Introduction
Every supply-and-demand comparison in higher ed runs through the CIP to SOC crosswalk, and the crosswalk has a reassuring property: it maps something to everyone. Pull the 648 detailed occupations in the SOC and the feeder programs behind each, and not one occupation comes back with zero graduates. That sounds like good news and mostly is. But the floor of that reassurance is thin. Forty-four occupations, together employing 1.65 million people, drew fewer than 100 graduates each in the 2024-25 award year, and almost all of them through a single program code. Some of that is the taxonomy working exactly as designed. Some of it is a measurement blind spot worth understanding before your next program review cites it.
Nobody is unmapped
Start with the headline finding, because it corrects a common assumption. We expected a list of occupations with literally no pipeline, the way "Other" CIP codes map to nothing. There are none. The crosswalk is many-to-many and generous: every SOC-6 occupation has at least one mapped CIP, and at least one of those CIPs has reported completions somewhere in the country. Even the smallest pipelines light up. The thinnest occupation in the entire directory, Hearing Aid Specialists, still shows four graduates a year through one program at one community college. So the honest question is not "which occupations have no pipeline" but "which ones live on a pipeline so thin that the crosswalk number is nearly meaningless."
The 44, and what they have in common
The list sorts by employment, and the top of it is striking: Painters, Construction and Maintenance, 225,190 people employed, 28,100 projected annual openings, and 29 completions a year. Cement masons: 206,170 employed, 79 completions. Butchers and meat cutters: 136,430 employed, 48 completions. Roofers: 135,490 employed, 42 completions. Flight attendants: 131,650 employed, 70 completions. Read as a labor shortage, those numbers are absurd. Read as a taxonomy, they are exactly right: 42 of the 44 occupations have a typical entry education of high school or less. Nobody goes to college to become a painter; the crosswalk's 29 graduates are the handful who did anyway, through the Painting/Painter and Wall Coverer program code (46.0408) at four institutions. The pipeline is thin because the occupations are entered through on-the-job training, apprenticeships, and employer certification, none of which IPEDS counts.
The 12 largest employers on sub-100 pipelines
| Occupation | Employed | Annual openings | Graduates 2024-25 | Entry education |
|---|---|---|---|---|
| Painters, construction and maintenance | 225,190 | 28,100 | 29 | No formal credential |
| Cement masons and concrete finishers | 206,170 | 14,300 | 79 | No formal credential |
| Butchers and meat cutters | 136,430 | 16,900 | 48 | No formal credential |
| Roofers | 135,490 | 12,700 | 42 | No formal credential |
| Flight attendants | 131,650 | 19,800 | 70 | High school |
| Slaughterers and meat packers | 69,950 | 8,400 | 48 | No formal credential |
| Structural iron and steel workers | 68,380 | 5,500 | 5 | High school |
| Glaziers | 58,480 | 5,100 | 11 | High school |
| Concierges | 49,240 | 6,800 | 52 | High school |
| Railroad conductors and yardmasters | 46,440 | 3,000 | 87 | High school |
| Service unit operators, oil and gas | 43,140 | 4,100 | 61 | No formal credential |
| Medical transcriptionists | 41,550 | 7,400 | 30 | Postsecondary nondegree |
The one-school pipelines
Several occupations in the list are fed by literally one institution in the entire country. Structural iron and steel workers: 68,380 people employed, five graduates a year, all from Moraine Park Technical College in Fond du Lac, Wisconsin, through the Metal Building Assembly code. Glaziers: 58,480 employed, eleven graduates, one school, Ivy Tech in Indianapolis. Hearing aid specialists: 11,270 employed, four graduates, Spokane Falls Community College. Watch and clock repairers run a slightly wider operation: five institutions, 78 graduates, led by Paris Junior College in Paris, Texas, which is, by most accounts, American horology. These are not failures of the labor market; they are niche credential markets the taxonomy happens to make visible. But if your program review ever cites one of them as a benchmark, know that the "national pipeline" is one department at one college, and its closure would take an entire occupation's measured pipeline to zero overnight.
The two exceptions
Two occupations in the 44 do require postsecondary credentials, and they fail thin for different reasons. Medical Transcriptionists: BLS lists a postsecondary nondegree award as typical entry education, 41,550 people are employed, and the entire feeder pipeline produced 30 graduates at 9 institutions in 2024-25. That is not a taxonomy artifact; it is a collapsing program field, automated out of the workflow it trained for, and the program inventory has correctly collapsed with it. The other exception is stranger and more instructive, and it gets its own section.
The invisible residency
Oral and maxillofacial surgeons: 4,910 employed, typical entry education a doctoral or professional degree, and a feeder pipeline of 39 graduates a year. A four-year dental surgery specialty producing 39 people a year nationally would be a crisis; it is actually a measurement artifact, and the crosswalk record shows it directly. The occupation maps to two CIP codes. One is 51.0507, Oral/Maxillofacial Surgery, the dental school degree, which carries the 39 completions. The other is 60.0101, the Oral and Maxillofacial Surgery Residency Program, which is where the actual clinical training happens, and its record is empty, because the entire 60 and 61 residency family is not valid for IPEDS reporting. We verified this from the other side in our CIP 2020 adoption analysis: zero rows in any completions file, any year. The pipeline exists, every surgeon went through it, and the federal completions data will never see it. Any occupation whose real training lives in a residency, most of medicine and dentistry, has an artificially thin measured pipeline, and no amount of crosswalk refinement can fix it while residency codes stay outside IPEDS.
What to do with this
- Before quoting a graduates-per-occupation number, check the entry education. If it is high school or less, the thin pipeline is the taxonomy telling you college is not the gate, and the number should never appear in a program viability memo as demand evidence.
- Check the feeder count. A pipeline running through one or two CIP codes at one or two institutions is a fragile benchmark; name the institutions if you cite it at all.
- For medical and dental occupations, say plainly that residency training is outside IPEDS and the measured pipeline is the degree, not the training. The 60/61 family is the reason.
- For genuinely postsecondary occupations with thin and shrinking pipelines, like medical transcription, the thinness is the finding: the market and the program inventory are agreeing with each other.
How we got this
The occupation side comes from our occupations build: 648 SOC-6 detailed occupations, each carrying OEWS May 2025 employment and wages, BLS 2024-34 projected national openings, typical entry education, and feeder completions from IPEDS 2024-25 summed across every CIP the NCES crosswalk maps to the occupation. We filtered to occupations with 1 to 99 feeder completions, then pulled the full record for each named case to identify the single feeder codes and institutions. The completions figures cross-check against the IPEDS warehouse directly; the residency-family zero was verified there too. The full extraction log and per-occupation records are in the evidence folder for this post. The one reading rule that matters: feeder completions are a supply measure through the crosswalk's theory, not a flow of graduates into the occupation.
Sources
- BLS OEWS, May 2025 release: employment, wages.
- BLS Employment Projections, 2024-34 round: annual openings, typical entry education.
- IPEDS Completions, 2024-25 award year (provisional C2025_A): feeder completions, first majors.
- NCES CIP 2020 to SOC 2018 crosswalk: the feeder mappings.
Where to go next
The crosswalk's rules and limits are covered in the CIP to SOC crosswalk, explained, and the reading rules for the occupation side are in what is a SOC code. For the opposite end of the pipeline spectrum, graduates per opening measures the occupations where supply runs ahead of demand. To browse any occupation's feeder programs directly, start at the occupations directory.
Check any occupation's pipeline in seconds
Ask for the feeder programs, employment, and openings behind any occupation, with the source and vintage on every number.
Book a demo