In the academic year ended in 2020, our governments at all levels spent more than $1.2 trillion on primary, secondary, and higher education. Most of this paid for government-owned and -operated schools, which subsidize education by charging zero or below-market prices. Calls for even greater subsidy are common.

In Libertarian Land, governments would subsidize far less, and they would never operate public schools.

The standard argument for subsidizing is that one person’s education raises everyone’s productivity (spillovers). An economy might be more efficient if most people share basic math and language skills or common knowledge of relevant laws, business practices, and scientific ideas. Under these conditions, the privately chosen amount of education might be less than what is desirable overall; education might be a public good.

This reasoning is plausible but less than compelling. Measuring education “spillovers” is difficult, and views are often “in the eye of the beholder.” Do spillovers arise from studying Shakespeare, or computer science, or auto mechanics? The particular subsidies adopted might easily reflect lobbying from vested interests rather than logic or data. 

Existing evidence, moreover, suggests only modest spillovers. This does not deny that education improves income or happiness for those acquiring it, merely that these individuals capture the main benefits.

A different argument for subsidies is that some people underinvest in education because they are too short-sighted to recognize the future benefits. Paternalism might then suggest subsidizing to lower the price of education, encouraging more. 

Paternalistic justifications are problematic, however; they open a Pandora’s box over who gets to define the paternalistic view. Should schools ban “evil” books to protect the innocent? Who decides which books are evil?

Paternalism can harm those it aims to help. Subsidies might encourage excess education by individuals who forget that acquiring education implies foregoing income, or who overestimate the income-enhancing benefits.

The degree of “bad decision making” is easily overstated. Some parents no doubt make bad education decisions, but parents make many choices that determine their children’s well-being: TV, homework, sports, friends. Thus improving parental decisions about education might have little impact. At most, paternalism implies subsidizing early years of education, not PhDs in computer science or classics.

Still a different justification for subsidizing education is that some cannot afford it, even when the higher income might justify the costs. Further, many individuals cannot borrow against their future income, since some lenders avoid loans where the only collateral is a borrower’s potential earnings.   

This argument is plausibly the best of the three, but it is also easily overstated. Private education lending exists, and many can diminish the impact of credit constraints via apprenticeships or on-the-job training. Credit constraints are quantitatively important mainly for low-income households.

Thus the arguments for subsidizing are plausible but not decisive; they mainly apply to young children in low-income households. And subsidies generate serious costs, beyond the direct expenditure.

If government subsidizes education, it must define education. This promotes standardization and discourages innovation. The right kind of education for some might be college prep but for others vocational or technical training. The most efficient format for some students might be hybrid, or year-long schooling for four days a week, rather than the traditional calendar. Government control does not foster such experimentation and variety.

When subsidy comes from government-owned and -operated schools, government must address controversial issues like speech codes, affirmative action, single-sex education, controversial faculty, sex education, and more. Imposing one approach is polarizing, and in the extreme amounts to thought control. Totalitarian states routinely monopolize their education systems.

In Libertarian Land, the federal government plays no role in education, leaving any such intervention to states. This reduces the risk of standardization and polarization that arises when a national government imposes one rule or system.

The only subsidy for education is vouchers for early years of education, such as K-6. These pay for education at private schools, since government schools do not exist. This reduces controversy, especially since the subsidy (voucher) goes directly to parents.

Will this education policy ever happen? On verra. Even recognizing the libertarian perspective, though, can focus education subsidies toward approaches with relatively large benefits and relatively small costs.