The flaw at the heart of Keynesian economics

The term ‘Keynesian economics’ means different things to different people.  But one thing almost everyone can agree on is that Keynesians believe that the 2009 fiscal stimulus boosted aggregate demand, at least compared to the no-stimulus alternative.  Alternatively, Keynesians believe the fiscal multiplier is positive.  In my view that’s a sort of sine qua non of modern Keynesianism.  Yes, it has many other things to say, but without the fiscal multiplier there’s nothing particularly “Keynesian” about the rest of the theoretical apparatus.

Even if you don’t agree with me, surely you’ve read dozens of pundits, reporters, economists and politicians make the argument for a positive multiplier, and call that argument “Keynesian.”  Indeed many go further and suggest that only unenlightened conservative ideologues could question something so obvious, so well established by both theory and empirical studies.

I’m here to tell you that it’s all a fraud.  There is no empirical study that shows the 2009 stimulus was effective.  It’s not even clear new Keynesian theory implies it was effective.  It might, but it also might not.  There is nothing scientific about “the multiplier.”  And many of the arguments made by pundits (with a few notable exceptions like Avent and Yglesias) are deeply misleading to readers.

Let’s start with the easy part.  New Keynesian theory predicts that the fiscal multiplier will be zero if the central bank is targeting inflation in a forward-looking fashion.  That is, increased deficit spending will not increase expected future growth in aggregate demand.  The smarter Keynesians know this, but the “smarter Keynesians” are a very, very small group.  If you polled PhD trained economists in America, I’d guess less that 10% know this, maybe less than 2%.

[Update:  The zero multiplier refers only to demand-side effects.  In the comment section it was noted that fiscal actions can also boost output by increasing labor supply, and this would not be offset by monetary policy.  This post addresses the demand-side effects of fiscal actions, which is also the focus of Keynesians.  I guess I’m not in the 2% either. :)]

Even worse, many of the Keynesians who do know this fail to mention it in most of their “pro-stimulus” screeds, thus giving average readers the impression that a positive multiplier is the default assumption, and that it’s up to those who disagree to explain why.  No, it’s up to those who believe fiscal stimulus would cause the Fed to stop targeting inflation (or stop Taylor Rule-type policies) to explain why they believe this.

Long time readers of this blog know that the monetary offset problem quickly degenerates into debate over whether the Fed would “sabotage” fiscal stimulus.  I don’t like that framing, because it implies monetary policy forces the Fed to “do something” to offset fiscal stimulus, or more precisely to “do something that looks like it’s doing something” to offset fiscal stimulus.

In a recent comment section, Statsguy asked if I thought the Fed would have offset a lack of fiscal stimulus in 2009.  I do think so, indeed I believe that the current unemployment rate would probably be lower than 7.7% if there had been no fiscal stimulus in 2009.  I will explain why, using what I hope everyone will see as eminently reasonable assumptions.  I’m not claiming that I know exactly what would have happened, but rather that quite plausible counterfactuals can leave us with a negative multiplier, contrary to the claims of Keynesians.  Here are my assumptions:

1.  In early 2009 there were fears of depression.  Stock prices had collapsed and unemployment was soaring.  People were frightened.

2.  If the fiscal policymakers provided no stimulus, Bernanke and the Fed would have felt an incredible burden to save the economy.

3.  However, I would not expect the Fed to do anything that looked radical or dangerous.

4.  I would have expected them to take off the shelf the most highly regarded models of how to save an economy with monetary policy when stuck at the zero bound.  That would be the Woodford model, which calls for level targeting of prices.

5.  I would have expected them to use at least some of the ideas that Bernanke suggested the BOJ employ.  One of those was level targeting of the price level.

6.  They would have been reluctant to abandon the 2% inflation target, with level targeting of the price level they would not have had to.

7.  Do you see where I’m going?  Level targeting of the price level along a 2% path is the overwhelmingly most likely “nuclear option” to be employed by the Fed if the economy is falling off the cliff and the fiscal policy makers are doing nothing.

8.  The level targeting theory suggests you want to make up ground lost in the crisis period where monetary stimulus is (supposedly) ineffective.  I think they would have started the trend line for the price level in September 2008, which is both the peak month for the price level (it started falling in October) and also the month the financial crisis blew up.

9.  The Fed uses the PCE price index, so we’ll assume they targeted that index (although the CPI or the core PCE would give you similar results.)  The PCE on September 2008 was 110.275.  That means the PCE for January 2013 should have been 120.165.

10.  The actual PCE in January 2013 was 116.342, well below the Fed’s likely target under PLT regime.  In other words, under the most likely alternative (no fiscal stimulus) policy, prices would have likely risen much faster than the actual path of prices, assuming the Fed was able to hit its target.

11.  Because both fiscal and monetary stimulus impact the demand-side of the economy, the depressed price level of January 2013 means, ipso facto, that AD and NGDP have also risen more slowly than under the no-fiscal stimulus PLT alternative.  We are worse off after having wasted hundreds of billions of dollars in fruitless deficit spending.

I can anticipate the objections:  “What makes you think the Fed would have hit its price level target?”  Initially I think they would have failed, but recall that even in the NK model a higher inflation target is far more effective than QE.  As prices fell in 2009, the expected inflation rate would tend to rise under PLT.  So even using the assumptions of the NK model, the Fed would have been doing a policy that is far more effective than the policy they actually conducted.  Surely that counts for something!  Indeed, with a higher inflation target you don’t even need to do as much QE, other things equal.  If you add in the lack of fiscal stimulus, then it’s unclear whether the PLT regime I propose would have required more or less QE than what we actually saw.

To be clear, I am not claiming that I “know” that this counterfactual would have occurred.  But I think any fair-minded person would admit that it’s a plausible counterfactual.  Here’s what I am claiming:

1.  My counterfactual is plausible.

2.  Almost all Keynesian discourse on fiscal stimulus, multipliers, neanderthal conservatives, etc, implicitly implies that my counterfactual is not plausible.

3.  Hence Keynesianism is nothing but a sandcastle built on a pile of flawed assumptions.

Maybe the 2009 fiscal stimulus helped a lot.  That’s possible.  But don’t pretend that your belief in the 2009 fiscal stimulus is any different from blind faith in some sort of religious dogma.  There’s no science to back it up, and none of the empirical studies I’ve seen have any bearing on the counterfactual I just presented.  None.

PS.  Remember those Keynesians telling us that higher payroll taxes would slow retail sales in Q1?  Looks like they might want to revise their models:

WASHINGTON (Reuters) – Retail sales expanded at their fastest clip in five months in February, the latest sign of momentum for an economy facing headwinds from higher taxes and pricier gasoline.

The solid sales last month comes on the heels of strong gains in employment and manufacturing. But the improvement in the economic picture is likely insufficient to shift the Federal Reserve from its very accommodative monetary policy stance.

“The economy in February is looking solid. None of this, however, is likely to cause the Fed to change tack in the near term,” said John Ryding, chief economist at RDQ Economics in New York.

Retail sales increased 1.1 percent, the largest rise since September, after a revised 0.2 percent gain in January. That was well above economists’ forecasts for a 0.5 percent advance.

So-called core sales, which strip out automobiles, gasoline and building materials and correspond most closely with the consumer spending component of gross domestic product, rose 0.4 percent after increasing 0.3 percent in January.

The upbeat report helped to lift to the dollar to a seven-month high against a basket of currencies. Prices for U.S. government debt fell and stocks on Wall Street slipped after a recent rally.

The healthy gains in retail sales came despite the end of a 2 percent payroll tax cut and an increase in tax rates for wealthy Americans on January 1.

Spending is being supported by the stock market rally, rising home prices and sustained job gains which are starting to push wages higher.

GROWTH FORECASTS RAISED

The gains in core sales in the first two months of the year offered hope that consumer spending, which accounts for about 70 percent of the U.S. economy, would probably not slow much this quarter after growing at a 2.1 percent annual rate in the fourth quarter.

A second report from the Commerce Department showed business inventories rose by the most in more than 1-1/2 years in January.

Retail inventories, excluding autos – which go into the calculation of gross domestic product – recorded their largest increase since August 1995.

That and the rise in core retail sales should help boost economic growth after output barely expanded in the last three months of 2012.

Economists raised their first-quarter growth estimates by as much as eight tenths of a percentage point after the reports.

Despite paying 35 cents more for gasoline at the pump, consumers also bought automobiles last month.

Receipts at auto dealerships rose 1.1 percent after falling 0.3 percent in January. Excluding autos, retail sales increased 1.0 percent, also the largest increase in five months. That followed a 0.4 percent advance in January.


Tags:

 
 
 

85 Responses to “The flaw at the heart of Keynesian economics”

  1. Gravatar of marcus nunes marcus nunes
    13. March 2013 at 09:01

    An adaptation of Lennon´s lyrics for “Imagine…there´s no fiscal stimulus”:
    http://thefaintofheart.wordpress.com/2013/03/08/its-not-a-fiscal-trap-but-an-inflation-targeting-trap/

  2. Gravatar of nickik nickik
    13. March 2013 at 09:05

    Its really pissing me of that people like Krugman, who know this stuff, never say anthing about it. Krugman just wants more goverment spending (on the things he likes), he uses the fiscal stimulus argument and keynsiansm in order to get large spending increase not because of any keynsian line of argument.

    Also arguing this with keynsians is so frustrating because they dont understand there own theorys. The can not name a liquidity trap, nobody ever observed one.

