Study: Obama Stimulus Destroyed Million Private Sector Jobs
Ohio State researchers: ARRA created/saved approximately 450 thousand state and local government jobs and destroyed/forestalled roughly one million private sector jobs.
Economists Timothy Conley and Bill Dupor are out with a study (PDF) on the impact of the $800 billion stimulus package passed in the early months of the Obama Administration:
Our benchmark results suggest that the ARRA created/saved approximately 450 thousand state and local government jobs and destroyed/forestalled roughly one million private sector jobs. State and local government jobs were saved because ARRA funds were largely used to offset state revenue shortfalls and Medicaid increases rather than boost private sector employment. The majority of destroyed/forestalled jobs were in growth industries including health, education, professional and business services. This suggests the possibility that, in absence of the ARRA, many government workers (on average relatively well-educated) would have found private-sector employment had their jobs not been saved. Searching across alternative model specifications, the best-case scenario for an actual ARRA has the Act creating/saving a net 659 thousand jobs, mainly in government.
Moreover, the study suggests that most of the funding from ARRA was used by states to avert their own budget crises rather than fund new projects:
A large fraction of the Federal ARRA dollars was channeled through and controlled by state and local governments.4 This is important for two reasons. First, it opens the possibility that states might receive direct ARRA allocations due in part to exogenous capacities to channel or attract Federal funding. Approximately two-thirds of all ARRA spending is formulary, of which there is substantial exogenous state-level variation in formula `parameters.’
Second, channeling through states creates an environment where Federal dollars might be used to replace state and local spending. The Act legislated ARRA funds to go to state and local governments for specific programs, such as schools in high poverty neighborhoods and highway construction. Importantly, as depicted in Table 1, states and local governments were already spending significant amounts of their own dollars on many of these programs before the ARRA.Often state spending was substantially higher than nominally targeted ARRA funding.
Upon acquisition of ARRA funds for a specific purpose, a state or local government could cut its own expenditure on that purpose. As a result, these governments could treat the ARRA dollars as general revenue, i.e. the dollars were effectively fungible.
Economists such as Paul Krugman have suggested that the problem with the stimulus is that it wasn’t big enough (even though there’s no record of Krugman criticizing the size of the package while it was being debated in Congress) but this suggest that the entire plan was fundamentally flawed in the manner that funds were allocated. It would have been better, it seems it would have been better to just give the money directly to the people. Or, you know, not spend it at all.
Finally, this chart shows the extent to which the manufacturing and goods-producing failed to benefit at all from the stimulus:
The authors defined “HELP Services” as health and education, leisure and hospitality and business and professional services and its no surprise they’ve done well, as has the government sector. But manufacturing? The lifeblood of the economy? It didn’t benefit at all and it is, as the unemployment figures can attest, still largely stagnant.
The stimulus was, in other words, a monumental failure.
H/T Power Line
“This suggests the possibility…” Well, yes, and I’ve seen studies that suggest the possibility that if Napoleon had had tanks at Waterloo, he would have won.
Not, the first study to show that the stimulus money went to states and localities not based upon high rates of unemployment, but based upon historic ability to attract federal money.
Actually, the Obama stimulus destroyed the entire world. This is just a hellish afterlife where people like Palin & Trump are taken seriously…
Here is a link to the study from last year. Key graph:
IOW, the stimulus didn’t go where it was needed, often going to where it was not needed. It went where Congressional pork producers (from both parties) were located.
If two government employed economists say government funding of jobs actually destroys jobs, we should believe them!
Anjin-san, we can always count on you to deflect the truth if it shows Obama may have screwed up.
Sam, can you show us a study to “suggest” the stimulus did what it was intended to do?
Only the Keynesian Herd truly believes in the miracles wrought by central planners.
Good post, Doug.
“Sam, can you show us a study to “suggest” the stimulus did what it was intended to do?”
I dunno. Haven’t really looked. I was only responding to the mealy-mouthedness of the claim — a claim, btw, that two minutes thought would show is goofy.
Frankly, I’m skeptical of this study. I think that John Taylor’ has it right: the ARRA didn’t do much one way or another.
