In 2024, a Japanese university in Osaka announced it would begin tracking students' "campus engagement" by monitoring how often they swiped their ID cards at various buildings. Library visits, cafeteria check-ins, gym entries — all fed into an algorithm that generated a "Student Vitality Score." A professor I spoke to said the score was shared with parents during consultations. Nobody shared it with the students themselves.
That detail — the parents get the number, the student doesn't — tells you everything you need to know about how institutions actually think about measurement. It's not a tool for learning. It's a tool for appearing to manage learning.
Having the data is the easy part. The hard part is building an institution that will actually do something uncomfortable when the data says something uncomfortable. - The School Traitor
This is the disease eating through education systems worldwide, and it's getting worse, because the technology for quantifying human behavior is getting cheaper and faster while the institutional courage to ask "does any of this matter?" remains exactly where it's always been: at zero.
The number goes up, so we must be doing something
There's a sharp new book review making the rounds right now about the pathology of metrics — the way measuring something inevitably changes the thing being measured, and rarely for the better. The reviewer spent over a decade quantifying their own life before realizing the data was mostly telling them what they wanted to hear. That's an individual making that mistake with a Fitbit. Now imagine an entire ministry of education doing it with children.
Japan's MEXT has been on a metrics binge for years. University rankings. Research output scores. Enrollment targets tied to funding. The logic is always the same: if we can't count it, it doesn't exist. And if we can count it, we can optimize it. And if we can optimize it, we're doing our jobs.
But here's what the metrics crowd never wants to discuss: the incentive structure that measurement creates is almost always at war with the thing it claims to measure.
When you fund universities based on enrollment numbers, you get institutions that will admit anyone with a pulse and a tuition check. When you rank researchers by publication count, you get a flood of garbage papers sliced thin enough to maximize entries on a CV. When you track student "engagement" by ID swipes, you get kids walking into the library, swiping their card, and walking right back out. The metric improves. The student learns nothing. Everyone's KPIs look great.
This is not a bug. This is the system working exactly as designed.
The Titan submersible disaster — the one where a billionaire's experimental sub imploded on its way to the Titanic — just got a new report showing that Canadian regulators had multiple chances to inspect the vessel and didn't. The agencies that should have been talking to each other weren't. The information existed. The warning signs were quantifiable. Nobody acted.
That's the part about metrics nobody admits: having the data is the easy part. The hard part is building an institution that will actually do something uncomfortable when the data says something uncomfortable. And universities — like regulatory agencies, like governments — are structurally allergic to discomfort.
A startup called Subquadratic just claimed it solved a fundamental mathematical bottleneck in large language models. Maybe it did, maybe it didn't — the receipts are still coming in. But the AI industry at least has a culture where someone can say "this entire approach has a ceiling" and get funded to try breaking through it. Education doesn't have that. In education, pointing out the ceiling gets you called negative. Or worse: gets your funding reviewed.
The students know. They always know. They know when a class is graded on attendance rather than understanding. They know when a professor is publishing for points rather than teaching for people. They know when the institution's "student success initiative" is really a retention-rate initiative dressed in nicer language. Students are not fooled by the metrics. They're just trapped inside them.
So here's the challenge, and it's directed at every dean, every ministry official, every ed-tech vendor selling dashboards with green and red dots: name one metric your institution tracks that, if the number went down, would cause you to celebrate because it meant students were actually being better served.
If you can't name one, your measurement system isn't serving students. It's surveilling them.
And you already knew that.