The Debt Limit is Not a Spending Limit
Reviewing a Fed-initiated alternative to the platinum coin.
Few aspects of American fiscal politics are as frustrating as perennial negotiations over the debt ceiling. Currently, the federal government is creeping up on its $36.1 trillion ceiling, with negotiations to raise it already underway. To understand why these debates are so absurd, recall a few salient facts:
As part of the federal budget, Congress orders the Treasury (via the president) to spend a particular sum of money. Call this amount S.
Congress also determines tax rates, which results in businesses and households paying a certain amount of money to the Treasury. Call this amount T.
In most cases, S > T, meaning that tax revenue is insufficient to pay for government spending.1 To bridge the gap, Congress authorizes the Treasury to issue federal debt.
But how much debt? In any reasonable world, Congress would allow the Treasury to issue as much debt as necessary to execute mandated spending. Unfortunately, we don’t live in a reasonable world — Congress sets a hard limit on government borrowing, known as the debt ceiling.2
Call the remaining limit on the debt ceiling D. Clearly, if D < S - T, problems arise. This contradictory condition would mean that the Treasury is barred from issuing a sufficient volume of debt to execute payments that it is required to make. The most significant danger in such a crisis is that the Treasury could fail to make interest payments on the national debt, triggering default and inviting financial catastrophe.
The potential for a debt ceiling crisis is seen as forcing the president into an impossible trilemma. Unless the debt limit is raised, the president’s only options are to unilaterally cut spending, raise taxes, or order the Treasury to ignore the debt ceiling — all three of which would directly violate Congressional law.3 But while this analysis is compelling, it is also somewhat incomplete.
Implicitly, the trilemma analysis assumes that taxes and debt are necessary to pay for government spending. Although this assumption has some roots in reality, a detailed look at the mechanics of federal payments shows that it doesn’t hold up to closer scrutiny. Because the federal government can always satisfy its payment obligations, the debt limit is not a spending limit.
A Brief Primer on Government Payments
To understand how government payments work, we need to start with the source of nearly every single federal disbursement — the Treasury General Account.
The TGA is the government’s bank account, held on deposit with the Federal Reserve. In many respects, the TGA functions exactly like a normal bank account. Just as you might own a checking account at JPMorgan, the Treasury owns a checking account at the Fed.
When the government pays for stuff, such as Social Security transfers or interest on the federal debt, the balance of the TGA decreases. When the government earns money, such as through tax receipts or proceeds from debt sales, the balance of the TGA increases. At its height, the TGA hit a balance of more than $1.8 trillion in the run-up to the Covid fiscal stimulus, and currently sits around $576 billion.
When we talk about a debt ceiling crisis, what we’re really talking about is a TGA crisis. The TGA is drained by government expenses and is frequently topped up with ongoing debt sales (plus whatever amount is collected in taxes). But if money stops flowing in, there’s a risk that the TGA could drop to zero, meaning the government can no longer fulfill its obligations.
At least, that’s the common view. But it’s not mine. To understand why, we need to look at the actual three-step process that occurs when the federal government executes a payment. As an example, consider what happens if the Treasury sends the Fed instructions to pay $10,000 to a contractor who banks at Citi:
First, the Fed reduces the value of the TGA by $10,000. The TGA is simply an electronic entry on the Fed’s computer, so it can be marked up or down.
Then, the Fed increases the value of Citi’s account by $10,000. Like the Treasury, commercial banks have direct accounts with the Fed.
Finally, Citi increases the value of the contractor’s checking account by $10,000. Once again, the contractor’s account is an electronic entry, so Citi can just mark it up.
When the system is running smoothly, we can basically think of a bag of money coming from the Treasury’s account, being handed off to a commercial bank, and then eventually being given to its ultimate recipient. But as the steps above show, this isn’t literally what happens. No money changes hands — only computer entries are adjusted.
For the debt ceiling, this distinction really matters. In particular, there is no necessary logical link between Steps 1 and 2 above. Technically, the Fed can mark up the value of Citi’s account without marking down the TGA. As a current operational practice, federal payments may reduce the TGA, but this isn’t an inviolable law of nature.
