On the basis of anthropologists' area-specific research of primogeniture, I propose a Becker-Barro type of dynastic model in which primogeniture may emerge as family heads' optimal policy to minimize their respective lineal extinction probability. I show that parents' bequest division will be affected by the intrinsic mobility structure of the society, contrary to previous causality conjecture. Although an unequal division of bequests widens the within-generation inequality, it may increase the intergenerational upward mobility for the well-endowed child, which in turn improves the originally rigid upward mobility of the poor and middle-income groups and, as a result, reduces the steady-state income inequality.

On the basis of anthropologists' area-specific research of primogeniture, I propose a Becker-Barro type of dynastic model in which primogeniture may emerge as family heads' optimal policy to minimize their respective lineal extinction probability. I show that parents' bequest division will be affected by the intrinsic mobility structure of the society, contrary to previous causality conjecture. Although an unequal division of bequests widens the within-generation inequality, it may increase the intergenerational upward mobility for the well-endowed child, which in turn improves the originally rigid upward mobility of the poor and middle-income groups and, as a result, reduces the steady-state income inequality.

One question development economists are especially interested in, but so far left unanswered, is: how would the societal income distribution be affected by introducing a family-planning program to reduce the reproduction rate of the poor, which is usually high in developing countries? The purpose of this paper is to search for analytical answers to this question. We are able to make definite comparisons about some class of inequality measures of the steady-state societal income distributions, and these comparisons provide strong theoretical support in favor of the above-mentioned family-planning program.

We treat each age-specific income-earning member of the family as an income “source,” and use the source-specific Gini decomposition approach as well as the Lorenz comparison approach to study the impact of the changing population age structure on family income inequality. Empirical analysis using Taiwanese data shows that the pattern of Gini coefficients is significantly affected by the above-mentioned age composition factor. The general implication is that for many developing countries which have recently gone through rapid demographic transition, family income inequality indexed may implicitly embody information as to the age-specific composition of family members, which is irrelevant to the general notion of inequality.