中文摘要：

基于辽宁省农民工抽样调查数据,采用劳动合同签订和职业流动频次作为农民工就业稳定性的代理变量,对农民工就业稳定性的收入效应及其性别差异进行分析。OLS回归结果显示：是否签订劳动合同变量对农民工工资有正向显著的影响,但对男性农民工的影响更大;职业流动次数对农民工工资的影响呈现非线性关系,但是对男性和女性的影响存在差异。采用Treatment-effects模型校正变量选择性偏差问题,倾向得分值匹配法控制变量内生性问题之后,职业流动次数的估计结果与OLS吻合,是否签订劳动合同度量的就业稳定性对农民工工资有正向显著的影响,但对女性农民工工资的提升效果更大,与OLS结果不同。同时发现,参加过培训、有高层次社会网络、参加工会的农民工就业稳定的概率更高。

英文摘要：

Based on the 2014 survey data of migrant workers in Liaoning Province,The article used the labor contract signed and job mobility frequency as agent variables of migrant workers＇ job stability,and analyzed the income effects of the migrant workers＇ job stability and its gender differences. The OLS regression results showed,whether to sign labor contract variable has a positive significant influence on migrant workers＇ wages,but influence the male migrant workers more; Job mobility frequency＇s influence on migrant workers＇ wages presents non- linear relationship,But the influence between male and female is differences. After apply Treatment-effects model to correct the variable selection bias and control the endogenous variable problem by PSM,the estimated results of job mobility frequency are consistent with OLS,job stability measured by whether sign labor contract have a positive significant influence on migrant wages,different from the OLS results,its effect is larger for women migrant workers＇ wage promotion,The article also found,the job stability probability of migrant workers who participated in the training,had a higher level social network or participated in the union is higher.