The debt deal that changes almost nothing

The 27 May agreement suspends the federal debt ceiling through 2025, and its authors have hailed it as a triumph of bipartisan restraint. Budget summits in Washington tend to be exercises in cosmetic accounting, and this one is no exception. Strip out the rhetoric and the spending deal barely touches the path of the structural deficit. Caps on discretionary spending look reassuring, while mandatory entitlements and net interest, the fastest-growing items, are left alone.

What the caps actually cover

The heart of the deal is a limit on the spending Congress votes on each year. Non-defence discretionary outlays are held roughly flat in the fiscal year that begins in October and allowed to rise by 1 per cent in the year after, while defence spending keeps growing. Beyond those two years the limits turn into targets with no enforcement behind them. The package also claws back some unspent pandemic relief money and trims part of the extra funding the Internal Revenue Service received last year, a cut that may well cost more in lost tax receipts than it saves. Social Security and Medicare, which together dwarf everything the caps touch, go unmentioned.

Much of the rest of the package is policy rather than money. It ends the pause on federal student-loan repayments this summer, widens work requirements for some recipients of food aid, and speeds up permitting for energy projects. Those provisions may matter to the people and industries affected, but none of them moves the deficit by much, and several were included to let each side claim a win.

Interest eats the savings

Interest costs on federal debt are rising along with policy rates, and within three quarters interest expense alone will exceed the projected savings. Freezing non-defence discretionary spending does nothing about the basic problem: the United States is running a wartime deficit at a time of peacetime full employment. Bond investors can see that the debt path has not been flattened. Congress has simply protected it from challenge for another eighteen months.

The repricing works fast. Treasury bills now yield more than 5 per cent, against almost nothing in early 2022, and a bill rolls over within months. Longer notes and bonds turn over more slowly, but each issue sold at pandemic-era coupons is being refinanced at several times its old rate as it matures. The deficit came to about $1.4 trillion in the last fiscal year, while the unemployment rate stood at 3.4 per cent in April, close to its lowest since 1969. Deficits of that size normally belong to recessions or wars, not to an economy with more job openings than unemployed workers.

The budget office's February projections showed net interest costs more than doubling over the coming decade. On that path interest becomes one of the largest items in the federal budget, and unlike the programmes the caps restrain, it cannot be trimmed in an appropriations bill. Higher yields raise the deficit, a larger deficit means more issuance, and more issuance puts upward pressure on yields.

The 2011 caps went the same way

Washington has run this experiment before. The Budget Control Act of 2011, also the product of a debt-ceiling stand-off, imposed a decade of discretionary caps backed by automatic cuts. Congress then raised those caps in bipartisan deals in 2013, 2015, 2018 and 2019, each time because the limits bit harder than either party would tolerate. The 2011 fight also cost the United States its top rating from Standard & Poor's. Fitch put its own AAA rating on watch for a possible downgrade on 24 May, and two years of caps followed by non-binding targets look unlikely to persuade it that the trend has turned.

The Congressional Budget Office's projections in February already had federal debt held by the public overtaking its post-war record before the end of the decade. Nothing in the deal alters that direction, because the lines driving it are the ones the negotiators left alone. Whatever the final score of the savings, they trim the level of borrowing without bending its slope.

When caps did work

There is one episode in which spending limits helped. The Budget Enforcement Act of 1990, agreed by George H.W. Bush at the cost of breaking his pledge not to raise taxes, combined caps on discretionary spending with a pay-as-you-go rule that required new tax cuts or entitlement expansions to be paid for. A further package of tax increases and spending restraint followed in 1993, and with a long economic boom lifting revenues, the budget moved into surplus by the end of the decade.

The lesson is that caps worked when they were one part of a package that also touched revenue and entitlements, and when strong growth did much of the rest. The 27 May agreement has only the first ingredient. It leaves taxes untouched and does nothing about the retirement programmes whose costs rise with an ageing population; the trustees of Social Security reported in March that its main retirement trust fund will be exhausted by 2033.

What the optimists say

Defenders of the agreement make a fair point. Holding discretionary spending flat in cash terms while prices rise amounts to a real cut, and if Congress honours the targets after the binding years expire, the savings would compound over a decade. Averting default also spared the Treasury market a far more expensive shock. The objection is one of scale: the capped programmes are the smaller and slower-growing part of the budget, and the later targets that would do the heavy lifting bind nobody.

The first test comes with the appropriations bills due before the fiscal year starts on 1 October, where the new ceilings meet the usual fights over individual programmes. The Treasury's quarterly refunding announcement at the start of August will show how much more longer-dated debt it plans to sell, and the budget office's updated projections will show whether net interest keeps outrunning the savings. A downgrade from Fitch would not change who buys Treasuries overnight, but it would put an official label on the doubt.

Washington has pushed any serious fiscal reform past the next presidential election. Until then the Treasury will keep selling debt with no real limit on supply, and bond investors will probably demand a steadily higher term premium to buy it.