  3. Gravatar of dbeach dbeach
    13. March 2013 at 09:07

    I find this post unpersuasive, which is odd because I more or less agree with the principles behind it. I’m not a big believer in fiscal stimulus, and I believe that better monetary policy could potentially have resurrected the economy more quickly.

    Nevertheless I find it highly unlikely that the Fed would have pursued PLT in the absence of the 2009 stimulus package. First, even with the stimulus (whether or not you think it was effective) the economy was in horrible shape in 2009 and 2010, so if the Fed “would have felt an incredible burden to save the economy,” they still should have felt that way in 2010, and we got QE, not PLT. Second, Ben Bernanke himself repeatedly advocated fiscal stimulus and said there were limits to what monetary policy could accomplish; in the absence of the stimulus, I find it hard to believe the Fed would have done more. Finally, I would strongly dispute the notion that PLT would not have seemed “radical or dangerous” to the guys like Fisher and Plosser who see inflation around every corner.

    I wish we lived in a world where it was just assumed that the central bank would target the price level or the level of NGDP and macroeconomic stabilization was not considered the business of Congress, but I don’t think that’s the world we live in.

  4. Gravatar of J.V. Dubois J.V. Dubois
    13. March 2013 at 09:10

    Interesting. Scott – if you are at this business of inventing plausible conunterfactual universes, please can you also invent one that makes ECB adopt price level targeting? What level of unemployment would cut it – 12% is clearly not enough. And then maybe we can test if there may be another even more plausible universe where ECB does not give a crap and just lets Eurozone burn.

    PS: yes, I think in a our own universe – where even enlightened people on the FED board claim that FED ran “extraordinary accomodative monetary policy” – believing that there exists a posibility of them adopting PLT during 2009 is to put in mildly highly implausible

    PPS: However your other criticism of Keynesians ignoring Sumner Critique is perfectly valid. I am also getting sick of deafening silence of keynesians on this – especialy in terms of porposals to change monetary policy rules.

  5. Gravatar of TallDave TallDave
    13. March 2013 at 09:39

    Thanks for this post, the “monetary vs. fiscal counterfactual” is a confusion I’m starting to see more often.

    We should keep a close eye on Japan, they may provide a valuable natural experiment these next couple years.

  6. Gravatar of TallDave TallDave
    13. March 2013 at 09:48

    dbeach — it helps to imagine the government was controlled by a (currently nonexistent) party that both eschewed fiscal expansion and favored monetary stimulus.

    I keep hearing arguments like “the Fed is politically constrained” which seems to mean the Fed tries to avoid politically unpopular policies that might be effective, even though it isn’t really “constrained” the way Democrats and Republicans are constrained by divided control of the government and courts. I think it’s a cop-out, but for the counterfactual you could get around it that way.

  7. Gravatar of John Papola John Papola
    13. March 2013 at 09:58

    Scott,

    Favorite. Post. Ever.

    Now, can we unpack the way NGDP gets discussed by many Keynesians (and non-Keynesians) as a RGDP+Inflation. Ex post, that’s certainly true. But that order and summary implies that real growth creates a level of NGDP and then “inflation” somehow “adds” on top. This is wrong, is it not? Instead, isn’t it basically the reverse? At a given level of NGDP, faster real growth will REDUCE inflation. We could, for example, have 0% growth in NGDP and 5% growth of Real GDP, pushing the so-called “price level” down 5% (to be a hydraulic quantity theorist for a moment).

    I see this confusion manifest all the time as a belief that real growth is “inflationary” when it is, in fact, “deflationary” and quite healthily so. Again, I see this confusion as stemming from a failure to understand both the microeconomic process which gives rise to economic growth (productivity growth that increases output per input). I also see this failure as a byproduct of an economics discipline bewitched by Keynesian-style analysis.

    This post should be linked to your Say’s law post where you rightly asserted that Macro students should only be taught Say’s law and monetary equilibrium / NGDP targeting… and flush “Keynesianism” down the toilet. Damn right.

    In a better future, there will be no “Keynesians”. The debate will be between nominal income targeters, free bankers and “hard money” Rothbardians. Talk of countercyclical fiscal policy will be remembered as this horrendously silly dark age we thankfully escaped. One can dream.

  8. Gravatar of J Mann J Mann
    13. March 2013 at 10:01

    A couple unfrozen caveman questions:

    1) The difference between price level targeting and inflation targeting is that in PLT, you try to get back on your trend line (correcting for past overinflation or underinflation/deflation), and with IT, you just shoot for a given percent from wherever you are now, right?

    2) Given that QE is the non-traditional tool the fed has repeatedly reached for since the crash, why would they have been more likely to reach for PLT without the stimulus?

    3) Given the political pressures (that large segments of Congress and the population are inflation hawks), QE is at least deniable – as I recall Econtalk in 09, the question then was could the Fed “mop up” the excess reserves to stop inflation once the economy got going, and QE proponents promised us that it could.

    On the other hand, if the Fed said something like “we are shooting for 4% inflation for the next two years because we had 2% deflation this year,” you’d get a lot of angry Congresspeople calling hearings about “have you seen how much milk costs in my district?”

  9. Gravatar of SilentKz SilentKz
    13. March 2013 at 10:18

    Scott,

    Your posts are a great counterbalance to the New Keyensian blogs I read. However I am failing to see what the difference is between price level targeting and QE. What is the transmission mechanism for price level targeting besides setting expectations? Expectations are a powerful tool, but only get you so far. Eventually there needs to be a mechanism that allows the Fed to reach its price target, or the emperor will have no clothes.

    Or would price level targeting in ’07 and ’08 merely call for greater amounts of QE to get back to the price target?

  10. Gravatar of TravisV TravisV
    13. March 2013 at 10:27

    Dear Market Monetarists,

    New question for improving my understanding:

    You reject representations of the supply and demand for money like these two, right?

    http://www.compilerpress.ca/ElementalEconomics/images/R&L%202012/Ch.%2028/fig28-03.gif

    Contrary to this graph, you believe that if demand for money shifts to the right, then interest rates go DOWN, correct?

    Do market monetarists have an alternative graphical representation available illustrating the relationship between supply of base money, demand for base money and interest rates?

  11. Gravatar of John Thacker John Thacker
    13. March 2013 at 10:30

    SilentKz:

    I think that Scott’s post definitely acknowledges that PLT would have QE as a tool of last result, viz. “If you add in the lack of fiscal stimulus, then it’s unclear whether the PLT regime I propose would have required more or less QE than what we actually saw.”

    Does anyone know what the Fed’s actual target is with the current regime? They claim to have a 2% inflation target, but they’ve consistently missed it by going under– which is then cheered by all the people who reflexively think that any inflation is bad (even in a time of low real growth.) The argument, I believe, is that the uncertainty of the Fed’s intentions and unwillingness to meet its supposed target causes it to have to do more QE. If the banks and investors were convinced that the Fed was willing to do what was necessary to reach a higher target, they’d start lending and investing in order to hunt yield higher than that target.

  12. Gravatar of pc pc
    13. March 2013 at 10:36

    Scott – great post. I feel like I will be going back to this one in the future.

    I have two comments:
    1. You’ve probably seen this, but this Forbes post brought me through a range of emotions ending in laughter (even though I couldn’t finish reading it): http://www.forbes.com/sites/johntamny/2013/03/10/the-comical-central-planning-fantasy-that-is-nominal-gdp-targeting/

    2. At some point, I’d like to have an investment adviser who supports NGDPLT, if only as a litmus test of his/her understanding of the economy. How do you reconcile majority belief in seemingly inferior economic frameworks with the efficient market hypothesis? How to profit?

  13. Gravatar of Aidan Aidan
    13. March 2013 at 10:53

    nickk – Yes, Krugman is just using the fiscal stimulus to get the alien invasion he’s wanted all along.

  14. Gravatar of Michael Michael
    13. March 2013 at 10:54

    I don’t buy this argument, for three reasons which I don’t think you have addressed in your post:

    1. Bernanke’s public comments throughout the past four years on short-term fiscal stimulus. He has never offered the argument you are making; in fact, he has consistently and repeatedly called for a combination of short-term stimulus and long-term measures to address the deficit. Either he doesn’t believe monetary accomodation could do the job alone or he doesn’t have the power to push through the amount of accomodation that would be necessary, perhaps because of objections by the FOMC.

    2. As you have pointed out, the Fed has not achieved a price-level target since 2008. So you are making an argument that they would have adopted a more accomodative policy target in the absence of fiscal stimulus than in its presence. This would be highly irrational and disappointing policy. If true, the fault lies with the Fed and not with the stimulus package.

    3. The stimulus bill passed in mid-February, 2009, and the crisis had already been going on for months.

    I think this would be a more plausible counterfactual:

    “If Greg Mankiw had been Fed chairman, he would have announced a price level target.”

  15. Gravatar of Michael Michael
    13. March 2013 at 10:57

    John Papola:

    “I see this confusion manifest all the time as a belief that real growth is “inflationary” when it is, in fact, “deflationary” and quite healthily so. Again, I see this confusion as stemming from a failure to understand both the microeconomic process which gives rise to economic growth (productivity growth that increases output per input). I also see this failure as a byproduct of an economics discipline bewitched by Keynesian-style analysis.”

    Does this also stem from targeting inflation rather than NGDP? If you target 2% inflation, then you need to inflate to offset any “deflationary” growth.

