The general libertarian position on education has long been that it should be privately funded, but I suspect libertarians have tended to be somewhat quiet on the issue because they suspect that some subsidization of education is good. I agree with this notion, and so I aim to provide a justification for limited public subsidization of primary education. 

The main benefit from primary education is that those educated improve their communication skills. Improvements in skills like reading, writing, basic math, and spoken language are obvious, but more subtle improvements in social skills are also important. Communication skills are a good because they lower the cost in time and effort of sharing information with other people. The benefits of those communication skills accrue to both the individual with the skills and all those who communicate with them because any time or effort saved through good communication is usually saved to all parties.

Investment into communication skills, without subsidization will be socially sub-optimal because individuals don’t get most of the benefit from those skills. Individuals have very little capacity to charge for the benefit others derive from their communication skills for several reasons. First, the benefits to each party from communication skills are closesly bound, in most cases, it is impossible for the skilled individual to deny the benefit of his skills to others without denying himself the same benefit. Second, in everyday life, the time and effort wasted trying to charge for the benefits others derive from one’s communication skills would easily be more than the time saved by having good skills. Because individuals cannot get reimbursement for the benefits to others from their investment into communication skills, society as a whole should be willing to pay individuals to make those investments.

Public subsidization of communication skill improvement internalizes the benefits of those skills so that people make decisions that are much closer to social optimal. Unfortunately the benefits derived from communication skills are very hard to measure, even though they are probably very large, so it will be very difficult to find a socially optimal level of subsidization except through trial and error.