1. Skip to navigation
  2. Skip to content
  3. Skip to sidebar
Source link: http://archive.mises.org/19161/government-stimulus-and-jobs/

Government Stimulus and Jobs

November 12, 2011 by

This will be no surprise for Mises.org readers, but a recent study published by the Phoenix Center looked at the empirical evidence for the US over the last 50 years and found that government spending/stimulus had zero positive impact on private sector job creation.


Giovanni P November 12, 2011 at 6:10 pm

I didn’t read the article, but they can’t know this stuff by means of a “study”. It’s impossible.

Hemlock Grace November 12, 2011 at 6:33 pm

And it is possible to leave a credible criticism without reading the article?

Artisan November 13, 2011 at 8:33 am

If you cannot tell even by means of a study, why do it?

Daniel November 13, 2011 at 9:13 am

why build one when you can have two at twice the price?

Walt D. November 12, 2011 at 9:27 pm

As with Bastiat’s “Broken Window Fallacy”, there are two aspects:
1) What is seen.
2) What is not seen.
In this example what is seen is that government stimulus does not create jobs.
However, what is not seen, is the consequences of the foregone opportunity resulting from the mis-allocation of the stimulus money.
The Obama Administration makes the same mistake. They point to jobs saved in the public sector by the various stimulus packages. What they overlook is the jobs lost or not created in the private sector as a result of the various stimulus packages.

FDominicus November 13, 2011 at 2:17 am

@Walt .D. I guess they will argue as usual. If we di not have done anything it would be much worse. Just because we spend a trillion we could “save” 100 Jobs not beeing lost. Or something alon that “argumentation line”.

Brian Drake November 13, 2011 at 5:33 am


Best way to handle that (“if we didn’t do anything it would be much worse”) I’ve found:

“How do you know?”

Mark Thornton November 13, 2011 at 7:43 am

There is a link to the entire paper in the press release. It is just an empirical study. Other studies have found that government spending actually decreases private sector job creation which we would expect.

Daniel Kuehn November 13, 2011 at 8:17 am

Do you have some thoughts on this criticism of the study: http://www.econbrowser.com/archives/2011/11/i_killed_some_b.html

What do they do to identify government spending? It’s a highly endogenous variable. This doesn’t seem like a very high quality paper.

Inquisitor November 13, 2011 at 1:34 pm

As flawed as the notion that gov’t spending can “create” jobs that have any relation to market-driven demand.

Daniel Kuehn November 13, 2011 at 4:16 pm

Well why do you think it’s flawed?

John H. Penfold November 13, 2011 at 9:10 am

If we accept the Keynesian model, we must accept the necessary and restrictive conditions that must hold for Keynesian spending to stimulate an economy. These are: (1) Keynesian spending must directly hire unemployed resources. If not, thereby failing to use unemployed workers and excess capacity, it merely shifts resources from an ongoing activity to a politically-motivated one. (2) The spending must not thwart adjustment needed to restore healthy growth to the economy. If not it will prolong recovery. (3) The spending must earn a higher return than its cost. If the first condition is met this is a sure thing, but if not, it is highly problematical. And 4), the country must run a current account surplus. If there is a current account deficit it means the state is spending domestically all of the nation’s savings making the multiplier infinity. Which is clearly nonsense, or, alternatively we count the current account deficit in the multiplier and it becomes negative. It is not an accident that econometricians cannot demonstrate a positive multiplier.

wibee November 13, 2011 at 10:34 am

“The Phoenix Center seeks to demonstrate that consumer welfare is best maximized by promoting free markets, competition, and individual freedom and liberty.”

Of course they would say government spending would have no positive impact on private job creation… :)

John H. Penfold November 13, 2011 at 2:38 pm

The study is serious, if you have a criticism of the methodolgy or data, please make it. Since all of us suffer biases and make mistakes, it is not serious to dismiss the results because of the source, rather one must critique the substance. Since the overwhelming majority of economists remain trained in the Keyensian models, more interesting would be to find a Keyensian model that finds a positive relationship between government spending and jobs growth during periods of slow or negative growth. I don’t know of any, but if they exist I would like to see their methodology.

