Faith in fiscal stimulus

Suppose there was a huge stock market crash that destroyed lots of wealth.  Would it reduce AD?  One would think so, and some might cite the 1929 crash.  But the Fed reduced the monetary base between October 1929 and October 1930 at one of the fastest rates of the 20th century.  So perhaps tight money caused NGDP to fall.

Sometimes I point out that a virtually identical stock market crash in 1987 had no significant impact on AD, not even a tiny slowdown followed the crash.  But it’s hard to convince people, as you’d think a crash would reduce AD.  Some people argue that 1987 was different, because Greenspan eased monetary policy to prevent a fall in AD.  But when I hear that argument I think to myself; “That just proves my point, monetary policy drives NGDP.”

In 1981 we had a rare exogenous fiscal policy shock.  Reagan was elected and took office in the midst of a business cycle expansion.  During 1980:4 and 1981:1 NGDP was soaring at an 18% annual rate.  He quickly introduced an expansionary fiscal policy; big tax cuts and huge military spending increases.  And a few months later NGDP growth plunged sharply.

Once again, people will point to the fact that it wasn’t a fair test–Volcker was simultaneously engaged in squeezing inflation out of the economy.  (Of course Keynesians weren’t using that excuse in 1981, when they predicted the Reagan deficits would lead to high inflation.)  But when I hear that argument I think; “That just proves my point, monetary policy drives NGDP.”

I do understand why fiscal stimulus has a powerful intuitive appeal, and would concede that one can find episodes that seem to support that assumption (like 1940-41.)  But I also think that people may sometimes let this intuitive appeal overcome their better judgment.

This past February I criticized (here and here) a paper by Feyrer and Sacerdote that used regional data to estimate the fiscal policy multiplier.  A number of prominent Keynesians like Brad DeLong, Dean Baker, and Paul Krugman cited this evidence favorably.  But it doesn’t really tell us what they thought it did, as even those models that predict no fiscal multiplier at the national level, would predict a fiscal multiplier at the regional level.  If the federal government pours $10 billion into a nuclear weapons base in North Dakota, then North Dakota’s GDP will rise.  We all get that.  The question is whether America’s GDP will rise.

Now Ambrosini seems to have spotted a similar problem in a paper by two Colombia University economists (Emi Nakamura and Jon Steinsson):

So while this paper claims to show the effect of military expenditures on output, it really shows that military expenditures create jobs and attract capital (they have an IV but as far as I can tell it doesn’t address this issue).

But can we salvage the overall message of the paper? Isn’t creating jobs and attracting capital exactly what we’re trying to do with fiscal stimulus? Yes, but not if it means that labor and capital is coming from other States. GDP is increasing in the States that have more military expenditures but is necessarily decreasing in the States that are losing workers and capital. Because these two effects cancel each other out, the “true” fiscal multiplier is less than the 1.5 estimated by this paper.

NGDP normally grows 15% over three years.  Since mid-2008 it’s up just over 4%.  How much less would the growth have been if there had been no ARRA?  How should I know?  It depends on what the Fed would have done if all the responsibility for stimulus had been put in their hands.  I could easily see the number being less than 4%; the Fed might not have made up for the shortfall.  Or maybe they have a 0.6% floor in core inflation (as in 2010), and we would have ended up in the same place.  Or maybe they would have been very aggressive and done level targeting of the price level in the absence of any fiscal stimulus, and we’d actually be a bit ahead of where we are now.

I’m not saying the fiscal stimulus advocates are wrong; I lean toward the view that the net effect on NGDP is probably slightly positive.  But anything more seems like an act of faith to me.

BTW, I do know that Bernanke talks like fiscal stimulus works.  And I also know that he’s uncomfortable pushing monetary stimulus, because of opposition within and outside of the Fed.  But I also know that he does move when NGDP growth is clearly unacceptable.  So his discomfort at monetary stimulus, and his wish that fiscal stimulus would do more, may not tell us all that much about counterfactuals.  Level targeting is much more powerful than QE, but might well be less politically controversial.


Tags:

 
 
 

13 Responses to “Faith in fiscal stimulus”

  1. Gravatar of Benjamin Cole Benjamin Cole
    8. October 2011 at 19:21

    Another deeply interesting post by Scott Sumner. I think the Market Monetarism Movement (M-Cubed) is gaining strength and maturing, and this post reflects some of that.

    Perhaps Bernanke accepts fiscal policy can be stimulative, yet his statements and advice to Congress are oblique and can be read in a few ways (this Olympian opacity is supposed to convey wisdom and reserve, I think).

    Bernanke could be telling Congress to lower taxes and regs, the famous impediments that are supposedly flummoxing monetary policy (i.e., we get only inflation not growth, when we go to NGDP targeting).

    I think the Market Monetarists need to show that monetary expansion worked before, in an economy arguably with more impediments than today.

