Inflation stickiness, demand, and judging the success of monetary policy

I am still a fan of flexible inflation targeting.  I agree with Nick Rowe that explicit inflation targeting has made inflation outcomes “stickier” – and that knowing inflation is in a range of the inflation target is therefore insufficient for telling if the central bank is truly achieving “socially optimal policy”.

For all the time I’ve spent reading monetary economics books and sitting in classes where macroeconomics has been discussed, I still remember a clear (albeit very partial) description of why we do inflation targeting that came in 100 level economics.

I distinctly remember my tutor saying that the point of inflation targeting in the form we use it in New Zealand was to pin down “inflation” (price growth that is shared between all goods and services that is independent of the “relative” value of these goods and services).  With people setting prices based on this view of inflation, central banks with a clear very of the economies “ability to produce” can move around “aggregate demand” in order to prevent recessions that are due to short-falls in demand.  The flipside was often that inflation stickiness was “asymmetric”, with excess demand translating into rising prices, while insufficient demand translated into falling output – this is our good friend the upward sloping, U-shaped, short run AS curve.

This story is massively oversimplified, especially in its treatment of expectations.  But an overarching goal of anchoring inflation expectations was always part of what central banks were aiming to do.  Given this, then in so far an central banks had an idea about economic capacity (which is a very debatable point in itself) they could help to manage “demand” in a macroeconomic sense.  When I was tutoring, this was very much how active monetary policy is taught to first years, and I doubt terribly much has changed.

And this is the point – economists have always know that, if their announcement of a 2% inflation target was the sole determinant of inflation, this does not mean “success” … it just means that they can focus solely more heavily on how the actions of monetary policy have a short-run impact on output.  Deviations of inflation from their target provide information that is useful information about the state of “demand”, but as the NGDP targeting proponents point out it does not capture the whole story.  Variables such as NGDP, and unemployment, provide significant information … something central bankers already recognise and incorporate, contrary to the “narrative” of them being closed minded (Nick isn’t saying this – I’m talking about the more general stories from the media).

This is consistent with flexible inflation targeting – and it does come with one massive hole.  Judging the “success of monetary policy”.  As of course, flexible inflation targeting can only be judged on a forecast, forecasts that are determined by the central banks themselves – and filled with unobservables such as “future economic capacity”.

NGDP targeting differs from this in only three ways:

  1. It changes a partially discretionary rule based process with a fully rule based process.
  2. It targets levels instead of growth rates – making policy “history dependent”.
  3. It gives guidance, and will make “stickier” growth in nominal income – compared to flexible inflation targeting which does this for price growth.

On that final point, at the moment inflation targeting lets a firm say “with competition and the such, I was able to increase my prices 2% this year – this is the same as inflation, and so in reality my “price” is the same”.  With NGDP targeting the firm will say “I increased revenues by 5% this year – this is the same as the growth in nominal spending/value add, as a result my real “revenue” is performing as well as the average firm”.  The right “guidance” will depend strongly on what we think is the most important factor for firms and households to have certainty about (to extract appropriate “market signals”).

To put all this another way – inflation stickiness isn’t an unintended consequence, it is a feature of central banks trying to improve the “sacrifice ratio” associated with active monetary policy!

  • Blair

    As an investor I suppose I would tend to use an inflation assumption more often in valuation models, but the success of the investment always hinges on the revenue line! Pragmatically I’d rather be in Australia where inflation runs a bit hot than in Europe.
    I didn’t follow this bit:”As of course, flexible inflation targeting can only be judged on a forecast, forecasts that are determined by the central banks themselves – and filled with unobservables such as “future economic capacity”.
    I agree that from a theoretical point of view FIT and NGDPLT are very close. I thought the critique of flexible inflation targeting flowed from empirics – in short, most countries running it had AD-induced problems recovering from the GFC. This in turn may stem from the fact that when AD is too high, the inflation rate (relative to target) usually sends the CB a very clear message about what to do, but when AD is too low it sends a weak signal and the CB starts bickering about what to do.
    Coming back to your quote, it seemed there were two related issues with CB behaviour. One, the CB’s forecasts were generally inaccurate, and two, even on their own forecasts, they didn’t seem to be forecasting a return to normal demand-supply balance in any kind of reasonable time frame (this was more apparent with the Fed where they have a dual mandate).
    Even when you have a single target like the ECB, you need to consider whether it’s desirable to have Trichet pointing to the inflation outcome, patting himself on the back and putting the whole load on fiscal policy. Though that’s kind of a political issue.

    • http://tvhe.co.nz/ Matt Nolan

      Hi Blair, good points.

      “As an investor I suppose I would tend to use an inflation assumption
      more often in valuation models, but the success of the investment always
      hinges on the revenue line!”

      Indeed, it is an interesting question. I’m not sure the idea of what CB management of the money supply is supposed to achieve in terms of certainty has been clearly defined – and that has led to discussions being quite all over the place. Originally I studied microeconomics, so that was just the way I saw things – and in many ways I think the NGDP discussion has made that point clearer to people discussing macro.

