September 2022

Sun Mon Tue Wed Thu Fri Sat
        1 2 3
4 5 6 7 8 9 10
11 12 13 14 15 16 17
18 19 20 21 22 23 24
25 26 27 28 29 30  
Blog powered by Typepad

« Best Paragraph I've Read So Far Today | Main | The Economics of the Stimulus in Comic Strip Form »


Feed You can follow this conversation by subscribing to the comment feed for this post.

People don't want more tables at any price?

Not too likely, if you ask me.

Can't find enough workers? Why not pay more? Of course, that raises costs. Have to charge prices. And that means less demand for the product. Which means you don't need so many workers.

Of course, if you think that no one wants more tables at any price, then I suppose that market clearing process just isn't something one would consider.

I think that changes in the composition of demand work much better when "aggregate demand" grows at a slow stable rate. That doesn't mean that there is no structural unemployment.

But where in this story was an account of how the economy gets back to equilibrium? If they don't want more tables at any price, then there is no equilibrium process.

We just wait until someone comes up with something that someone wants to buy?

Well, this "theory" ignores Says law, the role of the interest rate in coordinating saving and investment, and the relationship between the supply and demand for money.

Money expenditures on output are 14% below trend. That isn't because people are trained to produce the wrong thing.

But admittedly, my approach doesn't quite fit in with the no tables at any price--that is, when you through out the law of demand, what is left of economics?

1 out of 5 letter carriers has a college degree ....

I guess if I just keep lowering my asking price for buggy whips eventually my whole inventory will sell.

They still use buggy whips here in Charleston.

I think it depends on how many you have.

Sorry Steve Horwitz, fungibility is not the same thing as substitutability.


How I read that story is that fiscal stimulus is a poor substitute for monetary stimulus ;)

What I also took away from this, but also the entire 2007-2011 recession, is that government is better off focusing on public finance instead of trying to rescue the economy. It would have been far easier to just say: "Look at those low rates, we can issue debt on the cheap and renew all that infrastructure that needs renewing, let's just issue contracts and let people bid on those".

Similar arguments could have been made for tax breaks on investment spending. "The revenue raised does not weigh up against the distortion at these low rates, we're better of financing through debt now and raise taxes later.".

I think the entire rhetoric on jobs and stimulus is needlessly polarizing, but I believe that all parties can agree if we want government to do x, now is the time to do it cheap.

And leave whatever the monetary causes are up to the Fed and focus the rhetoric there.


if you're a Econ student with some time to kill - let's say a week - what (recent) work provides an accessible introduction to Capital Theory? Is Bohm-Bawerk a good place to start or are there better places to start that compare Mises, Hayek, Rothbard, Bohm-Bawerk, on Capital Theory?

It would perhaps be a good idea every time that you post something like that, that there would be a short section at the end with: "Interested? read work [x] by [y]". Or perhaps a section on this blog, the FAQ perhaps, where you have questions such as "I am interested in - for example - Liberalism, where do I start?". And then as the answer: "Ludwig von Mises 'Liberalism' is a good start, then 'Constitution of Liberty' by Hayek, and Hayek's three volume work 'LLL'. And for a contemporary account: Pennington's work on Robust Political Economy.".

Sorry Daniel, but every dictionary I look at indicates that the two words are near perfect synonyms.

From the American Heritage in my office: "being of such a nature or kind that one unit or part may be exchanged or substituted for another unit or equal part to discharge an obligation." Check a few online dictionaries too.

And I have NO desire to start debating definitions with you. You made an assertion that seems to me to be clearly wrong. Trying to split philosophical hairs to find a difference between fungible and substitutable seems like the last resort of those who have lost the debate on the merits.

Sorry - I was under the assumption that we were using teh terms as economists use them. Substitution is a technical term. Don't turn this into a philosophical discussion or accuse me of doing that Steve - fungible means perfectly substitutable. We know things are marginally substitutable even if they're not perfectly substitutable.

You seem to be suggesting that substitution between different types of labor isn't perfect. Isn't that what we all think? Do you think Chait is challenging that simply by pointing out that there's not zero substitutability?

I posted here earlier, but my comment seems to have vanished - I assume there was a posting problem, as I got some kind of error on my first attempt.

This is surely a matter of degree - there is no absolute answer.

Yes, labour is not entirely fungible; but people do have some flexible skills. Half of the current unemployed had jobs three years ago, so they're not completely unemployable. Their skills are not all applicable to just any construction project, but they're not entirely useless either.

Simple hydraulic Keynesianism won't work, because it does distort incentives; but if designed right it will work much better than no intervention at all. The Austrian insights of heterogeneous labour supply and price signals certainly have some truth; so do the Keynesian insights that some components of demand are driven by aggregate income and that money spent out of labour income does have an effect on whether resources suboptimally sit idle.

Does unemployment create value by incentivising people to retrain and redeploy themselves to new industries? Sure. Is it value-destroying because it makes it harder for people to demonstrate their skills credibly to new potential employers? Yep. Which effect is stronger? That's surely an empirical question, and the data in studies like the Mercatus one might help us to answer it.

My philosophical view is that there is no such thing as no intervention: even if the government doesn't do anything directly, that project manager in the quoted comment might still get hired away by another private company; value is always being both created and destroyed by factors outside of our control. I think that (well-designed, carefully targeted) intervention can have positive effects; but it does need to be applied with caution, recognising the risks of misdirected resources and bad investments. Government should be humble, using market signals as far as possible to help it work out which public goods have a positive net value.

Steve it comes down to this - you clearly don't think there is perfect substitutability. You presumably think there is some degree of substitutability (I hope you do at least). Chait never seems to say here that there is perfect substitutability, so let's not pretend he's an idiot for pointing out second order effects that are standard fare in thinking about fiscal policy.

Austrians can argue that substitutability is low and take a stand on that claim, but don't pretend that Chait's point isn't a valid one.

