More Sniping at Sumner (and Cochrane)
[UPDATE below.]
As longtime readers know, I initially loved Scott Sumner but then came to regret my early praise. Yes, he is the smart, consistent, logical, and funny blogger that I identified early on. But unfortunately, he uses his gifts to promote the idea that Bernanke just needs to start writing more checks, and everything will soon be fine.
Anyway, normally I grapple with Scott on technical matters, or on his interpretation of historical events. But today I caught him making (what sure seems to be) a silly error, goaded by John Cochrane:
The same point [that you can’t gauge monetary tightness or looseness from interest rates–RPM] is made in Mishkin’s textbook. And Mishkin is a respected “scientific economist” by anyone’s standards. So why is it that 90% of the respected scientific macroeconomists don’t understand this? Why do most keep insisting that the Fed has conducted an “accommodative” or “easy” money policy since 2008? Maybe they think that doubling the base is easy money, and are unaware that the Fed started paying interest on base money in October 2008. As Cochrane pointed out, this means that reserves are now effectively bonds, not money.
Here’s my question to Scott: If–in some alternative universe, let’s say–commercial banks paid positive interest rates on checking deposits, would that turn $20 bills into bonds, not money?
This is yet another point where, I hate to say it, that Keynesians make a lot more sense than the “free market” inflationists. People like Brad DeLong have been pointing out that when interest rates get pushed to zero, then Treasury bonds effectively become money. But that’s quite a different thing from saying that positive interest rates paid on money kept on deposit, turns money into bonds.
UPDATE:Scott emails me to say:
Thanks Bob, My view is that people can define money in any way they want. But in monetary models of liquidity traps, or anything else, money is a non-interest brearing asset. If they start paying interest on cash, then we need to change the models, but in that case there will be no liquidity traps, because nominal interest rates can go below zero.
I don’t consider checking account balances to be money–but lots of economists do.
Any article you can recommend discussing “monetary tightness” vs interest rates?
Hm, I don’t understand: if banks paid interests on checking deposits, that would make checking account dollars into bonds, not paper money.
By the way, that reminds me of an idea I had: what if people gradually shifted away from using federal reserve notes as money and started using inflation-indexed bearer bonds? That is, they eventually get to the point where the money they use automatically increases in value (with inflation), as measured by USDs. Would that just throw everything into a tail-spin? Make the hoarding “problem” that much worse?
Hm, I don’t understand: if banks paid interests on checking deposits, that would make checking account dollars into bonds, not paper money.
Sorry, I meant my $20 bill once I deposited it into my checking account. In other words, Scott is making it sound as if paying interest on deposits is some wacky thing Bernanke invented in 2008.
But … but … but what about my inflation-indexed bearer bond idea?
I guess I don’t understand either. The paying of interest seems to cause currency to be tied up and not circulating. If a dollar’s velocity is zero, is it really a part of the money supply?
Take the Japanese, for example. Their citizenry have plowed their savings into JGB bonds at miniscule interests for 20 years. They hold literally colossal amounts of these bonds, and yet Japan has suffered from persistant deflation for the entire time. At some point, for demographic reasons, all these bonds will be turned into circulating Yen- then, I would expect that inflation to explode, but not until then.
They started paying interest on base money? Damn the fed owes me interest payments for all the cash I have under my mattress.
Sovereign debt and fiat currency have a lot in common. Neither of them represent claims on capital, neither are backed by assets, both are denominated in terms of the fiat currency. The major differences are:
1. Bonds are harder to spend and earn interest, meaning by Gresham’s law, they have a lower V than currency.
2. Bonds are eventually destroyed by taxation. In a sense they are a temporary monetary injection, thus having a lower impact on inflation than pure currency. Higher future taxation may lead people to save more further reducing the inflationary impact of issuing debt compared to issuing fiat currency.
Two factors have lessened these two effects recently. First, the interest on reserves policy means that the opportunity costs of holding bonds vs. currency are far closer than in the past. Secondly, the government debt is so large that no one could reasonably expect it to be paid via taxation. Bonds will not be destroyed by taxing people to pay them off. They will either be continually rolled over, or monetized, which will lead to people thinking of bonds less as temporary and more as permanent creations.
Paying interest on fiat currency is not like a bank paying interest on an account. The banks are not lending money to the Fed in exchange for interest, they are simply collecting a check. Excess reserves do not represent a claim on real physical capital (just like government bonds). You’ve consistently said M0 is money, everything else isn’t, but many economists don’t think that way. Post-Keynesians don’t even put much effort into distinguishing currency from bonds, instead emphasizing net financial positions. Money is what people treat as money.