Measuring college effectiveness by graduation rates is an important new discussion in higher education policy. Ohio, Pennsylvania, Indiana, Tennessee, Washington state, and Louisiana have now incorporated some performance measurements, based on graduation numbers, to their funding for public colleges and universities

There are very good reasons to evaluate colleges, and sanction them, based on performance. The graduation rates at many American colleges are abysmally low, and funding colleges based on enrollment alone (which is how most state colleges work) doesn’t encourage high performance.

But the graduation rate measure is an awfully crude tool, and it might encourage some pretty terrible behavior in college administrators. This appears to be what happened at one school in North Carolina. Shira Hedgepeth writes at the Pope Center for Higher Education Policy that something went very wrong at Winston-Salem State University (WSSU), a public, historically black college in North Carolina:

Many of the entering students were not prepared for college and required remediation before they were able to take regular college courses. Thus, many of these students were unable to make it through to graduation and that put pressure on the university and faculty to increase success rates. The university’s four-year graduation rate for the university was a mere 14 percent. Since funding from the state is based on graduation rates and a school’s accreditation can be jeopardized by low rates, campus officials were extremely concerned.

Because of WSSU’s low graduation rate, school officials had to decide whether to change its admission standards and move to attract a higher quality of student, or to lower course standards to make it appear that students were more successful. During faculty and staff meetings, this was discussed and the chancellor stated that he did not want to change the “look” of WSSU. Raising admission standards would mean fewer students on campus and therefore less state money flowing in.

Well that’s a pickle.

The path the administration took was to make it seem as though students were doing better. Of course, there was no official statement to that effect, but I realized that was the case after faculty members began to share their concerns and frustration with me.

Many stated that they were under pressure to decrease the material in the courses and increase the passing rates…. Some said that they thought school officials were changing grades they had entered.

Eventually, Hedgepeth said, she discovered obvious cheating by students that faculty felt afraid to address.

Before I could learn more about the erosion of academic integrity at WSSU, I was fired. In 2011, after I had been promoted to director of academic technology, one of workers in the department, unhappy that a white person had been chosen, repeatedly attacked me verbally. After mediation with her failed to bring results, and because of other complaints about her performance, I recommended that she be released. I was stunned when, shortly thereafter, I was notified that I had been terminated.

The facts of Hedgepeth’s case are a little unclear. This is her impression and the truth may be a little more complicated. But she brings up something pretty disturbing about the graduation rate performance measure. Funding colleges based on their graduation rates doesn’t just lead colleges to improve if their rates are low; it also leads them to cheat.

Oddly, Hedgepeth puts the choices colleges with low graduation rates face between two things: “change its admission standards and move to attract a higher quality of student, or… lower course standards to make it appear that students were more successful.”

Note that “do a better job educating students” is not, in Hedgepeth’s mind, considered an option.

Daniel Luzer

Daniel Luzer is the news editor at Governing Magazine and former web editor of the Washington Monthly. Find him on Twitter: @Daniel_Luzer