Daniel Kuehn November 13, 2011 at 4:19 pm

I have a criticism of the methodology – it has absolutely no identification strategy for estimating a value (the fiscal multiplier) that is widely known to be highly endogenous. It never should have been published and it certainly shouldn’t be endorsed here.

How can you give any credence at all to a paper that doesn’t even attempt to address this?

Walt D. November 14, 2011 at 12:40 am

You make a good point. It is not enough to be right, you have to be right for the right reason. Also, if we are looking at the last 50 years, it is hard to believe that military spending did not produce any jobs in the military contractor such as Boeing, Lockheed, Martin Marieta or even boondoggles like NASA. (Whether or not they created more jobs than would have been created had there been lees military spending, or whether military spending in general is a good thing, is another question.
What is clear is that when Bill Clinton cut defense spending, there was a long recession in the LA area where a lot of the companies were based.)

Daniel Kuehn November 14, 2011 at 9:10 am

Well the coefficient on whatever spending variable you use SHOULD give you whether more were created than would have been created without the spending.

That information communicated by the coefficient is not the issue – it’s the bias in that information due to the inability to identify the model.

FDominicus November 14, 2011 at 6:30 am

I read the report (but am not fuly through with it) Just one number has stuck with me. One regulator less should yield 6.2 millions! It would suite me well if that was true. I’m writing against this “regulations” since ages. I really would appreciate a state not taking 50 % of the BIP but just the old christian tenth of all. That would be remarkable in itself and if it should work, the argument that the state simply can’t do it, would be confirmed impressively. I guess it will be the other way, the state expands to well over 2/3 and after the break-down ALL BUT the deledefs would be responsible….

Mark Thornton November 14, 2011 at 8:38 am

The study does not say that no jobs were created from government spending. That would be silly. It suggests that no jobs were added to the private sector by government spending increases during slumps.

Daniel Kuehn November 14, 2011 at 9:11 am

Do you have thoughts on Chinn’s critique, Mark – or thoughts on the absense of an identification strategy?

I’m just trying to get a sense of why you shared this with us. Do you think the methodology is sound? It seems very problematic to me.

Walt D. November 14, 2011 at 12:52 pm

One objective way to see whether jobs are created is to look at FICA (Social Security and Medicare) receipts. These results are difficult to falsify.

Mark Thornton November 14, 2011 at 3:10 pm

I’m not endorsing the methodology of the paper. I posted it because the results back up what economic theory (not various macroeconomic models certainly) suggests should happen. Its nice to have something to reference on such matters. That is all.

Mark Thornton November 14, 2011 at 4:09 pm

Chinn’s critique does not bother me. To say that such models have statistical weaknesses is to say the obvious. The whole point of the paper was to isolate the impact of government spending during slumps on private sector employment so you can’t criticize a guy for trying to establish a result on this issue rather than another issue. The notion that government spending creates government jobs is obvious and not really helpful for the economy. Furthermore, private investment should not have much of a lag, but should see results in the short run because investment creates jobs in the firms that are investing and in firms that they are buying their investments.

Daniel Kuehn November 15, 2011 at 6:32 am

But what about the study’s ability to grapple with the endogeneity of government spending???

Whether we’re really concerned about total or private employment is a secondary issue, it seems to me. The real problem is that there is no identification strategy at all in these regressions, which means it’s not offering convincing evidence for your position.

Daniel Kuehn November 15, 2011 at 6:33 am

Am I crazy here? Is there something wrong with this point?

It seems to me that judged by the standards of empirical studies of fiscal multipliers, this is a very low quality study and you are not doing yourselves any favors by citing it.

Mark Thornton November 15, 2011 at 10:35 am

Good question. Why don’t you go about doing your own study and get it published somewhere.

Pammi November 15, 2011 at 3:00 pm

Daniel Keuhn’s role here is to hammer the same question repeatedly — even though the question has been answered satisfactorily, or shown to be irrelevant — in order to make it appear that his inquisition is indomitable and that therefore his position is superior or correct when in fact the opposite is true. Another fraudulent routine of a defeated state-worshiper.

Comments on this entry are closed.

Previous post:

Next post: