Monday, August 31, 2009

And It Ain't Shinola...

by the Sandwichman

Wherein the Sandwichman documents the appalling intellectual vacuity of Lee E. Ohanian's "What -- or Who -- Started the Great Depression?" so the referees at the Journal of Economic Theory don't have to. Not that they would anyway.

First, let's cut to the chase. Ohanian summarizes his operative theory on pages 48 and 49. I will quote the section in full at the bottom of this post. For now, the two key sentences are "Any monetary explanation of the Depression requires a theory of a very large and very protracted monetary non-neutrality.... The non-neutrality is quantitatively large in the Hoover economy because Hoover's wage maintenance and work-sharing program reduces steady state hours and capital stocks [emphasis added]."

So what is the mechanism by which Hoover's program reduced capital stocks? For the answer to that, we take you back to pages 28-29 to an explanatory footnote about Ohanian's specified model. Again, I will present the full footnote at the bottom, but the key phrases are: "Capital input in this model is variable, and is equal to the capital stock scaled by the length of the workweek, or hours per worker." and "This treatment is also reasonable because there is evidence that worksharing that reduces the number of days an employee works, even keeping the length of the workweek fixed, also reduces output per hour."

The evidence Ohanian refers to is a 2001 article by Lanoie, Raymond and Shearer (hereafter Lanoie), "Work sharing and productivity: Evidence from firm level data." It is fair to point out that Lanoie is the only article in the paper's reference list dealing with the productivity effects of work-sharing. The article in question deals with a single, year-long "experiment" at a Canadian telecommunications firm in the early 1990s.

On first impression, the relevance to Hoover and the Great Depression of this single 1990s example may seem remote. But on closer inspection, it becomes laughable. The program was not simply a work-sharing arrangement. It was also a change in schedule to a nine-hour day precipitously imposed by management. Furthermore, the productivity impacts were found to be task specific. To generalize from this particular example to work sharing during the Hoover administration strains credulity, let alone plausibility. But here, in the authors' own words, are the peculiarities of the Canadian experiment that would be enough to disqualify it as in any way typical or representative of the Hoover-era experience:
These results suggest that the impact of work sharing on productivity is 'task specific' and that longer operations (both types of installations), for which the coordination cost is likely to be higher, are broadly more affected. As discussed earlier, another possible contributing factor to the decrease in productivity is the change in the work schedule that was introduced along with the work sharing programme. Namely that workers changed from working 8 hours a day for 5 days a week to working 9 hours a day for 4 days a week. It is possible that the extra hour tacked on to the end of the day was much less productive than the hours worked on the fifth day of the week. Unfortunately, without information on daily production, the data set does not permit identification of such effects. Certain officials also mentioned that managers were not well prepared to operate in this work sharing environment (the whole operation was implemented with a very short notice), and that coordination problems occurred not only between technicians, but also between technicians and dispatchers. One further possibility is that worker morale may have been negatively affected by the work sharing programme. Given that technicians were not given a choice of whether or not to participate in the programme whereas other types of workers were, technicians may have felt they were being unfairly treated (Akerlof, 1982). The fact that absenteeism increased following the introduction of the programme lends support to this interpretation.
To reiterate:
  • impact on productivity was task specific;
  • change in the work schedule was introduced along with work sharing; from 8 hours a day for 5 days a week to 9 hours a day for 4 days a week;
  • managers were not well prepared to operate in this work sharing environment (the whole operation was implemented with a very short notice;
  • technicians were not given a choice to participate (other employees were); morale may have been negatively affected.
What is absent from Lee Ohanian's article is any explanation of why he thinks the above 'evidence' is remotely relevant to his Great Depression theory. What does this say about the peer review process at the Journal of Economic Theory? Perhaps they should take a lesson from Navin Johnson's father:

Summary of Ohanian's theory, pages 48-49:
Any monetary explanation of the Depression requires a theory of a very large and very protracted monetary non-neutrality. Such a theory has been elusive because the Depression is so much larger than any other downturn, and because explaining the persistence of such a large non-neutrality requires in turn a theory for why the normal economic forces that ultimately undo monetary non-neutrality were grossly absent in this episode. That is, if the Depression is largely the result of monetary forces, then the size and the duration of the monetary non-neutrality were remarkably well outside estimates from any other period.

This paper provides such a theory for a large and protracted monetary non-neutrality. The non-neutrality is quantitatively large in the Hoover economy because Hoover's wage maintenance and work-sharing program reduces steady state hours and capital stocks. The non-neutrality persists in this model because it is a transition from a non-distorted steady state to the Hoover distorted steady state.
Footnote, page 28-29:
Capital input in this model is variable, and is equal to the capital stock scaled by the length of the workweek, or hours per worker. In the model, utilization falls in manufacturing, which is consistent with actual manufacturing utilization during the Depression. It is worth pointing out two issues about tieing [sic] the decline in utilization to hours per worker. One is that some of the decline in utilization was due to plant closings, rather than a shorter workweek across all plants. Another is that some worksharing was such that workers were employed for fewer days, but the plant could have had the same workweek length. I am unaware of data that can provide any type of detail on these distinctions, however, so I will treat the model as a parsimonious tool for capturing low capital input during the Depression, as it will allow the model to be consistent with actual manufacturing output per hour. This treatment is also reasonable because there is evidence that worksharing that reduces the number of days an employee works, even keeping the length of the workweek fixed, also reduces output per hour (see Lanoie, Raymond, and Shearer).


Caroline Abbott said...


