摘要：This paper is the first empirical framework that explains the phenomenon of fast
growth combined with the demographic transition occurring in the United States
since 1860. I propose a structural model that unifies those events through the
role of education: the key feature is that parental education determines
simultaneously fertility, mortality and children's education, so that the
accumulation of education from one generation to another explains both fast
growth and the reduction of fertility and mortality rates. Using original data,
the model is estimated and fits in a remarkable way income, the distribution of
education and age pyramids. Moreover, some historical data on Blacks, assumed to
constitute the bottom of the distribution of education, show that the model
predicts correctly the joint distribution of fertility and education, and that
of mortality and education. Comparisons with the PSID suggest that the
intergenerational correlation of income is also well captured. Thus, this
microfunded growth model based on human capital accumulation accounts for many
traits of American economic development since 1860. In a second step, I
investigate the long-run influence of income inequality on growth. Because
children's human capital is a concave function of parental income, income
inequality slows down the accumulation of human capital across generations and
hence growth. Simulations show that this effect is large
关键词：Unified Growth Theory, Human capital, Technological Progress, Inequality and
Growth
