How ineffective was the stimulus really?

Kevin Drum has a very thoughtful response to the new Jones and Rothschild papers.  I am still more skeptical about the job-creating magic of stimulus, however, and here are a few reasons why:

1. Labor market polarization.  This is a very popular idea among the Progressive Left and rightly so; it seems to be true and increasing in importance.  Yet it gets dropped like a hot potato when discussions of stimulus come up.  A simple interpretation of the data, consistent with labor market polarization, is that we have a larger sum of money chasing the same set of well-qualified, easily-employable workers.  Polarization also means not so much substitutability and there is plenty of evidence in the Jones-Rothschild paper of employers finding labor markets — for what they want — somewhat tight.

2. The Jones-Rothschild paper has an estimate that only 42 percent of the job offers went to the unemployed.  A lot of the money also was spent on capital, land, raw materials, and other factors of production.  I’ve never seen good estimates here, but labor’s share is about seventy percent of gdp, actually a bit shy of that.  Let’s say seventy percent of the stimulus gets spent on labor at all, and only forty-two percent of that gets spent on unemployed labor.  (It’s actually worse than that because it is 42 percent of the job offers and may well be less than 42 percent of the revenue, most likely so if you think of the unemployed as bringing lower wage offers.)  That’s less than thirty percent of the initial expenditure being spent on unemployed labor and that is before any other problems with the expenditures kick in.  It’s hard for me to see that as a triumph of the program (NB: we are only talking about one part of ARRA here); would direct government employment have overhead costs that high?  How about monetary stimulus?  What’s the new calculation for cost per job saved per year?

3. Cutting nominal wages of workers hurts their morale and firing people hurts the morale of everyone left behind.  Employers weigh these morale costs carefully when making personnel decisions.  To get to the matter in question, when times are tight employers are often quite relieved when workers leave the firm voluntarily.  It eases their cash flow, prevents a firing, and everyone is happy, sort of.  Bad times are precisely when replacements of these workers do not happen.  (Another version of that argument: If Keynesians are right about labor hoarding, job shifters don’t get replaced very often.)  So the claim that an ARRA hire of an already-employed worker led to a replacement for that worker at the original firm is not so strong.  This happened during down times and very often replacement is postponed, perhaps indefinitely.  The Keynesian view, after all, stresses how AD problems hit virtually the entire economy.

4. Very often when the replacement does happen, the replacement is drawn from the pool of workers who are doing well.  Some of those workers will be unemployed.  But they are the unemployed who least need the help.  Their average search time goes down, and that is somewhat of a social gain, but it is hardly the goal of a fiscal stimulus program.  We’ve failed very badly at reemploying the hard-core unemployed and that is borne out by other numbers.  So of the forty-two percent of offers going to the unemployed, how many were going to the “don’t so much need the help but were searching” unemployed?  That will make the calculation look uglier yet.

Labor market polarization, labor market polarization, labor polarization.  Market-oriented economists don’t like to stress this theme, but it’s true and it’s a big reason why the stimulus didn’t work better.  It’s not an idea we should suddenly leave behind.


The data-free relentless paid pursuit of a completely failed ideological narrative gets so tedious. More posts about buildings and food, please. And music.

Plenty of data in the Jones-Rothschild paper, oodles of new data and results.

Am still waiting for data on the characteristics of firms which claimed they did not add or save jobs, and whether they were complementary suppliers to firms that did add or save jobs in the same project which employed both firms. Data disclosed for a different purpose in one report showed, for example, that engineering firms did not add/save employees for an ARRA project, probably as one would expect, but did not disclose whether the construction company that took the blueprints for that project did add/'save jobs. You could certainly see that in a project where several firms were employed to serve a small or large role that some would and others would not add/save employees, and unless you looked at the project as a hole, you would be either be feeling the tail or the trunk of the elephant.

Waiting for an oodle.

"unless you looked at the project as a hole..."

Wait, are we back to discussing "shovel-ready" again? Because a "hole" project would definitely qualify.

George, Ya got me. Hat tip to you.

Unless those firms don't consume any of the money spent on stimulus, this is still going to substantially change your estimates of the effectiveness of stimulus.

For some values of "data."

Come on. Twenty-minute interviews with 85 not large, unidentified, firms in unidentified geographical areas in unidentified sectors? A nice little matrix of high unemployment, etc, but with the Mid-Atlantic area thrown in "for convenience" and no expalnation of where it fell in the matrix. Were the numbers reported by the interviewees even checked against actual company records?

This paper is useless.

It is worse than...what? Economists...alternatives...get some.

I imagine that you and Bill are also among those trashing recently appointed Krueger for his minimum wage studies? After all, those were based on survey data, and were contradicted by studies that checked against the actual company payroll data?

What papers contradicted Card and Krueger?

Let's make the same calculation for tax cuts. How much of them lead to increased investments vs. buying treasuries?

How Leading Are The Headlines, Really?

I love how people give Tyler crap in the comments when he is probably infinitely fairer than the people he regularly "argues" with in the blogosphere, i.e., DeLong, Krugman, Yglesias, etc.
Read the title of Yglesias' post on this issue.

Is it giving him crap, really?


At a trial, what you would do with that headline is stand up and say: Inadmissable, Assumes a Fact Not In Evidence.

This is Tyler's site and he can headline all he wants. It's free speach. Just as it is to point out how you may be led by the head, line that is.

Trials are bullshit. You know that Bill.