  16. Gravatar of ssumner ssumner
    13. March 2013 at 11:05

    dbeach, The Fed has been continually reacting to the changing situation. I don’t see how if the situation in 2009 had much much worse they wouldn’t have done much more. That’s there MO.

    Bernanke has continually said they have more powerful tools if they need them.

    Travis, It depends what you mean by demand for money. If it’s a function of the interest rate, then more demand means higher rates. If it’s a function of the price level, then more demand at any given price level might well lead to higher rates. It also depends on the time frame, long rates often go in the opposite direction from short rates.

    Silentkz, The higher inflation target is a tool, as is QE. You use the two in combination as needed to hit the inflation target. But as we’ve recently seen in Japan, higher inflation expectations are a more powerful tool than QE.

    pc, I linked to that nonsense a few posts back. Tamny has a history of this sort of thing.

  17. Gravatar of Aidan Aidan
    13. March 2013 at 11:17

    Scott – is Bernanke *currently* arguing that the Fed has more powerful if they need them, post-QE3/infinity?

  18. Gravatar of Tyler Joyner Tyler Joyner
    13. March 2013 at 11:20

    PC,

    I’m an investment advisor, and I support NGDPLT. That said, I don’t know any other advisors who spend much effort on monetary economics at all, and the ones that do are what Scott might call “internet Austrians”. Anyone who isn’t an Austrian is probably a Keynesian of some variety, to the extent that they care.

    As you might imagine, the EMH doesn’t get much air time around here either.

  19. Gravatar of Vivian Darkbloom Vivian Darkbloom
    13. March 2013 at 11:22

    “1. Bernanke’s public comments throughout the past four years on short-term fiscal stimulus. He has never offered the argument you are making; in fact, he has consistently and repeatedly called for a combination of short-term stimulus and long-term measures to address the deficit. Either he doesn’t believe monetary accomodation could do the job alone or he doesn’t have the power to push through the amount of accomodation that would be necessary, perhaps because of objections by the FOMC.”

    That’s more or less the point I was going to make; however, I don’t see how this negates Sumner’s argument. It simply suggests that Bernanke may not (completely) agree with the counterfactual scenario. That’s hardly a reason for not “buying the argument”.

    He (Sumner) wrote:

    “2. If the fiscal policymakers provided no stimulus, Bernanke and the Fed would have felt an incredible burden to save the economy.”

    That’s right. Bernanke’s a smart guy. So, why has he warned against removal of the fiscal stimulus (at least not right away)? Unlike Michael, I see not two, but three possible answers:

    1. He really does believe in the power of fiscal stimulus, at least as an adjunct to monetary stimulus and is therefore a true “Keynesian”;

    2. He believes in Sumner’s counterfactual scenario but is not willing to say so because he does not have the political power to back up that belief within the FOMC; or

    3. Like nearly everyone else, he’s not sure and is hedging his bets.

    I’d put my money on a combination of 2 and 3.

    And, frankly, I don’t blame him for hedging. It’s not a burden I would want to have on my shoulders, particularly given the amount he’s paid for the trouble.

    Interestingly, I think if the sequester holds, this will be one of the better opportunities to test the theories. It’s not likely it will hold; however, if it does, or Bernanke and Co think it will hold, they may need to call themselves into action (or not).

  20. Gravatar of ssumner ssumner
    13. March 2013 at 11:26

    Michael, You said;

    “If true, the fault lies with the Fed and not with the stimulus package.”

    Keep in mind that any argument for fiscal stimulus presupposes an incompetent central bank. That’s the only time fiscal stimulus works. But yes, my negative multiplier does hinge on Fed incompetence.

    I’ve heard the argument about Bernanke’s fiscal policy recommendation many times. I’ve offered two responses. Either Bernanke overrates the relative power of QE vs level targeting, or he simply doesn’t want all the pressure on the Fed’s shoulders as he’s unsure if they can rise to the occasion. Well I’m unsure as well, but I think it highly likely that the policy they would adopt with their back to the wall is far more powerful than actual policy. At the same time I can’t blame Bernanke for not wanting to take the chance.

    Also keep in mind that 99% of economists disagree with me, and Bernanke is a mainstream economist. Does his answer reflect what he and the Fed would do, or is he simply repeating the NK mantra that every bit of stimulus helps? I don’t know.

  21. Gravatar of Aidan Aidan
    13. March 2013 at 11:28

    Are you arguing that Bernanke and the did not feel an incredible burden to save the economy?

  22. Gravatar of Scott N Scott N
    13. March 2013 at 11:30

    I follow the markets and I am so sick of headlines about how austerity is the road to ruin and poverty. NO! Tight monetary policy is the road to ruin and poverty.

    If the Fed keeps its current policy in place, then Republicans may end up (unwittingly) winning the austerity argument. The Fed’s current monetary policy will cause a big economic expansion despite the horrible, unconscionable sequester cuts (Obama’s rhetoric not mine) and strip naked the lie that austerity is bad. Joe public won’t understand that the Fed engineered the expansion and will instead think that Republicans are right, that smaller government equals a better economy. I can’t say that this is a bad outcome.

  23. Gravatar of Tyler Joyner Tyler Joyner
    13. March 2013 at 11:31

    “Also keep in mind that 99% of economists disagree with me”

    If you had to take a WAG, what would you say the breakdown of economic schools is right now? Are most economists New Keynesians?

  24. Gravatar of ssumner ssumner
    13. March 2013 at 11:33

    Aidan, I’m sure Bernanke would agree that shifting to a 4% inflation target would be even more stimulative, at least if you asked him privately. Obviously he doesn’t like to come right out and say the Fed could do more, instead he usually states that they have other tools. Suppose they changed to 6.2% unemployment and 2.8% inflation thresholds, how is that not more expansionary?

    Vivian, I mostly agree, and would add two points:

    1. I think about these issues using a more “radical” or contrarian model than Bernanke does. I’m more “autistic” if you will. He relies more on common sense. That’s part of it.

    2. It’s easy for me to say “do more,” I don’t have to wake up every morning worrying about twisitng the arms of Fisher, Plosser, etc.

  25. Gravatar of Matt Rognlie Matt Rognlie
    13. March 2013 at 11:34

    Let’s start with the easy part. New Keynesian theory predicts that the fiscal multiplier will be zero if the central bank is targeting inflation in a forward-looking fashion. That is, increased deficit spending will not increase expected future growth in aggregate demand. The smarter Keynesians know this, but the “smarter Keynesians” are a very, very small group. If you polled PhD trained economists in America, I’d guess less that 10% know this, maybe less than 2%.

    Perhaps “less than 2%” know this because it’s not actually true. In general, the moral of the basic New Keynesian model is that if the central bank is targeting inflation, then the responses you’ll see to various policies will just be the neoclassical ones. And the “neoclassical fiscal multiplier” is never exactly 0, unless government spending is a perfect substitute for private consumption. Under typical assumptions it will be somewhere in the range 0.5 to 1.

    Don’t just take it from me – take it from pages 15-16 of Woodford’s paper on the multiplier, where he considers the case of a strict inflation targeter and concludes “Hence the multiplier will be given by (1.7), just as in the neoclassical
    model.” Formula (1.7) gives the neoclassical multiplier as a function of elasticities, and shows that it must lie within the range [0,1]. Surely if anyone embodies “smart New Keynesians”, it’s Woodford, right?

    Of course, the fact that the “multiplier” here is greater than 0 is not a justification for stimulus spending; after all, the neoclassical multiplier is always operative, and always greater than 0 (even during booms), and surely NK theory does not provide a justification for stimulus spending during booms. And from a practical perspective I am probably much closer to you than Krugman on this issue anyway. But it still grates me to hear that the multiplier is literally “zero”; unless you have a very different set of assumptions, it’s generally not.

  26. Gravatar of ssumner ssumner
    13. March 2013 at 11:37

    Aidan, Yes, but without fiscal stimulus the burden on the Fed would have been even greater. The spotlight of the profession would have been on the Fed. Maybe a new post.

    Tyler, Most economists don’t know what new Keynesianism is. But most that do are NKs of one form or another.

  27. Gravatar of Vivian Darkbloom Vivian Darkbloom
    13. March 2013 at 11:40

    “2. It’s easy for me to say “do more,” I don’t have to wake up every morning worrying about twisitng the arms of Fisher, Plosser, etc.”

    I was more thinking about waking up to the prospect of saving or ruining the entire economy. After that, he can worry about how he can coax Fisher and Plosser into bed with him. Now, *that’s* a worrying thought….

  28. Gravatar of ssumner ssumner
    13. March 2013 at 12:07

    Matt, I took it for granted that people would understand that I was excluding supply side effects of fiscal stimulus. Obviously there are fiscal policies that might shift the AS curve to the right. But in terms of demand-side stimulus, if you don’t increase inflation then you don’t increase AD, unless you assume the SRAS curve is flat, which it obviously isn’t.

  29. Gravatar of ssumner ssumner
    13. March 2013 at 12:29

    I should add a point for readers who don’t follow Rognlie’s argument. There are two ways that fiscal policy can boost AS:

    1. Cuts in MTRs give people more incentive to work.
    2. More government output causes people to work harder so that private output doesn’t fall one for one with the extra government output. Imagine a huge boost in military spending. If it caused private output to fall one for one with the extra military spending, then even in a neoclassical model people would want to work harder, because the marginal utility from an extra unit of private consumption rises relative to a marginal utility of an extra unit of leisure.

