Dropping the degree requirement from a job posting became one of the easiest wins in corporate hiring. It costs nothing to announce, it reads well in a press release, and it lets a company claim it's opening doors to candidates who never had, or never wanted, a four-year degree. Hundreds of large employers made this change over the past several years, often citing a tight labor market or a genuine belief that a diploma was screening out capable people for no good reason. The harder question is whether anyone actually got hired because of it.
A study from Harvard Business School and the Burning Glass Institute looked at what happened at large firms that made this change, and the answer is, mostly, no. At some of these companies, fewer than 1 in 700 new hires were actual non-college graduates, despite a formal policy that no longer required a degree at all. The posting changed. The org chart didn't.
That gap matters because it's rarely audited. A company can announce a policy change, get the press coverage, and never circle back a year or two later to check whether hiring outcomes actually shifted. The Harvard and Burning Glass research did that follow-up work, and the honest answer it found is that removing a line from a job description is not, by itself, a hiring strategy. It's a starting point that most companies never finished.
The policy was never the bottleneck
Roughly 18 percent of U.S. job postings still explicitly list a degree requirement today, which means the other 82 percent, on paper, don't. If the posting were really what stood between qualified non-graduates and a role, that gap should have produced a visible shift in who actually gets hired. It mostly hasn't.
The reason isn't hard to find, and it isn't cynicism about corporate intentions. A degree requirement was never really the gatekeeping mechanism, it was a proxy: a rough signal that let a recruiter sort a stack of resumes without individually evaluating whether each candidate's actual experience held up. Remove the proxy and nothing replaces it. The recruiter is still under the same time pressure, still working through the same volume, still needing a fast way to decide who's worth a closer look. Absent a degree to filter on, the instinct reverts to the next closest thing: a familiar company name on the resume, a familiar school even when it's listed as some college, a referral, anything that feels like it reduces risk. The degree wasn't the wall. It was one brick in a wall built out of proxies, and knocking out a single brick doesn't bring the wall down. Boards and executive teams that want a different outcome from the one this study found need to reckon with that: a policy memo changes what a job posting says, not what a recruiter reaches for when the applicant pile is due in an hour.
What would actually move the number
If the goal is genuinely to hire more people on the strength of what they can do rather than where they went to school, the fix isn't a second announcement restating the first one. It's giving the people doing the screening something they can trust in the time they actually have, which for most recruiters isn't much. A GitHub history that can be checked against the account that produced it. A certification that can be confirmed with the body that issued it. A reference from someone who genuinely worked alongside the candidate, answering specific, comparative questions instead of offering a generic character reference. That's roughly the bet a platform like Dealt is built on: not another self-reported claim to take on faith, but a signal a screener can verify quickly enough to actually use under time pressure.
Skills-based hiring, as a stated policy, was never wrong. The theory holds up fine on paper. What's missing is the infrastructure that would let a policy survive contact with an actual hiring funnel, where speed and risk-aversion beat good intentions almost every time. Until that infrastructure exists at scale, expect more of the same: another round of announcements about dropped requirements, and another study a few years out finding the hiring data barely moved.