Florida may start varying tuition by major:

Tuition would be lower for students pursuing degrees most needed for
Florida’s job market, including ones in science, technology, engineering
and math, collectively known as the STEM fields.

The committee is recommending no tuition increases for them in the next three years.

But to pay for that, students in fields such as psychology, political
science, anthropology, and performing arts could pay more because they
have fewer job prospects in the state.

Alex Tabarrok agrees with the conclusion, but not the official rationale:

The task force has the right idea but the right way to target subsidies
is not to the job market per se (let alone Florida’s job market), wages
already reflect job market needs. Subsidies instead should be targeted
to fields where education has the greatest positive spillovers, benefits
that spill over wages and flow to the public at large. Overall, this
likely means subsidizing the STEM fields more than anthropology which is
why the taskforce has the right idea. If the task force wants to
explain the idea, however, they should make it clear that the goal is to
focus subsidies on those fields where education most benefits the
taxpayer.

On reflection, though, Alex’s rationale tips the scales in favor of all high-earning majors.  New ideas generated by STEM majors are one kind of positive externality.  But the higher taxes paid by lucrative majors also qualify.  There’s a fiscal externality.  The more lucrative your major, the more likely your future taxes are to repay taxpayers for their subsidies.  But the higher the tax rate, the less likely you are to want to pursue a lucrative major.  Cutting tuition for lucrative majors relative to non-lucrative majors is a simple way to correct (or at least mitigate) this externality.

The same goes for employment rates.  Graduates who get jobs pay taxes.  Graduates who don’t get jobs consume taxes.  Once again, there’s a fiscal externality.  Taxpayers benefit if students focus on majors with high employment rates, and avoid majors with low employment rates.  And by almost all accounts, high-income majors are also high-employment majors.

But wait, there’s more.  Alex neglects another important efficiency consideration: signaling.  STEM majors spend a relatively high fraction of their time acquiring real world skills.  Other majors spend a higher fraction simply showing off their intelligence, work ethic, and conformity.  Both are privately rewarding, but the former is far more socially rewarding: Useful skills enrich the world, but signaling mostly just enriches the signaler.

Even if you don’t buy the signaling model, you should also consider the fact that STEM majors have relatively absolute standards.  When more students acquire STEM degrees, more people actually understand STEM.  Other fields, in contrast, have heavily diluted their standards to make room for marginal (and submarginal) students.  As a result, taxpayers are more likely to get their money’s worth for the former than the latter.

None of this means that I favor government support for STEM or high-earning majors.  The best education policy, as I’ve often argued, is a genuine free market.  But that doesn’t mean that all education subsidies are equally bad.  If any students are going to get taxpayer money, vocational studies should outrank mere navel-gazing.