    I think point two is the intuition behind Woodford’s result, but someone tell me if I am wrong.

  30. Gravatar of How do we know ‘anything’ is a “good story”? | Historinhas How do we know ‘anything’ is a “good story”? | Historinhas
    13. March 2013 at 12:34

    […] and being reminded of famous songs.  First it was Lars with “Believe it or not” and then Scott with a version of Lennon´s “Imagine (there´s no fiscal stimulus) it´s easy if you […]

  31. Gravatar of Bill Ellis Bill Ellis
    13. March 2013 at 12:38

    Scott says….”Level targeting of the price level along a 2% path is the overwhelmingly most likely “nuclear option” to be employed by the Fed if the economy is falling off the cliff and the fiscal policy makers are doing nothing.”

    You just described Europe…. Yet no level targeting.

    Which leads us to something like an actual observable counter factual… With no level targeting in the mix… Austerity (Anti-stim ) in Europe loses out to even our weak stim.

    Of course a lot of things cloud the comparison…chief among them the Euro.

  32. Gravatar of TallDave TallDave
    13. March 2013 at 12:44

    Bill Ellis,

    That’s only true for some Euro economies. Others are plugging along, maybe not doing great but not falling off a cliff either.

  33. Gravatar of colin colin
    13. March 2013 at 13:00

    Tyler Joyner–or whoever for that matter,

    Does anybody know why it seems to me like there are far more Austrians in investment planning and among traders than in the academy? I can at least understand why they would be pretty anti-EMH to at least some degree given what it can imply about their jobs, but they seem to skip right on past the freshwater school straight to Austrian stuff, often the paranoid variety. Is that just a sampling problem on my part?

  34. Gravatar of Niklas Blanchard Niklas Blanchard
    13. March 2013 at 13:17

    Colin: Probably intuition on their part…and internet Austrians are the ones yelling the loudest.

    Though it’s a fallacy of composition, negative news seems counter-factual because in most people’s frame of reference, they, personally, do everything they can to make things better. There is certainly a minority of people in the world who actively try to sabotage their own lives.

  35. Gravatar of Niklas Blanchard Niklas Blanchard
    13. March 2013 at 13:17

    Should have included that people are drawn to counter-factual claims.

  36. Gravatar of Adam Adam
    13. March 2013 at 13:20

    Aren’t you just asserting that you believe that price level targeting is more effective than fiscal stimulus? Is there evidence for that or is it just alternative dogma?

    There are also a lot of political dynamics around what would have happened in the absence of fiscal stimulus. I believe that you’re right about what would have been the right policy response by an unconstrained Fed, but I just don’t know what happens in the real world. What if the Fed feels that politically it can’t do what you think it needed to do?

  37. Gravatar of marcus nunes marcus nunes
    13. March 2013 at 13:23

    @John Papola
    John, you´re not just “infected with the ADD virus”, you´ve been “completely taken over by it”!
    Welcome!

  38. Gravatar of TravisV TravisV
    13. March 2013 at 13:45

    Dear Market Monetarists,

    Prof. Sumner wrote:

    “Travis, It depends what you mean by demand for money. If it’s a function of the interest rate, then more demand means higher rates. If it’s a function of the price level, then more demand at any given price level might well lead to higher rates. It also depends on the time frame, long rates often go in the opposite direction from short rates.”

    I’m still not understanding. Aren’t you guys arguing that when the Fed slowed NGDP growth in 2008, there was a massive shift to the right in the money demand curve? And don’t you guys think that that resulted in LOWER rather than higher interest rates?

    It seems to me that the Market Monetarist story is very different from these graphs:

    http://www.compilerpress.ca/ElementalEconomics/images/R&L%202012/Ch.%2028/fig28-03.gif

    But I’m having trouble visualizing an alternative graphical representation.

  39. Gravatar of ssumner ssumner
    13. March 2013 at 14:03

    Travis, There was a typo there, I meant to say :

    Travis, It depends what you mean by demand for money. If it’s a function of the interest rate, then more demand means higher rates. If it’s a function of the price level, then more demand at any given price level might well lead to LOWER rates. It also depends on the time frame, long rates often go in the opposite direction from short rates.”

  40. Gravatar of ssumner ssumner
    13. March 2013 at 14:10

    Bill, Good point–the ECB is far more conservative than the Fed. And I’m not sure why.

    BTW, I’ve always agreed that fiscal stimulus might have some positive effect in individual eurozone countries, as there is no obvious monetary offset. However the ones who need fiscal stimulus are broke. And it’s not clear if it affects the entire eurozone, as long as the ECB is targeting inflation at 2%.

    Adam, Call it “dogma” if you wish, but if Keynesian bloggers are going to keep assuming that those who doubt the fiscal multiplier are idiots, don’t they need a response to my “dogma.”

  41. Gravatar of Tyler Tyler
    13. March 2013 at 14:13

    Scott,

    I think your post is incorrect for a variety of reasons, but leaving that aside.

    Can you produce a counter-factual of what you would expect the path of overnight interest rates to have been and the path of inflation to have been in the absence of any fiscal stimulus at all, even automatic, up until your chosen date of March, 2013 (lets say for example the federal government ran the same policies as states ran more or less).

    I assume you think inflation rates would have initially fallen very dramatically and of course that rates would have been zero, so even more ground to make up,

    But I wont lead the witness. More importantly how would the Fed have achieved in concrete terms whatever inflation path you plug in, in the above scenario of complete balance budget.

    I believe when you actually situate your advice in this context you will perhaps, change your mind somewhat,

    By the way, the reason the Federal Reserve beleived/beleives its monetary policy is very stimulative is not complex, it simply rests on the absolute accomodativeness of negative real interest rates, and the ability to extend, without limit, those low money rates of interest into asset class after asset class. If the real economy can borrow at negative real rates, why wouldnt it? Although of course in your example the government shouldnt.

  42. Gravatar of Doug M Doug M
    13. March 2013 at 14:20

    I think that just about everything with Keynesianism is incorrect, and I have to rebut your take-down…..

    It is fundamental to the Keynsians that montary policy fails at the zero lower bound. The central bank was not able to target inflation in a forward looking fashion between 2009 and present. Hence, inflation has run below target for 3 years, and the fiscal multiplyer may be greater than zero.

    Even the true beleivers of Keynesianism don’t think that the government can create growth with no cost. At best, the growth created via fiscal stimulus is GDP that has been “borrowed forward” from future quarters. Fiscal action can smooth the most extreme points of the business cycle. This suggests that the anemic growth since 2010 is the cost we are paying for stimulus in 2009.

    Is slower long-term growth an acceptable price to pay for shallower recessions? Personally, I don’t think so! Maximal long-term growth minimizes the pain in the long-term. Embrace the bust.

  43. Gravatar of Adam Adam
    13. March 2013 at 14:22

    Just to be clear, I’m not dismissing your position. I just don’t have a way to sort out the competing claims, wish that I did, and wish that commentators would be clear about the basis for their beliefs and the extent to which they are grounded in evidence.

  44. Gravatar of Negation of Ideology Negation of Ideology
    13. March 2013 at 14:36

    This is one of the best posts on this blog – up there with the MOA vs. MOE post, the 1/NGDP post, and the one about there being no such thing as holding monetary policy contant. May I suggest adding to the recommended links section?

    This post really clarifies the issue of the only serious debate in economics – Keynesian vs. Monetarist, broadly defined. The key is that from the perpective of the Monetary Authority it doesn’t matter if a fall(or rise) in AD is caused by government or private spending. All it cares is that AD is off target. To the extent that the government is successful in pushing AD over or under the target, the Fed can offset it by issuing or retiring the base. If the Fed thinks AD is too low, it should just raise it directly.

    Scott, I do have one question. Tim Congdon advocates debt management operations by the Treasury when the short term rate approaches zero as an alternative to QE. That is, the Treasury shifts to short term borrowing in whatever amounts the Central Bank needs and buys back its own long term debt. In effect, the Treasury is doing the QE rather than the Fed. Do you think this would have been helpful in 2009? I’m thinking more politically here – I know it makes no difference economically.

  45. Gravatar of 123 123
    13. March 2013 at 15:05

    “You just described Europe…. Yet no level targeting.”

    What? Have you seen the chart? It is indistinguishable from level targeting. Remarks by Trichet about the “impeccable” success of inflation targeting confirms this too.

  46. Gravatar of 123 123
    13. March 2013 at 15:11

    Couple of things:
    1. Fiscal stimulus in the downturn reduces the volatility of NGDP. This means that Scott’s scenario will have higher peak job losses, but a much faster recovery afterwords, with a better job situation now.
    2. Retail sales – this is from John Hussman:
    Feb retail sales 2012: Unadjusted $376.7B, Adjusted $402.8B (+6.9% seasonal adjustment). 2013: Unadjusted $381.0B, Adjusted $421.4B (+10.6%); February retail sales: gas ($2,286m) and motor vehicles ($866m) represented over 70% of the February gain ($4,409m). This needs to be confirmed, as he also wrote that The Vatican just picked Ben Bernanke to be Pope in the next tweet.

  47. Gravatar of ssumner ssumner
    13. March 2013 at 16:18

    Tyler, You said;

    “But I wont lead the witness. More importantly how would the Fed have achieved in concrete terms whatever inflation path you plug in, in the above scenario of complete balance budget.”

