Guidelines for Empirical Project

The empirical project for ECO-6352 consists of identifying an interesting research question and making an attempt to find an answer using one of the econometric methods learnt in this course. The research question may involve exploring a relationship between two or more variables, investigating the change in relationship between two or more economic variables over time using suitable data or trying to answer a question by estimating a specialized econometric model taught in class. The econometric method used should consist of application of one or more of the following major topics covered in this course

• Instrumental Variables and Two Stage Least Squares
• Policy or Program Evaluation using Difference-in-Differences Methodology
• Panel Data with First Differencing, Fixed Effects or Random Effects
• Binary Choice Model Estimation using Probit or Logit
• Limited Dependent Variable Model using Tobit
• Sample Selection Model
• Any other methods covered in class

Major Sections of the Empirical Paper

Introduction: Explain clearly the question you are trying to answer and the reason for doing so and why it is interesting. It is often useful to present a chart that motivates the question you are interested in. State key results. Discuss the organization of the paper.

Previous Literature: Provide a brief literature review of previous research if feasible. Because you have time constraints, this section is optional.

Conceptual Framework: Explain any theoretical model you have in mind. Or simply state why you are interested in the question and expected sign of the relationship between the key variables. What biases or complications you expect.

Data: Is it a cross section or panel data? If panel data, how many individuals and what years does it cover? Discuss briefly the broad characteristics of the individuals/units included in the data e.g. females, males, manufacturing firms, emerging economies etc. What is the source of the data i.e. how did you collect it? It is recommended that you present a table of summary statistics of important variables of interest. The table may consist of mean, median, and standard deviation of the dependent variables, right hand side variables in your econometric specification and instruments if any.

Econometric Model

Results

You must have at least one table of results at the back of the paper. In this section you should discuss the direction, statistical significance and quantitative magnitudes of the relationship you are trying to explore. Make sure that you compare the difference between the results you got from the plain OLS and the advanced econometric method you used. Does it make a difference to use the advanced technique? Compare what you found with what you expected.

Robustness tests

You must have one more table of further results to discuss what happened when you added or dropped variables from the original specification. Are the estimates on the key variable sensitive to inclusion of more right hand sides variables? You may also discuss whether the results are sensitive to changes in the sample i.e. dropping certain observations from the sample.

Limitations and Caveats

Discuss any caveats or limitations in your analysis.

Conclusion

State the original question again and provide a brief summary of your findings. Also discuss implications of your findings in terms of its importance.

References

Mention any papers written by others that you may have referenced to write your own paper.

Tables and Charts

Present all the tables and charts you used to write the paper.