I was thinking about the trial of the lady who gave the kid hot sauce. Sentenced by a judge who would have you pepper sprayed in seconds if you questioned his authoritah.

Is that what you think of a right listed in the constitution?

"in" is stricken from the record.

Now sit down counselor.

Feel better now Bill?

Well, Andrew, you do not like the right to a trial, and you do not like the right of free speech apparently either.

Shall we pick one more constitutional right you do not like?

This narrative is a bit silly when it ignores what happens to the positions job switchers leave behind. Is the idea that their prior position goes unfilled? Seems unlikely to be true 100% of the time. If so many positions were really unneeded, is the argument that those employees who left staved off eventual layoffs?

Not sure how the non-switchers, non-unemployed numbers in the report argue against the stimulus, though they are lumped into this argument. The not-in-labor force but actually wanting to take a job component is just evidence that the unemployment measure is understated. That recent college grads were employed suggests that the stimulus helped the other main labor problem of the recession, permanently depressed wages for those entering the labor force during the downturn.

RE: bleh's first paragraph, "reading comprehension, how do that work?"

Go back and read Tyler's point 3.

Yeah, I skimmed pretty hard. Still, while #3 argument addresses this, it admits that by shifting a worker that an eventual layoff is averted. Furthermore, it was averted in such a way that the private companies in the sector were made more healthy: Payroll was reduced sooner than otherwise possible, no severance package need by paid, and finally morale costs of a layoff were avoided or delayed.

You actually don't have to look at the downstream jobs, although you could, although that would be a different paper, and the authors could do that, and of course I hope they do.

The point is, why would you expect the basic trend for the downstream employers to be different from the findings for the upstream employers?

You assume that eventually they will run out of employed workers, but that is not actually a mathematical fact.

The downstream employers don't know they are downstream. When they go to replace the employee, they don't go to a different labor market.

One "unintended consequences" memes kicking around is that the dramatic extension of unemployment benefits led to longer term unemployment through a combination causes (e.g., reward of the wrong behavior, causing people to look less hard for work which is unfortunate since the longer someone is out of work the harder it is to find, etc).

Did you have a clear view on the validity of this argument?

You lost our interest at "meme."

Can we please cease dignifying this policy action by calling it "stimulus." It's quite apparent there was nothing stimulating about it. At best, it allowed beneficiaries to tread water for a year, and now that the spending has evaporated, the markets are sinking to their fundamental levels.

Also please cease referring to "shocks." There was absolutley nothing exogenous about the causes of this crisis, no matter what your preferred explanation is. We didn't get struck by a meteor or invaded by Martians, and the Yellowstone supervolcano hasn't blown yet. The crisis was caused by policies and actions of parties responding to incentives and existing conditions. Exogenous shocks are nothing more than a modelling convenience, not a significant political-economic reality. Even natural disasters which are reasonably foreseeable should already have figured into most decision making processes.

Well, it is kind of pointless to debate all of this when we don't know what would have happened with the economy without the stimulus. The focus on 'why more jobs were not created' assumes that the stimulus was positive in the first place. It is not just a matter of capital crowding or even workers crowding. If we didn't have stimulus we would probably have more people laid off at first but would we have a bigger recovery right after that? How many workers were hired/kept working because of the stimulus and then laid off right after the money ran out? How does that impact long term unemployment?

If we didn't treat economic forecasting and counterfactual speculation as gospel truth, what the heck would bloggers talk about all day?

1.) Like a lot of well qualified and easily employable construction workers, for example? Because it's pretty clear a fair number of well-qualified and easily employable construction workers spent a lot of 2010 on there backsides in front of the TV (24.7%!!!!)....

...when they could have been building framework, operating backhoes and welding steel plates to fix any of about 6,000 bridges that are currently in some stage of falling down in Pennsylvania alone.

Why is this??? The simplest answer is that even a full year after ARRA was passed there still wasn't enough "money chasing" these people or these problems.

Unless your intention is to mislead people you should just nix #1. It is a demonstrably false assertion and you really should know better.

I think you're saying the stimulus wasn't "well-targeted" at skilled, unemployed workers. That's a major point of the Jones-Rothschild papers.

Better targeting-->Better stimulus.

Agreed but I am also saying bigger stimulus means more employment because the targetting is less important the more money you spend on things we can agree are important.


So, in your mind a framer or roofer is now qualified for structural welding? You've got to be kidding, right? So, balloon-frame construction workers are out of work, and your suggestion is to have them build bridges? Have you ever looked at the differences in framing and materials between the two?

Where did I say anything remotely close to any of this???

All I said was that you need people to do construction work and you have people with construction experience on their butts.

Yes, they need to be retrained. Yes that will take 3-5 years.

No, we didn't start that training three years ago when we should have.

Yes, thats because we low-balled the stimulus.

Yes, if there were more money chasing people there would be more people in the chase!!!

The broken bridges are not on a schedule. They are not going to fix themselves in five years. So a five year training period just means the sooner we start the sooner the job will get done. Same as for a ten year training period or a five week training period.

Why do adults need to be explained things like this?

I just can't leave well enough alone.

You're showing a massive elitest bias in not knowing the difference between stick building and road work. I guess to you office guys, all muscle work is interchangeable.

Most economists, and Keynesians and Marxists in particular, seem to assume that labor is fungible.

Then let them have their steel bridges built by carpenters.

Maybe, you should (leave well enough alone.)