    I’d encourage you to read Nick Rowe’s “People of the Concrete Steppes” posts. The price level target is far and away the most important “concrete step.” After that, you let the market determine how much base money they want to hold when expected inflation is on target. I have no idea what the path of interest rates (or the base) would look like.

    I believe inflation would have fallen less under my counterfactual.

    You said;

    “By the way, the reason the Federal Reserve beleived/beleives its monetary policy is very stimulative is not complex, it simply rests on the absolute accomodativeness of negative real interest rates, and the ability to extend, without limit, those low money rates of interest into asset class after asset class.”

    God I hope the Fed doesn’t believe that, as it’s wrong. If they do they should read Bernanke’s 2003 speech where he pointed out that real interest rates are not a good monetary indicator, and that NGDP growth and inflation are best. By those criteria money has been tight since 2008.

    Doug, I don’t agree. You think fiat money central banks are unable to debase their currencies? What evidence?

    Thanks Negation, but I’m not a fan of operation twist, which seems to be what you are describing. I’d prefer NGDPLT combined with ordinary OMOs.

    123, That’s a better response to Bill Ellis than mine.

    I don’t know what to make of the spending numbers. From what I read the figures were higher than expected even apart from gas. The point is that nominal expenditures rose pretty strongly—I don’t much care about the breakdown of which goods were bought. If fiscal drag was a problem shouldn’t the high gas price combined with fiscal drag led to a fall in other purchases?

    Bernanke would be an excellent Pope. And move Benedict to the ECB, he couldn’t do worse.

  48. Gravatar of Bob Murphy Bob Murphy
    13. March 2013 at 16:23

    Scott, this is an intriguing post. But let me play Krugman’s advocate for a moment: Why couldn’t he say something like, “Guys like Sumner talk about how the Fed can fix the recession. But that’s only holding fiscal policy fixed. Within any plausible range of Fed operations that the public would tolerate, the federal government technically has the power to offset it with fiscal policy, to keep AD constant.”

  49. Gravatar of Michael Michael
    13. March 2013 at 17:30

    Scott Sumner wrote:

    “I’ve heard the argument about Bernanke’s fiscal policy recommendation many times. I’ve offered two responses. Either Bernanke overrates the relative power of QE vs level targeting, or he simply doesn’t want all the pressure on the Fed’s shoulders as he’s unsure if they can rise to the occasion. Well I’m unsure as well, but I think it highly likely that the policy they would adopt with their back to the wall is far more powerful than actual policy. At the same time I can’t blame Bernanke for not wanting to take the chance.”

    My impression of your view on fiscal stimulus is this: if monetary policy is effective, fiscal stimulus has no effect on aggregate demand.

    If the Fed had adopted 2% PLT in Jan 2009, then whether the subsequent fiscal stimulus was $0, $1 trillion, or even $2 trillion, aggregrate demand would have been the same.

    2% PLT, with or without fiscal stimulus, would have been more effective policy than what the Fed has actually done since 2008. If Bernanke wanted this type of policy, why would fiscal stimulus have deterred him?

    His policies have seemed very geared towards preventing deflation and also preventing any of the “catch up” inflation that would be required to achieve 2% PLT.

    I am sure he would have undertaken more accomodation in the absence of ARRA, but don’t see any reason to think ARRA deterred him from setting a more aggressive policy goal. I find it more plausible to believe that – if he had gotten Bernanke’s job – Greg Mankiw might have done 2% PLT or announced a higher inflation rate target, since he has suggested both policies at various times.

    Also keep in mind that 99% of economists disagree with me, and Bernanke is a mainstream economist. Does his answer reflect what he and the Fed would do, or is he simply repeating the NK mantra that every bit of stimulus helps? I don’t know.”

  50. Gravatar of Doug M Doug M
    13. March 2013 at 17:36

    “Doug, I don’t agree.”

    You don’t agree that it is fundamental to Keynesianism that monetary policy fails at the ZLB? Or you don’t agree that the Fed undershot their inflation target in the 2009 to 2012 period?

  51. Gravatar of TallDave TallDave
    13. March 2013 at 18:33

    Within any plausible range of Fed operations that the public would tolerate

    I hate this line of reasoning more every time I see it. Why have an independent central bank if it’s just going to be subject to the whims of polling, even when that’s bad for the economy?

    And even then, why assume the public doesn’t want growth? The policy might initially be criticized, but if they’d done something more expansionary in 2009 would people be complaining more or less than they are now?

  52. Gravatar of TheMoneyIllusion » Here we go again TheMoneyIllusion » Here we go again
    13. March 2013 at 19:01

    […]  I appreciate the nice comments from people on my Keynesian-bashing post, but I regret my remarks claiming that most economists don’t understand the monetary offset […]

  53. Gravatar of Mike Sax Mike Sax
    13. March 2013 at 20:47

    Seems to me that the Sumner Critique-that fiscal policy only works to the extent that the CB is incompetent-only holds if the CB is an inflation targeting regime or some sort of policy that amounts to the same-level targeting, etc.

    In Marriner Eccles Fed it wasn’t true. It has really only been true in the post-Volcker era-though it may be less true now judging Bernanke by his own words.

  54. Gravatar of Mike Sax Mike Sax
    13. March 2013 at 20:48

    Seems to me that the Sumner Critique-that fiscal policy only works to the extent that the CB is incompetent-only holds if the CB is an inflation targeting regime or some sort of policy that amounts to the same-level targeting, etc.

    In Marriner Eccles Fed it wasn’t true. It has really only been true in the post-Volcker era-though it may be less true now judging Bernanke by his own words.

  55. Gravatar of Mike Sax Mike Sax
    13. March 2013 at 20:48

    Seems to me that the Sumner Critique-that fiscal policy only works to the extent that the CB is incompetent-only holds if the CB is an inflation targeting regime or some sort of policy that amounts to the same-level targeting, etc.

    In Marriner Eccles Fed it wasn’t true. It has really only been true in the post-Volcker era-though it may be less true now judging Bernanke by his own words.

  56. Gravatar of The flaw at the heart of Keynesian economics « Economics Info The flaw at the heart of Keynesian economics « Economics Info
    13. March 2013 at 23:00

    […] Source […]

  57. Gravatar of J.V. Dubois J.V. Dubois
    14. March 2013 at 03:23

    Scott: “123, That’s a better response to Bill Ellis than mine.”

    Actually it is a great response, but not the way that supports your counterfactual argument. The only reason why ECB is close to the actual policy of PLT is that it operates the fiscal-monetary policy torture machine: tighten monetary policy -> governments announce austerity meausres including higher taxes -> higher inflation -> tighten monetary policy ….

    Reasons? Because they target worog (HICP) inflation and not GDP deflator. The inflation rate measured by GDP deflator is close to 1%. The NGDP growth since 2008:2 is only 2,47% How I know it? I use your own data from “The Eurozone NGDP Catastrophy” post

    So if ECB did not even manage to switch to correct inflation measure, how on earth should they do something more radical?

    And as FED goes, what about counter-factual universe where Obama would cave in to and would enact “negative” stimulus – spending cuts + tax increases to combat recessions? Are you really so sure that FED in 2009 would behave that differenly than ECB?

    Obviously not – and again I will use your own words ” It’s easy for me to say “do more,” I don’t have to wake up every morning worrying about twisitng the arms of Fisher, Plosser, etc.”

    So yes, it is easy for you to construct counterfactual universe where everything is fine. But such a universe is a lot different than the one where we actually live in. So to conclude, I contest your premise 1 “My counterfactual is plausible.”. This is false and therefore your whole argument is false.

  58. Gravatar of Rademaker Rademaker
    14. March 2013 at 03:42

    My intuition on this is that trying to offset the 2009 decline with monetary promises is like trying to stop and 18-wheeler by waving a feather in front of it. Too much momentum behind it. The argument falls apart when viewed from a disequilibrium minded point of view, i.e. a realistic one.

    If monetary promises can stop a deflationary decline, why didn’t they prevent one from happening in the first place? Why doesn’t the market anticipate this reaction on the part of the Fed? Aren’t markets efficient? Can they predict anything efficiently EXCEPT these new Fed promises?

    All this in addition to the fact that you’re continuing to ignore the issue of inordinate credit expansion that more Fed action would fuel. To wish for a return to the great moderation is lunacy. Credit market growth as a proportion of GDP was on a vertical trajectory. Even monetizing the current excess away through Fed action would likely immediately induce an even greater credit boom. Banks will jump on the chance to set themselves up for the next monetary bonanza, acting as a colluding agglomerate through the use of OTC derivatives that allow them to control the rate of credit expansion together. I don’t think real GDP growth will exist under such a set up, but banks will have their guaranteed nominal returns. Meanwhile they’ll price in the monetization’s inflation in the fixed income market and flee into foreign economies, leaving the domestic one a spent husk. Please don’t prove the Ron Paul types right.

  59. Gravatar of Mike Sax Mike Sax
    14. March 2013 at 05:18

    Rademaker, I’m probably sympathetic to you but you seem to be offering two contradictionary arguments.

    1). The Fed can’t stop deflation with just a promise for a higher price level.

    2). Anyway, if they are successful we’ll have a huge inflationary bubble.

  60. Gravatar of Matt Rognlie Matt Rognlie
    14. March 2013 at 07:16

    ” More government output causes people to work harder so that private output doesn’t fall one for one with the extra government output. Imagine a huge boost in military spending. If it caused private output to fall one for one with the extra military spending, then even in a neoclassical model people would want to work harder, because the marginal utility from an extra unit of private consumption rises relative to a marginal utility of an extra unit of leisure.