      ” I thought the critique of flexible inflation targeting flowed from
      empirics – in short, most countries running it had AD-induced problems
      recovering from the GFC. This in turn may stem from the fact that when
      AD is too high, the inflation rate (relative to target) usually sends
      the CB a very clear message about what to do, but when AD is too low it
      sends a weak signal and the CB starts bickering about what to do.”

      Indeed this is very much how things are viewed in terms of the “output gap”. If demand falls below potential, sales/output falls and profit margins rise (assumes an upward sloping cost curve) – thinking this is weird, firms lower prices (relative to the inflation targeted increase) to try and sell more and hit their “targeted margin”. In this envrionment, measured inflation will fall. We cut interest rates to get people to bring “demand forward” and close this output gap – this is very a crude explanation, but its the sort of idea.

      CB’s do this more broadly by “managing expectations about nominal variables” in order to improve outcomes. The idea is that if they are targeting inflation of “X%”, and they forecast that they will meet that target, then they are forecasting that they are, and will, do what needs to be done in terms of demand management. Now forecasts are always wrong, the key point is that they are unbiased and do work as a way of credibly managing expectations. As a result, it makes complete sense to use market measures to judge the stance of policy (relying on the wisdom of the crowds), in fact this is really the most sensible way to do things.

      As you say, at some points in time some CB’s ignored these market signals AND their own models, which didn’t help. Another thing that didn’t help was that the ECB was more incompetent than other central banks had allowed for under Trichet – Trichet forgot that financial stability required a lender of last resort, and he forgot that the advantage of having a well anchored inflation target was lowering the sacrifice ratio … looking only at inflation was never sensible practical policy.

      To use the extreme case that proves the rule – if inflation was fixed in the short term, CB policy can be used directly to fix the “quantity” of output in this short term. Anchoring inflation expectations makes CB’s more able to explicitly deal with demand!

      • Blair

        BTW good article by Brian Fallow today – I think he has been reading your blog.
        One of the reasons I think NGDPLT is interesting because I as a patriotic Kiwi I like to think the savings-investment imbalance in the real economy (or bias against saving) will one day be fixed and I worry that there won’t be enough demand around. (I lived through the late 80s and early 90s recession, and since then have always had the view that if you want to restructure your economy you need to make sure there is plenty of demand around, otherwise your human capital stock will simply decay).
        I remember Lars Christensen saying he thought small open economies should use a NGDPLT with a floating peg exchange rate. That’s one way of doing it, but it’s a very big change. Or, we could take the view, which I think you do, that our current framework can handle it. As long as, one day when the overnight rate hits 0.25%, the RBNZ knows what to do!
        I might have a look through the RBNZ’s speeches for hints about this. But I think their attitude to date has been “we’re miles from the ZLB, so it’s a moot point”.

        • http://tvhe.co.nz/ Matt Nolan

          “BTW good article by Brian Fallow today – I think he has been reading your blog.”

          Hehehe – Brain Fallow’s articles are always great, but today’s was especially good.

          I know for a fact he doesn’t read the blog – so I can’t take any credit! He just uses the same sort of framework that I do, and then communicates it a hell of a lot better :)

          “As long as, one day when the overnight rate hits 0.25%, the RBNZ knows what to do!”

          This is very true. I really like the RBNZ, and think they do heaps of good things. But it would be encouraging to see them announce what they would do with policy when the cash rate hit zero – especially after the confusion we’ve seen around the world.

          I hope they would take the view that we can use forward guidance, and even a commitment to target a level/higher inflation, in the case of hitting the ZLB – but such a mechanism should be put in place and explained before it becomes as event.

  • http://www.facebook.com/people/Tom-Brown/1290763443 Tom Brown

    Matt: question from a layman here just trying to understand what these various policy proposals are. You write, in regards to NGDP targeting:

    “It targets levels instead of growth rates – making policy “history dependent”.”

    1. What do you mean NGDP targets levels? I thought Scott Sumner likes to refer to a 5% growth rate in NGDP as being desirable. That’s a growth rate, and if he’s selected a number (5%) isn’t he talking about targeting that?

    2. Can you expand on what you mean by “history dependent?”

    Thanks

    • http://tvhe.co.nz/ Matt Nolan

      Hi Tom,

      Scott is talking about targeting a “level” of NGDP that grows at 5%pa. So if you only achieve 3% one year, you have to have growth of just over 7%pa to reach your level target.

      If you are targeting the growth rate instead of the level, then if you have 3% one year, you only target 5% the following year and “ignore the shortfall”

      And this is what history dependent means – when you target a “level” you make up for past shortfalls, or tighten policy if you were too loose in the past. With growth targeting you ignore what has happened in the past.

      As a result, when we ask whether to target the level of growth, we need to ask what people use for making decisions etc etc, and that is a pretty complicated question that people will need to think about. Back in the day people decided on growth targeting, but that won’t necessarily always be agreed upon.

      • http://www.facebook.com/people/Tom-Brown/1290763443 Tom Brown

        thanks!