Economics
312: Theory and Practice of Econometrics
Reed College --- Jeffrey Parker
Spring 2007
Project #7: Debates to be held in lab on
Thursday,
April 5 and in class on Friday, April 6
Debate Teams
| Sex Team 1 | Sex Team 2 |
| Liz Alexy | Erica Strachan |
| Sakib bin Salam | Andrei Stephens |
| Amanda Bittaker | Eric Schmitt |
| Jose Fernandez | Kris Russell |
| Rachel Ferst | Daniella Morar |
| Race Team 1 | Race Team 2 |
| David Jackson | Josh Lingerfelt |
| Andreea Jurculet | Tolga Yilmaz |
| Upasana Khadka | Avery Ucker |
| Wyatt Laikind | Michael Turner |
| Karim Lakhani | Matt Summers |
Assignment:
On Thursday, April 5, in lab and in class on Friday, April 6, we will have two econometric debates. In the first (Thursday), the issue will be whether there is strong evidence of sex discrimination in wages. In the second (Friday), the issue will be race descrimination in wages. The teams above have been assigned to one or the other debate. I will draw a random number at the beginning of class to determine which team in each debate will be attempting to demonstrate that there is discrimination and which will be trying to show that discrimination is not a major problem. Each team must be prepared to argue either case!
You should begin by reading Chapter 5 from Ernst Berndt's The Practice of Econometrics, which discusses empirical studies of wage determination. You may also want to read some of the papers he cites in this chapter or some more recent studies. Berndt provides data sets from the 1978 and 1985 Current Population Surveys with which to test these hypotheses. The variables in these datasets are described in Berndt_CPS.pdf and the datasets themselves are available in Stata format as cps78.dta and cps85.dta.
Each debate will last 50 minutes. Each team will have 12 minutes to present its case and justify its procedure, with the team arguing for the existence of discrimination presenting first. After both teams have presented their cases, each will have an 8-minute followup to respond to the other team's arguments. The final 10 minutes will be for class discussion. Note that the presentation and discussion should focus on the validity of the statistical/econometric methods; on sensitive issues such as these it is important to maintain a dispassionate and professional attitude.
You may seek additional datasets with which to test your hypothesis and demonstrate your conclusion. However, as in a legal proceeding, any additional datasets you use must be made available to the judge (me) and to the opposition in a timely manner, i.e., as soon as you have them in usable Stata format.