社会阶层认知分化的实证研究——基于有序Probit面板模型的估计

文章基于2012与2014年中国劳动力动态调查数据,利用有序Probit面板模型对我国劳动力社会阶层认知分化进行影响因素分析。实证显示,具有先赋性和自致性因素优势的个体更倾向于作出较高的阶层认知;主观认知因素显著影响劳动力的社会阶层认知,"相对剥离感"不利于阶层认知的提升。上述结论关于主观认知解释变量的内生性和有序模型的非正态性设定都是稳健的。此外,这些因素对劳动力社会阶层认知分化的边际概率影响具有一定的地区差异性。这些结果对我们正确评估社会转型期的代际传承、弱势群体生存状态、青年人职业选择、劳动力身心健康等都具有政策意义。

Empirical Analysis on Social Status Cognition Division of Labor Force based on the Estimation of Ordered Probit Model with Panel Data

In this paper we study the mechanism which leads to the divergence of status in labor force based on the data from "China Labor-force Dynamics Survey" (CLDS2012&2014), and analyze the effect upon social status division from three kinds of factors by using ordered Probit model with panel data. Our empirical evidence shows that a person who is in an advantageous position in terms of ascribed factors and self-induced factors is more likely to get higher status cognition. Subjective cognitive factors have a significant impact on a person's status cognition. A sense of "relative deprivation" has a negative effect on the promotion of status cognition. The finding above is robust in the IV estimation and semi-parametric estimation of the ordered Probit model. Besides, the marginal probability effects of those factors on social status cognition of labor force show some regional difference in China. These conclusions can be of importance in policy guidance for us to evaluate the intergenerational transition, the living state of vulnerable groups, the choices of occupation made by young people as well as the physical and mental health of labor force in the period of social transition.