This Neil Irwin post is pretty good (I think I got this from Alex Tabarrok on Twitter):
[I]n the…market for Treasury bills, things are starting to get scary. These are short-term IOU’s of the U.S. government, bills issued for 30, 60 or 90 days. They enable Uncle Sam to manage cash flow much the way a homeowner might use a credit card. They also form the backbone of trillions of dollars in transactions: Major corporations and banks use them as a place to park short-term cash; they are held by money market mutual funds; and they serve as collateral for millions of transactions in markets around the world.
…[O]n Sept. 30, eight days ago, the interest rate on Treasury bills maturing Oct. 17 was a mere 0.03 percent. Nothing, in other words.
But since then, the possibility that the Treasury might have trouble paying or might not be able to pay its bills over the next few weeks has grown — and the interest rate has skyrocketed. It was at 0.16 percent at Monday’s close. On Tuesday the rate so far has been almost double that, as high as 0.297 percent.
Ironically, this can create self-fulfilling problems for the Treasury. Treasury bills roll over every week, on Thursdays. Here’s how it works: The government issues the bills for a “discount,” then refunds the par value when they come due. So, for example, an investor might pay $995 for a bill that returns $1,000 in three months, for an equivalent of about a 2 percent annual interest rate.
But if buyers don’t want to roll over their bills — if they don’t trust the government enough to pay the usual high price for that debt — then the government’s cash crunch becomes even more severe. If, for example, investors were only willing to pay $950 for a 90-day, $1,000 bill (about a 20 percent annualized interest rate), then the government would run into its legal debt limit even faster than it is now scheduled to. We’re nowhere near that point now. The rate has risen to 0.3 percent, not to single-digit, let alone double-digit, percent levels. But the abruptness of the move, in a market that usually is rock-solid and stable, is startling nonetheless.
In the 2008, post-Lehman Bros. crisis, major banks — even those that seemed to be fundamentally solvent — were facing a liquidity crisis, as short-term access to cash became a challenge. Buyers of Treasury bills have no evident concerns about the longer-term solvency of the United States. But the action on the bill market over the last several days, and especially Tuesday, looks like the early phase of a liquidity crisis.
So I’ll repeat my question for Keynesians: Why isn’t this at least a partially good thing, since it clearly gets us out of the liquidity trap? If the above trend continues, then clearly short-term nominal interest rates are no longer stuck against the zero lower bound.
In fact, why isn’t Krugman arguing that by definition, we can’t have a crisis unless interest rates rise, and yet if interest rates rise the Fed can gobble up bonds and push them back down? The only reason the Fed wouldn’t want to do that, is if (price) inflation starts rising, which will only happen once we’ve restored Aggregate Demand.
I have asked this three times now, and have yet to see a satisfactory explanation. Sure, the federal government will slash spending, but the Fed can counteract that by more expansionary policy. The only time this can’t happen–according to Krugman–is when we’re at the zero lower bound, and the central bank can’t offset fiscal austerity.
So I ask again: If we take the framework Krugman has developed over the last five years on his blog, how can a US government debt default possibly hurt the US economy?