In Understanding the Process of Economic Change Douglas North observed that governments are less efficient than markets at solving problems because the problems that are left for the government to solve are the ones that the market cannot solve efficiently. Or, to put it another way, governments are less efficient than markets because of the nature of the problems each is asked to solve, not the inherent nature of governments vs markets.
The implication here is that government may well be the most efficient mechanism for solving the sort of problems left for governments to solve—at least at the time that the government took on the solution. Markets, presumably, would solve them no more efficiently. Similarly, governments are less efficient at solving the sorts of problems that markets are good at solving.
What are some factors that make markets or government better fit for a problem? I don’t have an exhaustive list, but two factors mentioned in the book are transaction cost and alignment of incentives.
When transaction costs can be fully internalized into the price, the market is likely to be more efficient than a centralized solution imposed by the government. However, if costs are not fully internalized then a centralized solution can be more efficient. However, those are not the only two options. A hybrid option is that the government uses its authority to force the internalization of cost. This is the idea between markets in various emissions.
Incentive alignment is harder to describe concisely. Problems that are well suited to the market are those whose incentives are aligned to the rules of the market. That means that markets tend to work well when the main factors you care about can be included in the price. However, price cannot capture all of the incentives that we might want to consider. When non-price translatable incentives contribute a non-trivial share of the value of some action, then market mechanisms are likely to be ineffective.
As a concrete and timely example, the market would be a terrible way to set policy for distribution of covid-19 vaccines. That would magnify the disparities we already see in covid-19 deaths because it would favor folks who are already less likely to die from the disease. However, markets may well be an effective mechanism for setting up the actual distribution network, with more vaccination capacity allocated to those organizations that demonstrate an ability to more effectively deliver the vaccines to people according to the schedule that the state and national governments determined. As this article discusses, in West Virginia a public-private partnership is getting effective results—and a different public-private partnership with CVS and Walgreens is not doing as well.
However, we should not assume that this flow is static or only goes one direction. Some solutions taken on by the government may be better taken on by the market. When this is because the government chooses to not solve the problem—whether or not the government could theoretically solve it more efficiently—then it is easy to see that the market should solve the problem instead. When we talk compare efficient solutions, we only include those which are actually available.
The harder case is when capabilities and constraints change in a way affects the efficiency with which a problem can be solved. A problem that was once most efficiently solved by the government may now be more efficiently solved by the market. A problem that was once solved most efficiently solved by the market may now be more efficiently solved by government.
These cases are the most challenging because organizations, once given a power, tend to be reluctant to give up that power. The more monopolistic the organization that holds the problem solving power is, the more resistant it will be to change. This means that in markets where a player has disproportionate power shifting problem solving power will be harder—whether that shift is within the market to another player or to a government. Even harder is when something should shift from the government back to the market. Since the government generally puts the force of law behind its monopolies, even motivated new solution providers can be stopped from getting a foothold.
So the interesting debate is not government versus markets, but rather which problems are suited for which problem solving mechanism. This is not a tradeoff we should discuss once but one which we should be reevaluating regularly as the landscape of capabilities and constraints evolves. Sometimes, the best approach may be a hybrid solution where we utilize a public-private partnership to divide the problem we are trying to solve into parts that are best handled by the market and those which, due to externalized transaction costs or incentive alignment, are better solved by the government.