The Washington PostDemocracy Dies in Darkness

The math in Trump’s infrastructure plan is off by 98 percent, UPenn economists say

The Holland Tunnel in Jersey City. (Julio Cortez/AP)

President Trump is overselling the financial impact of his proposed $1.5 trillion infrastructure plan by about $1.3 trillion, according to economists at the University of Pennsylvania’s Wharton School.

Earlier this month, the White House unveiled a proposal that it said would stimulate $1.5 trillion in new infrastructure spending through a $200 billion federal investment. The White House said the other $1.3 trillion would come from new state, local and private spending unleashed by its spending plan.

But the Penn Wharton Budget Model team found that the new federal investment would lead at most to an additional $30 billion in state, local and private spending, or about 2 percent of the amount envisioned by the White House.

“We really tried to be generous here,” said Kent Smetters, faculty director of the Penn Wharton Budget Model. “But what the literature says is that states will figure out how to qualify for these grants without changing their existing behavior.”

Trump’s big infrastructure plan has a lot of detail on everything but how to pay for it

The White House derided the Penn Wharton study, arguing that it failed to account for global investments in U.S. infrastructure and included unrealistically low projections of the impact of new federal incentives for state investments.

“The author of the study clearly failed to read the President’s framework,” a White House spokeswoman said in an email. “Notwithstanding modeling challenges, they got the policy flat wrong. Furthermore, they view the U.S. economy as closed and assume there is no growth with this investment, even though it is abundantly clear that the economy is booming under President Trump’s pro-jobs, pro-growth agenda.”

Trump officials have proposed a variety of changes to America's infrastructure system — including new environmental permitting rules and streamlining other federal regulations — that they say will lead to a big influx of spending on roads, bridges, tunnels and other projects.

But both liberal and conservative economists have panned the White House’s $1.5 trillion number as based on rosy assumptions unlikely to come to fruition.

“I think it’s basically spot-on,” Michael Sargent, a policy analyst on infrastructure at the conservative Heritage Foundation, said of the Penn-Wharton study. “When you’re looking at total spending, it seems about right.”

Douglas Holtz-Eakin, who served as president George W. Bush's chief economist and remains influential in Republican circles, said he thought Penn-Wharton’s $230 billion number was probably on the low side. But he agreed that the model was essentially fair and that the White House’s $1.5 trillion number is way too high.

“I think they’re making a fair point: How the states respond matters, and there’s no guarantee that they’ll pony up an additional $1.3 trillion,” said Holtz-Eakin, who is president of the American Action Forum, a conservative advocacy organization. “The major problem with the plan is selling it as $1.5 trillion. If they sold it as, ‘We can identify some really good infrastructure projects without giving a number,’ everyone would look at it and say, ‘Yeah, that makes sense.’ ”

The White House proposes that $100 billion would go to competitive projects, administered by the federal government, with awards of up to 20 percent of their costs. Under the proposal, states and localities would have to find new sources of revenue — the White House has suggested property or sales taxes — to qualify for the federal help. White House officials have also floated raising money for the plan by increasing tolls or other user fees.

The other $100 billion in the plan is split into $50 billion for rural infrastructure programs, $20 billion for “transformative” projects like tunnels for high-speed rails, and $30 billion to expand loan programs.

Trump has called the plan necessary to replace and repair America’s roads and bridges. During the presidential campaign, he called for increasing infrastructure spending by $1 trillion, more than double the amount proposed by Hillary Clinton, the Democratic nominee.

During his State of the Union address to Congress, Trump said: “Together, we can reclaim our building heritage. We will build gleaming new roads, bridges, highways, railways and waterways across our land. And we will do it with American heart, American hands and American grit.”

Trump has struggled turning an infrastructure plan into a reality that matches the criteria he laid out in his campaign. Republicans, who control Congress, have been resistant to ratchet up federal spending, particularly after last year’s Republican tax law slashed rates on corporations and individuals. (Tax receipts are already projected to be $200 billion lower in 2027 than forecast last year, according to the White House.)

Mixed signals on infrastructure plan emerge from Trump retreat

But banking on states to come up with the rest of the cash is also an imperfect solution, economists said. Part of the problem is that even if they produce the money to get the new federal grants, they are likely to offset that increase by reducing infrastructure spending elsewhere.

“I’m not surprised at the conclusion that very little additional spending will be called forth by this plan,” said Douglas Elmendorf, who served as director of the Congressional Budget Office under President Barack Obama, adding that he had not personally reviewed the Penn-Wharton model. “State and local governments are constrained in their funds more than the federal government, because they generally have to balance their books on an annual basis.”

Wharton’s projections were based on more than two dozen academic studies on federal infrastructure spending, according to Smetters, the Wharton economist. One of the more optimistic academic papers was a 1974 study that found that states increase their total spending by $1.06 for every federal dollar received. But subsequent studies found much lower effects, Smetters said.