Suggestion for Carney and Osborne: A NGDPLT/inflation hybrid

As far as I can tell there is a pretty strong desire on the part of both Mark Carney and the Cameron coalition government for a more pro-growth policy from the Bank of England.  A policy that would allow them to continue with fiscal austerity, while still providing for reasonable levels of demand growth.  We know that Carney has some interest in NGDPLT, as do some people in the Cameron government.  But there is also substantial opposition to abandoning the 2% inflation target.  Doing so risks a loss of credibility and confusion among the public (some claim), as NGDPLT is not well understood.   I have a proposal that would deliver at least 95% of the anti-cyclical benefits of NGDPLT, while preserving 2% inflation as a long run target.  It’s actually a variant of flexible inflation targeting, with strong NGDP stabilizing properties.  First I’ll describe the policy if we were starting from full employment.  Here’s how it works:

1.  The Treasury estimates Britain’s long term sustainable trend RGDP growth each 2 years.  Suppose the initial forecast is 2%.

2.  The Treasury instructs the BOE to set a 4% NGDP growth target for the next 4 years, with level targeting.  This would be 2012 to 2016.  The target will be updated every two years as new long term growth estimates come in from the Treasury.

3.  Suppose in 2014 the Treasury estimates Britain’s long term trend growth has fallen to 1.8%.  They would instruct the BOE to aim for 3.8% NGDP growth between 2016 and 2018.  That gives markets plenty of time to adjust.  The growth target would remain at 4.0% from 2014 to 2016.

4.  Suppose that by 2016 the Cameron supply-side reforms seem to be working, and long term trend growth is estimated at 2.1%.  Now the Treasury instructs the BOE to aim for 4.1% trend growth between 2018 and 2020.  The target remains 3.8% between 2016 and 2018.

If we started from a position of high unemployment, the initial target path would involve some catch-up in NGDP, before leveling off at the steady state.  How much is a judgment call.  After that the only discretion is the Treasury’s trend growth estimates.

To some NGDPLT supporters it might seem I’m giving away the store with this compromise.  Not so.  Trend RGDP estimates evolve slowly over time, and hence the adjustments in the NGDPLT path would be very gradual.  Now imagine a trend line with those sorts of modest adjustments superimposed over the dramatic crash in NGDP that occurred in Britain and elsewhere in 2008-09.  It’s easy to see that you’d still get at least 95% of the anti-cyclical properties in a recession, maybe even more.  Remember that at any point it time a minimum of two years in trend NGDP growth would be “locked in,” and as much as 4 years of nominal growth.  The target NGDP growth path would not be perfectly smooth, but it would be smooth enough to dramatically stabilize the business cycle.

The other advantage is that the government could still adhere to its 2% long run inflation objective.  Note that 2% inflation is the only number built in to the policy.  The NGDP target paths are those believed to be most likely to produce 2% inflation in the long run.  I don’t see how it would be wrong to describe this as “flexible inflation targeting” if the government thinks that label is essential for political purposes.  Remember, it’s not as if the current policy rigidly adheres to aiming for 2% inflation at each and every moment in time.  The BOE has already been one of the most “flexible” of the flexible inflation targeting central banks.  It’s failure in 2009 (and again in 2011) was not due to a lack of flexibility, it was due to a lack of level targeting, and a lack of recognition that NGDP, not inflation, is the key short term variable to focus on.

All the agonizing decisions about how fast to proceed with fiscal austerity are not agonizing because of how fast they might reduce inflation, they are agonizing because of worry about how fast they’ll reduce NGDP.  The public cares more about NGDP than people think; it’s just that the public doesn’t really understand the term.  They think in terms of “the economy.”  But anyone who thinks the public doesn’t care about NGDP should consider the following:  Ask people if they care about their own personal nominal incomes.  About how big a raise their boss gives them, or whether they’ll see more investment income, or whether they lose their jobs.  Then think about the fact that NGDP is simply the sum of all (gross) nominal incomes in the economy.

Yes, people also care about real incomes over time, but the central bank can’t control that in the long run.  And in the short run NGDP and RGDP are highly correlated.  So tell me again why the public doesn’t care about NGDP instability.

PS.  Astute readers will notice that I get this seemingly miraculous compromise by applying “let bygones be bygones” to errors in RGDP trend growth estimates, but adhere to rigorous level targeting of NGDP.  That does lead to a policy that fails to stabilize the price level along a 2% trend line, but we already have that sort of base drift in the price level in the various inflation targeting regimes used throughout the world.

PPS.  If the proposal is not destroyed in the comment section, I would appreciate if influential readers would pass this along to important policymakers that they know.



58 Responses to “Suggestion for Carney and Osborne: A NGDPLT/inflation hybrid”

  1. Gravatar of Felipe Felipe
    14. December 2012 at 06:59

    Small detail: the central bank changes targets every 4 years, but RGDP estimates change every 2 years. Let me continue your hipothetical timeline:

    5. RGDP is estimated in 2018 to be 1.8%. What happens with the 2018-2022 target for the central bank? Is it changed? Or are this estimates pushed to the target for the 2022-2026 range?

