Introduction
Every comparison of program supply against labor market demand crosses the same seam: your completions are counted by CIP code, and the demand side is counted by SOC code. The only federal bridge between the two is the CIP to SOC crosswalk, a joint NCES and BLS product. It is also one of the most misused files in higher ed, because it looks like a lookup table and is actually a statement of expert opinion. This post covers what the crosswalk says, what it deliberately does not say, and the four mapping rules most likely to make your analysis wrong.
What the crosswalk is
The current edition maps 6-digit codes from the 2020 CIP to 6-digit codes from the 2018 SOC. It is the second edition of the crosswalk, jointly built by NCES and BLS statisticians, and it is distributed as a spreadsheet from the NCES CIP user site. The construction method matters: analysts at both agencies read the CIP program descriptions and the SOC occupation definitions and matched them where the program provides knowledge and skills directly applicable to the occupation. NCES states plainly that the matches are based on the content of the descriptions and the agencies' expertise, not on empirical data, and that NCES is still exploring data sources that could validate the matches. It is a qualitative file. That is not a flaw; it is just a fact that changes how the output should be worded. To see the mappings for a program without opening the spreadsheet, look it up in our CIP to SOC crosswalk tool, which rolls the federal 6-digit file up to 4-digit CIP series and lists every SOC occupation each one maps to.
What a mapping means
A CIP-to-SOC mapping indicates that programs in the CIP category provide knowledge and skills directly applicable to performing jobs in the SOC occupation, and that the programs typically satisfy the educational requirements for entry and advancement, or prepare students for licensure or certification. Registered Nursing (CIP 51.3801) maps to Registered Nurses (SOC 29-1141) because the program is the licensure pathway. That is the paradigm case: the program is the occupational pipeline, and everyone involved knows it.
What it does not mean
- It does not mean completers are qualified for every job in the occupation. The mappings cover the educational component of the work, not work experience or other prerequisites many jobs demand.
- It does not mean the program is required to enter the occupation. The crosswalk guidelines' own example: Stonemasons (SOC 47-2022), an occupation generally requiring no postsecondary education, is mapped to Mason/Masonry (CIP 46.0101) because the program content is directly relevant, not because it is a gate.
- It is not a record of where graduates actually go. It says nothing about any individual's outcomes. The federal file that follows actual graduates into the labor market is PSEO, which is built from tax records, not from crosswalk logic.
- It does not mean a one-to-one relationship. One CIP code can map to many SOC codes, several CIP codes can map to one SOC code, and a single occupation's pipeline can span the entire taxonomy; business degrees feed dozens of occupations.
The rules that trip people
"Other" CIP codes generally get no match
Residual CIP codes (format XX.9999) are by design wide-ranging, and NCES states it cannot be assumed they directly prepare students for specific occupations. Where there is no obvious match, the SOC assignment is "no match," coded 99.9999. If your program reports under an "Other" code, the crosswalk has nothing to say about its demand side.
SOC "all other" occupations map to nothing useful
On the occupation side, the residual detailed occupations are equally uninformative in reverse: a program mapped to an "All Other" SOC code tells you the taxonomy lost precision, not that a specific labor market exists.
First-line supervisors inherit their workers' programs
For groups like construction trades, supervisory occupations are generally assigned the CIP programs related to the workers they supervise. A supply count that includes supervisor SOC codes is partially double-counting the same pipeline at two rungs.
Teachers are mapped twice, on purpose
Teacher preparation programs (CIP series 13) map to teaching occupations at matching levels, and separately, programs that teach the subject matter map to postsecondary teaching occupations. The guidelines' example: Plant Physiology (CIP 26.0307) maps to Biological Sciences Teachers, Postsecondary (SOC 25-1042). Any education-related supply analysis has to decide which of the two mappings it means to count.
The credential-level caveat
The crosswalk maps programs to occupations; it does not encode the credential level each occupation requires. The guidelines warn about this directly: it may not be appropriate to treat associate degree completers as the supply for occupations that require a bachelor's or higher. The clean illustration is nursing. Registered nursing completions include associate degrees, bachelor's degrees, and RN-to-BSN completions, and the RN-to-BSN population is already licensed nurses being counted a second time. In our graduates-per-opening analysis, rebuilding the nursing numerator at the associate-only level changed the state rankings materially. Before quoting any graduates-per-opening number, decide which award level matches the occupational licence, and rebuild the numerator there.
Using it in a program viability memo
The workflow that survives scrutiny: count completions by CIP at the credential level the occupation actually requires, pull the crosswalk mappings for those CIP codes, then read demand, employment, wages, and projected openings, from the BLS side by SOC code. State the crosswalk as what it is: a theoretical bridge maintained by two federal agencies, not an outcome measurement. Label the credential level of the numerator. And where the mapping set is ambiguous, many-to-one or "Other" codes or supervisor rungs, say so in the memo rather than averaging it away. For a fully worked version of this pipeline, including the checks we ran before trusting the ratios, see graduates per opening: CIP completions versus SOC demand.
Where to go next
The two taxonomies the crosswalk connects are covered separately: what is a CIP code for the program side and what is a SOC code for the occupation side. On the outcomes side, why your program has no earnings number covers what PPD actually publishes at the field-of-study level, and the federal data release calendar tracks when each of these files drops.
Run the whole pipeline in one conversation
Completions by CIP, demand by SOC, and the crosswalk between them, with sources and vintages on every step.
Book a demo