Evan Soltas on fiscal policy in a recession

Vaidas Urba and Mark Sadowski directed me to an Evan Soltas post, discussing recent research on fiscal multipliers:

I have been doing some reading for my undergraduate thesis, which looks at the role of credit-supply shocks in the Spain during its housing boom and bust, and I came across some interesting thoughts from Bob Hall. Commenting on research by Alan Auerbach and Yuriy Gorodnichenko, Hall makes some useful points that contradict a lot of the received wisdom about the efficacy of fiscal policy:

I conclude that the chapter uncovers a proposition of great importance in macroeconomics—that the response to government purchases is substantially greater in weak economies than in strong ones. The finding is a true challenge to current thinking. The first thing to clear away is that the finding has little to do with the current thought that the multiplier is much higher when the interest rate is at its lower bound of zero…Standard macro models have labor and product supply functions that are close to linear over the range of activity in the OECD post-1960 sample. The simple idea that output and employment are constrained at full employment is not reflected in any modern model that I know of. [Bolding is by me, not Hall.]

On the economics blogosphere, the “current thought” is also that, because monetary policy is in certain respects (that is, if only by social convention) constrained when the policy rate hits zero, fiscal policy becomes discontinuously powerful at the zero lower bound. Once the policy rate is a quarter of a percentage point, time to turn off fiscal policy, one might infer. Scott Sumner is one of the clearest and most persuasive exponents of this view—see here, for instance.

I thank Evan for calling me “persuasive”, but there are days when I feel like I’m not making any headway.  Just to reiterate, here are my basic views:

1.  There is no such thing as “the” multiplier, in the sense of it being a deep parameter.  It depends on how monetary policy responds.  It’s an empirical question, and “the” multiplier may be time-varying.

2. If monetary policy is optimal, there will be no demand-side multiplier effects, even at the zero bound.  (There may be supply-side effects.)  When Evan refers to “this view”, he means that I don’t favor fiscal stimulus at positive rates, not that I do favor it at zero rates.

3.  Previous studies suggesting a positive multiplier at the zero bound are marred by the inclusion of countries with and without monetary offset (with the eurozone being a particularly significant problem).

4.  There was no fiscal multiplier effect when the Federal government suddenly reduced the deficit from $1060 billion in calendar 2012 to $560 billion in 2013, despite predictions to the contrary by over 350 Keynesian economists.

I took a look at the Alan Auerbach and Yuriy Gorodnichenko paper, as well as Robert Hall’s comments.  I see two potential problems:

1.  The empirical results seem quite weak to me, unless I’m missing something.  Hall says:

Their point estimate is that one added dollar of government purchases results in about $3.50 of added GDP when the economy is weak, with a 90% confidence interval running from 0.6 to 6.3.

I’m very weak at econometrics, but this doesn’t seem at all persuasive to me.  I’ve always thought the standard 95% confidence interval is way too lenient, and helps explain all the bogus studies that cannot be reproduced.  It’s an invitation to data mining.  So why did they use the 90% confidence interval, instead of the already lenient 95%?  Perhaps because at 95% the interval would include zero.  In other words there would appear to be no statistically significant evidence of any multiplier effect.

2.  Let’s say I’m wrong about the econometrics (quite possible.)  My other concern is that I don’t see any evidence that they discriminated between countries with and without independent monetary policies (no list of countries was included).  In other words, they should have excluded countries under fixed exchange rate regimes, as well as those in a single currency like the eurozone.  Studies by people like Mark Sadowski, Kevin Erdmann, and also Benn Steil and Dinah Walker, suggest that when the data is limited to countries with an independent monetary policy, then the multiplier effect seems to nearly vanish, at least during the recent recession.

I don’t want to sound too dogmatic here.  I have no doubt that WWII military spending caused measured RGDP to rise (although consumption and living standards fell.)  So it’s possible that this paper did correctly find a positive multiplier. But I am not convinced that the Auerbach/Gorodnichenko paper has established a positive multiplier for countries with independent monetary policy regimes.

As they say, “more research is needed.”


Tags:

 
 
 

53 Responses to “Evan Soltas on fiscal policy in a recession”

  1. Gravatar of Ironman Ironman
    22. March 2016 at 18:38

    I have a suggestion for a potential research direction….

    Why not? They worked in Spain, Greece and the U.S. – although in the latter case, I really do need to go back with the revised GDP data to see if the GDP multiplier is still about equal to 1 for QE.

  2. Gravatar of Gary Anderson Gary Anderson
    22. March 2016 at 18:43

    So, if WW2 military spending boosted GDP why would not Sanders’ massive infrastructure plan do the same thing and what are the pitfalls? Is the downside the day after the stimulus stops? I remember when my dad had a very old car, and when I was 4 in 1953 he bought a new car. So, it took awhile for things to get better after the post WW2 bust.

    Can we handle a bust on that scale?

  3. Gravatar of Cliff Cliff
    22. March 2016 at 18:59

    Post-WWII bust??

  4. Gravatar of E. Harding E. Harding
    22. March 2016 at 19:12

    @Cliff

    -Yes. Gary’s right. As for Gary’s comment, it’s a bit like comparing the U.S. quickly building a massive border wall v. becoming a garrison state on the matter of North Korea. The former is temporary and causes a short-term boost in RGDP, the latter is permanent and causes a long-term stagnation in RGDP.

  5. Gravatar of Benjamin Cole Benjamin Cole
    22. March 2016 at 19:26

    Nice post. It has been many a moon, but I did “serious social science work” at one point, and I can remember a professor telling me to “collapse some cells” to get a better result. Something moving from deciles to quintiles to prove people were not moving around in income brackets.

    Sumner raises interesting points about the selection of confidence intervals, or that no list of countries is provided.

    One can data mine on Wall Street with similar results. Search hard enough, and you will find that, say, all tech companies that reported Q3 EPS up more than 10% are good buys. It is a true observation. Doesn’t mean that it will repeat.

    I do think that supply-side tax cuts in combination with heavy-duty QE will have the desired effects. If not, then fine, we pay off the debt and have lower taxes.

    Dallas went into CPI-deflation in 2015, for parts of the year. Why do I mention Dallas? A boom city in a boom state in a nation with a seven-year and running recovery.

    And it hit deflation.

    Do you think the old demand-pull inflation model is dead? How about the Phillips Curve? How about Jimmy Hoffa? Jeb Bush’s presidential hopes?

    So why is the Fed (and academia) so squeamish about boosting demand?

    OT:

    “On Monday night, CNN’s Wolf Blitzer asked Republican presidential candidate Ted Cruz about Frank Gaffney, a conspiracy theorist Cruz recently added to his list of foreign policy advisers. Gaffney, who worked in the Pentagon during the Reagan administration, has a long history of accusing everyone from liberal Supreme Court Justice Elena Kagan to conservative anti-tax zealot Grover Norquist of bizarre, convoluted connections to “radical Islam.”

    Blitzer asked Cruz whether he agreed with a number of Gaffney statements, including his 2009 claim that “Barack Hussein Obama would have to be considered America’s first Muslim president.” But Cruz wouldn’t disavow Gaffney. “I appreciate his good counsel,” he said.”

    –30–

    So who is the American Schicklegruber?

  6. Gravatar of Evan Soltas Evan Soltas
    22. March 2016 at 19:53

    Thanks for the response, Scott. I agree with your basic views (1), (3), and (4). (2), I think, is not super obvious to me. In a world where the effects of policies are uncertain, we should (i.e. it is optimal to) conservatively use any single policy, and to mix the effects of multiple policies.

    And I guess (2) really gets at a more interesting issue than any single empirical finding. Particularly when the CB is likely to be “ZLB-constrained” (if only in their head) in the future, it makes sense to not offset fiscal stimulus, even if interest rates are positive.

    The bigger-picture point is that the existence of the ZLB actually has pretty important implications for monetary policy when the CB is at positive interest rates.

  7. Gravatar of TallDave TallDave
    22. March 2016 at 23:52

    Not a strong correlation and dependent on a false belief about CBs by CBs, but here’s probably a place for a small countercyclical fiscal policy.

    Unfortunately such efforts typically vanish in the noise of the welfare state, which voraciously shovels taxes into its maw in good times and in bad times has often so bankrupted the state that solvency requires procyclical policy.

  8. Gravatar of Daniel Daniel
    23. March 2016 at 01:36

    Yo Scotty,

    Given what just happened in Brussels, please tell us some more how Eastern Europeans are a bunch of racists and fascists for not wanting to take part in Merkel’s insanity.

  9. Gravatar of Major.Freedom Major.Freedom
    23. March 2016 at 03:49

    “I have no doubt that WWII military spending caused measured RGDP to rise (although consumption and living standards fell.)”

    This is also true for inflation.

  10. Gravatar of Major.Freedom Major.Freedom
    23. March 2016 at 03:53

    Evan wrote:

    “In a world where the effects of policies are uncertain, we should (i.e. it is optimal to) conservatively use any single policy, and to mix the effects of multiple policies.”

    There is no way to apodictically conclude any single policy is being used “conservatively” as opposed to something else though. The economy is complex, and what may have been conservative in the past might not be today, and vice versa.

  11. Gravatar of jonathan jonathan
    23. March 2016 at 04:23

    Thanks in part to your writings, I’ve recently become more skeptical of fiscal policy. I mostly agree with (2), and entirely agree with (1) and (3).

    That said, I’m not sure we can conclude much from (4) for three reasons. First, we don’t know the counterfactual path of GDP, so we can’t compute the multiplier. Second, if you eyeball the NGDP series 2013 actually looks a little worse than 2012 or 2014 (3.1% growth vs. 4.1%). And third, the growth rate of total government spending didn’t change much between 2012 and 2013.

    Here are the time series:
    https://research.stlouisfed.org/fred2/graph/?g=3TyE

  12. Gravatar of ssumner ssumner
    23. March 2016 at 05:05

    Evan, Thanks for the comments.

    You said:

    “In a world where the effects of policies are uncertain, we should (i.e. it is optimal to) conservatively use any single policy, and to mix the effects of multiple policies.”

    I suppose this depends on what you think of the significance of my point 4, and also whether you buy my claim that monetary policy is costless while fiscal policy is costly. I can’t say you are wrong, but all in all I’m not yet convinced. I also don’t think economists as a profession should be relying on fiscal policy as a tool, because future Congresses are far less likely than future Fed’s to do what’s necessary.

    As you know, I am a libertarian, so I hope I am wrong. I hope I will be able to advocate huge tax cuts in the next recession, with a clear conscience. But I’m not there yet.

    I hope the Fed has learned some lessons from 2008-09, and will be far more aggressive in the future. The Fed has generally learned lessons from past fiasco’s like 1929-33 and the Great Inflation, and I’d expect them to gradually improve over time. Level targeting would mostly eliminate the zero bound problem, for instance.

    Daniel, And I thought the left was the home of wimps that wanted big government to protect us from all the scary dangers out there. I hate to break it to you, but people die every day. Like Ben Franklin, I’m not willing to trade freedom for a bit more security. (Nor do I think it would work.)

    If you want to save 5000 lives a year, we could just legalize the kidney market. In a world of 7 billion people, the risk of being killed by terrorists is utterly infinitesimal—especially in rich countries. If it gets 10 times worse, I’d still say the same. You’d be better off focusing on a less emotional issue like auto safety. But then I guess you are probably a Trump voter, who thinks with his reptilian brain.

    Jonathan, NGDP rose faster in 2013 (4.05%) than 2012 (3.24%), (Q4 to Q4).

  13. Gravatar of james elizondo james elizondo
    23. March 2016 at 06:48

    Mr. Sumner in your article “why the fiscal multipler is roughly zero” you said “Congress passed a nearly $800 billion stimulus in early 2009, yet growth remained sluggish”

    Why would we expect $800 billion stimulus package to largely influence a 16 trillion dollar economy? Do we spit in the ocean and expect the tide to change course?

    you said “By now we know that central banks are not out of policy options when rates fall to zero.”

