We estimate the World Distribution of Income by integrating individual income distributions for 138 countries between 1970 and 2000. Country distributions are constructed by combining national accounts GDP per capita to anchor the mean with survey data to pin down the dispersion. Poverty rates and head counts are reported for four specific poverty lines. Rates in 2000 were between one-third and one-half of what they were in 1970 for all four lines. There were between 250 and 500 million fewer poor in 2000 than in 1970. We estimate eight indexes of income inequality implied by our world distribution of income. All of them show reductions in global inequality during the 1980s and 1990s.

The steady state and transitional dynamics of two-sector models of endogenous growth are analyzed in this paper. We describe necessary conditions for endogenous growth. The conditions allow us to reduce the dynamics of the solution to a system with one state-like and two control-like variables. We analyze the determinants of the long run growth rate. We use the Time-Elimination Method to analyze the transitional dynamics of the models. We find that there are transitions in real time if the point-in-time production possibility frontier is strictly concave, which occurs, for example, if the two production functions are different or if there are decreasing point-in-time returns in any of the sectors. We also show that if the models have a transition in real time, the models are globally saddle path stable. We find that the wealth or consumption smoothing effect tends to dominate the substitution or real wage effect so that the transition from relatively low levels of physical capital is carried over through high work effort rather than high savings. We develop some empirical implications. We show that the models predict conditional convergence in that, in a cross section, the growth rate is predicted to be negatively related to initial income but only after some measure of human capital is held constant. Thus, the models are consistent with existing empirical cross country evidence.