19,808,772,381,624.74. No, this is too small to be Avogadro’s number. Rather, it is the current ceiling on U.S. public debt subject to limit. Among the sorrier spectacles in our representative democracy are the dramatic standoffs in which the Congress refuses to increase this limit, constraining the Treasury’s operations and ultimately risking default. It is a hollow exercise because no one wants actually to default and jeopardize the status of the dollar as the world’s reserve currency. It is also a repeated performance, spread across the 74 times the Congress has raised the limit since 1962.
The show is back in town, with Treasury Secretary Steven Mnuchin writing to Congressional leaders of the urgent need “… to increase the nation’s borrowing authority by September 29, 2017.” This is an estimate, of course, since the U.S. government is large and complicated, with unpredictable inflows to and outflows from its coffers. The gold standard for calculating the drop-dead date when the Treasury would no longer have enough cash to fund its operations is provided by the Bipartisan Policy Center. Their current assessment is that this event risk becomes elevated starting October 2nd.1
As with any show that runs through endless sequels, the latest one does not seem to be getting much attention, at least in terms of Google news searches in the U.S. Judging by the location of those searches, the people who are interested are either fearful Inside-the-Beltway types or hopeful Western libertarians. Investors of any type, though, must be mindful because the four-week Treasury bill auctioned next week matures after Mnuchin’s circle on the Congressional calendar. There is a kink in the money-market yield curve, consistent with the intuition that toying with the threat of nonpayment is costly for an issuer. Indeed, researchers have found taxpayers paid for debt-ceiling crises in terms of higher borrowing costs.2
The debt ceiling (1) confounds stocks and flows, (2) is pinned to an unsatisfying concept of the public debt, (3) invites gimmicks, (4) requires a government official to violate the law if it were to bite, but (5) may have no market bite thanks to processing improvements at the Federal Reserve (the “Fed”), the fiscal agent of the Treasury. We take these issues in turn.
Stocks and flows of the same variable (such as the level and issuance of debt) are related arithmetically over time, but you would not know it from the Federal Register. The Constitution enumerates powers of the Congress, including the ability to authorize spending, levy taxes, coin money, and issue debt in the name of the U.S. Since the Second Liberty Bond Act of 1917, the Congress has delegated debt issuance to the Treasury subject to an overall cap on the amount outstanding. The amount of that cap has been raised and its form revised over the years, transforming that $15 billion limit of 1917 to the one binding today that has 14 digits to the left of the decimal point. The chart above plots its recent course, along with public debt subject to limit. Note that the line depicting the ceiling is broken in parts because, on occasion, legislation suspended the limit.
The logical problem is that the Congress also takes spending and revenue actions. If the outflows from the former exceed the inflows from the latter, debt must be issued as a matter of arithmetic. Imposing a restriction on top of this is either redundant or overdetermined (as right now) when the cumulated flows do not add up to the mandated stock.3
The concept of debt subject to limit includes, as in the table, both debt held by the public and intragovernmental holdings, mostly trust funds. Thus, the ceiling is not a measure of net exposure to the private sector. (And the public here includes the Federal Reserve, an agency of the U.S. government owning about $2.5 trillion of U.S. government obligations.) To some extent, the debt the government owes to itself embodies some of its contingent liabilities. But the trust funds do not come close to the total contingent liabilities of the government.
Gimmicks abound around these debt-ceiling events. If the Secretary declares a “debt issuance suspension period,” the Treasury can replace obligations from some trust funds with IOUs that do not count toward the limit, opening headroom for marketable borrowing. These unconventional devices are administratively costly and make the government accounts more opaque.4
Lawbreaking is not typically a legislated instruction, but it is in debt-ceiling events. Fundamentally, there are three governing instructions.
