The failed 2013 Keynesian experiment: even worse than you thought

When 2013 ended I couldn’t help pointing out that we passed Krugman’s famous “test” of market monetarism with flying colors.  GDP growth was initially estimated as being slightly higher in 2013 than 2012 (comparing Q4 to Q4.)

Now the annual revisions are in for 2012 and 2013, and the new numbers look even worse for the Keynesians.  Let’s start with the numbers that Keynesians always (wrongly) use as proxies for demand growth—RGDP:

2011:4 to 2012:4:   1.60%

2012:4 to 2013:3:  3.13%

So real GDP growth almost doubled under the weight of higher income taxes, higher payroll taxes and the famous “sequester” which reduced government spending.  I predict this will do nothing to shake the consensus Justin Wolfers discusses here.

The theoretically appropriate indicator is of course NGDP:

2011:4 to 2012:4:   3.47%

2012:4 to 2013:3:  4.57%

I was asked if the strong growth in Q2 caused me to revise my belief in the Great Stagnation.  Not really, as it is based on long term trends.  Or at least not very much.  (Don’t forget Q1 was very weak.)  I still think we are trending at about 3% NGDP and 1.2% RGDP growth over the next few decades.  With immigration reform, both those numbers might rise a few tenths of a percent.

Of course this means low nominal interest rates for the rest of my life.  I doubt I’ll ever again see T-bills yield 5%.  I hope I’m wrong.

Mark Sadowski missed on the GDP numbers this morning (he had RGDP up 1.6%) but given the massive revisions we often see, perhaps we should wait a bit.  He was far and away the best on Q1, but we didn’t know the actual Q1 RGDP figure was negative 2.1% until today.  It was originally up 0.1%.

When I started blogging I didn’t know we were in a Great Stagnation.  I still think I was right about the mostly demand-side nature of the 2008-09 recession, but in the early years of blogging I modestly overestimated the size of the output gap.  I think that gap is now mostly closed, and will be completely closed next year.  But keep in mind aggregate supply and demand can get “entangled,” and also that policies like unemployment insurance and Obamacare can slightly impact trend RGDP.  So these things are very tough to pin down.

PS. In fairness to the Keynesians, I don’t quite believe that 2012/2013 RGDP growth differential.  I think the BEA overestimated the speed-up in growth in 2013.

PPS.  I have another post on fiscal stimulus at Econlog.

HT:  Commenter Nick

Update:  Commenter Kailer Mullet added this:

The year of Austerity, 2013, had the highest Q4/Q4 (if you use absurd decimal precision) in 10 years.


Tags:

 
 
 

44 Responses to “The failed 2013 Keynesian experiment: even worse than you thought”

  1. Gravatar of TallDave TallDave
    30. July 2014 at 07:59

    More and more I’m wondering if the concept of “aggregate demand” has net negative utility to economic theory.

  2. Gravatar of Kailer Mullet Kailer Mullet
    30. July 2014 at 08:08

    I was just about to tweet you this, looks like I got scooped. The year of Austerity, 2013, had the highest Q4/Q4 (if you use absurd decimal precision) in 10 years.

  3. Gravatar of TravisV TravisV
    30. July 2014 at 08:14

    Dear Commenters,

    Is this observation from Vincent Cate correct or incorrect?

    “So really economists want to print money to make people poorer on their existing wages (in real terms), so companies can afford to hire more people and reduce unemployment.

    The plan of printing money is to reduce the real income of the average guy and it works every time.”

    https://www.themoneyillusion.com/?p=27180&cpage=1#comment-359488

  4. Gravatar of Major_Freedom Major_Freedom
    30. July 2014 at 08:14

    Even if there is no increase in productivity or supply, a rise in NGDP will cause a rise in the reported RGDP statistic.

    Also, it is premature to make the test sample end with 2013. Economists think in the long term. The after effects of past inflation cannot be known until much later.

  5. Gravatar of John Hall John Hall
    30. July 2014 at 08:17

    Scott, of course we didn’t know that Q1 GDP was -2.1% until today, but the last update was -2.9%…

  6. Gravatar of Matt McOsker Matt McOsker
    30. July 2014 at 08:19

    Scott, how does this reconcile with your belief that money is/has still been too tight? Why can’t things be a combination of fiscal and monetary?

    Looking back at 2008 PCE was moving up, as was gov spending. The trade deficit was a huge drag on NGDP, and investment was the item flattening and dropping.

