How often does Matt read TheMoneyIllusion? Again and again . . .

Four days ago I said this:

But interest rate targeting (which underlies all of New Keynesian economics) has been an unmitigated disaster for American workers.  And given that rates are likely to frequently hit the zero bound in future recessions (as trend productivity growth and population growth both slow) NK policy will fail us again and again in future recessions, i.e. when we most need it to be effective.  Our current monetary regime is roughly like a car with a steering wheel that works fine—except when driving on twisting mountain roads with no guard rail.  (emphasis added.)

Today Matt Yglesias said this:

Zero Bound Recessions: Again and Again

Let me just flag this post as something to think about that will play a role in questions I’ll be arguing further. The present economic dilemma in which setting short-term nominal interest rates to zero has failed to revive the economy looks exceptional from our present perspective but will in fact be typical of future recessions. 

That’s because of factors related to population aging, slowing population growth, and eventual population decline.

And he concluded as follows:

Long story short, unless you want to be in a permanent depression you either need to find a way to put nominal interest rates below zero or else to permanently increase the background level of inflation to 4 or 5 percent a year to give yourself headroom.

Those would work (although a cashless society is obviously decades away—if only for political reasons.)  But we don’t need 4% to 5% inflation, just 4% to 5% NGDP growth, level targeting.  We could have avoided this recession with higher trend inflation, but we also could have avoided it with level targeting of NGDP.

PS.  Enjoy your honeymoon in Argentina.



18 Responses to “How often does Matt read TheMoneyIllusion? Again and again . . .”

  1. Gravatar of dwb dwb
    17. April 2012 at 12:04

    trends in central banking:

  2. Gravatar of Morgan Warstler Morgan Warstler
    17. April 2012 at 12:04

    Sooner than later, Matt will have gotten the memo that NGDPLT is not his friend.

    Well it may be good for things he says he wants, but it won’t deliver an economy where DC flourishes.

    Perhaps, this is a sign Matt has grok’d what Sumner has in store.

  3. Gravatar of marcus nunes marcus nunes
    17. April 2012 at 12:12

    Did you see the Steve Waldman piece that builds on Yglesias and Ryan Avent? I did a short post on it. For a moment, I was “terrified”.

  4. Gravatar of Brito Brito
    17. April 2012 at 12:35

    Why pay attention to standard NK DSGE models in this case?

    When you study the NKPC, you can just combine that with the equation of exchange instead of DIS, lets assume V is constant. In this case, increasing M of the representative agent boosts output because of sticky prices.

    This model doesn’t have banks, so if you’re going to use NK intuition at all, your policy shouldn’t include banks either. Instead of standard monetary policy, the implication of NK models is that the optimal policy is a negative tax for everyone financed by printed money, dispersed such that the average of everyone weighted by their negative tax rate is equal to the representative agent.

  5. Gravatar of Jim Jim
    17. April 2012 at 13:28

    So why are New Keynesians so wedded to interest rate targeting? It’s not like they have real world policy constraints when modeling.

  6. Gravatar of Brito Brito
    17. April 2012 at 13:31

    Jim, because that’s just how the model works. Monetary policy IS interest rates, and nothing else, and the solution is some sort of policy rule that results in the real interest rate being equal to the natural real interest rate. New Keynesians love things like the Taylor Rule.

  7. Gravatar of Benjamin Cole Benjamin Cole
    17. April 2012 at 13:33

    Excellent blogging.

    Go Market Monetarism. I have yet to read a critique of Market Monetarism that is compelling. Most are peevish, quibbling, or laden with hoary shibboleths and pompous pettifogging masquerading as monetary policy.

    I like the Keynesians; I think they mean well. But even if they think they are right, they need to get real. The GOP will stomp on fiscal stimulus.

    What is Plan B, for Keynesians? Sniveling in the dugout?

  8. Gravatar of Brito Brito
    17. April 2012 at 13:38

    Plan B should be credit easing.

  9. Gravatar of ssumner ssumner
    17. April 2012 at 14:33

    dwb, Good article. Yes, the Europeans are beginning to realize that monetary policy is the only way out.

    Morgan, I’d be disappointed if he dropped NGDP targeting.

    Marcus, Thanks for that link, I may do a post on that.

    Brito, I’m all for not paying attention to NK DSGE models.

    You said;

    “Monetary policy IS interest rates, and nothing else”

    But that’s only because they assume it is interest rates. They could have assumed it’s M2, or NGDP expectations, or exchange rates, or lots of other variables.

    Thanks Ben.

  10. Gravatar of Brito Brito
    17. April 2012 at 14:47

    Scott, what I mean is that in all of the NK DSGE models I have seen, the only variable the central bank controls is i_t (nominal interest rate). M doesn’t explicitly appear in the model.