    Perhaps better is to produce RGDP estimates every 4 years, but with a phase of 2 years relative to the setting of the central bank’s target. In other words, the target is revised two years after setting it, but the setting of the new target takes 2 years to implement.

  2. Gravatar of StatsGuy StatsGuy
    14. December 2012 at 07:18

    You could move this to a continuous basis, and just use a rolling updated average. In other words, you have a prior estimate of what RGDP+2% will be, and the KEY is that the CB is not allowed to update that immediately if RGDP comes in low.

    2/4 year cycles create problems. Let them update it every month if they want, but the key is that the update is only partially. So each month, they specify a projection for the next 2 years (or 3 or whatever), and the NEW projection = 0.98*old projection + 0.02*new projection.

    This would get rid of cyclical uncertainty around update periods, and achieves exactly the same thing.

    The only problem I would see here is if the market expects a shock to last more than 2 to 4 years. NGDP targets, especially NGDPLT, deal with very long term failures or misalignments.

    For example, what if the markets suddenly discovered that Britain had a massively overvalued currency that had been building over time, and needed a significant adjustment – it would take years for this adjustment to be implemented. In the first few years, RGDP estimates would begin to drop, and we might end up in a low cyclical trap where RGDP projection is low but we’re still capped to 2% inflation, which would significantly slow down the path that would allow Britain to rebalance its currency.

    The longer the lag period in the RGDP projection updating, the less of a problem this is. If, for example, you moved to a system whereby each month the new RGDP projection is 0.99*Old Projection + 0.01 * New Projection, it might be better off – that gives you a half life of about 70 months.

    This is a classic exponential decay type function.

  3. Gravatar of Nick Rowe Nick Rowe
    14. December 2012 at 07:30

    Scott: not bad. I think Statsguy has a point. A second danger is hysterisis, where the time-path of the natural rate of RGDP depends on the past history of actual RGDP.

  4. Gravatar of Bill Woolsey Bill Woolsey
    14. December 2012 at 07:31

    I support this sort of thing, but with much less frequent adjustments.

    I have usually proposed 5 years.

  5. Gravatar of Mark Carney Mark Carney
    14. December 2012 at 07:39

    Great idea, I will pass along

  6. Gravatar of ssumner ssumner
    14. December 2012 at 07:50

    Filipe, The target for 2018 to 2020 does not change, the target for 2020 to 2022 becomes 3.8%.

    Bill, I originally proposed this with 10 year adjustments, but I’ve come to believe shorter adjustment is more politically acceptable, and does almost as well. I certainly would not oppose 5 years, obviously.

    Statsguy and Nick, I’m not opposed to Statsguy’s suggestion, or indeed almost anything reasonable along these lines. But I wonder if you are overestimating the likely size of changes in trend RGDP estimates. Recall that there is no trend revision to the actual RGDP path, and that these are very long run trend estimates. So economic problems that caused Britain to expect only 1% RGDP growth for 5 years, might very well depress the very long run trend growth estimate by a much smaller amount, say from 2% to 1.8%. That was the point of my final PS about base drift in the price level. There will be lots of price level base drift. Had this been in effect after 2008 the BOE would have had to allow much higher than 2% inflation between 2008 and 2012, as the trend growth during those 4 years was probably much less than 1%. But that much less than 1% number would have never become part of the target, as the British still expected close to 2% RGDP growth over the very long run–that number tends to change very little. Part of the temporarily slow growth, for instance, was sharply falling oil output. That does not and should not have much weight in estimates of very long term trend RGDP growth. So I still think I’m right as a practical matter, but am obviously willing to tweak the proposal in either Stastguy’s or Bill’s direction.

  7. Gravatar of Rob Rob
    14. December 2012 at 08:06

    Politically this may be difficult – maybe even more difficult than simply switching to NGDPLT.

    At present, although the Treasury sets the mandate for the Bank, there’s a widespread perception (which all parties wish to maintain) that the Bank operates entirely independently of the Treasury. After all, the mandate was set 14 years ago and hasn’t changed since. There’s a folk memory of the “bad old days” when the Chancellor could set interest rates directly, or could direct the Bank to behave in certain ways at will. Most people are agreed that the current situation, in which the Bank does what it believes to be best, without political interference, is better.

