BLS Says These Jobs Need a Degree. We Checked Who Has One.

The US Bureau of Labor Statistics publishes two numbers about every occupation in the American economy, in the same 118-kilobyte spreadsheet.

One says what education you normally need to enter that job. The other says what education the people currently doing it actually have.

For a lot of occupations, those two numbers are not close. BLS knows this and says so in its own documentation — but its explanation runs in one direction, and the published tables never measure the other.

We pulled twelve years of that file — eight editions, 2012 through 2024 — and compared them line by line. In 58% of the occupations where the government says entry normally requires a degree, at least a quarter of the people doing the job hold less than that degree. In thirteen of them, more than half do.

The gap is closing. It is closing slowly, it is closing from the workers’ side rather than the government’s, and in a few large occupations it is getting wider.

Two tables, one file, one disagreement

The file is the Employment Projections program’s education.xlsx. Inside it:

Table 5.4 assigns every occupation a “typical education needed for entry” — no credential, high school, a certificate, an associate’s, a bachelor’s, a master’s, a doctorate. This is the line that gets quoted in careers advice, usually without attribution and usually as though it were a rule.

Table 5.3 reports what employed people aged 25 and over in that occupation actually hold, in seven buckets that sum to 100%, built from American Community Survey microdata.

Put the second against the first and you get a number the tables never state: the share of people currently doing a job who hold less than the credential the government says that job normally needs.

BLS anticipates the comparison. Its technical documentation warns that the assignment “reflects typical entry-level educational requirements, whereas ACS data report the level of education attained by workers already in the occupation”, and that this “can lead to cases in which the attainment data reflect a higher level of educational attainment than the category selection.” The emphasis is ours, and it is the whole point: the documented explanation covers workers with more education than the bar. The direction we measure here — workers with less — is attributed to cohort, in a single sentence, with no figure attached. That sentence is testable, and twelve years of editions is what tests it.

Of the 831 occupations BLS assigns, 338 carry a degree-level bar — an associate’s degree or higher. We report the headline on the 136 of those that BLS marks as an exact match to the survey’s occupation codes, and give the all-338 figure alongside. More on why in the limits section, because it matters.

The headline survives the choice of scope exactly: 58% either way. The extreme tail does not — restricting to exact matches cuts the “more than half” group from 22% of occupations to 10%. That is why the careful number is the one we lead with.

What the gap looks like in jobs you have heard of

Two things are worth noticing in that list before anything else.

The largest gaps cluster in management and coordination roles — construction managers, facilities managers, cost estimators, logisticians, general and operations managers. These are jobs people arrive at from underneath, after years of doing the work, and the credential the government names for entry is frequently not the route anyone actually took.

The smallest gaps cluster in licensed and gatekept professions. Actuaries sit at 1.1%, veterinarians at 2.5%, optometrists at 2.6%. Where a board exam stands between you and the job, the stated requirement and the held credential converge, because they are enforced by the same mechanism.

The same file says some of these jobs also want five years first

Table 5.4 carries two more columns that almost nobody quotes: the prior work experience BLS considers typical for entry, and the on-the-job training that follows.

Of the 338 occupations with a degree-level bar, 21 also expect five years or more of prior experience in a related occupation, and a further 35 expect less than five. The rest expect none. That is the official description of a job you cannot walk into: a degree, plus half a decade somewhere else first.

There is a tempting pattern here. Among exact-match occupations, the gap grows with the experience the job expects: a median of 27.1% below the bar where no prior experience is typical, 31.0% where less than five years is, 34.7% where five or more is. That would fit the story the rest of this page tells — the more a job is something you work your way into, the less the stated credential describes the people in it.

We are not going to claim it. Those groups contain 106, 17 and 13 occupations. Widen to all 338 and the gradient breaks: the five-years-or-more group drops to 27.8%, below both of the others. A pattern that reverses when you change the scope, on thirteen data points, is not a finding. It is recorded here as unestablished, with the numbers that would let someone else settle it.

What does survive is the plain description, and it is the more useful half anyway. If you are writing an application for one of those 21 occupations, the government’s own account of the job says the credential is the smaller part of what it expects.

The gap is closing, and the closing is real

This is where BLS’s own explanation gets tested. Its documentation says that “because of changing entry requirements, individuals entering an occupation may need a higher level of formal education than for those persons who are already working in it.” In other words: the gap is old workers, and it should be draining away as they retire.

If that were the whole story, the gap would close briskly. Twelve years of editions let us watch it.

Twelve years of editions let us test that. We built a panel of 111 occupations present in every edition, exact survey matches, with a degree-level bar — and held the bar at BLS’s 2025 assignment for every year, so only attainment moves.

