Earnings Uncertainty and Aggregate Wealth Accumulation:Comment
评论并推广了Caballero(1991)关于收入不确定性解释总财富的模型，指出引入跨期替代动机后财富存量会显著不同，并强调财富模型应通过财富分布而非仅总量来验证。
There are many motives for saving. Until recently little attention had been paid to the role which earnings uncertainty might play in determining aggregate savings and the associated stock of wealth in a steady-state economy of identical individuals. Jonathan Skinner (1988), Stephen Zeldes (1989), and more recently in this Review, Ricardo Caballero (1991) have developed models which seek to address this. In each case the authors have argued that income uncertainty is capable of explaining a substanitial proportion of aggregate wealth-perhaps as much as half of the actual stock. Zeldes obtained his results by using numerical methods, Skinner by approximating the Euler equation resulting from a constant relative risk aversion (CRRA) utility function, and Caballero by choosing a specific utility function and income generation process which result in closed-form solutions. The present paper reexamines and generalizes Caballero's (1991) findings. It also corrects some presentational errors in his paper. Generalizing a model that focuses upon a single motive for wealth accumulation is important. In particular, we show that by introducing an intertemporal substitution motive for savings, aggregate wealth stocks emerge which can be considerably different from those presented by Caballero (1991) and Skinner (1988). This relatively simple result is important, because many authors have proposed models which focus upon a single savings motive and claim that their model is able to explain a large component of the wealth stock. Franco Modigliani's life-cycle work (e.g., Modigliani 1988) or the work of Lawrence Kotlikoff and Lawrence Summers (1981) on intergenerational transfers are examples. If we were to sum the explanatory power of all such models we could explain the wealth stock several times over! Therefore, building richer behavioral models is necessary. This paper enriches the uncertainty model. The second purpose of our paper is to argue that any model of wealth accumulation should be validated not simply by referring to the aggregate stock of wealth it can generate, but also by examining its associated wealth distribution.