    When people see phrases like “Now the Treasury instructs the BOE to aim for 4.1% trend growth”, they will probably imagine a return to the “bad old days”. After all, the Treasury still has an incentive to overestimate the potential for trend growth, because in the short run this will lead to more stimulative monetary policy even if the government hasn’t “earned” it with supply-side improvements. Given that it will take years for the estimates to be proved wrong and it’s unlikely that the Treasury ministers who are responsible for the projection will still be in office when this happens, they’ve every incentive to be overly optimistic. Even if the same politicians are still around when the forecast is proved wrong, it may be hard to hold them to account because responsibility is diffused between the Treasury and the Bank – people are accustomed to the idea that excessive, unexpected inflation can be blamed on the Bank, but in this scenario it’s less clear – in theory, the Treasury should be blamed. A possible way around this would be to give the forecasting role to the Office for Budget Responsibility, but this is a new body which is still finding its way, and has already demonstrated a tendency to provide overly-optimistic forecasts of economic growth.

    I’m not saying that I don’t like the idea, but if the objective is to make things easier to understand and to overcome objections about weakening the constraints on inflation, I don’t think that this proposal would work. For the average British political commentator (bears of very little brain indeed) it would probably be harder to understand than simply saying that the Bank is still in charge but with a new mandate.

    Another way to put it: if they government makes any change to the mandate of the Bank, they will receive the same amount of criticism from the same groups of people (hard money conservatives and the Labour party, the latter because their job is to oppose anything the government does irrespective of whether it’s a good idea). A lot of this opposition is instinctive or tribal rather than rational and as such it’s not possible to reduce it via compromise. On that basis, it’s better to just accept the criticism and shoot straight for the policy you really want.

  8. Gravatar of Brian Donohue Brian Donohue
    14. December 2012 at 08:16

    This is an outstanding post for anyone trying to understand what exactly you guys and your crazy theories, which may embody important truths/insights, say and do. Thank you.

    Really, Scott, would you say your apple hasn’t really fallen far from Friedman’s tree?

  9. Gravatar of jsalvatier jsalvatier
    14. December 2012 at 08:19


    You could have the bank itself estimate real GDP trend, it does that sort of thing already in the US.

  10. Gravatar of Saturos Saturos
    14. December 2012 at 08:40

    I think the defect with this approach is mainly that you can’t guarantee the level of NGDP several years out. And isn’t that supposed to be the way that NGDPLT gets an economy out of a liquidity trap? There is no credibility beyond the very short run, and perhaps not even then. Still, probably an improvement over present arrangements (at least it’s level targeting, of sorts).

  11. Gravatar of dtoh dtoh
    14. December 2012 at 08:56

    Stop diluting your message! It’s just a matter of time. Economists are slow learners and policy makers are slow to make changes.

    Don’t confuse them!

  12. Gravatar of Bill Woolsey Bill Woolsey
    14. December 2012 at 08:59


    “long term sustainable trend RGDP growth” means what exactly?

    I think some of the discussion interprets this to mean estimated average potential output growth over the next four years.

    For the U.S., you can go to the potential output estimate four years out, and come up with the average growth rate it takes to get there.

    That determines the growth rate (of the growth path) starting two years in the future and going on to 6 years in the future.

    That doesn’t make sense. And so, I think you have in mind someting like the average 10 year CBO forecast of potential income growth.

    Every two years, we see if it changed. If it did change, then the growth path for NGDP is adjusted starting in two more years. So in years 2 to 6 we will have a growth rate for our growth path consistent with the CBO estimate in year 1 for the average potential ouptut growth in year 1 to 10.

    My approach is that changes take place with a 5 year lag.

    So, if we decide that potential output is going to be growing slower or more quickly, or that we want a higher or lower trend inflation rate, or we want to targeting nominal final sales rather than nominal GDP, then when we pass the law making this change, it only goes into effect in 5 years.

    If we believe we are going to have slower growth in potential output for the next 20 years or something, then making this adjustment might be sensible.

    Anyway, I think inflation targeting is building extra inertia in to prices, and especially wages.

    What you are proposing is that if we expect slower potential output growth in the future, we have 2 years to slow nominal wage rate growth.

    (Well, if it is population growth impacting this, then that is less of a problem. We avoid the need to ramp up wage rate growth.)

    I am OK with adjusting the rule from time to time, but only with long adjustment periods. In other words, not due to transitory conditions, but rather to very long lasting changes.

  13. Gravatar of Richard A. Richard A.
    14. December 2012 at 09:25

    I actually do favor fine tuning NGDPLT. It should be adjusted for changes in the growth rate of the potential total labor supply and also for long run changes in inflation. For inflation, a long run moving average over several years could be used. The difference between the moving average and the desired inflation rate could be subtacted out from NGDPLT.

  14. Gravatar of flow5 flow5
    14. December 2012 at 09:26

    In the U.S., monetary lags are not “long & variable”. The lags for monetary flows -MVt (our means-of-payment money times its transactions rate-of-turnover), i.e. the proxies for r-gDp growth & for inflation, are mathematical constants (for the last 98 years). Rates-of-change (roc’s) in MVt are always measured with the same length of time as the specific economic lag (as its influence approaches its maximum impact (not an arbitrary date range).

