Such was the question discussed at the Liberty Fund conference I attended over the weekend. Our discussions ran the gamut of all of the theoretical and practical issues associated with the question, from the extent to which basic science is a public good to the costs of governments trying to pick technology winners and the politicization of science.
One thing the discussion highlighted for me was how the Samuelsonian/neoclassical argument about public good provision has permeated even the thinking of the non-economist. It has become common to argue, following Samuelson, that if something has public good characteristics it will be underprovided in equilibrium, and therefore necessarily cannot be provided optimally through market processes.
That logic is false. It ignores this logic, which is a fundamental, basic point in economics: if someone can capture enough benefit at the margin to make it worth it to them to incur the marginal cost, they will provide the good. If marginal benefit is greater than marginal cost, you do it. If others derive benefit and free ride on you, so what?
Of course, that is a static argument, so let’s introduce some dynamics and see what kind of strategic behavior we might see. Only here I want to keep my framework in which one agent’s marginal benefit exceeds his/her marginal cost. Suppose that is the case, and suppose further that government has bought into the false Samuelsonian presumption of inefficient free riding in equilibrium. In that situation, does the agent with MB>MC have an incentive to provide the good? No, because if someone else will pay for it, then I can enjoy the MB without having to incur all of the MC. So conditional on the existence of government funding, the agent who would otherwise pay for the good becomes a free rider.
My point is that the free rider problem is an empirical one, and it’s a dynamic and strategic question. We have all sorts of evidence, empirical and experimental, that inefficient free riding does not always occur in public good situations, but it can occur, depending on the context and institutions governing the transaction. We have empirical evidence that private agents will engage in basic research. But will they engage in enough? Of course, one of the difficulties is that we do not know the optimal amount of research.
Not only is the free rider problem in research empirical, it’s important to remember not to commit the Nirvana fallacy. Just because private agents may underprovide research relative to some theoretical government-funded benchmark, that doesn’t mean you are making the correct comparison. The correct comparison has to incorporate the transaction costs under both alternatives, the potentially negative effects of being too commercially focused in private research, the effects of short-term focus in both private and public funding, the dissipation of value through rent-seeking for public funding, and the costs of the politicization of research that can arise with public funding.