If it ever made sense to overinvest in training people now is that time. That costs money. That money doesn't dissapear into ether. People but things with that money. The people who get trained help fix our bridges faster than they would get fixed otherwise.

Where are you getting lost here?

(Not to mention that in PA a lot of the bridges are reinforced carpentry but we will leave that alone...)

Parton me, 'people BUY things with that money'

(I still cant do syntax and snark at the same time).

An assumption that was never completely true, and has become less true over time.

I do think there is an elitist bias here. Most economists are academics who have never worked in serious positions inside industry. They may not have a reasonable appreciation for the small differences that make all the difference.

GinSlinger is right. Nowadays, there is no such thing as a 'construction worker'. If you're building a bridge support, you don't need a 'construction worker'. You need a metal fabricator with a journeyman's certificate in welding and/or riveting, with experience in the type of steel being used and the specifics of construction construction materials, type of joinery, etc. Maybe someone certified for TiG welding, who's comfortable working from a height in a harness. Then you might need an AWS-certified welding inspector. Need to put down a simple road? Oh, now you need someone trained on an asphalt machine, several people who can do survey work, someone with a class 1 license and experience with a large dump truck, grader operators, A job foreman, people with safety certifications...

Even a seemingly-monolithic job like 'welder' is broken down into multiple sub-specialties, each of which can have its own certifications and requirements. Gas welding, resistance welding, solid-state welding, arc welding... And these days it gets even more specific, because there may be a need to be trained on specific robotic welding machines or be able to read radiographic printouts or other specialized skills. There are also special skills required of people who weld large bits of metal together, so a guy who can arc-weld a car body many not be able to do the same on a bridge support without significant training and experience under the eye of a senior guy.

Go look at the job listings. Get back to us if you can find one that says, "Wanted: Construction worker for bridge project"

How about a job opening that says: Entry level position? Training available?

Two weeks of training will get you pretty far in a lot of professions. I don't know how many years you imagine it would take you to learn to drive an asphalt machine, but given the lead time on all these stimulus projects you probably had the time.

Sure, you PREFER to hire workers with experience doing the exact job you need done. But if there is an elitist bias, it's the idea that blue collar workers a somehow deficient in the human capacity of learning.

I work in a specific field of software, if we could only hire people certified or even familiar with all the systems we use, we'd never hire another person again.

Two weeks of training will get you pretty far in a lot of professions? Name one. If you think any of the jobs I mentioned can be learned in two weeks, you're just proving my point about how clueless people who have never worked in industry can be about what it really takes to do these jobs.

Just to give you an idea: A basic welding ticket requires that you find a job as an apprentice. Then you sign up for an 8-week college semester with multiple courses. Then you work as an apprentice for a long time. Then you do another 8 week semester. Then you work as an apprentice for another long period. Then you do a third 8-week semester. That's 24 weeks of full-time instruction.

How much apprenticeship is required? If you want to work in industrial and construction, about five YEARS. After that, and after completing all the coursework, you can then write a very extensive series of exams. If you pass all those, and submit all the paperwork, you are finally qualified to do unsupervised welding on a bridge.

Working as an apprentice is not working? Do the things you weld then not get welded?

Like you said, step 1 the process is finding a job apprenticing. This job of apprenticing involves you being paid to weld things. Shocking, that supervision is required for untrained workers.. But if you are imagining what stimulus supporters had in mind was a bunch of unsupervised know-nothings assembling at a riverside and giving birth to infrastructure, you are mistaken.

As a software architect, I'm surprised at the incredible surplus of qualified programmers that you say exist in your subfield. I believe most people in our profession would argue that hiring an able person is extraordinarily difficult. The difficulties only multiply when you consider relevant experience. How could it be otherwise? Good software generally requires deep specialization and domain expertise to create, and the people who write it are not interchangeable. Have you heard of many filesystem folks who reinvent themselves as machine learning wizards? Search engine experts as Windows application designers? Better yet, casual phone app designers as any of the above? This is not to say that they couldn't--only that it would require a very large amount of time and energy on their part and patience on their employers'. Complex work in software is not handled by finding and retraining individuals in superficially related areas. It is simply too expensive and time-consuming to acquire the relevant knowledge. Our money is much better spent hiring someone who fits the bill from the start. If this has not been your experience, I would be very curious.

GinSlinger's and Dhanson's comments about engineering broadly comport with my experience as a programmer, so I am naturally biased in agreeing with them. You are correct when you point that work as an apprentice is indeed work. I doubt, though, that stimulus money chases construction apprentices when the benefits of knowledgeable, experienced employees are as high as the previous commenters claim.

"Working as an apprentice is not working? Do the things you weld then not get welded?

Like you said, step 1 the process is finding a job apprenticing. This job of apprenticing involves you being paid to weld things. Shocking, that supervision is required for untrained workers.. But if you are imagining what stimulus supporters had in mind was a bunch of unsupervised know-nothings assembling at a riverside and giving birth to infrastructure, you are mistaken."

You don't really know what you're talking about, I'm afraid. No one is going to hire a person, call them an apprentice, and let them start welding structural bridge components. In some cases, a person isn't even allowed to touch a structural member unless they not only hold a journeyman certificate, but have special endorsements. We spend a lot of development time building systems just to manage the extensive amount of extra qualifications, licenses, and certifications required in manufacturing and construction. It's not simple, and it's never fast.