    Investment before and during great recession (I threw in housing because that – residential – seems to account for a lot of the drop):
    http://research.stlouisfed.org/fred2/graph/?g=GF6

    NGDP Component Chart
    http://research.stlouisfed.org/fred2/graph/?g=GF7

    Seems to me the fiscal policy around the recession was too small to get a big boost but did prevented a larger drop in overall government spending. The recent tax hikes and government spending slowdown seem to small to have a material impact, and overall fed, state local spending has been slightly declining.

  7. Gravatar of ssumner ssumner
    30. July 2014 at 08:49

    kailer, Great, I’ll add that.

    Travis, The whole point of monetary stimulus is to boost the real income of Americans, but that’s a common mistake. He confuses real hourly wages with real total income. Real hourly wages rose sharply for American workers between mid-2008 and mid-2009. How’d that work out for workers?

    John, Yes, I recall that.

    Matt, Still a bit too tight, but rapidly reaching a point of indeterminacy.

    Component analysis can be misleading–what is the policy counterfactual?

    Housing is an overrated factor. It fell sharply between January 2006 and April 2008, and yet RGDP grew and unemployment merely rose from 4.7% to 5.0%. The big problem was other sectors, which tanked after mid 2008.

  8. Gravatar of Doug M Doug M
    30. July 2014 at 09:03

    ” I doubt I’ll ever again see T-bills yield 5%. I hope I’m wrong.”

    If you were of your grandfathers generation, you would have said this, as rates would never get that high.

    If you were of your father’s generation, you would have said this, as rates would never get that low.

  9. Gravatar of Brian Donohue Brian Donohue
    30. July 2014 at 09:05

    TravisV,

    That sounds right to me. Isn’t that the whole point of sticky wage theory?

  10. Gravatar of Steven Kopits Steven Kopits
    30. July 2014 at 09:10

    For those of you interested in oil price debates: Hamilton, Kemp, and me, here

    http://blogs.platts.com/2014/07/30/peak-oil-forecasts/

  11. Gravatar of Morgan Warstler Morgan Warstler
    30. July 2014 at 09:10

    More and more I’m wondering if the concept of “aggregate demand” has net negative utility to economic theory.

    +1 on this Tall Dave

    it helps explain why Digital Deflation is such a hard thing to deal with.

    Normally when people give up steak for chicken, we’re screwed trying to figure out how much worse chicken is than steak.

    But with digital deflation, the chicken is better!

    Cell phone = $600 > Car = $3K

    When kids give up the car, they are doing it because the damn phone is worth more to them.

  12. Gravatar of TravisV TravisV
    30. July 2014 at 09:11

    Prof. Sumner,

    Thanks! I think that confirms that the story in my head is correct.

    Vincent Cate is Major Misleading.

    I’m reminded of this excellent graph Prof. Sumner created decades ago:

    https://www.themoneyillusion.com/?p=4220

    Worth 1000 words!

  13. Gravatar of Cameron Cameron
    30. July 2014 at 09:35

    I’d love to see a post about the Great Stagnation sometime. Are you assuming growth will remain slow because it has been recently? Do you agree with Tyler Cowen’s arguments here? Anything to add? Etc. etc.

  14. Gravatar of Tom M. Tom M.
    30. July 2014 at 09:42

    And it absolutely destroyed the prevailing right wing myth that marginal tax increases on the wealthy hurt the economy. Not that the myth wasn’t already demolished after the Clinton boom.

  15. Gravatar of Mark A. Sadowski Mark A. Sadowski
    30. July 2014 at 09:47

    Just for everyone’s information, the BEA’s preliminary estimate for 2014Q2 GDP has the NGDP major components significantly higher ($28.2 to $38.5 billion at an annual rate) than my nowcast did for PCE, private nonresidential investment in structures, equipment, inventories and government consumption. Thus the differences are broadbased and not overly dependent on any single component. The biggest difference on a percent basis (not counting inventories) is in private nonresidential investment in structures which is 5.8% higher (nonadjusted) than I was projecting.

  16. Gravatar of Matt McOsker Matt McOsker
    30. July 2014 at 09:55

    Scott I see housing as simply the first domino of many to fall (housing bottomed in 2009) – investment collapsed all the way through and PCE certainly dropped mid 2008. My main point is that overall investment was what was falling significantly.

    So far as policy counter factual – the Fed started raising rates from 1% in 2004 to 5.24% until June 2007 when they started lowering rates, by that time the housing decline already started to have its effect on many related sectors. Still not sure why their raised rates at all as inflation and GDP growth did not seem to be “overheating”. Again the trade deficit was nearly 6% of GDP at the time – a huge drag.

