EC 204-A1, C1, E1 Empirical Project Information Fall 2025

The empirical paper is a very important assessment for this class. Doing a good job on this paper can have long term benefits in terms of your job and/or graduate school prospects. This is a group project; no individual papers are allowed. Unless necessitated by class size or other circumstances, all groups must have 3 students. Since this is a big project, there are huge positive externalities from group work. Additionally, learning to work in groups is very valuable for your future professional career, be it industry or academics. According to Finley (2021), ability to work effectively in teams is listed as the most important skill among college graduates by employers.

Everyone in the group gets the same points, unless group member(s) bring to my attention about non-participation of one or more members for any of the assessments related to the project. In that case, the non-participating group member(s) will not be earning the same grade as the group on the paper assessments. In the extreme case, their grade may even be a 0 on the entire paper. If you are unable to contact your group mates or they aren’t doing their assigned work on time, please bring this to my attention ASAP, so that I can intervene early on.

The empirical paper requires you to use Current Population Survey (CPS) data. This data is housed in the Integrated Public Use Microdata Series (IPUMS) data repository. All projects MUST  use  CPS  data,  no  other  data  set  use  is  allowed  for  this  project.  There  are  some exceptions, for example, you may use macro level price data to convert nominal variables into real terms. The following are some steps that need to be completed before starting work on your paper.

1.   Have access to Stata either by buying it or from public computers on campus (check options for this semester).

2.   Create    a    login   to    access    IPUMS   –   CPS    by    visiting   the    following   website https://cps.ipums.org/cps/.

3.   Start familiarizing yourself with the  CPS data by reading DOCUMENTATION on the left on the above website.

4.   Practice  downloading  data. You will need some basic familiarity with Stata, a data extraction software, and your login. Details about downloading data from IPUMS-CPS has been  explained  in  a  video  uploaded  under  “Videos  related  to  the  class” folder on blackboard.

5.   Understand the variables in the downloaded data by reading the variable definitions.

6.   Practice cleaning data, creating variables, and running some regressions in Stata.

7.   Learn how to read and analyze regression outputs.

There are three required assessments for this project and the project in total is 30% of your final grade:

1.   Submission of your paper proposal on blackboard (3% of your total grade) by 11:59 pm on W, 10/08

2.   Presentations on Simple Regression Models (7% of your total grade) in-class on W, 11/12; F, 11/14; M, 11/17

3.   Final version of the paper (20% of your total grade) due on turnitin by 11:59 pm on W, 12/10

Details on each of these assessments are below:

1.   Paper Proposal Submission:

You  must  spend  time  as  a  group  examining  the  data  set  before  submitting  your proposal.  A  sample  proposal  form  has  been  uploaded  on  blackboard  under  the “Empirical  Paper”  folder  with  the  questions  below.  Each  group  will  submit  ONE proposal form with a few lines for each of the questions listed below. The document must not exceed one page with a font size of 12. The second page can be for citations.

i)          What is the relationship that you will be examining in your paper? You must be able to examine this relationship using CPS data. (0.5 point)

ii)         Why is the relationship important? Evidence and citation must be presented to support this importance. (0.5 point)

iii)        Time point(s) for your analyses and whether it is annual or monthly and why choosing those time points would be interesting. Note: both x and y need to be available in the same time points. (0.5 point)

iv)        Appropriate main x and y variables to capture the relationship, and more

importantly, how you would clean your x and y variables (what observations would you drop, would you need to create dummy variables etc.). (1 point)

v)          Mention the other variables from the data set that you will be including to

control for endogeneity and omitted variables bias in your relationship. (0.5 point)

The points assigned above are approximate.

Please don’t choose both your main x and y as dummy variables, one of them as a dummy variable is fine. Some popular and doable topics are based on health, fertility (nchild), unemployment, citizenship, voting behavior, racial or gender differences. I talk about gender and racial differences in wages extensively in class, so ideally, you should choose a different dependent variable, if you are passionate about focusing on these  demographic  differentials.  Unique  topics  are  always  welcome,  increase  the chances of publishing, and will earn you some bonus points, but ambition needs to be matched  with  attainability.  To  fulfill  the  requirements  of  the  class,  it  is  more important to have a polished paper by the end of the semester.

This proposal submission will help me give you feedback on your topic and let you know early on whether you are on the right track, and potentially suggest a revised topic based on your interests.

2.   Presentations on Simple Regression Models:

The aim of these presentations isto gauge whether you have correctly estimated your simple  regression  models  and  know  how  to  interpret  the  estimates  from  them. Presentations  are  short,   no  more  than   5-6   minutes  with  some  additional  time allocated for my feedback.

Your slides must contain the summary statistics table, summary statistics according to your  main x,  all  possible  functional  forms  and  how you  cleaned  your x  and y variables. However, you can focus on explaining coefficient, intercept, slope, elasticity and semi-elasticity from only one main regression model. Presentations must include a title for your project.

