« Keynes's General Theory, chapter six | Main | Markets in everything, Xmas edition »
What is the best kind of fiscal policy shock?
Hot off the presses from the NBER, from Andrew Mountford and Harald Uhlig, the evidence is mounting:
We propose and apply a new approach for analyzing the effects of fiscal policy using vector autoregressions. Specifically, we use sign restrictions to identify a government revenue shock as well as a government spending shock, while controlling for a generic business cycle shock and a monetary policy shock. We explicitly allow for the possibility of announcement effects, i.e., that a current fiscal policy shock changes fiscal policy variables in the future, but not at present. We construct the impulse responses to three linear combinations of these fiscal shocks, corresponding to the three scenarios of deficit-spending, deficit-financed tax cuts and a balanced budget spending expansion. We apply the method to US quarterly data from 1955-2000. We find that deficit-financed tax cuts work best among these three scenarios to improve GDP, with a maximal present value multiplier of five dollars of total additional GDP per each dollar of the total cut in government revenue five years after the shock.
The emphasis is mine. I'm not saying you have to believe this paper in all its details (I don't), but over the next year you will continue to hear talk about the wonders of government spending as fiscal policy. The science isn't there. Here are ungated versions of the paper.
Posted by Tyler Cowen on December 16, 2008 at 06:49 AM in Economics | Permalink
Comments
Ugh. Why didn't they include the years before/just after the Great Depression?
Why not include 2001-2004? Oddly enough, I'm just not convinced that a tax cut today
that gives us benefits five years from now is the right idea. Agreed that the science
isn't there, but this is right up the alley of the global warming deniers.
Posted by: kxf_in_dc at Dec 16, 2008 8:06:49 AM
Then why weren't the Bush years more bitchin'?
Posted by: odograph at Dec 16, 2008 8:13:24 AM
I'm not saying you have to believe this paper in all its details (I don't), but over the next year you will continue to hear talk about the wonders of government spending as fiscal policy. The science isn't there.
Hang on a second, Tyler. In the NYT symposium of economists that ran about a week (?) ago, they asked you what the best form of fiscal stimulus would be. If I recall correctly, you didn't say anything about tax cuts; you said they should spend money on airports and to prop up state & local spending. (In contrast, Casey Mulligan, bless his heart, thought wage earners should get $300 billion back, if we're handing out free money in DC.)
So what's going on here? Were you being unscientific in your answer to the NYT, or is there some subtlety I'm missing? Like, "Oh, the spending on airports and state & local wouldn't 'boost' GDP, it would prevent it from collapsing further." ?
Posted by: Bob Murphy at Dec 16, 2008 9:03:35 AM
For those who missed it, here is the NYT sypmosium of economists. They were told they had to spend $500 billion on spending, tax cuts, or both. Now how would they spend it? Tyler's answer:
“I would modernize the few critical bottleneck airports in the U.S., most of all La Guardia and Kennedy. That would not cost a fortune.
“I would try to ensure that state and local governments do not cut funding which they will later restore. To me that is more important, and more conducive to macroeconomic stability, than embarking on new and potentially dubious programs. That will cost most of the money. It’s not that I think state and local governments are always so efficient and wise, but rather this is a very simple and direct way to prevent the economy from being hit by yet another sectoral shock when it is already reeling.
“There are many good ideas, such as electronic medical records, that will not benefit the economy as macroeconomic stimulus. And so they do not make the list as you have phrased the question.
“If there is money left over I would spend it on cutting the Social Security payroll tax for specified groups of lower- to middle-income workers, thus encouraging the resumption of hiring.”
Tyler rarely answers me (or anyone) in these comments, so if someone else sees why the present post is compatible with Tyler's answer in the NYT, please enlighten me. So far as I can tell, Tyler Today is saying Tyler on Dec. 15 is being unscientific.
Posted by: Bob Murphy at Dec 16, 2008 9:11:15 AM
So far as I can tell, Tyler Today is saying Tyler on Dec. 15 is being unscientific.
