Field of study decides which community college certificates clear the federal earnings bar

Five years after starting at a Texas public two-year college, students in construction trades certificate programs had pulled roughly $48,000 ahead of comparable peers who never enrolled. Students in information technology certificate programs had fallen about $21,000 behind. It is the same type of credential in the same state system.

Those figures come from a Texas study of more than 900,000 students, which Bruno V. Manno cites in an October 19 Community College Daily analysis of the new federal earnings test. A U.S. Education Department rule that took effect in July now puts one blunt question to programs nationwide: are their graduates out-earning workers whose education ended with a high school diploma? At community colleges, the answer mostly comes down to what a student studied.

How the earnings bar gets built

The Education Department's January 2026 explainer, built on preliminary analysis from its Office of the Chief Economist, lays out the mechanics. On the program side, the measure is the median earnings of federal-aid graduates who are working and not enrolled in college, taken four years after they finish. Those numbers come from a federal agency that holds individual income data, such as the IRS. On the other side sits a Census Bureau benchmark: median earnings of working adults aged 25 to 34 whose highest credential is a high school diploma.

In the department's first worked example, Sacramento City College draws 99% of the students in its associate degree in business administration from California. The program's median earnings of $53,872 exceed the in-state high school benchmark of $36,082, so the program passes. If a program misses the bar in two out of three years in a row, it is cut off from federal student loans.

The bar can bite, too. The explainer's second example is a master's program in music at The Juilliard School, which enrolls 77% of its students from out of state. Graduate programs are measured against bachelor's-level benchmarks rather than the high school wage, and that program's $46,372 median fell short of the $48,304 national same-field benchmark, so it fails. The test written into Section 84001 of the One Big Beautiful Bill Act covers associate, bachelor's and graduate programs. The department's flowchart lists non-degree undergraduate programs as not covered by that section, and the same presentation covers changes to the separate gainful employment rules to bring them into line with the new test.

According to Third Way, a bachelor's degree program passes only if its graduates out-earn workers whose highest credential is a high school diploma. The group places that level at just over $40,000 a year. The department builds its benchmark from working adults aged 25 to 34, and the example figures it gives come in below that. Where the line falls exactly varies by state and by data year.

The basic premise is popular. Third Way reports that more than 90% of voters support using high school earnings as a baseline, including 97% of Democrats, and the Institute for Higher Education Policy (IHEP) calls the rule an "important accountability safeguard."

Degrees pass, certificates split

For degrees, the numbers leave little to argue about. Third Way cites a PEER Center analysis by Jordan Matsudaira and Kristen Allen that projects more than 99% of bachelor's programs and 98% of associate and master's programs will pass. Over a longer horizon, the Texas study points the same way. Bachelor's students were roughly $87,000 ahead at year 15, a figure that already nets out tuition and the wages they passed up while enrolled.

Associate degree students in Texas took longer to get there. They broke even by their seventh year after enrolling and were about $25,000 ahead by their tenth. Those clocks start at enrollment, not graduation, so they don't line up neatly with the test's four-year post-completion window. They do suggest the associate payoff keeps building well after the point the federal snapshot captures.

Certificates tell a very different story. IHEP finds that just 55% of certificate holders attended programs that pass. Nearly every degree program clears the bar, so certificates make up about nine in 10 of the graduates in programs that miss it. Third Way puts the share of all college programs expected to fail at just 1.5%, and it says most of them sit at for-profit colleges.

Public colleges turn out about half of all graduates in the federal data, yet they account for only 6% of graduates from programs that miss the bar, according to IHEP. That makes the sector's overall exposure small but not zero. For a community college, the sorting takes place inside its certificate catalog.

The strongest early worry about the test was that it would push students away from lower-paid public service fields. The evidence so far doesn't back that up. IHEP finds nearly all bachelor's graduates in education and social work attended passing programs, and Third Way reports that every master's program in education and 99% of master's programs in counseling are expected to pass. For a community college offering transfer pathways into those fields, the degree side of the house looks secure.

Liberal arts transfer students land in similar territory. Third Way projects that all history programs, 99% of English programs and 92% of fine and studio art programs will pass. Art is the softest of those numbers, and it is still far above the 55% certificate pass rate.

Field first, then campus

Inside the certificate catalog, field of study explains most of the variation. In the Texas data, technical trades, health and protective services certificates did well alongside construction. IHEP finds that nearly every certificate holder in architecture and engineering attended a passing program. Information technology is the cautionary contrast: the Texas figure puts IT certificate students behind peers who didn't enroll at all.

The college still matters. Clive Belfield of Columbia's Community College Research Center made an early run at this question using federal College Scorecard earnings, drawn from tax records, for 896 community colleges. Measured 10 years after first enrollment, mean college-level earnings were $32,700 and the median was $29,100. Mean earnings for similarly aged high school graduates were $26,500. The spread between campuses was wide: attending a top-10% community college was associated with average earnings more than $12,000 a year higher than attending a bottom-10% one.

That data has its own scope. It counts every student who started at a college and received financial aid, whether or not they finished, which leaves out about 40% of students. Belfield also noted the existing evidence that student and program characteristics matter much more than the college itself, which fits the field-first pattern in the Texas and IHEP numbers.

That campus effect comes with a real counterpoint. When Belfield adjusted for where colleges sit (local demographics, the share of college-educated residents, the unemployment rate), the spread across colleges got much smaller. He describes the adjustment as illustrative, but the direction is clear. Some of what looks like program quality is the local job market.

The in-state benchmark handles part of this by measuring each program against its own state's high school wage. A state is still a much bigger area than one college's service district, though, so a program in a lower-wage region of a high-wage state could carry a burden the formula doesn't see.

The pandemic cohort goes first

According to IHEP, the first results are expected by July 2027. They will cover students who completed their programs in 2020 and 2021, when the pandemic was at its peak. So the opening verdicts will measure graduates who entered one of the strangest labor markets in recent memory. Programs that sent graduates into fields hit hard early on could post a weaker first reading than their usual track record.

The two-of-three-years rule matters here. A program loses federal loan access only after it falls short in two of three consecutive years, so a single weak year of data does not end loan access by itself. That gives a program time to show whether a poor first reading was a one-off.

A failing result does not shut a program down either. Third Way notes that a program that misses the mark loses access to federal loans, though students can still cover the cost through other means. Colleges that bring pricing more in line with what graduates go on to earn can also stay eligible for loans, the group adds.

When a program does fail, which is rare, most students can switch to a passing one in the same field and at the same degree level without leaving their current school. For those who have to search beyond it, the vast majority (83%) have a passing option nearby.

The stakes grow because this is where the students are going. Manno reports that about 750,000 certificate students were enrolled at community colleges last fall, a 28% increase over four years, while the colleges' overall enrollment rose 3.1% this year. Enrollment in undergraduate certificate programs grew 10% last spring, the most of any undergraduate credential. The department's new Workforce Pell program adds to that momentum.

Put that growth next to IHEP's 55% certificate pass rate and the practical takeaway for a workforce dean gets sharper. A college's exposure to the test increasingly runs through the field mix of its fastest-growing programs, which looks a lot more like construction versus IT than degrees versus certificates.