    Once the Central Bank quickly “catches up” to the optimum trend (roc) for r-gDp, then it must refocus on inflation. Under monetarism, this roc cannot be related to a fixed % target, rather it represents the historical monetary time lag in the proxy for r-gDp.

    That’s what “flexible inflation targeting” is actually about. Given that inflation’s lag is also constant, all prior price dislocations will revert-to-mean (prove transitory), when the roc in the proxy for inflation is brought back into long-term alignment.

    What’s subjective is the mean reversion. When roc’s in the proxy for inflation begin to exceed roc’s in the proxy for r-gDp, inflation begins accelerating faster than r-gDp. Otherwise this is cut & dry.

  15. Gravatar of ssumner ssumner
    14. December 2012 at 09:39

    Mark Carney, You’ve made my day!

    Everyone, I continue to be very busy, until Christmas, but will read all comments later. Don’t overestimate the change involved here, as trend rate adjustments in estimated long term RGDP growth rates are extremely gradual. At the level of cyclical stabilization there is little change. No hyperbole when I said at least 95% as good.

    If we win that’s great, but a fallback position is also useful. We’d get there eventually anyway, and need to also think about the US, where we are much further away from success. And then there’s the ECB . . .

    Once people get used to NGDP targeting, the inflation part might naturally fade away over time.

  16. Gravatar of Rien Huizer Rien Huizer
    14. December 2012 at 10:09


    Clever but clerly showing that NDGPLT is in essence an tempt to counteract market forces (like the desperate interwar economists a la Keynes). Bushfires and recessions should be allowed to burn out, not be contained etc. Unforunately bush fires threaten homes and recessions make life hard for politicians. Hence there is a need for plausible and simplifiable (ie communicatebale by politicians to their markets) policy. That is not NDGPLT targeting, despite the fact that you found a way to compromise between two forms of bureaucratic resistance to your policy (which I think is the next best solution to abstaining from any business cycle intervention). It may be good, but it is not what politicians need and for the economic purist (RBC) it is still interventionist. But for the Brits, it should be good enough!

  17. Gravatar of K K
    14. December 2012 at 10:10


    You didn’t address what would happen when they miss the level target over a two year period. I’d propose that any error is carried forward and added to the next period’s target. If so, I like it. Otherwise it’s really just a growth target, not a level target.   

  18. Gravatar of K K
    14. December 2012 at 10:14


    I was assuming that wasn’t actually Mark Carney???

  19. Gravatar of Liberal Roman Liberal Roman
    14. December 2012 at 10:40

    Krugman on the Fed move: “So philosophically, this represents a conversion to the Evans criterion for rates and the Woodford/KRUGMAN doctrine about monetary policy in a liquidity trap.”

    Is stealing people’s ideas and adopting them as your own a form flattery? If yes, I think he likes you a lot Scott.

  20. Gravatar of dtoh dtoh
    14. December 2012 at 11:38

    “Doctrine” ?

    What a pompous ass!!

  21. Gravatar of ssumner ssumner
    14. December 2012 at 13:05

    K, How many commenters named Mark Carney have “” email addresses. Maybe it was a prank, but I’m not so sure it’s not real.

    You guys don’t see the email addresses from where comments are sent, but I do. The address looked legit. But I know nothing about computers, perhaps all that can be faked.

  22. Gravatar of W. Peden W. Peden
    14. December 2012 at 13:13


    I’m pretty sure it can be faked, or so my advisors tell me.

  23. Gravatar of Felipe Felipe
    14. December 2012 at 13:14

    Thanks for the reply Scott, I evidently misread the original post.

    BTW, there is no checking of the email address used in the posts here, so they can be faked by just typing a fake email.

  24. Gravatar of Gregor Bush Gregor Bush
    14. December 2012 at 13:15


    Why don’t you send him an email and ask him. If it was him I’m sure he’ll respond – and would you mind letting us know if he does?



  25. Gravatar of Felipe Felipe
    14. December 2012 at 13:15

    In fact, I have now commented with a fake email address

  26. Gravatar of Mark Carney Mark Carney
    14. December 2012 at 13:19

    It isn’t hard to fake someone, even Mr. Carney.

  27. Gravatar of ssumner ssumner
    14. December 2012 at 13:20

    Rob, Good point, when you think about it there’s no reason not to have the BOE forecast trend growth.

    Thanks Brian.

    Saturos, You always know precisely where target NGDP will be at least 2 years out, and on average three years out. And beyond that you know roughly where it will be 5 years out, within a few tenths of percent. Actual errors in hitting the target are an order of magnitude larger. People are overestimating the amount by which long run trend growth changes from one year to the next. It’s one or two tenths of a percent.

    Bill, You said;

    “I think some of the discussion interprets this to mean estimated average potential output growth over the next four years.”

    No! If I’d meant that I would have said that. I meant over a very long time–say 20 years. But I certainly agree that 5 year periods are better than 2 years for adjustments-I was trying to be politically acceptable.

