The Third Way

The conventional wisdom says that as the economy recovers we need to wind down the fiscal stimulus as quickly as possible.  Krugman makes a very persuasive argument that it is much too soon to pull back on fiscal stimulus, as the economy has not even started to recover.

We do need much smaller budget deficits ASAP, but we also need much more stimulus.  How do we achieve these two seemingly incompatible goals?  With a much more aggressive policy of monetary stimulus we can get faster NGDP growth, and this will reduce fiscal deficits in two ways:

1.  The automatic stabilizer part of the deficit will shrink naturally.

2.   There will be less need for discretionary fiscal stimulus.

There are three ways to make monetary policy more stimulative.  Unfortunately the political viability of each approach is inversely related to its effectiveness.  The most effective but least likely option would be for the Fed to commit to a NGDP target path, with level targeting (i.e. a growth path that they commit to catching up to if they fall short (and vice versa).)  The second most effective option would be a modest interest penalty on excess reserves, perhaps 2%.  The least effective option is quantitative easing.  A bit of QE has been tried in the past few months, although less than many people realize.  Again, there are far more effective monetary policy tools, but as the Fed seems unwilling to use them, it looks like QE is all we have for the moment.

By the way, has anyone noticed what’s been happening to 5-year bond yields in the past week?  The recent drop in interest rates is really bad news.  I don’t know if the Fed fully understands the problem of low inflation expectations, but you can be sure that the stock market does.

US Treasury Bonds Rates Maturity Yield Yesterday Last Week Last Month
3 Month                                  0.13             0.14          0.15          0.13
6 Month                                  0.25             0.27          0.32          0.24
2 Year                                     1.23             1.27          1.41          0.85
3 Year                                     1.84             1.89          1.96          1.30
5 Year                                     2.71             2.78          2.93          2.00
10 Year                                   3.72             3.79          3.88          3.13
30 Year                                   4.58             4.64          4.62          4.08

It’s time for the Fed to stop hoping for things to turn around, and become much more proactive.  Right now they should be focused on the 2 to 5 year yields—which are signaling that near to medium term NGDP growth expectations are still far too low.

[BTW, if we reduce the fiscal deficits through a more expansionary monetary policy in the short run, it will reduce the risk of future inflation, as future governments won’t feel pressured to inflate away large nominal debts.]



23 Responses to “The Third Way”

  1. Gravatar of Alex Golubev Alex Golubev
    15. June 2009 at 09:33

    good call on the 5 yr, but political threshold for 5 yr yield decline hasn’t been reached, yet. CYA! there’s way more cover to err on the side of little stimulus right now. it’s wrong, but it’s true. Thow in the false S&P500 breakout last thursday and i think we have a decent bet of having seen the top of the green sh**t.

  2. Gravatar of Jeremy Goodridge Jeremy Goodridge
    15. June 2009 at 09:37

    Hi Scott:

    I have a question that has always puzzled me. Do you think it would be credible for the Fed to commit to an NGDP target? Certainly they can SAY they are committed to it. But can they actually carry it out? Imagine a Fed that everyone believed really would try to their absolute capability to drive up NGDP. So the issue of credibility was not one of whether they meant it or not, but rather one of their technical ability to carry it out. Will the money they print and inject into the economy actually result in an increase in NGDP? Or are people are SO pessimistic about the future, that there is nothing the FED can do to drive people to spend/invest more. I guess, in the extreme, they could certainly drive up spending by dropping loads of money into everyone’s savings account. But then again, would people anticipate inflation and just decide they have to keep all that in savings in order to preserve the real value of their savings. Or would the actions really required to drive up NGDP create hyperinflation. Maybe it’s one of those weird bi-polar equilibrium situations: you’re either in mild deflation or disinflaton or you have hyperinflation, but you can’t have what you want, which is in between those two states until people’s optimism improves.

    Thanks again,

  3. Gravatar of Alex Golubev Alex Golubev
    15. June 2009 at 09:58

    Scott, on complex systems, so you can get a sense of what i keep trying to “sell” you:

  4. Gravatar of Jeremy Goodridge Jeremy Goodridge
    15. June 2009 at 12:32

    Hi Scott:

    Another thing to point out: it seems that NGDP targeting is not just another tactic but a whole different framework. On the other hand QE and penalty interest rates are tactics, both of which might (and probably WOULD) be used under an NGDP targeting regime. If the Fed had an explicit NGDP growth target, it would do things that from an action standpoint look a lot like QE (and possibly involve the imposition of penalty interest rates). It’s true that the mere existence of the target makes a difference — because it tells market participants how far the Fed is willing to go. And in that sense it’s like a target. But it would certainly mean nothing if the Fed just said they had that target but didn’t ultimately print money and get that money into the economy.

