PARENTS HAVE gone to extreme lengths to get their children into highly ranked colleges. Charles Kushner, father of Jared Kushner, Donald Trump’s son-in-law, promised $2.5m to Harvard University in 1998 shortly before his son’s admission to the university. In recent years the “varsity blues” scandal landed a few famous parents in jail for bribing college staff with six-figure sums. But colleges have behaved badly too. A former dean at Temple University’s business school in Philadelphia received a 14-month prison sentence this month for gaming the rankings system. Now Columbia University is in the hot seat.
US News and World Report, a media organisation, began ranking America’s top universities in 1983, and has released it annually since 1988. It ranks over 1,400 institutions in various categories, based on two types of data: a survey that measures the school’s reputation among administrators at peer institutions (20% of the score), and quantitative measures submitted by the school (80%). These rankings matter because they influence prospective students’ behaviour. A study from Harvard Business School and the College Board found that a one-rank drop on US News led to a 1 percentage-point decrease in the number of applications to a college. (The Economist has its own MBA rankings that also use quantitative and qualitative measures to assess business schools.)
Columbia has ranked well for decades. It moved up this year from third place to second (tied with Harvard and Massachusetts Institute of Technology, and behind only Princeton). But a maths professor at Columbia, Michael Thaddeus, alleges the university may have reached the top dishonestly.
Last month Professor Thaddeus posted his analysis of the rankings on his website. He accuses the university of submitting inaccurate data about the university’s class sizes, the number of full-time faculty, the number of faculty with a PhD or similar degree, the student-faculty ratio and instructional spending. These five quantitative measures make up 23% of the score.
In several cases, the numbers Columbia reports seem implausible when compared with peer schools. The university claims that 100% of full-time faculty have “terminal” degrees—the highest degree in a field, such as a PhD or MBA. Its provost office reports that Columbia had 1,602 faculty members in 2020 (4,381 when including medical-school faculty). The website of Columbia College, one of the three undergraduate colleges at the university, lists faculty and their degrees. In that college alone, Mr Thaddeus says he counted about 66 faculty without terminal degrees.
Columbia disputes Mr Thaddeus’s methods. It claims his analysis does not reflect the fact that a terminal degree is not always a PhD, and it quibbles over who is counted as faculty: teaching fellows, graduate students and part-time faculty are excluded from its calculations. Yet its figures are very distant from those of its peers, adding to suspicion: Princeton and Harvard report 94% and 91%, respectively.
Mr Thaddeus also questions the amount of financial resources allocated per student (10% of the overall score). Columbia claims that it spent $102,000 per student in fiscal 2020. This is more than MIT ($82,000), Princeton ($63,000) and Harvard ($45,000). Mr Thaddeus claims that this is a result of including patient spending from the medical school in the calculations. Columbia says that it does not include expenses from its hospital (New York-Presbyterian) in its calculations. It did not explain why its figure differs so greatly from its near-peers.
Class size also differs greatly from peer schools, in Columbia’s favour. The university reports that 82.5% of undergraduate classes have 20 or fewer students. Princeton and Harvard report 77.6% and 76.3%. Mr Thaddeus says the actual percentage is probably somewhere between 63% and 67%, though he had to make several assumptions and had access to limited data, such as the class directory.
“Rankings create tremendous pressure on administrators to optimise their scores,” says Michael Sauder, a professor at the University of Iowa and co-author of a book about college rankings. The quantitative measures are not clear-cut. “There’s often some ambiguity about what counts and what doesn’t count,” he says.
Schools have changed their behaviour to perform better, says Mr Sauder. Some have reduced incoming classes to appear more selective (7% of the score). Others have sent glossy brochures to peer institutions to influence the reputation survey (20%). Universities have given away gifts to alumni to encourage donations, since alumni giving also counts (3%). Northeastern University’s former president, Richard Freeland, unapologetically focused on the ratings, even hiring faculty and capping class sizes to improve its score. The school rose from 127th in 2003 to 49th in 2022.
Columbia will probably pay a price for its suspicious behaviour. Some prospective students may turn to other options. If Columbia is at fault, it could experience far worse. US News says it will “unrank” schools that misreport data, accidentally or otherwise. (Several colleges, including Boston University, have been temporarily unranked for this reason.) Temple has spent millions on lawsuits prompted by the former dean’s fraudulent behaviour.