From TheMoneyIllusion to Sweden to the NYT

In a couple articles published in early 2009 (in The Economists’ Voice) I suggested a negative interest rate on bank reserves.  By mid-year The Swedish Riksbank had picked up on the idea (although there were loopholes in its implementation.)  Now the New York Times is on board.

For starters, the Fed could take modest steps, like shifting its portfolio toward bonds with longer maturities, which would help to keep long-term rates low and nudge investors into riskier investments. It could reduce the interest it pays on the banks’ huge reserves or even tax the reserves to try to encourage more lending. It could also resume buying Treasuries or other securities to provide additional monetary stimulus. A more aggressive strategy would be letting inflation rise above the Fed’s comfort level of 2 percent or so to, say, 4 percent. That could help the economy by easing the repayment of debt.

My only criticism is that the purpose of negative IOR is not to encourage bank lending.  After the sub-prime fiasco the last thing the government should be trying to do is encouraging bank lending.  The purpose is to reduce the demand for the medium of account.

HT:  JimP



27 Responses to “From TheMoneyIllusion to Sweden to the NYT”

  1. Gravatar of NotGregMankiw NotGregMankiw
    10. August 2011 at 18:33

    What about Greg Mankiw’s NYTimes editorial in early ’09?

    It May Be Time for the Fed to Go Negative

  2. Gravatar of tim tim
    10. August 2011 at 18:50

    Dear Scott
    Quick question, do you believe that the Bank of England’s monetary policy has also been too tight?

  3. Gravatar of Scott Sumner Scott Sumner
    10. August 2011 at 19:04

    NotGregMankiw, Yes, Mankiw did a piece as well (after the two articles I wrote, and some blog posts.) His proposal was somewhat different–more like Gesell’s idea during the Depression.

    Tim, Probably, but I am much less confident in making that judgment than I am for the US. My hunch is that a larger share of Britain’s problem is structural, compared to the US.

  4. Gravatar of John John
    10. August 2011 at 19:38

    What’s the difference in terms of results between encouraging lending and reducing demand for the medium of account. It seems like since the Fed works through the banking system, it is a bank afterall, that the results would be the same. Is there some place banks can put money besides as a deposit or as a loan?

  5. Gravatar of Joe Joe
    10. August 2011 at 20:30

    But how do you know that the recession is caused by an excess demand for the monetary base and thus reserves, as opposed to an excess demand for M1 and M2? Here are some criticism at Canuck’s Anonymous, who argues its neither. He has debated this issue back and fourth with Andy Harless.

  6. Gravatar of Charlie Charlie
    10. August 2011 at 21:05

    “The purpose is to reduce the demand for the medium of account.”

    Can you link up some explanations of this. I was just earlier today (before I read this) looking at my Recursive Macro Theory book by Sargent and Ljundquist and admiring what little useful monetary economics is in it. And this is one of the many things not covered.

  7. Gravatar of Donald Pretari Donald Pretari
    10. August 2011 at 21:32

    I’m going to post a link to one of those posts since it was one of the first times I commented on your blog. It still seems an interesting idea. I was influenced by Fisher, but wasn’t sure how it would work:

  8. Gravatar of FT Alphaville » Further reading FT Alphaville » Further reading
    10. August 2011 at 23:03

    […] – More cheers for negative rates. […]

  9. Gravatar of RebelEconomist RebelEconomist
    11. August 2011 at 00:33


    Why do you worry about demand for the medium of account/exchange/whatever? As long as the Fed supplies the demand for money for all purposes, there is no disruptive EXCESS demand. I have been asking a similar question of David Beckworth – it seems to me that he is wrong to see falling money velocity as evidence of excess money demand, as falling velocity is exactly what I would expect if the Fed fully met the demand for a (inherently low turnover) store of value.

