UBC Theses and Dissertations
Three essays on applied econometrics Xu, Jinwen
This dissertation consists of three chapters. Chapter 1 investigates how returns to education are related to occupation choices. Specifically, I investigate the returns to attending a two-year college and a four-year college and how these returns to education differ from a blue-collar occupation to a white-collar occupation. To address the endogenous education and occupation choices, I use a finite mixture model. I show how the finite mixture model can be nonparametrically identified by using test scores and variations in wages across occupations over time. Using data taken from the National Longitudinal Survey of Youth (NLSY) 1979, I estimate a parametrically specified model and find that returns to education are occupation specific. Specifically, a two-year college attendance enhances blue-collar wages by 24% and white-collar wages by 17% while a four-year college attendance increases blue-collar wages by 23% and white-collar wages by 30%. Chapter 2 and Chapter 3 study how to perform econometric analysis with complex survey data, which is widely used in large scale surveys. Although it is attractive in terms of sampling costs, it introduces complication in statistical analysis, when compared with the simple random sampling method. In Chapter 2, I study the properties of M-estimators when they are used with complex survey data. To undo the over- and under-representation effects of the complex survey design, it is typically necessary to use the survey weight in M-estimation. I establish the consistency and asymptotic normality of the weighted M-estimators. I also discuss how to estimate the asymptotic covariance matrix of the M-estimators. Further, I demonstrate serious consequences of ignoring the survey design in M-estimation and inference based on it. In Chapter 3, I consider specification testing with complex survey data. Specifically, I modify the standard m-testing framework to propose a new method to test if a given model is correct for a subpopulation. The proposed test has advantages over the standard m-testing, taking account of likely heterogeneity of subpopulation distributions. All of the three chapters deal with heterogeneity of subpopulation distributions, whether or not the subpopulation identity is known (Chapter 2 and Chapter 3) and unknown (Chapter 1).
Item Citations and Data
Attribution-NonCommercial-NoDerivs 2.5 Canada