Abolish Grades
The new school year is about to start!
To mark the occasion, I am sharing with you my reflections on rampant grade inflation at American universities—and the surprising step we should take to tackle the problem.
And for those of you who want to keep yourself—or your loved ones—sane in the coming academic year, I’m including a special offer: Take out a subscription for yourself, or for your daughter / nephew / family friend who is heading back to college, and get 25% off!
Grade inflation at American universities is out of control.
The statistics speak for themselves. In 1950, the average GPA at Harvard was estimated at 2.6 out of 4. By 2003, it had risen to 3.4. Today, it stands at 3.8.
The more elite the college, the more lenient the standards. At Yale, for example, 80% of grades awarded in 2023 were As or A minuses. But the problem is also prevalent at less selective colleges. Across all four-year colleges in the United States, the most commonly awarded grade is now an A.
Some professors and departments, especially in STEM disciplines, have managed to uphold more stringent criteria. A few advanced courses attract such a self-selecting cohort of students that virtually all of them deserve recognition for genuinely excellent work. But for the most part, the grading scheme at many institutions has effectively become useless. An A has stopped being a mark of special academic achievement.
If everyone outside hardcore engineering, math or pre-med courses can easily get an A, the whole system becomes vacuous. It fails to make distinctions between different levels of achievement or to motivate students to work hard on their academic pursuits. All the while, it allows students to pretend—to themselves as well as to others—that they are performing exceptionally well. Worse, the system as currently constituted creates bad incentives. To name but one example, it actively punishes those who take risks by enrolling in truly challenging courses.
All of this contributes to the strikingly poor record American colleges have at actually educating their students. As Richard Arum and Josipa Roksa have shown in their book Academically Adrift, the time the average full-time college student spends studying halved in the five decades after 1960, falling to about a dozen hours a week. A clear majority of college students “showed no significant progress on tests of critical thinking, complex reasoning and writing,” with about half failing to make any improvements at all in their first two years of higher education.
In one of the oldest jokes about the Soviet Union, a worker says “We pretend to work, and they pretend to pay us.” To an uncomfortable degree, American universities now work in a similar fashion: Students pretend to do their work, and academics pretend to grade them.
It’s high time for a radical reboot of a broken system.
Why Grades Matter
Grade inflation has deep roots.
Tuition fees are higher than ever. And so universities have come to see students as prized customers whose demands—from lavish dorms and gyms to teachers who don’t unduly interfere with their extracurricular commitments and busy social lives—better be heeded.
Meanwhile, tenure-track jobs are getting harder and harder to find. And so the young academics who make it through the system tend to be laser-focused on the metric that matters most to their careers: publishing in academic journals. Throughout my PhD program, the advice to us graduate students, implicit or sometimes explicit, was not to “waste” too much time on teaching. In the years since, multiple friends who got good academic jobs have been told the same by senior members of their departments who will one day decide whether to grant them tenure. When professors are too busy worrying about their own research to care about delivering truly excellent teaching, it’s easy to see why they won’t object when grade inflation starts to run rampant.
Another important reason why grade inflation has been able to take hold so easily is that many academics have lost faith in the utility of grades and grown uncomfortable with the authority entailed in their role as evaluators. That is in some ways understandable. In the classroom, professors’ primary preoccupation shouldn’t be to assess students for their suitability for influential positions in society; it certainly isn’t mine. Like many of my colleagues, I aspire to help students master complex ideas and develop the ability to think for themselves.
But as it happens, a good grading scheme can be useful in accomplishing our core pedagogical purposes in ways that sometimes go unacknowledged.1 Understanding this is key to understanding why a bad grading scheme does so much harm.
Möchten Sie meine Artikel und Gespräche auf Deutsch direkt in Ihre Mailbox bekommen? Klicken Sie diesen Link und schalten Sie unter Notifications “auf deutsch” an! 🇩🇪
When people debate the utility of grades, they mostly tend to focus on their signaling function to outsiders. Selection committees need to pick who wins a prestigious fellowship. Employers need to pick which applicant to invite for a job interview. Grades give them access to the unvarnished judgment of multiple professionals who have closely observed the candidate’s work.2
This makes meaningful grades a key component of a meritocratic society. There will always be positions in society that are associated with greater prestige and material reward. In the absence of other compelling considerations, the fairest way to distribute these positions is to give them to those who are best suited to excel at them. This is not just a matter of distributive justice; it also has implications for aggregate social outcomes. The principal point isn’t that the smartest student most deserves to become a surgeon; it’s that every patient deserves the best possible surgeon when they’re going in for a perilous medical procedure.
