养老金缴存与居民家庭风险资产替代  

The Replacement between Pension Payment and Risk Assets of Households

作　　者：任燕燕[1] 王纬 REN Yan-yan;WANG Wei(School of Economics,Shandong University,Jinan 250100,China)

机构地区：[1]山东大学经济学院,山东济南250100

出　　处：《数理统计与管理》2021年第3期463-474,共12页Journal of Applied Statistics and Management

基　　金：山东省社会科学规划研究项目(18CZKJ19);山东省软科学重点项目(2018RZE27029).

摘　　要：对我国机关事业单位2015年初启动的新一轮养老金并轨改革政策效应进行评价,基于中国家庭金融调查(CHFS)数据,利用双重或无偏机器学习政策效应评价方法,实证检验养老金并轨政策对并轨单位工作人员家庭投资决策的影响.从机关事业单位养老金缴存是否发挥积极的养老储蓄作用视角验证改革的效果;该评价方法假定条件较少,评价结论更加科学客观.结果表明:并轨单位工作人员缴存养老金与家庭金融风险资产投资之间存在显著的替代关系,养老金缴存减少了其家庭风险资产投资,并具有时滞效应;双重机器学习模型和Tobit模型稳健性检验验证了养老金缴存政策效应系数的可靠性;我国此轮养老金并轨改革效果显著.Based on the China Household Financial Survey(CHFS)data,this paper evaluates the effect of the new policy of pension reform policy initiated by China’s government and institutions in early 2015 to empirically test the pension consolidation policy to work with the integration unit,using the double or debiased machine learning method.The impact of personnel family investment decisions.The effect of reform is verified from the perspective of whether the pensions of institutions play an active role in pension savings;the method of evaluation assumes fewer conditions and the evaluation conclusions are more scientific and objective.The results show that there is a significant substitution relationship between the payment of pensions by the staff of the joint-track unit and the investment of family financial risk assets,and the payment of pensions reduces the investment of family risk assets;the double machine learning and the robustness test from the Tobit model verify the reliability of the effect coefficient on the deposit the effect of this round of pension reform in China is significant.

关 键 词：养老金并轨 金融风险资产 双重机器学习模型 

分 类 号：F224[经济管理—国民经济] O212[理学—概率论与数理统计]