国家间价格水平差异的成因——对富人社区效应模型的扩展和检验

The Reasons for the Deviation of National Price Levels: Expanded Rich Neighborhood Effect Model and Test

唐翔(2008)放宽巴萨模型假设条件,提出了更为现实的解释国家间价格水平差异的富人社区效应模型。然而,该模型基本形式与巴萨模型在计量上难以区分。文章将富人社区效应模型做了进一步扩展,构建了能够识别其与巴萨模型区别的计量检验形式。随后,通过放松非技能工人只能生产非贸易品的假设,又构建了符合当代世界外包生产情况的富人社区效应当代模型,使之能对高收入和低收入国家间物价水平的不同表现进行理论预测。通过对基于1990~2010年面板数据的联立方程进行估计发现,在高收入国家中,非技能劳动力比重对物价水平有极大的负向影响,并超过了其通过影响人均收入而对价格水平产生的间接影响;然而,在低收入国家中这种情况并不成立。这表明,富人社区效应模型对国家间价格水平差异的现实解释力是优于巴萨模型的,具有普适意义。

Tang( 2008) proposes a contrasting theory of the Rich Neighborhood Effect for explaining the national price levels. Rich Neighborhood Effect is a more general theory because it looses the strict hypothesis of homogeneous domestic labor force and intersectional labor mobility in Balassa-Samuelson model. However,it's difficult to distinguish Tang's model from Balassa-Samuelson model in econometrics. In this paper,we expand the rich neighborhood effect model and construct an econometric framework for testing it. We also develop a contemporary model predicting different behavior of the price level between high-income and low-income countries,which is confirmed by the empirical panel data from 1990 to 2010. These results are compelling evidence in favor of our model over the Balassa-Samuelson model.