    I think point two is the intuition behind Woodford’s result, but someone tell me if I am wrong.”

    Yes, this is exactly the intuition behind the neoclassical “multiplier” mentioned by Woodford. Again, not a strong argument for stimulus, since it is always greater than 0.

    The formula for this multiplier turns out to be very simple: in a model with separable utility, if -eta_u is the elasticity of the MUC with respect to C, and eta_v is the elasticity of the MUL with respect to total output Y, the formula is eta_u/(eta_u+eta_v). Intuition: if the MUC changes very rapidly with respect to C, then people will try to work harder to replace almost all the lost consumption when spending increases, and the “multiplier” will be close to 1. If the MUC changes slowly but the MUL changes rapidly, then people are completely unwilling to work much harder, and the “multiplier” will be close to 0.

    This is not really fertile ground for a Keynesian anyway – to really justify stimulus, generally people need to assume that the central bank’s inflation target is temporarily inactive at the margin, mainly because of the zero lower bound.* It’s not clear how to interpret a complicated intertemporal rate-setting and asset-purchasing strategy like the Fed has now – is it closer to inflation targeting being active or inactive? In my view, probably somewhere in between.

    * It’s actually even worse than this. In the basic NK model, if the Fed is inactive, then the multiplier is 1 (a benchmark if the expected real rate going ahead stays constant) plus a contribution from the stimulus-induced rise in expected inflation. The only reason the multiplier is ever higher than 1 is that inflation increases dramatically with stimulus, which then decreases the expected real rate. My view is that this (1) dramatically overstates the current short-term response of inflation to changes in government spending and (2) would surely be self-defeating if it ever did occur, because a big change in inflation from spending would provoke a big response from the Fed.

    To be fair, this might have been a fairly big channel during the Great Depression. And most Keynesians have other mechanisms for the multiplier in the back of their heads – “rule of thumb” consumers as formalized by Gali, Lopez-Salido, and Valles, and so on. I suspect that they overrate the power of these mechanisms, and they are still vulnerable to monetary offset.

  61. Gravatar of Michael Michael
    14. March 2013 at 07:38

    Rademaker wrote:

    “If monetary promises can stop a deflationary decline, why didn’t they prevent one from happening in the first place? Why doesn’t the market anticipate this reaction on the part of the Fed? Aren’t markets efficient? Can they predict anything efficiently EXCEPT these new Fed promises?”

    Because no such promises were made?

  62. Gravatar of Jon Jon
    14. March 2013 at 08:25

    Matt, in short your claim is that AS shifts when G increases. Scott’s claim is that AS stays fixed, so the multiplier is zero.

    Since we are discussing SRAS, you are almostly certainly seeing an illusion in your model. Consider oil, demand is inelastic in the medium term (months). Production is basically pegged to a narrow band years in advance. So any increase in delY certainly leads to and increase in delP. Given an inflation target, del Y cannot move.

  63. Gravatar of Adam Adam
    14. March 2013 at 11:09

    Bob – I believe Scott has previously responded to such comments by noting that the Fed acts last. I’d add that it also acts with the most flexibility and the fewest political hurdles.

    I’d also add that given the incentives elected officials face, I’m skeptical that Congress would ever intentionally use fiscal policy to slow economic growth.

    So, I don’t think that would be Krugman’s response. Having read enough Krugman, I’d instead say that his response would be that he only advocates for fiscal stimulus at the zero lower bound, where he believes that multiplier is positive (he is sometimes careful to note that is his position) and he believes monetary policy has limited effects (I believe he would have said no effects not terribly long ago and give Scott some credit for turning that around).

  64. Gravatar of InterestedDude InterestedDude
    14. March 2013 at 11:57

    What about the CBO report that states that ARRA was indeed effective? Couldn’t that be seen as empirical evidence? If not, why not?

  65. Gravatar of ssumner ssumner
    14. March 2013 at 11:57

    Bob, I agree with Tyler Cowen—the Fed always moves last.

    Michael, You said;

    “I am sure he would have undertaken more accomodation in the absence of ARRA, but don’t see any reason to think ARRA deterred him from setting a more aggressive policy goal.”

    This makes no sense to me, as setting a more expansionary policy goal is the primary way of doing more accommodation.

    Doug, I probably misinterpreted you. I meant that I don’t agree the Fed’s out of ammo when rates are at zero. New Keynesians have diverse views on that issue. Bernanke and Svensson think they can do more.

    Mike, Any Fed policy without a nominal anchor is incompetent.

    JV, That’s a good argument for Europe, but less so for the Fed (which has very different views on M-policy.) The Fed was doing QE in 2011 while the ECB was raising interest rates. I see no comparison.

    Rademaker, You said;

    “My intuition on this is that trying to offset the 2009 decline with monetary promises is like trying to stop and 18-wheeler by waving a feather in front of it. Too much momentum behind it. The argument falls apart when viewed from a disequilibrium minded point of view, i.e. a realistic one.”

    I’m guessing that you’ve never looked at monetary policy in 1933, when monetary policy sharply reversed NGDP from falling fast to rising fast, in far more difficult circumstantces than 2009. It’s more like Superman than a feather.

    Matt, Thanks for that info. This stuff was actually discussed when I was in grad school in the late 1970s. Larry Sjastaad argued that government output not valued by consumers would cause them to work more, for the reasons discussed above, but government output valued by consumers woud not cause them to work more, as their marginal utility from consumption would not fall. That always seems too conveniently anti-Keynesian, and I can see how there might be government output that is valued in some way by consumers, but not a perfect substitute for private output.

    I put a mea culpa in a more recent post.

    As an aside, I don’t like the way that Keynesians use RGDP as an indicator of the success or failure of stimulus or austerity. (I’d prefer NGDP) But the mechanism Woodford describes does provide some justification. On the other hand it’s never clear if RGDP reflects supply or demand side factors, whereas NGDP seems much more linked to the demand side via which fiscal stimulus is supposed to work.

    In Britain employment is hitting record highs, whereas RGDP growth lags well behind the US. Which is the appropriate indicator for success of fiscal policy?

    At least I can say that if the Fed is targeting NGDP, then the fiscal multplier when applied to NGDP is zero!!

  66. Gravatar of Tyler Tyler
    14. March 2013 at 12:30

    The following is much to flippant. You said:

    “God I hope the Fed doesn’t believe that, as it’s wrong. If they do they should read Bernanke’s 2003 speech where he pointed out that real interest rates are not a good monetary indicator, and that NGDP growth and inflation are best. By those criteria money has been tight since 2008.”

    Lets return to the empirics.

    It is true that since 2008, judging by inflation rates, monetary policy has been tight, but that is an arbitrary and retrospectively biased criterion.

    Why not choose any other date before 2008, which therefore includes it? I will choose a 10 year window in deference to your mentioning of Bernanke’s 2003 speech.

    Over the last 10 years,

    Core PCE has averaged 1.85%. Lets allow that to sink in, Core PCE has averaged, even inclusive of the period you mentioned,2008-now, exactly the Fed target.

    Headline CPI has averaged 2.45% over the same period.

    Now lets turn to monetary policy simply defined as inflation outcomes.

    Lets say that at time zero, January 1st 2013, the Market had an expectation that CPI would grow at 2%, which I get from 10 year Break-evens at that time. Lets also say that the major reason the market assumed that such an inflation rate would be achieved is because it is the Fed’s inflation target. Now the market in 2003 actually understated the CPI inflation rate…

    Again, notice the empirical facts, even inclusive of a dramatic recession, inflation rates have actually been higher over the last 10 years than the market anticipated.

    Probably a good reason for this is the unforeseen commodity price inflation which led to a greater than expected increase in Headline CPI over and against Core PCE, that is a difference between the relevant policy rate and the TIPS reference rate.

    How does this square with your argument that monetary policy, proxied by inflation, has been too tight?

    Lets turn to the present, Inflation Breakevens suggest that over the next 10 years, Headline CPI will be 2.6% per year.

    Now let us again assume that the market expects a large gap between CPI and the Fed relevant policy rate, Core PCE, to remain. (though this is a very charitable assumption as its unclear why the compositional dynamics that were exigent over the last 10 years would be so again.)

    By coincidence, or maybe not, this would translate to an expected path for Core PCE of exactly 2% (2.6%-60bps as derived from 2.45-1.85 above). And a 20 year average gain in core PCE of 1.925% (arithmetic average of the two 10 year preiods).

    I understand price level targeting can be different than inflation targeting, but if the Fed achieves 20 year inflation rates of exactly its target, it is profoundly incorrect and disingenuous to say that their monetary policy has been failed even on their own measure. It has not. Success at achieving your inflation target for that long a period is exactly equivalent to a price level target.

    Of course you know this and the above is not directly relevant for nominal GDP targeting, but it does flatly contradict this statement:

    “God I hope the Fed doesn’t believe that, as it’s wrong. If they do they should read Bernanke’s 2003 speech where he pointed out that real interest rates are not a good monetary indicator, and that NGDP growth and inflation are best. By those criteria money has been tight since 2008.”

    Bernanke has achieved a price LEVEL target, both retrospectively and forward looking.

    And finally we can turn to interest rates. In 2003 the market had a sense of what interest rates would be required in order to achieve their forward looking inflation estimate. And they also knew that if this interest rate forecast was wrong, the Fed would adjust the actual path of interest rates in order to ensure that at least the Price level target was met. Not surprisingly, they were exactly right.