    Regarding the duration of the SOMA portfolio, I keep wondering when someone (other than me) is going to wonder what contribution this made to the financial crisis. Despite regarding low long term interest rates as a “conundrum” associated with a “saving glut” from foreign exchange reserves inflows, the Fed itself was almost certainly the largest central bank holder of long-term treasuries until about 2007. If the Fed had considered low long-term interest rates to be a problem, it could have shortened the duration of the SOMA portfolio, as I argued at the time:

    I also have an off-topic, but I think vital, question that I would be grateful if you would address when you get on top of your blog comments. You emphasise the “hot potato” effect of money. Is there any mechanistic or theoretical explanation (eg microfoundations) of this effect? Given that money is always supplied by exchanging it for some asset, there has to be some value added in the process, otherwise it would seem that the real bills doctrine is correct.


  10. Gravatar of John John
    11. August 2011 at 03:04


    I like how you concede the Austrian point that encouraging lending or credit expansion leads to disastrous results. You may become a real economist someday (just kidding).

  11. Gravatar of JTapp JTapp
    11. August 2011 at 05:53

    Justin Wolfers posts over at Freakonomics that the Fed’s announcement was backdoor QE– after all, long-term rates decreased.

    “So yes, we got lower long-term interest rates. That’s what matters. And it doesn’t really matter how we got there.”

    Doesn’t matter how we got there???

  12. Gravatar of Britmouse Britmouse
    11. August 2011 at 06:01

    Willem Buiter was writing in the FT about negative interest on base money (and even currency) back in 2009 too – he had previously done a paper on Gesell & liquidity traps.

    UK news: Merv King was as gloomy as usual in the BoE quarterly press conference yesterday. Their 2yr CPI forecast has tracked below 2%, I want to believe they are getting ducks lined up ready to QEase further.

    An interesting development on UK stats: they are revising the methodology on the GDP deflator for the National Accounts in October; switching to use of CPI from the historic RPI series (the latter is typically 1% higher than the former). So we might get an upgrade to to real GDP for 2010 onwards, and an improvement in the real/nominal growth split, which would be very welcome.

    “The basic theory and expected impact of changing the basis of the prices previously used (RPI) to using price changes from the CPI is that it will raise volume GDP.”

  13. Gravatar of W. Peden W. Peden
    11. August 2011 at 06:04

    “The purpose is to reduce the demand for the medium of account.”

    Specifically, to stop distorting the market demand for the unit of account with subsidies.

  14. Gravatar of johnleemk johnleemk
    11. August 2011 at 06:04


    As W. Peden and I were discussing in another post, this is easily New Keynesianism’s biggest failure. It pains me to see so many intelligent economists, as Scott says, reasoning from a price change.

  15. Gravatar of JTapp JTapp
    11. August 2011 at 06:04

    As a follow up to the Wolfers quote, I notice that Ryan Avent has taken him to task for it over Twitter. It’s too bad Wolfers doesn’t read your blog, Scott.

  16. Gravatar of W. Peden W. Peden
    11. August 2011 at 06:05


    It does, because the Fed has greatly limited its scope of credibility-preserving action, as George Selgin has pointed out.

  17. Gravatar of Scott Sumner Scott Sumner
    11. August 2011 at 06:11

    John, They can buy T-bonds. (Not sure about equities–can only investment banks buy equities?)

    Joe, The base has tripled. If the demand for base money weren’t rising, then NGDP would have tripled.

    Charlie, Check out this post:

    Donald, Thanks, and also thanks for warning me not to give the impression I was recommending stocks–that would have been quite embarrassing. 🙂


    You are right that it is not necessary to reduce the demand for base money, you could increase the supply and get the same effect. But some people claim the Fed is reluctant to have its balance sheet get too large. The risk factor.

    You asked:

    I also have an off-topic, but I think vital, question that I would be grateful if you would address when you get on top of your blog comments. You emphasise the “hot potato” effect of money. Is there any mechanistic or theoretical explanation (eg microfoundations) of this effect?”

    It’s simply the supply and demand model. Supply more of X, and the value of X falls. That’s all–nothing mysterious. The only wrinkle is that the nominal price of X is fixed, so a fall in its value means inflation.

    John, Thanks, and the same to you.

    JTapp, On no! That’s awful. I need to do another post.

