1
Risky Business: Assessing Risk Preference Explanations for Gender Differences in
Religiosity
Scholars of religion have long noted that women are almost universally more
religious than men (see Walter and Davie 1998, and Stark 2002 for reviews). Many
explanations for this gender gap are centered on differential socialization, although Miller
and Stark more recently explained this gender gap as a consequence of men’s greater
biological propensity to take risks (Miller and Stark 2002). Defining irreligion as a risk of
divine punishment, they supported this claim using cross-national and inter-religion data
to reveal a larger gender gap in religions with a strong risk of posthumous punishment.
But we argue that Miller and Stark’s conceptualization and operationalization of risk is
inadequate to support risk-preference theory. We use a more precise measure of other-
worldly religious risk that is sensitive to belief in life after death, to test this theory. Our
findings fail to support and, in fact, contradict risk preference theory. We then use our
results to speculate about alternative explanations that are grounded in sociological
understandings of gender, arguing that socialization theories are not the only alternatives.
THEORIES OF THE GENDER GAP
Miller and Stark (2002) claim that most sociological explanations for the gender
gap in religiosity rely on differential socialization. They describe three types of
socialization arguments that dominate the existing literature. The first is that girls learn
traditionally “feminine” personality characteristics such as nurturance and
submissiveness during primary socialization. Because these characteristics are associated
with religiousness, they lead women to be more religious than men. The second is a
structural location argument that claims that women expect to occupy the role of
homemaker or at least to subordinate paid employment and civic engagement to their
Page 1 of 27 
Convention