Proposal Seven: Rank Universities on Basis of Costs Quality, Graduation Rate and Income of Alumni
In 2013, President Obama announced plans to create a federal rating system that would allow parents and students to easily compare colleges. He also planned to encourage Congress to pass legislation to link student aid to the rating system. President Obama’s proposal was bitterly opposed by college presidents.
On September 12, 2015, the Obama Administration abandoned its proposal to rank colleges let alone tie the ranking of the colleges to student aid and simultaneously introduced a web site providing raw statistics on university performance including
information on annual costs, graduation rates, and salaries after graduation.
Readings on the Obama Era Efforts:
New York Times Article on Announcement by President Obama on plan to rank colleges on results and costs and link financial aid to the college rankings
New York Times Article on Reaction to College Rating Proposal by University Presidents:
Ner York Times article on Obama Abandoning Plan to Rank College
A Discussion of the College Score Card web site:
The web site
introduced by the Obama Administration to provide information on the cost and value of colleges has some interesting raw data and a lot of limitations. The information on this web site for each school includes – average annual cost, graduation rate, salary 10 years after attending school, percent of students receiving guaranteed loans, typical loan amount for student borrowers who finished the program, percent of graduates who earn more than a high school graduate, SAT and ACT scores, demographics of students, and type of programs.
• The debt totals excluded private student loans and PLUS loans. The site also does not have separate information on percent of student using private loans and PLUS loans for each school. These loans often are cause financial hardship for borrowers. The use of these loans is likely to widely vary across schools. The omission of information on private loans and PLUS loans will have a substantial impact on financial risk measures across schools.
• The ScoreCard software only reports median debt levels by student borrowers at a school. It is likely the dispersion of debt levels is higher for some schools than for other schools. It would be interesting to know the percent of borrowers with debt levels exceeding certain amounts ($50,000 or $75,000) at each school or debt levels at the 90th percentile for each school. These alternative statistics are a better measure of the likelihood a student attending a school borrows too much.
• The repayment statistic in the database — the share of students who paid back at least one dollar on their student loan three years after leaving school — is not particularly useful. Initial repayment rates could be low at a school where many alumni go to graduate school. The default rate, delinquency rate, and the proportion of students not in graduate school who are on target to finish payment on their student loans in 10 years are all more useful measures of repayment difficulty.
• The benchmarks in the database based on national averages are not particularly useful. Common sense suggests that the College Score Card cannot be used to compare the value of a state university to the value of an elite private institution. The difference in post-graduate earnings is likely largely determined by the difference in the talent as measured by SAT score of the student body. The difference in debt is likely determined by the price tag and the amount of available aid. A ranking that found Harvard University graduates earned more than students from Ohio State but owed more at graduation would not provide new news. However, universities in the same category (highly selective, moderately selective, not selective) could and should be compared with universities in the same category.
Concluding Thoughts: President Obama’s proposal to rank colleges based on their costs and the value they give their students was slammed by university presidents who proved they have a lot in common with executives running tobacco firms or leaders in the financial services industry opposing regulation after collapsing the financial system. Statistical analysis can and should be used to compare similarly situated universities and statistical information should also be used to guide the allocation of resources.