In your statements, you have advocated tax cuts over government spending as a policy instrument for economic stimulus. Could you explain why tax cuts would be more efficient?

I would not use the words “more efficient,” but I will come back to that. My view is that spending increases would certainly make sense if there were projects available to the government that clearly had benefits in excess of costs. Standard examples would be bridges that are falling down and that need to be repaired. I completely accept that there must be some projects out there that the government needs to do, and there are clearly some parts of the interstate highway system that are in need of repair. However, I am skeptical that there is anything remotely near to US$400 billion of projects that make sense on standard cost-benefit terms, which the government would want to fund independent of the recession.

Now, if you use the standard Keynesian justification that government expenditure improves the economy, government spending is valid even it is totally wasteful—even it is digging ditches and filling it back up. Nothing in the model assumes that those expenditures are, or requires those expenditures to be, somehow productive. Instead, any kind of expenditure “works.”

Yet, it seems very, very unlikely that wasteful government spending could make sense in the long run. Now in the very short run, it might make some sense under some assumptions, but I think those assumptions are not so trivial. In the long run, we do not want to have devoted much government effort to digging ditches and filling them back up. That is crowding out, taking resources away from other, more productive uses of the economy and the government budget. If one argues that we could easily create government projects, even projects of questionable cost-benefit value that would be purely temporary and would get the economy going again—either in jumpstarting the economy away from recession or moving it toward full employment—and if one states that we could eliminate these once they were no longer needed for stimulus purposes, there would need to be a rational argument for that.

My assessment is that many allegedly short-term stimulus measures are still with us decades or even centuries later. Therefore, at a minimum, one should think both about the long-run efficiency effects of government purchases and not just about their short-run stimulus impacts. So, is there a way to try to stimulate the economy without taking the risk that we embark on some ill-conceived government expenditure projects? The answer is yes, we can cut taxes. In the standard Keynesian model, tax cuts are effective at stimulating private demand, creating either more consumer demand or more investment demand. We could certainly use more tax cuts.

Now, the conclusion in the standard model is that one gets a bigger “bang for the buck” with government expenditures—for a given size increase in the deficit, there is a larger effect on output from expenditure increases than from tax cuts. This effect is the government purchase multiplier, which is larger than the tax cut multiplier. The Keynesian model is relatively narrow, and there are other factors that play a role in determining what is more effective. In particular, if you think about the efficiency effects of lower taxes, then that might cause taxes to be more beneficial than government spending over a medium or longer-term horizon, even if not in the short run. Existing evidence suggests that tax cuts also have a large multiplier effect in stimulating output, and government purchases are somewhat more modest. Indeed, the evidence from the tax cut multiplier does not come from some right-wing hyena. It comes from Christie Romer, who is the Chairman of the Council of Economic Advisors under Obama, and should not be ignored.

Now there may be some other factors too, but at a minimum, the tax cut view seems completely logical and consistent. It clearly has the potential to have the desirable short-run Keynesian stimulus properties, and by reducing the disincentives from taxation, it has beneficial and sensible long-run properties. To a large degree, deciding on tax cuts versus expenditure increases is really taking a position on who gets to spend the increased spending, on who decides what the projects are. If we use expenditures by the federal government or the state governments, then legislatures, governors, and presidents decide. If we use tax cuts, then individuals decide. My perception is that for the most part, individual firms and businesses make better decisions than governments.

Now there are other parts of the stimulus where you might say something slightly different. You could make an argument, a quite rational argument, for government subsidies of research and development because there is potentially a beneficial externality. If we are to talk about policies to be more “green” and other such programs, many of these are legitimate, but there is a much better way to accomplish this than to have government try to monitor research projects or to have government decide that some jobs are green jobs and others are not. This could inevitably be the subject of all kinds of political abuse. There are other policy instruments, such as increasing carbon taxes. That might just mean a gasoline tax, but it would ideally be broader than just a gas tax in order to prevent a substitution with coal. In any case, the appropriate taxes can internalize the externalities. You have one simple, clean policy. Furthermore, it is very easy to see how big that policy is – you know exactly what you are taking in through tax revenue and how much you have raised the price relative to the market price. Furthermore, this forces people to think about the cost. There must be a more explicit form of pain, which may be worthwhile—it is going to be more upfront and explicit than saying “Oh, we will just do it with more expenditure.” When this expenditure comes from borrowing, from future generations, from thin air, it is very easy for politicians and some voters to say government spending is good and not focus on the fact that it has a real cost. Therefore, a tax is a better approach.