    K, It’s a 100% level target–no let bygones be bygones. My post describes the target path over time. That’s a level target. Thus if you end up the period of 4% growth target at 3.5% actual growth, then you add 0.5% to the next one. If the target for the next period is 3.8%, as in the example, then you aim for 4.3% next year.

  28. Gravatar of Mark Carney Mark Carney
    14. December 2012 at 13:20

    Or maybe it is, hahaha

    I don’t know much about computers.

  29. Gravatar of ssumner ssumner
    14. December 2012 at 13:25

    OK, Maybe it was fake. But what would be the motive? I once thought a Brad DeLong comment was fake, and when I asked him it was real.

  30. Gravatar of ssumner ssumner
    14. December 2012 at 13:32

    Gregor, But if I send it to the address, how do I know it’s him? If the address is fake the responder will pretend it’s him.

  31. Gravatar of Felipe Felipe
    14. December 2012 at 13:45

    If the address is, then the faker cannot[1] receive the email (unless the faker is an employee at the bank too). Given that, the faker cannot answer the questions you put in the email, since he doesn’t know them!
    Most likely scenarios are an “address not found” bounce or a “you got the wrong address” reply, though.

    The motive would be a joke, I imagine. Unfortunately, on the internet nobody knows when you’re telling a joke.

    [1] For relatively lax standards of impossibility. I’m assuming the faker is not really serious in maintaining the fake identity.

  32. Gravatar of Felipe Felipe
    14. December 2012 at 13:45

    Oops, I posted the above with the fake address. Sorry about that.

  33. Gravatar of Gregor Bush Gregor Bush
    14. December 2012 at 14:42

    As Felipe said is the form of the BoC email addresses. So it actually is Carney’s email address. The problem is that anyone can enter it.

    Given the subsequent responses it appears to be a fake. But you should email Carney anyway under the guise of trying clearing this up but with the secret agenda of getting him to read your blog post and respond to your idea.

  34. Gravatar of johnleemk johnleemk
    14. December 2012 at 14:47

    I agree with Gregor.

  35. Gravatar of James in london James in london
    14. December 2012 at 14:49

    Not sure why NGDPLT keeps getting called a more “aggressive” policy than inflation targeting. If implemented it would just be quite nice. As Market Monetarists have said, the market would do most, all, of the work.

    Rigid inflation targeting is aggressive and has had some pretty aggressive results once it succeeded in bring inflation and interest rates down.

    Aggressive policy is not really something the public will buy.

  36. Gravatar of ssumner ssumner
    14. December 2012 at 14:51

    Felipe, I just sent you an email at your fake Fed address. So why didn’t I get a reply email saying the message did not go through because the address doesn’t exist? I really don’t understand how this stuff works. Usually when I send an email to the wrong address (which doesn’t exist), I get an error message. Obviously your Fed address is fake, so why didn’t I get an error massage?

    Gregor, Normally I would have immediately assumed fake–but I’ve recently gotten some emails from actual people that are fairly important, so I suppose this got my hopes up. Let’s just say that things are looking up for NGDPLT. But I suppose it was too much to hope for that he’d reply here. Still if you go to all the trouble to fake one, why such a normal message?

  37. Gravatar of Gregor Bush Gregor Bush
    14. December 2012 at 15:16

    “Still if you go to all the trouble to fake one, why such a normal message?”

    I agree. Come to think of it, if you’re going to go to all the trouble of moving to a different central bank, why do it if you’re just planning on implementing existing (suboptimal) policy.

    I still think should email him. I think he’ll respond either way.

  38. Gravatar of Felipe Felipe
    14. December 2012 at 15:23

    E-mail delivery is not so simple and can be configured, I can think of the following explanations:

    1. Your mail provider is auto configured to retry a number of times if a delivery failure happens.
    2. The Fed mail provider is configured to accept every mail and then silently drop it (this is done by a large number of mail providers to combat spam).
    3. The Fed mail provider is configured to accept every mail and then try to deliver it internally (this can happen if the internal network is complex), thus delaying the bounce message.
    4. Your mail somehow got marked as spam, and thus is silently dropped.

    I’m sure there are other reasons, but 2 and 4 seem the most likely to me.

  39. Gravatar of anon anon
    14. December 2012 at 16:00

    Gregor, even if Carney did write that comment, he might choose to keep plausible deniability: after all, he hasn’t even moved from the BoC to the BoE. I don’t think it matters that much either way: we already know that plenty of academics and policymakers read Scott’s blog.

    I think Scott’s idea is a great example of how “flexible inflation targeting” can be operationalized and made more rigorous. Once we take for granted that NGDP (not “inflation” or “nominal rates”) is what steers the nominal economy, it’s not hard to come up with improvements to current practice.

