ECON EXPERT only - Econ task

1

Applied Econometrics -- Economics 350 Project Outline

Overview Part of your grade will depend upon the completion of an econometric project. The projects can be done individually or in a group, however no group should consist of more than 4 students. Although I do not expect a finished product that is of publishable quality for journals, I do expect that you will demonstrate your ability to convey technical information, to interpret your regression output, and to look at a problem that has some interesting econometric modeling aspects. I expect that you will choose your group by October 3 (Week 2) and your topic by October 10 (Week 3). The literature review for the topic of your choice is due October 24 (Week 5). I also expect you to have a data set, cleaned and ready to go by October 31 (Week 6) so that you can run your regression model(s) before November 12 (Week 8). The printed copy of the project is due by noon December 11 (Finals Week). Grading This project is worth 20% of your final grade and I will use a 100-point scale to grade it. Grading on the final project will comprise the following components: A. Content: what is the purpose of your paper and how do you meet that purpose? Is the economic theory appropriate and explained clearly? Is the literature review relevant and helpful? B. Model: is the econometric model appropriate? Are corrections made for the data limitations, violation of assumptions, etc? (Given what you have learned.) C. Interpretation: are the appropriate tests conducted? Are the results discussed clearly and concisely and tied back to theory? If the paper is policy-oriented, is there a tie between results and policy prescription? D. Presentation: is the paper well organized, do you use the right words, are there errors? Length I expect that your paper will be no longer than 9 pages of text. This number includes the title page, abstract, tables, footnote/endnote section and appendices. I do not want to see 2- or 3-inch margins or spacing larger than double-spaced so that the paper looks longer than it truly is. I expect that you need at least five pages to lay out your ideas and discuss your findings. Format and Style Even though you won’t be expected to publish the paper, I would like the paper to be written as though it were going to a journal. That means that you must have a certain look to the design and that you report your statistics in a particular way. This report must be typed and double-spaced. Use standard rules for spacing on bulletin points and footnotes. Because I hope this assignment will better prepare you for the “real world,” I will expect all papers to be grammatically correct and stylistically appropriate. The parts of the report are as follows: I. TITLE PAGE: includes title, your names, date, class. II. ABSTRACT (1-2 paragraphs): start with a sentence that tells the reader what the paper is about. Then add details (“using data on 425 Drexel students, we regress a student’s GPA on…” “The results indicate…” “Our conclusion is…”) Note: Unlike an executive summary, an abstract briefly tells the reader what the study is about and the final conclusions. You do not have room to write

2

details as you do with an executive summary. III. THE REPORT

A. INTRODUCTION (no more than 1 page) 1. Purpose of the Report 2. Motivation for the Study 3. Scope of the Study 4. General Background Information

B. BODY OF REPORT 1. Review of the Literature: what findings are out there in the literature? This will help you to build up a set of hypotheses. 2. Economic Theory: tell in words what economic theory says about your topic. You may interweave this with the literature review, but at the end of these two parts, you should have developed a set of hypotheses. 3. Econometric Model

a. Description of the Model b. Data Source(s) c. Special problems the reader should be aware of.

4. Interpretation of the Results a. Verbal Interpretation b. Testing

C. CONCLUSIONS (no more than 1 page) This is a summary of what your paper does. It indicates what information findings your paper gives the reader.

IV. ENDNOTES V. TABLES & APPENDICES VI. BIBLIOGRAPHY Research Tips: Data and Sources If you are struggling to get at the literature review or the model, here are some ways to get started: 1. Review the local and national newspapers and magazines. Use them mostly for getting ideas or general background information to motivate your study. 2. Use a textbook related to the topic to get an overview and a list of citations. I often use a labor economics test book to review a theory (e.g., discrimination) and to get some citations of famous works on the subject. 3. Consult the Social Science Index or the Journal of Economic Literature to get a list of citations. 4. Journals related to your topic. Try to avoid the highly technical ones (e.g., Econometrica). 5, Books on the subject area. 6. Possible Data Sources: A. The Internet. There are many datasets on the Internet. One example is the FREC dataset (Federal Reserve Economic Data) at http://www.stls.frb.org/fred/ This has a number of macro variables. The Bureau of Labor Statistics (BLS) and the World Bank also have some data that you can download for free. B. Our Econometric Test Book. It includes some data sets for class. The datasets are described in different part of the text.

3

C. Other Faculty. D. The Statistical Abstract of the United States. http://www.census.gov/compendia/statab/ Please note: When using other sources, be very careful about citing these resources. It is very easy to forget, but that forgetfulness constitutes plagiarism. The general rule I use is: When in doubt, cite! Finally, if possible, use The Chicago Manual of Style as your citation method. Use the parenthetical citations within the body of the paper – (Murtazashvili 2013) or (Murtazashvili and Bernanke 2013) or (Murtazashvili et al. 2013). Use endnotes only to elaborate on a point that does not need to be in the body of a paper. Reporting Your Results Usually students do a very good job at explaining the theory and what others have tested. However, many students do a poor job in explaining their own model(s) and interpreting their regression results. Make sure to have the model written out very clearly. For example, State Welfare Expenditures = B1+B2 # of Welfare Recipients + B3 Federal Block Grant Size + B4 Average Education Level of State + B5 Percent of State Population in the Rural Area + B6 State Legislature Predominate Party + B7 Governor’s party + B9 State Income + error The variables in the model should clearly be explained in terms of the theory you are testing or in terms of common sense. Next, you should include a table (or more than one table) of your results. DO NOT rewrite the above equation with the results. Use TABLES to report your results.

Table 1: Regression Results of State Expenditures on Welfare Coefficient Model 1: Basic Model Intercept 500.25** (200.19) # of Welfare Recipients 345.89*** (140.88) … other variable follow Notes: Data Sources: The National Welfare Database. Standard error in parentheses. Statistical Significance: *** = 0.01 level; **=0.05 level. You would interpret the second coefficient as follows: The coefficient on the number of recipients indicates that on average, states pay $345.89 per every additional recipient. This is expected because other studies have found similar results for the same time period. Blah, Blah,… Really cool interpretation here! You may want to add a second model if you make some sort of model correction or if you have different functional forms (e.g., heteroskedasticity). This will be discussed in class as we move along. Joint work I would like every group member to submit his/her own evaluations of other group members’ effort/work on this project directly to me. These evaluations will be confidential and I will be the only person who sees them.