Oops should have been Tyler on Dec. 12 was being unscientific. (There was an UPDATED DEC 15 on the NYT article that threw me.)
Posted by: Bob Murphy at Dec 16, 2008 9:15:28 AM
If deficit-financed tax cuts would bring economic salvation, we would be in nirvana right now.
The failure of the Bush tax cuts to bring the promised economic boom should put an end to the begging for ever-lower top tax rates, but it won't. Tax cuts as the answer to everything is not based on empirical evidence; it's a matter of self-interest looking for a pseudo-scientific justification.
Posted by: a at Dec 16, 2008 9:24:37 AM
Bob, "stop state and local spending from falling" is not symmetric with "boosts to federal spending will bring recovery." There isn't any contradiction. And another runway at LaGuardia would do wonders for productivity.
Posted by: Tyler Cowen at Dec 16, 2008 9:37:02 AM
I don't think consistence of what Tyler opines is important. Rather what this post says is. Greg Mankiw some days ago posted a summary of economic papers too that seemed to point to tax cuts as the most efficient way to boost GDP.
http://gregmankiw.blogspot.com/2008/12/spending-and-tax-multipliers.html
Posted by: cerasus at Dec 16, 2008 9:38:07 AM
Tyelr is only quoting a portion of what Mountford finds. He really throws Keynes under the bus. As Robert Wenzel points out http://tinyurl.com/5w6zhh, the new love affair with tax cuts as a stimulus by Mankiw, and now Tyler, seems to coincide with Obama's naming of Cristina Roemer to head CEA.
Posted by: Sydney at Dec 16, 2008 10:01:27 AM
Between folksy anecdotes, necromancy and regression analysis as guides to federal policy, it's pretty much a tossup.
Yes, even "vector autoregression...with sign restrictions..."
I will note in passing my ASA and INFORMS memberships.
Posted by: ZBicyclist at Dec 16, 2008 10:31:11 AM
Tyler wrote:
Bob, "stop state and local spending from falling" is not symmetric with "boosts to federal spending will bring recovery." There isn't any contradiction. And another runway at LaGuardia would do wonders for productivity.
Tyler, OK, that's in the ballpark of what I thought you would say--well actually, I'm flattered that you said anything!--but still, I think this sentence in the NYT article was misleading:
Tyler: “There are many good ideas, such as electronic medical records, that will not benefit the economy as macroeconomic stimulus. And so they do not make the list as you have phrased the question."
If I understand your latest clarification, you could have easily added, "Of course, nothing would make the list as you have phrased the question, so don't misconstrue the first things I listed as part of the empty list. Supporting state and local spending are also good ideas that would not benefit the economy as macroeconomic stimulus, but I decided to distinguish those items from electronic medical records because...[I don't know why--Bob.]"
Anyway, just pointing out something that confused me so you will know you have literal people like me reading your future columns. :)
Posted by: Bob Murphy at Dec 16, 2008 10:32:26 AM
Can we put this in the same bin as the recent banning imports to increase our multipliers idea?
Posted by: pytheian at Dec 16, 2008 11:14:53 AM
Haven't we learned that the best course of action is to do the opposite of whatever Mankiw thinks best?
Posted by: at Dec 16, 2008 12:26:14 PM
a at:
The big message I see in these Keynesian multiplier studies is not that tax cuts create a magical world of unicorns and rainbows, but that no matter who is writing the study, tax cuts are better than (a.k.a. less awful than) spending hikes.
Of the papers that Tyler and Mankiw have discussed in recent days, some have found big multipliers and some have found small multipliers. But ALL have found that spending multipliers are smaller than tax multipliers.
Thus, a reasonable conclusion would be that if the Bush tax cuts only boosted GDP a little, then a similarly-sized spending hike would have boosted GDP even less.
So reality-based people should be keeping a tally: In quality-weighted terms, what fraction of papers point toward dY/dG>dY/dT? So far, the pro-G fraction looks to be substantially less than 1/2…..
