Academic Statement of Purpose:

I am currently a PhD Student and wish to join your university due to my field of interest, my early research experience in applied econometrics has increased my quest to study econometrics at the doctoral level. I look forward to become a professional academic researcher in the field of econometrics and I am convinced that your university is the best for me to pursue econometrics at the doctoral level.

After joining the University of Michigan I began working on my first project which was a study on the effects of marriage policy in China in the late 1970’s using Duflo (2001) methodology, I was stunned by the dramatic change on marriage age after the implementation of this policy.

However, my result were challenged by the potential possibility of heterogeneity whereby agents participating and complying with the policy might also result in self-selection on marriage age even though the policy is compulsory, this is a common problem that plague most applied social programs (Handbook of Econometrics, Chapter 70). Later on, I encountered a similar problem on a project regarding the impact of social security reform on savings behavior in China, I estimated the substitution effect of compulsory individual pension and regular private saving, and results supported the later reforms on data selected from real individual account data to nominal individual account data which eliminated the heterogeneity problem.

These experiences instilled in me a keen desire to better understand the underlying theoretical analysis on treatment effects in studies and also existing econometrics theories. Keeping these questions in mind, I began my course study at the University of Michigan where I learnt a number of econometric theory courses. With outstanding performance on these courses, I started working as a research assistant for my Professor on the project “two-stage within-group estimator in dynamic panel”, this model extends the standard panel model when the regressors are endogenous and also integrates heterogeneity and endogenous variables separately. However, the standard within-group transformation will give a rise to another source of endogeniety in dynamic framework and therefore further bias the estimates. Therefore, we focus on tracing each source of endogeneity and asymptotical property of estimators. I therefore encountered other dimensions of treatment literature as well the two stage averaging parameter method in this project. Our problem was later solved using Donald and Newey (2001) method

Academic Statement of Purpose:2

whereby the instrument bias was eliminated by choosing subset of instruments and averaging the parameters from each selection.

These two projects were beneficial to me and they formed a solid foundation to work on more advanced theoretical topic, they encouraged me to struggle to rise to a higher and complex level of understanding of econometrics especially recent research on the subject. My current coauthor work with my Professor is the identification of quartile treatment effect when treatment is endogenous under non-separable panel framework. Recent literature on treatment effects emphasizes nonparametric identification of certain parameters, robustness, as well as certain forms of heterogeneity in responses to treatment. ((Handbook of Econometrics, Chapter 70, Chernozhukov and Hansen (2005), Chernozhukov, Fernondez-Val, Newey(2009)).

We are currently working on a project which is targeting quartile treatment identification in non-separable panel using bound restriction. This project is important given that it will help me gain more knowledge and also deepen my understanding of econometrics, the non-separable panel using bound restriction idea can be extended to censored regression, whereby the bound restriction is not automatically applied. In practice, the potential result can also apply to empirical work broadly when the data is selected from the lower and upper quartiles.

I am currently pleased with my performance at the university and particularly proud of working with my professor on the various projects, being an outstanding learner over the years has encouraged me to learn more and also fellow learners have always consulted me for assistance. I believe that my experience, academic qualifications and knowledge confirms my ability to pursue econometrics at the Ph.D. level in your university, my professor has all the time been supportive and I look forward to learn from distinguished professors at your university who will better my understanding of econometrics. I am confident that I will gain valuable research skills in your university and hope to add to the reputation of your university by publishing articles in journals and also the inclusion of my name in the list of prominent scholars at your university. I thank you for considering this application.


Academic Statement of Purpose:2

Chernozhukov and Hansen (2005) Econometrica, 73(1), pages 245-61

Chernozhukov, Fernondez-Val, Newey(2009)"Improving point and interval estimators of monotone functions by rearrangement," Biometrika, vol. 96(3), pages 559-575

Handbook of Econometrics(Chapter 70)

Esther Duflo (2001). The Quarterly Journal of Economics, vol. 122(2), pages 601-646