    But you assume the options that remain at the zero bound are optimal? Are they? As Summer argues these tools have experience diminishing returns and promote asset bubbles through by encouraging excessive leverage. Also lax monetary policy may push the future neutral real rates down even further by speeding up investments and pulling forward demand. Fiscal policy may reduce national savings thereby raising the neutral real interest rate and stimulating growth.

    The Fed as tools at the zero bound but I challenge your notion that they’re optimal.

  14. Gravatar of jonathan jonathan
    23. March 2016 at 07:04

    Scott:

    Interesting, this seems to be a case where it matters a lot whether you take the average or the difference first.

    The reason seems to be that 2012 Q4 GDP was relatively low compared to the rest of 2012, while 2013 Q4 GDP was relatively high compared to the rest of 2013. Thus computing YoY growth rates using Q4 makes 2013 look better than 2012, while the opposite is true if you look at all four quarters.

    Given how noisy quarterly data is, I prefer my method. Using a single quarter YoY growth rate isn’t very robust. If we use Q3, then 2012 GDP growth was 4% and 2013 growth was 3%, the opposite of using Q4.

    I also have concerns about the timing of the experiment. Since in the NK model the main effect of fiscal policy is through inflation expectations, we have to look at how expectations of future government spending changed. And there wasn’t a single announcement here, because there was a lot of uncertainty about government spending during this period.

    And this also doesn’t resolve the problem that there wasn’t a clear reduction in government spending in this period anyway. The main movement I see was a big spike in 2012 Q4.

    Here is the raw data:
    https://research.stlouisfed.org/fred2/graph/?g=3Va9

  15. Gravatar of Ray Lopez Ray Lopez
    23. March 2016 at 07:33

    Sumner: “I’m very weak at econometrics, but this doesn’t seem at all persuasive to me. I’ve always thought the standard 95% confidence interval is way too lenient, and helps explain all the bogus studies that cannot be reproduced. It’s an invitation to data mining. So why did they use the 90% confidence interval, instead of the already lenient 95%?” – for the same reason in Blanchard’s Econ 101 textbook he cites an author that shows monetarism works (the IS-LM model) but only to 60% confidence*

    Sumner: “Mark Sadowski, Kevin Erdmann, … suggest that when the data is limited to countries with an independent monetary policy, then the multiplier effect seems to nearly vanish, at least during the recent recession.” – pretty weak. A newly minted middle aged PhD (Sadowski), and a guy who is trying to prove there was no housing bubble (Erdmann) show that for certain countries the multiplier ‘seems to nearly vanish’ for the most recent recession (N=1 sample size). Nice.

    @the Thai Turkey farmer: Google “Hitler Never Really Was Schicklgruber” and educate yourself. Can’t believe Marcus Nunes is going to use your input for anything. Bet that gig don’t last long.

    RL

    * Olivier Blanchard “Macroeconomics”, 2nd edition., p. 96, 5-6, ‘Does the IS-LM Model Actually Capture What Happens in the Economy?’ Shows that for data 1960 to 1990, that a 1 % increase in Fed funds rate shows, over 4-8 quarters, a decrease in sales, decrease in output, decrease in employment, increase in unemployment rate up to 6 quarters (then a decrease after 8 to the original level). However, the “confidence band” is only 60% probability, not the usual 95% confidence.

  16. Gravatar of Student Student
    23. March 2016 at 08:00

    Excellent stuff all around… including many of the comments.

  17. Gravatar of bill bill
    23. March 2016 at 09:14

    @Jonathan,
    I went to the Fred graph, then downloaded the data. I believe the dip that is showing as 2013 is technically 1/1/2013. i.e., the little dip is 2012 annual growth. Regardless, such a small dip for a $500 billion deficit reduction would refute the concept of the multiplier. A small rise annihilates the concept.

  18. Gravatar of bill bill
    23. March 2016 at 09:18

    PS
    Here’s the data (last column is GDP). Looks like 2013 was the HIGH year between 2012 and 2014.
    2011-01-01 0.7 3.7
    2012-01-01 -0.6 4.1
    2013-01-01 -0.3 3.1
    2014-01-01 2.3 4.1
    2015-01-01 2.3 3.4

  19. Gravatar of E. Harding E. Harding
    23. March 2016 at 09:21

    “I hate to break it to you, but people die every day. Like Ben Franklin, I’m not willing to trade freedom for a bit more security. (Nor do I think it would work.)”

    -Sorry, those are not the options on the table. The options are immigration restriction v. national security state (though you can have both, as Trump prefers). I prefer the former alone.

    “Daniel, And I thought the left was the home of wimps that wanted big government to protect us from all the scary dangers out there.”

    -Those big scary dangers exist (though they were, in this case, deliberately created by big government). The Left desires to protect us from social insecurity. The Right wants to protect us from physical insecurity.

    ” You’d be better off focusing on a less emotional issue like auto safety.”

    -Aw, come on. You would be better off focusing on a less emotional political figure, like Scalia’s replacement or the Speaker of the House. Nobody called them Hitler (though they’re controversial, too).

    And focusing on tragedies irrelevant to Dan’s remark is a total cop-out and non-sequitur.

    “But then I guess you are probably a Trump voter, who thinks with his reptilian brain.”

    -You think solely with your reptilian brain, too, in the case of Trump. I think with both brains.

  20. Gravatar of Carl Carl
    23. March 2016 at 09:24

    Scott:
    The effect of terrorism is two fold. It kills and it creates terror. Counting the deaths is insufficient to measure its importance. The reaction to the terror threatens freedom and prosperity. It’s an analogue to how bad monetary policy leads to economic fear and to calls for more government intervention in the economy. For the former, you call for ignoring it, for the latter you call for NGDPLT to mitigate the risk to economic freedom (among other reasons).
    I think our involvement in Iraq and Libya were counter productive, but I also think that Salafi terrorism is non-trivial for the West.
    .

