Hillary Clinton and Donald Trump differ on many issues, but when it comes to making big investments in America’s physical infrastructure, they are in general agreement. Clinton proposes to increase spending by $275 billion over five years to rebuild roads, bridges, airports, schools, and water systems. Not to be outdone, Trump has vowed to double that amount.
Both candidates support more new infrastructure spending because, as Trump puts it, our bridges are “falling down.” Both argue it would put millions of Americans back to work. And both say it would boost US productivity, a matter of utmost importance as the country founders in a “productivity recession.”
This bipartisan support for more infrastructure spending is not surprising. The image of putting people with hard hats back to work building tangible projects makes for a great political optic – and many economists, particularly on the left, applaud the idea. For example, the liberal Economic Policy Institute calls infrastructure spending “the next ‘new thing’ for powering economic growth.” And former Obama administration economic advisor Larry Summers argues that America is in a period of “secular stagnation” and needs to spend an additional $2.2 trillion over the next decade on physical infrastructure to jump start growth. But why so much faith that it will actually work?
One answer is that these infrastructure proponents are adherents of Keynesian economics, which holds that weak consumer demand is holding back robust economic recovery. Spending more money, especially to build things, is seen as a good way to get people back to work. But if that’s all it does, then it risks emulating Keynes’ rhetorical quip about governments putting cash in bottles and burying them in abandoned mines just so people can work at digging them back up.
The real key is not just to put people to work, but to put them to work doing things that will raise US productivity and long-term growth. How well does investing in physical infrastructure fit that bill? Summers says quite well, and for proof he cites a McKinsey Global Institute study that pointed to a 20 percent rate of return on such investments.
But these rates of return occur only for investments in things like electricity, broadband grids, and transportation networks in developing economies where there is very little existing infrastructure. In nations where infrastructure is largely built out, adding more infrastructure spending has a much lower economic growth impact.
It also depends on what kind of infrastructure we are talking about. Summers advises that the lion’s share of the money should be spent on projects to repair existing infrastructure where maintenance has been deferred, projects that "do not require extensive planning or regulatory approvals, so they can take place quickly.”
But filling potholes actually has a lower impact on productivity than expanding roads — given that Americans spend 8 billion hours a year stuck in traffic. Economists also have found that the productivity benefits from investing in education buildings, water mains, and other nontransportation infrastructure are low. This is why a comprehensive analysis of over 33 studies on infrastructure spending found that a 10 percent increase in infrastructure leads to a productivity increase of only about 0.5 percent.
So any effective infrastructure plan has to focus on expanding, not just repairing roads and bridges. But any plan also needs to go beyond pouring more concrete and laying more steel. Indeed, the infrastructure we need for today’s economy is rapidly changing with advances in information and communications technology.
Around the world, nations are using these technologies to make infrastructure smarter and more efficient. This new “hybrid infrastructure” that integrates both physical and digital aspects is critical to delivering the next wave of innovation and economic growth. For example, water mains embedded with Internet-connected sensors can detect and transmit information on leaks. Smart traffic lights that sense ebbs and flows and adjust accordingly can reduce travel time in cities by 25 percent. Overall, studies find that investments in IT-enabled infrastructure can have 60 percent greater productivity impacts than investments in roads alone.
Finally, infrastructure includes not just tangible things – concrete, steel, or chips – but intangibles. And one of the most important kinds of intangible infrastructure is scientific and engineering research. Federal support for research has been a critical driver of growth, powering a host of technologies, from semiconductors to the Internet to life-saving drugs. Unfortunately, federal support for R&D is lower as a share of GDP than it has been since before Sputnik in 1957. To match the level of federal R&D investment we enjoyed in 1987, Congress would need to increase R&D spending by $110 billion – but the economic studies show that investing in research produces about 2.5 times more productivity and GDP growth than spending on transportation infrastructure.
The bottom line is that America faces an investment deficit. And if we are going to solve it in a way that puts people to work and boosts long-term growth, we need to do more than repave existing roads and repair existing bridges. We need to make sure that a sizable share of the $275 billion to $550 billion that the presidential candidates are proposing goes to building new roads and bridges, systematically making all infrastructure smarter with information technology, and investing in cutting-edge science and engineering research. That will go a long way to addressing our growth challenge.
Robert D. Atkinson (@RobAtkinsonITIF) is the founder and president of the Information Technology and Innovation Foundation, a think tank focusing on the intersection of technological innovation and public policy. He also chaired the National Surface Transportation Infrastructure Finance Commission.