    Also, I suspect penalty interest rates on reserves will end up looking very similar to QE with short-term Treasuries. If banks are penalized for holding reserves at the Fed, that will force them to buy short-term treasuries. Which, of course, is what the Fed would be doing in a QE regime.


  5. Gravatar of Jon Jon
    15. June 2009 at 13:51

    A bit of QE has been tried in the past few months, although less than many people realize.

    Its a bit dependent on your start date. Since January 15th, monetary base is flat, but versus March +200B. To the eye, the trend looks to be ‘in the noise’ of fluctuating demand.

  6. Gravatar of Tom Humphrey Tom Humphrey
    15. June 2009 at 17:01

    I’m very much in sympathy with your proposal that the Fed target NGDP. But I’m afraid the Fed cannot target that particular variable with any credibility. The Fed cannot control that variable as it can the monetary base. Only variables like the base, which the Fed has under its direct control, are credible Fed targets.

  7. Gravatar of ssumner ssumner
    16. June 2009 at 04:06

    Alex, I hope I am wrong (i.e. I hope the Fed has already done enough.)

    Jeremy, The bi-polar issue was discussed in a post back in February entitled “but wouldn’t that cause hyperinflation.” It is a sign people are thinking about the problem in the wrong way. The key is that if policy was on target, inflation expectations and real growth expectations would be higher, and thus there would be less hoarding of cash. So with an effective policy you won’t have to pump a lot of money into the economy.
    A more recent post called “spot the flaw in NGDP index futures targeting” (a month or two back) discussed how NGDP can be targeted.

    Alex#2, I gave up part way through, can you give a quick summary?

    Jeremy#2, The problem with QE in that in the absence of an explicit NGDP (or inflation) target it may not be viewed as permanent. And only the permanent part of monetary injections raises prices. So the same QE can have very different effects if expectations are different.

    Jon, Yes, the MB went down then up. I should do a post on this. The drop in January and February shows that the MB is currently endogenous. I doubt there is a single economist in the world who thought the Fed had a “tight money policy” or even “tight money intent” during that period (when the markets were all crashing.) Instead, I think we’d all agree it was endogenous money. So if cutting the MB 200 billion didn’t hurt, why should we expect that increasing 200 billion after March would help? The answer is that it may not help. QE is only effective if it leads to a higher future expected money supply. QE only works if it changes inflation expectations.

    Tom, There are a couple ways of addressing this. My preferred approach is for the Fed to target the future expected NGDP. Thus they should set the base (or interest rates) at a level where expected NGDP growth is on target. Ideally they would peg an NGDP futures contract. Or they could rely on internal forecasts. With futures targeting, the base is endogenous and automatically adjusts to the level where expected NGDP is on target. In years past when I presented this idea to the Fed, they said they could forecast just as well as the markets, so in that case they should set the MB at a level where their own internal forecast unit projects on target NGDP growth.

    A key part of the policy is “level targeting” or targeting a growth trajectory over time. Thus if in the first year they target 5% growth and it comes in at 4%, then the next year they target 6% growth. That makes velocity stabilizing. If the economy is drifting off target, velocity moves in such a way as to nudge you back toward the target (in anticipation of future monetary actions to get you back on target.)

  8. Gravatar of US producer prices: a new sign of deflation? | US producer prices: a new sign of deflation? |
    16. June 2009 at 09:07

    […] and the Fed is that they stop their stimulus too early. The central bank is currently using the least effective stimulus method, argues The Money Illusion […]

  9. Gravatar of JTapp JTapp
    16. June 2009 at 10:05

    Dr. Sumner,
    I’ve tried to keep up with your blog as best I can, but haven’t made time to catch up on every post, so forgive me if this question is answered elsewhere.

    Have you previously stated that a NGDP targeting could successfully get a country out of a liquidity trap by itself?

    I’ve ditched the Mishkin text and will be using Laurence Ball’s new M&B text in the fall (Mankiw’s recommendation). He makes the same point that Krugman does using the example of Japan:
    “With the interest rate at zero, the BOJ performed large epansionary open-market operations…M1…grew by 14 percent in 2001…24% in 2002. However…rapid money growth did not raise output or inflation.”
    He shows how in a liquidity trap demand for money goes horizontal, so that when money supply is increased there is no effect on nominal interest rates.

    Is there a specific post where you explain why exactly NGDP targeting would work better than QE given a nation is ALREADY in a trap? Wouldn’t Fed action for NGDP targeting be the same as expanding the monetary base and lead to the same result? Why/how does NGDP targeting in a liquidity trap affect inflationary expectations?


