Can Macroeconomic Policy Only Influence The Nominal? A Critique of NGDP Targeting

Update: When I am talking about nominal and real effects, insert the word ‘reliable’ as appropriate. My point was that MMs hold the distribution between RGDP/inflation to be impossible to determine, rather than that there can be no real effects.

At the centre of arguments for NGDP targeting is the implicit assumption that macroeconomic policy can only have nominal impacts. I believe this to be false.

‘Market Monetarists’ (MMs) hold that macroeconomic policy can only have nominal or ‘cash’ effects, and the effect on the actual production of goods and services is uncertain. To put it intuitively, they are saying (or believe they are saying) that the amount of cash we inject into the economy can only affect the amount of cash flowing around the economy. There are two reasons this is flawed.

Firstly, MMs use a standard AD/AS framework for their analysis. When you respond that fiscal stimulus can be spent on roads and the like, which benefits the economy in more ways than one, they suggest that you are confusing the AD side of the equation with the AS one. However, this is a problem for MMs: fiscal stimulus can simultaneously boost AD and AS (not to mention LRAS), which means that even if the CB offsets the increase in spending, fiscal policy can ensure that a higher percentage of the spending is a real increase in production. In other words, fiscal policy can impact RGDP better than monetary stimulus, at least in certain circumstances.

My second point, however, is more important and applies universally. MMs neglect of the role of the RoI in determining not only the rate but the type of macroeconomic activity that takes place (New Keynesians are also guilty of this; at least, I’ve never seen them mention it).

In a capitalist economy, the rate of interest not only determines the amount of investment, but also the nature of that investment. Here’s Adam Smith, whom I can never resist quoting on the subject:

If the legal rate of interest in Great Britain, for example, was fixed so high as eight or ten per cent, the greater part of the money which was to be lent, would be lent to prodigals and projectors, who alone would be willing to give this high interest. Sober people, who will give for the use of money no more than a part of what they are likely to make by the use of it, would not venture into the competition. A great part of the capital of the country would thus be kept out of the hands which were most likely to make a profitable and advantageous use of it, and thrown into the those which were most likely to waste and destroy it.

High interest rates mean that there is less investment and also that the investment that does take place is less sustainable. Speculative investment causes bubbles and price inflation, reducing the CB’s control over both the nominal and real economy. By keeping long term rates low, the CB is hence able to impact the real activity in the economy.

For an example, see this poorly taken photograph of long-term corporate bond yields from Geoff Tily’s excellent book Keynes Betrayed:

High interest rates are correlated with downturns, whilst steady low interest rates are correlated with the ‘Golden Age’. Rates weren’t massive in the recent crisis, though there were plenty of ‘hidden’ sources of funding such as the repo and Eurodollar markets, and expectations were not anchored downwards.

Of course this evidence is far from exhaustive but Tily’s book has much more, and I don’t want to flood this post with too many poorly taken photographs.

If macroeconomic policy can have real effects, that blows a significant hole in the arsenal of NGDP target advocates. For whilst most critiques of NGDP targeting focus on how it isn’t feasible, particularly at the ZLB, this one is compatible with its feasibility, but questions its efficacy.


, ,

  1. #1 by noiselull on January 24, 2012 - 1:27 pm

    So Adam Smith had moronic views regaring usury. What of it? The fallacies of his argument for a legal maximum on interest rates are well-known to most econ students. This has to be one of your worst arguments yet. (FD: I hate NGDP targeting because I’m a Free Banking guy.)

    • #2 by Unlearningecon on January 24, 2012 - 3:46 pm

      I’m not aware of much discussion of Adam Smith’s ‘moronic’ views in econ courses, at least not in non-ideologically charged ones.

      Also you haven’t engaged my points at all.

  2. #3 by Leigh Caldwell (@leighblue) on January 24, 2012 - 2:28 pm

    Scott Sumner definitely agrees that monetary policy can have real effects. So do, I think, all market monetarists – otherwise there’d be no point in pursuing the policy.

    There is disagreement over the mechanism whereby this happens. Some say it works because of sticky prices & wages, some because of the need to satisfy a desired increase in saving while avoiding the paradox of thrift, and some because a reduction in nominal expected returns (due to falling expected inflation) also reduces real expected returns while pushing the real interest rate up. But almost nobody disagrees that monetary targets (whether NGDP or inflation) make a short-run difference to the real economy. (There is, however, a mainstream belief that money is neutral in the long run – is this what you are referring to? Only Austrian economists argue from this that there should be no monetary targeting at all.)

    You can look at it like this: nominal problems (unexpected deflation, sticky wages) definitely cause real problems; so monetary policy is largely aimed at fixing these nominal problems and therefore unblocking the real economy.

    As you say, monetary policy can probably also affect the prices and quantities of real investment assets. This has certainly been discussed in the mainstream economic world – I remember in 2007 seeing arguments about whether the central bank should (or could) act to reduce asset prices even though consumer price inflation was under control – but I don’t think the mechanisms for this are well-understood or agreed upon. At any rate I haven’t seen a convincing study of it.