Posted by: Garett Jones at Dec 16, 2008 12:50:28 PM
From the post: "We find that deficit-financed tax cuts work best among these three scenarios to improve GDP with a maximal present value multiplier of five dollars of total additional GDP per each dollar of the total cut in government revenue five years after the shock."
If this were the case, then shouldn't we cut all taxes to zero and deficit-finance all government?
This suggests a number of difficulties: one is that there is no generic T for taxes. All kinds of things are taxed. And also the cost of deficit financing depends on the amount of the deficit and debt, doesn't it?
Posted by: indiana jim at Dec 16, 2008 12:54:36 PM
The way I understood Mr. Cowen's comments (both in NYT and here) is that we should be skeptical of fiscal policy as a GDP booster. That's about it, because Cowen, either in NYT or this post, does not specifically endorse FP as a good booster at all. But if we HAD to spend half a tril, keep state and local gov'ts from significantly cutting their spending, because if they did we'd be in the hole further. Maybe tax cuts weren't in the NYT symposium because some people are skeptical that they would significantly prop up GDP (and they seem to drive our government into deficits besides...). At least with giving money to lower levels we know there is some negatives being prevented. Any other way of spending it, the outcome is less certain, so at least put it into some close-to-surefire efforts instead of blowing it on spending and tax cuts that may or may not help.
Posted by: Tim at Dec 16, 2008 1:31:01 PM
this guy's a complete joke. He finds one NBER paper, and decides to use it as something that sounds atleast somewhat conclusive. He qualifies it by saying he doesn't believe in the DETAILS of the paper, but decides to use it as evidence against the wonders of fiscal spending.
Posted by: at Dec 16, 2008 3:51:10 PM
What is the difference between the shopping list and letter for Christmas Father and any crass Keynesianism? The multiplier effect, only for good economists...
Posted by: Massimo GIANNINI - M.G. at Dec 16, 2008 4:04:47 PM
Cut government spending, cut tax rates 50% and it will be cheaper than all those bailouts and to much better effect.
HT: Brian Wesbury
Posted by: jorod at Dec 16, 2008 4:34:48 PM
If deficit spending would bring economic salvation, we would be in nirvana right now.
The failure of the government deficit spending to bring the promised economic boom should put an end to
the begging for ever-greater spending, but it won't. Spending as the answer to everything is not
based on empirical evidence; it's a matter of political self-interest (vote purchasing) looking for a
pseudo-scientific justification.
Posted by: Superheater at Dec 16, 2008 4:45:27 PM
Does this paper assume that all output is created equal? I'd rather have another runway at LaGuardia than $50 billion in "output" from some ponzi scheme hedge fund populated via tax giveaways.
Posted by: Brian at Dec 16, 2008 4:54:33 PM
Runway spending. It passes the concrete test.
Posted by: Andrew at Dec 16, 2008 8:41:08 PM
We can end the deflation by having the Fed create new dollars and using them to purchase some of our $10 trillion in existing US debt.
To avoid inflating again, the Fed set the value of the dollar to some amount of a stable commodity like gold and just buy enough US debt to maintain that price. The Fed can always resell that debt later if the dollar's value begins to decline. We should have a target price and a range around that price that triggers the appropriate action if exceeded.
And we should avoid re-inflating. The housing bubble was caused by Fed-created inflation.
This approach would, unfortunately, not require any wasteful new spending or bailing out any doofusses.
Posted by: Alan Brown at Dec 17, 2008 6:31:22 AM
"The science isn't there."
The science isn't ANYWHERE in economics. It's not a science.
If you meant to say, the historically contingent tentative empirical evidence is slightly more conclusive in X direction rather than Y...
Posted by: J at Dec 17, 2008 11:50:20 AM
//And we should avoid re-inflating. The housing bubble was caused by Fed-created inflation.//
//having the Fed create new dollars//
??
Posted by: Tim at Dec 17, 2008 11:53:16 AM