  10. Gravatar of jtapp jtapp
    16. June 2009 at 12:08

    I’d also point out that the Fed is still actively buying 5 year Treasuries and longer this week, pushing yields lower.

  11. Gravatar of Jon Jon
    16. June 2009 at 12:26


    Under greenspan the Fed was always buying five year treasuries; that leaves the question of quantity.


    Why is inflation helpful if quantity itself is the transmission mechanism?

  12. Gravatar of Winton Bates Winton Bates
    16. June 2009 at 16:06

    It seems to me that it is possible to be sceptical about the Fed’s ability to prevent inflation expectations from rising sharply at some time over the next couple of years (as I am)and yet to feel that the consequences for the real economy will be much worse if it follows policies that would guarantee that this will not occur. Even for those of us who are sceptical about the Fed’s ability to manage expectations (based on its recent track record) the relevant question is whether outcomes will be better if the Fed aims to raise infation expectations a little above current levels or if it aims to reduce them below current levels.

    By the way, how about writing a post about the political economy of monetary policy? (I’m assuming you haven’t written one already.) I am wondering what constraints prevent the Fed from providing clearer signals of its intentions.

  13. Gravatar of ssumner ssumner
    17. June 2009 at 04:43

    JTapp, I don’t recall which posts, but let me say that both inflation and NGDP targeting can work, and they work in essentially the same way. They tend to lower real interest rates by raising inflation expectations, and also raise the Wicksellian equilibrium real interest rate by raising expected real growth. How can we do this if we are in a liquidity trap? Commit to a future monetary policy that raises future prices (i.e raises prices after we have escaped the liquidity trap.) Higher future expected prices raise current expected inflation (and current equity prices I might add.) I think most economists accept this view, it’s just a question of whether the Fed can make this commitment.

    JTapp#2, I am pretty sure that the five year yields are falling because of increasing bearish expectations. Look at the stock market. If easy money was depressing rates in the last few days, stocks would have been rising. I’m not saying you are never right, stocks rose and yields fell when QE was announced in March. But typically they go the same way when you are in a deep slump.

    Jon, Sorry, The term “quantity itself” confused me–can you be more specific? Is your question about NGDP, or quantity of money? BTW, In my newest post I should have also criticized Krugman for again saying the Fed normally bought only T-bills. He is still making that claim.

    Winton, I think you are saying that right now it is better to err on the side of a bit too much inflation. If so, I agree. I think 2% inflation is about right (although I prefer NGDP targets.) But if we erred, I’d prefer 2.5 or 3.0 to 1.5 or 1.0. After all, we do have 9.4% unemployment.

    I tend to assume the fed is idealistic, and give them advice on that basis. That may be naive, and I’ll try to think of a more political take on things for a future post.

  14. Gravatar of Alex Golubev Alex Golubev
    17. June 2009 at 06:05

    Scott, sorry i understand that 60 slides is a bit much and the intro was a bit basic. I suggest loooking at slides 35, 37, 38, 45-54 – that is where the meat of new material is. here’s the same link:

  15. Gravatar of ssumner ssumner
    18. June 2009 at 05:01

    Alex, OK, I looked at it, but what does it have to do with monetary policy? Does it have anything to say about targeting NGDP futures?

  16. Gravatar of Alex Golubev Alex Golubev
    19. June 2009 at 06:21

    not directly, but you asked me why i keep bringing up the need to model complex social systems with reinforcing processes and this is a much better presentation of the material.
    The way i see it is that leverage is directly correlated to “coupling”. normally your neighbors’ actions wouldn’t have too many implications for you. But if everyone around you is leveraged, a slight miscalculation on one’s part can lead to asset liquidations which will affect the rest of the agents leading to real loss of equity and a PHASE SHIFT (not gradual linear drift) into much slower growth state.
    Another degree of coupling is globalisation and higher integration. For example – the ability to access your electronic funds online give you that ability to more and REMOVE your money from riskier assets quicker than ever before. This enhances the degree of coupling once again (in my opinion). There’s a lesser “cooling off period”, so our “rationalizing” mind has even less time to make educated decisions/choices and we end up acting on emotions more.
    what do all these things have to do with monetary policy. Not that much directly, but economics is a study of the most efficient way of resource allocation and i wanted you to opine on something that i think has great application potential, but very little current acceptance and even exposure. (I want you to say EUREKA and share the presentation with the econ community. I mean, the one way communication with a teacher is great, but I think the blog can give something back to you as well)

  17. Gravatar of ssumner ssumner
    20. June 2009 at 08:46

    Alex, My general impression is that people who come at economics from the outside think that the economic system is very unstable. Thus they look for all sorts of deep reasons for that instability. In my view the economic system is only as unstable as monetary policy. Bad monetary policy makes the system look much more unstable than it really is.