  21. Gravatar of Derivs Derivs
    23. March 2016 at 10:20

    “the risk of being killed by terrorists is utterly infinitesimal—especially in rich countries”

    So I had shit luck I guess. I remember back in ’93, where I was in WTC all that happened was the lights turned on and off… then everyone had to evacuate and I remember seeing all these people on the lower level all covered in smoke, hearing people died, thinking what could be worse than this.

    ’01 I refuse to talk about.

  22. Gravatar of Art Deco Art Deco
    23. March 2016 at 10:35

    Daniel, And I thought the left was the home of wimps that wanted big government to protect us from all the scary dangers out there. I hate to break it to you, but people die every day.

    So you’ll be taking that cheap apartment in Roxbury?

    That aside, they do not wake up every day to find their country overrun with an indigestible 7-digit population of foreigners drawn from the most troublesome demographic segments. Importing the refugees into Europe was gratuituos. Seeing to their sustenance can be done just as well in camps adjacent to Syria, and their eventual repatriation can certainly be managed better if they’re concentrated in those loci.

  23. Gravatar of Britonomist Britonomist
    23. March 2016 at 11:19

    Ray, I don’t have that book. But a *multivariate model* fit is not tested using a standard “confidence interval”, but its explanatory power is measured typically with an ‘R squared’ value, an R squared of 60% for a simple multivariate model is highly respectable for social science (where data is very noisy), it means 60% of the data is explained by this simple model alone. So depending on what you mean, that 60% figure could either be very good or very bad. What you’re saying right now is useless without more context, or the study being referenced

  24. Gravatar of Ray Lopez Ray Lopez
    23. March 2016 at 11:31

    @Britonomist – it’s not what you say, it’s not a “multivariate model’ involving R^2 but what I said. The quoted text is (among other things): “The true value of the effect [referencing various graphs in response to a Fed funds rate increase] lies within the two dotted lines with 60% probability. For this reason, the space between the two dashed lines is called a confidence band”. (Technically, it’s not quite right to call this 60% probability, but that’s a small quibble). Why 60% and not 95%? Obviously there would be no effect at 95% confidence.

    The source for Blanchard: Lawrence Chambers, Martin Eichenbaum [Barry’s brother?!], and Charles Evans, “The Effects of Monetary Policy Shocks: Evidence from the Flow of Funds”, Review of Economics and Statistics, Feb. 1996, 78-1.

  25. Gravatar of bill bill
    23. March 2016 at 13:01

    I’m with Scott. I refuse to be terrorized. I hope that we are able to find a better balance as a society.
    On a related note, Marginal Revolution had a post about bus safety. Busses are 67 times safer than cars. Several mandated safety requirements slow busses down and slow bus speed reduces ridership.

  26. Gravatar of LK Beland LK Beland
    23. March 2016 at 13:29

    How many countries both:
    1-had independent monetary policy
    and
    2-were at the ZLB?

    I think an ideal study of multipliers at the ZBL should only include countries/currency zones that satisfy both conditions. Are there such studies?

  27. Gravatar of Anthony McNease Anthony McNease
    23. March 2016 at 13:57

    The only positive effects for the US I can think of would be in AS where fiscal funding increases the amount of stuff we have or in productivity by funding infrastructure or R&D into new and better technologies.

  28. Gravatar of james elizondo james elizondo
    23. March 2016 at 14:15

    Mr. Sumner

    “There was no fiscal multiplier effect when the Federal government suddenly reduced the deficit from $1060 billion in calendar 2012 to $560 billion in 2013, despite predictions to the contrary by over 350 Keynesian economists”

    Do you say this because during that time gdp grew at decent rate? If so please enlighten me why that is. Refuting spending multiples by observing that growth in gdp coincided with austerity ignores the business cycle and counter-factual comparisons (in my humble opinion)

    As for the business cycle even Keynes admitted that economies can be self-correcting (use, decay, and obsolescence can create scarcity of capital). Its just an extremely long and painful process. I think some of the growth in 2013 can be attributed to this self-correction. If that’s the case the economy would have grown with or without austerity/stimulus. The question then becomes what would have been the counter-factual outcome i.e less austertiy, more stimulus? In my eyes although we saw growth we significantly underperformed in terms of gdp potential output gap. If that’s the case I don’t understand how proponents of austerity can claim victory even if there was a monetary offset.

  29. Gravatar of E. Harding E. Harding
    23. March 2016 at 15:52

    james, what part of “despite predictions to the contrary by over 350 Keynesian economists” don’t you understand?

  30. Gravatar of Dustin Dustin
    23. March 2016 at 16:38

    Carl
    Well said – my thoughts exactly, including the MP analogy.

  31. Gravatar of ssumner ssumner
    23. March 2016 at 16:49

    James, You said:

    “As Summer argues these tools have experience diminishing returns and promote asset bubbles through by encouraging excessive leverage.”

    I strongly disagree with this claim. There’s not a shred of evidence to back this up, not a shred. Monetary policy needs to focus on macro stability. Period, end of story. If we start getting distracted with other targets like asset prices, we will end up stabilizing neither. In 1929 the Fed tried to stabilize stock prices, and ended up destabilizing both stock prices and the economy. it’s not skilled enough to stabilize asset prices, so don’t even try.

    The Fed had plenty of tools to use even before they started raising rates. The reason the Fed did not use them was because their forecast for future inflation and employment growth was on target. I think they were wrong, but Bernanke always insisted they were not out of ammo, and he was right. They ended QE1 and QE2 because they thought AD was adequate, not because they had run out of bonds to buy.

    If the zero bound really is a problem (which I doubt) the solution is a 4% inflation target, not fiscal policy (which the GOP would never approve in 100 years).

