In 2017, a Harvard Business School report, called Dismissed by Degrees, put a number on something a lot of people had been sensing. Drawing on a large database of job postings, the researchers found that in 2015, 67% of postings for production supervisors asked for a bachelor’s degree, while only 16% of the people actually working as production supervisors held one. They called the phenomenon “degree inflation,” and the finding made rounds in the usual places: the Atlantic, a few HR trade blogs, a LinkedIn think piece or two. Then, mostly, people moved on.
I’ve been thinking about that gap lately. Sixty-seven percent of postings versus sixteen percent of workers. Someone decided that the credential mattered, even though the evidence of the job itself said otherwise. And once enough employers made that call, it became self-reinforcing: the posting asks for a degree, candidates without one don’t apply, so the hired pool has degrees, so the next cycle of postings asks for a degree again. The number justifies itself.
This is the specific mechanism I want to talk about, because I don’t think the general conversation about credential inflation actually describes it accurately. The shorthand version goes: employers got greedy, they added requirements that don’t reflect the job, it’s discrimination against people who can’t afford college. That’s partly true and not unimportant. But it skips the part that I find both more unsettling and more useful to understand, which is that a lot of credential creep isn’t cynical. It’s lazy, and lazy is harder to fight.
Here’s what I mean. When a hiring manager opens a requisition and writes down requirements, they’re doing it under time pressure, with imperfect information, using whatever mental template came before them. The last person in the role had a degree. The other candidates who came in through the pipeline had degrees. The path of least resistance is to write “bachelor’s required” and move on. There’s no audit, no moment where someone asks whether the requirement actually predicts success at the job. It just accumulates, posting by posting, year by year, and the bar rises without anyone deciding to raise it.
There’s a second chapter to this research, and it complicates the picture in a useful way. In February 2022, the Burning Glass Institute (with Joseph Fuller, the same Harvard researcher behind the original report) published The Emerging Degree Reset. Analyzing more than 51 million postings from 2017 and from 2019 through early 2021, it found that employers had started moving the other direction: dropping the bachelor’s requirement across a wide range of roles. Some 46% of middle-skill occupations and 31% of high-skill ones saw what the authors called a “material” reset between 2017 and 2019, meaning the share of postings asking for a degree fell by more than five percentage points. The pandemic’s tight labor market accelerated the shift, but it had already begun.
The part that gets buried is what the authors concluded about why. It would be easy to read the reset as pure market leverage: employers drop the degree when they’re desperate for bodies and add it back when the pool gets competitive. Some of it is exactly that. But the report estimated that only about 27% of the changing occupations were “cyclical” responses to the labor shortage. The majority, around 63%, were “structural” resets that had started before the crisis and looked durable. So this wasn’t only employers flinching under pressure. A lot of them appeared to be genuinely rethinking whether the credential was ever load-bearing in the first place.
Which brings me to the chapter that matters most, and the one I’d want anyone reading a job posting to sit with. In 2024, the Burning Glass Institute and Harvard Business School went back to check whether the reset in postings had actually changed who got hired. It hadn’t, or not by much. Looking at more than 11,000 roles at large firms before and after they dropped degree requirements, they found the share of new hires without a bachelor’s rose by an average of just 3.5 percentage points. Across the labor market, that came to fewer than 1 in 700 hires in 2023 going to someone who benefited from the change. Almost all of the real movement came from about 37% of the companies studied. The rest had rewritten the posting and kept hiring the same people. The report’s own name for the gap between the announcement and the behavior was the distance “from pronouncements to practice.”
So the requirement on a posting turns out to be a strange kind of object. It moved, across millions of listings, which means it was never a fixed law of the role. But when it moved on paper, the hiring often didn’t move with it, which means it was never the whole gate either. What you’re reading in a job posting is not a neutral description of the job. It’s closer to a rough draft of the candidate a recruiter is picturing, and sometimes not even the operative version of that.
I want to be careful here about what I am and am not claiming. I’m not saying you should ignore requirements or that the degree-or-no-degree question doesn’t matter. It does, especially for roles with licensing requirements or specific technical training. And I’m not saying that if you’re a humanities grad without a credential stack you’re automatically competitive for anything that requires, say, a CFA. That would be overclaiming in a way that doesn’t help anyone.
