405 
sent explanations for the typical non-monotonic hazard func- 
tions of the "risk" of marriage and divorce, to evaluate the 
relative strength of effects of different socio-economic 
variables as well as cohort effects, and to estimate the 
effects of expansion in education on the marriage and divor- 
ce patterns. 
2. Theory 
While there is a long tradition of research on marriage and 
the family by psychologists and sociologists (Goode, 1963; 
Carter and Glick, 1970; Cherlin 1981) economists only re¬ 
cently became interested in this research area. Becker's 
innovative idea was to apply the apparatus of modern micro- 
economic theory, i.e. marginal analysis, to the process of 
family formation and dissolution (Becker, 
1981; Becker, 
1975; Becker, Michael, Landes, 1977; Michael, 1979) Starting 
with a sparse set of principles the theory allows derivation 
of many different hypotheses and integrates a lot of well 
known sociological findings in a common framework. 
The basic assumption of the theory is that a household pro- 
duction function with time for household work, time for 
market work, market goods, prices of market goods, and wage 
rates as inputs is maximized under budget and time restric- 
tions. People marry if the value of this function is higher 
for a common household than the utility of remaining single, 
i.e. if there is a gain through marriage. This gain through 
marriage is higher, the higher the similarity or complemen- 
tarity between traits that are used jointly in production 
and the more dissimilar are skills of the household members 
which can be substituted for each other. In accordance with 
the theory women with high income, occupational prestige, 
and human capital profit less from marriages because they 
lose by the traditional mode of division of labor in the 
household. Therefore, high income women with a high degree