causal links between early marriage and poor outcomes among women. Is it early marriage in and of itself that is the problem or is it the characteristics of those who marry early?

In contrast to the extensive documentation of female age at marriage, the literature on men is quite sparse (Malhotra, 1997). In part this limited attention to men is because demographic surveys, up until the last decade or so, have been restricted to women. But it is also due to the fact that across a wide spectrum of countries and cultures, relatively few men marry during the teenage years, and it is early marriage that is considered problematic and thus worthy of consideration.

In this chapter we will examine trends in the timing of first marriage or union for men and women. We define marriage broadly to include all socially recognized unions, including legal marriage as well as any other type of union that is recognized and reported in particular countries. The principal focus is on documenting trends in the age at marriage for the major regions of the developing world; however, the chapter also addresses a few subthemes: the current extent of early marriage, differences between men and women in trends in age at marriage, and the association between age at marriage and sociodemographic characteristics, specifically education and rural-urban residence. To the extent that changing patterns of behavior are revealed, we will try to identify to what such transformations might be attributed and draw on the demographic literature to provide insights.

A number of scholars have conducted research on marriage timing. We begin with a brief review of the contributions of various social science disciplines to an understanding of age at marriage.

Historical demographers have done an admirable job of documenting marriage patterns throughout Europe over the last few hundred years; however, they have fared less well in identifying a particular set of factors that explains trends across cultures. Hajnal (1965) first observed what he called a “European” pattern with late age at marriage and high proportions unmarried. In describing this distinctive pattern that existed from at least as early as the eighteenth century, he hypothesized that an association existed between marriage and household formation, arguing that when marriage involved the establishment of a new household, as it did in much of Western and Northern Europe, resource and skill acquisition were determining factors in the decision to wed. Wrigley and colleagues (1997, p. 122), in