However, I think it bears mentioning that both Taylor’s work and this study are empirical studies. To the best of my knowledge every report finding that the ARRA created jobs relies on models rather than data.
Many economists are government employed; even the ones at state universities. Then there are the ones who consider their economic literacy to include political insights.
Can’t blame that on bush, but Comrade Obama will find a way.
So, the stimulus saved/created ~660k jobs, largely in gov’t, especially state & local gov’t. And those state & local gov’ts used the money largely to forestall their own budget crises (which would have resulted in mass layoffs of gov’t workers). That much I can accept, but it sounds like they then go on to imply that if the stimulus had instead been thrown at private businesses,
a) all of those laid off workers would have successfully (and immediately) moved to the corporate world, and
b) even more jobs than laid off workers would have been created.
While I don’t doubt the stimulus money could have been better-directed, I don’t see the back-end for either of those conclusions. Am I missing something?
AFAIK, John Taylor hasn’t looked at the stimulus in light of state-specific factors like this study. I think it’s key propositions are sound: money went to states and localities with no correlation to the degree of unemployment. Also, the states took the fed money and reduced spending elsewhere (while in the process created large bureaucratic expenses to unsuccessfully protect against that very outcome).
Now, it may be that a state specific focus is limited — this study suggests the possibility of money disproportionately going to Georgia instead of Florida, might still result in Georgians buying vacation homes in Florida, a desirable outcome for Florida that’s not modelled. But I think it’s preferable to the other extreme that discounts the reality that it was state governments that received the money.
Like many professions, the good economists are rich and the bad ones work for the government.
legion, “That much I can accept, but it sounds like they then go on to imply that if the stimulus had instead been thrown at private businesses, a) all of those laid off workers would have successfully (and immediately) moved to the corporate world, and b) even more jobs than laid off workers would have been created.”
I don’t think that’s implied by the study. The authors indicate there may be shortcomings in their analysis for more study, but I think they simply conclude that the stimulus was ineffective in creating jobs outside of government. I think the study could even be used to argue that the money spent by the federal government directly was more effective, because it was more likely used to purchase goods and services from the economy. It’s also possible that the size of the national economy has gone beyond the ability of the federal government to do a keynesian stimulus of any kind.
“But I think it’s preferable to the other extreme that discounts the reality that it was state governments that received the money.”
Only two-thirds, right? The other one-third was tax cuts. (Just in the interests of accuracy).
I’m glad some of these commenters are actually looking at the study and considering it, as opposed to the idiots who accept this brand new study as either fact or fiction depending on their own particular viewpoint.
What tax cuts?
@ sam, true, there were tax cuts, but my point was that I distrust the studies/analysis that treat one dollar sent to South Dakota the same as one dollar sent to Detroit. Keynesian theory would assume the dollar would have not nearly the same value.
Isn’t that one of the problems with economics? The subject matter is very complex, and its very hard to experiment in macro-economics (we’ll have a control country and a country in which we try solution A and then another, close to identical country in which we try B etc) – what’s left seems to be studies in which the economists themselves disagree strongly with each other. its not something you find often in less complex fields like say civil engineering: most engineers will come to an agreement over whether a particular design will stand or not.
Its why its hard to take these studies seriously – they might be right, they might be wrong, you can’t tell just by listening to the experts, as they disagree. Which suggests we don’t know enough about economics yet to make such studies meaningful. Its still worth studying economics, so we get to the point where it is predictive, but at this point as a non-economist its hard to put much credence in its predictions.
Pete, the ARRA included $288 billion in federal tax cuts, and $499 billion in federal spending (from the subject report)
However, a crowding out story cannot be a full explanation as it is not consistent with our estimates that the ARRA destroyed/forestalled more jobs in the HELP services sector than it created/saved in the government. We leave completing an explanation for our estimates of ARRA destruction/forestalling of so many jobs as a question for future research.”
The study finds a neagitve correlation between job growth in one sector and the difference between stimulus spending and other budget decreases by state, and conclude that the stimulus destroyed jobs… They have no explaination of mechanism. I think they have committed the fallacy of equating correlation with causation. I don’t find it very compelling.
