Basil Halperin’s critique of NGDP targeting

Lots of people have tried to find flaws in NGDP targeting, but most of these posts are written by people who have not done their homework.  Basil Halperin is an exception.  Back in January 2015 he wrote a very long and thoughtful critique of NGDP targeting.  A commenter recently reminded me that I had planned to address his arguments.  Here’s Basil:

Remember that nominal GDP growth (in the limit) is equal to inflation plus real GDP growth. Consider a hypothetical economy where market monetarism has triumphed, and the Fed maintains a target path for NGDP growing annually at 5% (perhaps even with the help of a NGDP futures market). The economy has been humming along at 3% RGDP growth, which is the potential growth rate, and 2% inflation for (say) a decade or two. Everything is hunky dory.

But then – the potential growth rate of the economy drops to 2% due to structural (i.e., supply side) factors, and potential growth will be at this rate for the foreseeable future.

Perhaps there has been a large drop in the birth rate, shrinking the labor force. Perhaps a newly elected government has just pushed through a smorgasbord of measures that reduce the incentive to work and to invest in capital. Perhaps, most plausibly (and worrisomely!) of all, the rate of innovation has simply dropped significantly.

In this market monetarist fantasy world, the Fed maintains the 5% NGDP path. But maintaining 5% NGDP growth with potential real GDP growth at 2% means 3% steady state inflation! Not good. And we can imagine even more dramatic cases.

Actually it is good.  Market monetarists believe that inflation doesn’t matter, and that NGDP growth is “the real thing”.  Our textbooks are full of explanations of why higher and unstable inflation (or deflation) is a bad thing, but in almost every case the problem is more closely associated with high and unstable NGDP growth (or falling NGDP).  In most cases it would be entirely appropriate if trend inflation rose 1% because trend growth fell by 1%.  That’s because what you really want is stability in the labor market.  If productivity growth slows then real wage growth must also slow.  But nominal wages are sticky, so it’s easier to get the required adjustment via higher inflation (and steady nominal wage growth) as compared to slower nominal wage growth.

I said “most cases” because there is one exception to this argument.  Suppose trend growth slows because labor force growth slows.  In that case then in order to keep nominal wages growing at a steady rate, you’d want NGDP growth to slow at the same rate that labor force growth slows.  As a practical matter it would be very easy to gradually adjust the NGDP growth target for changes in labor force growth. I’d have the Fed estimate the growth rate every few years, and nudge the NGDP target path up or down slightly in response to those changes.  Yes, that introduces a tiny bit of discretion.  But when you compare it to the actual fluctuations in NGDP growth, the problem would be trivial.  I’d guess that every three years or so the expected growth rate of the labor force would be adjusted a few tenths of a percent.  Even if the Fed got it wrong, the mistake would be far to small to create a business cycle.

Say a time machine transports Scott Sumner back to 1980 Tokyo: a chance to prevent Japan’s Lost Decade! Bank of Japan officials are quickly convinced to adopt an NGDP target of 9.5%, the rationale behind this specific number being that the average real growth in the 1960s and 70s was 7.5%, plus a 2% implicit inflation target.

Thirty years later, trend real GDP in Japan is around 0.0%, by Sumner’s (offhand) estimation and I don’t doubt it. Had the BOJ maintained the 9.5% NGDP target in this alternate timeline, Japan would be seeing something like 9.5% inflation today.

Counterfactuals are hard: of course much else would have changed had the BOJ been implementing NGDPLT for over 30 years, perhaps including the trend rate of growth. But to a first approximation, the inflation rate would certainly be approaching 10%.

[Basil then discusses similar scenarios for China and France.]

Basil’s mistake here is assuming that there is a 2% inflation target.  As George Selgin showed in his book ‘Less than Zero”, deflation is appropriate when there is very fast productivity growth.  Isn’t deflation contractionary?  No, that’s reasoning from a price change.  Deflation is contractionary if caused by falling NGDP.  But if NGDP (or NGDP/person) is growing at an adequate rate, then deflation is an appropriate response to fast productivity growth.  Indeed if you kept inflation at 2% when productivity growth was high, then the labor market could overheat. (See the U.S., 1999-2000).

Let’s suppose that the Japanese decide to target NGDP growth at 3% plus or minus changes in the working age population.  In that case, the target might have been 5% in the booming 1960s, and 2% today (assuming labor force growth fell from 2% to minus 1%.  Or they might have chosen 4% per person, in which case NGDP growth would have slowed from 6% to 3%.  In the first scenario, Japan would have gone from minus 2.5% inflation to about 1%, whereas in the second scenario inflation would have risen from minus 1.5% to about 2%.  Either of those outcomes would be perfectly fine.

As an aside, I recommend that countries pick an NGDP growth target higher enough so that their interest rates are not at the zero bound.  But that’s not essential; it just saves on borrowing costs for the government.

