I once watched a hiring decision get made in four seconds. Not four minutes. Four seconds.
Someone scanned the top third of a CV, saw the wrong university, and slid it onto the reject pile. Nobody argued. Nobody asked what the person had built, shipped, or fixed at 2am. The filter did the arguing, and the filter had been set years earlier by a manager who no longer worked there.
This is what academic thinking does to a workplace. It does not announce itself. It arrives as a policy line in a job advert, a puzzle question in an interview loop, a promotion given to the person who argues best in the abstract. Then it quietly picks your team for you.
The filter measures schooling, not work
Start with the evidence, because the evidence is not kind to credentials.
In 2022, Paul Sackett and colleagues rebuilt the meta-analysis the whole hiring industry had leaned on for twenty years. They found the old numbers were inflated by bad statistical corrections, and they published a revised ranking of what predicts job performance. Structured interviews came out top at .42, followed by job knowledge tests at .40, biodata at .38, work samples at .33, and cognitive ability at .31.
Look at what those five things have in common. Every one of them measures behaviour or knowledge. A structured interview asks how you handled a real situation. A job knowledge test asks whether you know the thing. A work sample asks you to do the work.
None of them ask where you studied.
Your degree filter is not a weak version of these tools. It is a different measurement altogether. It tells you who had the money, the family stability, and the postcode to sit in a lecture theatre at nineteen. Useful for a university admissions office. Close to useless for predicting who ships good software at thirty-four.
I hire engineers. I have hired people with doctorates who froze the first time production went down, and people with no degree at all who read a stack trace like a doctor reads an X-ray. Both types exist in the pile you filtered in four seconds.
Most firms announced skills-based hiring and changed nothing
The uncomfortable part is not the filter. It is the theatre around removing it.
Harvard Business School's Project on Managing the Future of Work and the Burning Glass Institute studied firms who publicly dropped degree requirements. Their February 2024 report, Skills-Based Hiring: The Long Road from Pronouncements to Practice, found the promised opportunity "has borne out in not even 1 in 700 hires last year."
One in seven hundred. After the press release, the LinkedIn post, the values page update.
The breakdown is worse. Around 45% of the firms studied made no meaningful change to who they hired. Another 20% made early gains and then slid back. Only about 18% earned the label of leaders, and those firms raised their hiring of people without degrees by nearly 20%.
Here is the bit every engineering leader should read twice. At those leader firms, workers without degrees stayed. Their retention ran 10 percentage points higher than their degree-holding colleagues. And people hired into roles which previously demanded a degree saw an average 25% salary increase.
So the firms who did the work got better retention and changed lives. The rest wrote a policy and kept hiring exactly the same people.
The academic reflex does not stop at the door
Removing the degree line from the job advert fixes one symptom. The disease is deeper: we keep running work like a university seminar.
Think about your interview loop. Somebody at a whiteboard, no editor, no documentation, no colleague to ask, solving a puzzle with a trick answer. It is an exam. We built an exam and told ourselves it measured engineering.
Google ran the numbers on their own version of this and killed it. Laszlo Bock, then their head of People Operations, told the New York Times in 2013 brainteasers were "a complete waste of time". His words: "They don't predict anything. They serve primarily to make the interviewer feel smart."
Read the last clause again. They serve primarily to make the interviewer feel smart.
The same reflex runs through the working week:
- The design review where the loudest theoretical objection wins, and the person who has run this exact system in production gets talked over.
- The architecture document graded like an essay, sent back for structure, while the service it describes goes unbuilt for another sprint.
- The engineer with strong opinions on hexagonal architecture and no memory of the last time they were on call.
- The word "rigorous" used to mean slow.
Meanwhile the person who quietly noticed the retry loop hammering the payment provider gets no airtime, because they explained it in plain words and plain words read as unsophisticated in a room trained to reward complexity.
Common sense loses these rooms. Not because it is wrong, but because it sounds too simple to be clever.
Credentials pick your managers too
Now follow the same logic up a level, because this is where the damage compounds.
Ask most organisations how someone becomes a manager and you get a list of proxies. Time served. A qualification. An MBA. The ability to present well to a steering group. Almost none of it involves evidence of getting the best from another human being.
My own research on bad bosses found 99.5% of respondents said they had experienced one or more types of bad boss. Not 60%. Not 80%. Effectively everyone. Those bosses were not chosen at random off the street. They were promoted, most of them by a process which rewarded credentials and confident abstraction over the observable skill of leading people.
You do not get a certificate for noticing someone is drowning. There is no module for saying "my call, my mistake" in front of the team. So we do not measure it, and what we do not measure, we do not promote for. I write more about fixing this on Step It Up HR.
What to run instead
None of this is an argument against education. I love clever people. I love deep technical argument. The argument is against using schooling as a proxy for capability when better instruments sit right there, unused.
Five things to change this quarter:
1. Replace the puzzle with the work. Give candidates a real pull request from your codebase, ideally one with a bug you shipped. Ask them to review it and talk you through what they would do. You learn more in thirty minutes than a whiteboard gives you in a week.
2. Structure your interviews properly. Same questions, same order, every candidate, scored against a rubric written before anyone applied. It is the top predictor in the research, and most teams have never tried it.
3. Delete the degree line and then audit yourself. Six months on, count your hires without degrees. If the number has not moved, you removed a sentence, not a filter. The Harvard research shows this is the normal outcome.
4. Score the plain explanation. In design reviews, ask whoever proposes the complex answer to explain the simple one they rejected and why. Reward the person who makes a hard thing sound obvious.
5. Promote on evidence of leadership, not on credentials. Ask their team. Ask the people who left. Look at who grew under them. Those signals exist already. Nobody collects them.
The uncomfortable question
Look at your last ten hires. Then look at your last ten rejections at CV stage, if your applicant tracking system still holds them.
Do you know a single concrete thing any of those rejected people had built? Or did a filter someone set years ago make the decision for you, in four seconds, while you were in another meeting?
The talent is out there. Most of it never made it past your paperwork.