Again, can you name a stimulus job that can be done by any unemployed person with two weeks of training? Or even six weeks? I can help you out: On a road crew, you might be able to train the sign holder that holds the 'STOP' sign to do it right in two weeks. And I'm not even sure about that, because I don't know what safety training is required to direct traffic. But that's about all I can think of. Everyone else on a road crew is a specialist in something. And bridge construction crews are for the most part highly trained people.

Checking around licensing information, seems like with the right drivers licenses for heavy machinery, anybody who was a cement driver qualified to fill a job as asphalt paver or dump truck operator. Obviously they are not doing the work on day 1, but they are not on a journeyman system. Honestly, describe to me the first two weeks of employment of a cement truck driver who comes in with class II licenses but no experience specifically driving a cement mixer?

BLS says we employed 500,000 welders in 2000 and we only employed about 450,000 in 2008. Doesn't seem like crowding out somehow.

In 2008 we added 50,000 welders while 25,000 retired. We are spending $22 Billion on wages for welding labor. Agian not what you would expect from crowding out, huh?

Its not about labor being 100% fungible its about providing an incentive for a portion of 25,000 retiring welders to to help train another 15,000 welders a year. Its about subsidizing $10 Billion over five years in time and half for half a million welders to help train another 30,000 welders.

Last but not least - its about ponying up $500 Billion to fix the DAMNED BRIDGES!!!!

Yes, it is going to take 10-15 years - but it will never get cheaper than now to get started.

If terrorists had a plot to destroy 10,000 bridges in the US we would find that money. I guess their plot was to distract us for a decade while nature took its course.

This is a great thread. Very high-quality replies and discussion (mainly by GinSlinger and Dhanson)

But anyway, I'm only half joking here: If modern construction techniques require too much specialization, why not have a stimulus package to build a bunch of bridges and roads using 1930s techniques? Back then, it was a lot closer to "ten trained engineers, a hundred trained foremen, and five thousand unskilled labourers", right? And plenty of bridges from then are still standing.

What am I missing?

Opportunity cost

So, you're saying that you want a bunch of rickety wooden bridges that can barely hold up one Model-A at a time?

There is a reason that may older bridges were built so narrow that two cars traveling in opposite directions couldn't pass each other.

I'm back and sad to say this is pretty much what I expected. I think -bleh- hits it on the head while
GinSlinger, Dhanson and Michael are lost in the forest.

Take Welding... Nevermind the low haninging fruit of carpentry, and construction equipment operators.

I realize that a guy who can weld needs a few years of training. The thing is we have had a few years of good candidates for that training sitting on there asses!!! Don't forget that eventually we need to fix SIX THOUSAND BRIDGES!!!!!!!!

Does it make sense to:

A) Wait around for the same qualified welders to fix each and every one of them, or B,
B) Start training welders while potentially qualified people are sitting on their asses so we will be able to fix bridges sooner.

There would only be a justification for A if we risked some huge glut of unemployable welders. That is crazy.

The justificaitions for B are 1) there are potentially qualified people sitting on their asses and 2) We have negative interest rates. If we wait 15 years I guarantee we will not have as many.


Shocking, I know but...

this country had less than 500 tanks in 1940 and had produced almost 30,000 in 1942 alone. At the same time we launched the battleships New Jersey and Iowa. By 1945 some of the best welders were young women who were high school girls in bobby sox a few years earlier. Guys we figured out how to train an army of welders then (after a decade of depression when there wasn't a lot of welding going on). We can do it now.

Fortunately, back then we had a liberal elitist president who viewed labor as marginally fungible to kick everybody's ass until we got that sh*t in gear!!!

I'm sorry I said there are six thousand bridges that needed to be fixed. I meant there are SIX THOUSAND BRIDGES that needed to be fixed in PENNSYLVANIA ALONE!!!!!

Tyler, your last paragraph reminds me, again, of why macroeconomics based on aggregates of goods and services that are not good substitutes has never been able to explain the phenomena that most people and policy-makers are concerned (particularly in the context of a crisis). I know many market-oriented economists that have always been concerned about aggregating tomatoes and apples --you don't need to be a Hayekian not to aggregate them-- and therefore reluctant to engage in any sort of macro analysis. I hope labor economists reading your post can explain to MR readers how much the field of labor economics has progressed in the past 50 years in identifying reasonable criteria for the types of labor relevant to macro analysis.

For MR readers that have read Tyler's many posts on monetary theory and policy, I recommend to compare "money" and "labor", two of the aggregates of macro analysis. Macroeconomists have been manipulating the definition of "money" (M0, M1, M2, ......, M999) because they can rely on huge amounts of data and their only criteria for aggregation is a "good fit" --to be later mocked by Goodhart Law. With "labor" they have not been able to play this sort of research game because they don't have data, but more important because it is too obvious that there are "Not Two Alike". [One may clarify the idea that money is fungible, but there is not way to argue that labor is fungible].

More markets in everything and less tiresome Holy Warring, please.

Tyler, the great debate about ARRA's success/failure deserves a SNL break:

Can we keep doing the same thing over and over again and expect a different result? Yes, we can!

If your narrative is that highly skilled, highly employable workers are doing just fine in this recession, then your narrative has serious problems. Maybe you could bring some support to this claim, but unemployment has increased across all sorts of demographic groups from those lacking HS diplomas to PhDs. Even for those with jobs, my anecdotal observations are that managers are taking advantage of softness in the market to squeeze wages and benefits. Even for very good employees, looking for a new job in this environment is not an appealing project.

