Labor Economics (Fall 2005)

Professor Rizzo

Empirical Paper Project

A major component of this course (20%) is a research project that will lead to the submission of a research paper. These papers must include statistical analyses of data. You are free to work in groups of up to 3 students on this project, but it is perfectly acceptable to work individually or with one other classmate. My expectations for those that work in groups will be higher than for those that work individually.

I have listed below a set of “suggested” topics, many of which derive from my current teaching and research interests, the economics of higher education. At the end of each topic I indicate the section of the course to which the topic relates. These topics are “suggested” in the sense that I know that each is feasible to undertake. However, you are neither restricted, nor even encouraged, to work on these topics. They are meant primarily to be illustrative of the types of projects that I have in mind. You are encouraged to suggest other topics that interest you that relate to the course. You will then be asked to indicate your first two choices of topics and the members of the group in which you would like to work. This choice is to be made to me via e-mail by Monday, September 26th.

Once groups/individuals decide what they would like to work on, each group should meet with me by Friday, September 30th so I can confirm what topic you will be working on, help to get you started with the project and to establish what my expectations are for the paper. You should meet with me regularly during the semester to discuss your progress and any snags that you are encountering. The papers will be due the last day of the semester (Friday, December 9th). If you submit a draft of the paper to me by November 18th, I will provide comments to you by the beginning of Thanksgiving Break, which will give you time to make some revisions.

To adequately address each topic, you will need to have a familiarity with multivariate regression analysis and access to a statistical software package. Some of you may be familiar with SAS or E-Views, or a similar type of software. I do not care which package you decide to make use of. Copies of SAS and STATA can be found on many of the lab computers and instructions on how to them can be found on their respective web pages. Finally, if you are very good at using Excel you may able to do the whole project using it.

Examples and Suggestions for Topics

  1. If you Google the phrase, “Best Places to Live” you will find links to various magazines and websites that have compiled listings of cities/regions that have desirable (and undesirable) characteristics (such as parks, coffee shops, educated professionals, etc.). Furthermore, the government reports a great deal of data (including faculty and staff salary data) for every single college in the United States in a survey called IPEDS (Integrated Postsecondary Education Data System). Salaries of both staff and faculty vary widely across these institutions. To what extent can you attribute differences in salaries across institutions to the desirable/undesirable characteristics of living in or near the town where the school is located? Are there differences in salary responses to living conditions for faculty and staff? (Compensating Wage Differentials, Pay and Productivity)
  2. The Chronicle of Higher Education publishes information on the salaries of top administrators at many colleges and universities. I recall reading somewhere that on average university presidents last 7 years on their jobs. What factors affect how long academic administrators remain in their positions? Are these factors different for liberal arts colleges and research universities? (Mobility and Human Capital)
  3. College and university campus crime data is available from two major sources, the U.S. Department of Education which under the "Jeanne Clery Act" collects statistics from more than 6,000 schools, and the FBI's Uniform Crime Reporting program which includes data for about 400 schools. Both sources are available at What factors affect crime rates and the types of crimes at college campuses? Are crime rates and types of crime on campus related to crime rates and types of crime in the surrounding areas? What factors might have played a role in the changing composition of and occurrence of crimes on campuses over time? (Human Capital, Occupational Choice, Discrimination)
  4. Do crime rates and/or the types of crimes affect any easily observable outcomes for college campuses? These outcomes include graduation rates and time to degree (by race and gender), alumni giving, faculty and staff retention rates, tuition rates, other campus fees, etc.
  5. A very small percentage of National Hockey League (NHL) players are nonwhite. Are there pay differences between whites and nonwhites in the NHL? Do nonwhite hockey players have an effect on attendance at games? Do nonwhite hockey players play more or less than comparable white hockey players? (Discrimination)
  6. We will read a paper entitled “How Race Affects Dismissals of College Football Coaches” later in the term. Can you execute a similar paper looking at College Basketball coaches? Can you examine this question for other levels of athletics? Can you extend it to other areas? (Discrimination)
  7. Many of your Centre professors have been rated using web-sites like RateMyProfessor.com and you guys may also have informal rating systems that I am not familiar with. Are more highly rated professors those that offer higher grades, on average, than less highly rated professors? Do professor ratings correlate with a faculty member’s research productivity? Do average ratings vary systematically by department? By academic rank? Is there any relationship between a professor’s physical characteristics and their student ratings? (Investments in Human Capital)
  8. Each year average salary data is published in the AAUP magazine Academe for female and male faculty at each rank, by institution. The ratio of female to male salaries varies widely across academic institutions, as does the share of faculty that is female? Do these differences reflect discrimination against females? What are the factors that you might expect will lead to gender ratios in salaries and faculty numbers varying across institutions? Present econometric evidence that provide tests of your hypotheses (Gender, Race and Ethnicity in the Labor Market)
  1. There are many ways to measure the research productivity of faculty- two are the number of citations to an individual’s published work and the number of his or her publications. Human capital models predict that the productivity of an individual should vary over the life cycle. What do life cycle productivity profiles actually look like for Centre economists, sociologists, or psychologists? What are the reasons that at a point in time research productivity might vary across Centre faculty members in a field? Present econometric evidence that confirms or does not confirm your hypotheses (Investment in Human Capital)
  1. Salary levels vary widely across colleges and universities in the United States. Theories of turnover suggest that, other factors held constant, higher paying institutions should have lower turnover rates. What are the other factors that one should hold constant? Provide econometric evidence on the relationship of faculty turnover to salary levels (Worker Mobility)
  1. In its Yeshiva decision, the Supreme Court effectively barred collective bargaining for faculty in private institutions of higher education. However, state laws governing public employees permit faculty bargaining in public institutions in a number of states. What determines whether faculty members in a public higher education institution are covered by collective bargaining? Do unionized faculty earn more than comparable nonunion faculty? Provide econometric evidence to document your claims. (Unions and the Labor Market)
  1. Faculty/Student Ratios vary widely across colleges and universities in the United States. Faculty members are an input into the production of research and teaching in academia and the theory of the demand for labor suggests that higher the faculty salary level is, the lower faculty employment should be at an institution. However, if one looks at the data, one observes that higher priced faculty tend to be located at institutions with the highest faculty/student ratios. How can one reconcile this with the theory of demand for labor? Using institutional level data estimate demand curves for faculty to see if, when properly specified, the demand for faculty is a downward sloping function of their salary levels? (Labor Demand)
  1. The distribution of American college and university students, by choice of major, has changed widely over time. For example, the share of students majoring in the arts and sciences has declined over time and the share of students majoring in economics has increased in some years and decreased in other years. What are the factors that you think influence the shares of students choosing different majors, either at Centre, or nationwide? Using time-series data provide statistical evidence on whether the factors that you postulate are important prove to be important. (Occupational Choice)
  1. Over the last 30 years earnings inequality has increased substantially in the United States. This has taken place not only for the workforce as a large, but also for workers with a given level of education. Have earnings of faculty in colleges and universities similarly increased in inequality? If they have, what are the factors that may have caused this growing inequality in earnings? Provide statistical evidence as to whether your hypotheses are correct (Earnings Inequality)
  1. Every decade the National Research Council conducts a study that “rates” the quality of PhD programs and their faculty at major American universities. As part of the last study, they also collected average research productivity data, by graduate field, for faculty members in a number of Arts and Sciences and Engineering fields. Given knowledge of productivity differentials across institutions, would you expect to observe faculty salary differentials across the institutions? What theories of the labor market lead you to expect such differentials? Provide econometric evidence about whether, other factors held constant, higher paying institutions have more productive faculty? (Pay and Productivity, Mobility)

1