In fact, the concept of marking up a bank account without a corresponding decrease in an account elsewhere isn’t as esoteric as one might think. For instance, when Citi pays an employee their salary, they can do so by directly crediting the value of the employee’s Citi checking account. Just as in the Fed-TGA situation, the money for Citi’s salary payments doesn’t need to ‘come’ from anywhere — it is created out of thin air.
This concept is why I argue that a debt limit is not inherently a spending limit. Debt issuance is only necessary to refill the TGA if one assumes that spending drains the TGA in the first place. When push comes to shove, the Fed always has the power to execute the Treasury’s payment requests. In the contractor example above, the Citi credit is the payment — the TGA debit is nothing more than bookkeeping.
The Least Illegal Option
If the Fed decided to simply stop charging the TGA for corresponding payments to commercial bank accounts, there’s no doubt that it would invite a host of legal and political attacks. But as the president’s trilemma shows, the only options in a debt ceiling crisis are unpleasant ones. The aim is to find the least illegal way to prevent a catastrophe.
To be sure, there is at least one option that appears to be a perfectly legal way to avoid a debt ceiling crisis: minting an ultra-high-denomination platinum coin. For various reasons, however, this is a Treasury-initiated action, meaning that it requires a president who wants to keep federal payments flowing.4 This might have been a fine assumption for previous administrations, but it is no longer the case under Trump.
From this perspective, the Fed’s ability to stop draining the TGA provides a meaningful option to sidestep the debt ceiling that can be executed without relying on the executive branch. And while it might seem like an outlandish proposal, the Fed’s responsibility to pursue financial stability requires it to consider even extraordinary measures to prevent a US government default, which could easily be the most destabilizing event in financial history. Combined with the fact that this maneuver avoids the Treasury’s Congressional trilemma, pausing TGA reductions might be one of the least illegal ways to avoid a debt ceiling crisis.
Historically, most debt ceiling proposals related to the TGA have focused on the Fed allowing the Treasury to go into overdraft, permitting a negative balance. However, this idea is explicitly barred by the Fed’s prohibition on lending to the Treasury directly.5 What’s different about the above proposal is that there is no expectation, requirement, or even possibility of the Treasury ‘paying back’ the executed transfers. Thus, it’s hard to see how it would qualify as a form of lending.
In reality, one of the biggest hurdles would likely be from an accounting perspective. By crediting commercial bank accounts without a corresponding debit elsewhere in the system, the Fed would be booking an increase in liabilities, reducing the central bank’s net worth. But given the Fed’s penchant for creative accounting — including booking operating losses as a deferred asset — this hurdle is far from insurmountable.
Conclusion: Useful Chaos
Pausing TGA reductions is just one of the possible Fed-initiated mechanisms to overcome the debt ceiling. Elsewhere, JW Mason has discussed the possibility of the Fed prepaying remittances to the Treasury to bolster the TGA, an idea with origins in a 1969 internal Fed memo. And while techniques like this are often derided as accounting gimmicks, the debt ceiling itself is an accounting gimmick with no place in America’s fiscal policy.
Unfortunately, despite the fact that the debt ceiling itself is almost certainly unconstitutional, it’s probably here to stay.6 The chaos fomented by regular run-ups to the debt limit is politically useful to policymakers on both sides of the aisle, leading to little genuine motivation to repeal it. And as long as gimmicky problems remain, gimmicky solutions must be considered.
Of course, taxes aren’t really revenue, nor do they pay for government spending, but Congress doesn’t think about things that way.
The specific wording of the debt ceiling applies the limit to all government-guaranteed debt. Curiously, however, Federal Reserve liabilities such as reserves and term deposits have historically not counted toward the limit. Under a textualist reading, it’s not obvious to me that this should be the case.
Interestingly, through his use of tariffs by executive order, Trump may have inadvertently found a workaround to this trilemma.
An obscure law allows the Treasury Secretary to order the US Mint to create a platinum coin in whatever denomination the Secretary dictates. Like any other coin, it can be deposited at the Fed at face value, with the Fed crediting the TGA for the amount.
Per JP Koning, this hasn’t always been the case.
As should go without saying, nothing in this article is legal advice.