    Real rates were expected to be 2% in order to achieve the inflation target. In fact ex-post overnight rates averaged
    (-)60 bps real. What is amazing, although in many ways, of course not, is that the real overnight rate that on a forward looking basis the market believes will be required to achieve the forward looking inflation rate of approximately 2% (in core PCE, 2.6% in CPI) is -60bps.

    Think about that. In order to achieve the forward looking inflation rates that are almost equal to the achieved inflation rate, the market believes approximately the same interest rate will be required as ex-post was required over the last 10 years. Although of course this forward looking rate is much lower than the equivalent forward looking rate for almost all of the past period.

    (all of the above leaves aside term premium and many other subtle problems in estimating expected forward rates from spot rates)

    The provisioning of Base Money in order to achieve whatever nominal GDP target you want, works because money has a zero interest rate, which in the context of positive nominal GDP growth makes the quantity the public wants to hold of it scarce, though variable. This is no different from the usual effects of interest rates.

    So we see that society’s estimate of the interest rate required to achieve a price level target has changed dramatically,

    As society initially set the interest rate too high, given what subsequently happened, policy rates ex-post have been lower than the market anticipated,

    this was required in order to guarantee that the inflation rates society expected were in fact met, which they were.

    The real rate of return, short-run ex post, to achieve the inflation level target over the last 10 years was -60bps,

    The real rate of return, 10 year forward looking to achieve the inflation target, is now expected to be, lo and behold

    -60bps.

    Id say very few schools of thought would argue that such a forward looking monetary policy, where negative real interest rates are seen to be consistent with a credible price level target, is in fact unaccomodative.

    The interaction of nominal GDP targeting with the above I leave for another time.

  67. Gravatar of 10 Thursday PM Reads | The Big Picture 10 Thursday PM Reads | The Big Picture
    14. March 2013 at 12:30

    […] Time Factory Workers Toiled Longer (Bloomberg) – The flaw at the heart of Keynesian economics (The Money Illusion) but see Four Big Flaws in Progressive Attacks on Keynesianism (Next New Deal) – Inflated […]

  68. Gravatar of Eric Zuesse Eric Zuesse
    14. March 2013 at 13:03

    The IMF’s lead economist did a study of the size of the multiplier and found that he and the rest of the IMF staff had consistently been underestimating its size: http://www.imf.org/external/pubs/ft/wp/2013/wp1301.pdf

    Another IMF study found that “expansionary austerity” (the anti-Keynesian idea) is a myth: http://www.imf.org/external/pubs/ft/wp/2011/wp11158.pdf

    A Bank for International Settlements study found the same: http://www.bis.org/publ/work362.htm

    As regards “The Flaw at the Heart of Keynesian Economics,” its author doesn’t know what he is talking about.

  69. Gravatar of Counterparties: When debt becomes a problem | Felix Salmon Counterparties: When debt becomes a problem | Felix Salmon
    14. March 2013 at 14:19

    […] Wonks Scott Sumner on flawed Keynesianism and blind faith in the 2009 stimulus – The Money Illusion […]

  70. Gravatar of John John
    14. March 2013 at 15:22

    For goodness’ sake. No, no, and no. We are sinking deeper and deeper in this quicksand of theorising without any empirical grounds whatsoever, making up things as we go along and redefining concepts whenever needed. Why is it that people like you seem to think that the ‘macro’ in macro-economics means imaginary?

    “Let’s start with the easy part. New Keynesian theory predicts that the fiscal multiplier will be zero if the central bank is targeting inflation in a forward-looking fashion. That is, increased deficit spending will not increase expected future growth in aggregate demand. The smarter Keynesians know this, but the “smarter Keynesians” are a very, very small group. If you polled PhD trained economists in America, I’d guess less that 10% know this, maybe less than 2%.”

    No. You have just summarised that age old misconception called ‘the treasury view.’ It has been disproven, time and again. And the 98% of economists you deem misinformed know this, for they payed attention when they took History of Economics 101. The treasury view rests on a theoretical fallacy, a misapplication of accounting concepts, and a disregard for temperal factors, in particular inflationary tendedncies. But more importantly, it has been proven empirically wrong, again and again, most recently and comprehensively by Olivier Blanchard.

  71. Gravatar of Scott Sumner on Fiscal Stimulus | Across All Sports Scott Sumner on Fiscal Stimulus | Across All Sports
    14. March 2013 at 15:56

    […] Scott Sumner penned a post arguing against liberals’ blind acceptance that the stimulus worked. He presents a counter factual arguing that if Congress hadn’t passed the stimulus, that […]

  72. Gravatar of Joe Eagar Joe Eagar
    15. March 2013 at 00:15

    For what it’s worth, I think Scott Sumner is basically correct; the side effects of monetary stimulus””currency devaluation””would have been much more beneficial to the economy in 2009 than the side effects of fiscal stimulus (which if anything delayed the necessary adjustments for our economy to grow again).

    The U.S. basically experienced a balance of payments crisis, only without the currency part of a BOP crisis (in financial panics, investors always flee to U.S. Treasuries, even panics originating in America). More aggressive monetary stimulus would have simultaneously lowered the private debt burden and devalued the dollar, redirecting investment into export-producing industries and away from, say, housing.

    In that sense, the 2009 fiscal stimulus was completely unnecessary; by propping up private consumption and (temporarily) subsidizing the housing sector via the first-time home-buyers credit, it slowed the path of economic adjustment. This allowed structural unemployment to take hold.

  73. Gravatar of Graeme Bird Graeme Bird
    15. March 2013 at 01:55

    “At a given level of NGDP, faster real growth will REDUCE inflation. ”

    Yes this is true. But there is some confusion here, since you are equating economic growth with growth in GDP. I know everyone does this, and your basic point is quite correct. But if you went to try and prove this statistically you might run into a lot of problems.

    Here is the formula that is usually used with this sort of thinking:

    MV=PQ

    V is velocity of circulation and it is inferred after the fact. M is the money supply and that needs to be defined carefully. P is the price level. Q is the quantity of output. It is here with Q where people wind up getting in trouble even if the equation is a tautology and the logic behind what you are saying would seem unassailable.

    More output will indeed lead to falling prices. And therefore will be deflationary if falling prices is how you define deflation. But you’ve come up against a grave fault with our economists. They are one metric trick pony’s. Whereas in the old days investment analysts would use a range of metrics to make a decision, our economists, brutal in their simplemindedness insist on the one best metric in theory but always choose GDP or nominal GDP in practice.

    But GDP is a highly netted figure. So while you are right that more output leads to falling prices, we don’t want to always be stuck with output of finished goods and government spending. We might want to sometimes use Gross Domestic Revenue as our metric for output. And if we do so we’ll probably get a better statistical result between theory and practice.

  74. Gravatar of Graeme Bird Graeme Bird
    15. March 2013 at 01:59

    “For what it’s worth, I think Scott Sumner is basically correct; the side effects of monetary stimulus””currency devaluation””would have been much more beneficial to the economy in 2009 than the side effects of fiscal stimulus (which if anything delayed the necessary adjustments for our economy to grow again).”

    But this is a tales you win heads I lose situation. We don’t practice monetary policy. Monetary policy means adjusting the massive subsidy to the financial sector and seeing if they use it to ponzi up more money supply, which extorts value from the rest of the community.

    Real monetary policy is increasing cash in the hands of the public through debt reduction, or increasing the reserve asset ratio. All the things that the central banks actually do are not monetary policy. They are theft. So once again we’ve got to have this attention to detail when we go to the defining of things. Nominal GDP is not the same as aggregate demand. Nor even a good metric for it. Real GDP is not the same as Q or output. And “monetary policy” is not monetary policy properly considered.

    The economists need to get more serious and take more pride in their science.

  75. Gravatar of ssumner ssumner
    15. March 2013 at 06:57

    Tyler, You said;

    “It is true that since 2008, judging by inflation rates, monetary policy has been tight, but that is an arbitrary and retrospectively biased criterion.

    Why not choose any other date before 2008, which therefore includes it? I will choose a 10 year window in deference to your mentioning of Bernanke’s 2003 speech.

    Over the last 10 years,”

    It’s not arbitrary at all. I start the clock in mid-2008 becasue that’s when the economic crisis started. I’m claiming that the economic crisis was caused by money being too tight since mid-2008, not since mid-2203. Milton Friedman once talked about a man drowning in a lake that averaged 4 feet in depth. Averages don’t help if inflation was too high in 2003 and too low in 2008.

    So even by your own criterion you are wrong. But it’s even worse. You talk as if 2% inflation is the Fed’s policy goal. But it isn’t they have a dual mandate. They also focus on employment. If the dual mandate means anything it means inflation should run below 2% in boom years and above 2% in recession years. By that criterion the Fed is even worse.

    And you also ignore the fact that Bernanke mentioned both NGDP growth and inflation. If you average the two then money during the crisis has been tighter than any time since Herbert Hoover was President.

    Regarding PLT, would you average in the French near hyperinflation of the 1920s and near hyperdeflation of the 1930s and pronounce policy “sound” because the price level was maintained in the long run? Maybe if they had an announced policy of returning to the trend line, but certainly not if they did not have such an announced policy, and the big swings were unexpected.