But grades don’t just have an important signaling function to the outside world; they are, first and foremost, meant to give students a clear sense of how they’re doing. When excellent, good and poor students get very similar grades, it is hard for students to know whether they are doing excellent, good or poor work.
This is especially true in America because of a larger cultural emphasis on a certain kind of politeness and a greater fear of giving offense. One of the first things I was taught as a teaching assistant who had recently arrived from England, where feedback is traditionally much more blunt, was to present my assessment on student work in the form of a “poisoned Oreo cookie.” Irrespective of the quality of a student’s work, we were supposed to start and end our comments with chocolaty praise, smuggling the substance of a student’s shortcomings into the creamy filling in the middle. Some students know how to read this code; others focus on the faint praise and, taking comfort in their B+, fail to recognize their serious shortcomings.
The students I am lucky enough to teach are, for the most part, very smart. They find it easy to turn in competent work with relatively little effort. But when half-assed effort is enough to earn them an A or an A-, they fail to recognize how far they fall short of the true excellence to which they could aspire. Talented and hard-working though they may be, many of them never push themselves to live up to their true potential.
Grade Inflation Is Bad for Ambition—and for Social Justice
One of the smartest and most inquisitive students I ever taught at Harvard came to me for advice for what to do after graduation. Since Arvind3 was exceptionally talented, I assumed that he would have the grades to show for it, opening many doors to him. Sheepishly, he told me this was not the case.
How, I asked him, could that possibly have happened? The reason, he explained, is that he consistently took courses that challenged him. He would enroll in advanced classes across a wide range of STEM subjects, from astrophysics to engineering, even when they were far outside his area of specialization. In many, he excelled. But on two or three occasions, he struggled to catch up, and wound up with the Bs or Cs that lastingly hurt his GPA.
Arvind was far more gifted and intellectually ambitious than the average student at Harvard. But judging solely by his GPA, he was one of the worst students at the college.
This predicament comes directly downstream from grade inflation. Back when the most commonly awarded grade at elite colleges was a B- or a C+, this counter-intuitively lowered the stakes for any one course. If you messed up in one, you could compensate by doing very well on another for which you worked particularly hard or happened to have a special aptitude. As a result, a poor grade in one course had a relatively modest impact on a student’s overall GPA.
Now that the most common grade at most four-year colleges is an A, the stakes for each individual course are much higher. Since there is no way for students to distinguish themselves by doing exceptional work, a single negative outlier takes on outsized weight. To get a stellar GPA, a student doesn’t have to be exceptionally good at any one thing; they have to manage risk in every single course they take over the course of four years.4 As a result, today’s grading system has come to express a perverse set of institutional values: “We care much more about your ability to jump through any hoop we put in your path than about your ability to excel in your strongest subject or about your intellectual curiosity for challenging fields outside your main focus.”5
Souhaitez-vous—ou quelqu’un que vous connaissez—avoir accès à tous ces articles et conversations en français? Cliquez sur ce lien et activez “en français” sous Notifications! 🇫🇷
By the same logic, grade inflation also punishes students for uneven performance over time. If you are a middling student with few major life challenges and strong mental health, you will wind up with a high GPA. If you are a brilliant student who really struggles during one term because of a family crisis or some mental health problem, your GPA will tank, never to recover.
This is deeply ironic. Much of the explicit or implicit justification for grade inflation is a concern about social justice. And yet the current grading system favors mediocre kids from stable homes over talented ones from less stable backgrounds.
How to Fix Grade Inflation
As in the classic case of inflation, lenient grading practices are hard to contain once they creep in. When students have the expectation that minimally competent work will get them an A, professors risk upsetting them—and scaring future students away from their courses—if they insist on more demanding standards.6 And then of course there is a simple problem of fairness, one that creates a collective action problem from which no one is exempt: Why should any one professor give students worse grades than their colleagues would, effectively punishing them for taking his or her class?
A few professors have reacted to this dilemma with a sort of private rebellion. Harvey Mansfield, a conservative political theorist who recently retired after teaching at Harvard for half a century, used to tell his students on the first day of class that they would receive two grades: their “real” grade and their “ironic” grade. The real grade had not been affected by grade inflation; even smart and hard-working students had to expect to receive an occasional B. The ironic grade, which would eventually make its way to the registrar, was in line with the university’s inflated standards.