  40. Gravatar of flow5 flow5
    14. December 2012 at 17:57

    Monetary policy is principally a function of time lags, it is not a function of quantitative targets.

    Written on Mar 30 11:31 am prior to the MAY 6th FLASH CRASH:

    “Contrary to economic theory, monetary lags are not “long & variable”.

    Assuming no quick countervailing stimulus:

    jan….. 0.54…. 0.25 top
    feb….. 0.50…. 0.10
    mar…. 0.54…. 0.08
    apr….. 0.46…. 0.09 top
    may…. 0.41…. 0.01 stocks fall

    Been saying this for the last 6 months. Should see shortly. Stock market makes a double top in Jan & Apr. Then the real-output of final goods & services falls/inverts from (9) to (1) from Apr to May.

    Recent history indicates that this will be a marked, short, one month drop, in rate-of-change for real-output (-8). So stocks follow the economy down (with yields moving sympathetically?)”

    Ben Bernanke is directly responsible for the depth of the Great-Recession. I don’t forgive him just because he thinks Obama is stupid.

    POSTED: Dec 13 2007 06:55 PM |
    10/1/2007,,,,,,,-0.47,,,,,,, -0.22 * temporary bottom
    11/1/2007,,,,,,, 0.14,,,,,,, -0.18
    12/1/2007,,,,,,, 0.44,,,,,,,-0.23
    1/1/2008,,,,,,, 0.59,,,,,,, 0.06
    2/1/2008,,,,,,, 0.45,,,,,,, 0.10
    3/1/2008,,,,,,, 0.06,,,,,,, 0.04
    4/1/2008,,,,,,, 0.04,,,,,,, 0.02
    5/1/2008,,,,,,, 0.09,,,,,,, 0.04
    6/1/2008,,,,,,, 0.20,,,,,,, 0.05
    7/1/2008,,,,,,, 0.32,,,,,,, 0.10
    8/1/2008,,,,,,, 0.15,,,,,,, 0.05
    9/1/2008,,,,,,, 0.00,,,,,,, 0.13
    10/1/2008,,,,,,, -0.20,,,,,,, 0.10 * possible recession
    11/1/2008,,,,,,, -0.10,,,,,,, 0.00 * possible recession
    12/1/2008,,,,,,, 0.10,,,,,,, -0.06 * possible recession


    Some people think Feb 27, 2007 started across the ocean. “On Feb. 28, Bernanke told the House Budget Committee he could see no single factor that caused the market’s pullback a day earlier”. In fact, it was home grown.

    flow5 (2/26/07; 14:34:35MT – msg#: 152672)

    Suckers Rally

  41. Gravatar of ssumner ssumner
    14. December 2012 at 18:47

    Thanks for the info Felipe.

    So I’ve been learning more about email. I was told that I should check the IP address that accompanied the email. The IP address was “” It seems to be associated with Alliance Bernstein L.P. in New York. What does that mean? Spam?

  42. Gravatar of Steve Steve
    14. December 2012 at 18:57

    Alliance Bernstein website:

    “Our Mission

    To place our clients’ interests first and foremost

    To use our research to gain more knowledge than any investment management firm in the world

    To use and share our knowledge to help clients achieve investment success and confidence”

    Obviously, they are reading MoneyIllusion to “gain more knowledge than any investment firm in the world.” hahahahahaha

  43. Gravatar of Steve Steve
    14. December 2012 at 18:58

    Alliance Bernstein website:

    “Our Mission

    To place our clients’ interests first and foremost

    To use our research to gain more knowledge than any investment management firm in the world

    To use and share our knowledge to help clients achieve investment success and confidence”

    Obviously, they are reading MoneyIllusion to “gain more knowledge than any investment firm in the world.” hahahahahaha

  44. Gravatar of Benjamin Cole Benjamin Cole
    14. December 2012 at 19:47

    Thoughtful blogging, bit I still detest the idea of genuflecting to 2 percent inflation as a ceiling. Maybe as a floor would be better.

    Has anyone been looking at charts of sovereign debt yields? They are headed for zero bound, globally.

    Do you see Japan?

    Do you know how hard to is to get an economy up after recession-zero bound?

    Add on: Why a 2 percent inflation ceiling? Is there any evidence that 2 percent results in higher long-term real growth than 3 percent?

    All that said, I understand Sumner in this post. I don’t understand the economics profession and its fixation on microscopic rates of inflation.

    I would prefer a fixation on boom times.

  45. Gravatar of JimP JimP
    14. December 2012 at 20:45

    Cochrane would prefer a price-level target:

    I’m not such a fan of new-Keynesian models (here, with hard academic article warning) so this lack of real commitment doesn’t trouble me that much. I don’t think we would get immediate benefit even from a completely credible tied-to-the-mast commitment to buy trillions of dollars until unemployment hits 6.5%. (I do think rules-based policy in general is a good idea, not this sort of discretionary commitment-making. But I can think of a lot better rules, like “the price level shall be CPI=130 forever, period.”)