The paper is (tipping my hand) dismantled here, Did the stimulus really destroy a million private-sector jobs? (H/T, Tyler Cowen).
Ok, you know, read the cite and make up your own minds (unless you are sick unto death with dueling economists).
I agree, and I think that’s the problem here – it’s quite a step to go from “stimulus created jobs in gov’t” to “stimulus destroyed jobs”, so I think my issue is more with the way Doug is framing/interpreting the study than the study itself…
We know you don’t like Krugman, but if you’re going to insult him, try and have facts on your side:
Krugman was saying it was too small well before it ever passed.
Here is Krugman on January 6 of 2009:
You need to read some more Krugman; this is at least the second time you made passing reference to him not making criticisms he clearly made all along.
Pete, two seconds with google turned up this:
Most of the didn’t apply to me, but I do recall the “making work pay” tax credit.
And Doug — Krugman worried that the stimulus package was too small in Jan 2009, back before Obama was even sworn in.
Technically, that’s not “while it was being debated in Congress” but I’m pretty sure he didn’t shut up after that.
i have to say this doesn’t even make sense in the context of the numbers we know. Saved 600K jobs, mostly in government? we know that the public sector has shed 400K jobs. so how does this jive? the market was shedding 600k jobs a month in dec ’08/jan ’09 – then leveled off pretty much in sync with the stimulus. we know that since then we have added 1.7m private sector jobs – so where are the 1 million jobs lost because of the stimulus? again – it doesn’t seem to jive with the context.
my prediction – i suspect this paper will be thouroughly debunked. however, in the meantime, the right wing echo chamber will have shouted it from the treetops so much that it will have become fact to the right-wing…at which point it will have joined the list of other things believed without proof – like tax cuts pay for themselves, iraq had wmd, evolution is a hoax, 100% of peer-reviewed climate change science is faked, etc. etc. etc. etc.
legion, part of the issue with job destruction is time frame. The Wilson study cited by sam shows 2.3 million jobs created through February of 2010, but only a net of 800,000 jobs through October of 2010. That suggests crowding out is beginning to occur by the end of 2010.
Krugman also said the real problem was finding shovel ready projects to spend on.
Oddly, not one commenter was able to show the claim false. That is because the facts are the stimulus failed. Period. That is all. Government has never been good at creating jobs. Let us just wait and see if this President can get reelected with a 9% unemployment rate and a national debt he grew faster than any President in history.
Crowding out is kind of a promise by economists that something is happening that none of us can see. They don’t find it in the wild. Instead they descend to the basement and throw the bones. It is all very Terry Pratchett.
Fwiw, I find it very hard to believe that in this large consumption downturn anything real was really crowded.
actually wiley, sam linked to this which pretty much destroys the credibility of the paper.
you reallly do give stoners a bad name.
Put another way, crowding calculations are one of those things that require a stable framework. To judge them in the tides of crash and recovery emotion seems … the sort of classic macro lunacy.
Good link, hey norm. That’s a more basic error than what I guessed, based on crowding comments above.
jp, there are fewer highway, street and bridge jobs available despite the stimulas. I’m pretty sure that’s at least partly because mergers and buyouts by the companies getting stimulus jobs. The linked report says we’ll have to wait and see:
Notice the linked paper compares its findings with the Wilson paper; the advantage of being later.
That is a reach PD, in more ways that one, but to attack it directly:
1) did new technology suddenly allow fewer to build more?
2) or are you saying we lost middle management?
3) wouldn’t that be reducing waste, fraud, and abuse?
(I can’t imagine that you are suggesting that the stimulous was sized to replace and extend all RE boom construction.)
Was there an alternative that would have benefited manufacturing? Doing nothing, it seems to me, would have made things worse. More unemployed government workers surely would have meant less consumption. Distribute the funds directly to the people? Some of that was done via temporary tax cuts but the concern I heard was that most people would likely save the money if given a lump sum as they were understandably worried about the future.