  17. Gravatar of tesc tesc
    30. July 2014 at 10:02

    Travis V, Keynesians think that way because they are obsessed with inflation and with interest rates.

    MM is a lot smarter than that. Printing money increases MV=PY.=NGDP. With P being sticky because of wages Y has to rise, increasing real wages.

    If P rises more than expected, money printing has to slow down. P rising means there is not much Y in the econony to increase real wages.

    Stabilize MV and you stabilize real wages. MM solves the problem.

  18. Gravatar of Nick Nick
    30. July 2014 at 10:08

    Thanks for the update, Mark. Comforting, I think. Don’t be a stranger, the interwebs need you.

  19. Gravatar of Patrick R. Sullivan Patrick R. Sullivan
    30. July 2014 at 11:31

    Off topic for this post, but Oklahoma’s AG has just rubbed Jonathan Gruber’s nose in his friends’ amicus brief;

    http://www.cato.org/sites/cato.org/files/documents/motion_to_supplement_msj_record_file-stamped.pdf

    ‘Defendants will likely argue that Gruber’s statements are immaterial, but such an argument should be squarely foreclosed by the fact that in their cross-motion for Summary Judgment, Defendants themselves relied on evidence from Professor Gruber in an attempt to argue the supposed “implausibility” of Congress having made something as important as the subsidies hinge on the States’ willingness to establish exchanges.’

  20. Gravatar of Benjamin Cole Benjamin Cole
    30. July 2014 at 21:26

    Excellent blogging, but I pull out my highest soapbox and biggest megaphone and shout at Scott Sumner over this sentence:

    “I still think I was right about the mostly demand-side nature of the 2008-09 recession, but in the early years of blogging I modestly overestimated the size of the output gap. I think that gap is now mostly closed, and will be completely closed next year.” –Sumner.

    Egads. I think the USA economy could increase output by 10 percent, and the exercise would do it good.

    Yes, depress demand long enough and you crimp supply too–but in a global economy, supply can increase quickly.

    There is gobs of capital waiting to be invested in any bottleneck that emerges. There is plenty of labor.

    Sure every modern government has structural impediments, state, local and national. Is the USA that much worse than the 1990s? (I dislike Obamacare, but it does not apply to businesses with less than 50 employees, and most businesses with more than 50 employees provide health coverage anyway.)

    I will concede the huge “disability” programs, SSDI and VA (12 million “disabled” now between the two programs) may be crimping supply of labor, but when thousands of people apply for jobs as L.A. firefighters or Wal-Mart employees, I have to wonder just how tight the labor market is.

    I do not think MM’ers should subscribe to the current “slow growth is the best we can hope for” defeatism that defines the economics trade, a meme invented by sado-monetarists to justify endless asphyxiation by central bankers.

    MM should be about obtaining robust growth.

    I remind everybody: Structural impediments in the USA were far worse in the 1960s, and we had economic growth like gangbusters. We also had a growth-crazy Fed.

    Actually, prices did not get out of hand in the 1960s. It was the late 1970s and early 1980s that we saw the double-digit inflation. (The PCE deflator was 4.7 percent in 1969!).

  21. Gravatar of Chris Brown Chris Brown
    30. July 2014 at 23:52

    I thought you never reason from a price change? Isn’t NGDP just that?

  22. Gravatar of W. Peden W. Peden
    31. July 2014 at 00:49

    Chris Brown,

    NGDP isn’t a price index, but an index of net expenditure. And if social scientists “never reasoned from a BEHAVIOURAL change”, then there wouldn’t be much for them to do.

  23. Gravatar of TravisV TravisV
    31. July 2014 at 03:42

    TESC,

    Thanks!

    The key is to distinguish between the average wage of (everyone who is working) vs. the average wage of (everyone who is working + everyone who wants to work).

  24. Gravatar of Major.Freedom Major.Freedom
    31. July 2014 at 04:53

    Reasoning from a price change is flawed for the same fundamental reason as reasoning from a spending change.

  25. Gravatar of TravisV TravisV
    31. July 2014 at 05:10

    Kevin Erdmann wrote an excellent new post here:

    http://idiosyncraticwhisk.blogspot.com/2014/07/regulatory-predestination-and-right-to.html

    And we had a good exchange in the comments section.

  26. Gravatar of Brian Donohue Brian Donohue
    31. July 2014 at 05:25

    TravisV, great link. Erdmann is excellent.