    These theories that monetary policy is not an optimal stabilization tool at the zero bound are all ad hoc, and were not taken seriously even a decade ago. There is no serious research supporting most of the claims I’ve seen. One can always write down a model to get any policy objective you’ve reached for other reasons.

    Monetary stimulus is just as likely to raise as to lower real interest rates, in the medium to long term. (I agree it lowers them in the short run). If the monetary stimulus is strong enough, the economy will grow much more rapidly and real rates will tend to be pushed up by the strong growth. Conversely, the very tight money of 2008 led to much lower real interest rates in 2009, as NGDP growth plunged.

    Jonathan, You do realize that all of your criticisms of my argument, if true, completely invalidate almost all the pro-Keynesian blog posts of Krugman, et al, over the past 10 years. The Keynesians have used the same procedure as I have. I agree that expectations are important, but people like Krugman mocked that argument when conservatives pointed out that expectations of Obama’s stimulus was not helping the economy in early 2009. Krugman said it hadn’t taken effect yet.

    You are wrong about the GDP data. The year over year figures for Q4 already give exactly equal weight to each quarter of 2013. Your procedure mixes up growth in 2012 and 2013, which makes it a poor tool for evaluating an austerity that began on January 1, 2013.

    Bill, See my response to Jonathan, growth actually sped up in 2013.

    Harding, I never compared Trump to “Hitler” the modern symbol of evil, you misunderstood those posts. I compared him to the German nationalist running in democratic elections in 1929. The politician Hitler didn’t run on the platform of pushing Germany into WWII and creating a Holocaust, he ran on a platform of Making Germany Great Again. And demonizing unpopular minorities. Yes, Trump and Hitler are radically different personalities. Trump’s bad, but not that bad. And of course our democracy is too robust to be taken away by a President Trump. We’ll survive even that.

    Carl, You said:

    “The reaction to the terror threatens freedom and prosperity.”

    That’s exactly my point.

    I agree that ISIS is a big problem, but mostly to people in the Middle East. I’m not saying terrorism is not a problem at all, every death is a tragedy. But it’s not high up on the list, like the ban on organ transplants. Tat creates far more tragedies than terrorism. If people were less like human beings and more like Vulcans, the world would have fewer problems (although it might also be more boring.)

    Derivs, You said:

    “So I had shit luck I guess.”

    Yup. But look at the bright side, you are still better off than the soldiers in the trenches during WWI. When you live in a paradise (and we do) even small tragedies seem huge.

    LK, Yeah, that would be a good study.

    James, You said:

    “Do you say this because during that time gdp grew at decent rate? If so please enlighten me why that is. Refuting spending multiples by observing that growth in gdp coincided with austerity ignores the business cycle and counter-factual comparisons (in my humble opinion)”

    That’s fine, but you don’t seem to understand that the burden of proof is on the Keynesians. It wasn’t me that said 2013 would be a test of market monetarism, it was Paul Krugman. It wasn’t me who signed the letter warning that austerity could led to recession, it was 350 Keynesians. Every time an economy slumps after austerity, they say “see, I told you so.” But those examples also have problems, don’t they? I’m just following their method. Now if there are more scientific ways of determining the efficacy of fiscal stimulus, I’d love to see it. I’m a libertarian–I’d love to advocate massive tax cuts, if I could be convinced they’d work.

  32. Gravatar of Ryan Murphy Ryan Murphy
    23. March 2016 at 17:05

    Scott, the issues with weak statistical significance in the empirics side of spending multipliers is way worse than you think. Tons of papers can’t even achieve 90% and they actually fake it with 68%. Here is my paper at Econ Journal Watch documenting this fact: http://econjwatch.org/articles/unconventional-confidence-bands-in-the-literature-on-the-government-spending-multiplier

  33. Gravatar of Major.Freedom Major.Freedom
    23. March 2016 at 17:15

    In response to:

    “As Summer argues these tools have experience diminishing returns and promote asset bubbles through by encouraging excessive leverage.”

    Sumner wrote:

    “I strongly disagree with this claim. There’s not a shred of evidence to back this up, not a shred.”

    Of course YOU can’t see any evidence, because your theoretical premises make it impossible to even understand and identify the evidence as such. You see massive debt fueled bubbles and you are intellectually ill-equipped to link them to monetary policy.

    The fact you say “not a shred” more than once suggests you are motivated more by political ideology than truth. Indeed, the very next comment you made:

    “Monetary policy needs to focus on macro stability. Period, end of story.”

    Shows your concern is not over whether or not the Fed causes unstable debt fueled bubbles (it does), but that they merely should not worry about them. And why? Hitler of course:

    ” If we start getting distracted with other targets like asset prices, we will end up stabilizing neither. In 1929 the Fed tried to stabilize stock prices, and ended up destabilizing both stock prices and the economy. it’s not skilled enough to stabilize asset prices, so don’t even try.”

    Notice your inability to understand the world through anything but a control freak’s goggles? It is as if you believe the mere absence of thought on the part of the Fed regarding asset prices, is equivalent to the Fed not affecting those prices adversely.

    You say the Fed is not skilled enough to stablize asset prices. Sure, but they don’t have to be regarded as anything else simply by virtue of the fact that they adversely affect asset prices. You’re more worried about the Fed focusing on asset prices than you are with simply stating a truth that the Fed causes asset bubbles. Don’t say the emperor is not wearing any clothes because then there would be emperor clothing intervention!

    It is not any lack of evidence (there is in fact evidence) but rather a desire to avoid, at all costs, even truth, any validation or consideration of central banks focusing on something other than NGDP. It is hand waving, nothing more.

    The claim there is “not a shred of evidence” is really just a statement of your theory which prevents you from identifying the evidence.

    The easiest way for you to at least start on a path of understanding the evidence is to ask what would the evidence have to look like if it had looked as though monetary policy did cause debt fueled asset bubbles? It would look like….history wouldn’t it?

  34. Gravatar of Mark Mark
    23. March 2016 at 17:40

    Scott, are you at all familiar with (or have any opinion on) Valerie Ramey’s research on fiscal multipliers?