    Supporters of monetarism believe that, regardless of these effects, monetary policy is still much less distorting than fiscal policy. I tentatively agree with that view, but this does not mean there is no place for fiscal policy. Fiscal policy is of course designed specifically to have real effects, and as a society we may well choose that we want those effects.

    • #4 by Unlearningecon on January 25, 2012 - 1:15 pm

      Sorry, I should have written ‘reliable’ nominal/real effects throughout the post – MMs hold that it’s difficult to determine the impact of macroeconomic policy on RGDP.

      I disagree with two key points:

      – Sticky wages/prices explain downturns (I have a previous post on this here).

      – Money is neutral. I could go into the theory of liquidity preference and so forth but I’m sure you’ve heard it all before. In any case, empirical evidence has the final say, and professor Keen’s empirical link between the growth of private debt and unemployment/RGDP is too robust to ignore.

      Keynes discusses the impact of speculation in TGT, as does Adam Smith in TWN. Sorry to keep linking you to lengthy posts/papers, but read here if you’re interested.

      • #5 by Leigh Caldwell (@leighblue) on January 25, 2012 - 3:06 pm


        On the sticky wages/prices point: I agree that cutting all wages across the economy would have aggregate demand effects which are damaging. However the standard sticky wages argument is more subtle than this: it says not that the _average_ wage is sticky, but that _relative_ wages are sticky.

        Sometimes relative changes in wage are called for by demand changes. Imagine society’s need for computer programmers rises and the need for bricklayers falls. Then the wage for programmers will rise, and the wage for bricklayers _should_ fall…but often won’t, because of nominal stickiness. If this is the case, some bricklayers will be unemployed (though computer programmers probably won’t be).

        Inflation allows real wages in low-demand sectors to fall more easily, while real wages in high-demand sectors can keep rising. So it allows relative wages (and prices) across the economy to adjust, which leads to growth.

        (Some neoclassical economists will argue that a general cut in wages is required to rebalance the relative prices of labour versus money or bonds, but most recognise that this wouldn’t work in practice, and market monetarists are certainly not suggesting it)

        I don’t claim that this stickiness explains everything about downturns but it is plausible that it is part of the issue.

        Money is definitely not neutral in the short run and few would argue that it is. In the long-run…well, I agree with your overall point about the lack of an equilibrium, so the theory of the long run probably doesn’t matter.

        I don’t have time to read that whole Keen paper but I find it very plausible that growth in private debt _is_ indeed linked to RGDP, so no dispute there [though I suspect there is an open question about what happens when the gross debt load is reduced by inflation, rather than by deleveraging].

        However I don’t think this conflicts in any way with market monetarism. Increasing NGDP expectations is not intended to substitute for consumers taking out private debt; it is just meant to make it easier for them to repay it. So Keen and Sumner can both be right.

      • #6 by Unlearningecon on January 25, 2012 - 3:49 pm

        ‘Then the wage for programmers will rise, and the wage for bricklayers _should_ fall…but often won’t, because of nominal stickiness. If this is the case, some bricklayers will be unemployed (though computer programmers probably won’t be).’

        FYI I largely reject the neoclassical model of the wage/labour market (supply-demand, excessive wages necessarily cause unemployment). In fact, I’d argue that a reduced demand for bricklayers should not result in a lower wage but fewer bricklayers, as fewer bricks are required to be laid!

        Overall I’m not sure MM is compatible with endogenous money. Keen and others argue that the CB controls M0 and only M0 (I align myself with them because it appears to be well documented empirically) whilst credit and others are endogenously determined and do not rely on base money decisions made by the fed, certainly in the short run.

        Also, it doesn’t seem to be compatible with my overall point. Rising interest rates would make an MM say ‘look! Loose monetary policy!’ but it would make an (old/post-Keynesian) say ‘watch out! Get those rates down!’

        I have yet to see any discussion of the impact of the RoI on the type of investment that takes place, though perhaps that is just ignorance on my part.

  3. #7 by atm0spheric on January 24, 2012 - 5:21 pm

    It doesn’t matter very much which gage you are watching as you waggle your lever if you only have the throttle in your hand, and a bunch of politicians are arguing about how the rudder works. Sadly, politicians have discovered that it is infinitely easier to stop things happening than to make them happen.

    By definition, macroeconomic policy can only be an enabling or a disabling mechanism; It doesn’t make things happen.

    If you want to make things happen, you have to make them happen!

    • #8 by Unlearningecon on January 24, 2012 - 11:36 pm

      You’re a bit of an enigma, atm0spheric. I’m never sure if you’re agreeing, disagreeing or just offering a related viewpoint.

  4. #9 by atm0spheric on January 25, 2012 - 9:27 am

    I question whether the choice between targeting inflation and targeting NGDP is an oversimplification. I am all for simplification as an aid to learning, but I really do like the concept of unlearning rules of thumb, and returning to first principles, when the rules of thumb have got us into trouble.

