Population Research and Policy Review 23: 161–185, 2004.
© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
The role of religion in union formation: An economic
EVELYN L. LEHRER
Economics Department, University of Illinois at Chicago
Abstract. Previous research has shown that the faith in which a young woman is brought up
has important effects on the subjective costs and/or beneﬁts of many decisions that she makes
over the life cycle, including schooling, employment, and fertility. Based on this evidence,
the present paper develops hypotheses regarding patterns of entry into marriage and cohab-
itation for the main religious groups in the United States: mainline Protestants, conservative
Protestants, Catholics, Mormons, Jews, and the unafﬁliated. The empirical results, based on
young women from the 1995 National Survey of Family Growth, are generally supportive of
Keywords: Cohabitation, Marriage, Religion, Secularization thesis
This study examines the role of religion in two dimensions of women’s trans-
ition to ﬁrst union: the timing of such transition and whether it takes the form
of marriage or cohabitation. Previous research has shown that religious afﬁli-
ation has important effects on economic and demographic behavior: it has an
inﬂuence on educational attainment (Lehrer 1999a, 2003; Darnell & Sherkat
1997; Chiswick 1988), attitudes toward pre-marital sex (Sweet & Bumpass
1990), fertility (Lehrer 1996a, 1996b; Thornton 1979), female employment
(Sherkat 2000; Lehrer 1995) and the prevalence of divorce (Teachman 2002;
Lehrer & Chiswick 1993). Based on this evidence, the present paper develops
hypotheses regarding patterns of entry into marriage and cohabitation for the
main religious groups in the United States: mainline Protestants, conservative
Protestants, Catholics, Mormons, Jews, and the unafﬁliated.
Much of what we know about the effects of religion on marriage and
cohabitation is based on unions formed prior to the mid 1980s (Lehrer 2000;
Sander 1993; Thornton et al. 1992). The present study uses more recent data
on young women from the 1995 National Survey of Family Growth. This
An earlier draft of this paper was presented at the annual meetings of the Population As-
sociation of America, May 2002, Atlanta. I am indebted to Barry Chiswick, Carmel Chiswick,
Linda Waite, and anonymous referees, for many helpful comments and suggestions.