06 March 2016

History Dependence in the Natural Rate

One of the pet claims of Market Monetarists is that the Federal Reserve's failure to cut the nominal interest rate quickly enough in 2008 caused the natural interest rate to become negative in 2009 -- basically that the natural rate is history dependent. This claim does not immediately seem suspect; after all, if you cause a recession in the current period by setting the nominal interest rate above the natural rate, then it only makes sense that simply reversing that decision in the next period will not close the output gap.

The problem with this is that it is ignorant of what the Wicksellian natural rate actually is: expected inflation plus the real natural rate (which is completely independent of monetary policy, unless you believe in hysteresis, which, as far as I know, no Market Monetarists do). To argue that setting the interest rate above the natural rate in the current period results in a reduction in the natural rate in the next period is to argue that the current nominal interest rate affects the inflation rate expected to prevail two periods from now.

So, either agents are backward looking (which no market monetarists, who generally like the EMH, believe to my knowledge) or monetary policy affects potential output. Also take note that a lower natural real rate is either consistent with higher current potential output (which would mean that current tight money raises potential output in the next period) or lower future output (which would mean that current tight money causes potential output two periods in the future to fall).

Given the forward looking nature of the natural rate, it should be clear that the only way that a central bank can influence the natural rate is by changing the expected path of future interest rates relative to the expected path of future real natural interest rates (I guess Woodford is smarter than some Market Monetarists might like to admit).

Then what actually happened in 2008? Evidently the Fed allowed expectations of future interest rates to exceed expectations of future natural rates, something that they could not have prevented by cutting the nominal interest rate further in 2008 and something that could not have been prevented without a large increase in inflation expectations -- which could not have happened under a 2% inflation targeting regime or an NGDPLT regime that would be broadly consistent with 2% inflation.

Note: before you take issue with my last statement, I simulated a simple New Keynesian model in which the real natural rate goes negative for five periods under two different regimes. In one, NGDP is banned from being off-target and, in the other, the central bank sets the nominal interest rate equal to the (nominal) natural rate. In both cases, the only thing that prevented the zero lower bound from binding was an increase in the inflation target and/or the equilibrium growth rate of NGDP; NGDPLT on its own can't circumvent the zero lower bound problem.

If you want proof, I have all the pictures here (it may not be immediately clear what each one means, be careful not to misinterpret).


  1. "Then what actually happened in 2008?"

    I can honestly say that I don't think it had much to do with interest rates. Instead, it was an paradigm change that went from loans easily given (everything was OK) to the realization that many, many loans were un-serviceable (people had loans that were well past their ability to service).

    We had a bubble economy, which continues today in my opinion. Today's bubble is still credit driven and still the credit is issued by government but the benefits go to a different demographic from the housing bubble demographic.

    1. This post is agnostic to the real causes of the Great Recession, it's generally just assumed that the financial crisis somehow lowered the real natural interest rate, which dragged down the Wicksellian natural rate.

      What I was dealing with here is the Market Monetarist suggestion that the Fed caused the Wicksellian natural rate to go negative in 2009 by not cutting rates quickly enough in 2008. If this were the case, it would mean that, regardless of what was going on with the financial sector, a recession could have been avoided.

  2. John:

    1. MM's generally generally find "the natural rate of interest" to be a very slippery concept that may or may not be useful. It doesn't play the central role in our thinking that it does in New Keynesian (which should be called "Neo-Wicksellian") thinking.

    2. AFAIK, no MM has ever denied (some degree of) hyesterisis in the natural rate. It would be silly to deny it. I (and I'm pretty sure other MMs) have explicitly asserted it depends on history of monetary policy.

    3. What MMs add (though we may not be strictly alone in this) is that the current natural rate depends on expected future monetary policy too.

    4. Inflation may have some inertia, even if expectations are purely rational. Overlapping wage (or price) contracts can generate inertia, so it looks like backward-looking expectations. (That's an old result, that pre-dates MM.)

    5. The only remotely plausible thing to say about the natural rate is that if there are two parallel economies, otherwise identical in both past and future except that one targets a higher inflation rate or money growth rate than the other, they may both have the same time-paths of real rates (approximately). And that if one of those economies changes the inflation target to be the same as the other's, those two economies *may eventually* converge. (though that is not assured).

    If you search "natural rate" on my blog posts you will see a lot of posts: E.g.

    A short explainer: http://worthwhile.typepad.com/worthwhile_canadian_initi/2009/12/what-is-the-natural-rate-of-interest.html

    Where I discuss history-dependence and expectations-dependence:

  3. Nick,

    "AFAIK, no MM has ever denied (some degree of) hyesterisis in the natural rate. It would be silly to deny it. I (and I'm pretty sure other MMs) have explicitly asserted it depends on history of monetary policy."

    This is exactly what I want an explanation for. As it stands, there seems to be little theoretical evidence for it.

    From what I've learned so far, Market Monetarism either requires a ton of nominal wage rigidity or is a waste of time. Not only does an NGDPLT not mitigate negative natural rates in sticky price models, all the evidence I have seen about nominal wage rigidity is that it doesn't exist to a large enough extent to be relevant (to my knowledge, this is why everyone decided to switch to sticky prices back when NK was being created).

    This post is mainly a really late reaction to a Ponnuru and Beckworth column in the New York times from a while ago. Basically, all I want is a good explanation for why the natural rate is not completely forward looking.

    Of course, I don't actually like framing monetary policy in terms of interest rates, so don't get on my back about being too Keynesian. If you want to talk about the money supply, then my problems with MMism are different and mostly related to monetary policy impotence in liquidity traps.