人工智能是否会降低劳动收入份额——基于固定效应模型和面板分位数模型的检验

Will AI Reduce Labor Income Share——A Test Based on Fixed Effect Model and Panel Quantile Model

以2006—2017年的省级面板数据为样本,考察了人工智能对劳动收入份额的影响,检验了人工智能对劳动收入份额影响的边际效应。研究发现:人工智能对劳动收入份额具有显著的抑制作用,其应用能够显著降低劳动收入份额,并且该影响存在区域异质性,即对西部地区的负向影响最大,中部地区次之,东部地区最小,对东北地区的影响则不显著;在劳动力丰裕程度较低的地区,人工智能对劳动收入份额的负向影响较大;随着劳动收入份额的提高,人工智能对劳动收入份额的边际抑制效应逐渐减弱,该效应在中部和西部地区显著,在东部和东北地区则不显著。

Using the samples of the provincial panel data from 2006 to 2017,this paper empirically analyzed the impact of artificial intelligence on labor income share,and tested the marginal effect of artificial intelligence on labor income share.The empirical results showed that,artificial intelligence had a significant inhibitory effect on labor income share.And this negative effect had regional heterogeneity,which had a biggest effect in western region,followed by central region and eastern region,but not evident in northeastern region.Besides,artificial intelligence also had a greater negative impact in the regions with lower degree of labor abundance.Finally,the increase of labor income share may lead to the decline of the marginal inhibition effect,which had an obvious impact in central and western regions,while it was weak in eastern and northeast regions.