Budget Deficits and a Balanced Budget Amendment:

I received this email from a reader:

Prof.,

I have yet to encounter articles on the desirability of Fed budget deficits. Today's [Wednesday's] news states that due to increased tax revenue, the Fed deficit this year was reduced $100 B, but is still at $338 B. This in conjunction with a story I saw in Monday's WSJ (about corporate cash reserves being at an all-time high) had me thinking about the following: As corp debt is to some degree desirable, can gov't debt ever be?

Principally, corporations leverage themselves so that they can invest in what they are good at and make a considerably higher return than the interest rate on their debt. Could there ever be such an economic justification for gov'ts? Or is it just poor management, coupled with crisis spending, i.e., wars, hurricane relief, etc?

I would be interested to see you post on this topic on the VC.

This is something that I also haven't seen much commentary on and since this is something I have thought about a bit (its actually related to questions of household savings and that sort of thing that I work with all the time), so I'll give my thinking on it. I offer this to stimulate discussion and I hope other economists will offer their views too, especially if they have relevant empirical evidence that sheds light on the question.

The answer is yes, my view is that I think deficits could be ok but only under limited circumstances. In general, however, they are difficult to justify. Thus, it is likely that they are more defensible in theory than in practice. Whether the current budget deficits are defensible is an empirical question, I think, for reasons that I will elaborate, but that it is doubtful that much of recent budget deficits are justified.

First, as suggested in the email there is an analogy to corporate debt, if the government is making profitable investments in collective capital investment projects that the private sector would not undertake efficiently. So that if the government was building infrastructure, as local governments do when they go into debt to build schools or road. Or, arguably, investments in national defense or anti-terrorism could be these sorts of collective capital investments if they actually build a sort of anti-terror infrastructure. The federal government does not issue bonds in the same way that a municipal government does, but under this logic, they would be instrumentally similar.

There are, however, two qualifications to this. First, it must be that the investments are profitable investments, i.e., as mentioned in the email, where the rate of return exceeds the interest rate paid (which is actually the opportunity cost, as we will get to in a second). Unlike private businesses, however, there is no real accounting for the economic return on these projects, so it is pure speculation whether the return exceeds the interest rate. Given the amount of pork in infrastructure budgets (see the recent budget-busting highway bill) and the tendency to allocate these funds to political ends, rather than their highest-valued economic purpose, there is some doubt as to whether they recover their cost. As for investments in anti-terror and miliary expenditures, others have detailed the ways in which much of this money too has been diverted to other (pork) ends. So the WSJ definitely has a point in noting that the war against terror is certainly the sort of thing in which most people would agree it is appropriate to fund through deficit spending (just like wars are usually funded through deficits) and that this is probably a wise investment in theory, it also seems clear that in practice some of this money has been inappropriately diverted to nonproductive ends.

The second qualfication, of course, is that the money is actually being invested, not merely spent on consumption. To the extent that the deficit is caused by things like transfer payments (such as social security), etc., then it is difficult to see why this should be funded by deficits rather than taxes.

The second circumstance is what households do with the money. A "deficit" is nothing more than a deferred tax increase--either we pay today, or we pay in the future plus interest. So a deficit simply means that we have more money today, and less money whenever "we" have to repay the loan. There are three things that we can do with this "windfall." (1) We could simply save it in anticipation of higher taxes later--for instance, if we bought Treasury Bills then we would in theory perfectly break even vis a vis our future tax liability, or we could save the money and bequest it to our children to pay off our portion of the taxes. In this scenario (which is essentially similar to what is called "Ricardian Equivalence") deficits would be perfectly neutral. (2) We could spend our greater current income on current consumption, in which case this is a simple intergenerational wealth transfer, allowing us to consume today beyond our true wealth and trying to get someone else to pay for it.

Or, (3) We could invest it ourselves in some sort of individual investment and capital development. This is the most interesting arguemtn. It could be that if there are imperfections in certain capital markets for some reason, it might make sense for the government to borrow on our behalf and give us the money. For instance, the government essentially does this when it guarantees student loans--because of the riskiness of these sorts of investments in human capital, it is conceivable that such a private loan market might not work perfectly efficiently, and so a guaranteed student loan program might enable more efficient investments in human capital markets. So, if people are using their "extra" money, say, to quit working and go back to school to develop more valuable skills, and for some reason this is the most efficient way to enable them to do this, then it might make sense.

Again, however, the qualifications are substantial. My recollection from the Ricardian Equivalency literature (and any economists out there can set me straight on the current empirical state of the world on this question) is that people tend to save some of this and consume a substantial portion of this. Perhaps they invest some of it as well, but for most of those investments my guess is that the case for funding them through deficit spending is pretty week.

In the end, the case for budget deficits boils down to an empirical question about what the government and private individuals do with the money they get today, and the recognition that deficits today are just future taxes. In general, however, I am skeptical that capital markets are so incomplete that it makes sense to fund individual human capital investment through government deficit spending. And much of the deficit spending by the government today seems unrelated to true investment ends as well.

Two other arguments have also been made that I won't really discuss much, as they are more political and philosophical arguments, rather than economic arguments.

If the problem with deficits, roughly stated, is excessive government spending that funds current consumption (rather than investment), some have argued that deficits have the practical utility of constraining the overall size of the spending budget. This is Milton Friedman's famous justification for Reagan's budgets. This assumes, however, that there is some limit on public tolerance for budget deficits--it is no longer clear that this is the case. I had a long talk with James Buchanan a few weeks ago, and he strongly believes that this is the case, and that today, spending would be lower if we forced it to be financed by taxes rather than permitting deficits.

Some have also made a philosophical argument that even if there is an intergenerational wealth transfer, it is justifiable, because our children and grandchildren will almost certainly be wealthier than us, and (oversimplified) that they "wouldn't mind" lending us some money today, just as we leave money for them in our wills. This doesn't seem to be particularly persuasive to me, but it is one argument that is occasionally made to justify even deficits that fund current consumption.

As for a balanced budget amendment to the Constitution, the idea seems to be an idea that has disappeared as a political matter, nonethless I think that the intellectual argument is interesting. Leaving aside practical enforcement problems for a moment, I understand the strongest argument in favor of a balanced-budget amendment to be a public choice sort of argument, rather than an economic one. The idea would be that to the extent that a balanced budget amendment created some constraint on spending, it would be to create a sort of prisoners' dilemma game among politicians--i.e., I could only get more for my project if you got less. The logic of the argument would be that through this competition for limited funds, politicians would have the incentive to identify and plunder one anothers' pork projects, thereby increasing the public scrutiny on pork. Under the current system, by contrast, they can logroll each others' pork and shove the cost into the deficit. The argument would be, therefore, that the effect of a balanced-budget amendment would be to reduce expenditures on pork more than expenditures on public goods.

My sense of this argument for a balanced-budget amendment is that it too seems theoretically sound, but that the enforcement problems with a balanced-budget amendments, and the need for exceptions, would probably render it unenforceable in practice. In particular, it seems that any such BBA would probably put most entitlement spending off-budget, so it would not constrain exactly that which it would be intended to constrain in theory. Which is why I think the idea has fallen out of favor--in addition, of course, to changes in the political winds in Washington that no longer see it as, um, "necessary".

Update:

Professor Larry White notes in the Comments that he and Roger Garrison wrote a more extended discussion of this issue in the late-1990s in The Free Market, "Do Deficits Matter?" They also discuss Steven Landsburg's argument (also referenced in the Comments). I especially commend it for its discussion of Ricardian Equivalence, which I lacked the space to go into in the main post.