Also I should point out that Jones and Rothschild's estimate of the percent of workers coming from the ranks of the unemployed is fairly high. Other studies of programs have that figure down in the teens and still come out with a net positive impact.

When talking labor, fungibility and substitutability are the same thing, unless you have some weird theory about the homogeneity of production methods and techniques, and homogeneity of internal structures across an industry or something. Otherwise workers could be hired from competitors and start to work directly without any sort of training, instructions or adaptation period, and without any hit on their productivity. It is definitely not the case.

Mathieu - fungibility is the interchangability of some good. We do not think labor is fungible. We do think it can be substituted at some discount. Am I wrong? Is there any difference in Chait and Steve's understanding except that Steve thinks the discount is steeper (but presumably not total - so that even under Steve's construction of the situation there is bound to be some indirect job creation in addition to the direct job creation)?

To challenge Chait's point you have to think that the firms whose workers were poached by ARRA are unable to hire any new workers to replace the lost workers. If you don't think that you have to agree with Chait that there is some additional indirect job creation.

How much and whether it offsets the cost of ARRA is an empirical question (and a tough empirical question at that).

Daniel, how I understood Steve's post is as followed:

1. Keynesian policy assumes that a stimulus can fill the output gap and restore full employment.

2. #1 assumes that labor is fungible

3. Study shows that labor is not fungible as vacancies are filled by both employed and unemployed workers.

4. Therefore labor is not fungible and therefore #1 is false and potentially destructive see #3.

Now I think he has a point there: fiscal stimulus is a very expensive and destructive way of conducting monetary policy. I am not sure though what kind of monetary policy Steve would recommend within the existing context.


I doubt very seriously that Steve Horwitz is unfamiliar with the relationship between price and quantity demanded.

And it is true that an aerospace engineer can drive a taxi cab -- which did happen from some of the stories out of Houston at the time when the NASA budget was first cut after the moon landings, way back when.

But if, say, Northwood University were to terminate my employment(and assuming no alternative employment immediately available in another institution of higher learning), I am not qualified to work on a construction project (roads, bridges, repairing school building roofs, etc.) due to the fact that I have no "human capital" skills or experience to do that kind of work.

I could, I suppose, get a job in some service industry sector, a sales position, a host at the entrance of a WalMart.

So I do have degrees of "substitutability," but my degree of substitutibility is (human) capital specific.

I might have to significantly lower my "reservation wage" demand and my acceptable type of employment to a level of human capital "non-specificity" that delays a person's willingness to accept a job; especially, if there is government unemployment payments that reduce part of the financial cost of being unemployment.

Besides the type of Austrian capital theory that Steve is emphasizing, still a very good guide to understanding why unemployment of men and goods may occur is W.H. Hutt's "The Theory of Idle Resources" (1939) -- which Axel Leijonhufvud once referred to as the "locus classicus" on the theory of unemployment.

Richard Ebeling

Right Richard. And I believe Chait is equally aware that workers aren't perfectl substitutable.

But we don't think substitutability is zero, so I don't see how anyone could balk at the idea that ARRA employment of currently employed workers implies some indirect job creation.

If you want to argue that it's not enough, fine. But let's not pretend that indirect job creation is some sort of voodoo idea.

It's like the hubub over the phrase "jobs created and saved". The idea that you have to compare to a counter-factural and that saving jobs that would be lost "counts" should not be controversial.

The idea that there are both direct and indirect job creation effects should not be controversial.

The idea that labor is not perfectly substitutable but is also not unsubstitutable should not be controversial.

Chait makes a good point that many other people have made.

Richard is right. And further, someone like Richard working as a greeter at Walmart would be a misallocation of human capital, to say the least. (I became sensitive to this phenomenon when I was working at a hotel at night; now I'm not misallocated so much as unallocated.)

Also, the comment from the blog (someone should write a paper with that person) really addresses what is at issue. Since labor is not fungible, meaning that even if you have identical backgrounds, etc., you will still have to get caught up on where the project is -- meaning, there will have to be some degree of training -- meaning the project is going to lose money. Yes, this sort of thing happens all the time -- but government stimulus is like the cash for clunkers, drawing away many more at one time than would have occurred otherwise and -- in the case of stimulus -- for projects that may boil down to hole digging.

So, yes, there is substitutibility. With training. Or with misallocation of human capital. But certainly not without cost. Someone does need to calculate such costs.


Who is "we"? You can't pretend Chait doesn't believe labor is fungible when the piece is titled "Employment Is Fungible". And the title isn't likely to be from an editor.

The difference between fungibility and substitutability when talking workforce isn't simply one of substitution rate because that would conceal the adjustment period, the loss of productivity and other important phenomenons. Perfect substitutability would be saying masons can replace accountants without any decrease in productivity, perfect fungibility is saying masons are instantly interchangeable without any impact on their productivity. That is not true, not even when releasing the "perfect" hypothesis for a substitution rate.

Wow - I didn't notice the title. That's pretty bad. And it's wrong.

But read the post. Does he give any indication he thinks labor is interchangable?

Fine - I didn't do due diligence and neglected the title of the post. Let's call Chait wrong on that basis alone. Chait is beside the point.

There is still indirect job creation. Lot's of people have criticized the paper on this basis over the last 24 hours without using the word "fungible". If you all agree with that point, then great.

re: "...perfect fungibility is saying masons are instantly interchangeable without any impact on their productivity. That is not true"

Agree 110%.

Now would you agree that there is indirect job creation associated with the replacement of the people hired by ARRA projects and that the magnitude of this indirect job creation is an empirical question?

Because I don't see how one couldn't agree with that, and if you can agree with that I don't see how we aren't saying "Chait was right to point this out despite his dumb title".


"Who is "we"? You can't pretend Chait doesn't believe labor is fungible when the piece is titled "Employment Is Fungible". And the title isn't likely to be from an editor."

Daniel has a very bad habit of getting offended when critics of people whose views he shares take said people at their word. When we quote Krugman chapter and verse, Daniel says "well he really means this instead." Sorry, I can't read minds, just English.