  18. Gravatar of RebelEconomist RebelEconomist
    11. August 2011 at 06:42

    Thanks Scott, but I’ve been got at by Mike Sproul, and I don’t think a simple supply and demand model will do! If X is worthless except as a substitute for the backing asset it was exchanged for (eg gold, government debt etc), then supplying more of X in return for more of its backing asset will not increase the exchange value of X. The question remains, why is money worth more than its backing. Otherwise, as Mike Sproul says, you are assuming the essence of the quantity theory in order to prove it.

  19. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    11. August 2011 at 08:43

    Rebel economist, money has a unique characteristic; demand for it is infinite. No matter how much is created it will be purchased with goods and services. All those stories of Germans carrying money to the grocery store in wheelbarrows after WWI demonstrate that.

  20. Gravatar of George Selgin George Selgin
    11. August 2011 at 09:38

    “Rebel economist, money has a unique characteristic; demand for it is infinite.” Sorry, Patrick, but this observation gets you an “F” in Monetary Economics. Reichsmarks became worthless precisely because the demand for them, that is, for real balances made up of them, was not only finite but shrinking, while the nominal supply exploded.

  21. Gravatar of John John
    11. August 2011 at 09:49


    By reducing the demand for the medium of account, we can get banks to shove their money into T-Bonds, funding socialism. Lovely. Yeah I remember that banks can buy equities after glass-steagal was repealed so my fault for asking a dumb question.

  22. Gravatar of RebelEconomist RebelEconomist
    11. August 2011 at 09:59

    I have no doubt that you are right that money has a “unique characteristic”, Patrick, but economics ought to be more rigorous than that, especially when it is being used to choose policy.

    By the way, I would say that hyperinflation is a special case which is not relevant to, say, the US at the present time. I am asking about a normal situation in which money is supplied by buying assets at market value, as opposed to being “printed”, meaning supplied to the government in return for non-marketable or even non-existent debt.

  23. Gravatar of Scott Sumner Scott Sumner
    11. August 2011 at 18:21

    Rebeleconomist, Why would anyone assume cash and bonds are close substitutes? They don’t seem like close substitutes to me. When I go shopping I don’t think “hmmm, should I take cash or T-bills to the grocery store.”

    And we know that if we monetize a big debt we get hyperinflation–so the quantity of money clearly matters. Sproul’s model can’t explain why Canada and Australia have similar price levels. The QTM can.

    I’d go even further. You can control the price level merely by controlling the supply of nickels–as long as the nickel market doesn’t have a shortage or surplus.

  24. Gravatar of Scott Sumner Scott Sumner
    11. August 2011 at 18:25

    Britmouse, Thanks for that interesting info about Britain. I did know about Buiter–like Mankiw he published his proposal after me, and his was also more like Mankiw’s as I recall.

  25. Gravatar of Philo Philo
    13. August 2011 at 19:30

    Have the excess reserves of the banking system been increasing lately? I would expect so: the Fed is paying 0.25% interest on reserves, while short-term interest rates in the general economy have fallen considerably below that. Banks have an ever-more-powerful incentive to increase their reserve deposits with the Fed. Are they visibly acting on that incentive?

    (I asked this question of Bill Woolsey, but I am less confident of a reply from him than from you–when you get time.)

  26. Gravatar of ssumner ssumner
    14. August 2011 at 08:17

    Philo, That’s not clear. Recall that the Fed, not banks, determines the total monetary base. Any more ERs must come from cash, or from RRs. But low interest rates also increase the demand for cash, and the demand for RRs.

  27. Gravatar of Philo Philo
    14. August 2011 at 18:48

    Bill Woolsey *did* reply, referring me to a graph from the St. Louis Fed, showing that excess reserves have been increasing quite sharply this year.

    – July, 2011: 1618.197 (billions of $)
    – June, 2011: 1588.808
    – May, 2011 : 1512.702
    – April, 2011: 1452.121
    – March, 2011: 1362.678

    It appears that since near the end of 2010 excess reserves have increased from a mere $1 trillion to over $1.6 trillion. Doesn’t this make the Fed uncomfortable? Are they ever asked about it–for example, in Bernanke’s press conferences?

Leave a Reply