Louis Woodhill on QE3

Here’s Louis Woodhill in Forbes:

So, how is this QE3 thing working for us?

During three months of QE3, 4Q2012, nominal GDP (NGDP), which is precisely the variable that QE3 is supposed to stimulate, hit a wall.  Annualized NGDP growth slowed sharply, falling from 5.78% in 3Q2012 to only 0.46% in 4Q2012.  This was the lowest quarterly NGDP growth rate since the recession ended in 2Q2009.

Strike one.

The growth of total employment, which was 526,000 during 3Q2012, fell to 331,000 in 4Q2012.

Strike two.

Let me remind readers that I much prefer NGDPLT to QE3, but I did predict that QE3 would have a very small but positive impact on growth, certainly enough to offset fiscal contraction.  The evidence so far suggests that there is no reason to reject that prediction.  Let’s start with Woodhill’s jobs numbers, which are incorrect.  Job growth was larger during the 4th quarter than the 3rd quarter.  Indeed the five months since QE3 have seen the creation of an average of 196,200 jobs per month.  During the preceding 5 months the average monthly increase was only 133,600.  I certainly don’t view that acceleration as being statistically significant, but if Woodhill thinks jobs are the right measure, will he will now change his mind and declare QE3 a success?  Here’s the link in case anyone wants to check my math.

The GDP number was very low in the fourth quarter, but recall that NGDI is a much more accurate estimate of NGDP, than NGDP itself. And the NGDI number won’t be announced for a few more weeks. I’ll go on record predicting it will show much higher growth than the NGDP estimate.  The biggest and most important components of NGDP are total wages and salaries, plus profits.  The jobs numbers suggest total wages are growing (and indeed 4th quarter wage data is already in, and confirms that fact.)  Profits in Q4 would have to be horrible to offset the wage gains.

I don’t think macro data tells us very much about QE.  Even my previous post says little more than “no reason to move away from my theory-based prior of a zero multiplier.”   You really need to look at market reactions to QE rumors, where the data is overwhelming clear to anyone who pays the slightest attention to the financial markets.  But if people insist on using macro data, they need to get it right.

Woodhill has done some great work on the folly of interest on reserves. But monetary policy is very complex and I believe he puts too much weight on that single factor.

PS.  Lars Christensen recently commented on my Canadian AS shock post.  Here’s my reply:

Lars,  Just to be clear, I claimed it was BOTH a supply shock and a demand shock.  I relied on Nick’s claim that core inflation was stable, in which case both curves shifted left.  I also relied on introspection.  It defies common sense that unemployed auto workers in Ontario would be immediately rehired as construction workers in Vancouver—it takes a while to move and to be retrained.

I did blame the BOC for also allowing a negative demand shock, which increased unemployment.

The fact that a drop in demand for Canadian exports SEEMS like a demand shock is Keynesian reasoning–mixing micro and macro perspectives. It might be true, but in this case the evidence suggests that Canadian AS declined—presumably for reallocation reasons.

BTW, I believe the same applies to the US.  Under NGDP targeting we would have had stagflation in 2008-10; perhaps 1% RGDP growth and 4% inflation.


Tags:

 
 
 

17 Responses to “Louis Woodhill on QE3”

  1. Gravatar of Rob Rawlings Rob Rawlings
    8. March 2013 at 12:33

    “The fact that a drop in demand for Canadian exports SEEMS like a demand shock is Keynesian reasoning”

    Why is it Keynesian reasoning ?

    Income falls due to a reduction in exports.

    If its a demand shock the correct monetary response is to increase the money supply.

    If its a supply shock the correct monetary response is to do nothing.

    Are you saying that the correct response iwas for the BoCscenario was to do nothing ?

    (To say it was a supply shock would make some sense if the fall in exports put upward pressure on prices like a real supply shock (say a reduction in productive capacity due to a natural disaster might do) but is there any evidence that that happened ?)

  2. Gravatar of ssumner ssumner
    8. March 2013 at 12:58

    Rob, I’ve discussed this in recent posts.

