The Frontier Theory And Application Of The Treatment Effect Models

This paper concludes the basic theory and the frontier development of the treatment effect model and uses the endogenous treatment effect model,the quantile treatment effect model and the multi-valued treatment effect model to solve the problem of the influence of marriage on income.Moreover,for the first time,the unconditional endogenous quantile treatment effect is used in empirical analysis and multi-valued treatment effect is expanded to research the influence of marriage time.The empirical results show that marriage has negative influence on females,not through their withdrawing from labor market,but through their decreased labor time.Furthermore,the influence is different for females with different incomes.Marriage time has a great influence on females’ income,and the late married women have more income than the early married women.Moreover,for the early married women,their income increases as their marriage age increases.But for the late married women,there is not a clear rule.Therefore,this paper has practical significance,which could be referred to study females’ late marriage or even no marriage.The innovative contributions include:(1)applying the endogenous quantile treatment effect model to empirical analysis(2)expanding the application of multi-valued treatment effect model in Labor Economics and Welfare Economics(3)providing a new aspect for research in marriage and income,through constructing three new instrument variables to solve the endogenous problem(4)providing a Chinese empirical evidence about marriage and income.This paper first introduces the research background and content,followed by the research significance,innovations as well as the structure of the entire paper.In the second chapter,the literature review of this paper’s theory method,the effect of marriage and methodology,the treatment effect model,is outlined as a brief summary.And the frontier model theory will be specifically analyzed in the following chapters since it involves comprehensive content.In chapter 3 the ordinary least squares regression and instrumental variables regressions are applied to study whether marriage has impact on female’s income.Without regard to individual heterogeneity,the negative effect of marriage on income of marriage appears significant after controlling for the endogenous problem by using the average age of family members as the instrumental variables.And because cannot identify the effect of marriage out of the difference of individual heterogeneity,we have to apply the exogenous and endogenous treatment effect models to investigate the influence of marriage on female’s income level.Taking the individual heterogeneity into account,the results can yield a much more accurate analysis on very similar females.Results indicate married women earn significantly less than unmarried women after controlling the endogenous problem using whether the women is born in a later married family as the instrumental variable.Since the focus is on the effect of marriage on income instead of the difference between married and unmarried women,the result is expected and less than the instrumental variable regression.The results also show that marriage reduces women’s working hours but does not influence their involvement in employment.Chapter 4 analyzes the effect of marriage on the income of different income levels of women through the use of conditional and unconditional endogenous quantile treatment effect models.In this part the instrumental variable is whether her husband has a mother alive.Results from conditional quantile treatment effect shows that among all income levels,the income of married women are far less than unmarried women.But in the conditional endogenous quantle treatment effect model,with the rising of income levels,the negative impact of marriage on women gradually decreases.Results from unconditional endogenous quantile treatment effect indicates that women with the highest and lowest income had the largest decline on their income level after marriage,while the effect on middle-income women is less and at the 0.5 quantile level,the effect is not significant any more.Chapter 5 uses the multi-valued treatment effect model to discern the effects of the age of marriage on women’s income.Results show that women who marry relatively later on average have a higher income,but not the later the better for the late married women.Also,women with late marriage are more likely to leave the labor market at the time of marriage,but the remaining women work more time.Chapter 6 is the conclusion,explaining marriage’s impact on real life based on this paper’s finds.Also it summarizes the problems from an individual’s welfare perspective.