Chait said "fungible" without any explanation or adjectives attached to it. That was his core criticism of the Mercatus studies. It was the central assumption of his whole argument. I simply took him at his word.

The person you should be criticizing Daniel is not me, but Chait. He's the sloppy one, not me. Same with Krugman.

Steve -
I'm not offended at all. Stop saying I'm philosophizing or that I'm offended.

If Chait is being confusing he's not the only one - I find your whole argument here very confusing as well. You seem to only be saying that labor isn't perfectly substitutable. Fine. I've said multiple times now I agree with that.

Would you agree with what I asked Mathieu, "that there is indirect job creation associated with the replacement of the people hired by ARRA projects and that the magnitude of this indirect job creation is an empirical question"

Chait is not confusing to me. Chait simply seems to be saying this to me. You seem to think he's saying something else. OK, there's some confusion here. But can you agree with that statement I made above?

If so then the only problem seems to be that you think Chait is relatively less clear and I think you're relatively less clear.

But that's what comment threads are for - clarification and questioning.

So do you agree?

The problem is, Daniel, you disagree with all those you support on all the details of the things they actually say, but support them "in general" -- on what you think they ought to have said. If you believe in heterogeneity of capital, labor, etc., then you don't believe in Keynesianism.

Frankly, I get a little tired of playing football with people who keep moving the goalposts.

I don't think labor fungibility is a standard assumption of any school of economics. And the only times I can recall hearing the word in the popular press is when people point out that oil is fungible rather than nations serving as monopolistic competitors with distinct varieties of oil.

Can we get some real science on this? I could just as easily argue that because labor is not fungible, if the stimulus project hired John away from his current job, that employer will actually have to hire two new employees to replace John. In that case, the stimulus creates a ton of jobs (even before other potential multiplier effects kick in).

That comment gets absurder and absurder as it goes on. Granted there are transition costs for retraining and education, some of which can take place on the job, but by the time we get to "none of these table makers can build a chair" you're talking about skill sets that largely overlap. "Labor" might not be fungible, but skills are.

"Stop saying I'm philosophizing..."

Steve accused you of philosophizing?! That's vicious!

"If you believe in heterogeneity of capital, labor.."

I believe in one capital, heterogeneous, complementary, the maker of all satisfactions seen and unseen, and I believe in one labour, not perfectly substitutable, that proceedeth from the right hand of the capital...

Just one aside. The goal of stimulus is to increase the flow of capital. If a stimulus project does cause more money to be spent/wasted overall due to stealing away workers, that doesn't matter, or it even be beneficial --- for the purposes of the stimulus: to increase the flow of money. All that wasted money has to change hands a few times and get taxed along the way. It went from some firm with lots of money (or debt) and into someone else's hands who used it to buy food or an iPod or a new cell phone or to feed the voldestreet market... I think I just heard GDP tick up a few cents as I was writing this...

However, great article and cited blog comment! Austrian school is just as idealist as the Keynesian, except that the Keynesian system is a much HUGER unknown --- a chaotic system seemingly. At least in the Austrian system, we know what will happen (it's not what the Austrian's say what will happen though --- they are idealists). The Keynesians are just in denial of what their system ultimately ends up as --- or are at least experimenting to see how long it will last and under what conditions it ultimately fails.

I wonder just how much time and effort this particular discussion is worth. Presumably the government could create jobs just by going to the unemployment roles, drafting those people into the army, and paying them freshly printed greenbacks to do various kinds of work. If some of the people drafted weren't unemployed or didn't show up or didn't get their paychecks, it would be an embarrassment, but so what? The real impediment to economic recovery is the distortions in the structure of production caused by previous interventionism, and the inhibiting of the necessary processes of adjustment by the new intervention.

Oops -- "rolls."

Steve no one's moving the goal posts. Mathieu brought information to my attention which I am happy to use to change my opinion and say that Chait is cookoo. Done.

I still find your post confusing, so I'm curious what your answer is to that question. I still find your position confusing except that I know you don't think labor is perfectly substitutable, which I agree with you on.

Troy - I don't follow your logic. I don't know of any Keynesians who think that labor or capital is homogenous.

"I don't know of any Keynesians who think that labor or capital is homogenous."

Perhaps you're right. But they sure talk as if they do.

I don't know what's so confusing about my post.

1. Chait argues labor is fungible and that this undermines the Mercatus analysis.

2. I say labor isn't fungible and that this point is made particularly clear by Austrian capital theory applied to human capital.

3. The MR commenter gives some real-life examples of the point.

4. Chait's criticism doesn't have the legs he thinks it does.

End of story.

Of course labor isn't *perfectly* specific (neither is most capital). Nor did I say that recognizing the specificity of labor is the only argument one needs to suggest that stimulus programs don't work. I've offered others elsewhere here. Again, what's so confusing here?

Meanwhile: "I don't know of any Keynesians who think that labor or capital is homogenous."

Exhibit A for my earlier comment.

Do Keynesian models have a "K" in them? They often do. I take that seriously. If it's "K", then capital is all one big blob of homogenous stuff. Just easily to hand, here's a New Keynesian model with a nice fat K in it:

And this one even assumes physical capital and real government bonds are perfect substitutes. No heterogeneity in capital there if any "dollop" of it can be swapped out for bonds perfectly.

I'm now naming this the Kuehn sequence:

1. Austrian claims Keynesian says X
2. Austrian says X is wrong
3. DK says "I don't know of any Keynesian who believes X"
4. Austrian points to textual evidence of X
5. DK insists that's not what the Keynesian meant
6. Austrian does facepalm, wastes a bunch of time arguing about it and eventually realizes it's pointless, and goes back to working on something more productive.

You think the inclusion of K in a model means they believe capital is homogenous?

Does Garrison think investment and consumption are homogenous? He has I's and C's in his model. Be serious Steve. Models are not reality, and you know that.