  3. Gravatar of josh josh
    8. March 2013 at 13:04

    “…monetary policy is very complex and I believe he puts too much weight on that single factor.” From the perspective of a non-economist, that seems like a major theme of this blog. A lot of economic frameworks – Keynesian, RBC, etc. – seem to work in some circumstances, with the right exogenous variables/assumptions. MM steps back and can handle various types of supply-side and demand-side shocks (however defined), various fiscal and monetary responses, and feedback interactions between all of these. If there’s a Sumner’s Law, maybe it’s, “To understand a complex system, try the most general level of analysis that is still useful.” I guess the word is “robust.”

  4. Gravatar of Rob Rawlings Rob Rawlings
    8. March 2013 at 13:09

    I read the recent posts and still don’t see how it makes any sense to say it was supply shock.

    I guess I agree with Lars on this one.

  5. Gravatar of Rob Rawlings Rob Rawlings
    8. March 2013 at 13:14

    “It defies common sense that unemployed auto workers in Ontario would be immediately rehired as construction workers in Vancouver””it takes a while to move and to be retrained.”

    This is a true statement but way would cause the BoC to tighten monetary policy unnecessarily which is what you seem to be saying happened?

  6. Gravatar of ssumner ssumner
    8. March 2013 at 13:16

    Rob, If they hadn’t tightened monetary policy then inflation would have risen.

    I do agree that if a central bank was targeting the money supply then a drop in export demand would show up as less AD.

  7. Gravatar of Bill Ellis Bill Ellis
    8. March 2013 at 14:08

    Prof Sumner…Ceteris paribus ? How do the jobs numbers compare to the expected trend before QE3?

    Off the top of my head, the consensus was that they were projected to be improved by now. Is the 195,000 average jobs created much different than what was expected before QE3 even happened ? I doubt it.

    If jobs can not be clearly attributed to QE3… can anything ? What Stat can supporters of QE3 clearly point to as a ” This would not be happening now if not for QE3″ ? How can we tell if it has worked or is working ?

    When The Fed did QE3, I asked on these threads what should we look for to see how QE3 will work. The answer I got was unnecessarily cryptic… “It has already had it’s effect.” ( You may have had a point, but it seemed like a dodge to the question everyone was gonna ask. )

    I guess now we can tell that its effect was/is as nebulous as your explanation was and your evidence is…

    (I still support NGDPLT, I think under normal economic conditions it will be a superior tool… Not a Magic wand. NGDPLT as a way to “End this Depression Now”… clearly does not cut it. )

  8. Gravatar of Bill Ellis Bill Ellis
    8. March 2013 at 15:36

    How about the FED just sets a 3% inflation target ?

  9. Gravatar of Rob Rawlings Rob Rawlings
    8. March 2013 at 18:28

    “Rob, If they hadn’t tightened monetary policy then inflation would have risen.”

    Can you explain the mechanism by which this would happen?

    I can see (using the AS/AD charts from yesterday) how a “standard” supply shock would have this affect. But what is the transmission mechanism that leads from a fall in export demand to a rise in prices that would lead an IT targeting CB to tighten ?

    From the point of view of a specific industry whose demand falls it makes no difference if its buyers are domestic or foreign. So if a decline in export demand is a supply shock then isn’t any fall in demand a supply shock ?

    I suspect I’m missing something here because as Lars says all you need is an AD shock to explain the Canadian scenario. The fact that the IT could be achieved even with declining RGDP can surely be explained as a result of the severity of the shock combined with sticky wages.

  10. Gravatar of Dustin Dustin
    8. March 2013 at 19:50

    Woodhill is using the number employed from the Household survey for his jobs number instead of the payroll data from the Establishment survey.

  11. Gravatar of Benjamin Cole Benjamin Cole
    8. March 2013 at 23:16

    This is a major fear. That Bernanke and the Fed are too timid in their use of QE, and that also there are too many loose cannons on the FOMC desk spouting off about the perils of inflation.

    So, as a result, QE “doesn’t work.”

    If the Fed abandons QE, basically we will do the Japan thing and it will take at least 20 years and maybe much longer for deflation to accommodate a flat money supply. And the ever-lessening hopes of people who own property of equities–they will throw in the bowl, and give up.

    Of course, by then our culture might have changed also, to one in which seeking material gains is frowned upon.

    Think not?

    Recall that societies in which meat was hard to come by often invented the idea that vegetarianism was sacred.