    After all, the highest federal marginal tax rates are lower today than in the 1970s, transportation and finance and telecom have been deregged. International trade is much larger, meaning demand in the USA is met globally, thus eliminating many bottlenecks. Labor, capital, services and goods cross into the USA easily, damping down prices. Unionized labor is a sliver of the workforce today. The case that we have more impediments today than in the 1980s or before is very weak.

    It is a great time to try NGDP.

    Why can’t Obama and Bernanke be making this case, instead of Scott Sumner? (Personally, I like Sumner more, but Obama and Bernanke would have more influence).

  2. Gravatar of Benjamin Cole Benjamin Cole
    8. October 2011 at 19:21

    Another deeply interesting post by Scott Sumner. I think the Market Monetarism Movement (M-Cubed) is gaining strength and maturing, and this post reflects some of that.

    Perhaps Bernanke accepts fiscal policy can be stimulative, yet his statements and advice to Congress are oblique and can be read in a few ways (this Olympian opacity is supposed to convey wisdom and reserve, I think).

    Bernanke could be telling Congress to lower taxes and regs, the famous impediments that are supposedly flummoxing monetary policy (i.e., we get only inflation not growth, when we go to NGDP targeting).

    I think the Market Monetarists need to show that monetary expansion worked before, in an economy arguably with more impediments than today.

    After all, the highest federal marginal tax rates are lower today than in the 1970s, transportation and finance and telecom have been deregged. International trade is much larger, meaning demand in the USA is met globally, thus eliminating many bottlenecks. Labor, capital, services and goods cross into the USA easily, damping down prices. Unionized labor is a sliver of the workforce today. The case that we have more impediments today than in the 1980s or before is very weak.

    It is a great time to try NGDP.

    Why can’t Obama and Bernanke be making this case, instead of Scott Sumner? (Personally, I like Sumner more, but Obama and Bernanke would have more influence).

  3. Gravatar of Dan Kervick Dan Kervick
    8. October 2011 at 21:39

    Just one preliminary note: the scope for fiscal action is not exhausted by versions of a one-shot “stimulus”. I believe we should have a permanent program in place for countercyclical employment and income support. We could still create such a program. It would do wonders to restore confidence and security to our frightened, penny-pinching middle class workforce.

    Personally, I think Bernanke wants Congress to authorize a serious helicopter drop, and was basically screaming to Congress, “What are you morons waiting for?” – at least as close to screaming as is allowed by the decorous and reserved norms of discourse among politically independent and neutral central bankers. By a “serious” helicopter drop, I mean one that spends new money directly on real economy consumption and investment purchases – not one that just trades dollars for other financial assets inside the financial industry, and not one that just randomly showers money on people and gives them the option to sock it away in inactive savings. If the government is to inject money into the economy, it needs to go directly into spending and investment. The best way to assure that happens is for the government to spend it or invest it themselves.

    If the spending is money-financed, rather than tax-financed or debt-financed, then given that we are so far below capacity, there is not much of a chance that it will only effect a spending tradeoff from one part of the economy to another. But a helicopter drop of this kind requires the active cooperation of the fiscal authorities since only Congress is authorized to spend. Congress could establish the creation of a special new recovery account for the Treasury to at the Fed. The could order Bernanke to credit that account by a certain number of dollars, and then authorize specific spending programs to pay out that amount of money.

    I know a lot of the market monetarists are inflationists. But the political context is such that most public officials are in no mood to experiment with intentionally boosting inflation while so many people have stagnant incomes and are already feeling the pinch of price rises. Part of Bernanke’s purpose in his recent comments was to emphasize the fact that the coast is clear for money-financed fiscal expansion, since inflation is quite low. The hyperinflationists and the gold bugs have already been proven wrong anyway.

  4. Gravatar of woupiestek woupiestek
    8. October 2011 at 23:43

    Fiscal versus monetary stimulus: doesn’t this one belong in Bryan Caplan’s “Myth of the rational voter”? Doing fiscal stimulus looks better than pressing the central bank to do more monetary stimulus to the electorate.

  5. Gravatar of Mark A. Sadowski Mark A. Sadowski
    9. October 2011 at 00:09

    Scott wrote:
    “If the federal government pours $10 billion into a nuclear weapons base in North Dakota, then North Dakota’s GDP will rise. We all get that. The question is whether America’s GDP will rise.”

    Here’s a thought experiment. What if we were trapped at ZIRP. In other words suppose you have a really stupid Fed (very much like the current one) that is targeting the fed funds rate as its monetary policy. And suppose a small amount of expenditures (say $10 billion) were added to the Federal budget such that this did not change the Fed funds rate at all given the economy’s extremely depressed level. Wouldn’t that elevate the NGDP as well?

    Duh.

  6. Gravatar of Scott Sumner Scott Sumner
    9. October 2011 at 06:12

    Ben, Those are very good questions.

