Last week, the Faculty of Arts and Sciences at Harvard University voted — by a pretty wide margin — to cap the number of A grades in undergraduate courses at 20 percent per class. They did it as a way of managing grade inflation, which has become a big problem at universities around the United States. Simply put, an A doesn’t mean much anymore, and this policy is meant to address that.

As you’d probably expect, the change is getting mixed reactions. Some people think it’s a long-overdue correction, while others think it’s going to create a more anxious and cutthroat student culture. And, to be fair, both sides have reasonable points.

But I keep getting stuck on something everybody in the debate seems to be overlooking…

If Harvard — or anyone, really — actually believed grades were a meaningful measure of learning, the solution to grade inflation would be obvious: Make the courses harder to master. After all, if we raise the bar for what earns an A, then the grade would reflect genuine achievement, and the distribution would sort itself out naturally.

Problem solved!

But that’s not what Harvard did. Instead, they capped the number of A’s that can exist in a class regardless of how well the students in it actually perform. And this tells us the policy isn’t really about learning at all. It’s about managing a signal. Harvard wants to make sure the grade distribution looks a certain way so employers and graduate schools can better use it as a sorting tool.

In other words, Harvard’s faculty just voted to “fix” grades by making them less connected to individual performance, not more. And I think that’s worth discussing because it underscores everything I keep trying to explain in Learning to Fail about the strange relationship all of us develop with grades as a result of spending so many years in school.

The System Measuring Itself

The news out of Harvard is a great reminder that grades aren’t a precise instrument for measuring what any individual student knows or can do. Instead, grades are a tool for managing a very large, complicated, and imperfect system.

Consider, for example, that twenty years ago, A’s accounted for roughly a quarter of all Harvard grades. By last year, that number had climbed above 60 percent. What does that tell us about Harvard students? Did they suddenly get dramatically smarter? Or maybe the faculty got dramatically more effective at teaching?

Obviously, neither of those things are true. What changed was the incentive structure for students, for professors, and for everyone operating inside a system where grades carry enormous downstream consequences and nobody particularly benefits from giving a bad one.

That drift happened slowly, invisibly, and through thousands of individual decisions that each made sense in context and collectively produced something obviously problematic. The system drifted away from measuring learning and toward measuring something more like performance within the system, which is a different thing entirely. And the solution the faculty landed on is, in its own way, further evidence of the same dynamic. Rather than reconnecting grades to learning, they’re recalibrating the distribution so the signal works again as a sorting mechanism.

To be clear, I don’t mean any of this in a cynical or malicious way. It’s just what happens when a bureaucratic system tries to serve millions of people across decades. The tools they deploy drift from their original purposes, and eventually the institution has to make adjustments. That’s not just normal… it’s healthy! But it should also tell you everything you need to know about what grades actually measure and how seriously any of us should take them in terms of being a meaningful representation of someone’s actual capabilities.

What Grades Never Measure

After teaching for nearly two decades, I’ve had students who were genuinely brilliant and original thinkers, but they struggled enormously with grades. I’ve also had students who were exceptionally skilled at getting good grades but who I’d hesitate to confidently call knowledgeable about anything hireable or valuable.

None of this means grades are worthless. But it does mean grades don’t reflect what most people think.

Overwhelmingly, people assume grades equate to things like intelligence, potential, and worth. But none of that is true, and Harvard’s new policy proves it.

If a school like Harvard can simply limit the number A’s, they’re telling us grades aren’t actually a system for accurately measuring people. They’re a system for helping institutions manage large numbers of students. As a result, they measure the pressures and incentives of the institutions.

And I’m telling you this because, if you made it this far, there’s a good chance you (or someone you know) struggles with that reality. You equate getting bad grades with being a failure.

Those aren’t the same things.

If you’ve ever lost sleep over a GPA, or let a rejection letter convince you you weren’t smart enough, or internalized a grade as evidence of something flawed about yourself, take a closer look at that Harvard vote. The people at one of the most prestigious “givers of grades” in the world just admitted grades are a bureaucratic instrument that were never really about students in the first place.

And while that doesn’t mean grades are worthless, it does prove they’re not the objective measure of human potential we were raised to believe.

In the Moment…

A helpful reframing for those moments where failure feels possible.

You don’t have to be a Harvard student to let the lessons of the Harvard vote impact your response to critical assessments.

The next time a grade, a score, or someone else’s evaluation of your performance lands harder than you’d like, try asking a few questions before you let it settle into something permanent:

• What is this actually measuring, and what was it designed to measure?

• Who built this system, and what problem where they trying to solve when they built it?

• What does performing well here actually require — and how much of that overlaps with what I care about getting good at?

• Is this a verdict on me, or is it feedback about my fit with a particular set of rules?

• What would someone who knows me well say this result says about me?

The goal isn’t to dismiss every evaluation that stings. Some feedback is genuinely useful. It’s just worth knowing the difference between a measure and a mirror.

The Debrief…

Reflecting on our experiences is where the most important learning happens.

After you’ve had some distance from an experience where you were evaluated — and especially one where the result didn’t reflect how you felt about your own performance — try asking:

• What did I actually learn here, separate from what the grade or score said?

• Where did my performance within the system diverge from my actual understanding, and why?

• What was I optimizing for, and was that the right thing to be optimizing for?

• What would I have done differently if nobody was going to evaluate the outcome?

• What does this experience tell me about the system I’m operating in, versus what it tells me about myself?

Systems need to measure something. But you don’t have to let someone else’s measurement be the last word on what you’re capable of.