The debt ceiling and glitchy word-processors do not mix well
Once upon a time, Congress didn't want to raise the debt ceiling, and sent the country into default. It was bad, and we shouldn't do it again. The end.
Oh, you wanted to hear the rest of the story? Okay, here it is. Back in 1979, Congress waited, and waited, and waited to lift the debt ceiling, because Congress never likes taking responsibility for the tax and spending decisions it's already made. Now, Congress usually does the right thing after it's exhausted every other possibility, at least when it comes to paying our bills, and this debt limit increase was no exception. Congress did raise it right before defaulting on our obligations would have been unavoidable ... but that didn't let us avoid defaulting on our debt. At least not $120 million or so of it. That's because the logistically and technologically-challenged Treasury couldn't get the checks out in time on such short notice. As Donald Marron of the Tax Policy Center explains, the Treasury got swamped with an inordinately high demand for Treasury bills, which it couldn't meet due to a word-processing error. So we defaulted on some of them.
The government eventually paid back everyone what it owed with interest, but that didn't erase this accidental default from the market's memory. Short-term interest rates shot up 0.6 percentage points after the Treasury missed its payments, as you can see in the chart below from Terry Zivney and Richard Marcus, and remained higher than they otherwise would have been for a few months at least. It wasn't the end of the world, but it wasn't exactly something we want to repeat either.
But it would be much, much closer to Armageddon if we did this today. Blame shadow banking. The past few decades, there's been a shift in the financial system towards things that, economically-speaking, look like a bank, act like a bank, but technically aren't banks. Institutions like hedge funds, structured investment vehicles, and money-market funds all borrow short and lend long, just like a bank, but do so outside the web of regulations that control, and safeguard, regular, old banks. In other words, they trade FDIC insurance and access to the Fed window for complete financial freedom.
They are the shadow banking system, and they didn't really exist back in 1979. At least not anywhere near today's scale. And they, along with conventional banks that have gotten into the game, use Treasury bonds as a kind of money. They use Treasuries as collateral for cash in repurchase agreements (repos) to fund their daily trading, with those same Treasuries often getting "rehypothetecated" -- that is, reposted as collateral by whoever first got it as collateral -- in a dizzying chain of financial connections. It's almost impossible to predict what would happen to these collateral chains if there was any kind of default on Treasuries, but it would almost certainly be 1) bad, and 2) very bad. Think about it this way. Treasuries are supposed to be the safest of safe assets, and as such are the lifeblood of the financial system, which has been running low on safe assets since mortgage bonds turned out not to be so. Removing the system's blood is not something we want to try. The last time something like that happened was, of course, back in 2007-08 with subprime bonds, and it set off and old-fashioned bank run on the uninsured assets of the shadow banking system that nearly brought down the world economy.
There's one last lesson from 1979. It might seem funny, at least in retrospect, that we defaulted on our debts in part due to a word-processor glitch, but it's not clear things would be any better now. As Brad Plumer of the Washington Post points out, it would be a technical, and not just legal, nightmare for the Treasury to prioritize payments if the debt limit isn't increased and we have to immediately cut 40 percent of spending. The Treasury's computers are set to pay bills as they come in -- because why would they do anything else? -- and it's no sure thing it could reconfigure them without a costly hitch. In other words, we could default this time because of a faulty program instead of a faulty word-processor.
This story has three possible endings. Either Congress lifts the debt ceiling, and we all live happily ever after, at least until the upcoming battle over the continuing resolutions; or Congress doesn't, and we only get a massive dose of austerity; or Congress doesn't, we get a massive dose of austerity, and then we default, with interest rates shooting up and the financial system melting down. This shouldn't be a difficult choose-your-own-fiscal-adventure.