非平稳面板数据共同因子与结构变化的交互效应探讨

Interaction Effect Analysis of Common Factors and Structural Changes in Non-stationary Panel Data

数据结构的准确判定是模型设定与参数估计的前提,是保证实证研究结论可靠性的关键。文章对非平稳面板数据模型中共同因子、同质和异质结构变化的最经典检验方法进行分析,再结合实际经济数据给出共同因子与结构变化交互效应的实证结果,结果发现,非平稳面板数据模型中结构变化并不一定独立于共同因子,而有可能是共同因子发生了结构变化。因此,构建非平稳面板数据模型时应通过恰当的检验方法准确判定数据结构,避免因直接套用理论而得到的伪结论。

Accurate determination of panel data structure is the premise of model setting and estimation, and is also the key to ensuring the reliability of empirical research conclusions. This paper analyzes the most classic test method for common factors, homogeneous and heterogeneous structural changes in non-stationary panel data model,and then proposes the empirical results of the interaction effect between structural changes and common factors by using the real economic data. The result finds that structural change in the non-stationary panel data model is not necessarily independent of common factors, which probably means the change in the structure of common factors. Therefore, in the course of constructing a non-stationary panel data model the data structure should be determined accurately by use of appropriate test methods to avoid the false conclusions drawn from direct application of the theory.