Winner of the New Statesman SPERI Prize in Political Economy 2016


Wednesday 5 September 2012

Zero Lower Bound Denial


There is a great deal in Mike Woodford’s Jackson Hole paper. What was new to me was his comprehensive discussion of Quantitative Easing (QE). I hope I’m being objective in reading his account as being very sceptical of what QE can do, if it is not signalling intentions about future interest rate policy. The meat here is in the discussion of portfolio balance effects in section 3, but this paragraph from the conclusion is worth quoting in full.

“Central bankers confronting the problem of the interest-rate lower bound have tended to be especially attracted to proposals that offer the prospect of additional monetary stimulus while (i) not requiring the central bank to commit itself with regard to future policy decisions, and (ii) purporting to alter general financial conditions in a way that should affect all parts of the economy relatively uniformly, so that the central bank can avoid involving itself in decisions about the allocation of credit. Unfortunately, the belief that methods exist that can be effective while satisfying these two desiderata seems to depend to a great extent on wishful thinking.”

While I believe macroeconomists practicing demand denial represent a minority (albeit still a distressingly important minority), I think what I might call Zero Lower Bound (ZLB) denial is far more prevalent. What I mean by this is a belief that somehow monetary policy alone can overcome the problem of the ZLB. It is in many ways a perfectly understandable belief, reflecting what I have called the consensus assignment developed and implemented during the Great Moderation, which was (rightly) seen as an advance on the bad old days where fiscal policy was routinely used for demand stabilisation. Nevertheless the belief is incorrect, and damaging.

It is incorrect for two reasons. The first is summed up in the quote above. Monetary policy that involves temporarily creating money to buy financial assets is of an order less effective and reliable than conventional monetary policy, or fiscal policy. The second is not addressed in Mike’s paper, but is just as important. Even if you follow the Krugman/Woodford idea of using commitments about future interest rate setting to mitigate the recession today (which is equivalent to permanently creating more money), this does not mean that you can forget about fiscal policy. To put it another way, fiscal policy would still be a vital stabilisation tool at the ZLB even if the central bank targeted nominal GDP (NGDP). It is reluctance to accept this last point which is a particular characteristic of ZLB denial.

I have elaborated on this second point before, but let me sketch the reasoning in a different way. Suppose we only think about demand shocks, and suppose the central bank is all powerful and prescient, so absent the ZLB any demand shock can be completely offset through monetary policy. Inflation and output are related through a Phillips curve with no cost-push shocks, so keeping output at its natural rate keeps inflation on track, and NGDP on track. What the central bank actually targets does not matter here, because inflation, output, price level or NGDP targeting would all be perfectly successful in terms of desired levels of output and inflation.

Now suppose we have a large negative demand shock so that we hit the ZLB. We will hit the ZLB whether we have an inflation target or NGDP target. In both cases output falls below the natural rate and current inflation is too low (below its desired level). We will miss today’s NGDP target. What the NGDP target tomorrow does is reduce the impact of the shock on current output and inflation, because inflation that is too low today means the central bank will aim for inflation and output above their normally desired levels tomorrow (to hit the NGDP target tomorrow), which supports output today through expectations effects.[1]  

We can now make two key points. First, the ZLB still matters. NGDP targets help reduce (but not eliminate) the cost of the ZLB today, but only by incurring costs in the future. Second, fiscal policy could in principle eliminate all of these costs. A fiscal stimulus today could eliminate the ZLB constraint, allowing desired output and inflation to be achieved today and tomorrow. Equally, fiscal austerity today moves inflation and output further away from desired levels both today and tomorrow, even if we have NGDP targets.

This is why I keep irritating some by going on about fiscal policy when commenting on the current conjuncture. Apart from a few academic caveats, I’m a fully paid up member of the ‘monetary policy is all you need’ club outside of a potential ZLB or monetary union.  My views on monetary policy targets are very similar to, and have been heavily influenced by, those of Michael Woodford. But the ZLB does make a difference. It is a feature of the real world, not a consequence of any particular monetary policy strategy. ZLB denial, particularly in the hands of independent central banks, leads not just to wishful thinking, but encourages governments to make bad fiscal policy decisions.