  18. Gravatar of Alex Golubev Alex Golubev
    22. June 2009 at 08:49

    Scott, I’m not that much of an outsider, 4.0 in all econ classes i ever took, 3.8 in finance (major), CFA, and years of watching financially EDUCATED people makign real financial decisions. But i’ll give you that i’m not fascinated by econ, i’m fascinated by markets (since high school – thank you dotcom bubble!) so a close “sibling”. I see econ more like a game of monopoly, because of what I believe to be unrealistic assumptions. Yes, we can explain a lot within that framework (and i did just fine in college), but i don’t think the “captains” (fed, hedge funds trying to take the other side of a “misbalance”, educators…) are well equipped right now to handle the unfair reality of life/business/politics due to assymtric incentives and leverage.

    I think that we’d want economic activity to be stable. However i would argue that economic activity/resource allocatio has to be fair and if it’s unfair, imbalance will be created.

    Now high dedgees of LEVERAGE, are highly unstable for any system and you cannot say that leverage is not a concern of monetary policy, because one of fed’s tools is the ability to control the reserve ratio, not to mention the creation of money out of thin air is also “leverage”. So… all i’m saying is that leverage can be one of the most creative and destructive forces on earth and not monitoring the NEW WAYS it manifests itself is WRECKLESS on the part of the captain, whoever thinks they’re watching over STABLE (5%) economic growth. It actually baffles me how little you talk about leverage, because i always though it was way more IMPORTANT than interest rates.

    Another point – pace of techonoligical progress and innovation is hardly linear (5% or not). exponential jumps and phase shifts as new technologies and energy sources are discovered. new ideas don’t change the fact that FUNDAMENTALS and MARKET LEVELS can get disconnected, but how many people do i hear every day tell me that there’s an “green premium” for living in seattle. i bet 95% of the people dont’ understand that only NEWS aren’t already reflected in prices. And that’s with boring concepts. Whenever the next internet is invented, we’ll ahve another blowout.

    We will ALWAYS misallocate resources as new things are invented. THUS another aspect of striving for 5% growth is that it doesn’t jibe with what political and economic pressures will exist for the agents within the economy as well as the captain. Obviously an all knowing being doesn’t exist to tell us when to slow down and you task the markets with this, but i think that WAY in which growth is accomplsihed is incredibly unfair. I’ve only been in this country for two full cycles (’93-now) and it baffles me that we’re so far ahead of other coutnries in fairness, yet so far behind simple fairness, because of the incredible assymtery in incentives and leverage. Growth shoudl take a distant 2nd to fairness in my world.

    (i’m sorry for being Yoko, but argument/discussion – creative destruction – are much more productive than me/you talking with people that agree with me/you. so thank you/me :))

  19. Gravatar of Current Current
    22. June 2009 at 09:01

    Alex Golubev you could do better with fewer capital letters.

  20. Gravatar of Alex Golubev Alex Golubev
    22. June 2009 at 12:07

    I disagree, any sort of writting could do much better with tone. It’s a subset of speech which is a subset of communication. unless of course you’re being a d1ck, then i’d have to say something mean back about you not having a last name. Or if you were joking, then i’d say LOL. BTW, i dont’ capitalize i’s and first letter of some sentences to make up for it. Capitalization redistribution.

  21. Gravatar of ssumner ssumner
    23. June 2009 at 06:32

    Alex, I’ll concede one point. I think the “Great Moderation” from 1982 to 2007 did lead to more leverage. The paradox is that the more stable the Fed makes the economy, or seems to make it, the more risks people take. So that is an issue. But I still think we should go for stability, even if it leads to more risk taking, as it will make the consequences of bubbles much less bad. The subprime bubble never should have caused this severe recession. BTW, one way to reduce leverage is to reduce moral hazard by getting government to do less bailing out.

    Current and Alex, I tend to use caps more from my home computer, as there is no italics button (or maybe I’m too ignorant to find it.) Otherwise I use italics at school. But I have no problem with caps.

  22. Gravatar of Why Negative Nominal Interest Rates Miss the Point Why Negative Nominal Interest Rates Miss the Point
    11. July 2009 at 18:24

    […] Buiter, Greg Mankiw, and Scott Sumner have all recently proposed negative nominal interest rates on reserves or currency as a way to […]

  23. Gravatar of TVHE » On penalty cash rates TVHE » On penalty cash rates
    12. July 2009 at 19:37

    […] Scott Fullwiler from New Economic Perspectives (ht Economists View) describes some issues he has with the “negative interest rate” idea being put forward by Willem Buiter , Greg Mankiw , and Scott Sumner […]

Leave a Reply