- The Second Liberty Bond Act of 1917 (and its successors) set an overall limit on the public debt
- The 14th Amendment to the U.S. Constitution directs that “…the validity of the public debt of the United States, authorized by law…shall not be questioned;”
- The Federal Reserve Act of 1913 forbids the central bank to lend directly to the U.S. Treasury
The Fed, as the fiscal agent of the Treasury, processes all payments, both for goods and services and for interest and principal, and offers the Treasury a deposit account for its cash.
The drop-dead date for a debt-ceiling crisis occurs on the day that the Treasury does not have enough funds in its Fed account to make all of its payments. The Fed cannot let the Treasury overdraft its account, so it suspends payments until sufficient funds flow in. If those scheduled but suspended payments are of interest or principal, the result is default. On that day, officials have a choice:
- The Secretary of the Treasury could sell marketable debt above the limit, violating the Second Liberty Bond Act;
- The Chair of the Federal Reserve could allow the Treasury to overdraft, violating the Federal Reserve Act; or
- The Secretary of the Treasury could allow default, violating the 14th Amendment by calling into question the validity of the debt.
None of these are good choices, any are grounds for removal with cause, and all would probably lead the credit rating agencies to opine about the status of the U.S.
Processing efficiencies may make it possible for the Fed to make some payments and not others at the Treasury’s instruction. There was a special briefing to the Federal Open Market Committee in 2011 during the debt-ceiling crisis marked by a spike in the first chart. The money paragraph from that discussion5 is:
With respect to the first, the principal on Treasury securities that are maturing would be funded by having auctions that would roll over those maturing securities into new issues, so the new issues would be able to fund the redemption of the maturing securities. With respect to interest payments, the way the Treasury planned to ensure that it would be able to pay interest payments timely [is] by holding back other government payments and accumulating sufficient cash balances in its Fed account to pay upcoming coupon payments. The implication of this approach would be that the Treasury would be delaying non-P&I payments even on days when it may have ample balances in its Fed account to have been able to make those payments if it had so chosen. Instead, the Treasury would be conserving that cash to be able to ensure that it would be able to pay future-dated interest payments.
The ethics and optics of not paying, say, the military or Social Security recipients to conserve cash for debtholders is, to put it mildly, problematic. Of course, this also involves violating other laws—the ones in which the Congress authorized the suspended payments to be made. But, if the Treasury Secretary agrees to prioritization, default can be avoided. His impeachment will not, even if he is led to this as the lesser of multiple evils imposed by his political masters. Despite the histrionics, we have never as a nation hit the drop-dead date of running out of cash while the debt ceiling binds, presumably because the stakes associated with failure are so high. We think that the Congress will use part of its 12 working days in September to raise the ceiling, most likely buried in hard-to-vote-against legislation providing disaster relief and extending the life of theNational Flood Insurance Program.
The sad conclusion is that this is a repeated game of chicken. It is a costly contest, though, both in terms of administrative expenses and borrowing costs. It is a risky one, too, because of the low level of trust among the parties involved. Remember that the prototypical game of chicken was filmed in Rebel Without a Cause. That movie ends with Sal Mineo going over a cliff, his coat sleeve caught in the car door. Mistakes happen even when the stakes are high.
12017 Debt Limit Analysis, August 24, 2017, https://bipartisanpolicy.org/library/2017-debt-limit-analysis/
2 Take It to the Limit: The Debt Ceiling and Treasury Yields, March 2017, https://www.federalreserve.gov/econres/feds/files/2017052pap.pdf
3 The “Gephardt Rule” from 1979 made the redundancy explicit: this parliamentary rule automatically raised the debt ceiling when a budget was passed.
4 Analysis of 2011-2012 Actions Taken and Effect of Delayed Increase on Borrowing Costs, GAO-12-701: Published: Jul 23, 2012. Publicly Released: Jul 23, 2012. http://www.gao.gov/products/GAO-12-701
5 Conference Call of the Federal Open Market Committee on August 1, 2011, https://www.federalreserve.gov/monetarypolicy/files/FOMC20110801confcall.pdf 2018