All group members must be participating equally, not necessarily presenting equally, but must be able to answer all questions that I direct at them.

At this point, copying and pasting the Stata tables on the slides would be enough, make sure the font size is big enough to be visible at the back (my eyes are not great). It is your group’s job to note down my comments.

You will be graded on the following. The points assigned are approximate.

i)           Clarity of your slides (1 point)

ii)         Presentation of tables – this includes steps before getting your tables like data cleaning, paying careful attention to sample size etc. (2 points)

iii)        Interpretation of coefficients (3 points)

iv)        Groupmates participating equally (1 point)

3.   Final version of the paper: The final paper should have the following sections in the order listed:

Title Page:

Must have:

i)          your name(s)

ii)         title

iii)        abstract (no more than 200 words): Abstract must have your research question, the relationship that you are examining, anything new that you did with respect to past research, and your findings. You must also include information on your data set and variables. When you explain your findings, choose the ones that are most interesting/statistically significant. Do not exclude findings from your interaction terms.

Introduction:

Must include the following:

i)           Research question, which is the relationship that you are examining in your paper

ii)         Why that question is relevant (motivation). The motivation needs to be backed by citations. In this section, citations can be from news, policy reports etc.

iii)        A brief description of the economic model(s) used in the analyses

iv)        Summary of the major findings ofthe analyses.

v)          Final paragraph should have the outline of the different sections in the paper (section II is lit review, section III is econometric model etc..)

Literature Review:

Must contain a summary of 3 similar studies conducted in the same area as your paper. When you summarize a paper, explain the innovation in the paper relative to past studies, data set, important variables and results that are relevant to your paper. You should also mention how your paper is similar and different from these studies that you are reviewing. Try to include regression-based papers, which will make it easier to compare similarities and differences with your paper. Explain in terms of the data set, variables, research question etc. If you cannot find papers that analyze the  exact  relationship  that  you  are  looking  at,  try  to  find  close  enough  papers. Literature review should be at most 2 pages.

Econometric Model:

Write the equation for the simple and multiple population regression models and state the parameter of interest (which is the β associated with the main x). Explain the econometric models that you are running – log linear, log log etc, interaction terms, hypothesis tests, specification tests (if we are able to do) that you are doing. Use Microsoft Equation Editor for writing equations.

Explain the economic theory behind the econometric model that you are estimating. How do the additional x variables help you control for omitted variables bias and endogeneity? Why have you chosen to create these interaction terms? Why are you running the hypothesis tests? If you can support your reasoning with citations, that is even better.

There  is  no  need  to  write  regression  equations  for  all  the  models  that  you  are estimating.  In  this  section,  you  should  be  describing  all  the  analyses  that  you’re conducting in your paper. Do not explain the results in this section, that is done in the “Description of Results” section.

Data and Descriptive Statistics:

Explain your data  (years  chosen, description of data set, sample size etc) and the description  of  the  dependent  and  independent  variables  used  in  your  analyses. Explain how you cleaned data from missing observations. Do not devote too many pages in describing the variables, since Table 1 in appendix will have the variable definitions. All tables are in the appendix section. Explain tables 2 and 3 highlighting the  interesting  trends/findings.  You  can  tie  this  to  past  studies/news/general observations. Write in paragraphs.

Estimation Results:

Explain the relevant and interesting results in your analyses (Tables 4 – 7). Say how the main estimate changes in different models. Refer Table#s and model#s in the tables when explaining. Pay careful attention to the units when you interpret. You must  link  your   results  to  past  studies,  wherever  possible.  Results  must  be  in paragraph form. You can choose to focus more on one result but try to mention all tables in your explanations.  In  other words,  don’t just present tables without mentioning them in the text of your paper.

Conclusions:

i)           Reiterate the relationship that you are examining

ii)         Briefly describe how your study is related to other studies

iii)        State your important result(s) again and the broad implications from your results

iv)        Drawbacks/limitations of your study

v)          Interesting extensions from your paper

References:

•   You should include a detailed reference section citing all your work. You can use any of the acceptable formats for academic work (but all references must follow the same format). Please see the following link for APA formatting:

APA Style Introduction // Purdue Writing Lab

•   The references that you cite at the end, should also appear in the main text of your paper. See ‘other considerations related to the paper’ for more information.

•    References must be in alphabetical order.

•   Cite CPS. The citation is as follows. It is also available in this link IPUMS CPS

Sarah Flood, Miriam King, Renae Rodgers, Steven Ruggles, J. Robert

Warren, Daniel Backman, Annie Chen, Grace Cooper, Stephanie Richards, Megan Schouweiler, and Michael Westberry. IPUMS CPS: Version 12.0

[dataset]. Minneapolis, MN: IPUMS, 2024.

https://doi.org/10.18128/D030.V12.0