Product1 publisher3 min readPublished
Underemployment data cuts against the story that the CS degree stopped paying off
A Fast Company piece concedes recent computer science graduates carry higher unemployment than other new grads, then puts their underemployment substantially lower, which is the measure of whether a degree gets used at work.
The Product Desk · Product desk
What happened
- A Fast Company piece describes a narrative that has taken hold over the past year that the computer science path is collapsing, citing coders working at Chipotle and the bursting of a CS bubble.
- It also says those graduates continue to experience substantially lower underemployment than most of their peers, meaning more of them hold professional, college-level roles that use the degree.
- Research from the National Association of Colleges and Employers puts computer science first for highest projected new-graduate salaries and third in overall employer demand.
Compiled by The Product DeskSomething wrong?How this is made
Why it matters
- decision A manager weighing a junior requisition is choosing between two indicators that point opposite ways. The article supplies neither one's percentages, so the choice cannot be checked against its numbers.
- constraint If employers want proof of applied work instead of a transcript, the limit on junior hiring sits in the screening step, and a resume pile of coursework cannot clear it.
- contradiction Employers telling a survey that entry-level STEM graduates are the hardest group to attract does not sit comfortably with the claim that the market has stopped wanting them.
- exposure Programs that graduate transcripts and nothing demonstrable are the ones exposed here. CUNY and Howard moved on AI coursework ahead of any recovery in demand.
The image the collapse story travels on is a computer science graduate working at Chipotle [1]. That graduate has a job that does not use the degree. The statistic for that is underemployment, and on underemployment the Fast Company article says recent CS graduates still do substantially better than most of their peers [3] [14]. The unemployment gap it concedes runs a couple of points above the average for other recent graduates [2].
The article does not include the percentages behind either comparison, or any count of junior roles employers have cut [13]. It is evidence about graduate outcomes and about what employers said in surveys, not about how teams are actually staffing.
Two surveys sit under the present-tense case. Strada, described in the piece as an IT services and consulting firm, asked employers nationally which entry-level graduates were hardest to attract. Math-intensive STEM graduates, CS among them, came out top, as they also did for critical thinking across all fields [5]. Research from the National Association of Colleges and Employers puts computer science first for projected new-graduate salaries and third for overall employer demand [6].
The forward-looking half is a different kind of claim. Fast Company invokes the Jevons paradox, where a capability getting cheaper expands demand for it [8]. It also cites "a growing number of technology leaders" calling for forward-deployed engineers who sit inside an organisation, apply AI tools to a specific business problem, and rebuild the workflow around the results [7]. Demand for people who can do that, the piece says, "is only going to grow, and the talent pipeline to fill that demand does not yet exist" [9].
The part that lands on a hiring loop is the gap the article names between what schools graduate and what employers can evaluate. Businesses increasingly want evidence that a student can apply skills in a real environment, not a transcript, and the piece pushes hands-on and experiential programs such as internships and apprenticeships [10]. CUNY recently fast-tracked new AI fluency requirements, and Howard University is expanding AI coursework across its campus this fall [11].
For a team holding an open junior requisition, that splits into two questions worth answering separately: can you source the candidates, and can you tell which of them can apply the skill. The Strada respondents answered the first, and entry-level math-intensive STEM graduates were the hardest group to attract [5]. On the second, the article puts the real disconnect in evaluation [10]. A team that never recorded how long its last few junior hires took to ship something a senior would have signed off has nothing measured to decide on, so it is deciding on the Chipotle anecdote.
What to watch
- Whether the next National Association of Colleges and Employers survey keeps computer science first for projected new-graduate pay and third for demand.
- Whether CUNY's AI fluency requirements and Howard's expanded AI coursework produce work samples employers can evaluate, or only new transcript lines.
- Whether anyone publishes the underlying underemployment percentages for recent CS graduates against other majors.