一种新的购买力平价汇总方法

A New Purchasing Power Parities Aggregation Method

在国际经济比较中,"可加性"和"特征性"是购买力(PPP)汇总方法应该具有的重要性质,而ICP所用过的所有汇总方法均不兼具这两条性质。本文根据特征性的定义,借鉴GEKS法通过逼近Fisher指数以满足特征性的方式,基于用一个非线性最优化过程逼近Fisher指数的视角,构建了一种最大化特征性的PPP汇总方法——最大化双边特征(MBC)法。实例分析充分说明,MBC法相比当前8种满足可加性的汇总方法,在降低替代偏差的效果上最优。MBC法结合了GEKS法的特征性和GK法的可加性优点,且能降低替代偏差,因此可以改变由于PPP使用目的不同需要选择不同汇总方法的现状。

In international economic comparison, additivity and characteristicity are two important properties that purchasing power parities aggregation methods should have,but all the methods that ICP used before don ＇t conform to these two properties. According to the definition of characteristicity,this paper references the way that the GEKS method approximates the Fisher index to satisfy characteristicity,and finally based on the viewpoint that using a nonlinear optimization process to approximate the Fisher index to build a PPP aggregation method（ named MBC method） which has maximized characteristicity. An empirical example shows that the MBC method is better than the current 8 aggregation methods that satisfy additivity in decreasing the substitution biases. The MBC method combines the merits of the characteristicity property that the GEKS method possesses and the additivity property that the GK method possesses,and also can decrease substitution biases. Therefore,the MBC method can change the situation that we should choose different aggregation method when the purpose of using PPP changes.