Basil does correctly note that New Keynesian advocates of NGDP targeting don’t agree with market monetarists (or with George Selgin):

Indeed, Woodford writes in his Jackson Hole paper, “It is surely true – and not just in the special model of Eggertsson and Woodford – that if consensus could be reached about the path of potential output, it would be desirable in principle to adjust the target path for nominal GDP to account for variations over time in the growth of potential.” (p. 46-7) Miles Kimball notes the same argument: in the New Keynesian framework, an NGDP target rate should be adjusted for changes in potential.

Basil points out that this would require a structural model:

For the Fed to be able to change its NGDP target to match the changing structural growth rate of the economy, it needs a structural model that describes how the economy behaves. This is the practical issue facing NGDP targeting (level or rate). However, the quest for an accurate structural model of the macroeconomy is an impossible pipe dream: the economy is simply too complex. There is no reason to think that the Fed’s structural model could do a good job predicting technological progress. And under NGDP targeting, the Fed would be entirely dependent on that structural model.

Ironically, two of Scott Sumner’s big papers on futures market targeting are titled, “Velocity Futures Markets: Does the Fed Need a Structural Model?” with Aaron Jackson (their answer: no), and “Let a Thousand Models Bloom: The Advantages of Making the FOMC a Truly ‘Open Market’”.

In these, Sumner makes the case for tying monetary policy to a prediction market, and in this way having the Fed adopt the market consensus model of the economy as its model of the economy, instead of using an internal structural model. Since the price mechanism is, in general, extremely good at aggregating disperse information, this model would outperform anything internally developed by our friends at the Federal Reserve Board.

If the Fed had to rely on an internal structural model adjust the NGDP target to match structural shifts in potential growth, this elegance would be completely lost! But it’s more than just a loss in elegance: it’s a huge roadblock to effective monetary policymaking, since the accuracy of said model would be highly questionable.

I’ve already indicated that I don’t think the NGDP target needs to be adjusted, or if it does only in response to working age population changes, which are pretty easy to forecast.  But I’d go even further.  I’d argue that the Woodford/Eggertsson/Kimball approach is quite feasible, and would work almost as well as my preferred system.  The reason is simple; business cycles represent a far greater challenge than shifts in the trend rate of output.  Because NGDP growth is what matters for cyclical stability, it doesn’t matter if inflation is somewhat unstable at cyclical frequencies.  That’s a feature, not a bug.  And longer-term changes in trend growth tend to be pretty gradual.  In the US, trend growth was about 3% during the entire 20th century.  Since 2000, trend growth has been gradually slowing, for two reasons:

1.  The growth in the working age population is slowing.

2.  Productivity growth is also slowing.

Experts now believe the new trend is 2%, or slightly lower.  I think it’s more like 1.5%.  But I fail to see how this would add lots of discretion to the system. Imagine if the Fed targets NGDP growth at 5% throughout the entire 20th century, using my 4% to 6% NGDP futures guardrails.  No Great Depression, no Great Inflation, no Great Recession.  Then we go into the 21st century, and the Fed gradually reduces the target to 4.5%, then to 4.0%.  And let’s use the worst case, where the Fed is slow to recognize that trend growth has slowed.  So you have slightly higher than desired inflation during that recognition lag.  But also recall that only NKs like Woodfood, Eggertsson and Kimball think that’s a problem.  Market monetarists and George Selgin think inflation should vary as growth rates vary.

Who’s opinion are you going to trust?  (Don’t answer that.)

Seriously, even in the worst case, this system produces macro instability that is utterly trivial compared to what we’ve actually experienced.  Or at least if we hit our targets it’s highly successful.  And Basil is questioning the target, not the Fed’s ability to hit the target.  You would have had 117 years with only one significant alteration in the target path.  Yes, for almost any other country, the results would be far worse.  But that’s why you don’t want to adjust the NGDP target for changes in trend RGDP growth.

Further, level targeting exacerbates this entire issue. . . . For instance, say the Fed had adopted a 5% NGDP level target in 2005, which it maintained successfully in 2006 and 2007. Then, say, a massive crisis hits in 2008, and the Fed misses its target for say three years running. By 2011, it looks like the structural growth rate of the economy has also slowed. Now, agents in the economy have to wonder: is the Fed going to try to return to its 5% NGDP path? Or is it going to shift down to a 4.5% path and not go back all the way? And will that new path have as a base year 2011? Or will it be 2008?

Under level targeting there is no base drift.  So you try to come up to the previous trend line.  In 2011 you set a new 4.5% line going forward, but until you change that trend line, the existing 5% trend line still holds.  If you drop the growth path to 4.5% in 2011, then by 2013 the target for NGDP will be 1% less than people would have expected in 2008, and by 2015 it will be 2% less.  In fact, NGDP was more like 10% less than people expected.  So even if a gradually adjusting path is not ideal, it’s a compromise worth making to satisfy the NKs who are far more influential than I am, but have yet to read Less Than Zero.  (George may not agree with the compromise, he’s less wimpy than I am.)

Before I close this out, let me anticipate four possible responses.

