusions.

II

An Econometric Model Of Religious Giving

In an excellent review of the theoretical and empirical literature on factors related to religious giving, Hoge and Griffin (1992) document the importance of specific economic, demographic, and attitudinal characteristics in determining patterns of religious giving. With this as a foundation, we formulate a single equation model of religious giving in which monthly religious contributions are a function of the givers' socioeconomic and demographic characteristics, religious behavior, and attitudes. Algebraically, the equation is written as:

GIVE = [a.sub.0] + [a.sub.1]TAX_DED + [a.sub.2]LNINCOME + [a.sub.3]LNINCOME * TITH

+ [a.sub.4]ATTEND + [a.sub.5]MARRY + [a.sub.6]HH_SIZE + [a.sub.7]HH_AGE

+ [a.sub.8]COL_GRAD + [a.sub.9] COMIT + [a.sub.10]CONFID + [a.sub.11]SECULAR

+ [a.sub.12]GV_BACK + [a.sub.13] CANT_DO [1]

where:

GIVE = the household's monthly religious contributions;

TAX_DED = 1 if the household intends to use the tax deduction for charitable contributions, 0 otherwise;

TITH = 1 if the household tries to give a certain percentage of income to charity, 0 otherwise;

LNINCOME = natural logarithm of gross monthly household income, measured as the midpoint of the annual gross household income range specified by the survey respondent divided by 12;

ATTEND = 1 if the respondent attends church services once or more per month, 0 otherwise;

MARRY = 1 if the respondent is married, 0 otherwise;

HH_SIZE = the number of household members;

HH_AGE = the age of the respondent if the respondent is single or the average age of the respondent and spouse if the respondent is married;

COL_GRAD = 1 if the respondent is a college graduate, 0 otherwise.

COMIT = 1 if the respondent answered that meeting religious beliefs or commitments is a major factor in motivating charitable contributions, 0 otherwise;