[1] Having a NGDP target tomorrow only becomes useful (and different from an inflation target) if we miss the NGDP target today. We could eliminate the ZLB constraint today by having a much more expansionary monetary policy tomorrow (which exceeded any NGDP target), but that has greater costs tomorrow (and is also less credible today). 

10 comments:

  1. "I hope I’m being objective in reading his account as being very sceptical of what QE can do, if it is not signalling intentions about future interest rate policy."

    I read it that way too.

    Sure, if there were a really big exogenous real shock, the natural rate could go below (say) minus 2%, so a (say) 2% inflation of price level path target wouldn't work. A 5% NGDP target has a bit of insurance built in, because lower natural rates tend to be correlated with lower growth rates of potential GDP.

    But that's unlikely. Storage technology starts to kick in at negative real rates for some goods.

    But right now

    ReplyDelete
  2. What if we complete the market for future NDGP by creating a NGDP futures market, and the CB commits to buying/selling unlimited quantities of futures at the target level. Wont non-arbitrage prevent deviations from NGDP target?

    K

    ReplyDelete
  3. Monetary policy has to do with more than just interest rates. Focusing on ZLB makes the problem harder. Legislation permitting, it's possible to be far more direct -- and to make monetary policy far more effective.

    Give the US Fed (and other central banks) the authority to add a rebate to all commercial transactions. By "putting the economy on sale" the Fed would increase demand and help the economy get out of its slump. Usefully, the rebate would not be inflationary since it lowers prices rather than raises them. (If the Fed also had the authority to add a surcharge to commercial transactions, it could stifle inflation when needed.)

    This would be monetary policy applied directly. It would have an immediate effect, it would achieve its goals, and it would do so with minimal distortion.

    Instead of all the complex interest-rate and QE arguments, let's try something simple and effective.

    ReplyDelete
  4. I take a different position on the Zero Lower Bound. I'd appreciate any criticism on my position. In brief, my thought is that a central bank buying foreign assets would depreciate the domestic currency, leading to lower relative prices of domestic goods, which would increase aggregate demand through a monetary channel.

    http://socialmacro.blogspot.com/2012/09/now-let-me-be-liquidity-trap-zlb-denier.html

    ReplyDelete
  5. PROF. WREN LEWIS, FIRST OF ALL LET ME CONGRATULATE YOU FOR THE WAY YOU HAVE PUT YOUR OPINION IN HERE. IT IS AMAZINGLY CLEAR AND SIMPLE. I THINK I COULD NOT DO SUCH A GREAT JOB EVEN IF DEEPLY DEDICATED MYSELF TO IT.

    BUT THE REASON I AM WRITING TO YOU IS THAT THERE IS ONE POINT I DON'T UNDERSTAND IN THE VIEW YOU ARE DEFENDING:

    1. I BELIEVE WE ALL AGREE THAT MONETARY POLICY HAS ITS EFFECT EXAUSTED IN THE SOHORT RUN, THAT' S WHAT WE CALL IT "ZERO LOWER BAND".
    2. I BELIEVE THAT THIS IS A DIRECT RESULT OF A LONG PERIOD OF APLICATION OF KEYNESIAN POLICIES THAT HAD THE OBJECTIVE OF DIMINISHOING PRODUCT FLUCTUATIONS - THIS FIXING BUBBLES BURSTING BY REINFLATING BUBBLES ONCE POTENTIAL PRODUCT IS NOT SUCH A NEAT MEASURE.
    3. THE RESULT OF THIS PROCESS IS THAT AT SOME POINT THE CAPACITY OF GOVERNMENTAL ACTION ON GDP LEVELS REACHED A POINT OF EXAUSTION WHERE MONETARY POLICY BECOMES LARGELY UNEFFECTIVE.
    4. IF GOVERNMENT STARTS TO DO THE SAME WITH FISCAL POLICY, THE PROCESS WILL CONTINUE TO ITS EXAUSTION, LEAVING GOVERNMENT WITHOUT AVAILABLE TOOLS TO DEAL WITH RECESSIONS. IT IS IMPORTANT TO STRESS THAT MONETARY POLICY TOOK A WHILE BEFORE BEING EXAUSTED AND THAT FISCAL POLICY WOULD LIKELY DO THE SAME.

