What Would It Take To Raise $2 Trillion for Infrastructure?

May 10, 2019

Last week, we pointed out how stupefyingly large a $2 trillion infrastructure plan would be – assuming that is a ten-year number, it would almost triple all current federal non-defense capital spending (total public physical capital outlays for non-defense in fiscal 2019 are expected to be $130 billion, but $20 billion of that is equipment purchases, leaving $110 billion for the entire universe of capital spending which could conceivably be infrastructure, including all grants to state and local governments for same).

The upshot of the Trump-Pelosi-Schumer meeting at the White House on April 30 was that they would reconvene at the end of May, at which time President Trump would give proposals for how to pay for a $2 trillion infrastructure spending plan. So we thought we would put that number in perspective from a revenue point of view.

Motor fuel taxes. Suppose you wanted to raise $2 trillion entirely from increasing federal gasoline and diesel excise taxes. How much would you have to increase those taxes to raise $2 trillion over ten years?

Over $1.50 per gallon. Immediately.

(Ed. Note: Obviously, a $2 trillion plan would include many types and modes of infrastructure, and just as obviously, any attempt to use an increased gas/diesel tax to pay for infrastructure unrelated to highways and mass transit would instantly lose the support of the truckers and the motorist lobby, and probably the Chamber of Commerce as well. But it’s useful to put the amount raised by fuel taxes in perspective.)

In January, the Congressional Budget Office estimated that each penny of federal gasoline tax was going to bring in $13.3 billion to the Highway Trust Fund over the next ten years, and every penny of diesel fuel tax was going to bring in an additional $4.2 billion to the HTF. However, any increase in those tax rates would also cause a decrease in federal income tax receipts (every dollar taken out of the economy via an excise tax means that someone, somewhere, has a dollar less in income to declare on their taxes), and this is estimated to cost the general fund between 22 and 25 percent of the amount of the excise tax increase, depending on the year. So while the gross money raised by each additional penny of gasoline and diesel tax would be $17.5 billion over ten years, the net revenue increase to the Treasury would only be $13.4 billion per penny over ten years.

As a result, even if you assume that demand for gas and diesel will be completely, 100 percent inelastic to changes in price (which it would not be, at this level, but more below), you would ned to increase gas and diesel taxes by at least $1.50 per gallon (an 815 percent increase in the gas tax and a 701 percent increase in the diesel tax) to net $2 trillion in increased revenues for the Treasury over ten years.

Estimated Tax Receipts from a $1.50/Gallon Gas/Diesel Tax Increase
(Assumes complete inelasticity of demand, billions of dollars.)
FY20 FY21 FY22 FY23 FY24
Gross Receipts 274 271 268 265 263
Income Tax Offset -60 -59 -59 -59 -59
Net Receipts 214 212 209 207 204
FY25 FY26 FY27 FY28 FY29 10-Year
Gross Receipts 260 257 255 253 251 2,619
Income Tax Offset -58 -64 -63 -63 -62 -607
Net Receipts 201 194 192 190 189 2,013

Economists have long noted that demand for gasoline, as a commodity, is very inelastic – people are going to use what they are going to use despite what the cost is, up to a point. Most Congressional discussion of a gas tax increase involves incremental increases of 5 cents per gallon per year or less, and since the price of gasoline usually varies by at least 5 cents per gallon on a week to week basis with no noticeable effect on demand, it was easier just to assume complete inelasticicty of demand for tax increases of that level than it have been to find an economist to perform elasticity calculations. But you can be certain that if the federal tax on gasoline went from 18.4 cents per gallon to $1.68 per gallon on October 1, the demand for gasoline would indeed show some elasticity, and a lot of people would change their driving habits post haste. So the immediate $1.50 per gallon increase is the absolute lower bound for the increase needed to raise $2 trillion over ten years. The actual amount of the necessary tax increase, counting for depressed demand caused by drastically increased prices, would be a good bit higher.

Tax the incomes of the rich. Looking outside the realm of transportation user taxes, there are plenty of ways for the government to raise money for new projects and programs. Fortunately, the Congressional Budget Office (and the Joint Committee on Taxation, but we’re just saying CBO for short this time) periodically publishes an analysis of how much money would be raised from a variety of revenue increase options, and the most recent version was published just five months ago. The estimates are all from the FY 2019-2028 timeframe instead of the FY 2020-2029 timeframe, but that’s close enough for back-of-the-envelope calculations of ways to get to $2 trillion over ten years.

Using the post-2017 brackets, the two highest brackets (in 2017 dollars) are the 35 percent bracket that starts at $200,000 of regular income for single filers and $400,000 for joint filers. The top bracket (37 percent) is for regular income over $500,000 for individuals and $600,000 for joint filers. These rates revert back to the pre-2017 rates after 2025, under current law.

CBO says that if you increase the regular income tax rate for those top two brackets by one point, you will increase federal revenues by $123 billion over ten years. Using static analysis, you would have to raise those income tax rates on the rich by 16 percentage points of income (from 35 percent to 51 percent for the $200/$400K bracket, and from 37 percent to 53 percent for the $500/$600K bracket) in order to raise $2 trillion in extra tax receipts over ten years.

Tax capital gains. CBO estimates that if you increase the tax rate on long-term capital gains and dividends by 2 percentage points, you would raise $69.6 billion over ten years. Using that math, in order to raise $2 trillion over 10 years, you would have to raise the capital gains/dividend tax rates by 57 percent (by brackets, the lower capital gains bracket would go from the current 15 percent tax rate to a new 72 percent tax rate, and the upper income rate would go from the current 20 percent rate to a 77 percent tax rate).

CBO also adjusts aligning the two top brackets to reflect the capital gains tax brackets, and in this case, the amount of necessary increase of the capital gains tax would be a bit lower.

Eliminate itemized deductions. Now we’re getting to serious money. CBO estimates that if you get rid of all itemized deductions (mortgage interest, health expenses, state/local taxes, charitable contributions, all of them), you would increase tax receipts by $1.3 trillion over ten years. Because of the end-of-2025 expiration of a lot of the 2017 tax cuts, the effects of this option change a good bit before and after that point, but it is still clear that this option would be serious and fundamental tax reform (the kind of tax reform that some in Congress talked about, but chickened out of, during debate on the 2017 bill).

And that still only gets you two-thirds of the way to $2 trillion over ten years.

Tax everyone’s income. The CBO report says that if you increase everyone’s income tax rates – all the way down to the 12 percent bracket that covers incomes of $9,525/yr through $19,050/yr – by one percentage point, you would raise $905 million in extra taxes over a decade. So in order to raise $2 trillion that way, you would need to raise everyone’s rates, from poor to plutocrat, by 2.2 points.


None of this is to say that any one way of raising new revenues is better or worse than any other way – only that $2 trillion over ten years is a tremendous, gigantic, amount of money, particularly when you look at it from the point of view of federal revenues and what it would take to increase those revenues.

For that reason, it seems likely that one of three things will happen:

  1. Nothing will come of this infrastructure discussion at all; or
  2. Any paid-for infrastructure package will mostly consist of Highway Trust Fund reauthorization (which has $400+ billion in excise tax pay-fors over ten years already written into the baseline, allowing more than $3.50 of “new” spending for every dollar of pay-for) and the face value of increased federal loans and loan guarantees, which only need a “pay-for” of less than 10 percent of said loan face value; or
  3. Spending will be increased with few, if any, pay-fors.

Search Eno Transportation Weekly

Latest Issues

Happening on the Hill