Deficit Spending on Infrastructure in a Depression

Last week we blogged on whether a deficit-spending program--a stimulus package currently estimated to cost $825 billion--is appropriate to deal with the current economic crisis. I will not repeat the points I made. If there is a risk of deflation, and increasing the supply of money is not considered an adequate response, then there is an argument for the stimulus. It is essentially Keynes's argument: if private demand falls short of the supply that the economy is capable of producing, public demand--public expenditures on projects that will put people to work--can fill the gap.

Our focus this week narrows to the infrastructure portion of the program, which is the most conventionally Keynesian. The stimulus package proposed by Democratic Congressmen--the "American Recovery and Reinvestment Act of 2009"--allots $90 billion to infrastructure: $30 billion for highway construction, $31 billion for energy savings in federal and other public facilities, $19 billion for flood control and environmental improvements, and $10 billion for transit and rail. Other parts of the Act, however, allocate additional funds to other construction projects, and from an anti-depression standpoint it is really construction that is the relevant category. There is a lot of unemployment in construction--110,000 construction workers were laid off in December--as part of the continuing fallout from the housing bubble, which increased the demand for new houses; the demand has now collapsed.

One objection to public-works spending as an anti-depression measure is that by the time work on the public projects actually begins, the depression will be over and all that remains will be the bill for the projects, in the form of an increased national debt, since public-works spending that is financed by taxes rather than borrowing has no effect on increasing demand for goods and services. What is given with one hand is taken away with the other. But construction projects, especially those interrupted or postponed because of the economic collapse, can be started up (or resumed) pretty quickly. Moreover, this depression (as I think it is, and not merely a recession) is likely to last at least two more years, and that should be time enough for much of the $90 billion (plus additional money allocated to construction) to be spent.

Another advantage to infrastructure, and construction generally, as an emergency measure is that it may not add significantly to the deficit in the long run. The reason is that the costs can be recouped out of user fees, such as tolls for highways and taxes on airline and railroad tickets. This presupposes that the projects create some real value, unlike the "bridge to nowhere" that was proposed for Alaska, or else the user fees will really just be taxes. To the extent that the money allocated to infrastructure in the American Recovery and Reinvestment Act (which doubtless however will be changed before it is actually enacted and signed by the President) is for interrupted or deferred projects, or merely accelerates projects planned for a later date (accelerated to increase demand now), there will not be incremental waste--that is, waste beyond what is already built into approved projects. With new projects, the risk that costs will exceed benefits is greater, but we must be careful not to view costs and benefits in too narrow terms. Even a worthless project, if it puts people to work, can reduce the economic impact of a depression, as Keynes argued long ago.

Another advantage of construction-oriented public-works spending is that the risk that it will crowd out private investment is small. If labor and other resources used in construction were being fully employed, then the only effect of the government's launching construction projects would be to increase construction prices, because it would be bidding against private employers for a limited stock of labor. But given the high unemployment rate of construction workers, there are plenty of them to hire without an employer's having to lure them away from other employers by the promise of higher pay.

One thing that must give one pause, however, is the question of substitutability across construction projects. Building a house and building a highway are not interchangeable construction activities. Unemployment in the construction industry may be concentrated in residential construction, and residential construction workers may not have the right skills for highway or bridge construction, let alone for flood control. (What use is a plumber or a carpenter in building a highway?) The more specialized the American workforce has become, the less effect Keynesian public-works projects are likely to have on employment. And if they do not reduce unemployment but instead compete with private employers for workers, the only effect may be to increase the national debt and engender inflation (though inflation is one way of combating deflation). This is a general concern rather than anything peculiar to construction. In fact it is a greater concern with some of the other projects proposed in the American Recovery and Reinvestment Act. Projects designed to promote efficient use of energy (in order to limit global warming and dependence on foreign oil--worthy objectives, however) will create inflationary pressure by bidding for scarce resources (such as scientific and engineering skills and complex, novel technologies) against the private sector. That is a compelling reason for concentrating the stimulus in the industries in which unemployment is greatest.