A facile argument we often hear is that because education has “positive externalities,” it needs to be subsidized, if not provided altogether, by the state — because the free market supposedly underproduces good that have such externalities. In this Pileus post, however, British scholar Mark Pennington draws a distinction between externalities that have policy implications and those that don’t. The externalities associated with education, he argues, don’t have policy implications. We would get the optimal degree of education without any government intervention, and therefore all of the manifold inefficiencies that accompany government intervention are unbalanced by any gains.