This is an archived article that was published on sltrib.com in 2016, and information in the article may be outdated. It is provided only for personal research purposes and may not be reprinted.

The bidding to be the next "infrastructure president" has kicked into high gear. Hillary Clinton has released a plan calling for $275 billion in infrastructure spending, on top of the $305 billion highway bill that President Barack Obama signed last year. Clinton's rival, Donald Trump, has promised far more, though his trillion-dollar figure is unlikely to be realized.

Sadly, but perhaps predictably, a lot of people on the political right have come out against infrastructure spending. Some question the need for road and bridge repair. Others deny the need for federal infrastructure spending in the first place. Still others claim that while spending is good in theory, the U.S.'s high infrastructure costs mean we should hold off until those costs can be brought down. A few simply ridicule Clinton's proposals and call it a day.

There are countries in which opposition to infrastructure spending would be helpful. In the 1990s, for example, Japan dramatically overbuilt roads, bridges and other construction projects. The country really does have bridges to nowhere.

But in the U.S., the automatic pooh-poohing of infrastructure spending does us a disservice. As a group, our elites downplay the benefits and inflate the cost of what we used to call internal improvements.

Things like transportation networks aren't like other forms of capital. They're not like factories, or machine tools or office buildings, because they have elements of what economists consider a public good. This is something that the private sector, left to its own devices, can't or won't provide enough of. And indeed, if you look around the world, although there are certainly private toll roads, the government is a big player in road- and bridge-building almost everywhere. There's almost certainly a reason for that.

First, government can solve the coordination problem involved in building a road. Roads cross many different people's land, and there's always the danger that one last holdout landowner could scuttle an entire project. But government, unlike a private company, can use the power of eminent domain to force that one holdout to sell.

Second, roads have an interesting property. If the road isn't congested, the cost of adding one more car to the road is essentially zero. That means that unless roads are very full, private companies will have trouble reaping the full economic benefit from building them. They can charge tolls, but since the benefit of the road goes down when it's less congested, the toll ideally should change depending on the number of cars on the road - a difficult system to implement, especially when a different company owns each different road in the network. This makes it tricky for free markets to handle the job of building road networks.

Finally, there's a subtler benefit. Transportation networks, and other infrastructure like water and electricity, allow businesses to cluster together. That produces benefits known as agglomeration externalities that go beyond any single business' incentive to move to an area. Those benefits also can't be captured by private companies, which is one reason we almost never see cities where the roads are privately owned.

So for all these reasons, infrastructure is often — though not completely — a job for government. It represents a way that government spending complements private business activity rather than crowding it out.

This principle is fairly easy to see in economic models. In 1993, economists Marianne Baxter and Robert King showed that if government can make a kind of capital that the private sector doesn't, then government investment gives a big boost to the economy, because it increases private investment as well. That benefit will look like a fiscal stimulus, but it's not Keynesian in nature — it comes from public goods.

In a recent article, I discussed how to analyze the costs and benefits of infrastructure spending. One potential cost comes from the fact that building or fixing roads and bridges requires using resources that private companies could otherwise use. That's called "crowding out." But one big benefit of infrastructure spending is that after it's finished, private investment can be "crowded in," due to the complementarity between government capital and private capital. The World Bank has found clear evidence that crowding in occurs in developing countries.

And a developing country is what the U.S. might become if it lets its infrastructure deteriorate more. The danger of spending too much is that you get bridges to nowhere, but the danger of spending too little is that cities degrade into economically isolated slums. Japan spent too much on infrastructure, it's true, but almost any country would rather be Japan than Nigeria.

As for the U.S.'s outsized infrastructure costs, these are very real, and they're a huge problem. The U.S. should do everything it can to find and eliminate the source of excess costs, whether these come from burdensome regulation, costly land acquisition, inefficient environmental review processes, inflated union wages or a combination thereof.

But bringing down costs is a difficult, long-term task. If the U.S. waits until costs decline to repair its crumbling roads and bridges, the cost of maintenance will rise due to the increased physical degradation, canceling out much of whatever cost savings can be made by improving the system.

So the people who are leaping to attack infrastructure spending bills should think again. The benefits are greater than they realize.

- - -

Noah Smith is an assistant professor of finance at Stony Brook University and a freelance writer for a number of finance and business publications.