1. NGDP variability is more important than inflation variability

Nick Rowe makes this argument here and Sumner also does sort of here. Ultimately, I think this is a good point, because of the problem of incomplete financial markets described by Koenig (2013) and Sheedy (2014): debt is priced in fixed nominal terms, and thus ability to repay is dependent on nominal incomes.

Nevertheless, just because NGDP targeting has other good things going for it does not resolve the fact that if the potential growth rate changes, the long run inflation rate would be higher. This is welfare-reducing for all the standard reasons.

The “standard reasons” are wrong.  The biggest cost of inflation, by far, is excess taxation of capital income.  That’s better proxied by NGDP growth than inflation. Things like “menu costs” are essentially unrelated to inflation as measured by the government.  The PCE doesn’t measure the average amount that the price of “stuff” changes; it measures the average amount by which the price of “quality-adjusted stuff” changes.  Hedonics.  If the government were serious about targeting inflation, they’d need to come up with an inflation measure that actually matches the supposed welfare costs of inflation in the textbooks.  We don’t have that.  We have nonsense like “rental equivalent”. The standard welfare costs also ignore the massive costs of nominal wage stickiness.  And Basil mentions the incomplete financial markets problem.  Please, can macroeconomists stop talking about inflation, and use NGDP growth as their nominal indicator?  It would make life much simpler.

2. Target NGDP per capita instead!

You might argue that if the most significant reason that the structural growth rate could fluctuate is changing population growth, then the Fed should just target NGDP per capita. Indeed, Scott Sumner has often mentioned that he actually would prefer an NGDP per capita target. To be frank, I think this is an even worse idea! This would require the Fed to have a long term structural model of demographics, which is just a terrible prospect to imagine.

Actually, it’s pretty easy to predict changes in working age population, because we know how many 64 year olds will turn 65, and we know how many 17 year olds will turn 18.  And immigration doesn’t vary much from year to year.  The Fed doesn’t need long range forecasts; three years out would be plenty.  As long as the market understands that the NGDP target path will be gradually adjusted for population growth, they can form their own forecasts when making decisions like buying 30-year bonds.

I want to support NGDPLT: it is probably superior to price level or inflation targeting anyway, because of the incomplete markets issue. But unless there is a solution to this critique that I am missing, I am not sure that NGDP targeting is a sustainable policy for the long term, let alone the end of monetary history.

I still think that NGDPLT, combined with guardrails is the end of macroeconomics as we know it.  All that would be left is discussions of supply-side policies to boost long-term growth. The freshman econ sequence could be reduced to one semester. Or better yet a full year, with a more in depth discussion of micro.

PS.  In this new Econlog post I make some forecasts.



53 Responses to “Basil Halperin’s critique of NGDP targeting”

  1. Gravatar of Benjamin Cole Benjamin Cole
    22. February 2017 at 16:13

    Excellent blogging!

    Yes inflation at 2% to 3% used to be the US norm associated with robust real GDP growth.

    Only quibble: Labor participation rates, and productivity, are malleable. Strong demand probably improves both.

    Err on the upside!

  2. Gravatar of John Hall John Hall
    22. February 2017 at 16:19

    Great post!

  3. Gravatar of dtoh dtoh
    22. February 2017 at 16:51

    Excellent post.

    1. I don’t think anybody knows what optimal trend inflation is (other than no hyper-inflation and avoid sticky wages.)

    2. So IMHO you don’t really need to worry about adjusting for productivity growth, population growth, LFPR, etc. Might be an interesting topic for pundits and academics, but under NGDPLT, it would not be an important real life practical problem.

    3. The FOMC could just meet in a room once very 4 years and flip a coin (er.. I mean have a serious discussion and adjust the NGDP policy rate.) With NGDPLT, they won’t have anything else to do.

  4. Gravatar of Ray Lopez Ray Lopez
    22. February 2017 at 17:13

    Lousy post. Strawmen set up and ‘defeated’ by our host. Apparently our host thinks stability is a big deal, when it’s not. See below, ‘off-topic 2’ on how Moldova lost 1/8th of their money and actually grew. Brazil after WWII had high-teens inflation and also grew. In both cases (deflation and inflation shocks) there were no bad effects. NGDP = RGDP + inflation is a backwards-looking accounting identity and has no predictive value. Read that again slowly. It’s an accounting identity: X = X. You don’t win arguments (unless you’re Sumner) saying “X = X”. The best our Fed can do is like D. Luskin suggests below and simply adopt a sort of ‘real-bills doctrine’ and try and track the natural rate of inflation. (While money is neutral, nevertheless gradually changing prices are on balance a good thing, less need for indexing inflation and other such minor hassles). – RL

    Off-topic 1: Influential Monday WSJ op-ed by Donald Luskin that Trump simply order the Fed to follow Knut Wicksell’s R-star (let the Fed follow the free market interest rate) is an implicit acknowledgement by Luskin that money is neutral.

