Did It, or Didn't It ...

|Peter Boettke|

Work ... fiscal stimulus that is.  Many studies have been released that claim the fiscal stimulus programs that have been pursued in the US since 2008 have not worked.  I tend to side with those studies because of my priors.  But there are alternative studies, and some claim that the stimulus has indeed worked.  Perhaps the reasoned debate over this should be carried on in a reasonable fashion.  See this Washington Post blog post and the linked to studies that argue that stimulus has in fact worked.

If you are a critic of fiscal stimulus programs, then what is your criticisms of these studies? What has been omitted in the analysis that leads them to conclude it works, while you still insist it didn't work?  Assume for sake of argument you have no ideological priors about "big government", and you simply want to look at the full array of costs and benefits of the programs pursued, what would your argument be?

This by the way is connected, not in the negative way that some may want to imply, is the reason I have used the old phraseology of the "economics of illusion".  And in the context that I have used this phrase it is about short-run and long-run, and in particular about fiscal policy and the shifting of the burden intergenerationally.  Starting in 2008, I stressed the need for analysts to re-read Hayek's Tiger by the Tail and Buchanan and Wagner's Democracy in Deficit.  I do think there are issues associated with crowding out, and even deeper issues associated with who gets the benefits of government stimulus programs (hint: they don't pass the generality norm rule for government policy), but the bigger issue is the long-run costs.

Since 2008, we have engaged in "triage economics", and while the medicine applied may have provided some short-term relief, the long-run costs need to be taken into account.  This I fear, they often are not. If you trade-off long-run growth for short-run relief, then you may very well be engaged in the sort of "fiscal child abuse" (to use Kotlikoff's phrase) and the embracing of the very "juggling tricks" that Adam Smith warned about so long ago as the natural proclivity of governments ancient as well as modern.  Stressing these points in economic analysis is not ideological, it is however recognizing that one must always put the political into political economy if you want to have a full analysis of the problem under examination.

Public choice is not a footnote to serious economic analysis, it must be at the core of any analysis that seeks to understand economic policy decisions and consequences.  Economic policy, as Miltion Friedman stressed in his review of Abba Lerner's Economics of Control, does not take place in a vacuum, and it is not designed by Edwin Cannan's Inca God who is omniscient, benevolent and omnipotent.  Public policy emerges out of the political machinations of real life and blood human beings (fallible yet capable as they are) who interact with each other in real institutional environments (historically contingent as they are).  This process of political decision making has costs and benefits that must be accounted for if an analysis of something like fiscal stimulus is going to strive to be complete.

Anyway, I think it is important for a reasoned debate in economics that individuals who are engaged in this debate serious be able to do what is done in the blog post linked above.  Understand the studies, not just their conclusion, but the way the authors reached their conclusion and what might be potential problems with their reasoning.

Comments

[Daniel Kuehn]

I have some thoughts on the (statistical) biases of particular multiplier studies, but on your own point about worrying about long-term consequences I definitely agree with you (with the concern - not necessarily with your answer to that concern). No model or policy proposal that assumes omniscience or omnipotence is worth much of anything and no discussion that focuses only on the short term is sufficient.

I think this is why you see so many people looking at the ultimate cause of the deficits right now (is it spending increases or revenue shortfalls from the recession), the DeLong-Summers work on what conditions allow the stimulus to "pay for itself" in a deep recession, and the work differentiating contributions to the long-term deficit from fiscal policy vs. contributions to the long-term deficit from entitlement problems (which are completely unrelated to the response to the crisis). That along with the fact that every economist and budget watcher in the D.C. metropolitan area is obsessed with getting the deficit down reassures me. In fact sometimes I think they're too obsessed with that.

[Richard Ebeling]

It seems to me that reading over the summaries of these various studies on the impact of the "stimulus" spending the following are issue can may be raised:

All of them focus on the "macro" impact on "employment as a whole." That is, irrelevant to the analysis, it appears, is the issue of what types of jobs were created, in which sectors of the economy, and whether or not such job creation would be sustainable in the longer-run once the stimulus and its residual effects were gone.

This is the importance of Hayek's "Tiger by the Tail." The very essence of "Austrian" monetary policy has been on the "micro" aspects of macroeconomic policies. The "older" or pre-Keynesian economists never disputed that government spending money ("created" or borrowed "idle deposits") may very well put people to work. The issue has been whether these are employments that in the longer-run represent sustainable jobs in the sense that they are involved in producing and supplying goods and services that reflect the patterns of market-based consumer and private investment demand.