  35. Gravatar of Christian List Christian List
    23. March 2016 at 19:14


    And I thought the left was the home of wimps that wanted big government to protect us from all the scary dangers out there. I hate to break it to you, but people die every day.

    Image Trump would say something as cynical and heartless. Scott and the media would be outraged. But when someone like Scott says something like this, it’s supposed to be a great idea.

  36. Gravatar of Saturos Saturos
    23. March 2016 at 21:47

    Yeah I’m pretty sure Evan is wrong. There is a ZLB in the Fed Chair’s head, but it never seems to stop them from doing monetary offset of fiscal policy. I think economists can carry any amount of cognitive dissonance so long as they stop themselves thinking about it too much. (Kind of like all scientists do with philosophy.)

  37. Gravatar of Postkey Postkey
    24. March 2016 at 01:09

    “In 1929 the Fed tried to stabilize stock prices, and ended up destabilizing both stock prices and the economy. it’s not skilled enough to stabilize asset prices, so don’t even try.”

    Prof. R. Werner thinks that, in the U.K., the monetary authorities could distinguish between ‘productive’ and ‘unproductive’ credit.

    “Importantly for our disaggregated quantity equation, credit creation can be disaggregated, as we can obtain and analyse information about who obtains loans and what use they are put to. Sectoral loan data provide us with information about the direction of purchasing power – something deposit aggregates cannot tell us. By institutional analysis and the use of such disaggregated credit data it can be determined, at least approximately, what share of purchasing power is primarily spent on ‘real’ transactions that are part of GDP and which part is primarily used for financial transactions. Further, transactions contributing to GDP can be divided into ‘productive’ ones that have a lower risk, as they generate income streams to service them (they can thus be referred to as sustainable or productive), and those that do not increase productivity or the stock of goods and services. Data availability is dependent on central bank publication of such data. The identification of transactions that are part of GDP and those that are not is more straight-forward, simply following the NIA rules.”
    http://eprints.soton.ac.uk/339271/1/Werner_IRFA_QTC_2012.pdf

  38. Gravatar of Postkey Postkey
    24. March 2016 at 01:16

    @james elizondo.

    You may be interested in this?

    “warren mosler writes:
    First, I had been looking for 4% growth for 2013 and scaled back to 2% due to the tax increases and sequesters, and I thought it would continue to weaken until deficit spending increased.
    Turns out there was an increase in private sector deficit spending/credit expansion on oil and gas exploration and production that offset the 2013 fiscal adjustments and further expanded in 2014 to further support GDP growth.
    The increase in energy related deficit spending is now over due to the Saudi price cut, and unless deficit spending elsewhere accelerates seems to me GDP growth will quickly evaporate.”
    http://econlog.econlib.org/archives/2015/01/the_keynesians.html

  39. Gravatar of Ray Lopez Ray Lopez
    24. March 2016 at 01:47

    http://krugman.blogs.nytimes.com/2015/06/22/2013-and-all-that –> on the 350 Keynesian economists letter and why Krugman is right (general equilibrium argument)

    http://econlog.econlib.org/archives/2015/06/the_other_thing.html –> Sumner’s rebuttal to the above argument

    http://www.nationalreview.com/article/415394/krugmans-fatal-conceit-ramesh-ponnuru –> national review article supporting Sumner

    After reviewing these articles, I declare the winner is: Paul Krugman. Why? Because if you accept Sumner’s argument you must somehow assume that a proposed tax increase must of had an immediate effect on US consumers (Lucas critique), which is not clear it did. Instead, as the National Review article points out, the Feds continued to spend, even more than in previous years (in actual numbers), just not as much as before relative to GDP. Look at the raw data: 2011 (before austerity), US Fed Govt received (and spent more than this amount): 2.3T (income tax was 1.3T, rest was largely social security taxes); 2012 (during austerity) US Govt rec’d 2.4T (1.4T in income taxes), and, 2013 (after austerity) 1.6T rec’d (and spent more than) and of this amount 1.6T in income taxes.

    Long story short: Krugman is largely right. The so-called austerity is not really such. The US Fed gov’t continued to spend more in 2012 than they did before austerity, i.e. in 2011, but just not as much “relative to GDP” since GDP increased after austerity.

    Sumner of course is also technically right, in that this kind of austerity did not hurt the economy, contrary to the “350 Keynesian Economists” letter. But there was no actual year-to-year cut in US Fed gov’t receipts or expenditures, just a decline relative to GDP.

  40. Gravatar of Ray Lopez Ray Lopez
    24. March 2016 at 01:51

    @myself- typo: 2014 is 2.6 T rec’d, and 1.6T was income tax. The point being: a steady increase in US Fed revenue rec’d since 2011, before ‘austerity’ (and even more spent, recall the US always spends more than it receives since that one year in the Clinton administration). Also you can arguably call what I termed the “Lucas Critique” something more like “Ricardo-Lucas equivalence”, but they’re close cousins of one another actually.

  41. Gravatar of Ray Lopez Ray Lopez
    24. March 2016 at 01:59

    BTW, Kevin Drum disputes my argument above, as he uses inflation adjusted real GDP/capita, which shows a decline in US gov’t total expenditures since 2009 to end of 2013 (“Non-Chart of the Day: Where’s the Austerity? —By Kevin Drum | Sat Jan. 10, 2015 7:00 PM EST “) from 19.5k $/capita to about 19k, from 2011 to end of 2012, or $500 per capita in the period in question. Still, it’s a small drop so arguably this is not really austerity, the way most people consider the word austerity. A $5k/capital drop is austerity, not 500 dollars per capita, IMO.

  42. Gravatar of Art Deco Art Deco
    24. March 2016 at 02:43

    The politician Hitler didn’t run on the platform of pushing Germany into WWII and creating a Holocaust, he ran on a platform of Making Germany Great Again. And demonizing unpopular minorities.

