Benjamin Studebaker

Our new president-elect has a plan that has serious potential that we should all get really excited about–he wants $1 trillion in new spending on infrastructure. This sounds like a lot of money, but our crumbling infrastructure could use even more–the American Society of Civil Engineers estimates we’d need to spend $3.6 trillion to bring our infrastructure up to speed by 2020. Beyond the substantive benefits provided by the new infrastructure, there are tremendous economic advantages. An additional $1 trillion in spending would generate roughly somewhere between 6% and 10% GDP growth over the life of the program, depending on the size of the fiscal multiplier. It would also create piles of new construction jobs in the process. There are however some potential issues with how Trump wants to fund the plan, and we should talk about how to get the maximum benefits from it.

When’s the last time we threw big money at infrastructure? In 1956, Dwight Eisenhower created the interstate highway system, spending about 2.4% of GDP. Trump’s plan costs about 6%. Ike paid for his plan primarily with gasoline taxes, which contributed to the highway trust fund. This fund continues to pay for maintenance work on the system, but it’s not large enough to meet our needs today–it only funded about $47 billion in maintenance work last year on both the highways and mass transit systems.

Trump is wary of raising taxes and concerned about Republican congressional opposition. After all, this kind of infrastructure spending is really a much bigger, glitzier version of plans previously proposed by Hillary Clinton and Barack Obama. So to avoid right wing opposition, Trump has suggested that the government could induce the private sector to fund the projects with tax credits costing just $140 billion, or 0.8% of GDP.

The trouble is that the private sector is only going to be willing to fund projects that are profitable, even with a tax credit to help them deal with the large up-front cost of building infrastructure. This means that if firms do decide to build roads and bridges, they’re going to charge you to use them. If they can’t charge users, they have no reason to build. This means that pure public goods–like new pipes for Flint, Michigan–are not going to get built with this money. It also means that the funding mechanism will have regressive effects on the distribution of wealth, because tolls are flat fees which penalize low income users. Of course, gasoline tax is also a regressive funding mechanism, but at least low income motorists can buy cars that get better mileage. There’s no way to escape tolls unless you avoid the toll roads. That said, if we build new roads or expand the capacity of existing roads, low income motorists might still benefit from reduced traffic on those routes that do remain toll-free.

What would be a better way to fund this project? Earlier in the campaign, Trump observed that US borrowing costs are low and the government can take on additional debt at very little cost right now:

This is a time to borrow and borrow long term. Normally you would say you want to reduce your debt, and I like to reduce debt as much as anybody. The problem is, you have a military problem, you have an infrastructure problem — a tremendous infrastructure problem — and you have other problems. The asset is your rates are so low.

This is true. US 10 year bonds currently trade at around 2%, which means the amount of interest the government has to pay on new loans is a pittance:

So Trump could just borrow the money. The government would get some of it back–the increased economic growth the infrastructure plan would create would also increase tax revenue by about $200-$300 billion, depending again on the multiplier. Trump could also coordinate with the Federal Reserve to fund the infrastructure spending through quantitative easing–when the Fed prints money to buy back government debt from bondholders, the treasury pays the Fed and the Fed returns its profits to the treasury, effectively allowing the government to print money to pay for infrastructure. When Obama passed his $831 billion stimulus package, quantitative easing helped the government get much of the money back–the Fed paid the treasury more than $500 billion in the six years between 2008 and 2014 alone:

Indeed, for all of these reasons the IMF argues that the United States has plenty of fiscal space and should have at it, arguing explicitly that a lot of new spending on infrastructure is just the thing we need. We can and should freely encourage Trump to borrow more money, especially when he wants to spend it on things we desperately need, like functional bridges:

But even if Trump is unable to get congress to agree to the funding we’d need to do this the right way, it’s better to have a few new roads and bridges you have to pay to use then nothing at all. If he prioritizes this policy, we should back it. It might not be enough to make our country’s infrastructure great again, but it can stop some of it from falling apart, and that’s something.