What I am saying is that the requirement on the posting is often a description of the candidate a recruiter imagines, not a genuine minimum for the job. And there’s a meaningful difference between those two things. One is a wall. The other is a prior.
But the same research that lets me say that should keep me honest in the other direction too. A prior can be sticky. The 2024 findings are, in a way, a measurement of exactly how sticky: even when a company formally removes the wall, the prior often stays standing behind it, in the habits of the person actually making the call. So “it’s a prior, not a wall” is not the same as “the door is open.” It means the barrier is more negotiable than it looks and more persistent than a rewritten posting would suggest, both at the same time.
The question that actually matters, then, is how you figure out which one you’re looking at, and how much of it is still operative once you’re in the room. I don’t have a clean answer to this, and I want to sit with that rather than paper over it. The honest thing to say is that it involves reading signals that are genuinely ambiguous: the age and specificity of the posting, whether the company has a structured HR function or whether the hiring manager is writing their own requisitions, whether the role is being filled from within and the posting is a formality. None of these are certain. All of them are worth reading.
What I do think is that understanding the mechanism changes how you read a job posting, and that matters before you’ve sent a single application. If credential inflation is partly a lazy default, then the question isn’t only “do I have what they’re asking for” but also “is what they’re asking for actually load-bearing.” Some of it is. Some of it is scar tissue from the last hire who didn’t work out, or cargo-culted from a similar posting, or a recruiter’s shorthand for something they haven’t articulated. You can’t always tell which. But you can stop reading every requirement as though it were a locked door.
Dismissed by Degrees found something else that showed up less in the coverage, and it cuts against the intuition that the credential is a safe bet. In the survey of six hundred employers behind the report, two-thirds agreed that requiring a bachelor’s degree made a middle-skills job harder to fill. And once those roles were filled with graduates, employers reported that the graduates turned over faster and were less engaged than the non-graduates doing the same work. The degree was screening people in, but it wasn’t selecting for the thing the job actually needed. That’s its own kind of data point. The credential wasn’t predicting success. It was predicting presence.
I’ve only been in the workforce for about eight months, so I’m not handing this down from anywhere elevated. But I do sit inside a firm where I’ve watched requisitions get written, and the process is exactly as unglamorous as the research suggests. The requirements reflect what’s easy to specify, not always what’s genuinely required. That doesn’t mean the game is rigged against you. It means the posting is a rough draft of what they want, not a final answer, and sometimes not even the last word on who they actually hire.
Sources
- Joseph B. Fuller and Manjari Raman, Dismissed by Degrees: How Degree Inflation Is Undermining U.S. Competitiveness and Hurting America’s Middle Class (Accenture, Grads of Life, and Harvard Business School, October 2017). The 67% / 16% production-supervisor figures describe 2015 job postings and are drawn from Burning Glass Technologies data. The report also reports that two-thirds of surveyed employers said a degree requirement makes middle-skills roles harder to fill, and that college graduates in those roles showed higher turnover and lower engagement than non-graduates.
- Joseph B. Fuller, Christina Langer, Julia Nitschke, Layla O’Kane, Matthew Sigelman, and Bledi Taska, The Emerging Degree Reset (The Burning Glass Institute, February 2022; direct PDF). Based on more than 51 million postings from 2017 and 2019 to early 2021. Source of the 46% (middle-skill) and 31% (high-skill) material-reset figures, and of the split between cyclical resets (about 27%) and structural resets (about 63%).
- The Burning Glass Institute and Harvard Business School, Skills-Based Hiring: The Long Road from Pronouncements to Practice (February 2024). Based on more than 11,000 roles at large firms before and after degree requirements were dropped. Source of the 3.5-percentage-point average increase in non-degree hiring, the “fewer than 1 in 700 hires in 2023” figure, and the finding that nearly all real change came from about 37% of the companies studied.
- For additional context: Preston Cooper, How Unnecessary College Degree Requirements Hurt the Working Class (FREOPP), which notes that degree inflation measured by workers’ actual education levels has not meaningfully slowed since 2017, even as postings changed. This reinforces the gap between what postings ask for and who gets hired.