The ‘Sumner Critique’, or Why Not to Ignore Keynes

I thought I’d offer a brief note on Scott Sumner’s latest offering to the field of economics – the ‘Sumner Critique.’ Sumner offers an apt example of why macroeconomists who ignore TGT are basically wasting their time – virtually every macroeconomic insight is already in The General Theory. Sumner says he has ‘never been able to take the book seriously.’ Maybe he just needs to read it properly.

The ‘Sumner Critique’ states that if the path of NGDP is stable, all macroeconomic effects become classical in nature. Sumner and others appear to think this is new and original, but, unfortunately for them, it was stated 76 years ago by Keynes:

Our criticism of the accepted classical theory of economics has consisted not so much in finding logical flaws in its analysis as in pointing out that its tacit assumptions are seldom or never satisfied, with the result that it cannot solve the economic problems of the actual world. But if our central controls succeed in establishing an aggregate volume of output corresponding to full employment as nearly as is practicable, the classical theory comes into its own again from this point onwards. If we suppose the volume of output to be given, i.e. to be determined by forces outside the classical scheme of thought, then there is no objection to be raised against the classical analysis of the manner in which private self-interest will determine what in particular is produced, in what proportions the factors of production will be combined to produce it, and how the value of the final product will be distributed between them. Again, if we have dealt otherwise with the problem of thrift, there is no objection to be raised against the modern classical theory as to the degree of consilience between private and public advantage in conditions of perfect and imperfect competition respectively. Thus, apart from the necessity of central controls to bring about an adjustment between the propensity to consume and the inducement to invest, there is no more reason to socialise economic life than there was before.

There you go. Replace ‘NGDP’ with full employment, and Keynes said it a long time ago. Keynes’ primary policy prescription of long term interest rates also has the benefit of being tried, and of working, after WW2. Conversely, NGDP targeting relies on expectations fairies and continually pumping up the value of asset prices. In other words: Keynes said it, but better.

Addendum: The riposte the NGDP and full employment are sufficiently different to make my criticism void does not hold. As Jonathan Catalan points out, for Keynes, ‘full employment’ was synonymous with ‘maximum effective demand, given potential output constraints.’ It’s hard to deny that Sumner uses something very similar.


, ,

  1. #1 by Ritwik on May 31, 2012 - 3:24 pm

    Full employment is not a nominal variable. An aggregate demand manipulator needs a nominal anchor. That’s the biggest macro learning from the 70s.

    • #2 by Unlearningecon on May 31, 2012 - 3:28 pm

      This is an unsupported assertion as you have stated it, and Keynes’ theories dispute it.

      And even if what you said held, Sumner would still be paraphrasing Keynes.

      • #3 by Ritwik on May 31, 2012 - 4:50 pm

        Sure, but that logic is at the core of most non-Keynesian (i.e. monetarist, of various hues) aggregate demand thought. So Sumner is not channelling Keynes in any true sense, except the most literal and trivial. The claim that if something removes obstacle xyz, effects become classical/neoclassical is at the heart of all public policy analysis. Nothing specifically TGTian about it.

        The TGT makes a specific claim, of the CB/gov’t keeping real output at the level of full employment. Sumner makes a specific claim, of NGDP. This can be construed as a paraphrasing only if you remove the nominal/real distinction. And I’m saying you can’t. Or at least, in Sumner’s worldview you can’t.

      • #4 by Unlearningecon on May 31, 2012 - 5:53 pm

        Let’s leave the nominal/real distinction for another time, as it’s not really the point I’m trying to make.

        The claim that if something removes obstacle xyz, effects become classical/neoclassical is at the heart of all public policy analysis. Nothing specifically TGTian about it.

        This is somewhat true, but Sumner is talking about it as if it is something new and important, and something that he came up with. Also, are there any examples of somebody saying it before Keynes?

    • #5 by Min on May 31, 2012 - 3:41 pm

      Full employment may be operationally defined. Let the gov’t guarantee a job at minimum wage to everyone. The result will be full employment.

  2. #6 by rob on May 31, 2012 - 10:17 pm

    “But if our central controls succeed in establishing an aggregate volume of output corresponding to full employment as nearly as is practicable, the classical theory comes into its own again from this point onwards.”

    While here may be a vague similarity in the wording that Keynes used to what Sumner says the important difference is that Market Monetarists just want to fine tune the economy via adjustments to the money supply while Keynes in the General Theory wanted to introduce policies that would essentially change the plumbing for the whole economic system because he perceived the old classical plumbing as fundamentally broken.

    Classical economic theory is all about how an optimal level of employment will be arrived at by market forces alone. Keynes thinks this does not work – though apparently he thinks that within a framework where Keynesian policies have “fixed” capitalism and stabilized overall output at the chosen level there may be a small role for “classical theory” in terms of what the actual structure of production will look like.

