Trading Places : Intertemporal Choice , Consumption Dynamics and the Poor
Trading Places : Intertemporal Choice , Consumption Dynamics and the Poor
复制标题
交易场所:跨期选择,消费动态和穷人
DOI：
--
复制
发表时间：
2001
期刊：
影响因子：
--
通讯作者：
Taimur Hyat
中科院分区：
文献类型：
--
作者：
Taimur Hyat
巴基斯坦农村个人消费流动性极高。近60%的人口每年从一个消费十分位数移动到另一个十分位数。本文使用了一个新的14年的家庭面板数据从农村巴基斯坦，复制Jappelli和Pistaferri（2000）的方法，检查跨期消费选择的模型能够解释我们观察到的数据中的消费流动性的高度。在第一阶段，我们研究了实际的转移矩阵和相关的一组流动性指数。我们还利用一个独特的数据集的消费流动性的主观看法，以测试我们的首选类的流动性指数的测量误差的鲁棒性。然后，我们概述了跨期选择的四个一般理论--完全保险模型、经验法则模型、信贷约束模型和永久收入假说--并考察了这些模型对消费流动性的显著不同的影响。虽然Jappelli和Pistaferri（2000）在意大利的背景下研究了这四个模型中的三个，但以前的研究没有明确建模在消费流动性背景下对消极和积极收入冲击具有不对称消费反应的信贷约束模型。我们使用经验分布中的收入和消费参数来模拟消费分布的转移矩阵，以校准模拟。然后，我们比较实际和模拟的过渡矩阵，允许在消费的测量误差。这使我们能够面对的假设，我们已经从理论中得出的数据，从而测试跨期消费选择的四个模型在预测消费流动性在巴基斯坦农村。我们还简要地研究了收入测量误差对我们结果的影响。我们的实证结果拒绝严格版本的完全保险和永久收入假说。然而，强有力的支持版本的永久收入假设时，模型被允许考虑到适度的消费和一定程度的过度敏感性的测量误差，特别是在模型中，建立在流动性约束。在短期内，还为经验法则模式提供了数据支持。我们还研究了1989年至2000年之间的流动模式：这些长期的结果提供了一些额外的支持永久收入假说，而完全保险模型再次强烈拒绝。JEL分类：D31; D91; I30
Consumption mobility among individuals in rural Pakistan is extremely high. Nearly 60 percent of the population moves from one consumption decile to another every year. This paper uses a new 14-year household panel dataset from rural Pakistan, replicating the methodology of Jappelli and Pistaferri (2000), to examine which models of intertemporal consumption choice are able to explain the high degree of consumption mobility we observe in the data. In the first stage, we study the actual transition matrix and a related set of mobility indices. We also utilize a unique dataset on subjective perceptions of consumption mobility in order to test the robustness of our preferred class of mobility indices to measurement error. We then outline four general theories of intertemporal choice—the full insurance model, the rule-of-thumb model, the credit-constrained model, and the permanent income hypothesis—and examine the strikingly different implications of these models for consumption mobility. While Jappelli and Pistaferri (2000) examine three of these four models in the Italian context, no previous study has explicitly modelled creditconstrained models with asymmetrical consumption responses to negative and positive income shocks in the context of consumption mobility. We simulate the transition matrix of the consumption distribution using income and consumption parameters from the empirical distribution to calibrate the simulations. We then compare the actual and simulated transition matrices, allowing for measurement error in consumption. This allows us to confront the data with the hypotheses we have derived from theory, and thus test the four models of intertemporal consumption choice in predicting consumption mobility in rural Pakistan. We also briefly examine the effects of measurement error in income on our results. Our empirical results reject strict versions of the full insurance and permanent income hypothesis. However, strong support is provided for versions of the permanent income hypothesis when the models are allowed to take into account moderate degrees of measurement error in consumption and some amount of excess sensitivity, especially in the model that builds in liquidity constraints. In the short-run data support is also provided for the rule of thumb model. We also examine mobility patterns between 1989-2000: these longer-term results provide some additional support for the permanent income hypothesis, while the full insurance model is again strongly rejected. JEL Classification: D31; D91; I30