Are women more generous than men? Evidence from the US Consumer Expenditure Survey

Abstract

This paper examines how gender, age, education, income, race, and ethnicity affect giving behaviors using the 2006 US Consumer Expenditure Survey. The testable hypotheses are based on theories of human capital and social capital. The research suggests that gender differences in philanthropic behavior are non-existent. Education, annual income, wealth, and being Hispanic increase the probability of giving, but they had no effect on the amount gifted. It is estimated that age and race interact with gender to affect differences in giving – older women are more likely than younger men to donate but give smaller shares of their income, while white women, black women, and Asian women are less likely to donate and give smaller amounts than men of ‘other’ races.

Keywords::

• philanthropic behavior
• giving
• gender
• race
• ethnicity

Acknowledgements

Any errors are the sole responsibility of the authors. Earlier versions of this paper were presented at the 2010 Annual Meetings of the Chinese Economic Association in North America, the 2010 Annual Meetings of the Eastern Economic Association, and at the 2010 Consumer Expenditure Survey Microdata Users' Conference at the Bureau of Labor Statistics. We thank Geoffrey Paulin, the anonymous referees, and all participants at conferences for their comments on earlier versions. We also thank Mark Kolakowski for his editorial assistance.

Notes

^1. Annual income before tax, defined as FINCBTXM, is the most reliable single data item on a respondent's income in the CE data without addressing the multiple imputed data issue, which requires a special method to estimate standard errors, when it is used as a categorical variable.

^2. Respondents who are in a single-member consumer unit include those who are unmarried and those who are the only member of the consumer unit (i.e. FAM_SIZE = 1).

^3. Total donations exclude contributions to: (1) college students living away from home, and (2) educational institutions, child support, and alimony. Since the survey contains contributions per quarter, total donations per year are derived by summing the data for all four quarters for each respondent identification number (newid).

^4. The sample selection model is also known as the type 2 Tobit model (Amemiya 1985), the Probit selection model (Wooldridge 2002), or the bivariate sample selection model (Cameron and Trivedi 2005).

Additional information

Notes on contributors

Chu-Ping Lo

Sanae Tashiro, PhD, is an Assistant Professor of Economics at Rhode Island College and a visiting research scholar at the Claremont Institute for Economic Policy Studies. Professor Tashiro's research interests include wage and income inequality, fertility, and the economics of race and gender. She has published in the American Journal of Economics and Sociology, the Review of Black Political Economy, China Agricultural Economic Review, and the International Journal of Economic Issues.

Sanae Tashiro

Chu-Ping Lo, PhD, is an Assistant Professor of Economics at National Taiwan University. Professor Lo's research interests include international trade, international outsourcing, and growth models. He has published in China Economic Review, China Agricultural Economic Review, the International Journal of Management & Enterprise Development, the Journal of Agricultural Science and Technology, and the Journal of International and Global Economic Studies.