You want to know what I find so confusing? What's confusing is that you're making a big deal about Chait's title when his point seems to be a very valid claim about the paper. 58% re-hiring of pre-employed labor does not give us reason to believe that the stimulus was ineffective. You seem resistent to acknowledging this point which confuses me even more as to whether you even acknowledge this. So is your answer "yes"? If it's "yes" then I'm less confused. But don't pretend that's obvious. You've been talking about fungibility the whole time and failing to address Chait's point.

I resent being treated like I'm being unreasonable here. I'm genuinely confused about what you think of Chait's actual point - not his title. If you can't accept that some people occasionally find your points confusing, why have a comment thread?

I'm the author of the comment in question. The little aside about tables at the end was just a simplified thought experiment - don't take it literally.

The main point I was raising was rather specific to certain types of stimulus - primarily "infrastructure investment". Government projects to build new roads, bridges, invest in new technologies, and the like.

The key concepts I was trying to get across in that rather long message were:

- When new projects are started, the kinds of people needed are highly unlikely to be unemployed. On the contrary, they are already in highly-valued positions, and taking them out of those positions is very expensive.

- These people are scarce, even in a recession. Companies often have chronic shortages of such people. They may not be replaceable at any cost, or if they are they'll have to be poached from yet another company. You won't find them in an unemployement line.

- Removing such people from their current positions may leave gaps that cannot be filled in a timely manner. This can have significant costs, and those costs are not captured in estimates of jobs created by stimulus. Those costs can be many multiples of the person's actual salary.

- Fungibility is much lower in blue-collar and technical work than I think some Keynesians have an appreciation for. A welder capable of welding a car body may not be qualified to weld a bridge support. An electrician who can do home wiring may be completely unqualified to work in a high voltage power station. Retraining such people can take months or years, not days or weeks.

Unstated in the message, one of the underlying beliefs I have is that as stimulus gets more specific, fungibility declines rapidly. If I designed the stimulus from the bottom up, by first looking at the nature of the unemployed, and then designing projects tailored to their specific skills and in regions where the most unemployed are, I might be able to employ a significant number, but the value of what they build will be low. But if I start by saying that I'm going to stimulate the economy by building a high-speed train in a powerful Congressman's district, the odds of that project employing ANYONE who isn't already highly productive are almost zero, at least during the 'in-years' of such a large project. In fact, such a project is likely going to have to go outside the country to find enough talent to fill the initial requirements, design, and implementation phases, which will bleed off some of the stimulus money. The people who can do work at that level are always scarce resources.

Even a bridge project for a city may find it difficult to find workers, even if there are thousands of unemployed 'construction workers' in the city. Bridge building requires specialization that just may not exist in quantity among the ranks of the unemployed in a given region.

These details matter. But I don't see Keynesian models taking this into account. They assume one use of stimulus money is like another. In implementation, the results can be radically different.

Another +1 for Mr. Hanson.

I finally agree with something Daniel Kuehn has said: he is "genuinely confused." Many of us have thought that for a long time.

Couple of things. Chait's piece is straightforward and there is nothing in it to suggest he disagrees with his title. In other words, he does appear to really think L is L. Also, since when is Mercatus right-wing?

Second, unless you are specifically looking at unemployment, most models that have production in them use L and K. If you are looking at business cycles, you have a couple of sectors, but then that means K is used to produce two things - but it's still K. The Cobb-Douglas production function is still the work horse of the modeling world. This doesn't apply to Keynesians or New Keynesians only. It's Neoclassical, in fact. I think Austrians are the only ones to not model K as schmoo.

Yes, some researchers look into things like non-substitutability, or lumpy capital and so forth. But that is far from the norm. Maybe the future will bring more of that, but the literature isn't there yet.

Of no great significance, but for the record --

"Workers, like physical capital, have to fit into the jigsaw puzzle. Their human capital must be complimentary to the human and physical capital in the firms where the labor has left"

Unless politeness is important, the word is complementary.

Regards, Don


Keynesians design the stimulus so the money goes to where it is most likely to be spend again and again. In an indirect way this is supposed to stimulate employment. The goal is to get the circular flow going and then let the market sort out the rest.

Jerry -
There's no shame in being confused from time to time.

I still am, Steve - if you want to clarify that question. I would hope your answer would be "yes", but I'm still honestly not sure.

I found the comment that Horwitz quoted irritating. If you go back to the thread, he made other comments complaining about how the U.S. hasn't trained enough engineers. Too many lawyers.

In the section Horwtiz quoted we had at least two uneconomic claims--the no more chairs at any price and the claim that apple can sell all the ipods it wants. At what price?

I have no doubt that Horwitz understands economics, as does Ebeling. But this engineer (or whatever) doesn't.

I don't favor increasing government spending to create jobs ever.

Still, suppose people choose to live in their parents' basements and drive bigger cars. Or, to pick a more controversial issue, suppose people choose to buy Chinese toys and pay for it by selling lumber to the Chinese construction firms.

Where are the jobs going to come from for the unemployed contruction workers or toy workers?

Well, in the first case, it is the auto industry and in the second, the lumber industry.

But when we look at the expansion in the car industry, we see that they didn't mostly hire the unemployed. What would be my first response? Well, that created vacancies elsewhere, which will be filled.

However, it isn't simply that the employment impact of these government contracts are being held to a different standard, and that specialized human and physical capital is suggested to be so brittle that it would appear to be on the edge of collapse, it is rather the comment that Horwitz quoted seemed to ignore scarcity.

Who responded to my first remark with something about buggy whips? Is the U.S. economy really in the state where the typical good or service is like buggy whips? Perhaps there are goods with perfectly inelastic demands. Perhaps the demand for ipods is perfectly elastic at the current price and apple is producing under conditions of perfect competition--all they want at the given price.

There are tens of thousands of products in the U.S. economy today, and my vesion of the world doesn't involve "fungability," of resources, but the notion that the demand for nearly all of those things is perfectly inelastic at the current quantity seems unrealistic.