It seems to me that you are starting from the assumption that this is an Austrian recession and working from there. Then it's of course we can conclude that when the government hires some people away from other jobs, those other jobs just go unfilled. Where is the evidence that there are classes of workers doing just fine today in comparison to other recessions or time periods of 9% unemployment?

The average organization who responded to our survey said it was as hard or harder to find good people compared to before the crisis. That's one piece of evidence that the stimulus was not targeted at weak sectors of the labor market.

Another example: This SF Fed study notes that ARRA's transportation funds were not targeted based on state unemployment rates--they were planning to use a pre-ARRA allocation formula:

A modestly well-targeted spending program would spend more in high-unemployment states, or in states with the largest increases in unemployment. Doing better than that is hard---it's a real Hayekian knowledge problem....

Re: "A modestly well-targeted spending program would spend more in high-unemployment states, or in states with the largest increases in unemployment."

Which is why Garret should support Shiller's proposal to have a list of projects in place in the future so that in the next downturn you could pick the targeted projects? Correct?

Did Shiller steal that idea from me?

I'm joking, although it is plausible. I never saw it before I said it and now it's Shiller's new idea. It's shocking, partly because it is such a no-brainer that it's embarrassing that Keynesians hadn't thought of it and it hat to come from me/Shiller.

By the way Bill, Garrett doesn't have to be for anything. You are confusing advocacy with investigation.

Wouldn't it be funny if he did? You're right that it is strange that Keynesians hadn't thought of it before.

Not every PhD is highly employable. Gender studies experts etc.

When you take resources from one part of the economy to another, all you are doing is moving demand around. You're not creating new demand.

The only reason people think it does is that they are confused about what money is.

Yes, if you end up hiring people who can't normally earn more than the minimum wage and firing people whose skills are suited to work that makes economic sense, those more competent people will have an easier time finding work.

But as soon as the stimulus is gone, that effect reverses itself and we are back where we started at great expense.

Much better to address the fact that one minimum wage for every area, despite the disparity in living costs, is preventing demand from rising to meet supply for workers that need retraining and who's work currently is worth the minimum wage.

Policies that don't address this market problem won't work. Even cutting taxes on profits will fail if there is no profit to be had in hiring a worker.

Stimulus is specifically for when resources are under utilized & you are borrowing money to do the work.

If you decide to build a road instead of not building a road, you're by definition creating more demand for road-building. Taking some resources to do so may reduce the SUPPLY of resources available to meet other demand but you aren't making the resources or their products any less in demand.

Regarding your simple solution to everything, there are already local minimum wage laws.

Let's make this even easier: Who was or knows of someone, hired under the ARRA?

(I mentioned the question of hiring the already employed to a family member getting a stimulus grant for an environmental project, and she essentially told me to go fuck myself.)

My brother kept his teaching job thanks only to stimulus funds. If such aid had been continued or expanded 2 teacher cousins of my girlfriend wouldn't be recently graduated unemployed teachers.

Anyone know a person that lost a job because ARRA was discontinued or stimulus ended?

Surely if we wanted to help the worst off and unskilled, we would have waived Davis Bacon requirements instead of Congress imposing them on new programs.

The Democrats chose to support union members (mostly those with jobs) and the existing firms that know the rules, rather than the unskilled.

Yup. Democrats don't like to admit it, but FDR's Keynesian programs were hurt tremendously by his crazy pro-labor policies. He thought that people needed higher salaries to spend money, but actually they need salaries, period, to spend money. Artificially making salaries higher than their true market value just hurts economic growth and increases unemployment, with deadweight loss overshadowing the higher salaries to a select few workers.

Matt, How can labor rates be higher than the market if the requirement is that they pay prevailing market labor rates? You can look at this a different way: the government wants to award contracts to those who pay the prevailing labor rate for the project, denying a subsidy to a contractor who pays less than the prevailing rate. Otherwise, the government would be subsidizing with stimulus dollars the growth of firms which paid less than the prevailing market rate. You could view it simply as trying to take wage rates out of the equation so long as they are prevailing market rates.

I am not an expert on Davis Bacon, but you could see the problem if a contractor who paid squat kept winning government contracts. Stimulus dollars would be used as a wage suppresant to prevailing wages. Some may think that good, or not, but it doesn't make the constituents too happy if federal dollars are used to give a haircut. Opinions may vary.

Huh? There is no such thing as a wage below the market rate. The market rate is the rate that somebody is working for. In the case of Davis-Bacon, the "prevailing wages" often become wages set above a true market level. In the 30's, the prevailing wages often closed out blacks who were willing to undercut whites. Many Southern Congressmen voted for the bill specifically so that the blacks would not undercut whites.

Your argument only makes since in old "wage slave" terms, where it's assumed that companies can pay employees whatever and still get the same amount of work. Indeed, in my arguments with union supporters, they seem to think that Andrew Carnegie's pinkertons are coming out of the shadow at any moment to bust up a strike. That's now how it works today though. If you pay employees minimum wage, you will get minimum wage employees. The government won't accept your shoddy work and then you'll lose far more on the rework paying true market wages to employees who can actually do the work.

The only reason any "prevailing wage" clause is necessary is to enforce some sort of cartel on labor, where people with certain connections get paid more than they would if they didn't have that protection (i.e. whites stopping blacks undercutting them). Conservatives think I'm crazy because I support Keynesian liquidity trap arguments, but they're absolutely right that unions and Davis-Bacon today do far more harm than good.

