Would a Downgrade be Expansionary?

 As anyone who's been following the news knows, S&P is threatening to downgrade the US government if Congress doesn't come up with a budget plan that they like.  Even though S&P is full of crap generally, and especially when it comes to rating sovereign debt, Nick Rowe has an interesting post about what happens if the US gets downgraded in a liquidity trap (when short-term debt pays 0 interest, essentially making it a cash equivalent) and markets believe the ratings agencies.  Rowe suggests that the standard IS-LM model tells us that a 1% annual chance of default (as represented by a downgrade) would be the same as generating 1% more expected inflation (the analysis is in Rowe's post, which I linked).  And that would actually be good for the economy in the short run, as people holding cash who thought the purchasing power of their cash would erode in the future would reinvest it into higher-yielding assets or spend it on consumer goods, which in turn would stimulate the economy by getting cash-rich corporations spending.  So Rowe suggests that the IS-LM model, given these assumptions, says that a default would actually help get us out of recession in the short run.

I've been thinking about this all day, and I think I agree with Paul Krugman on this one.  The assumption where the analysis begins to fall apart is that the Fed will be able to keep interest rates at 0 in this situation.  Right now, they can keep interest rates where they are simply because investors don't see any default risk in US Treasuries.  If they believe a ratings downgrade, they will always take cash in exchange for Treasuries that have that default risk, since cash doesn't have any "default risk".  It might carry inflation risk, but Treasuries carry the same inflation risk, since they're denominated in the same dollars as cash.  So IF the market were to believe the downgrade, the Fed would literally have to buy all the Treasuries in order to keep the rate on them at 0.  So, where does that leave us? Well, as Krugman notes, there's still a rate at which investors would be willing to buy Treasuries in exchange for cash, and that rate is higher than it was pre-downgrade, and that higher rate is also reflected in longer-term debt, which means that borrowing costs rise across the economy.  since plenty of other borrowing costs are still linked to Treasuries (Fannie/Freddie borrowing rates being one, so the housing market would be impacted).

The crucial detail here though, I think, is that, even though the real value of Treasuries is eroding, the purchasing power of the dollar isn't (because, as established in the last paragraph, the Fed can no longer turn Treasuries into a cash equivalent without buying ALL of the short-term Treasuries outstanding).  So there isn't the kind of pressure to reinvest dollars in higher-yielding assets the way there is if expected inflation rises.  I think all that would really happen is a flight to cash instead of Treasuries, and some higher borrowing costs throughout the economy that would actually choke off recovery even more.  Especially in the long run, when the cost of servicing the debt is higher than it would be without the downgrade.

So I think my conclusion is that Rowe's stylized scenario falls apart in two places.  First, experience tells us investors don't care about what S&P has to say (they downgraded Japan in 2002 and nothing happened, though, as I wrote before, US debt may be different), but second, it's virtually impossible for the Fed to hold interest rates on government bonds constant in this scenario without buying ALL outstanding bonds.  So the scenario is effectively impossible.