    Dan, If that’s what you want then I’d suggest using a different term than ‘helicopter drop.’ You say you oppose money financed tax cuts–but that’s what a helicopter drop is.

    woupiestek. I’m not sure the public has an opinion on monetary policy. I tend to talk to highly educated people, and most of them don’t even have a clue as to what the Fed does. They think low interest rates are easy money. Also, most people say they oppose big budget deficits. That’s why the right is winning most elections now.

    Mark, I don’t agree with your premise. The Fed is no longer using interest rates as it’s policy instrument, it’s using QE, QE2, Operation Twist, vague promises of future policy intentions, etc.

  7. Gravatar of W Peden W Peden
    9. October 2011 at 06:38

    dtoh,

    I see. It’s just that, usually, ‘banks’ means commercial banks, and very often even excludes entities like investment banks.

  8. Gravatar of Mike Sandifer Mike Sandifer
    9. October 2011 at 07:27

    I don’t understand how some could have thought anyone was arguing against the idea that if you send money to a given district or state, for example, it’s GDP(income) would be higher than otherwise.

    Of course the issue is whether it raises GDP in the country as a whole. I don’t understand Krugman, et al’s point.

  9. Gravatar of Jason Jason
    9. October 2011 at 08:03

    I left a comment awhile ago regarding the Feyrer and Sacerdote paper here:

    http://www.themoneyillusion.com/?p=9213

    When you said: “… as even those models that predict no fiscal multiplier at the national level, would predict a fiscal multiplier at the regional level.”

    I went through and calculated that the national effect of adding up all of the regional effects would have a 1 sigma significant positive effect at the national level. It does not rule out zero or negative effect at the national level at more than 1-sigma, but it is less likely and the paper is only claiming a 1-sigma effect.

  10. Gravatar of Dan Kervick Dan Kervick
    9. October 2011 at 12:19

    You say you oppose money financed tax cuts-but that’s what a helicopter drop is.

    Scott, I never said I opposed money financed tax cuts. But a tax cut is obviously a fiscal operation, even a money-financed one. And depending on how it is financed with new money, the financing might itself require additional legislative action.

    Either Congress has to authorize the Fed to allow Treasury overdrafts on its accounts – or equivalently direct the Fed to credit Treasury accounts by a certain amount – or it has to raise the debt limit so that Treasury can issue new debt, even if the Fed then buys up the debt.

    The latter seems wasteful and inefficient to me. Yes, the Fed returns interest payments to the Treasury. But even at zero interest, the Treasury is still responsible for the principle, and the bond issuance amounts to a wasteful and completely unnecessary transfer of public funds to bond purchasers.

    In any case, what we are talking about is congressional action. What I have opposed is the monetarist obsession with central bank operations. Ben Bernanke obviously cannot announce following some meeting of the governors board that he is doing a tax cut. The Fed has no authority or operational control over tax policy and spending policy.

  11. Gravatar of Floccina Floccina
    10. October 2011 at 05:57

    My problem with understanding how fiscal policy could work is that you are taking just as much money out of the economy by selling t-bills as you are putting in.

    Now, I heard one plausible argument for fiscal stimulus. It said that said if Gov announces say a big 10 year road building project a road construction company might borrow money to buy equipment and hire people and that borrowing would be monetarily expansive. I do not think that at any reasonable level could be enough to jump start the economy nor do I see it as good, maybe in the future there will be less borrowing/debt. That IMHO would be a good thing so if monetary policy can get the economy going with less debt that would be doubly good IMHO. Also the gov must at some point pay back the money that it borrows.

    It seems to me that monetary policy is far better.

  12. Gravatar of anon anon
    10. October 2011 at 06:44

    Floccina, in monetarist terms, fiscal policy aims to stabilize NGDP by adjusting demand for money balances, thus varying velocity. Yes, a transfer from A to B will keep money supply unchanged, but if B has lower demand for holding money, she will increase spending and NGDP will rise. In practice, the monetary authority can react to shocks more easily and with fewer side effects, but if the central bank is pursuing a silly objective (k% money rule, gold standard, targeting interest rates etc.) fiscal policy can be second-best optimal despite its costs. As they say, it takes a heap of Harberger triangles to fill an Okun gap.

  13. Gravatar of Scott Sumner Scott Sumner
    10. October 2011 at 09:31

    Mike, I thought it was pretty obvious too. I guess not.

    Jason, You can’t add up the regional affects to get the national effect—that’s the fallacy of composition. The negative effects of fiscal stimulus (such as tighter money) are not regional.

    Dan, I meant you favor spending not tax cuts, so you’d prefer something other than a helicopter drop.

    Floccina, Yes, it’s better to do stimulus without massive deficits, especially given the demographic time bomb we are facing.

    anon, It’s not clear that fiscal stimulus will do much under a gold standard–perhaps in a closed economy model it would help.

Leave a Reply