Are you considering wages the pay rate or pay plus benefits?

I think we need more crazy up in here. I agree that FDR's policies would shoot themselves in the foot. For me the most egregious was price controls.

If there's no such thing as a wage below market rates then there's also no such thing as a wage above market rates, since if an employer agrees to pay a given wage that is (locally, for that one transaction) the market rate. And no, the minimum wage law does not vitiate that logic, since the employer always has the option of now hiring anyone at all if the thinks the wage is too high, just as the worker has the option of not working if he thinks the offered wage too low.

Now true, employers may wish that wages were lower and gripe that they are being forced to pay above-market wages, but they are paying those wages. By similar logic I might think that gasoline should be $1.19 a gallon again, and 90% of the country may well agree, but that does not make $1.19/gal the market price for gasoline.

Let me elaborate a bit on that post. I've worked on some government projects which have had Davis-Bacon requirements. Your point makes some sense about how a contractor with the lowest wages could keep winning contracts. My point is that the government would not accept their likely shoddy work and the government can not only switch contractors, but also go after the contractor to recover any sunk costs.

I've seen such fly-by-night contractors try to low-bid contracts with private companies without Davis-Bacon requirements and it never works out well. The inherent turnover, rework, etc. of such contractors means they often go bankrupt within a couple of years once they can't meet payroll. Such a contractor without Davis-Bacon requirements may try the same stuff with the government, but they would quickly fail the same way they do with private companies. Unless the government did not check their work at all, they would run out of money because the government would not check off their work.

Basically, if your scenario without Davis-Bacon was possible, it would have already happened with private contracts. In any case, if there's that big of a labor surplus for masonry workers or whatever where a contractor could name their wage, then perhaps the wages should be depressed until the market picks back up. Once the market does pick back up (or enough people quit masonry work), the government and the private sector will bid up prices for skilled workers so they do not get shoddy work.

Matt, How can labor rates be higher than the market if the requirement is that they pay prevailing market labor rates?

Bill, how can labor rates be lower than the market if people will work for that rate and companies will pay them that rate?

I'm highly confused as to why you think that market rates are something best determined by a six month study and a government report, rather than the market.

You can almost always do work with more people paid less, or fewer people paid more. Explicit policies designed to emphasize using fewer highly skilled, highly paid people increase unemployment, while helping those people out. In a time of high unemployment, it doesn't seem like the right policy. It's worth having a few contracts that get people unemployed retrained with new skills, even if they have to be paid less than the people who already have the right jobs who could do it with fewer workers.

For all the wrong that FDR did, it's Democratic Party policies now that make things like the WPA impossible.

So, we should use federal money to subsidize wage reduction?

So, we should use federal money to subsidize wage reduction?

You're arguing that we should use federal money to subsidize the well-off who already have good jobs by imposing protectionism preventing the poor and unskilled from taking those jobs and competing.

I'm arguing that we should use federal money to subsidize the poor, unskilled, and unemployed who don't have jobs to let them get new ones, by opening the playing field.

The people you're arguing for make more than the median income. Since they're paid out of tax money, you're arguing for the have-nots to subsidize the haves. I think that's indefensible in a time of high unemployment.

Do you similarly argue in favor of the ABA's power over bar admission, or the power of the AMA and med schools to limit the number of doctors, because you hate the idea of "wage reduction" for lawyers and doctors?

John, I'm all for opening up the professions to more competition. Glad you're on my side, but I don't understand why you think the grand polity would like it if they subsidized companies hiring below the prevailing wage.

Kevin Drum's argument about hireability makes a lot of sense to me.

At first, the stimulus only gets the most desirable workers. The already-employed with long track records. Then, as demand continues to increase, employers lower their standards. Eventually employers have enough demand that they'll go for the people six+ months unemployed in a different field.

Indeed, if the stimulus was too small as Drum says, you would expect it to leave behind the "hard-core unemployed." The increase in demand first works through the employed and the semi-ubemployed before getting people who hiring people who have been idle for months or need training.

I still think monetary stimulus can do this same thing without the issues of fiscal stimulus, but Drum is right on his interpretation of the paper. It's a myth that somehow a huge cross-section of the population became worthless as workers, but you would expect a too-small stimulus to hire other people before hiring the needy workers.

What you are talking about sounds like a cyclical expansion.

That's the question, can you borrow your way to a sustainable economic recovery.

Have the average and the median time it takes businesses to fill vacancies increased since 2008?

Sounds like we need to stimulate through direct government employment of the unemployed, rather than through tax cuts then, doesn't it?

Directly hiring unemployed people is not the fundamental purpose of stimulus. It might, for some projects, be the biggest bang for the buck, though.

Dear god no. We don't need more TSA employees, which would bring more unnecessary "security measures" to airports.

This is just one more way to say that heterogeneity matters, which is to say the heterogeneity refutes Keynesianism ... in other words, Hayek defeats Keynes ...

See, heterogenity is exactly what I don't understand about anti-Keynesian arguments. According to this theory, in late-2008 for whatever reason, 5-15% of Americans who were working suddenly became unhireable. What, exactly, changed in employers minds for them to say these workers are suddenly worthless at any wage? Why do many of the unemployed include, say, recent college graduates who have no issues using computers? Why, for example, would states be laying off so many young teachers and other workers?