    BUT STILL, IN COUNTRIES LIKE SPAIN AND GREECE THIS EXAUSTION MEANS A NIGHTMARE WHERE THE SOCIAL SAFETY NET IS COMPLETELY ERRODED WHEN IT IS NEEDED THE MOST.

    WHY DO YOU THINK THIS IS A GOOD SOLUTION FOR AMERICA, BRITAIN OR ANY OTHER COUNTRY? WHERE DO YOU THINK I AM WRONG? CAN NANYONHE ELSE ENLIGHT ME ON THE ISSUE?

    ReplyDelete
  6. OK, I think I now have the right metaphor (and my computer isn't trying to re-boot either).

    There are two questions:

    1. Is the ZLB wall a real wall, that you can't cross, and if the economy wants to cross it and can't, bad things happen?

    Yes, of course.

    2. Is the ZLB wall made of velcro, so if you hit it you stick to it and can't get off it, and you just have to hope some exogenous non-monetary policy force comes along and pulls you off it or pulls it off you?

    No, IMHO. It's only velcro if you think that monetary policy really really is setting interest rates.

    Plus, we (probably) only hit the wall in the first place because monetary policy wasn't good enough, and an NGDP level path target (unless we had set a very low target growth rate) would have meant we never hit it in the first place. The IS curve slopes up (if you have Ygap on the horizontal axis). That means the economy can be in recession even with the actual rate of interest below the natural rate.

    ReplyDelete
    Replies
    1. Nick – Do you agree with the following? There are really two issues here. One is whether QE works. The other is about the difference NGDP targets would make. I think it is clearer if we separate the two, which we can do by thinking about what NGDP targets would do in a world where QE does not work.

      In that world, I do not understand why – as you and lots of NGDP advocates seem to suggest (your final para above) - NGDP targets would stop ZLB shocks happening. My understanding of Woodford, and my own analysis, is that it would reduce the impact of these shocks at the ZLB, but not eliminate all the impact. As a result, I take your remarks about Velcro to really be a belief that QE does (or could) work. Is this right?

      Delete
    2. Simon: I'm sorry. I only just now saw your reply.

      Yes. I agree that there are two issues here. (Or rather, it seems to me that it may be useful to separate those two questions).

      1. Yes, I believe that an NGDP *level path* target would reduce the risk of hitting the ZLB compared to an Inflation Target. (A Price Level Path target would reduce the risk compared to an IT, and an NGDP level path target would reduce the risk still further). See my comment on your latest post, where I explain why I believe that.

      2. Yes, my remarks about Velcro are a way of saying that some other instrument could work (coupled with a communications strategy that talks about some target for e.g. NGDP level path that is consistent with an equilibrium above the ZLB, of course). The nominal interest rate on very short term safe loans is just one out of a very large number of prices that a central bank could use as its immediate target. It is perfectly possible to imagine a central bank setting the time path of the price at which it will buy and sell e.g. gold in some NGDP level path variant on Irving Fisher's "compensated dollar" plan. (I'm not a gold bug -- I'm just using gold as an example.) Central banks don't have to set an interest rate, and we don't have to think of monetary policy that way.

      "Hmmm, nominal interest rates on short safe loans seem to have hit 0%. OK, no problem, we had better announce that we are going to be raising the price of gold for a bit, until that nominal interest rate rises above 0%." (That's what Roosevelt did).

      Delete
  7. Have you seen Cochrane's response to Woodford? It seems to be a rebuttal based on the view that nominal shocks can never, ever, have real effects? Are we really still debating thios kind of stuff? Intellectual progress?

    ReplyDelete
  8. The Canadian is right. under a LT we're not gong to get half as much action because that 4.5% annualized to month over month growth means that no one pays attention anymore to the charts, they know exactly where the charts are going. ten years from ow we know exactly what NGDP will be.

    The bigger issue is that Monetary policy where it exists as a tax on savings, is a PREFERABLE type of tax to Fiscal, because it is a far flatter tax, than letting the government pick winners and losers under the guise of "fiscal" stabilizing the economy.

    If the only form of Fiscal is tax cuts for the wealthy, and under those exclusive circumstances Simon DeKrugman STILL want Fiscal, please wake me back up.

    ReplyDelete

Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time. In addition, I cannot publish comments with links to websites because it takes too much time to check whether these sites are legitimate.