    Read Mein Kampf. Hitler wasn’t shy about his rejection of parliamentary institutions or his crank racialism. All of this was in public print in 1928 and 1930 and 1932 (there were no federal elections in 1929 in Germany). Trump’s never written anything remotely like that. Trump does not ‘demonize’ anyone. The Anointed have their mascot groups and anyone discussing their mascot groups and not expressing deference to their mascot groups at the same time is conceptualized as ‘demonizing’ them. That’s not Trump’s fault; it’s yours.

  43. Gravatar of Art Deco Art Deco
    24. March 2016 at 02:47

    Image Trump would say something as cynical and heartless. Scott and the media would be outraged. But when someone like Scott says something like this, it’s supposed to be a great idea.

    It doesn’t seem to occur to libertarians living in tranquil suburbs that much of the world is not a serene place (or maybe they fancy the unserene portions are populated by losers getting what they deserve).

  44. Gravatar of Bonnie Bonnie
    24. March 2016 at 05:04

    It’s impossible to prove a negative, but if there were persuasive evidence it might be found in market reactions to the passage of the American Recovery and Reinvestment Act. As I recall, it didn’t appear to stem the markets steep decline in early 2009. So if there is any benefit at all from increased government purchases in a weak economy, at least in that case, it appears negligible.

  45. Gravatar of james elizondo james elizondo
    24. March 2016 at 08:05

    Mr. Sumner

    My arguments are all tied to secular stagnation. I think Summers makes good points. If secular stagnation does have merit many of your arguments can be disputed. If it doesnt well you know. Your quotes aren’t in order btw and may be taken out of context sorry. I try to be fair though

    “The Fed had plenty of tools to use even before they started raising rates……..” and

    “Monetary stimulus is just as likely to raise as to lower real interest rates, in the medium to long term. (I agree it lowers them in the short run). If the monetary stimulus is strong enough, the economy will grow much more rapidly and real rates will tend to be pushed up by the strong growth. Conversely, the very tight money of 2008 led to much lower real interest rates in 2009, as NGDP growth plunged.”

    Yes the Fed has plenty of tools even at the zero bound but the question is are these the best tools? You assume they are but there’s reason to believe they are not. You agree that monetary policy could make the neutral real rate needed to get full employment even lower. Or it could make it higher. I think the weight is on the former with monetary policy accelerating investment and pulling demand forward making the problem worse. We could just keep going full blast with monetary policy but as Summers says we’ll have to keep increasing the dosage to maintain the effect.

    However fiscal policy may have this problem. (or it might I have look further). It reduces national savings which may raise the neutral real interest rate.

    “I strongly disagree with this claim. There’s not a shred of evidence to back this up, not a shred.”

    you’re right of course there isn’t evidence. But is reasonable to argue that that lax monetary policy and low rates encourages people to take on too much debt which was a huge problem with the financial crises? I think so. In the absence of strong financial regulation we can argue that the polices you like would encourage excessive borrowing leading to financial meltdowns.

    “If the zero bound really is a problem (which I doubt) the solution is a 4% inflation target, not fiscal policy (which the GOP would never approve in 100 years).”

    There’s a lot of problems with this one in my eyes. First don’t we already have a target of 2%? When was the last time we hit that target? Why would a target of 4% work when we can’t even hit 2? Fiscal policy obviously has practical concerns mainly with the gop. But remember the backlash the fed faced when it implemented QE? Congress entertained auditing the fed….even when it’s already audited! I argue that unconventional (really just unfamiliar) monetary policies put the Fed’s independence at risk. Do we really want Janet Yellen to take orders from the tea party?

    “That’s fine, but you don’t seem to understand that the burden of proof is on the Keynesians.”

    I disagree. Self correcting mechanisms are built into keynesian theory. Keynes said something like use, decay, and obsolescence can create scarcity of capital. Not exact quote. For the 350 economists who predicted a double dip recession yes they got that wrong. But this doesn’t disprove that austerity hurts in depressed economy. If austerity would have never taken place in 2013 i.e the counterfactual gdp would have grown 3.7% according to Simon Wren Lewis. (I know you argued with him but I think the point still stands. If he’s right the opposition against austerity still wins.

    Lastly….umm I got nothing else.

  46. Gravatar of james elizondo james elizondo
    24. March 2016 at 08:15

    Postkey I’ll check it out thank alot.

  47. Gravatar of Carl Carl
    24. March 2016 at 08:28

    Art Deco:
    You make a good point about the difference between “The Art of the Deal” and “Mein Kampf” but go easy on the accusation that we suburban libertarians don’t realize there are any towns in the world without Starbucks and Nordstroms. Some of us just think that there are other kinds of blow-back besides the kind that happens to the Donald’s hair in a convertible.

  48. Gravatar of Britonomist Britonomist
    24. March 2016 at 08:42

    @Ray

    “it’s not a “multivariate model’ ”

    Uhh no.. it *is* a multivariate, look:

    ” data 1960 to 1990, that a 1 % increase in Fed funds rate shows, over 4-8 quarters, a decrease in sales, decrease in output, decrease in employment, increase in unemployment”

    There are *multiple variables* there, it’s not a simple regression of one variable on another, by your wording at least.

    ““The true value of the effect [referencing various graphs in response to a Fed funds rate increase] lies within the two dotted lines with 60% probability.”

    This is estimating the likeliest magnitude of the effect – THAT IS NOT WHERE CONFIDENCE INTERVALS IN HYPOTHESIS TESTING APPLIES: confidence intervals are for deciding whether the null hypothesis (that the effect of a change in one variable on the other is not *significantly different from zero*) should be rejected.

    Here’s the actual paper actually: http://faculty.wcas.northwestern.edu/~lchrist/research/fofa/flowoffunds.pdf

    It looks like they’re using vector auto-regressions which make things even more complicated (but also makes this paper much more rigorous). If you look at the results tables, an awful lot of the results for their coefficients have significance values (which are presumably P-values) below 0.05, which means they are within 95% confidence or more. In other words, this paper shows that the federal funds rate change affects many things with high statistical significance.