    And this is your problem here. The Fed did not have a 2% PLT policy in effect in 2003-08, thus the public didn’t see the price level in 2008 as being well above trend, and hence in need of deflation. Instead they thought the Fed would keep generating 2% inflation going forward. When we actually got deflation in 2009 it was a big problem precisely because it was unexpected, not built into wage and loan contracts.

    Since 2009 prices have resumed their rise, and unemployment has fallen from 10% to 7.7%, but it would have fallen even faster with a bit of catchup inflation to offset the huge unexpected shock of 2009.

    BTW, I’m not claiming that a modest rise in unemployment in 2008 was inapproriate, I’ve always argued it was to be expected after the overheating of 2006-07. But we moved way too far in the other direction in 2009.

    I have many other posts explaining why interest rates (both real and nominal) are an almost useless indicator of monetary policy.

    Eric, You are in way over your head here. Those studies don’t even address the monetary offset issue, which is the subject of this post.

    John, See my reply to Eric.

    Thanks Joe.

  76. Gravatar of Tyler Joyner Tyler Joyner
    15. March 2013 at 07:17

    Getting to be too many Tylers around here.

    Scott’s counterfactual seems pretty plausible to me. Government policy often seems to be defined in response to public reaction, not in response to a measured set of values. All it would take is a few well placed articles or media spotlights, and people would have been howling for the Fed to take more action. And it is pretty clear that the Fed is not immune to outside influence, since Bernanke has not followed his own recommendations.

    I see no reason why any potential outcry over inflationary monetary policy would have been any worse than the outcry over using huge deficits for a program of fiscal stimulus.

  77. Gravatar of Scott Scott
    15. March 2013 at 13:07

    I may absolutely be reading this wrong, which will no doubt result in my abject embarrassment. But it seems to me that most Keynesian commentators (the good one’s at any rate) grapple quite well with the monetary offset issue – it’s entirely the reason one would assume crowding out through interest rate changes during periods of full capacity utilization (while not an economist, Yglesias has described this very well, as you say).

    The counterargument is that at the zero lower bound conventional monetary channels are ineffective, and absent other accommodative Fed policies fiscal spending will indeed result in a higher multiplier.

    Your supposition, if I read it correctly, is that even in a liquidity trap stimulus is crowded out because its absence would have encourage the Fed to be more accommodating, in this case through PLT.

    I’d like to believe that, but I just can’t reconcile it with the behavior of the current crop of Fed governors. You have a group of people who are still quite hawkish, even having failed at both of their mandates for large swaths of the recession and recovery. Which is to say that I agree with you that the monetary channel is a theoretically more elegant and effective policy tool – I just can’t bring myself to embrace the counter factual given the visible actions of nearly ever central bank currently operating.

    Absent the belief that the Fed is currently crowding out stimulus through monetary inaction, I’m not quite sure the argument that the demand side multiplier must always be zero holds water.

  78. Gravatar of Secondary Sources: Economic Problem, Retirement, Keynesian Flaw – Real Time Economics – WSJ Secondary Sources: Economic Problem, Retirement, Keynesian Flaw - Real Time Economics - WSJ
    15. March 2013 at 14:50

    […] –Keynesian Flaw: Scott Sumner sees a flaw at the heart of Keynesian economics. “There is no empirical study that shows the 2009 stimulus was effective. It’s not even clear new Keynesian theory implies it was effective. It might, but it also might not. There is nothing scientific about “the multiplier.” And many of the arguments made by pundits (with a few notable exceptions like Avent and Yglesias) are deeply misleading to readers. Let’s start with the easy part. New Keynesian theory predicts that the fiscal multiplier will be zero if the central bank is targeting inflation in a forward-looking fashion. That is, increased deficit spending will not increase expected future growth in aggregate demand. The smarter Keynesians know this, but the “smarter Keynesians” are a very, very small group. If you polled PhD trained economists in America, I’d guess less that 10% know this, maybe less than 2%. Even worse, many of the Keynesians who do know this fail to mention it in most of their “pro-stimulus” screeds, thus giving average readers the impression that a positive multiplier is the default assumption, and that it’s up to those who disagree to explain why. No, it’s up to those who believe fiscal stimulus would cause the Fed to stop targeting inflation (or stop Taylor Rule-type policies) to explain why they believe this.” […]

  79. Gravatar of Secondary Sources: Economic Problem, Retirement, Keynesian Flaw – Real Time Economics – WSJ Secondary Sources: Economic Problem, Retirement, Keynesian Flaw - Real Time Economics - WSJ
    15. March 2013 at 14:50

    […] –Keynesian Flaw: Scott Sumner sees a flaw at the heart of Keynesian economics. “There is no empirical study that shows the 2009 stimulus was effective. It’s not even clear new Keynesian theory implies it was effective. It might, but it also might not. There is nothing scientific about “the multiplier.” And many of the arguments made by pundits (with a few notable exceptions like Avent and Yglesias) are deeply misleading to readers. Let’s start with the easy part. New Keynesian theory predicts that the fiscal multiplier will be zero if the central bank is targeting inflation in a forward-looking fashion. That is, increased deficit spending will not increase expected future growth in aggregate demand. The smarter Keynesians know this, but the “smarter Keynesians” are a very, very small group. If you polled PhD trained economists in America, I’d guess less that 10% know this, maybe less than 2%. Even worse, many of the Keynesians who do know this fail to mention it in most of their “pro-stimulus” screeds, thus giving average readers the impression that a positive multiplier is the default assumption, and that it’s up to those who disagree to explain why. No, it’s up to those who believe fiscal stimulus would cause the Fed to stop targeting inflation (or stop Taylor Rule-type policies) to explain why they believe this.” […]

  80. Gravatar of TallDave TallDave
    15. March 2013 at 18:05

    I meant that I don’t agree the Fed’s out of ammo when rates are at zero. New Keynesians have diverse views on that issue. Bernanke and Svensson think they can do more

    I’m increasingly baffled by the fact anyone thinks the Fed could be out of ammo when they’ve never changed the policy.

    Expectations are far more important than any near-term action the Fed takes or announces. It’s like saying “We’rer helpless before the enemy, we used all our bullets and all we have now are these tactical nukes, which we’re busy dropping on ourselves.”

    At any point since 2007 they could have simply announced a 5% inflation target and thereby poured Ex-Lax down the throat of the monetary system.

    I see no reason why any potential outcry over inflationary monetary policy would have been any worse than the outcry over using huge deficits for a program of fiscal stimulus

    Agreed. People seem to keep forgetting this is the worst economy in living memory for most Americans. The only way the outcry would be worse is if monetary stimulus had failed as badly as current policy, that is if we had the same slow growth and same employment numbers.

    Anyways, the “outcry” argument depends on monetary stimulus being the wrong policy. Either it works and outcry doesn’t happen or it doesn’t work and the outcry is irrelevant.

  81. Gravatar of ssumner ssumner
    16. March 2013 at 06:30

    Scott, You said;

    “I may absolutely be reading this wrong, which will no doubt result in my abject embarrassment. But it seems to me that most Keynesian commentators (the good one’s at any rate) grapple quite well with the monetary offset issue – it’s entirely the reason one would assume crowding out through interest rate changes during periods of full capacity utilization (while not an economist, Yglesias has described this very well, as you say).”

    No, it’s completely unrelated to the crowding out issue. One has to do with the effect of fiscal stimulus on real interest rates (holding M constant), and the other has to do with the effect of fiscal stimulus on inflation and NGDP (allowing M to change). And the zero bound doesn’t make the problem go away. Japan’s been at the zero bound for 15 years and their fiscal stimulus has not worked.

    I agree that the Fed has been too hawkish, but surely you must agree that they’ve been quite ACTIVE (QE1, QE2, Twist, QE3, Evans Rule), nudging policy this way and that in response to changing macroeconomic conditions. That’s my point.

  82. Gravatar of Scott Scott
    16. March 2013 at 08:00

    See? In my very first sentence I said I’d make an appallingly basic mistake, and so it proved. That’s truth in advertising.

    I think where we sharply differ, though, is in predicting the willingness of the Fed to embrace a PLT. Again, it’s not an intellectual disagreement – I agree that they really, really should.

    But while the Fed has been incredibly accommodative in terms of their actions (QE and such) the one thing they have not been is accommodative in results. This is a profoundly reactive Fed – moving into action when economic expectations disappoint, but pulling back once forward-looking indicators show some strength.

    To me, the one thing this Fed has been critically unwilling to tolerate – a rise in inflation expectations markedly inconsistent with their 2% target – is also the one thing needed to allow the PLT a period of catch up growth. Maybe this changes with the recent forward guidance they’ve offered. But that, more than anything, leads me to a healthy skepticism of your counterfactual.

  83. Gravatar of Potpourri Potpourri
    16. March 2013 at 13:36

    […] ==> I can tolerate Scott Sumner when he’s bashing Keynesianism. […]

  84. Gravatar of Potpourri « How to Articles Potpourri « How to Articles
    17. March 2013 at 06:22

    […] ==> I can tolerate Scott Sumner when he’s bashing Keynesianism. […]

  85. Gravatar of The Fiscal Multiplier: Sumner vs. Bernanke | Last Men and OverMen The Fiscal Multiplier: Sumner vs. Bernanke | Last Men and OverMen
    18. April 2017 at 01:11

    […] be lower than 7.7% if there had been no fiscal stimulus in 2009.”     http://www.themoneyillusion.com/?p=20017     He provides a rather detailed argument with 11 assumptions-feel free to check it […]

Leave a Reply