The median share below the bar fell from 31.6% to 26.6% — five points in twelve years. It narrowed in 91 of the 111 occupations, so this is a broad movement and not a handful of cases dragging an average.

Five points in twelve years is a real trend and a slow one. At that rate the gap does not close in this working generation.

So the cohort explanation is doing real work — the gap moves, in the direction BLS implies, across four fifths of the occupations we can follow. It is just not doing all of it. Twelve years of workforce turnover bought five percentage points, while a quarter of the workforce in most of these occupations still sits below the line. Whatever is holding the rest in place is not simply people who entered long ago.

The government did not move its own bar

The other way a gap can close is that the bar comes down to meet the workers. It did not.

Comparing BLS’s entry assignments in 2014 against 2025, across the 718 occupations that appear in both: 26 changed. That is 4%. Thirteen went up, thirteen went down. A wash.

The downgrades are mostly at the bottom of the ladder — postal service clerks and mail carriers moved from “high school diploma” to “no formal educational credential”, semiconductor processing technicians from an associate’s to high school. The upgrades are mostly clinical: athletic trainers and mental health social workers both moved from a bachelor’s to a master’s.

Whatever else is happening, it is not that the government quietly decided fewer jobs need degrees. Over twelve years it barely changed its mind about anything, and where it did, it went up as often as down.

Some occupations professionalised. Others did not move at all.

The panel average hides very different individual histories.

Nursing and pharmacy are the two clearest cases of a profession pulling its own workforce up. Pharmacy’s shift is the doctorate: the PharmD became the entry degree, and the 28% still below it are largely people who qualified before that. Nursing’s is the long campaign for the bachelor’s-prepared nurse — and it is worth saying that a quarter of working registered nurses still entered through an associate’s degree, which is a fact about a real and common route, not a deficiency.

Construction management did not move. Two thirds of construction managers held less than a bachelor’s in 2012 and two thirds do now. A twelve-year national trend toward credentials passed this occupation by entirely.

And general and operations managers went the other way, from 48.4% below the bar to 53.1%. It is one of the largest occupations in the country, and over twelve years the share of people doing it without the stated credential rose by nearly five points. We record that; we cannot explain it from this data, and we are not going to guess.

What this cannot tell you

Four limits. Each one rules out a sentence that would otherwise have been easy to write.

Attainment is not a hiring requirement

Table 5.3 measures what people have. It does not measure what any employer asked for. A 53% gap for operations managers does not mean 53% of employers waive a degree requirement — it means 53% of the people in the job do not have one, which is a different and weaker statement. Nothing here supports “employers are dropping degree requirements.”

The bar is an assignment, not a measurement

BLS assigns the entry education itself. Its documentation is explicit about how: “BLS economists assigned occupations to categories based on analyses of qualitative and quantitative information”, drawing on ACS attainment data, O*NET, postsecondary completion data from the Department of Education, and “qualitative information obtained from educators, employers, workers in the occupation, training experts, and representatives of professional and trade associations and unions.”

It is the government’s considered view of how people normally get in, not a requirement collected from employers. Treating it as a rule is the error that makes this comparison attackable, so we treat it as what it is.

The two tables are not fully independent

Read that sourcing list again: ACS attainment data is one of the inputs BLS uses to set the bar. The two numbers we are comparing are not arrived at separately — the assignment is made with some knowledge of what workers actually hold.

That cuts against the finding in the obvious way, and it is the strongest objection anyone can raise to this page. It also makes the result stranger rather than weaker: even with attainment among its inputs, the assignment still lands above where a quarter of the workforce sits in 58% of these occupations, and above half the workforce in thirteen. The other inputs — O*NET, completion data, what professional bodies say the job requires — evidently pull hard enough to override what the survey shows.

BLS names this problem itself, in the occupation we lead with

The technical documentation’s worked example is registered nurses: the typical entry path “is a bachelor’s degree, with no work experience and no on-the-job training requirements. However, the OOH profile notes that registered nurses can also obtain an associate’s degree in nursing or a diploma from an approved nursing program.”

So the single largest occupation in our table is one BLS has already flagged as a case where the assignment does not capture a real and common route in. That is a point against reading any single row as a scandal — and a point for reading the distribution, which is what this study is about.

Incumbents are not entrants

Everyone aged 25 and over is in the attainment figure, including people who entered before the bar was what it is now. That is exactly why the trend matters more than the level, and why we built the constant panel. But it does not disappear: some of every gap in the table above is history rather than current practice.

