Uninsurable Shocks and International Income Convergence
探讨外生不可保险收入来源如何影响储蓄行为和收入动态，在假设回报率恒定下，解释为何收入差距会随时间缓慢消失，对研究跨国或地区收入趋同的学者有参考价值。
The dynamics of production and income growth across countries, or across regions within the same country, are often thought to give evidence on whether returns to accumulation are increasing, constant, or decreasing at the local level. When poor countries or regions are found to grow no faster than richer ones, such lack of is viewed as unfavorable evidence for models in which technology offers high returns to accumulation in relatively capital-poor locations (see e.g., Paul Romer, 1987). When population growth, saving (or investment) rates, and other determinants of long-run income levels in the neoclassical model are controlled for, residual income differences across countries and across regions within the same country appear to vanish slowly, but significantly, over time. This suggests that local returns to accumulation are decreasing if the reference theoretical model takes technology to be uniform across locations, features unexplained and permanent differences in saving behavior, and eliminates or limits capital mobility across locations.' These assumptions clearly oversimplify reality in various respects, and a particular one is considered here.2 The theory that motivates empirical work on income convergence either takes saving propensities as exogenously given or else models them in terms of individual optimization under certainty. Random variability is essential to every empirical regression equation, of course, and real data are better represented by a stochastic steady state with ongoing uncertainty than by steady convergence toward steady-state uniformity (see e.g., Danny Quah, 1993). To explore the possible role of exogenous and uninsurable income sources in the joint determination of savings behavior and income dynamics, here I allow consumption and savings to draw on possibly stochastic location-specific income sources which are independent of past accumulation at the local level and which remain relevant as the aggregate economy grows. Returns to accumulation are assumed to be constant both over time and cross-sectionally, thus ruling out decreasing returns to accumulation as the driving force of income convergence and ensuring that potential capital mobility need not induce actual capital flows. Under the common assumption of constant saving propensities, this framework predicts that incomes should converge over time if the relative importance of nonaccumulated income is crosssectionally heterogeneous. When individuals optimize their savings in light of the size and composition of their accumulated and nonaccumulated income sources, consumption-smoothing behavior leads to persistent, but stationary dynamics for relative incomes and consumption levels. The concluding section briefly discusses the extent to which these saving-based insights may be applicable to international or interregional income dynamics.