Underemployment is the share of recent graduates working in jobs that do not typically require a bachelor's degree. It is the failure the unemployment rate cannot see, and it is wildly uneven by field. This tracks all 73 fields from 2018 to 2024, so you can see which ones funnel graduates into sub-degree work and how stable that pattern is. Companion to the snapshot tool, Recent Graduate Outcomes by Field of Study.
Underemployment is when someone with a bachelor's degree works in a job that does not require one: a biology graduate managing a retail floor, a communications graduate waiting tables. They are employed, so the unemployment rate records them as a success. The degree, though, is not being used, and that is precisely what the unemployment rate cannot see.
This matters because for most students a degree is an investment meant to unlock degree-level work and the pay that comes with it. Underemployment is the gap between that promise and the result, and the gap is expensive. By one estimate, a graduate in a degree-level job earns about 88% more than someone with only a high-school diploma, while an underemployed graduate earns just 25% more (Talent Disrupted, Burning Glass Institute and Strada).
It is neither small nor temporary. The New York Fed puts recent-graduate underemployment around 41%. Talent Disrupted, which tracks individual graduates over a decade, finds 52% underemployed a year after graduation and shows the condition sticks: 73% of those who start out underemployed are still underemployed ten years later. Their year-one figure runs higher than the Fed's because it captures the peak right after graduation, while the Fed's 41% averages over everyone aged 22 to 27, some of whom have had a few years to climb into degree-level work. Both agree the problem is large and persistent, and that the first job matters enormously.
What has been missing is a clear, current view of how underemployment differs by field of study, and whether those differences move. That is what this tool adds: it tracks all 73 bachelor's fields from 2018 to 2024, so you can see which majors funnel graduates into sub-degree work, which protect against it, and how stable the pattern is over time.
The figures come from the Census Bureau's American Community Survey, the same microdata the New York Fed draws on. We count graduates aged 22 to 27 who hold a bachelor's, and treat a job as below the degree when most workers in that occupation do not hold one. The cutoff is calibrated so our 2023 figure matches the New York Fed's. The full method is at the bottom of the page.
Two things stand out. First, the ranking is remarkably stable. The fields with high underemployment in 2018 are the same ones in 2024, and the fields that convert a degree into degree-level work stay low throughout. Criminal justice and medical technicians sit near 70% every year; nursing (about 11%), special education, and the engineering disciplines stay low year after year. A graduate choosing a major is not facing a roll of the dice that resets each year; they are stepping onto a track that has looked the same for seven years.
Second, the overall rate fell, then turned back up. It eased from about 45% in 2018 to 40% in 2023 as the labor market tightened, then rose to about 42% in 2024. Both moves are statistically real, not sampling noise. When we test each field for a genuine trend, only 14 of 73 clear the bar, and nearly all of those point down: the post-pandemic improvement was broad. The 2024 uptick is also broad rather than concentrated, showing up across large fields like biology, psychology, performing arts, and even normally stable nursing.
One caution worth stating plainly: most eye-catching one-year jumps in a single field are within the margin of sampling error, so we do not read them as signal. Computer science is a useful example. Its underemployment has been essentially flat near 23% the whole time. Computer science graduates struggle to get hired at all, which is an unemployment problem, not an underemployment one.
Recent-graduate underemployment overall, with four illustrative fields to show the spread and how little the ranking moves. The 2019-to-2021 step spans two years because the Census released no 2020 file.
Source: Opportunity Data analysis of Census ACS 1-year PUMS, 2018-2024.
Underemployment rate by field and year, with the 2018-2024 average and the fitted trend. The trend column flags only fields whose movement is statistically distinguishable from sampling noise. Click a column to sort; type to find a field.
Trend is a weighted least-squares slope in percentage points per year; "rising" or "falling" appears only where the change is significant at the 95% level (binomial standard errors inflated by a 1.5 design-effect factor). Fields with small annual samples are flagged and carry more noise. Download the data (CSV).
Built from Census Bureau ACS 1-year microdata (2018-2024), counting graduates aged 22 to 27 with a bachelor's and classifying a job as below the degree when most workers in that occupation do not hold one, calibrated to the New York Fed. Read the full methodology for sources, the occupation cutoff, the field crosswalk, validation, and limitations. Download the data (CSV).