Six in ten of these occupations are not exact survey matches

BLS flags an occupation when its attainment figure is shared with neighbouring occupations rather than measured separately. 202 of the 338 degree-level occupations carry that flag — 60%. It is worst exactly where the numbers are most quotable: of the occupations where more than half of workers are below the bar, 83% are flagged. The most spectacular figures in this dataset — orthotists at 90%, biochemists at 78% — are all shared-cell estimates, and none of them appear anywhere above. Every headline on this page is computed on the exact-match subset.

How to read a claim about degree requirements

Method

Source: education.xlsx, Employment Projections program, US Bureau of Labor Statistics. Table 5.3 (educational attainment of employed workers 25 and over, by detailed occupation, from American Community Survey public-use microdata) and Table 5.4 (education and training assignments by detailed occupation). Both are keyed to 6-digit Standard Occupational Classification codes, so the join between them is exact.

The file sits at one stable URL and is overwritten each year, so the time series comes from Internet Archive snapshots. Fourteen archived copies reduce to eight distinct survey periods: 2012-13, 2014-15, 2016-17, 2018, 2019, 2021-22, 2022-23 and 2023-24. The period for each was read from that file’s own source note rather than inferred from its snapshot date, because in several editions the sheet title and the source note disagree.

Sheet names changed from Attainment and Assignment to Table 5.3 and Table 5.4 in 2020; the reader handles both and reports any file it cannot map rather than guessing.

“Below the bar” is the sum of the attainment buckets strictly below the assigned entry education. Occupations whose bar is below an associate’s degree are excluded: being below a postsecondary certificate is not a shortfall worth reporting. All over-time figures use the constant panel with the bar held at the 2025 assignment.

Two mixed vintages to note: 2018 and 2019 are single-year survey files while the rest pool two years, so edition-to-edition wobble is partly a sampling artefact. Only the twelve-year shape is claimed, never a single step.

The dataset below carries all 338 degree-level occupations, their bar, their gap in every edition, and flags for exact-match and panel membership, so any figure on this page can be re-walked.

FAQ

Do you need a degree to become a manager?

BLS lists a bachelor’s degree as the typical entry education for general and operations managers, and for construction, facilities and administrative services managers. In every one of those occupations, more people currently doing the job hold less than a bachelor’s than hold one or more: 53.1% for general and operations managers, 65.3% for construction managers, 60.9% for facilities managers, 42.6% for administrative services managers. That describes the people already in the role rather than what any particular employer will ask you for.

How many registered nurses have a bachelor’s degree?

Among employed registered nurses aged 25 and over, 54.7% hold a bachelor’s, 12.0% a master’s and 2.8% a doctorate — about 69.5% at bachelor’s level or above. 24.9% hold an associate’s degree, which remains a standard route into the profession and is reflected in our registered nurse resume examples. That share has fallen substantially: 43.4% of nurses were below the bachelor’s level in 2012-13, against 30.5% in 2023-24.

Is the US really dropping degree requirements?

Not according to this data, which measures something adjacent and finds very little movement. The share of workers holding less than their occupation’s stated entry credential fell about five points over twelve years, and that came from workers gaining education rather than from requirements being lowered — the government changed its stated entry education for only 26 of 718 occupations in that period, and raised it as often as it lowered it. Claims about employers dropping requirements need job-posting or hiring data, which this is not.

Which jobs have the biggest gap between stated and actual education?

Among occupations measured exactly: construction managers (65.3% below a bachelor’s), rehabilitation counselors (63.3% below a master’s), cost estimators and healthcare social workers (61.1%), facilities managers (60.9%), agricultural inspectors (59.1%), logisticians (54.5%) and general and operations managers (53.1%). The pattern is coordination and management roles that people reach from below, after doing the underlying work.

Where does this data come from?

One public file published by the US Bureau of Labor Statistics, education.xlsx, containing both the attainment measurement and the entry-education assignment. It is 118 kilobytes, public domain, updated annually, and anyone can download it and repeat this analysis. The twelve-year series comes from Internet Archive copies of the same URL.

Sources

• US Bureau of Labor Statistics — education.xlsx, Employment Projections program. Table 5.3 (educational attainment by detailed occupation) and Table 5.4 (education and training assignments). The single source for every figure on this page.
• US Bureau of Labor Statistics — Educational attainment for workers 25 years and older by detailed occupation, the web version of Table 5.3, for the published notes and footnote definitions.
• US Bureau of Labor Statistics — Education and training assignments: technical documentation. The source of every quoted passage about how the entry assignment is made, why it can diverge from attainment, and the registered nurse example. Last modified 27 August 2026.
• US Census Bureau — American Community Survey public-use microdata, the underlying source of the attainment estimates, as stated in the file’s own source note.
• Internet Archive — snapshots of the same URL, used for the 2012-13 through 2022-23 editions. Retrieval timestamps for each edition are recorded in the dataset’s provenance notes.