    Off-topic 2, money neutrality: Moldova had 1/8th of their money supply stolen in 2014 yet no real affect on the economy. This week’s Economist on Moldova: “A leaked report revealed that up to $1bn, equivalent to more than an eighth of the country’s GDP, had been stolen from three banks [in 2014] … yet since then the country has coped remarkably well. GDP shrank by a mere 0.5% in 2015 (whereas Russia’s fell by 4% and Ukraine’s tanked by10%). Last year Moldova grew by 2%, fast by European standards.”

    How much evidence do you need for money neutrality?

  5. Gravatar of ssumner ssumner
    22. February 2017 at 17:20

    Ray, When money is stolen it doesn’t mean the money supply goes down.

  6. Gravatar of morgan warstler morgan warstler
    22. February 2017 at 17:49

    WAIT, let’s back up.

    We’ve switched to a market monetary system and things have been going for quite some time.


    The level target has been hit BAM, BAM, BAM on the GDP predicted and contracted on by the players in the Economy for every given month, with only slight variations – sometimes a couple Billion higher, sometimes a couple Billions lower.




    “Perhaps there has been a large drop in the birth rate, shrinking the labor force. Perhaps a newly elected government has just pushed through a smorgasbord of measures that reduce the incentive to work and to invest in capital. Perhaps, most plausibly (and worrisomely!) of all, the rate of innovation has simply dropped significantly.”

    Now wait a minute….


    To hit the damn target, month after month, the fed has had to be NON-POLITICAL.

    Remember, when we adopt NGDLT, who is responsible for too much inflation and not enough growth?

    The FISCAL party – the Congress, the POTUS, etc.

    And for years and years the brutal NGDPLT machine has forced them all to OWN THEIR FAILURES.

    A shitty fiscal govt? Has 1% growth and 4% inflation.

    A Great one? has no inflation and 5% RGDP.


    So NOW suddenly after years of all these different elected governments having to to eat their failures and run on their successes…


    We jump into a magical model where suddenly

    “A newly elected government has just pushed through a smorgasbord of measures that reduce the incentive to work and to invest in capital. Perhaps, most plausibly (and worrisomely!) of all, the rate of innovation has simply dropped significantly.”

    And this is govt is SPECIAL???



    Underneath this is the very basic question of what is gained by the fed losing discretion?

    Because we can have level inflation targeting or NGDPLT – and once you go level, we’re basically on Friedman’s PC.

    The point of loss of discretion is twofold:
    1) to keep idiots from doing stupid stuff.

    We do not want DRUDGE coming after Yellen. We do not want Krugman OR SUMNER (now you weirdo) insinuating the Plutocrats are trying to keep trump re-elected.




  7. Gravatar of Major-Freedom Major-Freedom
    22. February 2017 at 17:53

    NGDP targeting can be debunked in 4 words:

    Heterogeneous effects from inflation.

  8. Gravatar of Jeff G. Jeff G.
    22. February 2017 at 17:54

    When did this become a monetary policy blog? Where’s the deranged Trump-bashing? (I kid of course, great post!)

  9. Gravatar of Benjamin Cole Benjamin Cole
    22. February 2017 at 18:33

    Jeff G:

    Okay, here is what I say: For guidance on monetary policy, ask yourself, “WWTD*?”

    What Would Trump Do?

  10. Gravatar of Jerry Brown Jerry Brown
    22. February 2017 at 20:51

    Great post- Thanks! It cleared up many of my questions.

  11. Gravatar of Doug M Doug M
    22. February 2017 at 21:44

    “But maintaining 5% NGDP growth with potential real GDP growth at 2% means 3% steady state inflation.”

    A few years of inflation at 3% is not a big deal. We are hardly talking hyper-inflation. 20 years of 3% inflation is slightly bigger. Despite claims to the contrary on this blog, some things are pegged in nominal dollars and money is not neutral.

    However, in the NGDPLT even if the design is that the Fed has no discretion, the Fed still has discretion. If, over the long-term, its models for growth are clearly wrong, the Fed can abandon or change any target it sets for itself.

    I agree with what you have to say along the lines that deflation is not a problem in periods of high productivity growth.

  12. Gravatar of Maurizio Maurizio
    23. February 2017 at 00:20

    “The biggest cost of inflation, by far, is excess taxation of capital income. ”

    Is that why high inflation countries experience capital flight?

  13. Gravatar of Lorenzo from Oz Lorenzo from Oz
    23. February 2017 at 05:03

    Maurizio: I would say, yes, in part–and because a government which is so dysfunctional that it has to significantly rely on the inflation tax will tend to be bad at quite a lot of things.

  14. Gravatar of Lorenzo from Oz Lorenzo from Oz
    23. February 2017 at 05:03

    Scott: very clear and informative post.

  15. Gravatar of Ray Lopez Ray Lopez
    23. February 2017 at 07:09

    Fake news from Sumner who says: “Ray, When money is stolen it doesn’t mean the money supply goes down.”

