HP426 Applied Health Econometrics

HP426 Applied Health Econometrics
STATA Summative

Project 2: Study the effect of being unemployed on health and wellbeing

This assignment requires the empirical analysis of the dataset: British Household Panel Survey:

Waves 1-11, 1991-2002: Teaching Dataset (Work, Family and Health). Description of the data are provided in the BHPS User Guide, description of each variable is in the code book, and the questionnaire is also provided for more details about each variable. Hint: Look at the variable ajbstat.

INSTRUCTIONS

Based on your data analysis, write a short report. There is 3,500 words limit (tables and figures excluded).

Upload also your do file so that we can see how you have prepared your data and the details of your analysis.

The following steps might help you in the analysis.

.     Identify the outcome or outcomes of interest. What variables can be considered outcome variables?

.     Identify your main explanatory variable or variable of interest.

.     Which are the most important controls? Justify your choices. You might want to consider, for example, the most important socio-demographic information and the variables relevant for this specific analysis. Be careful about multicollinearity.

.     Note: This is a real dataset, and real data implies missing data, coding errors, etc. Use your common sense to pick and modify the variables if needed.

.     Once your variables are ready, use a regression technique to find the effects of X on Y(s).

.     Interpret your results carefully. Your regressions need to make sense (for example, do you find what you expected?).

.     Can your results be interpreted in a causal way? If not, why? How can you obtain causal effects? Note: If you cannot solve this empirically, provide the rational of an empirical model that would allow you to find causal estimates for this specific research question.

.     Write a report that contains a carefully done data analysis structured as follows:

Introduction, Data (a nice table with descriptive statistics on your main variables is always welcome), Empirical Strategy (write down the model that you will estimate), Results, and Conclusions.

.     Find in the literature 2-3 papers that study a similar research question and use them to complement your findings. Note: to search for a paper you can go to scholar.google.com or search, for example, in Journal of Health Economics, Journal of Development Economics, American Economic Review, Journal of Political Economics, The Quarterly Journal of Economics, American Economic Journal:   Applied Economics, Journal of Development Studies, World Development, etc.

.     If you want to paste Stata output into word use courier font 8 to make it legible. If you want to include in the same table multiple regression outputs you can use the following code:

发表评论

电子邮件地址不会被公开。 必填项已用*标注