    I suspect that the result of either policy, either in theory or in a particular historic circumstance, will depend on what else is going on at the time.

    In our present situation, with economies around the world frozen like frightened rabbits, monetary measures do not propagate very far into the system. So in the present circumstances I believe their effects on the wider economy will be heavily attenuated. That is not the same as saying it can never have an effect, only that where we are today it will not be the dramatically effective medicine it might be on other occasions in other circumstances. I say that irrespective of whichever target we choose.

    So I suppose I am offering a related viewpoint. I am not sure it matters which of the two we target; in our present circumstances, its effect is unlikely to get beyond the banking system. Sooner or later something (else) will get things moving again; what else and how are a different debate. But when things start moving, will the pent up effect of our monetary actions have an explosive effect on output, on inflation, or both? It depends on how we come out – Roubini or Keene?

    If I were in charge I would not try to use macro inputs to target anything. By its very nature of dealing with aggregates, macro measures things rather than controlling them. Macro inputs, as I said earlier, are enabling/disabling measures. You can certainly mess things up with macro inputs, but sadly you can’t necessarily put them right again.

    Very much simpler, but possibly a helpful analogy, is the situation a pilot finds himself in when needing to make a correction to height or speed while landing an aircraft. The throttle and the control column are both available – but things are not a simple as they may seem to the lay man. The throttle affects both speed and height; it controls total energy. The control column trades off speed energy and height energy. So the novice pilot has to assess first whether to trade one for the other, then whether to add or subtract from the total. Targeting either with stick or throttle just won’t work. The experienced pilot, familiar with the aircraft, applies a balanced amount of each to control a wayward variable.

    But without the correct rudder input you still can’t make a soft landing in a cross wind.

    And if you have messed up the stick and throttle inputs too much you cannot simply put those mistakes right. You must go round again – or crash and burn.

    • #10 by Unlearningecon on January 25, 2012 - 1:17 pm

      Your post seems to be based on a lot of intuition and analogy, and whilst it is a feasible PoV, Keynes’ theory of monetary activity and the real economy is incredibly rigorous and suggests that macroeconomic policy can affect both real activity and the type of activity.

      • #11 by atm0spheric on January 25, 2012 - 7:02 pm

        I agree totally – but you haven’t claimed that the results of macroeconomic policy are predictable.
        I submit that both quantitatively and qualitatively the effect of macroeconomic policies can be subject to variables which do not feature in the macroeconomic measurements characterizing the state of an economy.
        I further suggest that for this reason (inter alia) measurement of past correlation coefficients is of more limited value than some economists seem to believe.

      • #12 by Unlearningecon on January 25, 2012 - 10:18 pm

        Well, they are predictable to a certain extent but obviously there are limitations. Have you read the Lucas Critique?

      • #13 by atm0spheric on January 25, 2012 - 11:35 pm

        Haven’t read it but recognize it as a special case of my abstraction. Also lived through the Lamont debacle – confirmed my belief that measuring instruments should not be used as control levers! Saw the benefit of micro-based models subsequently Years of living through target setting regimes has made me a fan of (my second-hand understanding of) his thinking. Keep checking the assumptions – especially the first part of rational agents 🙂

  5. #14 by Luis Enrique on January 25, 2012 - 12:49 pm

    “Market Monetarists’ (MMs) hold that macroeconomic policy can only have nominal or ‘cash’ effects”

    what the hell? No they don’t. Citation please.

    ” Firstly, MMs use a standard AD/AS framework for their analysis. ”


    • #15 by Unlearningecon on January 25, 2012 - 12:58 pm

      Insert the word ‘reliably’ before every ‘nominal’ in this post. Apologies.

      Sumner DEFINITELY uses an AD/AS framework and regularly accuses Keynesians of not understanding their own models. He posts at a rate of knots so I’d find it difficult to get something, but surely if you read him you’ll know that he does both of the these things?

  6. #16 by Luis Enrique on January 26, 2012 - 9:06 am

    Sumner regularly says he doesn’t like the AD/AS framework; it is definitely not the basis for his support of NGDP targeting.

    • #17 by Unlearningecon on January 26, 2012 - 1:02 pm

      I wouldn’t say it’s the basis but I’ve definitely seen him use it to get his point across. Perhaps having Sumner as my main source of NGDP targeting knowledge isn’t a great idea, though, as he can be confusing sometimes.

  7. #18 by John77 on January 26, 2012 - 10:50 pm

    “At the centre of arguments for NGDP targeting is the implicit assumption that macroeconomic policy can only have nominal impacts. I believe this to be false.”
    Well, Duh, as Homer Simpson would say. To say that macroeconomic policy cannot have any real effects is to claim that it is dimensionally separated from the economy.
    This is where I as who is mad? Gladstone? Keynes?

    • #19 by Unlearningecon on January 26, 2012 - 10:58 pm

      Reliably. As in, it’s difficult/impossible to determine the distribution between inflation and RGDP.