Finally, I just spoke with a former student of mine who was in the infrastucture construction industry--building roads, sewer and water lines, storm sewers, and the like for new neighborhoods.

The entire firm closed down.

I was mayor of a small town that could use repair of roads, replacement of storm sewers and the like.

More importantly, I have been to meetings where road maintenance funds are allocated, and how much to catch up on deferred maintenance is a regular item for discussion. (With ordinary repairs being traded off for new projects explicitly.) I was there to speak in favor of paving a road in my town, and was more than a little troubled that this project, like those of other municipalities, meant that less of the backlog of ordinary maintenance would be done.

Now, perhaps the new suspension bridge industy, or something, wasn't much impacted by the recession. And having everyone try to build the new giant bridge that was on their wish list doesn't do much good.

And, my experience in local government does suggest a bias to go with the new sexy project rather than ordinary maintenance.

But there is unemployment of infrastructure workers and there is a demand for the sort of infrastructure these guys can produce. I know that for sure.

Like I said, I have no use for infrastructure spending as a jobs program. But I don't think that the experience related by the comment that was quoted was helpful.

Martin: “what (recent) work provides an accessible introduction to Capital Theory?”

I’m not Steve, but I really liked Roger Garrison’s “Time and Money” because he helps the reader relate Austrian theory to mainstream econ. Mark Skouzen’s “Structure of Production” is really good. But if I had to choose just one, and you had a week, I would suggest Jesus Huerta de Soto’s “Money, bank credit and economic cycles” because he draws from all of the best Austrian economists and does a good job of summarizing their main points.

On the fungible nature of labor, I remember the problem oil companies had getting crews to explore for oil when the price shot up five years ago. Of course we had low unemployment at the time. The industry was in a slump throughout the 90’s due to low prices, then in the early 2000’s prices took off. Congress and the media demanded that oil companies invest their profits in finding more oil.

And oil companies wanted to do just that, but 1) they couldn’t get the labor they needed and 2) they couldn’t get rigs.

All of the experienced rough necks had gone into trucking and house construction. The oil companies could attract a lot of greenhorns, but they aren’t much good for months. They need a lot of training, without which they get killed or kill someone because drilling is very dangerous work. For the first time I know of some oil companies set up a school for roughnecks in Oklahoma City.

And the oil companies needed experienced hands to train the greenhorns, but they had a severe shortage of those, not to mention shortages of engineers. They eventually solved the worker shortage, but it took about 5 years. And a lot of guys are reluctant to go into oil field work because it is so cyclical.

And they couldn’t find rigs at first. Rigs had set in fields and rusted for about a decade so available rigs were useless. Rig manufacturers had to ramp up production, but they faced a similar shortage of experienced workers.

Bill: "But there is unemployment of infrastructure workers...."

I don't see that. And if so, it's very small. The largest groups of unemployed are in housing, Wall Street, and autos, none of which are suited for infrastructure work. They could be trained, but it will take years.


New housing involves infrastructure work--roads, water and sewer pipes, storm sewers.

My student's firm went from $400 million to less than $1 million in 4 years.

Infrastructure work involves paving roads, repairing and replacing storm sewers, and the like.

What do you mean by housing? What do you mean by infrastructure?

Any model that treats labor and/or capital as homogeneous is a completely useless model.

Maybe Daniel believes in the fungibility of language.

@Martin: Keynesian economists do not design any stimulus. That is the job of politicians and lobbyists. Unsurprisingly, politicians tend to be more receptive to political incentives than to the abstract noodlings of any academic Keynesian economist.

@McKinney, thanks for the suggestions!

Bill, usually the types of infrastructure considered in Congress are interstate highways and bridges. That's the foci of the "infrastructure bank."

I don't recall Congressional spending on the infrastructure for housing developments. That's usually a city/county/developer thing.

Anyway, the skills and experience for building a house are very different from those required for road/sewer/bridge building.


The best introduction to modern Austrian capital theory is Peter Lewin's *Capital in Disequilibrium* recently released in Epub by the Mises Institute.

Also Lachmann's *Capital and Its Structure* is accessible and worth reading.

Bill Woolsey said:

"I found the comment that Horwitz quoted irritating. If you go back to the thread, he made other comments complaining about how the U.S. hasn't trained enough engineers. Too many lawyers."

I wasn't taking a shot at lawyers. Feel free to substitute any other job market that has been distorted by bad incentives. The point is that if incentives get out of whack, you get misallocations of resources and labor. This isn't particularly controversial.

"In the section Horwtiz quoted we had at least two uneconomic claims--the no more chairs at any price and the claim that apple can sell all the ipods it wants. At what price?"

Once again: The chair analogy was intended to be a simplified thought experiment. I wasn't making a real-world claim.

When writing on blogs full of economists, is it really necessary to state that of course Apple wouldn't be able to sell all the iPods it wants if it tried to charge $10,000 for them? I would think everyone would be implictly understand the context of the statement. If you'd like me to state it more accurately, how about "Apple manages to sell all the iPads it can make while still maintaining a profit." Or, "There is a market-clearing price for iPads which is profitable for Apple, even at full production."

"I have no doubt that Horwitz understands economics, as does Ebeling. But this engineer (or whatever) doesn't."

I'll let my blog posts speak for themselves. You're welcome to your opinion.

"Still, suppose people choose to live in their parents' basements and drive bigger cars. Or, to pick a more controversial issue, suppose people choose to buy Chinese toys and pay for it by selling lumber to the Chinese construction firms.

Where are the jobs going to come from for the unemployed contruction workers or toy workers?

Why, I guess from the same place the jobs came from for all those unemployed telephone switchboard operators, or all those people displaced off farms by mechanized agriculture, or all those draftsmen and typesetters displaced by desktop publishing and CAD software. That is, unless the government steps in and props up those dead-end jobs with stimulus money.