  49. Gravatar of ssumner ssumner
    24. March 2016 at 13:06

    Ryan, Thanks, I’ll add a link to that in a future post.

    Mark, I don’t have a good memory for names, do you have a link?

    Christian, My comment was not directed at you, it was directed at an obnoxious commenter.

    Saturos, Agreed.

    Ray, Austerity does not mean lower government spending—surely even you know that.

    Art, You said:

    “Trump does not ‘demonize’ anyone.”

    Sorry, I was under the impression he was demonizing Mexicans and Muslims. My mistake. And when did I ever say Trump was identical to Hitler? Trump will not become a dictator.

    You said:

    “It doesn’t seem to occur to libertarians living in tranquil suburbs that much of the world is not a serene place (or maybe they fancy the unserene portions are populated by losers getting what they deserve).”

    When I had said:

    “In a world of 7 billion people, the risk of being killed by terrorists is utterly infinitesimal—especially in rich countries.”

    I guess you don’t know what the word “rich” means. And as far as “getting what they deserve”, I’m a utilitarian. It’s stupid right wingers and stupid left wingers who are constantly obsessing about bad guys getting what they deserve, not me.

    James, You said:

    “We could just keep going full blast with monetary policy but as Summers says we’ll have to keep increasing the dosage to maintain the effect.”

    Actually it’s just the opposite. A more aggressive monetary policy leads to faster NGDP growth, allowing lower dosages going forward. It’s the laggards like Japan and the eurozone that have to keep increasing the dosage, as they did too little earlier on.

    Monetary policy is not about interest rates, it’s about NGDP growth.

    You said:

    “But is reasonable to argue that that lax monetary policy and low rates encourages people to take on too much debt which was a huge problem with the financial crises?”

    Why? I don’t see the logic at all. More likely poor banking regulation (moral hazard) led to too much debt.

    You said:

    “First don’t we already have a target of 2%? When was the last time we hit that target? Why would a target of 4% work when we can’t even hit 2?”

    You need to pay more attention to Fed policy. They target the inflation forecast, not the past inflation rate. They think we’ve already hit the target for the forecast, that’s why they are raising rates. Why do you think the Fed is raising rates? Seriously, I’d like to know your take on recent events. Suppose the Fed has a bias that causes them to undershoot inflation by 50 basis points, on average. OK, so with a 4% target we end up with 3.5% instead of 1.5%. So what’s your point?

    You said:

    “Do we really want Janet Yellen to take orders from the tea party?”

    You have it backwards, it’s tight money that leads to low NGDP growth that leads to Fed controversy. There was little Tea Party objection to the Fed in the Bush years when rates were positive and there was no QE, even though policy was far more expansionary.

    If there is no austerity, then there is less monetary stimulus, and 2013 is about the same, plus or minus a little bit.

  50. Gravatar of Christian List Christian List
    24. March 2016 at 16:10

    But is reasonable to argue that that lax monetary policy and low rates encourages people to take on too much debt which was a huge problem with the financial crises? – Why?

    Maybe we can separate the issues here.

    First: Low rates might encourage some people to take on too much debt. (I don’t know if this is true and I don’t actually care too much).

    Second: “Lax” monetary policy leads to the inflation goal (maybe with a bit of overshooting). Inflations leads to money losing its value at a certain rate. Money losing its value leads to people spending their money faster – and without as much careful thought. Or in other words: They might buy things they wouldn’t have bought at total price stability (=0% inflation). Right?

  51. Gravatar of james elizondo james elizondo
    24. March 2016 at 16:19

    Mr. Sumner thx for the response. Again I largely agree with Summers and the theory of secular stagnation. My overall argument is that the policies you advocate may not the best solutions.

    “Why do you think the Fed is raising rates? Seriously, I’d like to know your take on recent events.”

    to get ahead of the curve so the economy wont overheat. Same thing as saying they believe they reached they’re target. I think we both agree that they’re wrong but I get your point.

    “OK, so with a 4% target we end up with 3.5% instead of 1.5%. So what’s your point?”

    Alright perhaps I should pay more attention to fed policy. But I’ll make another point. your big on price stability right? I think you said the fed should just focus on marcoeconomic stability and nothing but. Well targets like 4% inflation target may be in direct conflict with price stability. Why 4%? Why not 6%? Why not 8? Can you see how it may get out of control?

    (About monetary policy creating asset bubbles)“I strongly disagree with this claim. There’s not a shred of evidence to back this up, not a shred.”

    Since QE reduces liquidity and risk premiums cant lax monetary policy lead to assets much above fundamentals?

    With monetary policy constantly attempting to reach the neutral interest rate wont the fed run the risk that it could up buying the entire stock of securities in a given maturity. Bernanke even says that converns about “losing control of the balance sheet was factor of not choosing rate targets over qe.

    We need public infracture. Fiscal policy may provide better options than monetary policy at the zero bound.

  52. Gravatar of Art Deco Art Deco
    25. March 2016 at 09:22

    Sorry, I was under the impression he was demonizing Mexicans and Muslims. My mistake.

    Your point was addressed anteceddently. It does not ‘demonize’ some segment to acknowledge there’s a crime or a security problem emanating from that segment. It only seems that way to people addled by what the disreputable Mr. Sailer calls ‘leapfrogging loyalties’.

    And when did I ever say Trump was identical to Hitler? Trump will not become a dictator.

    I think it was at some point before you decided in this post to make a dog’s breakfast of all your previous posts on this matter.

  53. Gravatar of ssumner ssumner
    25. March 2016 at 15:21

    James, If we need public infrastructure then build it, but don’t pretend it has anything to do with countercyclical policy or secular stagnation. But why have the public sector build infrastructure? I prefer Europe’s privatized airports to our publicly owned airports.

    No, I don’t believe in price stability, I oppose it. I favor a NGDP target, level targeting.

    Art, It was only in your fevered imagination. You and Harding aren’t too good with nuance.

Leave a Reply