70% of Employers Say They Do Skills-Based Hiring. Fewer Than 1 in 700 Hires Show It.

Two numbers came out of the same hiring season this year, and they do not agree with each other. Ask employers whether they practice skills-based hiring, and 70% say yes — up from 65% a year earlier, according to the National Association of Colleges and Employers' Job Outlook 2026 survey. Ask what actually happened when companies dropped a degree requirement, and a Harvard Business School / Burning Glass Institute study of more than 11,000 U.S. job postings between 2014 and 2023 found that fewer than 1 in 700 hires — about 97,000 workers out of roughly 77 million yearly hires — can be attributed to it.

Both numbers are real. Neither is wrong. The gap between them is the story, and it is bigger than a survey-methodology footnote — it is a preview of what happens to any hiring policy that changes the words on a job posting without changing the mechanism behind it.

Two studies, two very different questions

The NACE survey asks employers what they believe they do. Collected from NACE member institutions between August and September 2025, it reports that 70% of responding employers now use skill-based hiring in some form, that this approach shows up most heavily at the interview stage (87%) and the screening stage (65%), and that GPA screening — a proxy that skills-based hiring is supposed to be replacing — has fallen from 73% of employers in 2019 to 42% today. Read at face value, that is a five-year trend line moving in exactly the direction the skills-based hiring movement promised.

The Harvard Business School / Burning Glass Institute study asks a harder question: when a company actually removes a degree requirement from a posting, does its hiring change? The answer, examined across a decade of postings, is mostly no. Non-degree hiring rose only 3.5 percentage points across the full sample. Of the firms that formally dropped degree requirements, the researchers found 45% did so "in name only" — the posting language changed, the applicant pool and the people actually hired did not. A smaller group, about 37% of the firms making changes, were genuine "skills-based hiring leaders" whose share of non-degree hires rose by roughly 20% — proof that the mechanism works when a company actually operates it, and proof by contrast that most companies are not.

That is not two studies contradicting each other. It is a survey of stated policy next to an audit of outcomes, and the distance between them is the size of the gap between a hiring team's intentions and its screening mechanism.

Where the gap actually lives: the first filter, not the interview

NACE's own breakdown points at where the disconnect is likely to originate. Skills-based hiring shows up in 87% of employers at the interview stage but only 65% at the screening stage — meaning a meaningful slice of employers are asking behavioral or technical questions in the room while still filtering who gets into the room by the same credential-shaped signals as before. A job posting that says "skills, not degrees required" but is still parsed and ranked by a resume screen looking for a specific degree keyword, a specific list of past employers, or a GPA line that officially "isn't used" but still correlates with everything the model was trained on, will produce exactly the Harvard Business School / Burning Glass Institute result: language changed, funnel unchanged, outcome unchanged.

This is a mechanism problem, not a sincerity problem. A recruiting team can be entirely honest in its stated intent to hire on skills and still run a pipeline whose first filter has no skills signal in it at all — because the tooling never asked the candidate to demonstrate one. If the only artifact in the pipeline is a resume, "skills-based" screening is, structurally, still resume-based screening with different search terms.

What a pipeline needs to actually be skills-based

The distinction the research draws — between the 45% that changed language and the 37% that changed outcomes — maps directly onto a testable design question: does a candidate get evaluated on anything other than what they wrote about themselves, before a human ever sees the file?

At NiceHire, every candidate who reaches a structured pipeline is scored against the same three fixed dimensions — technical, communication, and cultural fit, each out of 100 — against a threshold the hiring team configures for that role. That is a mechanism claim, not an outcome claim: it says what gets measured and against what, not that the measurement is unbiased or that it predicts job performance. What it does establish is a second, independent signal in the funnel — one that did not exist in the "language changed, mechanism didn't" firms the Harvard Business School / Burning Glass Institute study found, and one a team can point to when a director of talent acquisition asks the question those researchers were effectively asking: not "does our job posting say skills-based," but "does anything in our funnel actually evaluate a skill before a resume-shaped judgment does."

We would not put it more strongly than that. A structured scoring stage is a necessary piece of an actually-skills-based pipeline; it is not sufficient on its own, and a scoring stage bolted onto a funnel that still rejects on a degree filter upstream inherits the same 45%-in-name-only problem the research describes. The fix the data points to is upstream of any vendor: know where in your funnel a candidate is filtered before any skill is assessed, and treat that point — not the interview questions, not the job-posting language — as the one that decides whether "skills-based" describes your hiring or just your careers page.

The practical read for a hiring team this quarter

1. Map your funnel's first filter. Before asking whether interviews test skills, ask what removes a candidate before an interview is ever scheduled. If that filter is a keyword match on a resume, the interview-stage numbers in the NACE survey do not apply to the candidates who never got there.
2. Treat the GPA/degree-drop distinction as a warning about any surface-level policy change. The Harvard Business School / Burning Glass Institute finding — that dropping a formal requirement changed outcomes for only a minority of firms — generalizes beyond degrees. A policy change that only edits a job description, without touching the scoring or filtering logic behind it, is the "in name only" pattern regardless of which requirement it targets.
3. Ask what "skills-based" means as a specific claim about your own pipeline, not a category you belong to. The 70% adoption figure in the NACE survey is an average across employers doing very different things under the same label. The number that matters is not whether your company is inside that 70% — it's whether your funnel has an independent skills signal before the point where most candidates are already gone.

Skills-based hiring is not a failed idea; the Harvard Business School / Burning Glass Institute researchers found real evidence it works exactly as intended for the minority of firms that actually operate it that way. The lesson of the 1-in-700 finding is narrower and more useful than "skills-based hiring doesn't work": a policy statement is not a pipeline change, and the only way to know whether yours is one is to look at the filter, not the wording.

---

Sources: NACE, "Skills-Based Hiring Grows, but College Students Don't Fully Understand It" (Job Outlook 2026 survey, data collected August–September 2025); Harvard Business School Institute for Business in Global Society, "Harvard Research: Fewer than 1 in 700 get hired without a college degree", summarizing the Harvard Business School / Burning Glass Institute report "Skills-Based Hiring: The Long Road from Pronouncements to Practice"; corroborating figures also reported by HR Dive and Higher Ed Dive. All accessed via search-result content on 9 August 2026.