Mansfield’s practice was a neat send-up of the university. And it did fix one pedagogically important problem with grade inflation: the “real” grade he gave to students provided them a meaningful sense of how they were doing, preserving some incentive to push themselves beyond the mediocre level required to earn an A on the official transcript. But his solution is nevertheless insufficient. As Mansfield himself would readily admit, it maintained the charade that GPAs are a meaningful metric of performance to the outside world and didn’t fix the pernicious incentive for students to avoid challenging STEM courses which might result in a poor grade.
This is why the only true solution to the current malaise would be to reintroduce practices that make meaningful distinctions between students. Only when grades actually capture the full range of performance do they fulfill all their useful functions, including their role in sustaining meritocratic institutions; their use as a feedback mechanism for students keen to understand their own performance; and the creation of incentives that encourage intellectual curiosity rather than the management of downside risk.
The most straightforward way to accomplish this would be to grade students on a curve, something that makes grade inflation impossible. Another solution, already implemented by some departments at elite colleges, is to introduce limits on the number of high grades professors can award; this gives academics more flexibility than a strict curve, but blocks them from awarding As to most of their students.7 And just as countries often change the denomination or even the name of their currency after periods of hyperinflation, so it might make it easier for universities and colleges to get serious about kicking bad habits if they introduce a novel grading scheme. If students are used to receiving an A for mediocre effort, a system that places them on a 20-point or 100-point or 26-letter scale might make it easier to impose fine distinctions.8
But here’s the thing: I have pretty much given up hope that any of this is going to happen at scale. Grade inflation serves the short-term interests of the main stakeholders of American universities too well—and the first college president to wage a serious battle against grade inflation would have to navigate too many minefields.
Most university administrators are keen to keep their students happy, ever conscious that the first campus that takes on grade inflation could easily slide down the rankings published by U.S. News & World Report. Most students are perfectly happy with a system that ensures they get a high GPA and leaves them to focus on other aspects of student life. And of course the flattening of grades also avoids the risk of all kinds of embarrassing revelations that might ensue when actual gaps in performance are uncovered. For example, it is surely helpful to any college which fails to raise up to the same standard of performance categories of admitted students with lower average SAT scores—whether they be recruited athletes, the children of rich donors, or members of certain demographic groups—to obscure this fact by awarding virtually all students high grades.
So is there anything we can do?
A Modest Proposal
The grading system at American universities is an embarrassment. The best solution would be to take the simple, if somewhat brutal, steps to end grade inflation. But if that is not in the cards, then it’s time for universities to admit that the emperor has no clothes. If reestablishing more demanding standards turns out to be impossible, then the second best option may be to put an end to the whole charade.
According to my “modest proposal,” universities would make all of their courses pass-fail, a practice that has already been adopted by some elite law and business schools. Students would still have to submit their assignments and meet the minimum standards that are now expected of them. But they would no longer be able to pretend that they had been recognized for exceptional achievements.
Abolishing grades is much worse than a grading system that makes real distinctions between students. But by the same token, it is much better than the status quo. If universities no longer awarded grades to their students, they would:
Stop misleading outside institutions, including prospective employers, into believing that GPAs are a meaningful metric of achievement.
Stop communicating to students that mediocre performance deserves praise or constitutes the limit of what they can achieve.
Stop creating a disincentive for students to take the most demanding courses or taking advantage of the full variety of course offerings at their universities.
As it happens, there is some prominent precedent for such a measure of last resort. In the 1950s, 30-40% of Harvard students had their academic excellence recognized by inclusion in the Dean’s List. By the early 2000s, 92% of upperclass students were included in the list, making it useless as a way to recognize special achievement. As the absurd consequences of grade inflation became difficult to ignore, the university opted for an easy but effective solution: it abolished the Dean’s List.
Sometimes, a system becomes so irredeemably broken that the least bad option is to give up on it, at least for the time being. The grading system at American universities has now reached that stage. Imperfect though that solution may be, it’s time to bin the whole damn thing. And perhaps, in ten or thirty or fifty years, that will allow us to start from scratch.
Some academics don’t believe in grades on deeply ideological grounds. They have principled objections to the whole enterprise. But in this context, I’m thinking of academics whose views are more moderate. They don’t dismiss the idea of grades as irredeemably racist or sexist or “capitalist.” They would probably even prefer a grading system without rampant inflation. But nor have they reflected on how much harder it is to achieve the pedagogical goals we share without the assistance of a clear signaling system that can demarcate different levels of achievement.
