ing about fertility: only if females have a strong influence on household decisions about fertility do genetic influences on fertility motivations reveal themselves in fertility outcomes. While it is probably undisputed that females in contemporary cohorts exercise a strong influence on household decisions the strong influence of females in cohorts born around 1885 (i.e. in cohorts facilitating the early fertility decline) may be more surprising.
Implicit in such an argument would be the notion that women were genetically more effective at bargaining over remaining childless than over family size. There is no justification for why women born in the 1880s stood out in terms of genetically mediated bargaining power over fertility control, compared with those born in immediately subsequent decades. Nor is it readily apparent why women’s liberation on fertility decision making should not have shared-environment components and sources.
An important feature of this extended citation is that it acknowledges that childbearing (and marriage) involves both partners. This carries several implications for future research on fertility, marriage, and divorce. First, twin studies will never permit us to address and disentangle genetic components of variation for the dyad involved. The only way to get leverage on this aspect will be through molecular genetics (or linking back through neuroendocrine protein pathways to genetic markers), and that will require much sharper elaboration of likely pathways and identification of markers for these pathways. Second, shared environment predominantly relates to shared family of origin effects and is a concept largely introduced as a means for separating out narrow-sense heritability (that nevertheless incorporates all gene-environment interplays). In particular, it always seems odd to pay exclusive attention to shared family-of-origin environment components for adult behavior, although nevertheless a source of important and interesting pathways to adult development. A full analysis in the context of fertility would surely involve separating out shared-environment components from the family of origin of both partners and for the couple themselves, in addition to examining assortative mating by genes and by childhood shared and nonshared environments, and the roles of genotype-genotype interplays. The study of such complex differentiation would both necessitate molecular genetic indicators and new tools of analysis, although again endocrine and neural pathways may also hold some promise.
We have examined the studies of sources of variation in fertility over time in considerable detail because these are potentially extremely interesting and could lead to insights that go beyond the “black box” partitioning of variance that is usually the only result from such studies. The studies using the Danish twins are the most sophisticated yet in this context. The innovative attempt to use linkages for kinship in the National Longitudinal