  46. Gravatar of Dilip Dilip
    15. December 2012 at 05:40

    “So I’ve been learning more about email. I was told that I should check the IP address that accompanied the email. The IP address was “”³ It seems to be associated with Alliance Bernstein L.P. in New York. What does that mean? Spam?”

    What that means is the person who made that comment made it from a PC which has that IP. and per your comment it looks like that PC is part of network inside Alliance Bernstein. Safe bet the real Mark Carney had nothing to do with it.

    Many people forget when they leave a comment their PCs IP is also registered alongside.

  47. Gravatar of Saturos Saturos
    15. December 2012 at 09:07

    Dilip, or cellphone. And IPs can be scrambled, in case Scott didn’t know.

  48. Gravatar of RebelEconomist RebelEconomist
    15. December 2012 at 11:45

    Wow Scott! Even I could live with this. As you know, my fear about NGDPLT is that the authorities would use the change to get off the hook of above-target inflation mandating unpopular monetary policy tightening, but with a target as low as around 4%, with an adjustment of trend RGDP from time to time (for which purpose persistently over 2% domestic inflation would presumably suggest a downgrade), the authorities would be kept on a short leash.

    Unfortunately, I fear that this means that they won’t go for it. I still expect the UK adopt NGDP targeting, but either with a target more like 5%, or a trend based on pre-crisis NGDP. What you have to understand is that these people are as cunning as cornered rats! (By the way, did I ever mention that I have worked at both HMT and the BoE?)

  49. Gravatar of ssumner ssumner
    15. December 2012 at 15:14

    Could the owners of that company determine which employee sent the comment?

  50. Gravatar of Morgan Warstler Morgan Warstler
    16. December 2012 at 00:23

    Scott, you need to admit outright people who want a raise want it WHILE prices stay the same, they don’t want prices to go up, and then they need a raise just to keep up.

    Remember that who touches the money first thing, that you say doesn’t matter?

    The bond guy MAKES HIS QUOTA when the Fed comes shopping. He doesn’t get fired and have to sell his boat to retiree for 35% of what he paid for it.

    The retiree and the employee who wants the raise have the same motives, more REAL CONSUMPTION for them, less for the bond guy.

    Stop pretending more nominal income is outright viewed as a overall positive.

  51. Gravatar of Bonnie Bonnie
    16. December 2012 at 01:17

    From a public policy point of view, the political debate over NGDPLT vs. inflation targeting is a debate worth having – very publicly – considering the circumstances under which the Great Recession happened and the effects on public debt, spending, the fabric of society, political stability, and peripheral considerations such as it happening during a war. The sooner central bank discretion is locked down to a rules-based regime the better, and inflation targets allow for too much wiggle room to demagogically expand the scope of the central bank’s power. We don’t want the tail wagging the dog.

    It might be tempting to take a short cut by coming up with a way to do NGDPLT without changing mandates, but I think that at least the politicians should understand what went wrong and the full impact to the political system from the choices the central banks made. That is certainly one cat that needs to be let out of the bag. It isn’t the time to get cold feet and leave the current sort of nebulous bunch of discretionary policies stand. You have everything you need to tear that sucker down and ensure that something like the Great Recession never happens again. It would be a shame to miss that opportunity by keeping the failures of the status quo from entering wider public debate and letting these people who have been so incredibly wrong, acting as if they are above the law make all the choices and avoid consequences for what they have done.

  52. Gravatar of W le B W le B
    16. December 2012 at 02:46

    Most helpful, Scott.

  53. Gravatar of Saturos Saturos
    16. December 2012 at 07:39

    Scott, I see, so it’s just NGDPLT dressed up to look like inflation targeting. I’m on board with that!

    Damn, so were those Eliezer Yudkowsky comments fake as well? I was really excited about those…

    For the record, no one is faking this account. This is 100% Saturos, right here 😉

  54. Gravatar of Eliezer Yudkowsky Eliezer Yudkowsky
    16. December 2012 at 21:41

    The Mark Carney comment is almost certainly (a) fake and (b) untraceable. Sadly this is standard Internet behavior, and (’tis but a guess) the commenter probably didn’t expect you to take their joke seriously or realize that you might be disappointed – a serious fake would’ve tried to add more detail, a brief comment like that was probably meant as a joke. Alas.

    Motives to fake an Eliezer Yudkowsky comment are much less clear, but I’ll be happy to answer yudkowsky@gmail if that’s really an issue. 🙂

    (Yes, I read this blog – seems like a strong Correct Contrarian Cluster candidate. Though I’m still trying to figure out what happens if e.g. Zimbabwe adopts NGDPLT tomorrow – it seems like NGDPLT is mathematically consistent with rapidly increasing inflation and rapidly decreasing economic activity, but I don’t know if this is economically plausible.)