But when we look at the expansion in the car industry, we see that they didn't mostly hire the unemployed. What would be my first response? Well, that created vacancies elsewhere, which will be filled.

My point is that employment churn is expensive, and that cost isn't captured in Keynesian models, because they assume stimulus money is only affecting idle resources. The minute the stimulus starts crowding out labor or resources, it imposes costs on the economy that have to be examined and captured in the models if you want to get an accurate prediction of the potential for job creation.

In addition, you picked auto workers, which is probably one of the more fungible categories of blue collar work if you restrict it to the assembly line. But if you want to talk about hiring new automotive designers, or top-level project managers, you're now in a realm where it is extremely expensive to the market if you pull people out of existing positions, and where it's not necessarily true that replacements can be found, or if you do find replacements that they will be capable of operating at the same level of efficiency.

So, my other point was that at the beginning of an infrastructure project, the human assets you need are among the least fungible in society, are likely to be already employed, and pulling them from the private economy to work on a stimulus project is a very expensive thing to do.

And if we're talking about a large stimulus project like HSR or wind farms or new bridge projects, those initial phases employing high-valued people can last years, not weeks or months. By the time you're at the point where you can hire large quantities of lower-skilled labor, you'd hope that the recession will either be over or you're in a balance sheet recession or other restructuring recession that will last a long time, and for which Keynesian stimulus is probably not appropriate in the first place.

However, it isn't simply that the employment impact of these government contracts are being held to a different standard, and that specialized human and physical capital is suggested to be so brittle that it would appear to be on the edge of collapse, it is rather the comment that Horwitz quoted seemed to ignore scarcity.

Since you claim to have read the complete thread where that post came from, I'm sure you saw my follow-up post where I said that there's nothing unique about government contracts in this regard - every new opportunity has these costs. You always have to weigh the cost of pulling people off of existing projects to chase new ones, or the risks of delays to projects because you can't find qualified people. But in a market situation, this is acceptable because either the new demand brings more value than the cost, making it profitable for the company AND good for the economy, or it won't get done. When government subsidizes that process with stimulus, it may still be good for the company, but the net value to the economy can be negative.

But more specifically, the point I was making was that if you're trying to calculate the benefits of stimulus or estimate jobs created, you simply can't ignore these costs, because they are significant. The stimulus may still be a good idea if the numbers work out, but you can't put your head in the sand and assume that manipulating the economy in this way has no transactional costs.

Is the U.S. economy really in the state where the typical good or service is like buggy whips?

When you have homes in major cities like Detroit selling for $10,000, and entire neighborhoods in good locations are empty, and there are half-built apartment blocks sitting idle, and a significant portion of homes in the country have underwater mortgages, then yes I'd say that the misallocations are pretty severe.

There are tens of thousands of products in the U.S. economy today, and my vesion of the world doesn't involve "fungability," of resources, but the notion that the demand for nearly all of those things is perfectly inelastic at the current quantity seems unrealistic.

Who said that?

But there is unemployment of infrastructure workers and there is a demand for the sort of infrastructure these guys can produce. I know that for sure.

The fact that you treat 'infrastructure worker' as a monolithic group suggests that you still don't get my point. Just how many sub-specialities do you think there are in that category, and how easy do you think it is for workers to move from one to the other? If your experience is with small-town road crews, you probably have a biased perspective.

In addition, you have to look at the fact that in infrastructure, much of the cost has to go to raw materials. The Keynesian models assume idle resources. But in fact, there are shortages of many of the materials in question. The Chinese are using up ongodly amounts of concrete. Rare earth elements are in short supply. Steel is in short supply. Energy is in short supply. Go have a look at what's happened to steel prices in just the past two years.

Adding infrastructure projects will bid up the price of these scarce resources, which will impose costs on others because of crowding out. Again, that happens with new projects all the time, but in the market case we assume that the value of the new projects exceeds those costs, or they wouldn't be done. In the government case, that's not necessarily true. And in any event, those costs and their effect on employment in the greater economy cannot be ignored. Any policy can be shown to 'create jobs' if you ignore one half of the balance sheet.

@Steve: Thanks, I will start with Lewin's work.

@Fearsome Tycoon: Yes, this was a problem overlooked by Keynes. If I recall correctly, in his letter to Hayek regarding the Road to Serfdom, he argued for more planning on the assumption that it would be done by the people sharing Keynes' and Hayek's moral conviction.

Keynes wasn't entirely wrong though, if you look at Singapore for example, you can see what "planning" can do if it is done by the people sharing their convictions. Singapore though is slightly autocratic by our standards.

Just want to put this out there:

A straight-forward public works program, employing low-skilled workers at a relatively low wage, avoids these problems. Projects to be selected by vote among Chamber of Commerce members. Eh? If we're going to do stimulus anyhow, which kind is the least bad?

Do you have an example of such a program? Unless you're talking about employing armies of people to walk along roadways picking up trash or something, I can't think of any such projects. The closest thing I've seen would be the 'green jobs' initiative in Seattle, which attempted to hire unskilled workers, train them in the application of weatherstripping, and employ them by subsidizing the weatherization of poor people's houses. It was a total bust. They found they couldn't find enough people willing and able to do it, enough homeowners will to take part in the program even with heavy subsidies, and they couldn't train people fast enough. Thousands of homes were supposed to have been weatherized by now, using thousands of employees. As of last month, I think 14 people had been hired and no more than a handful of homes weatherized. Asked why, the people involved said (paraphrasing), "well, it seems really simple to do, but once you get down into the details of it all, it turned out to be a lot harder than we thought."

Details actually matter. Who knew?

As for 'relatively low wage', that's not going to happen, because the Davis-Bacon act requires all federal projects to employ workers at prevailing wages, which are generally taken to be union wages. So you can employ unskilled workers to pick weeds, but you'll still be paying them $20/hr or more. And you'll be generating nothing of value.

If you want to go that route, you'd be better off just dropping money from a helicopter, Milton Friedman style.

