Peer Effects, Learning, and Physician Specialty Choice
Peter Arcidiacono and
No 1553, Econometric Society World Congress 2000 Contributed Papers from Econometric Society
Sixty percent of medical students switch specialties between the first and fourth year of medical school. These changes have a profound impact on the students' future earnings because there are substantial income differences between specialties. In 1997, for example, the mean income ranged from $323,000 in orthopedic surgery to $133,000 in psychiatry. In this paper we use a unique data set that contains the universe of students who graduated from a U.S. medical school between 1996 and 1998 to examine factors that influence specialty choices. There are possible several reasons why so many students switch specialties. One possibility is that switching occurs because residency positions in high-income specialties such as orthopedic surgery and dermatology are rationed, based in part on a student's performance during medical school. As students learn about their own performance during medical school and learn about the rationing rule -- students who perform well in medical school have a higher probability of entering competitive, high-income specialties -- they adjust their specialty choice accordingly. Another possibility is that the ability and specialty preferences of a student's peer group affects his own performance and specialty choice. Finally, schools might exert their own influence on a student s performance and specialty choice. Empirical Model We divide the specialty decision into two parts. A student arrives at medical school aware of his ability, as measured by the Medical College Admission Test (MCAT) score, and his preferred specialty. Medical students take the Step 1 National Board of Medical Examiners test after their second year of school. In the first part of the model, first-year medical students forecast their performance on the board exam based on their ability, the average ability of their medical school classmates (the ability peer effect), their current preferred specialty, and the school they attend. The first-year specialty preference is included to account for the possibility that students with similar levels of unobserved ability will select the same specialty, and students who plan to enter competitive specialties might work harder in order to increase the likelihood of entering the specialty. The error term from the regression of the board score on the forecasting variables is interpreted as a `performance shock' -- new information a student receives regarding his or her performance in school. In the second part of the model, the probability that a fourth-year student chooses a high-paying specialty is assumed to be a function of their predicted performance on the board exam, the performance shock, their first-year specialty preference, the first-year specialty preferences of their classmates (the specialty preference peer effect), and the school they attend. Data The sample consists of 37,000 medical students who graduated from one of the 127 U.S. medical schools in 1996, 1997, or 1998. We have information on students' ability prior to medical school, their performance during medical school, and their specialty preferences at the beginning and end of medical school. A student's pre-matriculation ability is represented by their test score on the biology, chemistry, and reading components of the MCAT, a uniform exam given to all medical school applicants. The Step 1 board score, also a uniform national exam, measures a student's performance in medical school. Students are surveyed by the Association of American Medical Colleges in their first and fourth years of medical school and asked to indicate their preferred specialty. Because we have the universe of U.S. medical students, we can measure ability and preference peer effects: the average pre-matriculation ability of each school's students and the proportion of each school's first-year students who prefer each specialty. Results Among first-year students, there are no substantial differences in MCAT scores between specialties. By the fourth year, however, students with high board scores are more likely to be in high-income specialties. Students who receive a positive performance shock -- a higher board score than they expected -- are much more likely to switch to a high-income specialty relative to other students. Women are more likely than men to choose a low-income specialty in the fourth year, conditional on their board score and initial specialty preference. Peer effects are an important determinant of board scores and specialty choices, but these effects disappear when we include school-specific fixed effects.
References: View references in EconPapers View complete reference list from CitEc
Citations: Track citations by RSS feed
Downloads: (external link)
http://fmwww.bc.edu/RePEc/es2000/1553.pdf main text (application/pdf)
This item may be available elsewhere in EconPapers: Search for items with the same title.
Export reference: BibTeX
RIS (EndNote, ProCite, RefMan)
Persistent link: https://EconPapers.repec.org/RePEc:ecm:wc2000:1553
Access Statistics for this paper
More papers in Econometric Society World Congress 2000 Contributed Papers from Econometric Society Contact information at EDIRC.
Bibliographic data for series maintained by Christopher F. Baum ().