    Yes it does. We’ve had this conversation before, as you don’t seem to think the Fed buying bogus commercial paper has any bad effect on the economy. For example, when the Fed triples its base money from $1T to $4T, as it did from 2008 to now, the money supply changes (goes up). Likewise, if an agent –call it the Theft Fairy– were to destroy every eighth note by shredding it, as de facto happened in Moldova (see my prior post for you readers who don’t read me but should),then the money supply changes (goes down).

    You don’t have to lie to beat me, you have natural advantages, but it’s Sad! that you don’t use them. Cryptic one line appeals to authority is all you got?

    PS–when do I get ostracized aka the silent treatment from you, like Major Freedom does? Then I know I’ve won. But I’m getting there… Oh, the stock market is up. Trump has, and worldwide this is true, increased animal spirits (started trending up worldwide when Trump ran for presidency, ditto PH’s Duterte, much as I h8 both). Just like FDR’s fireside chats increased animal spirits and bottomed the Great Depression in 1933/34, with a confounding variable being devaluing the gold dollar, as you (don’t) know.

  16. Gravatar of Ray Lopez Ray Lopez
    23. February 2017 at 07:15

    PPS– (Sumner does this, why can’t I?): an economist, I think it was the Freakonomics guy, once said that if a billionaire was to take his entire say $10B net worth and convert it to cash, or say 100 billionaires converting $1T net worth into cash, and then burn up the cash in a bonfire of vanity, the net effect of this is to lower prices somewhat for the entire population (money supply decreases). Yet if you read Sumner’s reply to me above, you would think the billionaire(s) action has no effect, which defies logic. Defies logic, sounds like our host. Oh wait! I forgot, I reasoned from a price change (hocus-pocus catch phrase that means nothing).

  17. Gravatar of LK Beland LK Beland
    23. February 2017 at 07:45

    I still feel that GDP (nominal or not) is fairly difficult and long to compute. Wages seem like a more appropriate target, as obtaining high-quality and high-frequency data is easier.

    The issue I see is that while targeting wages most likely solves the cyclical unemployment problem (i.e. sticky wages), it does not directly solve the cyclical debt problem (i.e. nominal debt contracts). That said, I would guess that wages and NGDP are correlated well enough that the debt problem would not be much of an issue.

  18. Gravatar of Philo Philo
    23. February 2017 at 08:35

    The Fed could, in effect, delegate the task of estimating the future size of the labor force to the market: the so-called “NGDP futures market” that you call for should actually deal in futures on *NGDP/worker*.

  19. Gravatar of Randomize Randomize
    23. February 2017 at 11:15

    If I might suggest a marketing improvement, it’d be better to call the target “Income Per Person” rather than NGDP/worker. Since NGDP is fungible to NGDI, it should be the same result but a much more appealing and widely understandable concept.

  20. Gravatar of Randomize Randomize
    23. February 2017 at 11:17

    LK Beland, wages are only half the formula. As we saw during the recession, wages were actually rising while income fell because hours worked were falling dramatically. What do you care more about: your hourly rate or your monthly income? Which does your mortgage depend on?

  21. Gravatar of Basil Halperin Basil Halperin
    23. February 2017 at 13:04


    Thanks for the detailed and interesting response. I’ll put together a response of my own in the next few days when I have a chance.

  22. Gravatar of jj jj
    23. February 2017 at 13:36

    To avoid causing business cycles, you just need the NGDP target path to adjust more slowly than the time horizon of market actors. Of course there are many relevant horizons: sticky wages has one, every contract has another, business investment decisions have their own… I’m guessing something like 10 years is pretty safe.

    If the NGDP target path adjusts too quickly, you’re back to causing business cycles. If it adjusts too slowly, inflation will be biased from where you want it. But an extra point of inflation is a trivial problem, so I’d err on the side of too slowly.

  23. Gravatar of James Alexander James Alexander
    23. February 2017 at 13:53

    If you are really convinced by Level Targeting why have you stopped berating the Fed about the current rate of NGDP Growth and think they have monetary policy set just about right?

  24. Gravatar of bill bill
    23. February 2017 at 14:26

    One thing I like about NGDPLT is that the consequences of choosing a slightly less than optimal NGDPLT target rate are very minor compared to the results of getting NAIRU wrong. Halperin seems to think that choosing the proper NGDP growth rate is really hard, but in fact, it’s simpler than trying to find the ever changing rate of NAIRU.

  25. Gravatar of George Selgin George Selgin
    23. February 2017 at 17:01

    “Maintaining 5% NGDP growth with potential real GDP growth at 2% means 3% steady state inflation! Not good. And we can imagine even more dramatic cases.”

    This conclusion by Halperin really disappoints me: I thought he got it! Steadily declining TFP growth MEANS steadily increasing unit production costs. That makes the scenario itself unfortunate. But allowing prices to rise 3% per year doesn’t add to the misfortune: it simply reveals the sad reality in a readily-apparent way, instead of disguising it as a percentage-point increase in the rate of steady-state FACTOR PRICE inflation, which is what you get if the monetary authorities insist on sticking to a 2% outputinflation target.