    Also, back to my comment above. Doesn’t ‘sticky wage’ theory hold that nominal wages don’t go down, which is what produces unemployment, but, if you have some inflation, real wages can adjust down, avoiding the unemployment? I mean, isn’t this the whole kit and kaboodle?

  27. Gravatar of TravisV TravisV
    31. July 2014 at 05:38

    Brian,

    Yes. However, I think the key reason why real wages adjust down is that thousands of people go from looking for work ($0 wage but not included in the average real wage calculation) to working (at a relatively low wage).

    As I said, I think the key is to distinguish between the average wage of (everyone who is working) vs. the average wage of (everyone who is working + everyone who wants to work).

  28. Gravatar of ssumner ssumner
    31. July 2014 at 05:46

    Cameron, Mostly due to 3 factors:

    1. We know that working age population growth will be very slow over the next few decades, barring a surge in immigration.

    2. Productivity growth has been slowing quite a bit–and I see no signs that will turn around soon.

    3. The markets appear to forecast very slow growth for as far as the eye can see.

    I’d add that this is a global phenomenon–even affecting Germany, so it can’t be brushed off by pointing to Obama.

  29. Gravatar of ssumner ssumner
    31. July 2014 at 06:00

    Tom, Supply-side is a long run claim, not cyclical like Keynesianism.

    Matt, Investment always or almost always falls much more than the other categories, regardless of what causes a recession.

    Patrick, That’s amusing.

    Ben, We can do better, but we need more than monetary policy–supply side reforms would help.

    Chris, No NGDP is not a price–I’m reasoning from an AD shift.

    Travis, Yes, that’s a great post.

    Brian, Wages are assumed to be sticky in both directions, in the short run.

  30. Gravatar of Brian Donohue Brian Donohue
    31. July 2014 at 06:17

    TravisV and Scott,

    OK, one more time. I understand your point about misleading averages. We see the same thing generally during recessions (average wages go up) and recoveries (average wages stagnate) because the denominator changes.

    I’m talking about individuals. The story is that, at least temporarily, real wages do go down for some individuals who would otherwise face unemployment. This is a feature, not a bug.

    I see the UK experience as a triumph of monetary policy. The important thing is to get people into jobs and keep them there. Liberals have a view of ‘career minimum wage’ employees that is wildly at variance with reality.

    The real tragedy of the crisis is how it ‘lands’ on certain cohorts. By dint of year of birth, people are condemned to lower lifetime earnings. The arbitrariness of this should be evident to everyone. Much better a temporary reduction in wages than this outcome, right?

  31. Gravatar of Michael Michael
    31. July 2014 at 06:20

    I’m curious why you think the output gap is mostly closed. It still looks close to $1T to me.

    http://research.stlouisfed.org/fred2/graph/?g=GJM

    Also, all things being equal, you would expect growth to be higher when the output gap is this large, wouldn’t you?

  32. Gravatar of TravisV TravisV
    31. July 2014 at 09:31

    PIMCO’s McCulley expects Yellen to let the economy run until labor gets a bigger piece of the pie

    http://www.valuewalk.com/2014/07/pimcos-mcculley-introduces-richs-ratio

  33. Gravatar of TravisV TravisV
    31. July 2014 at 09:38

    Brian Donohue,

    You wrote: “at least temporarily, real wages do go down for some individuals who would otherwise face unemployment. This is a feature, not a bug.”

    I think we’re on the same page. What market monetarists are really concerned about is the optimal response to a negative supply shock (such as an oil shortage, hurricane, etc). If a major oil shortage happens, real wage growth will inevitably slow. However, that doesn’t mean that millions have to have their wages fall to $0 (i.e. lose their jobs). That only happens if the central bank responds to the negative supply shock by inflicting a negative demand shock.

  34. Gravatar of Jason Jason
    31. July 2014 at 19:12

    I do not think this calculation shows anything meaningful. The average growth rate for the entire post 2009 Q4 period is 3.8 ± 1.6%. In particular, the average growth rates (as opposed to just using the end points, which introduces a lot of error) for the two periods are:

    2011 Q4 to 2012 Q4: 3.7 ± 1.3%
    2012 Q4 to 2013 Q4: 3.9 ± 1.7%

    http://informationtransfereconomics.blogspot.com/2014/07/i-do-not-think-that-calculation-means.html

    This is totally inconclusive.

  35. Gravatar of Jason Jason
    31. July 2014 at 22:46

    Scott,
    why do we even BOTHER with NGDP if it keeps getting revised like this? Why not use NGDI or wage level targeting and be done with it?