Doug- You should update this post. As Sam noted, the results in this study are not statistically significant. As a lawyer, you may not read scientific studies very much, but this paper proves nothing. Mankiw is a smart guy, but he does have his own political biases. I think his citing this paper is proof that an econ professor is not above partisan politics. Again, here is the link. You should consider updating this or writing another post correcting it.
jp, I believe part of the answer is that highway construction firms realized that long-term growth in their sector was limited and used the stimulus as an opportunity to merge with maximum profitability. I think that’s a part of the inefficiencies of the stimulus, which is supposed to increase government spending over the norm, when it ends up being deployed into what are seen as shrinking sectors (construction, autos, etc.) The government ends up paying money to facilitate job shrinkage.
It’s the problem with Obama and Bush’s temporary tax cuts, they didn’t promote long-term investment, because they were short term gimmicks.
Is the Keynesian position premised on the view that crowding out is a fiction? I always thought it was that crowding out is minimalized, though on the basis of conditions that seem increasingly idealized in today’s economy (or at least U.S. federalism). The Keynesian prescrpition that government spending be reduced during growth seems to suggest he believed in crowding out.
PD, I just don’t get the stimulus causality. Surely, sans stimulus these firms would have contracted more. The only alternative to mergers in contraction was downsizing or bankruptcy.
Hard to see how adding money and projects makes the job equation worse.
“paying money to facilitate shrinkage” would only make sense if someone was paid (a la farm subsidies) not to build something.
(I’m not a Keynesian, I just don’t believe in immaculate spending.)
Perhaps it is just as fair to start speaking of “jobs deferred or destroyed” as it is of “jobs created or saved.” Each is equally difficult to quantify accurately and just as prone to initiate partisan attacks.
On another note is the biggest problem of Keynesian stimulus is it utterly ignores the expense side of the balance sheet and focuses solely on what it projects as the benefits. I like to call the ignored factors the divisor effect. TANSTAAFL.
Don’t you have some more daughters to threaten?
this type of study has the same problems with it that papers saying x amount of jobs where created or saved – none of them are actually measurments. Just some statistical simulation based on past experience – in short these mean really nothing. We will not emperically know what this bill actually did.
The one paper I’ve seen that actually looked like it meant something was one that actually compared across states with different levels of stimulus spending so you could acually try to isolate a measured effect. That paper concluded total job increases where posative but very minamal.
The observationt hat states mostly just used it to pay off their own budget whoels ratehr than create new projects though is valid. Others have noted this too.
jp, I didn’t think you were a Keynesian either, which makes your defense of the stimulus odd. I supported a Keynesian stimulus, though it’s implementation fails me, much as I supported the Iraq war.
As i understand Keynesian theory, shocks to the system create excess capacity and with low interest rates, the government can borrow to make up for the capacity and achieve efficient, humane outcomes. The money borrowed can be paid paid before it constitutes a drag on the economy. The first problem is that as this study and the other study I linked to indicate, the stimulus money, no doubt due to our political system and not ill will, didn’t necessarily go to where excess capacity exists. It also was confounded by the reality that the excess capacity existed in economic sectors (construction, auto, finance) that are due to shrink. Our economy may be too advanced for the theory. The third problem is that we aren’t going to pay the borrowing back before interest rates begin to hike. Private sector foresees this and starts hoarding nuts for the winter.
I confess here that I was in error to be ambivalent about extending the Bush tax breaks. I thought we were due for real tax reform, along the line of the deficit commission, and a partisan argument defined around pro and anti- Bush tax breaks was irrelevant. Our leadership fails us, and we’ve failed to elect leaders.
I see myself taking a middle line on the stimulus. I see it as weakly positive, but I think the best attack is on costs, not benefits.
Government spends money on goods and services, someone gets paid, goes shopping. Hard to avoid unless “spending” is “deleting” money from government accounts, with no recipient.
As Josef Stalin might have said:
One lost private sector job is a tragedy. A million? Merely a statistic.
“The first problem is that as this study and the other study I linked to indicate, the stimulus money, no doubt due to our political system and not ill will, didn’t necessarily go to where excess capacity exists.”