    I frankly am dumbfounded that even clever economists cannot grasp the different implications of demand vs. productivity-driven inflation and deflation. I guess we just haven’t communicated well enough to them, and must keep at it.

  26. Gravatar of LK Beland LK Beland
    23. February 2017 at 17:15

    I should have specified monthly, not hourly wages.

  27. Gravatar of Scott Freelander Scott Freelander
    23. February 2017 at 17:17

    George Selgin,

    I started paying attention to some of your work on free banking recently and am impressed. I hope there’s a way a pilot program can be developed, perhaps in a small country somewhere. Unless I’m mistaken, it would be very difficult if possible at all to run a pilot program in the US.

  28. Gravatar of LK Beland LK Beland
    23. February 2017 at 17:17

    Or, actually, total wages per working age adult.

  29. Gravatar of ssumner ssumner
    23. February 2017 at 17:26

    Jeff G., Thanks for reminding me, I need to get back to Trump bashing.

    Maurizio, I’m not sure. I imagine that really high inflation is correlated with a number of anti-saving policies.

    Ray, You said:

    “PPS– (Sumner does this, why can’t I?): an economist, I think it was the Freakonomics guy, once said that if a billionaire was to take his entire say $10B net worth and convert it to cash, or say 100 billionaires converting $1T net worth into cash, and then burn up the cash in a bonfire of vanity, the net effect of this is to lower prices somewhat for the entire population (money supply decreases). Yet if you read Sumner’s reply to me above, you would think the billionaire(s) action has no effect, which defies logic. Defies logic, sounds like our host. Oh wait! I forgot, I reasoned from a price change (hocus-pocus catch phrase that means nothing).”

    Please find me that post, as it’s wrong. I’d like to do a post. The billionaire would have essentially donated $10 billion to the Treasury. And if someone steals my wallet, the money supply doesn’t change, it’s just held by someone different.

    LK, We are targeting one year forward values in any case.

    Philo, Yes, that’s actually my preferred system, but I’m fine with either.

    Randomize, Yes, “income” is easier for the public to understand.

    Thanks Basil.

    JJ, And again, I think NGDP stability matters more than inflation stability in any case.

    James, Here’s an analogy. Suppose the Fed has a 2% inflation target and I favor a 4% inflation target. If inflation is 3%, what do I do? I recommend tighter money, as the most important thing is that policy be stable and predictable, not that it be my preferred target.

    Right now the Fed does not do level targeting. So my advice is to help them to hit the target that they are in fact trying to hit.

    Bill, I agree.

    George, I think the problem here is that the standard models used today are so different from what you and I are used to that it becomes easy to overlook that distinction. Or to assume that “flexible” inflation targeting solves the problem.

  30. Gravatar of Carl Carl
    23. February 2017 at 17:44

    In times of high productivity growth, the inflation component of the NGDP target will be negative. Does that mean that the central bank won’t be able to conduct NGDP targeting during times of high productivity growth if the country has a free-banking system?

    I ask this question because I assume that a central bank cannot constrain the money supply using open market operations for a prolonged period of high productivity because they would run out of securities to sell. That leaves reserve ratio and discount rate management, but in a free banking regime the central bank doesn’t control those.

  31. Gravatar of Major-Freedom Major-Freedom
    23. February 2017 at 18:13

    NGDP targeting can also be debunked in 2 words:

    Capital malinvestment.

    Actually, one word:


  32. Gravatar of Ray Lopez Ray Lopez
    23. February 2017 at 23:46

    @Sumner – here it is (it was the Armchair Economist, not Freakonomics). Also stealing 1/8th of the money supply ignores that one thief cannot spend all that money (as could millions of Moldovians) hence it must be contractionary at full employment under monetary theory:

    For the purposes of macroeconomics, burning money is equivalent to removing the money from circulation, and locking it away forever; the salient feature is that no one may ever use the money again. Burning money shrinks the money supply, and is therefore a special case of contractionary monetary policy that can be implemented by anyone. In the usual case, the central bank withdraws money from circulation by selling government bonds or foreign currency. The difference with money burning is that the central bank does not have to exchange any assets of value for the money burnt. Money burning is thus equivalent to gifting the money back to the central bank (or other money issuing authority). If the economy is at full employment equilibrium, shrinking the money supply causes deflation (or decreases the rate of inflation), increasing the real value of the money left in circulation.
    Assuming that the burned money is paper money with negligible intrinsic value, no real goods are destroyed, so the overall wealth of the world is unaffected. Instead, all surviving money slightly increases in value; everyone gains wealth in proportion to the amount of money they already hold.[2] Economist Steven Landsburg proposes in The Armchair Economist that burning one’s fortune (in paper money) is a form of philanthropy more egalitarian than deeding it to the United States Treasury.[2]

  33. Gravatar of James Alexander James Alexander
    24. February 2017 at 00:03

    I thought you’d say that. Just helping them do their existing job
    But …
    If NGDPLT is so superior to IT, as we agree on, maybe just helping them isn’t nearly enough? Maybe the existing job is impossible because of the contradictions inherent in IT? Are you just “collaborating” in their continuing failure, and that’s why you now think 3% NGDP growth just about acceptable. Still sounds like you really don’t think LT that important anymore.