  12. Gravatar of ssumner ssumner
    9. March 2013 at 06:27

    Bill, You need to read my posts more carefully. I must have mentioned 100 times the right way to determine whether QE has an effect.

    Here’s what I said:

    “I don’t think macro data tells us very much about QE. Even my previous post says little more than “no reason to move away from my theory-based prior of a zero multiplier.” You really need to look at market reactions to QE rumors, where the data is overwhelming clear to anyone who pays the slightest attention to the financial markets. But if people insist on using macro data, they need to get it right.”

    If you disagree, fine, but please don’t pretend I’m not offering any criteria.

    In any case I think we both agree that NGDPLT is a far superior policy to QE.

    A 3% inflation target would have been better in retrospect, but inflation targeting is a flawed idea, and adjusting the inflation target up and down doesn’t solve the problem.

    Rob, Good question. You are thinking in micro terms and you need to think in macro terms. If the central bank stabilzes NGDP, and if export demand falls, then the central bank must assure NGDP doesn’t fall. But real output will fall, as the unemployed export workers won’t immediately be as productive in other industries. If RGDP falls, and NGDP is stable, then prices must rise. Think of it in terms of demand for other goods rising enough to soak up some of the unemployed export workers. House prices might rise, for instance. Indeed that happened.

    Dustin, OK, but why would he do that? Everyone knows those numbers are meaningless in the short run.

  13. Gravatar of Rob Rawlings Rob Rawlings
    9. March 2013 at 07:06

    Thanks.

    I see that NGDPT will be the best policy in that scenario for the reasons given.

    I think it is misleading to call that a supply-shock (as nothing has changed on the supply-side initially). Rather, the relative demand for 2 goods has changed (I don’t think it matters if export goods are involved – the same thing would apply between 2 domestic goods). While the supply-side transitions relative prices will likely change in a way that will be seen as inflation by an IT targeting CB.

    This is a very interesting effect and definitely is a scenario where IT is non-optimal – there is way too much noise in the system for the rate of inflation to be correlated with the demand for money.

    I think it needs a special name – something like “relative demand shock”.

  14. Gravatar of Dustin Dustin
    9. March 2013 at 09:25

    Scott, obviously he used those numbers because they luckily supported his claim. He couldn’t use the more reliable payrolls data now could he?

  15. Gravatar of Tom Tom
    9. March 2013 at 18:52

    Under NGDP targeting we would have had stagflation in 2008-10
    You do keep repeating yourself on “NGDP”, but what policy makers need is a clear mechanism.
    QE3 is one. Lower rates, the previous usual way to loosen money, is the “expected” way.
    Had the Fed dropped the interest rate to 0% in 2008, we would likely have had a less steep fall, but it would probably still be recession. But, like the Great Depression of the 30s, the House Bubble Pop/Great Recession will never again occur in a similar enough fashion to clearly claim what tools would have really worked.

    Unless the Fed is talking about other tools than QE3, or now- wimpy interest rates, or the small but very silly interest on excess reserves (why not more posts against that foolishness?), your call for NGDP remains weak. Maybe also more repeat calls for an NGDP futures market — there are prediction markets slowly gaining momentum.

  16. Gravatar of ssumner ssumner
    10. March 2013 at 07:52

    Rob, Yes, “real shock” would also be a better term.

    Dustin, Don’t be so cynical!

    Tom, You said;

    “but what policy makers need is a clear mechanism.”

    No, they need to determine where they want to go. The mechanism is relatively easy, but only once you figure out where you want to go.

    You are missing the point entirely. Your claim that QE was weak is completely besides the point. Market monetarists said it would be weak. The strong tool is setting the target, everything else follows from that decision. If the Fed had done that in 2008 we might well have never hit the zero bound. QE is not a “good policy,” it’s a signal that policy has already failed.

  17. Gravatar of Scott H. Scott H.
    17. March 2013 at 15:50

    QE results in lowering long term rates. Which, as far as bank lending is concerned, is like raising short term rates. Banks will lend less. Money supply will shrink. Ceteris peribus NGDP will shrink.

    Where is the small but positive effect on growth here?
    Maybe the markets are just becoming less worried about the health of US banks instead? (Due to increases in excess reserves)

Leave a Reply