-
406
of education marry later or not at all and are more prone to
divorce. A strong sex-specific interaction effect is ex¬
pected for these varables. The theory predicts that men with
high income, occupational prestige, and human capital gain
much from marriage. These men marry early after finishing
education and are less prone to divorce.
Of course, education increases the age of marriage for both
sexes because the "risk" of marriage is low for both men and
women as long as these are in the educational system
(Galler, 1979). However, after finishing college or univer-
sity the degree of education has a positive impact on the
tendency to marry for men and a negative impact for women.
Hence, the shift of age at marriage by education must be
stronger for women than for men (Keeley 1977: 245).
Costs of divorce is another component influencing both age
at marriage and marriage stability. The longer the expected
duration of marriage the longer is the period of search for
a mate if information is imperfect. Because it is assumed
that catholics have higher costs of divorce, a higher age at
marriage and a lower risk of divorce is expected for the
catholic population other things being equal. On the other
hand if people's search-period is short the likelihood of a
"mismatch" increases. This might be a reason for the mar-
riage-destabilizing-effect of early marriages. As can be seen
from this short discussion Becker's theory is much more a
framework than a closed deductive theory. The core of the
theory is the hypothesis of utility maximization under
restrictions. However, additional auxiliary hypotheses are
necessary to bridge the gap between theoretical terms like
"gain through marriage" and the observed socio-economic
characteristics.