Is infrastructure spending a free lunch?

The International Monetary Fund is having its annual meeting in Washington, DC this weekend, which includes a variety of talks and panels (opens in a new tab) about the state of the global economy. Before attending one of these panels focusing on public spending and economic growth, Paul Krugman posted (opens in a new tab) a graph of construction spending in the United States in recent years. The clear downward trajectory of the spending juxtaposed with declining inflation-adjusted interest rates demonstrates the clear case for increased infrastructure spending, according to Krugman.

Making this point before attending an IMF panel on the topic is fitting given recent work by the international economic agency. As part of its World Economic Outlook for this year, IMF staffers argue that spending on infrastructure can actually pay for itself by boosting growth in the short term and the long term. For an organization well known for advocating for fiscal austerity, it’s quite the turnaround.

At The Washington Post, Larry Summers, the Harvard University professor and former Treasury Secretary, lays (opens in a new tab) out part of the IMF’s argument. In essence, he says the investments can fund themselves when interest rates are very low. Summers gives a numerical example: if the real return on infrastructure investment is 6 percent, the government would receive about 1.5 percent, or 25 percent of the total return. And since real interest rates, or the cost of borrowing, for the United States is about 1 percent, this means the government would earn an extra return of about 0.5 percent.

The IMF’s conclusion is similar to one made by Summers and our own Brad DeLong in the Brookings Papers on Economic Activity (opens in a new tab). The case made by DeLong and Summers, however, is even bolder. They argue that all expansionary fiscal policy can be self-financing—not only infrastructure spending but also other forms of government spending and transfers. DeLong and Summers’s argument rests on the idea of hysteresis, the idea that workers and other resources idled in a recession can become unproductive and reduce the long-term growth rate of the economy. If this hysteresis effect is large enough then current fiscal policy that quickly puts the economy back toward its long-run potential will be paid for by the future output it created.

Crucially, these arguments rely on low interest rates and a very weak economy. These facts are the reality at the moment, but they haven’t always been that way and mostly likely won’t be that way in the future. The IMF and DeLong and Summers aren’t arguing that infrastructure investment or fiscal policy is self-financing everywhere at all times. They work best in periods of economic slack and unused resources. Sadly, our times fit that description all too well.