This Article  • PDF  • Send to a friend  • Save in My Folder  • Save to citation manager  • Permissions  Citing Articles  • Citation map  • Citing articles on HighWire  • Citing articles on Web of Science (17)  • Contact me when this article is cited  Related Content  • Related letter  • Similar articles in JAMA  Topic Collections  • Medical Practice  • Medical Education  • Alert me on articles by topic  Social Bookmarking   What's this?

To the Editor: There is ongoing litigation about whether the National Resident Matching Program (NRMP) restrains competition for residents and suppresses wages.¹ Although the legal issues have been discussed in detail,^2-3 the basic economic question is whether a market organized by a match results in lower wages than if there were no match. We compared similar markets for postgraduate medical fellowships that operate with and without a match.

Methods
Some subspecialties of internal medicine use the Medical Specialties Matching Program (MSMP), while others use no centralized match. We purchased from the American Medical Association the Graduate Medical Education Directory, 2002-2003 data set of fellowship wages by program,⁴ and compared all internal medicine subspecialty fellowships in the United States requiring 3 years of prior residency. Of the 1249 directory entries representing 14 subspecialties in 209 hospitals, 1178 include wages. Sports medicine was excluded because there are only 2 internal medicine programs in that subspecialty, bringing the final number of directory entries including wages to 1176.

Results
The mean compensation of the 4 subspecialties that use a centralized match is $41 963 ($41 357 weighted by number of first-year fellows), while it is $41 171 ($41 248, similarly weighted) for the other subspecialties that do not use the MSMP (Mann-Whitney U test, P = .28) (Table 1). Furthermore, the median, minimum, and maximum wage, and the SD of the distribution of wages of subspecialties that use the MSMP, are all marginally greater than those that do not use this match (Mann-Whitney U test, P>.05 for each).

View this table: [in this window] [in a new window] Table. Compensation and Characteristics of Internal Medicine Subspecialty Fellowships Using and Not Using a Centralized Match

To test the hypothesis that, within hospitals, wages for subspecialties that use the match are different than wages for subspecialties that do not, we computed the average wage within each hospital for subspecialties that use and do not use the MSMP. The Wilcoxon matched-pairs signed-ranks test for 169 hospitals was nonsignificant (P = .96).

Comment
Internal medicine subspecialties that use the MSMP do not offer different salaries than those that do not. Thus, a match is not the feature of these markets that determines wages.

The market for fellowships is not the same as the market for residencies: there are similarities and differences. Although both the MSMP and the NRMP use the same Roth-Peranson matching algorithm,^5-6 stipends for residents are presently funded largely by Medicare, while fellowships are mostly funded by the National Institutes of Health. In each case, additional funds would need to be found if wages were to be increased. Unlike residency training, fellowships are an optional part of a career path. Thus potential fellows have market alternatives; fellowship programs must compete not only with other programs, but with less-specialized medical positions, because fellows could practice medicine without pursuing a fellowship. Relatively low wages accepted by fellows, viewed in this light, represent their willingness to continue to invest in specialized training.

The fact that wages for fellowships are unrelated to the presence of a match suggests that eliminating the resident match would not necessarily increase residents' wages.

Funding/Support: This work was partially supported by a grant from the National Science Foundation.

Muriel Niederle, PhD
Department of Economics
Stanford University
Stanford, Calif

Alvin E. Roth, PhD
Department of Economics
Harvard University
Cambridge, Mass

1. Jung v Association of American Medical Colleges 02-CV-00873 (DDC 2002). 2. Miller FH, Greaney TL. National Resident Matching Program. The National Resident Matching Program and antitrust law. JAMA. 2003;289:913-918. FREE FULL TEXT 3. Chae SH. Is the match illegal? N Engl J Med. 2003;348:352-356. FREE FULL TEXT 4. American Medical Association. Graduate Medical Education Directory, 2002-2003. Chicago, Ill: American Medical Association; 2002. 5. Roth AE, Peranson E. National Resident Matching Program. The effects of the change in the NRMP matching algorithm. JAMA. 1997;278:729-732. FREE FULL TEXT 6. Roth AE, Peranson E. The redesign of the matching market for American physicians: some engineering aspects of economic design. Am Econ Rev. 1999;89:748-780.

Letters Section Editor: Stephen J. Lurie, MD, PhD, Senior Editor.

JAMA. 2003;290:1153-1154.

CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter     What's this?

RELATED LETTER

Effect of a Match on Salaries for Medical Fellows
Amnon Schlegel
JAMA. 2003;290(18):2408.
EXTRACT | FULL TEXT  

THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES

AOA Symposium. Current State of Fellowship Hiring: Is a Universal Match Necessary? Is It Possible?
Harner et al.
JBJS 2008;90:1375-1384.
FULL TEXT  

Effect of a Match on Salaries for Medical Fellows
Schlegel
JAMA 2003;290:2408-2408.
FULL TEXT