Published Online: November 9, 2004
Published in Print: November 10, 2004, as New Rankings Rely on Choice


New Rankings Rely on Choice

Article Tools
  • PrintPrinter-Friendly
  • EmailEmail Article
  • ReprintReprints

U.S. News & World Report magazine’s college rankings have guided students for nearly two decades, amid complaints from college administrators. Now a group of academics has devised what it says is a more accurate system based on where students decide to enroll.

The system ends up with results similar to those of U.S. News, although in slightly different order: Harvard University is at the top, followed by Yale and Stanford universities. U.S. News ranks Harvard and Princeton University in a tie for No. 1.

The developers of the method, who outline it in a paper published last month by the National Bureau of Economic Research, say they have used a foolproof barometer: colleges’ relative desirability in the eyes of students, instead of criteria used by U.S. News, such as SAT scores and admissions rates.

The authors are Christopher Avery, a professor of public policy, and Caroline M. Hoxby, an economics professor, at Harvard; Andrew Metrick, an associate professor of finance at the University of Pennsylvania, and Mark E. Glickman, an associate professor in the department of health services at Boston University.

Under their model, each time a student who has been admitted to multiple colleges makes a decision on where to enroll, he or she has chosen which college “wins” in head-to-head competition. The college that wins the most such “tournaments” gets the top spot in the rankings. The model resembles those used to rank competitive tennis and chess players, the authors say.

Mr. Avery said that the matriculation and admissions rates used by other rankings can be easily manipulated by universities.

“The U.S. News & World Report rankings actually pretend to be more scientific than they actually are,’’ he said. But it would not be possible for colleges to manipulate the authors’ model, he said.

Although only 43 states were included in the research, Mr. Avery said that did not detract from its accuracy because of the large number of schools in the survey.

Vol. 24, Issue 11, Page 6

Back to Top Back to Top

Most Popular Stories