  36. Gravatar of TravisV TravisV
    1. August 2014 at 04:04

    U.S. Dollar Has Gained 4.3% Against Euro Since Early May

    http://online.wsj.com/articles/dollars-rally-bolsters-investor-expectations-that-fed-will-raise-interest-rates-1406834494

    Could this explain the weakness in U.S. stocks?

  37. Gravatar of ssumner ssumner
    1. August 2014 at 05:50

    Brian, Don’t confuse real wages for individuals and real incomes for individuals. There is a big difference.

    Real wages do not have to fall for monetary stimulus to be effective. What needs to fall is W/NGDP.

    Britain doesn’t really show the typical effect of monetary stimulus, as the big story in Britain is the fact that output has not risen with employment. The weak real wages are a supply-side story, for the most part.

    Michael, Yes, the government thinks the output gap is $1 trillion, but they’ve been wrong about almost everything in this business cycle, including the output gap. It will be revised lower.

    Jason, Not quite sure what your point is regarding the “inconclusive” growth rates. Where do the “error” terms come from?

    I agree that wage targeting might be better, but there are issues there as well. Targeting aggregate labor compensation might be a good compromise.

    Travis, I suppose, but I don’t see much weakness in US stocks–they are near record highs, and dramatically higher than few years ago.

  38. Gravatar of Jason Smith Jason Smith
    1. August 2014 at 09:33

    Hi Scott

    You asked: “Not quite sure what your point is regarding the “inconclusive” growth rates. Where do the “error” terms come from?”

    The errors come from a (fairly benign) null hypothesis of constant NGDP growth with fluctuations being a result of random measurement error and the EMH (random economic changes and price fluctuations). You can’t reject the null or confirm your hypothesis, so the result is “inconclusive”.

    However, sequestration did seem to have a (statistically significant) impact based on this model (graph at link):

    http://twitter.com/infotranecon/status/495257596144545792

    or here for non-Twitter:

    http://informationtransfereconomics.blogspot.com/2014/08/prediction-update-not-bad-for-five.html

    Also, that second Jason wasn’t me; I’ve updated with my last name.

  39. Gravatar of ssumner ssumner
    2. August 2014 at 15:41

    Jason, I guess I don’t follow that. I’ve never seen someone estimate error terms by assuming no volatility in the times series, so that everything is an error.

    BTW, if you assume actual NGDP growth is stable, that that proves monetary offset is true. So I’d like nothing more than to be wrong here. But I just don’t see it.

    Recall that my “hypothesis” is not that NGDP growth sped up in 2013.

  40. Gravatar of flow5 flow5
    2. August 2014 at 17:56

    I am the best market timer in history (bonds & stocks). That’s why I called for a stock market “zinger” when unlimited transaction deposit insurance was due to expire. I.e., savings would reverse and flow back through the non-bank financial intermediaries where they are “put to work”.

  41. Gravatar of Cory Cory
    4. August 2014 at 02:14

    The case is undoubtedly pretty damning for Keynesians. It is safe to say that none of the predictions about the sequester et all causing a recession came true.

    My question as a layman is what particular policies on the FED can be said to have properly offset the contractionary fiscal policies?

    Was it because of Janet Yellen? The Evans Rule? What was the forward guidance that allowed for this to happen?

    Seems to me that Conservatives might just as well argue that the “confidence fairy” is a real thing…that the confidence generated by fiscal consolidation is what is the cause of this growth, etc.

  42. Gravatar of ssumner ssumner
    4. August 2014 at 04:33

    Cory, The Fed did two policies in late 2012—QE3 and forward guidance. My view is that the forward guidance was more effective.

  43. Gravatar of Roosevelt Institute | In Blowout Aftermath, Remember GDP Growth Was Slower in 2013 Than in 2012 Roosevelt Institute | In Blowout Aftermath, Remember GDP Growth Was Slower in 2013 Than in 2012
    20. October 2015 at 05:18

    […] a lot of reasons people haven’t discussed it this way. I saw a lot of people using year-over-year GDP growth for 2013, proclaiming it a major success. A problem with using that […]

  44. Gravatar of In Blowout Aftermath, Remember GDP Growth Was Slower in 2013 Than in 2012 – Roosevelt Institute In Blowout Aftermath, Remember GDP Growth Was Slower in 2013 Than in 2012 - Roosevelt Institute
    10. November 2015 at 09:47

    […] a lot of reasons people haven’t discussed it this way. I saw a lot of people using year-over-year GDP growth for 2013, proclaiming it a major success. A problem with using that […]

Leave a Reply