If not, they will be unsustainable as the form and degree of the government spending is ended or reduced at some point in the future.

Another way of saying this, is that it is not sufficient that if the economy is operating "inside" the production possibility frontier that government spending moves the economy's resource use to "some point" (or "any point") along the frontier.

Whether the "full employment" use of labor and resources is sustainable will depend up whether or not the point along the production possibility frontier influenced by the government "stimulus" is that point reflecting the underlying "real" savings and spending preferences of the income-earning consuming and investing public in a market-based and generated full employment.

If not, then the economy will have to "re-balance," again, with an inescapable period in which the economy temporarily operates "inside" the production possibility frontier, as it is doing so.

As Pete correctly points out, Public Choice theory is essential in this analysis, as well, because it highlights that the direction of the government spending will be motivated and guided by what appears to be politically useful or desirable by those in political office. Which, of course, may have little or nothing to do with market-based employment and resources use. And, which, again, means that politically generated employments have little longer-run sustainability once the government spending stops or slows down.

This is separate from the more usual macroeconomic question as to whether or not fiscal policy without accompanying "monetary accommodation" has any ability to have a noticeable employment impact.

(An excellent piece on this latter point is by Fritz Machlup, 'The Effects of Fiscal Policy and the Choice of Definitions,'in, Harry Greenfield, et. al., eds, "Theory for Economic Efficiency: Essays in Honor of Abba P. Lerner" [Cambridge, MA: The MIT Press, 1979], pp. 92-109. It appears to have been ignored, including by Austrians, because it seems not to have been reprinted in any of the collections of his essays.)

Richard Ebeling

[Daniel Kuehn]

Some big "it didn't work" studies left off the list are of course Barro and Redlick as well as Ramey.

I think the problem with these estimates is related to your concerns here about acting like policy is made in a vacuum (less related to the long-term issue). They produce what they allege to be exogenous variation in spending by using war spending. Lot's of econometricians are happy to sign off on that, but like you said policy isn't made in a vacuum. Military spending is perhaps the least random kind of spending that governments do - in fact it is highly correlated with perceived security threats or the potential for war. Perceived security threats and the potential for war are going to substantially depress output relative to its potential, so while there may be a stimulative effect during a depression (say, in 1940 and 1941 as many economists have found), any multiplier you're going to calculate from that is going to be biased downwards because of the relationship between security concerns and output.

And what do you know: these studies (Barro & Redlick, and Ramey) produce the lowest multipliers out there.

Understanding these problems isn't just an intellectual exercise. It helps you account for the variability in the estimates.

[Daniel Kuehn]

I'm curious about peoples' thoughts on that take on Barro/Redlick and Ramey, as I'm trying to write something up on it now.

[Peter Lewin]

Richard, how can we get a copy of the Machlup article? Is there an electronic scan anywhere that you know about?

[Daniel Kuehn]

Very coincidental... Economics in the Long Run: New Deal Theorists and Their Legacies just arrived in the mail!

[Richard Ebeling]

Peter Lewin asked about a scanned version of Fritz Machlup's article that I mentioned in my comment, above.

I do have a scanned PDF of the piece. If anyone wants a copy, please email me, and I'll send it along.

Richard Ebeling

[casquette new era]

Votre blog est très bonne, je l'aime! Merci pour votre partage! Votre blog est vraiment aide dans ma recherche et j'aime vraiment it.I arrêt biseaute juste la lecture de ce. Son si froid, si plein d'informations que je ne savais pas.

[Yuka]

OK, I understand what you're gtniteg at Steven.Harry Browne, again granting the premise that he's an authority, thinks that "working to acquaint people with our [sic]* message ... is a lot more productive."Relying on empirical evidence and not intuition for proof, I again ask how this can be measured. I contend that not only can it not be measured, but that all evidence points to libertarian efforts to educate being futile, and perhaps utterly counter-productive** in the net.There is a meta-productivity that I will agree exists -- proselytizing libertarianism certainly gives libertarians something to do, and to the extent that they profit from it, it is as productive as any form of entertainment, such as pro football, filmmaking, dance, etc.However, if one is measuring political productivity, it is as time-wasting an enterprise as can possibly be conceived.* I do not permit my inclusion in collective pronouns such as "we" or "our" without my express written consent, therefore Browne's use of the word "our" has no meaning to me personally.** I admit I have no empirical evidence for this claim, anymore than a proponent of libertarianism has for the claim that it does net good. However, my intuition based on world events leads me in this direction.