Standard economic theory maintains that, in the absence of externalities, private investment works pretty well. Entrepreneurs tend to acquire the capital necessary to take on valuable projects because they stand to gain when those projects succeed and lose when those projects fail. Public investment, in contrast, is not subject to the same profit-and-loss mechanism. The relevant public sector decision makers have a hard time knowing whether a project is worth pursuing and have little incentive to act in accordance with that information when it is available.
If the resources required to take on various projects are scarce, we usually want the private sector to choose how those resources will be used. Private entrepreneurs will tend to ensure that resources are used to produce the most valuable goods and services in the least costly ways. Handing these investment decisions over to politicians is likely to result in less desirable projects. The more desirable projects that private entrepreneurs would have taken on will be “crowded out” by public sector investment.
In a recent article, Robert Skidelsky boldly claims that “the crowding-out argument is wrong.” He rejects the assumption that resources are scarce, claiming instead that “most market economies normally have underemployment or spare capacity.” If that is the case, he writes, “public investment can ‘crowd in’ resources that otherwise would be idle.” He then attempts to demonstrate that “the state has in practice always played a leading role in allocating capital.”
Skidelsky’s argument is insufficiently supported. His supposed refutation of standard economic theory, as it pertains to crowding out, does not actually refute the theory. And, more importantly, his mustering of the empirical evidence to show that the state has “always played a leading role” leaves a lot to be desired.
Skidelsky claims to show the crowding-out argument is theoretically wrong because one of its key assumptions—that resources are scarce—does not hold in the real world. But showing that an assumption does not hold is not enough to demonstrate that a theory is wrong. At best, it would only show that the theory is inappropriately applied.
A theory is essentially an if-then statement. If the assumptions hold, then the predictions of the model will follow. Suppose one were to make the following statement:
If my mother’s sibling is male, he is my uncle.
There is no denying the statement is correct, as “uncle” is defined as “mother’s male sibling.” Noting that, in fact, my mother’s sibling is female does not make the statement wrong.
Similarly, Skidelsky’s noting that “[i]n fact, most market economies normally have underemployment or spare capacity” does not make the crowding-out argument wrong even if his claim about underemployment or spare capacity is correct. The crowding-out argument takes the assumption of resource scarcity as given. If resources are scarce, then public investment will crowd out the private sector. In order to show that the crowding-out argument is wrong, Skidelsky would need to demonstrate that the private sector is not crowded out by public investment when resources are scarce. He does not do that.
But let’s not let basic logic get in the way of what’s really important here. What Skidelsky presumably means is not that the theory is wrong, but rather that the theory is inappropriately applied.
The strict assumptions of a theory rarely hold in the real world. Fortunately, a theory can be useful even if its assumptions are not met. Theories identify causal mechanisms and, in doing so, help one think more clearly about cause and effect. The relevant question is not whether public investment always crowds out the private sector; it is whether the public sector tends to channel resources into more valuable projects than the private sector.
To this end, Skidelsky offers a parade of examples. Toyota benefited from “tariff protection and state subsidies.” Silicon Valley benefited from government-financed research and development. China’s “economic ascent is the apotheosis of state-led development today.”
As illustrative as Skidelsky’s examples might appear, they are anecdotal. One should consider whether they have been cherry-picked. That the public sector occasionally channels resources into more valuable projects than the private sector would not indicate that it tends to do so.
The bigger problem with Skidelsky’s examples, however, is that they do not even show that the public sector occasionally channels resources into more valuable projects than the private sector. That Toyota, Silicon Valley, and China look like good outcomes do not tell us whether outcomes would have looked worse without the corresponding public sector efforts. Indeed, we do not even know whether these apparently good outcomes happened because of or in spite of public sector efforts.
To avoid drawing the wrong conclusion from anecdotal evidence, we should take a step back and consider whether the public sector tends to channel resources into more valuable projects than the private sector—and, if so, at what margin or under what conditions.
In extreme cases, where investment decisions have primarily been made by the public sector, the projects taken on appear to be much worse than what likely would have resulted if those decisions had been left to the private sector. The Soviet Union, China (before reforms), India (before reforms), Cuba, Venezuela, and North Korea all fared poorly when investment decisions were made by the public sector.
Even in less extreme cases, however, those countries with more state interference in the allocation of capital have tended to experience lower levels of real income and lower rates of economic growth. In general, the evidence suggests that private investment usually works better than public investment.
How about when labor is underemployed or the economy has spare capacity? Might public investment productively employ idle resources in that case, as Skidelsky claims?
It might. But we should not expect it to. In this case, the causal mechanism identified by the crowding-out argument is worth considering even if all of the assumptions do not hold.
Private investment usually works better because entrepreneurs are subject to the profit-and-loss mechanism. They have an incentive to seek out the relevant information and take on worthwhile projects. The incentives are much muted for public sector decision makers, as they are not subject to that same profit-and-loss mechanism. Hence, even if the public sector might improve matters because there is scope for improvement, we should not expect it to improve matters because it lacks the incentives to do so. Instead, we should expect public sector decision makers to allocate capital to advance their own personal and political ends.
Even if public investment could be expected to productively employ idle resources, however, that would amount to a fairly limited scope for public investment. We are not generally in a recession. And, oftentimes, resources appearing to be idle are not actually idle.
While Skidelsky rests his argument for public investment on a theory of idle resources, he does not seem to accept the limited scope for public investment that such a theory implies. He has repeatedly called for a national investment bank. Such a bank would not merely productively employ resources that might otherwise remain idle during occasional short run slumps. Rather, it would be “lending for the long term,” he writes, to “help long-term growth.”
Why should one expect public investment to outperform private investment on the margin in normal times? Skidelsky doesn’t say. Standard economic theory and the available empirical evidence says we shouldn’t.
Recognizing that private investment usually works better than public investment does not mean that public investment is never desirable. But it does put the burden of proof on those who champion public investment projects. Rather than offering faux refutations of the crowding-out argument and citing anecdotal evidence, as Skidelsky does, those in favor of a public investment project should make the case for why it is desirable. They should explain why their preferred public investment project will yield superior outcomes, in contrast to the many underperforming public investment projects that have been taken on in the past.