What you should understand is that this study shows nothing. Its results are not statistically significant. As pure theory, I think you get a lot of it right. If you look at the reports put out by the big forecasters, they show the stimulus creating jobs and increasing GDP in the short run, but they also follow it out and show that it has costs in the long run. The hope is that by then a stronger economy can absorb those costs and also pay off the debt incurred by a Democratic stimulus, spending, or a GOP stimulus, tax cuts. A policy that indulges only in deficit spending is not truly Keynesian IMHO.
Ill will? No…. mere favoritism. Areas that tend to vote Democrat.
steve, if you want to be insulting, please be more direct about it.
Truth is Pete, you were ignorant of the tax cuts in the stimulus, which were a huge component of the package. If you are that clueless, you should consider sitting the thread out. Your “reasoning” seems to be if it makes Obama look bad, it must be true. Not impressed.
I think that was George W.
A mere pittance, comparatively. But of course the small tax cut and not the uge government spending is the issue, to you.
It must be a bear of a job, defending your fellow leftists once their plans are put into action, and as invariably happens, disaster results… as in this case. I guess the measure of the true believer. I’m reminded of Baghdad Bob…. “There are no enemy tanks.”
Just under 40% of the stimulus is a mere pittance?
Also it’s really hard to ignore that the disaster that was the ’08 crash began under a Republican president, and was arrested by a Democrat.
It must be fun living in your fact-free world.
Oh, and to Doug’s subhead:
Conley is from Canada. So there you go….
Key take away from this before this drops out of sight…the paper fails to produce any statistically signifigant results…so to put it in terms that Doug used in the original post…this study was “fundamentally flawed”, not the Stimulus Package.
I’m always distrustful of claims regarding counterfactuals (the stimulus saved/created X jobs! The stimulus actually cost X jobs! Etc). Though one can postulate what would have happened w/o a stimulus package, with a larger one, with one containing a higher or lower % of tax cuts… ultimately we don’t know.
I find the basic idea of countercyclical government spending logical and, thus, I support it. The obvious problems are: a) we don’t do countercyclical spending, since we tend to run deficits even during booms and though the Feds might ramp up stimulus spending in recessions the state and municipal governments do the opposite and negate the Federal stimulus; and b) it’s fairly easy to say “let’s spend $X to stimulate the economy” but it’s harder to spend that money well. Also, when individuals have as much debt as Americans do now, stimulus money (especially tax cuts) is likely to be saved or used to pay down debt (good things, certainly, but not economically stimulative in the short term).
If, by disaster you mean the recovering economy we are now experiencing,well I guess it is a disaster. From your perspective. All you have left is praying for 12% unemployment and $6.00 gas.
Kind of sad how when they were handing out brainpower, you got a mere pittance…
Here’s a clue for you bit, when you get to the part that says “This suggests the possibility…” It’s time to move on to something serious…
I don’t know which is more frightening; That the White house is selling that we’re in a recovery, or that you believe it.
And you think everything’s just grand.? Oh, wait. “Think” doesn’t apply to you, apparently.
Good summary, Rob in CT.
Bit… never said all was grand. Is your argument so weak that you have to make things up and attribute them to me? apparently so. the economy is recovering from the disaster Obama inherited. it is a painful process, the economy nearly collapsed under Bush. Stop reading “the American thinker”, and pick up a copy of the wall st. Journal.
BTW bit, if you go back and look at posts from before the real estate bubble collapse, you remember, when conservatives were crowing about the great Bush econ, no? you will find a lot of of posts from me stating we were in a credit bubble.
Given your track record for being wildly, spectacurly wrong about so many things, you might want to stick to your own blog, where the bar rests comfortably upon the ground. After all, you are the guy who flatly stated “Obama can’t win”.
So the take away is either the stimulus destroyed 550k jobs or maybe it saved 660k jobs. Or somewhere, anywhere, in between.
Only in economics would this be considered a result to report out. Instead of a just cause to fire someone for gross incompetence.
It should probably matter that the authors of said study chose to completely ignore fully one third of the stimulus program. In their model (which has marvelously broad confidence intervals) they didn’t include the $280M in tax cuts…all they considered was the spending portion. It does seem odd that an analysis of the stimulus would exclude such a huge portion of the stimulus.
This is clearly Bush’s fault. Or the Koch brothers.