Reed College
Reed College

Economics 312

Spring 2012
Econometric Project #7

In-Class Debate on Monday, April 9

Sex discrimination in wages

Do women earn less than men in the United States when other relevant determinants of wages have been taken into account? This has been the subject of a huge literature. (For a dated methodological survey, read Chapter 5 of Ernst Berndt's The Practice of Econometrics, which is on reserve.) The usual dataset that has been used to test this question is the Current Population Survey, conducted monthly by the Bureau of the Census. Two teams of four students are assigned to this topic:

  • Team A: Casey Anderson, Brett Beutell, Martis Buchholz, Jess Delaney
  • Team B: Anya Demko, Lauren DeRosa, Allie Hemmings, Zach Horváth

Value of higher education

How much does attending college increase one's wages? Like sex discrimination, this question has been much debated and is often tested using Current Population Survey data. Two teams of four students are assigned to this topic:

  • Team A: Sean Howard, Svetoslav Ivanov, Paige Leishman, Mischka Moechtar
  • Team B: Nick Pittman, Joan Wang, Joseph Warren, Sunny Yang

Which team will argue which side of each of these questions will be determined by a coin toss in class on Monday, April 9.

Available datasets

Econometrics textbooks provide numerous sample datasets that could be used to test these hypotheses. You may use any or all of these that you find helpful to your case, or use other datasets that you find. However, if you use any additional datasets, you must make these available to your rival group no later than classtime on Friday, April 6. You should note that the datasets below have been extracted from a diverse set of studies, many of which were looking at questions quite different from those you are exploring. You are responsible for determining whether the sample observations and variables present in the datasets are appropriate for your analysis. Even though there is not a lot of information given in the source documents, you can infer some information about the data from inspection. For example, if there is a wife age and a husband age for every observation, the sample must consist only of married couples.

  • From Ernst Berndt's The Practice of Econometrics (sources and variables are documented in Chapter 5)
  • From Wooldridge's Introductory Econometrics, 2nd ed.
    • CARD.dta
      Data set is taken from paper on returns to education: David Card (1995), "Using Geographic Variation in College Proximity to Estimate the Return to Schooling," in Aspects of Labour Market Behavior: Essays in Honour of John Vanderkamp,  ed. by L.N. Christophides, E.K. Grant, and R. Swidinsky, (Toronto: University of Toronto Press), 201-222.  This book is not in the Reed Library. I've ordered it through Summit and will make an electronic copy of the final paper available to you when I can. The link above takes you to the NBER Working Paper version, which is probably quite close to the final version.
  • From Hill, Griffiths, and Judge, 4th ed.
    • CPS4.dta  Description
      This is the full version of the CPS4small dataset that you have used in projects. In the text, HGL say that it is from 2008. Their data description file is less than completely informative.
    • CPS.dta  Description
      I believe that this dataset is a leftover from an earlier edition of the text. The third edition discusses a cps.dta dataset that is from the 1997 CPS; I think that this is it.
  • From Stock and Watson's Introduction to Econometrics, 2nd ed.
    • CPS04.dta  Description
      The Stock and Watson description file describes this as "for 2004 (from the March 2005 survey)." There are only a few variables in this dataset.
    • CPS92_04.dta  Description
      This dataset appears to combine the CPS04 dataset with additional observations from 1992, but for the same limited number of variables.

Format of debate

To make sure that we have enough time for each side to make their cases and rebut their opponents' cases, we will begin at 8:45. Each side will have 10 minutes to make a presentation of evidence in favor of their position, with a 10-minute discussion to follow the second presentation on each question. Time limits must be rigidly enforced in order to complete both debates within a 65-minute class session. You may use the classroom projector for your presentations, but you must have your materials loaded and ready to go before 8:45 or the time required will be subtracted from your allocated 10 minutes. 

A written summary of your analysis and results should be submitted to the instructor on the day of the debate. This need not be elaborately literate, but should include the relevant tables, Stata results, and a copy of your presentation materials.