大学生性别工资差异分解研究

发布时间：2018-03-21 10:01

本文选题：性别工资差异　切入点：分解　出处：《南京财经大学》2014年硕士论文　论文类型：学位论文

【摘要】：高校毕业生就业难已成为不争的事实。女大学生受自身因素限制和歧视的影响,遭受更大的冲击。有关数据显示,女大学生在就业和工资方面与男大学生存在明显差距。国内学者多从定性角度研究此类问题,较少从定量角度分析。本文从理论和实证方面对大学生的性别工资差异进行分析,并提出针对性的政策以解决这一性别歧视现象。首先,本文对女大学生就业和工资现状进行分析,并通过人力资本、歧视、社会学和专业隔离理论详细解释女大学生工资劣势的理论依据和客观性。其次,对微观数据进行OLS和分位数回归分析。实证结果表明：一方面男女大学生在人力资本、社会资本等个人禀赋特征方面存在差异,女大学生的人力资本回报率远低于男大学生。另一方面,性别为“女性”对大学生的工资存在显著负影响,并随工资水平提高而增大。接着,本文选取均值分解方法中最基本的Oaxaca-Blinder分解,分布分解方法中选取基于无条件分位回归模型的Melly(2006)分解、基于经典线性回归模型的JMP1993分解和基于半参数模型的DFL分解方法以便进一步分析性别工资差异中歧视造成的差异所占比重。分解结果显示,大学生的性别工资差异主要来自于系数差异(歧视)。分解结果从均值层面而言,JMP1993的分解结果中,歧视造成的工资差距占总差距的比重最大,其值为99.96%;其次是DFL和Oaxaca-Blinder分解方法,其值分别为93.81%和91.66%；Melly(2006)分解方法得出的平均歧视程度在72.48%～85.27%之间。分解结果从分布层面而言,随着工资分位点的上升,禀赋特征差异由正变负,这表明男大学生的个人禀赋特征优势逐渐减小,最终转化为女大学生的禀赋特征优势；大学生的性别工资差异逐渐减小,而歧视程度却逐渐增加,这说明位于工资高端的女大学生遭受更大的性别歧视,即只有非常优秀的女生才能获得与次优秀男生同样的工资。最后,本文解释歧视的原因,一方面是由于偏好歧视和统计性歧视的双重作用,另一方面是由于双重劳动力市场分割和专业隔离所造成。因此,本文提出五点政策建议：第一,完善法律监督和社会保障体系；第二,将女性生育成本社会化；第三,消除专业性别隔离；第四,努力提高个人人力资本积累；第五,建立现代性别文化氛围。
[Abstract]:It has become an indisputable fact that the employment of college graduates is difficult. Female students, affected by their own restrictions and discrimination, suffer a greater impact. Relevant data show that, There is a significant gap between female college students and male college students in employment and wages. Domestic scholars study this kind of problems from a qualitative perspective, less from a quantitative point of view. This paper analyzes the gender wage differences of college students from the theoretical and empirical aspects. And put forward targeted policies to solve this gender discrimination phenomenon. First, this paper analyzes the employment and wage status of female college students, and through human capital, discrimination, Sociology and professional segregation theory explain the theoretical basis and objectivity of female college students' wage inferiority in detail. Secondly, OLS and quantile regression analysis are used to analyze the microcosmic data. The empirical results show that, on the one hand, male and female college students are in human capital. There are differences in the characteristics of individual endowment such as social capital. The rate of return on human capital of female college students is much lower than that of male college students. On the other hand, gender as a "female" has a significant negative impact on college students' wages. Then, we select the most basic Oaxaca-Blinder decomposition in the mean decomposition method and the Mellyn 2006 decomposition based on the unconditional quantile regression model in the distribution decomposition method. JMP1993 decomposition based on classical linear regression model and DFL decomposition based on semi-parametric model are used to further analyze the proportion of discrimination in gender wage difference. The gender wage difference of college students mainly comes from the coefficient difference (the result of decomposition of discrimination.) in terms of the average level, the wage gap caused by discrimination accounts for the largest proportion of the total gap, its value is 99.96; the second is the decomposition method of DFL and Oaxaca-Blinder. The average degree of discrimination obtained by the decomposition method is between 72.48 and 85.27%. In terms of distribution, the differences in endowment characteristics change from positive to negative with the increase of wage score. This indicates that the advantages of individual endowment characteristics of male college students gradually decrease, and finally turn into the endowment characteristics advantages of female college students, the gender wage difference of college students gradually decreases, but the degree of discrimination increases gradually. This shows that female college students at the high end of the salary suffer greater gender discrimination, that is, only very good girls can get the same salary as the second best boys. Finally, this article explains the reasons for discrimination. On the one hand, it is due to the dual role of preference discrimination and statistical discrimination, on the other hand, it is caused by dual labor market segmentation and professional segregation. Therefore, this paper puts forward five policy recommendations: first, Improve legal supervision and social security system; second, socialization of female reproductive costs; third, eliminate professional gender segregation; 4th, strive to improve personal human capital accumulation; 5th, establish a modern gender culture atmosphere.
【学位授予单位】：南京财经大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F249.24

【相似文献】