Defining 'Infrastructure' Is a Nonsense Argument


The debate over President Joe Biden’s infrastructure bill has become preoccupied with the unimportant question of what counts as infrastructure. This is a pointless distraction; what really matters isn't how to define infrastructure, but what the government needs to spend money on to improve the country.

Less than half of Biden’s proposed $2.25 trillion in government spending would go to things that we typically refer to as “infrastructure,” such as roads, bridges, airports, water systems, transit and electrical grids. The rest includes spending on research and development, schools, housing, elder care, education, and support for manufacturing and small business. Some opponents are using this fact to attack the bill:

This could backfire politically; regardless of whether it is or isn’t infrastructure, people like the idea of elder care. But it’s possible that the chattering classes will get distracted by a meaningless debate over whether the legislation ought to limit itself to a narrow definition of infrastructure. In fact, there’s no good reason to waste time thinking about this question. The government should simply spend money on things that are good to spend money on.

And one of those things is providing public goods. Public goods are things that the private sector usually can’t do on its own because it can’t capture all the value it creates. A network of roads, for example, creates lots of economic activity in the towns near the roads. That value can’t all be recouped by tolls. That’s why every country that builds a good road network does it with extensive government involvement; the private sector simply doesn’t build enough roads when left to its own devices.

But lots of government activities fall into this category without being traditional “infrastructure.” The best example is research and development. Since scientific ideas can be copied at essentially no cost, research creates huge economic spillovers. But while the private sector does do a lot of research, it’s almost all of the applied, practical variety. Basic science, which provides the foundation for a lot of those other discoveries in the private sector, is mostly done with government funding through agencies like the National Institutes of Health and the National Science Foundation. It’s the same story as with the roads:  private companies just can’t monetize basic research, so they don’t do enough of it.

So to criticize Biden’s bill for spending hundreds of billions of dollars on research, simply because research isn’t something we typically call “infrastructure,” makes zero sense. The economic rationale is exactly the same.

Plenty of public health improvements fall into the same category. There’s volumes of evidence that childhood exposure to lead damages people’s brains in ways that makes them commit more crime as adults. That crime imposes a grievous cost on society. Cleaning up lead pipes would thus create plenty of social value that utility companies have no way of recouping. Since utilities don’t have an economic incentive to clean up the lead, Biden wants to do it for them. Again, this is the same public-goods rationale as building roads. Some might even argue that schools are a public good, based on the idea that a better-educated populace cooperates more effectively, generating broad social benefits that go beyond what individual students would ever pay for with their tuition dollars.

What about housing, though? It’s very hard to argue that’s a public good. But with housing, there’s a political impediment to development; powerful homeowners called NIMBYs dominate local governments across the nation and block new construction. The federal government’s power to compel localities to build new housing is highly limited under our constitution. But it can spend money directly on creating housing, to help cancel out the effect of the NIMBYs.

OK, so how about elder care? That’s not a public good, and there are no NIMBYs blocking it either. The $400 billion that Biden would spend on taking care of the old and disabled is a form of in-kind redistribution: giving people money by providing them with specific services. In many cases, people would rather get cash than in-kind services from the government, but in the case of health care -- and thus probably elder care as well -- they often prefer getting the service itself.

So is in-kind redistribution an appropriate thing to put in an infrastructure bill? I struggle to see a reason why it shouldn’t be. The rationale for redistribution is that it creates a more equal society; by definition, that is something private markets won’t do on their own. If subsidized elder care is something people want, and it’s something only the government can do, then do it.

And since the political system forces us to lump elder care in with public goods in a single piece of legislation (because Republicans could block multiple bills with filibusters), then go ahead and put it all in one bill. There is no practical, political, or moral justification for creating some kind of rule that only public goods should go in an “infrastructure” bill. Let’s not let semantics distract us from the central task of building a better country.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Noah Smith is a Bloomberg Opinion columnist. He was an assistant professor of finance at Stony Brook University, and he blogs at Noahpinion.

©2021 Bloomberg L.P.

BQ Install

Bloomberg Quint

Add BloombergQuint App to Home screen.