Accountability is getting more and more attention in higher education. Democrats will soon release new gainful employment regulations, while Republicans are considering a range of accountability approaches, including risk-sharing that would put colleges on the hook when students can’t repay their student loans. With so many new ideas for accountability systems in the air, it is easy to get overwhelmed. To help distinguish the promising ideas from the noise, I look for whether a proposal is consistent with these accountability dos and don’ts.
Focus on Programs Rather than Institutions
Determining which entity we ought to hold accountable is of the utmost importance. Historically, we’ve used institution-level accountability. But program-level (e.g., the bachelor’s degree in accounting at Texas A&M) accountability is a vastly superior approach. Institution-level accountability lets bad programs at good schools off the hook. For example, when the Obama administration introduced its gainful employment regulations, a theater program at Harvard failed, and Harvard terminated the program. With institution-level accountability, this underperforming program would have stayed under the radar and would probably still be alive.
Institution-level accountability also punishes good programs at “bad” schools. For example, I recently applied a new hypothetical accountability metric to public colleges in Texas, and one of the worst-performing schools was Texas Southern University, which has six programs with poor outcomes for graduates. Yet the school also has eleven programs with good or acceptable outcomes, and those eleven shouldn’t suffer due to the six under-performers. Program-level accountability can treat the eleven successes differently than the six problematic programs. Institution-level accountability cannot.
Use Multiple Metrics
Using multiple metrics is highly advantageous. Multiple metrics will help avoid problems with special cases and are less prone to gaming. For example, if there is only a level target (a minimum level of performance), programs will screen applicants to ensure they only admit likely future successes. If there is only a growth target, high-performing programs may struggle to show continued growth. But adding a growth-rate target to a minimum-level target would approve both high-achieving programs (meeting the level target) and rapidly improving programs (meeting the growth-rate target).
Use Tiers and Sliding Scales
The traditional accountability system is binary. You’re either approved, in which case you can enroll an unlimited number of students using federal financial aid, or you’re not approved, in which case you can enroll zero such students. These radically different outcomes are needlessly arbitrary. A better approach would use sliding scales or tiers of aid access. One tier might qualify for Pell grant aid but not student loans. Or caps on enrollment growth could be tied to performance on an accountability metric.
Use Carrots and Sticks
Historically, accountability systems have only used sticks. There have been three accountability metrics for higher education: the cohort default rate (applies to all institutions), the 90-10 rule (applies to for-profits), and the gainful employment regulations (applied to vocational programs—ended by the previous administration but coming back in modified form under the current administration). They only used a stick—identifying which institutions would no longer be able to participate in federal aid programs like Pell and federal student loans. No wonder higher education is reflexively hostile to accountability—it can only hurt. But if we used carrots as well as sticks, successful programs and colleges would see an upside to accountability. The carrots could be monetary (performance bonuses) or regulatory relief (waive accreditation or state authorization requirements).
Don’t Use Subjective or Easily Manipulated Metrics
While graduation rates and student satisfaction matter a great deal, both would be terrible metrics to use in an accountability system. Student surveys are too subjective. One student’s five-star rating might be another student’s three-star rating. Moreover, survey results could be easily swayed by selectively misplacing or purging the contact information of those likely to give a school a low rating. And colleges can also manipulate graduation rates. For example, when Chile started holding colleges responsible for the debts of dropouts, the graduation rate rose as colleges lowered their graduation standards.
Don’t Try to Rank Colleges
An accountability system should safeguard taxpayer, student, and parent investments in education. It is not supposed to be, nor can it succeed as a ranking system. As Beth Akers said, “The federal government should not be in the business of sussing out and celebrating the ranks of good, better, and best colleges. Instead, it should aim to identify and levy punishment on the ranks of bad, worse, and unacceptable colleges.”
Don’t Unintentionally Incentivize Bad Behavior
No accountability system will be perfect, but we ought to consider what actions and behaviors a proposed accountability system will incentivize. Scholars have long complained that the dominant college rankings reward higher spending, and thus tend to spur a spending arms race among colleges seeking to climb up the ranking ladder. If you don’t want to encourage higher spending, then make sure your accountability system doesn’t incentivize or reward higher spending.
Hopefully, this list of accountability dos and don’ts will help you evaluate which accountability proposals have potential, and which ones don’t. The push for accountability is a good thing, but we need to ensure we pursue it in the right way.