    Market Monetarists essentially are supporters of the free market who believe that central banks are bad at stabilizing the price of money (many believe that free banking would do a better job). They believe that to maintain economic stability then one variable (the money supply) needs to be adjusted in order to accommodate changes in another variable ( the demand for money). The level of NGDP has been identified as a way of measuring changes in the relationship between these two variables. This is clearly a much lighter-weight level of intervention than what Keynes had in mind.

    • #7 by Will on June 1, 2012 - 5:43 am

      Here I see a dynamic whereby a difference in framing is made to appear as a revolutionary discovery. The foreground and background switch places, and a new school is born. This approach seems to be more common in economics than other disciplines, but I don’t know.

      In Keynes’s account, the foreground is the fact that market economies are not in fact stable and self-correcting. He discusses changes that might be made to correct this dynamic, and that’s most of the book. He then draws away to the background, the investment and consumption decisions of individuals and capitalists, considered apart from the macro implications. Once the aggregate demand problem has been fixed, Keynes says, this background marketplace will be self-adjusting and stable, and Marshall a good guide to it.

      Friedman instead gives us a tableau with the individual decisions of consumers and capitalists maximizing whatever it is they’re maximizing. He dotes on how harmonious and beautifully self-adjusting it all is — provided the big bad government doesn’t screw up the incentives by intervening. He then draws to the background: a stable, growing money supply. As long as the monetary authority does not blunder like bureaucratic government idiots like to do, Friedman says, the foreground of mutually beneficial freedom will be undisturbed. In fact, let’s go one better and replace those bureaucrats with a numerical rule so that this background detail will be completely trivial and automatic.

      I’m a bit reminded of Joan Robinson’s quip about Schumpeter’s writing being Marx with the adjectives reversed.

      • #8 by Unlearningecon on June 1, 2012 - 11:14 am

        That’s a great response from will.

        Market Monetarism – and monetarism in general – appears to revolve around effectively denying that they are having to intervene in an economic system that is inherently unstable. You can see it with Friedman ‘blaming’ the federal reserve for the Great Depression, when in reality his argument was that they didn’t do enough; there wasn’t enough government intervention!

        Keynes was relatively free of ideological bias and as such he did not care, within reason, how much state intervention his theory would produce. He advocated an international stabilisation mechanism to allow countries autonomy over their own monetary and fiscal policy, and advocated low long term rates to prevent depressions, with fiscal stimulus should the economy happen to be in one. Market monetarist’s silence on such a stabilisation mechanism leads me to believe that they are often more driven by ideology than anything else.

      • #9 by Justin on June 2, 2012 - 2:46 pm


        But that’s completely wrong. Friedman did prove that the Fed caused the Great Depression by their policy blunders. But what you’re ignoring (or maybe you don’t know) is that Friedman said the Great Depression would have never happened if the Fed didn’t exist – you can easily Google his paper on it. He actually stated that ultimately he would have liked to seen the Fed ended.

      • #10 by Unlearningecon on June 2, 2012 - 4:43 pm

        Your use of the word ‘proof’ is questionable as Friedman relied on an exogenous money multiplier model that is highly questionable at best. He looked at M1 and M2 figures as if the CB could control them. Nicholas Kaldor’s studies indicate that base money – the only one the fed can feasibly control – actually rose by 15% between 1929 and 1931.

        I’m also not sure how Friedman could advocate free banking given his earlier criticism of ABCT.

  3. #11 by Bob Murphy on June 1, 2012 - 6:00 pm

    By the same token, if you replace “e=mc^2” with “NGDP,” then Einstein anticipated Sumner by decades.

    • #12 by Unlearningecon on June 1, 2012 - 6:15 pm

      I assume you’re completely kidding, right? I mean I’m not even how that is true as you state it, but for Keynes full employment was synonymous with effective demand; the latter is synonymous with NGDP for Sumner

  4. #13 by pe on June 1, 2012 - 6:14 pm

    Econ (or the neoclassical perversion) says allocate resources effectively. I put it to you that using valuable time reading Sumner (or Williamson or Cochrane for that matter) is a terrible waste of time. I stopped reading their total bs 12 months ago, its just not worth it.
    You could be reading Godley and Lavoie for Gods sake man!

    • #14 by Unlearningecon on June 2, 2012 - 12:07 pm

      Good point. I wish Godley weren’t so expensive…

  1. The Keynes Critique | Economic Thought
  2. Link roundup « Negative Interest
  3. TheMoneyIllusion » It’s worse than it looks
  4. On The Similarities Between Austrian and Neoclassical Economics « Unlearning Economics