Oh Daniel, Daniel. My dearest Daniel. Ever the contrarian, ever the hair-splitter, ever the medium for understanding what your fellow-travelers *really* meant. I haven't read an economics blog for months. The one day I do, there you are...doing the same a T. Over and Over.

I was once told here to stop calling you a contrarian. I'd like to think that phase of unfamiliarity with you has now weakened and passed with the experience of reading your posts.

As for the subject matter, ha, laymen like me in the private sector with a mild understanding of Austrian Economics and Capital Structure were saying this stuff before the stimulus even came out back in early 09. But what do pseudo-armchair-economic-dabbler-ignoramuses like me know anyway?? I should just take the news at face value and take all these technocrats and serious public servants and pundits seriously when they talk about what passes for economic policy and expertise in DC and the major news channels. Right? Sigh.

A little Austrian Econ goes a long, LONG way in understanding and distinguishing Scientism BS from good economic thinking. Quiet observers like me in the private sector get really bored with economics news. Very little ever surprises in the broadest sense....and that's well over half the battle to understanding.

There's plenty I don't know. PLENTY. I just keep the basic notions of Austrian Econ front and center and it's never failed me. Try it, Daniel. Maybe then you can reorganize your vast stock econ knowledge and reprioritize it all while putting some of it to the bottom of the pile where it belongs.

I don't favor infrastructure projects to create jobs.

I'm not a Keynesian. And I am not very worried about estimating how many jobs will be created by added government spending.

However, I don't think unemployment exists because some workers can't produce anything of value, and there are just a few worthy workers who really have everything they want.

We live in a world of scarcity.

If the demand for ipods was three times as high, what would that do to the market clearing price and the capacity that apple is building?

If the market price of ipods is higher, what does that do to the demands for substitute goods?

Of all the thousands of goods in the U.S. today, is it only ipods (or some few goods,) that people value greater amounts?

I don't deny that there is such a thing as adjustment costs. Who denies the existence of structural unemployment?

But when expenditures on output are 14% below trend, imagining that the entire 10% reduction in output below trend is due to structural adjustments seems implausible to me.

Get spending back up to trend, and if output and employment stay the same, then it must be true.

But get spending back up to trend.

Bill Woolsey,

I want to address only one issue that you've raised. Once again, if Rogoff and Reinhart are correct, what you ask is impossible. Balance-sheet, financial crises of the type we have gone through lower the trend rate of growth for as long as 10 years.

Printing money cannot compensate for real changes. If you target NGDP, you will get more inflation and less growth.

If you want to be taken seriously, you must address their findings.

Technically, people receiving checks from unemployment insurance are employed and meet the Keynesian demand for stimulating aggregate demand. As many Keynesians have written, the quality of the work doesn't matter; only the spending matters.

Keynes advocated make-work projects such as building pyramids only because of the aversion to charity that reigned in society at the time. He probably would have seen unemployment insurance as an automatic stabilizer of AD, just as most Keynesians today see it.

I realize I have often complained about people tweaking definitions in order to make an argument work and that's what I'm doing here. I don't really want to define those receiving unemployment benefits as being employed.

But viewing them as employed can shed some light on the success of Keynesian economics. They receive a check from the federal government, so technically they can be viewed as employees of the government. Instead of building pyramids, they are doing whatever work suits them - playing golf, looking for a job, getting more training, etc.

We don't need more make-work projects, such as building bridges to nowhere, to see how Keynesian economics works: just look at unemployment benefits.


What specifically did Rogoff and Reinhart say?

I read "This Time is Different." Is your claim that their statistics prove that balance sheet recessions cause years of lower real growth?

My view is that the reduced real growth in the episodes they describe isn't due to "balance sheet recessions," but rather to net capital outflows.

I don't think that net debt repayment is a "real" factor that causes recession. This is just a version of the paradox of thrift. While it is possible that there could be some kind of adverse productivity shock associated with that, there isn't necessarily such a shock.

And if there is an adverse productivity shock, I don't think that stable growth of NGDP prevents lower real growth. As for more inflation, compared to what? Compared to inflation targeting, it results in more inflation and less reduction in real output (not no reduction in real output.)

While I think that stabilizing the growth path of NGDP is the best policy in response to a net capital outflow, I think that a net capital output will have adverse effects for real growth.

Of course, the U.S. hasn't had a net capital outflow yet and so that isn't relevant. (I think the R&R research project was looking forward to just such an event and the actual problem we had was then a square peg hammered into the round hole)

If a country did have a net capital outflow (all of those foreign investors want to quit lending short and repatriate funds,) I think it is likely that some of the long term projects that were being funded with those funds we need to be abandoned, and so there will be a negative productivity shock. And so, that would result in a higher domestic price level and a reduced volume of output. The exchange rate falls and imports get more expensive. Using a CPI type deflator, the domestic output is worth less. That could last for years. Meanwhile, there is a shift from these investment projects that were supposedly funded with foreign investment to export and import competing goods. So, more structural unemployment. It is like an adverse productivity shock. And finally, the real interest rate is higher. And while equity investors should take a big loss, there should be a shift away from labor to capital going forward, and so nominal wages need to go to a lower growth path.

So, I think a net capital outflow is going to have an adverse impact on real incomes (and real wages) no matter what. While Rogoff and Reinhart didn't point to any episodes where NGDP was kept on a stable growth path, there were lots of episodes where NGDP grew rapidly. And real output was still depressed. While it is likely that some of this was do to going to far, with massive inflation being desruptive in and of itself, I think I have explained all sorts of reasons why a net capital outflow will have adverse effects on real growth for a long time, and so, why NGDP targeting would result in a rising price level for some time.

But a net capital outflow isn't a balance sheet recession.

Deleveraging by households is an increase in the supply of saving.

The natural interest rate falls.