  34. Gravatar of bill bill
    24. February 2017 at 06:00

    NGDP per worker. While it makes sense to adjust the NGDP target over time for workforce growth changes, we should not target NGDP/worker per se. During a crisis, NGDP could drop say 4% and if unemployment rose by 4 points, then NGDP/worker would be steady and not call for any loosening. And heaven help us if unemployment rose 5 points and NGDP/worker rose.

  35. Gravatar of LK Beland LK Beland
    24. February 2017 at 06:39

    “LK, We are targeting one year forward values in any case.”
    (In regards to higher quality data and frequency; ngdp vs nwages)
    I would still prefer steering the macro-economy using forward values of a high-quality, high-frequency dataset, especially if level targeting.

  36. Gravatar of mpowell mpowell
    24. February 2017 at 09:14

    I kind of agree with Selgin. Yes there is a lot of thought that went into Halperin’s critique, but it mostly seems to be that he just misses (somehow) perhaps the most basic point. Namely, that NGDP targeting is about targeting NGDP, not inflation, because targeting NGDP will be better for the economy. If Sumner really wanted to target inflation instead, he’d be talking about inflation level targeting and targeting the forecast, etc.

  37. Gravatar of George Selgin George Selgin
    24. February 2017 at 10:19

    “it makes sense to adjust the NGDP target over time for workforce growth changes, we should not target NGDP/worker per se.” Bill, I believe the idea is that the assigned growth rate target itself is meant to allow for the expected rate of growth of the labor force. In my original ’97 proposal, I recommended a spending growth rate equal to the anticipated growth rate of factor (labor and capital) input, weighted by factor shares. In other words, part of the NGDP growth rate target is to compensate for anticipated factor input growth, while the remainder, if there is one, is there to achieve a long-run inflation target.

  38. Gravatar of George Selgin George Selgin
    24. February 2017 at 10:22

    Scott Freelander, RE: pilot trial. “A kingdom, a kingdom; my horse for a Kingdom!”

  39. Gravatar of George Selgin George Selgin
    24. February 2017 at 10:28

    Carl, I have written at length concerning how, in a free banking system, there is a tendency for the money multiplier to adjust to stabilize MV when the base is constant. When productivity grows, on the other hand, the multiplier doesn’t.; So prices fall as a matter of course. For a brief discussion see the appendix of Less than Zero.

    this isn’t to say that a frozen base itself is ideal. The point is that free banking makes it easier rather than harder to maintain a desired NGDP target in any given base regime. There is no such thing as a “free banking” base regime. So in assessing the consequences of free banking one must specify what base regime one has in mind. That is, some B and mechanism for its control must be specified.

  40. Gravatar of ssumner ssumner
    24. February 2017 at 10:48

    Ray, Ah, he also describes it as a gift. And was the theft of money in the form of currency? I doubt it. Bank reserves? If not, the central bank can simply print more money at a trivial cost.

    James, That flawed monetary regime worked pretty well during the Great Moderation. If we get to a point where they get far off course (as in 2008) I will be speaking out loud and clear.

    Bill, Hourly wage growth usually slows if there is a fall in NGDP. But wages can be sticky, which is why I favor NGDP per person or labor force, not employed worker. It gives you a more timely signal than hourly wages.

    LK, Fair point. In the past I’ve worried about the political problems associated with wage targeting. But I’m certainly open to the idea. Also, see my reply to Bill.

    mpowell, In his defense, he certainly understood the cyclical argument, he was disagreeing on the trend inflation issue (which is the place where Woodford, Eggertsson and Kimball also disagree.) Thus Halperin, (who I believe was an undergrad when he wrote the post) had the same view as Woodford, the top monetary economist in the world. I don’t agree, but it’s not a simple mistake.

    Carl, You may want to check out George’s post here:

  41. Gravatar of George Selgin George Selgin
    24. February 2017 at 11:52

    “Market monetarists and George Selgin thin inflation should vary as growth rates vary.” Careful, Scott: should vary with TFP growth rate, not with output growth rate per se. In principle, as you and others observe elsewhere, one wants to allow for changing rates of factor input. But those are much easier to predict than TFP growth. Here some of our critics have it wrong: it’s the inflation targeting that calls for anticipating changes in potential GNP. NGDP growth targeting calls for anticipating changes in factor input, which are less volatile than potential GNP, because the later reflects the combined changes in (relatively predictable) factor input growth and (relatively unpredictable) TFP growth.