If there are answers from Austrian theory that actually match the empirical data, I haven't seen them. Such free market purism should have no space for much of the labor force suddenly becoming idle. Why, exactly, can't these idle workers find any work in their field for any wage? I know that an electrician can go work at McDonald's, but if the typical wage for electricians is still $15/hour, why can't an unemployed electrician find work for $10/hour? These discontinuities have no explanation in Austrian theory because Austrian theory simply isn't true.

First, no one's saying they're worthless at any wage. They're saying that there are no jobs for these people that are A) in their area, B) within the scope of their expertise, and C) worth paying at least $7.25/hr for, that being the minimum wage.

I don't think many would argue that if you dropped the minimum wage and dropped all government financial aid, some of those that are unemployed would find some kind of work in short order. We've just made a social agreement that it's better to support people with govenrment money than to make them work for less than $7.25/hr.

Austrian theory is all about heterogeneity. How can a whole lot of people be suddenly unemployable? Easy: just have an asset bubble pop which changes the calculus of supply and demand. Projects which were viable and profitable when people assumed a 4% growth rate may not be viable at a 1% growth rate. Housing projects which employ carpenters, plumbers, electricians, bricklayers and others are cancelled when demand for housing ceases.

The Keynesian argument is that if you just throw more money into the mix, there will be a rise in aggregate demand which will cause those people to be hired. The Austrian argument is that the problem is structural - there are more carpenters than the economy demands - and these people aren't going to be employable at previous pay levels (representing a certain level of productivity) until they retrain and develop skills in an industry that needs them. A stimulus that artificially props up the housing industry with borrowed money prevents that necessary adjustment from happening and just freezes the economy in a low-productivity state.

Does a small stimulus do nothing because it does not meet a feedback threshold? What is that feedback threshold?

Does a small stimulus do the most per dollar because it avoids diminishing returns? Where are the diminishing returns?

Both? Where are the thresholds?

It seems to me that a lot of people here want to increase the total number of jobs. Why? it seems to me that the economy is just readjusting and this may take some times, perhaps years. I dont understand why we need to throw tons of money at the system in order to hope that the aggregated number of jobs does not decrease. Is it really that bad for an economy to go through an adjustment period?
Let all the little micro economies do their own part to help the macro economy.

Anecdote: I am an electrical engineer in my mid 30's with a master's degree, 5 years of engineering experience and 5 years of military experience. I was voluntarily unemployed (to look after my young daughter/earn my MSEE) until the summer of 2009. I spent 3 months looking for work, blasting resumes anywhere it looked like they might stick, following up with phone calls and messages, etc, and was finally hired by a relatively large company to a position created by ARRA funding. In the spring of this year, ARRA funds had dried up and there'd been no growth in my department, so the company laid off employees. I was retained, but at the expense of someone else's job. So, my experience shows another side of the stimulus wrt labor markets: ARRA funds could be used essentially to replace less productive employees with more productive ones, with a long period of overlap that provided for job-specific training and evaluation of the new hires.

So education and retraining should have been the target of stimulus?

Yes. But practical, on-the-job retraining is more useful than many "retraining" programs that don't actually teach job skills. But Davis-Bacon requirements make it impossible to hire unskilled labor for lower wages while you teach them a valuable skill. It means that you can only hire the people who already have the skills and can command the higher wages.

You can almost always do a job with fewer, more skilled people who are paid more, or with more, less skilled people who are paid less.

Too much of the stimulus was aimed at helping the former.

Maybe this was covered already in the thread, but I feel like Tyler is missing another huge benefit of ARRA: the continued employment of existing employees who would otherwise have been fired. The prevention of job destruction is at least as valuable as job creation.

Only if it's a cyclical issue. If not, then you're preserving unsustainable jobs that will go away later.

It's also worse than useless if all you do is preserve bonuses and raises for people who would have had jobs anyway, and there's good evidence that that's what happened. Lots of money went to bonuses and raises for those highly skilled who were in no risk of losing their jobs (both inside the government and with favored contractors); little went to the unskilled and those needing retraining.

It's a case of the have-nots subsidizing the haves.

I'd rather use the money to hire new teachers than preserve the raises of teachers (when private sector workers aren't getting any) in a time of high unemployment.

I'd rather use the money to put unemployed people lacking the right sort of skills to work at lower wages, than use the money for signing bonuses and raises to lure away people that already have jobs.

If you really believed what you write then why not borrow at negative real interest rates and hire unemployed people to pick up litter? I am sure your cousins would approve. It would hit the unemployed, it would be stimulative. You could restrict the policy to only long-term unemployed people in fact. No polarization. Reducing litter has to have some marginal benefit, and it is stimulative. Can you argue against this without contradicting your previous arguments?

There is really an overlooking of distortions of the market. In the post, it seems the stimulus not going to unemployed is lost. What is true is that stimulus not going to unemployed (or to save employed) isn't effective as a stimulus, but goes back in the hands of Americans, it's not burned. It's like a transfer, it will not increase the consumption rate as hoped. It will instead build more capital, more saving. Since that is the GOP alternative plan, the only difference is in the distortion.

You are trading an increase of consumption rate for an endowment distortion. What's the effect of the distortion? Probably, zero. Looking at the size of the stimulus, the size of USA's economy, and remembering that distortions are second order effects, the effect should be way below 1% of the stimulus, if we were at the optimum without it.
On the plus side instead, 50% of the stimulus went to people that are now working, increasing the consumption rate. Another unknown % went to save work places.