  55. Gravatar of ssumner ssumner
    17. December 2012 at 07:59

    Saturos, What this exercise really shows is that even with NGDP targeting the long run inflation rate would be pretty stable. There just isn’t that much variation in long term RGDP growth rates. Do it in per capita terms (which is preferable) and they become even closer.

    Eliezer, Thanks for the info. I can’t do my job if I can’t trust emails to be legit. I just flew half way around the world to give a talk in Singapore. What if that invitation had been faked? Sending out phony emails impersonating someone else should be a crime, and anyone doing so from a company computer should be fired.

    And the weirdest thing of all is that it wasn’t even funny–you’d think someone doing so would at least create a zany message.

    Monetary policy can’t solve real problems of the sort that Zimbabwe faced. But as a practical matter you never get real GDP delines big enough where a NGDP target would produce hyperinflation. What NGDP stability does is prevent the monetary authority from adding insult to injury, by piling demand-side unemployment on top of supply-side unemployment.

    And if it’s really you, I am honored to hear that you read my blog.

  56. Gravatar of Major_Freedom Major_Freedom
    17. December 2012 at 11:04

    Sending out phony emails impersonating someone else should be a crime, and anyone doing so from a company computer should be fired.

    While I don’t condone fake email messaging, I’ll just say this:

    It needn’t be a crime if we started calling it “stimulative communication policy”.

    Think about it. Every recession is marked by a pronounced decline in communications (e.g. spending communications), right? It’s clearly the cause of recessions. For imagine if the government shut down the internet! Or the phone lines! It would cause increasing unemployment and declining output. So clearly I’m not off base for saying that negative “communication shocks” can be causes of recessions.

    Back to fake emails, the central banks can just start sending out fake emails to everyone during periods of lulls in communications, coaxing people into doing things they otherwise would not have done, and with enough communications of people and resulting movements of bodies and resources, maybe we can “get the economy moving again”. We can reduce the stimulative communication policy later on when the economy heals and communications in the market rise back up.

    Yes, fine, there will be some Cantillon effect of misleading people with this stimulative communicative policy. But these costs in the macro context will be more than outweighed by the gains. Part of living in society is to be sacrificed for the benefit of others. Just consider all the stimulus to employment and output that would result from this communication stimulus! If your Singapore email was indeed fake, and you flew to Singapore anyway, only to find that the market you thought was there, was not there, then don’t fret! You have to think in terms of macro. Such problems can be avoided if the Singaporean government also sent fake emails to Singaporeans saying you were coming to teach them about…I don’t know, how to construct solar cells. If that doesn’t work, then send more fake emails!

    Look at what a fake email would do. Your actions boosted the airline industry. The Singaporeans actions boosted the local transportation industry. How could you possibly be against it? If you were mislead by the email, then you should have used EMH and discerned which emails are valid and which emails are done by the policy makers. I mean, if you want to guard against being mislead, then you can invest in IP lookup programs. It doesn’t matter WHO gets the fake emails. It matters WHAT the fake emailers are saying.

    Why are you being such an ideologue that you want to put the stimulative communication policy makers in a cage? They’re just working with what they have. Their actions are boosting communications, the diminutions of which, as proved above, are causes of recessions. If their emails are misleading you into doing things you otherwise would not have done, then you have to stop thinking of yourself only, and consider the “macro” implications instead. Fallacy of composition and all that. What is bad for you is good for the economy as a whole.

    Yes, maybe “stimulative communications policy” is a very small fringe theory right now, but with enough blogging, we’ll make it a reality! Who’s with me!

  57. Gravatar of Eliezer Yudkowsky Eliezer Yudkowsky
    18. December 2012 at 00:50


    When you *send* an email, it almost always goes to the person intended. If you *get* an email saying, there’s indeed a possibility it might not be me. If you reply, and then I reply again quoting what you just said to me, then that’s much harder to fake.

    (Roughly, what goes on is that the email network is a relic of the earliest days of the Internet when there wasn’t much security on anything, and it’s made up of computers forwarding email to each other. When computer A calls up computer B and says it has email from Mark Carney for Scott Sumner, computer B will look up where Scott Sumner is, and then deliver this email, with its fake From line, to Scott Sumner. But when you tell your computer C to deliver an email to Mark Carney (or rather his email address), that email will indeed go to B and B will again look up Mark Carney and faithfully deliver it to him – there’s no way for Evil Computer A to reach out and pluck it from the ether, unless they’re the NSA or they’ve guessed your account password. So if Mark Carney replies to your reply, you can basically trust that it’s him, unless he tells you to send money to Nigeria in which case you should probably start being suspicious again.)

  58. Gravatar of Felipe Felipe
    18. December 2012 at 11:26

    Scott: that would depend on the IT policies of the company in question. It seems unlikely that, even if they had the information, they would share it with you.

    Impersonation is a crime, at least in the US (and several other places).

Leave a Reply