Will the Trigger Make Only a Small Dent in the Debt?

(Or Why You Should Pay for the Tax Cuts in the First Place)

There has been much talk in recent days about a new trigger that is apparently being added to the Senate’s tax legislation and that is meant to serve as a backstop in case economic growth doesn’t pick up to offset the costs of the tax cuts. Last night and after Senator Corker agreed to vote the tax legislation out of the Budget Committee having apparently reached a deal on the trigger, CQ reported a few details of a trigger that was apparently circulating on the Hill yesterday. (As an aside: With a vote only a few days away, we deserve better than a report in CQ based on unnamed sources about an important element of what this bill does.)

According to that CQ report, the trigger would go off after 5 years — based on whether revenues hit a desired target in 2022. The trigger, which reportedly would impose additional taxes on corporate income, would then raise up to a maximum of $350 billion over the ten years starting at that point, according to CQ. So, the maximum of $350 billion would apparently be raised from 2023–2032.

There are a lot of things that could be said about triggers like this, including how credible they are or aren’t, interactions with business cycles and possible pro-cyclicality, the timing of the trigger (sooner would be better if you use one), and so on. These are all important issues.

But, there’s one overriding fact about this trigger if the CQ report is accurate: it’s just too small and would offset only a small share of the expected costs of these tax cuts even if it were allowed to work exactly as envisioned.

Through 2027 (the normal ten-year window), the trigger would apparently offset only in the range of 10 percent of the amount this bill is reasonably expected to add to the debt. Through 2032, the effect would be somewhat larger. Maybe an offset in the range of 25 percent or so.

To state the obvious: 10 or 25 percent is not 100 percent. The trigger — even if implemented as designed — simply isn’t that much of a back-stop.

How do I arrive at these numbers? I take figures from the Joint Committee on Taxation and the Congressional Budget Office to arrive at the amount that the Senate bill is expected to add to the debt through 2027 under “static” scoring — not assuming any effect on the size of the economy. I then use an average of the “high” and low” estimates from the Penn-Wharton Budget Model of the Senate bill’s effects on economic growth and the dynamic revenue (and reduction in debt) generated by that. And, I then compare this to the savings a trigger would generate assuming it raise $35 billion per year starting in 2023. The figures extending out beyond the ten-year window (from 2028–233) require some rough approximations. Additional details on these calculations are provided in the addenda.

Now, some trigger proponents might say that this is comparing to too high a cost estimate. Perhaps the trigger is only intended to offset the cost of these tax cuts relative to a world where current temporary business tax breaks are made permanent (a “current policy baseline”). That is wrong-headed thinking — unless we also counted all of the new temporary tax breaks in this bill as permanent and added them to the cost of the legislation. But, in any case, even using this gimmicked arithmetic, the trigger still can’t offset the expected cost of the legislation.

In the end, there is no substitute for actually paying for these tax cuts using the estimates from JCT and CBO and based on the best available economic evidence, and this version of the trigger doesn’t come close to being an effective backstop.

Addendum: Trigger in Numbers

The table below shows the effect of the Senate bill on debt through 2027 and 2032. It separates out several elements: the cost of the tax bill, before interest; the interest cost (using CBO-based interest rates); the dynamic effect of the tax cuts — from higher GDP — on the federal debt based on the Penn-Wharton Budget Model; and finally, the effect of the trigger including the $35 billion per year in revenue and related interest savings.

Addendum: Projecting Through 2032

The trigger as described by CQ apparently lasts only through 2032, at which point it shuts off. So, I’ve also projected the effect on the debt through that year. This projection is very rough.

I assume that the $30 billion in net deficit reduction under the Senate bill as of 2027 continues for the next five years. There are many potentially offsetting factors there; some of the raisers in 2027 are largely revenue shifts (like amortizing research costs) while other raisers like the chained CPI really are growing even as costs grow. I have not done a detailed analysis and the exact path wouldn’t have much of an effect on the conclusion.

I extend CBO’s interest rate assumptions for the next five years.

Finally, and probably most roughly, I calculate a dynamic effect on the debt — based on an average from the Penn Wharton Budget Model — based on figures given by PWBM for 2027 and 2040. I assume the dynamic effect linearly rises between those years to calculate a figure for the end of 2032. That’s probably somewhat optimistic in terms of the pattern of additional revenue, and front-loads the dynamic effect relative to the actual path.