Why Infrastructure Spending Is a Bad Bet

by Veronique de Rugy

As we know, one of the things the president will call for tonight is more infrastructure spending. We have heard many times that infrastructure spending, in one form or another, is the key to growth and job creation — and, in the president’s defense, he certainly isn’t the only one who refers to stimulus and government spending as “investing” in infrastructure.

No one disputes that American public works need improving, and economists have long recognized the value of infrastructure. Roads, bridges, airports, and canals are the conduits through which goods are exchanged. However, whatever its merits, infrastructure spending is unlikely to provide much of a stimulus — and it certainly won’t provide the boost that the president will promise the American people tonight. 

For one thing, even though Mark Zandi claims that the bang for the buck is significant when the government spends $1 on infrastructure ($1.44 in growth), that’s just his opinion. The reality is that economists are far from having reached a consensus on what the actual return on infrastructure spending is. As economists Eric Leeper, Todd Walker, and Shu-Chum Yang put it in a recent paper for the IMF: “Economists have offered an embarrassingly wide range of estimated multipliers.” Among respected economists, some find larger multipliers and some find negative ones. (Thanks Matt Mitchell for this great paper).

Second, according to Keynesian economists, for spending to be stimulative, it has to be timely, targeted, and temporary. Infrastructure spending isn’t any of that. That’s because infrastructure projects involve planning, bidding, contracting, construction, and evaluation. Only $28 billion of the $45 billion in DOT money included in the stimulus has been spent so far.

We know that the stimulus money wasn’t targeted toward the areas that were hit the most by the recession, but even if the funding were targeted, it still might not be stimulative. First, the same level of job poaching from existing jobs would have happened; construction workers tend to be highly specialized, and skilled workers rarely suffer from high unemployment. Many of the areas that were hardest hit by the recession are in decline because they have been producing goods and services that are not, and will never be, in great demand. The overall value added by improving their roads is probably a lot less than that of new infrastructure in growing areas that might have relatively little unemployment but do have great demand for more roads, schools, and other types of long-term infrastructure.

As for being temporary — which stimulus spending needs to be to work — what the president will propose tonight is likely to cost the American people money for a very long time.

Infrastructure spending tends to suffer from massive cost overruns, waste, fraud, and abuse. A comprehensive study examining 20 nations on five continents (“Underestimating Costs in Public Works Projects: Error or Lie?” by Bent Flyvbjerg, Mette K. Skamris Holm, and Søren L. Buhl) found that nine out of ten public-works projects come in over budget. Cost overruns routinely range from 50 to 100 percent of the original estimate. For rail, the average cost is 44.7 percent greater than the estimated cost at the time the decision was made. For bridges and tunnels, the equivalent figure is 33.8 percent, for roads 20.4 percent. 

I should also add that I think it’s a mistake to assume that it is the role of the federal government to pay for roads and highway expansions. With very few exceptions, most roads, bridges, and even highways are local projects (state projects at most) by nature. The federal government shouldn’t have anything to do with them.

In fact, I would argue that taxpayers and consumers would be better off if these activities were privatized. And if states aren’t ready for privatization, they can do what Indiana did a few years back when it leased its main highways to a private company for $4 billion. The state was $4 billion richer, and it was still the owner of the highway. Consumers in Indiana were better off, because the deal saved money. Experiences in other countries have also shown that privatization leads to innovation and reduced congestion. (I really like this paper by Randy O’Toole on these issues.)

Spending more on infrastructure right now, and expecting miracles from it, is a very risky bet.