    Here again, I can’t help expressing my surprise at the obtuseness of those who thing that targeting NGDP calls for a greater ability to model and predict potential GNP than targeting inflation. It’s not hard: potential GNP growth rate = Factor input growth rate plus TFP growth rate. Long-run inflation rate = NGDP growth rate – potential GNP growth rate = target long run inflation rate – TFP growth rate (since NGDP growth rate is = target inflation rate plus long-run factor input growth rate). To the extent that factor input is incorrectly forecast, the actual inflation rate will also differ from the target – TFP growth rate by the forecast error.

  42. Gravatar of James Alexander James Alexander
    24. February 2017 at 12:07

    “In fact, NGDP was more like 10% less than people expected.”
    10% off course (on your own estimate) not big enough, then?

  43. Gravatar of Blair Blair
    24. February 2017 at 14:42

    My only concern is in the event there is a mistake that ends up with the nominal income index several percentage points above the target level( e.g. due to supply shock). How do you get back on the trendline? Severe Volcker style recession or gradual glide path?

  44. Gravatar of George Selgin George Selgin
    24. February 2017 at 14:53

    Blair, exactly how does a supply shock alter NGDP? Saying it does is like saying that shifting the AS schedule shifts the AD schedule. Why should it? Alternatively, if NGDP = MV = Py, why should an exogenous shock to y (that is, a shock to natural y) lead to a change in Py = MV when Py = MV is, ex hypothesi, the thing that is being kept stable?

    If I could make a wish (make that a monetary economics wish), it would be that no one would reason about these things in print without first checking their logic with either and AS-AD diagram or the equation of exchange, or (best of all) both!

  45. Gravatar of George Selgin George Selgin
    24. February 2017 at 15:03

    To clarify: under NGDP targeting, the CB does NOT have to anticipate or respond to AS shocks; instead, it allows the shocks to raise prices (goods are scarcer and cost more, oh my!). In contrast, under P (or inflation) targeting, it has to respond–and the response, if it is what the rule calls for (M reduction to shrink MV = Py enough to keep P from rising), makes everyone worse off! in one case, fewer good. In the other, fewer goods and a (short-run) money shortage so people can’t even but what goods there are!

  46. Gravatar of Carl Carl
    25. February 2017 at 09:14

    Thanks George Selgin and Scott.

    George’s article helped. I’m not sure if the tools of the central bank–e.g open market operations, reserve ratios–are considered part of the monetary regime. If they are, then the banking system does influence the monetary regime. The central bank can’t, for example, use the reserve ratio tool to hit its NGDP target if the country has a free banking system. A free banking system may lead to a more stable money supply and, therefore, less need for the central bank to use the tools at its discretion, but that’s a separate point. I was just trying to explore how a free banking system and an NGDP targeting central bank monetary regime could interact under different conditions.

  47. Gravatar of George Selgin George Selgin
    25. February 2017 at 09:32

    Carl, central banks have long given up employing changes in mandatory reserve requirements as means for monetary control, for all sorts of good reasons. Under free banking (assuming a fiat regime, of course) they would still be able too employ the tools they rely upon at present, and open-market operations in particular.

  48. Gravatar of ssumner ssumner
    26. February 2017 at 06:16

    George, The only factor I would allow for is labor. That’s because I view sticky wages as the key problem.

    James, Not sure what your question is asking. Would a 5% NGDPLT have been wise in retrospect? Yes.

    Blair, If you are above the trend line the economy would be overheating. Getting back to the trend line would not involve a recession, as long as the policy was credible.

  49. Gravatar of HH HH
    26. February 2017 at 19:06

    “we know how many 64 year olds will turn 65, and we know how many 17 year olds will turn 18”

    No, we don’t.

    I kid, I kid. I know Scott knows we don’t know who’s going to die, and that the margin of error is still very small. I just get so few chances to be pedantic…

  50. Gravatar of George Selgin George Selgin
    27. February 2017 at 06:54

    Scott, I agree that allowing for labor input is more important than allowing for capital input. A case can be made for the last, on the grounds that doing so makes P changes more transparent indicators of changing average unit production costs. That is, while perhaps no better on stickiness grounds, setting NGDP growth to roughly equal capital as well as labor input growth minimizes signal-extraction costs.

  51. Gravatar of ssumner ssumner
    27. February 2017 at 13:32

    George, It’s not clear to me who cares about that particular signal.

  52. Gravatar of Reply to Sumner | Basil Halperin Reply to Sumner | Basil Halperin
    14. March 2017 at 07:23

    […] Sumner has a thoughtful reply to my critique of NGDP targeting from 2015. Allow me to restate my comment from a slightly […]

  53. Gravatar of NGDP targeting and the Friedman Rule | Basil Halperin NGDP targeting and the Friedman Rule | Basil Halperin
    9. April 2017 at 15:58

    […] post continues the discussion from Scott Sumner’s thoughtful reply to my critique of NGDP targeting from 2015. (Note to frequent readers: I previously published a […]

Leave a Reply