Harvard Tries Grade Deflation

This academic year, Harvard has endured a dark night of the soul, including (among other troubles) facing down the consequences of grade inflation and searching for curricular absolution. In a much-publicized fall report, the university confessed that its current grading system is, in fact, a “problem”—an admission that reportedly left students “soul-crushed” and crying in their beds.

Undeterred, America’s oldest institution released in February a new plan to deflate grades. Will the new plan lead Harvard from the dark of inflated GPAs and into the light of more accurate grading, or will the institution “continue in Oblivion lost?”

Harvard Has a Plan

The primary recommendation of the new plan is a simple cap on the number of A’s. Professors can now assign only 20 percent of their students (plus four additional A grades, regardless of class size) the top mark. While I applaud just about any effort to curb grade inflation, Harvard’s reform gives reason for skepticism. Harvard is not the first university to try such a scheme.

First, Harvard is not the first university to try such a scheme. In 2004, Princeton noted similar problems with grade inflation and prohibited professors from rewarding A’s to more than 35 percent of their undergraduate classes. And it worked—at least a little. In the decade in which the policy existed, the percentage of A’s dropped from 47 to 41.8. A comparable policy, capping most course averages, led to similar results at Wellesley College.

Alas, both colleges reversed course. A report from a faculty committee at Princeton bemoaned increases in student anxiety and poor comparisons to peer universities that had not adopted similar grade caps. In response, the faculty voted in 2014 to remove the cap, and grades once again crept inexorably upward. If past is prologue, then this current initiative from Harvard could be destined for similar tragedy.

More substantially, however, the policy at Harvard is simply too weak. Professors cannot give more than 20 percent of their students A’s. A-’s and B+’s, though? They can award as many as their hearts desire. That’s akin to replacing a diet of Big Gulps and Big Macs with large Slurpees and Quarter Pounders—technically better but an improvement too small to matter. Despite my criticisms, I want this Harvard policy to succeed.

Despite my criticisms, I want this Harvard policy to succeed. Grade inflation is a very real problem. Readers of the Martin Center are likely familiar with the arguments against it, so I won’t rehearse them at length here. But they’re worth emphasizing briefly.

Fundamentally, grade inflation creates poor incentives; when students expect rigorous academic trials, they spend more time studying and thereby learning the material. If an A is expected, why apply the effort? In addition, inflated grades are fundamentally dishonest. Students, parents, and future employers shown inflated grades have a poor sense of an individual student’s ability and, importantly, how he or she compares to others.

With that in mind, Harvard’s modest reform does allow a differentiation between the exceptional and the merely adequate, to its credit. When everyone gets an A, there’s little incentive to strive towards excellence. Capping A’s while allowing A-’s will allow for the recognition of a smaller pool of high-performing students.

Experimenting with Other Solutions

While I commend Harvard for facing the issue head-on and leading a charge against grade inflation—however flawed—other campuses needn’t wait until this experiment plays out to adopt their own policies. Universities’ leaders should take a “try everything and see what works” approach. Fundamentally, caps can do only so much. Incentives must change, or else any institutions attempting correction will find themselves in the place of Princeton, succumbing to pressure. Fundamentally, caps can do only so much. Incentives must change.

At the institutional level, this can and should begin with a simple policy: Don’t do bad things. For example, a recent piece in the Chronicle of Higher Education reveals how “Every Student Graduates” initiatives and faculty performance reviews are often tightly tied to student evaluations. Such modi operandi all but guarantee inflation. Give everyone an A, or else both your boss and your students will shame you and threaten your tenure.

The Chronicle article also describes so-called equitable grading practices, such as those offering no penalties for late work and unlimited opportunities to redo tests. Neither mad scientists nor fringe critical theorists could devise a better set of policies to not only disincentivize high-quality work but actively facilitate and enable bad academic habits. Without consequences for laziness or irresponsibility, a college can expect only laziness and irresponsibility.

The fall report from Harvard also notes that an increasing number of professors have long relied on lower-stakes assignments over the course of a semester instead of high-stakes final exams. Low-stakes, regular quizzes are great for learning. They encourage dispersed studying instead of cramming, and there’s substantial research that the act of recall is a far greater boon to learning than is simple reviewing or rereading.

Professors can implement low-stakes assessments without ditching rigorous final exams. That being said, a learning tool is different from an evaluative tool. When teaching, I could assign homework and expect everyone to get an A without issue, because the intent of the exercise was practice. But, at a unit’s end, a demanding exam pressed students to their cognitive limits and so created a bell curve of mastery and ability. Summative assessments have different purposes than homework and low-stakes quizzes do and should not be conflated with them. Professors can implement low-stakes assessments without ditching rigorous final exams, which possess great utility and will create a broader spread of final grades.

That same Harvard report also found that “faculty shifted from exams and papers to alternate modes of assessment, such as creative assignments and group projects, in the hopes of increasing student engagement with their courses.” Such alternative assessments are fundamentally more subjective, and thereby more prone to inflation, than are exams. You can either balance organic equations or you can’t. You either read and thereby remember the plot points of Shakespeare’s Henriad or you didn’t and therefore don’t. How does one precisely grade a spoken-word poem or a group presentation?

As for correctives outside of stopping unwise practices, instead of caps the application of a curve to final exams could more substantively address grade inflation and might make it easier for professors to score such work. When I had to grade a stack of essays, for example, it was hard to discern and justify in some objective way what exactly warranted an A- versus a B+, but I could certainly tell that Sally’s essay was better than Jimmy’s.

But reform cannot come solely from within. There’s a collective-action problem that necessitates change from outside of Harvard. As Princeton’s experiment exemplifies, if only one institution changes, there’s strong pressure to revert back to inflation. And if an elite 18-year-old can go elsewhere for an Ivy League degree at a fraction of the effort and stress, why wouldn’t he or she? One institution cannot likely withstand such pressure alone.

To encourage more colleges and universities to follow suit, there needs to be public pressure on all institutions towards reform. Last year at the Martin Center, Scott Yenor summarized well what such external pressures could include and what they might accomplish:

The gathering of grade-inflation data—when combined with return-on-investment numbers and other factors—might form the spine of an additional effort at higher-education reform. If grades no longer serve as a signifier of mastery at certain universities or in certain disciplines, institutions could be forced either to drop those disciplines or to develop pre- and post-tests as external checks on accomplishment. Perhaps exit exams could be required to earn a degree in certain disciplines. Accreditors have dropped the ball on this.

I would add to that list ranking systems. If U.S. News & World Report includes academic rigor and ROI in its evaluation systems, not just graduation rates and faculty salaries, then universities that don’t reform will face public pressure and even potential humiliation.

With the long view in mind, grades have been slowly rising since at least the 1940s. It won’t be one policy or a single vibe shift that returns us to pre-World War II-era deflation. It will be the culmination of various successful and failed experiments at hundreds of different colleges and universities, as well as numerous policies and pressures that slowly redirect our institutions of higher education not only toward tougher grades but toward a commitment to academic excellence and rigor. 

Daniel Buck is a research fellow at the American Enterprise Institute, the director of the Conservative Education Reform Network, and a former teacher and school administrator.