    What model do you use?

  11. Gravatar of Morgan Warstler Morgan Warstler
    17. April 2012 at 16:01

    “What is Plan B, for Keynesians? Sniveling in the dugout?”

    Benji, a funny!

  12. Gravatar of Rien Huizer Rien Huizer
    18. April 2012 at 05:23


    Good to see someone raising the topic of models here (I raised this a while ago wrt to ECB policymaking on the Kantoos blog).

    Central Banks (especially the multi-stahelolder ECB) are bureaucratric intitutions that need to justify their actions by reference to formal rules that are part of or logical consequences from, their mandates. That leads to or is combined with policy rules. And those policy rules require models for two things: action design and simulation of a action consequences.

    In the CB’s bureaucratic context created by the gvt it is more imporatnt that policy rules are complied with that that the policy make anyone happy: politicians have the CB as a fall guy and the CB governor is appointed for a very long term, almost like a judge.

    The current policy rules coexist with NK (DSGE) models, because that is the most advanced technology acceptable to both gvts and CBs.And DSGE models have technical advantages for policy support, apparently.

    In order for market monetarism (and I guess MM policies such as EFFECTIVE level targeting of NDGP are more desirable than following a Taylor rule) to be adopted by grown-ups, it must acquire the modeling technology that keeps the people in the legal department happy and allows for try-outs that the CB can use for confidential discussions with relevant politicians (despite independence) prior to going public with significant departures from business as usual. Central banks must be credible, not necessarily predictable and by defining the mandate in terms of the highest level macro outcome (NDGP level over time), politicians do two things: they remove discretion from the Fed) and replace the current lack of discretion at the ECB by another type) All of that would require very reliable modelling based on consensus regarding the inclusion of relevant variables as well as key assumptions.

    You put the finger onto something important..

  13. Gravatar of ssumner ssumner
    19. April 2012 at 06:23

    Brito, I use monetarist models of the transmission mechanism, augmented by the NK assumption that it is the future path of policy that matters. Thus I am modeling the future path of money supply and demand, or money supply and the Cambridge k.

    The long run impact of M on NGDP is caused by the hot potato effect, not interest rates driving investment.

    Rien, What we need is for central banks to outgrow their belief in the supernatural, their belief that “models” can predict NGDP better than markets. Market monetarism is actually the “grown-up” approach, not NK DSGE models.

  14. Gravatar of Rien Huizer Rien Huizer
    20. April 2012 at 02:14


    Of course CBs need to be not superstitious, but they need models for completely different reasons. They are large bureaucracies that may be independent but not unaccountable. Mathematical models are an essential tool for the engineering side. It is a bit like building a bridge: everyone who could be held responsible if it collapses would want documentation that the best possible care has been taken in the design. That requires widely accepted quantitative methods. One obstacle to better policymaking is the lack of tools that allow all those responsible to do PYA. So priority #1 for market monetarism is to develop a suite of policies and tools that will function in a bureaucratic context and enable institutionally acceptable decision making.

    Even then I suspect politicians would prefer an independent CB with the capacity to make mistakes. Not an autopilot. Furthermore, politicians will never give up the capacity to make discretionary spending and taxing decisions under the guise of fiscal policy and for that they need a suitable doctrine. NK is perfectly OK for them, especially when you keep in mind that democracies tend to attract politicians (I consider Thai politics as a good naturally occurring example of raw democracy, and there it is all about buying votes and selling influence. In the US politicians may ne more institutionally constrained, but probably their instincts are similar) rather than bureaucrats who watch in awe how the automatic pilot keeps NDGP on a monomonous trajectory.

    But all jokes aside, there is a need for bureaucraitc implements, and I do not see them yet.

  15. Gravatar of ssumner ssumner
    20. April 2012 at 14:12

    Rien, I don’t really see your point. Monetary policy is nothing like building a bridge, it’s like steering a ship. You might need a model to steer a ship in the fog, but not if the captain can see a lighthouse (futures market.)

    Of course governments won’t give up the ability to spend and tax, but during the Great Moderation they most certainly did give up on fiscal policy. If we do NGDP targeting they will give it up again.

  16. Gravatar of Rien Huizer Rien Huizer
    21. April 2012 at 20:23


    Only if there was a lighthouse..

  17. Gravatar of Rien Huizer Rien Huizer
    21. April 2012 at 20:26


    If NDGP targheting means they have to give it up (as I agree that would be a consequence) again, they will not want it and sabotage it. Unless someone finds a way to compensate the losers. Key aspect of reform.

  18. Gravatar of ssumner ssumner
    22. April 2012 at 07:56

    Rien, Give them some other job, like regulating banks. I heard there’s a need for it.

Leave a Reply