Since grades no longer allow such institutions to distinguish between most applicants, they are forced to rely on letters of recommendation. But these are far worse as a signaling mechanism. They rely on a student’s ability to forge a close bond with two or three particular professors; on the skill those people happen to have at expressing a student’s strength in an eloquent way; and on the imperfect attempt to decode the language in a genre of writing that itself has come to be replete with flattery and diplomatically omits any negative information.
Name changed.
The current system also disincentivizes students from taking risks within the courses they take. Want to go for an unusual presentation that may make you stand out but could also go wrong? Or are you wondering whether to challenge the political priors of your professor during class discussion or in your final paper? If there is little upside to standing out, and much to lose from one bad grade, the incentive towards conformity becomes all the stronger.
The same problem also rears its head in other contexts. Rotten Tomatoes displays the percentage of reviews that are positive for a particular movie. This favors inoffensive movies which everyone likes even if nobody loves them. Metacritic, by contrast, displays the mean score of each movie. This favors more adventurous movies that are beloved by many even if they are disliked by some. Both systems probably have their uses, but an exclusive focus on one is likely to lead to serious defects.
As many a professor and teaching assistant will be happy to attest, the best way to get a student to complain—and create a needless bureaucratic headache for yourself—is to give them a B+ for subpar work when they are used to getting an A-.
In principle, a solution which reduces the incentives for grade inflation is better than one which creates an artificial cap on grades. One of the underlying reasons for grade inflation at American universities is that professors grade their own students, for the most part without any attempt at anonymization. This not only makes it socially awkward to give a poor grade; it can also feel pedagogically counterproductive to signal that a student who you are trying to motivate to do better has despite their efforts fallen short of expected standards. British universities mostly get around this problem by ensuring that exams are anonymous; that multiple people grade a student’s work; that at least one of these people was not responsible for teaching the relevant student for the relevant module; and by having external examiners from other universities audit their grading practices in a regular manner. But this system comes with drawbacks of its own, such as requiring academics to spend a much higher proportion of their time on assessing student work, and would require massive structural changes to implement in the American context.
Like radical solutions to monetary inflation, such as pegging a local currency to the dollar, each of these schemes come with serious drawbacks and distortions. For example, some advanced courses really do attract a self-selecting cohort of exceptional students; grading these on the same curve as everyone else would unduly punish students in that class.
I used to teach at both an elite private high school and at an elite college. I am by nature an easy grader, but even I chafed under the pressure to inflate grades, which came not only from parents, but also from the administrations. (Elite private high schools use their records on elite college placements to attract families who will pay the high tuitions and maybe also donate, and colleges’ US News rankings improve directly with students’ average GPAs.)
I love your idea of pass-fail grades, but there is an even simpler fix: blind grading. My son went to Oxford, where all student work is graded anonymously by other professors in each department. So professors aren’t tempted to give out inflated grades in order to boost their evaluations (or just because they like and sympathize with their students and want to cut them a break), and parents and students aren’t able to beg for leniency (or to call in favors). Oxford, to its credit, expects professors to maintain traditional percentages of students who receive a first (about 10 percent), an upper-second, etc. I think this is a good system that removes a number of perverse incentives that have led to grade inflation.
An Immodest Proposal
Of course we agree this is a real problem. perhaps it needs a real solution. Here’s a thought, based on an old management saying that engineers often quote: You cannot improve what you do not measure. Tesla measures the holy crap out of his rockets, and Boeing (not managed by engineers) doesn’t.
We know quite well how to measure, it’s spelled SAT or GRE. So here’s the proposal.
1. Abolish grades.
2. Set up a bureau of education standards (BES)
3. BES commissions private standardized tests in all important fields, and tests at least a random sample from each university (students tested are paid well for their time taking the test).
This forces universities to teach to protect their reputations. If they won’t cooperate, it will be assumed they don’t believe their students learned much.
But here’s the magical part. Employers, who actually do care about what students have learned, will be more likely to hire students who voluntarily take BES’s standardized tests and score well. In this way students volunteer to be graded accurately, or if they don’t (or hide their test scores) they are implicitly admitting they didn’t learn much.
This would only need to be done for a limited number of key subjects in each field. Students who scored well in these, could be counted on to have learned a good deal in the untested courses.
This abolishes compulsory grades but gives everyone about the same incentive to reveal how much is learned or taught, by volunteering to be graded.
This would also drive down the cost of education dramatically. Instead of private scam colleges, we would get efficient learning centers that helped students learn on their own.
Now let's figure out how to build a Persuasion Community to stop identity politics. Yes we can!