If monetary intitutions are such that market interest rates fail to decrease enough to match the lower natural interest rate, then nominal expenditure shrinks--you have a "balance sheet recession." The problem, however, was the monetary institutions. There is an excess demand for money at the natural interest rate, keeping the market interest rate too high. Printing money solves the problem.

It could be that the reason for the deleveraging was some real productivity shock. That will reduce output. With NGDP targeting, it will result in a higher price level.

But the problem isn't deleveraging. I just don't understand why you appeal to Keynesians like Rogoff and Reinhart. What is wrong with basic loanable funds interest theory?

Further thought from me on this, discussing writing and literature:

dhanson - if you're still reading this - Bill Woolsey is exactly right.

You make claims like this: "My point is that employment churn is expensive, and that cost isn't captured in Keynesian models, because they assume stimulus money is only affecting idle resources."

That are completely wrong. Far from being "not captured" these are an important part of Keynesian models. It's a big part of what the Nobel Prize was awarded for in 2010, and while these concerns aren't exclusively treated by Keynesians, it's a major part of Keynesian modeling. You are simply uninformed. As is Troy when he repeatedly makes this claim that Keynesians argue that capital and labor is homogenous. It's not true. It's great that you two are interested in this stuff, and it's great to throw out ideas, but if someone like Bill Woolsey points out that you're wrong, consider that there might be something to it.

There are things that people can legitimately disagree on, and then there are blatant factual errors. It is a factual error to say that Keynesians don't consider these costs. It's just wrong.

Bill, I didn't get that from Rogoff and Reinhart at all. I think you may be reading your assumptions into it. Of course, I may be, also. It's hard to read the minds of others.

One of the main points of the book was that external debt isn't all that matters, but total debt, domestic plus external. Countries with large external debts sometimes didn't fare badly if they had low internal debt.

Anyway, the problem of net capital outflows could be handled with higher interest rates, as was often done in the past when gold flowed out of a country.

My understanding of the book is that the authors didn't give a reason for why high debt levels reduced growth. They merely showed the correlation. They made some Keynesian style interpretations of the data, but offered no evidence. When I read their book I thought at the time that it would be very easy to reinterpret the data to arrive at Austrian conclusions.

In fact, you have to apply Austrian econ to understand why high internal debt stretches out depressions. People lose wealth in depressions. Stock markets decline, housing values plummet, businesses fail. The greater the debt burden before the depression, the greater is the loss.

People want to rebuild that wealth and that requires saving more. If Keynesians want people to start spending again, they need to first help them save, primarily through higher interest rates. Once they have restored some of their lost wealth they will start spending again.

Restoring lost wealth takes times and no amount of fiscal stimuli or Fed monetary pumping will change that.

Bill: "There is an excess demand for money at the natural interest rate, keeping the market interest rate too high. Printing money solves the problem."

It doesn't solve the problem because you have cause and effect backwards. Demand for money doesn't cause the depression; the depression causes an increase in the demand for money.

Demand for money is a symptom of a much larger problem. Treating the symptom by pumping more money is like giving someone aspirin to reduce the fever caused by the flu.

To be more exact, the loss of wealth caused by malinvestment during the boom is revealed in the falling stock market and housing prices and failing businesses.

Increased uncertainty and the desire to rebuild lost wealth cause people to save more and pay down debt. The fact that increased demand for money and the depression coincide doesn't mean one causes the other. That's a post hoc fallacy.

Part of the empirical problem is using gdp as a measure. It is quarterly data and a lot of activity takes place in quarters that the data hides. Quarterly gdp data reveals the net results at the end of the quarter of many decisions that took place during it.

Someone needs to explain the difference between aggregate and homogeneous, then, because if you aggregate a set of things, you are necessarily treating them as homogeneous. Any model that treats capital or labor in an aggregate fashion is thus treating them as homogeneous, and cannot help but come to wrong conclusions. Even if one then goes on to discuss labor of capital as heterogeneous, you are starting from a model that requires one treat them as homogeneous, since one is aggregating them into a single pile. The problem is that Keynesians cannot tell the difference between a pile and an interactive process. That is the only way one can move from one notion to the other, from treating capital as homogeneous, then treating it as heterogeneous, and not have severe cognitive dissonance.

"discuss labor of capital" -- "of" should be "or"

On Troy's point, it's even more subtle.

Austrians argue that even if you think (in fact, especially if you think...) capital is heterogeneous you can only create a meaningful aggregate if the economy is in equilibrium. Only there can we assume plans are in coordination and therefore that the prices of capital goods can be meaningfully summed. Otherwise we are aggregating using prices that are imbued with error because they may be elements of conflicting plans.


You say:

"We need higher interest rates for people to rebuild their wealth and then they can spend more."

You need to review basic supply and demand. You are focusing on an income effect and are completely ignoring substitution effects.

Lower interest rates reduce saving and expand consumption adn investment.

I don't really care whether increases in money demand are a "cause" or an "effect" of depression. Increases in the quantity of money to meet that additional demand will prevent unnecessary disruption.

All I can say is think about how there is a borrower for every lender.

Think about interest rates as market prices that coordinate (or fail to coordinate.)

And, to develop Steve's point, the economy is never in equilibrium, but is in fact always in a far-from-equilibrium state.

Bill: “You need to review basic supply and demand. You are focusing on an income effect and are completely ignoring substitution effects.”

I don’t see how that applies to my statement at all.

“Lower interest rates reduce saving and expand consumption adn investment.”

Except when they don’t, as in the latest depression. One of the problems with depressions is explaining why lower interest rates don’t reduce saving and expand consumption and investment.

Again, I fail to see how your other statements relate to my posts above at all.

"One must provide the capital goods lacking in those branches which were
unduly neglected in the boom. Wage rates must drop; people must restrict their consumption temporarily until the capital wasted by malinvestment is restored.
Those who dislike these hardships of the readjustment period must abstain in
time from credit expansion"—L. von Mises: Human Action, pp. 575-6.

Who is saving in this recession? Everyone is paying off debt.

The comments to this entry are closed.

Our Books