Is it a good deal? As Kevin pointed out, a bigger stimulus would have had a bigger effect (in proportion), but since no one is able to correctly compute and compare the pros and cons, whatever narrative is just that: a story.

"What is true is that stimulus not going to unemployed (or to save employed) isn’t effective as a stimulus, but goes back in the hands of Americans, it’s not burned. It’s like a transfer, it will not increase the consumption rate as hoped. It will instead build more capital, more saving. "

"On the plus side instead, 50% of the stimulus went to people that are now working, increasing the consumption rate. Another unknown % went to save work places."

Inconsistent much?

Besides, if you've ever spent time around the bulk of a road crew, you'd know that the money isn't going to increase anything. It's going to go into college savings for those at the top of the crew, into beer/whiskey for those in the middle, and to Latin America for those at the bottom.

"Inconsistent much?"

Not everything you don't understand is inconsistent. What I still don't know is: did you miss my point because of my bad English, or is it your prejudice? I care because I write hoping someone gets what I'm saying; if people don't understand it I'm losing my time. You can try to read it again, and maybe point out the inconsistency.

On the final part,
- Beer/Whiskey is consumption (and mostly made in USA), and I'm prone to believe a lot of those will also eat. And make their children eat. Maybe buy a car.
- Of course someone will also save those money. The point is that moving money to the unemployed, his consumption rate will be higher: it's not that it will be 1.

Since most people advocating against stimulus are pushing for investment/saving, they should be happy for money avoiding consumption and going back to savings: As I was saying in my previous post, the only negative effect (with respect of leaving money to the people) will be the distortion. Basic math says the negative effect on the economy is below the 1% of the stimulus size (I'd say 0%, but I'm now playing by your axioms). In the same figure of the lost opportunity cost (in this economy), I'd say 1-2% total. In exchange of a 10% more of consumption rate.

Is it worth? I'd say so. A bigger stimulus would have had less distortions (but the same opportunity cost) and a bigger increase in the consumption rate (since it would have gone to people with much higher consumption rate, and less money would have go in savings), and it would have been more rewarding. Since no one has a model to compute pros and cons consistently, it's really speculative.

While this open the road to same hand waving, it doesn't mean we have to refuse numbers and proper arguments: in these comments I really see a lot of narrative and very few numbers. When you basically say we are financing the alcohol problem of the road workers, it's not an argument. It's just bad (slightly offensive) rethoric. The point is consumption/savings, pointing out the bad consumption habits of the working class is only meant to undignify it.

"A lot of the money also was spent on capital, land, raw materials, and other factors of production. I’ve never seen good estimates here, but labor’s share is about seventy percent of gdp, actually a bit shy of that."

Insofar as you're just buying and transferring around the stock of machinery that already exists that won't have much effect of employment. But if the machines were idle and are now put to use (requiring labour), or your purchase meant that new machines were needed (the production of capital goods requiring labour) that could still boost employment. What am i missing?

I think its great that we have a lot of data on the stimulus and I also think its great that the administration tried to track things all the way down to level of individual jobs created. But a real danger of this is that we forget the scope of what it is that we are trying to do.

Trying to track individual jobs created is like adding a hundred gallons of water to a swimming pool at 6 AM and coming back at 6 PM and asking where those gallons are distributed. One way to answer that question is with a model (i.e. the 100 gallons are evenly distributed throughout the pool). Another way is to try to use some measuring tools (i.e. somehow tag water molecules with some radioactive element, measure where those radioactive atoms are etc.).

The model is likely to be wrong. (The 'new water molecules' are unlikely not to be perfectly evenly spaced throughout the pool). But your tools are likely to be very shoddy. You may only be able to see, say, 1% of the 'tags' you put in there. This is OK as long as you keep in mind the limits of measurement can do here. Just because you find you can only measure 1% of the water doesn't mean "adding 100 gallons of water only results in 1 new gallon staying in the pool", for example.

So here the fact that its hard to see the exact jobs created by stimulus can mean two things: One that few jobs were created. Or that its hard to measure which specific jobs were created/saved by the stimulus. Ideologues are attracted to the first option but the facts indicate the second is more likely. This is esp. true when you factor in that only a tiny portion of the stimulus was actual projects and a large portion of it was tax cuts, consumption support (unemployment, food stamps etc.), helping states with Medicaid costs etc. To even begin to try to track jobs created by this you'd have to be able to see into individual bank accounts and consumption decisions and then see the impact that consumption had on retailers, then suppliers to retailers, and so on and so on.

Fascinating discussion - but two points I'd like to make:

1. President Barack Obama said at a fundraiser held in a private residence in Washington, D.C., on Thursday evening that the $447 billion American Jobs Act he has proposed will create 1.9 million American jobs—which works out to a cost of about $235,263 per job.

I own a small business. Give me $235,263 and I'll hire FIVE people.

Of course, there is absolutely no assurance that once the stimulus money is gone, they will still have a job! Counting new jobs is pointless if many of them are only short-term, as many of the stimulus-created jobs were. Bleh mentioned that if stimulus money had continued, teaching jobs would have continued. How long should government (meaning me) pay the wages for a position that obviously is not needed?

2. Jobs are an old paradigm anyway. More and more work can be done by fewer people. Folks need to start looking at creating their own jobs rather than depending on someone else. Just look at self-check out lanes in the supermarket, and see the future of cashier jobs.

Comments for this post are closed