Note: Shinola is no longer being produced. Not the case, obviously, with the paired substance.

Bernadette said...


Anonymous said...

He definitely made a mistake with "Such a theory has been elusive because the Depression is so much larger than any other downturn...", when it should have read "Such a theory has been elusive because the Depression is so much larger than any other downturn before now..." Because this recession is not incorporating innovative work sharing solutions it has the capacity to dwarf anything seen before. Most likely some future scholar of questionable intellect will claim it was not a two front war, sub-prime mortgages, deregulation, credit default swaps, or just plain old greed that caused the depression of the late 2000's. Instead in his questionable wisdom he will point to health-care reform, infrastructure improvement, and any possible work sharing that is to arise as the "real" problems. Do these guys get paid to teach? We can see why the university trained economists are doing such a good job.

marycolleen said...

In reference to people who just didn't get it, my mother used to say "he doesn't know shit from shinola". I hadn't seen this clip before -- it's great!

John S. Cline said...

Yes! Thanks for this! This guy is with NBER? Dean Baker, send this guy back to the Cato Institute where he belongs. Next he'll be saying we all need to deregulate banking and ... oh, we did that, guess that didn't work out so well either.

Anonymous said...

Maybe Mr. Ohanian's been practicing Canadian-style work sharing. It would seem that his 9th hour productivity is in the toilet.

Anonymous said...


Anonymous said...


Anonymous said...

Doyle Saylor

Martin Langeland said...

Ohanian is a prime candidate to experience the tender mercies of a netless meritocratic free market -- which is to say he is a strong argument against tenure.

A better informed reader said...

Dear Sir:

Your comments on Ohanian's paper only reveal your uttermost lack of familiarity with modern macroeconomics.

You are confusing a reduction in capital stock (the point made by Ohanian) with a reduction in capital services input. Capital services inputs are, in the model, equal to capital stock scaled by workweek. In a model such as this one, the increase in market power reduces capital stocks in the new balanced growth path because we have higher mark-ups that work in the same way that a tax on inputs.
Even in the absence of varying capital utilization, we will have exactly the same reduction that Ohanian mentions but only slightly smaller in quantitative terms.
All this is well know to first year graduate students if they want to pass their quals and it is not really even worthwhile to mention it. Obviously, neither you nor the previous readers seem to be able to understand this. Instead you offer a egregiously childish reading of the footnote.

So, please, before throwing darts in the future, do your homework.

Sandwichman said...

Nice try, "better-informed"-reader,

How on earth do you come up with the inference that the Sandwichman confuses capital inputs with capital stock? Just makin' shit up, eh? Get cornered? Throw mud!

Funny thing is, there was a zinger I could have used if I wasn't concerned to be scrupulous about the difference between stocks and inputs.

Lanoie et al. themselves appear to (possibly) confuse the two and I was tempted to quote them in "contradiction" to Ohanian's argument. But I could only have done that by either 1. pulling a fast one on my readers (assuming they wouldn't notice the difference) or 2. going into a long, boring explanatory digression that would have been pedantic and mystifying.

But since you brought the matter up (you brave anonymous know-it-all, you), here's the cite from Lanoie et. al., that I didn't use specifically because I didn't confuse capital stocks and capital utilization:

"It is implicitly assumed that productivity is not a function of the capital stock (building, equipment, material, etc.) since discussions with the company's officials showed that, in the context of this study, it was constant throughout the period." (p. 1213)

So in the above quote do they mean capital stock was constant or capital input was constant? Gosh, you're the "better informed" one. Maybe you can tell me.

Meanwhile, I'll leave you with a fine comment from Brad DeLong's blog in response to an earlier discussion of an equally funky Ohanian fantasy:

"One of the central methods of economics as a discipline involves assuming your conclusion. The task of the theorist is then to conceal the conclusion somewhere in the paper where reviewers will either miss it or fail to understand its significance - kind of like the 'Find 6 Differences' comics that appear in your local newspaper. It makes logic easy, and fun!"

The commentator, "memory", left out the part about when your reviewer does find how you've assumed the conclusion, accuse him of "uttermost lack of familiarity with modern macroeconomics." I'm sure Bernie Madoff accused his critics of uttermost lack of familiarity with modern finance. You don't have to be an expert in modern macroeconomics to smell a bald-faced fraud.

But please, please stick around, "better-informed-reader"! I want to get your views on my response to the part II declaration by Lanoie, et al. (Ohanian's only source on the key issue of work-sharing) that "little empirical work has been done to measure the consequences of work sharing."

Me again said...

You still did not answer the main point of my comment: in a model such as Ohanian's, a raise in market power reduces capital stock in the new steady state EVEN when workweek and TFP are constant. Having a changing workweek only increases the effect a little bit quantitatively.
You may not like the model: you may think that their assumptions are unrealistic or just plain wrong and silly, you may think is not the right model to think about the issues. All those are valid criticism. What you cannot do is claim as you do in your post that capital stock only fails because of the footnote you refer to. This is just not true and it denotes a deep misunderstanding of how a modern DSGE model works.
Again, we are just talking about a property of the model.
Let's stick to the model

Just some code said...

By the way, in case you are interested in checking this out:

Run the following code in Dynare.

It is a simple growth model with monopolistic competition, where market power is controlled by 1/epsilon. If you reduce epsilon from 10 (a mark-upo f around 10%) to say, 8 (a mark-up of around 14%), capital in the steady state falls from 8.00425 to 7.48675, a reduction of around 6.5%. Again, note that in this simple model none of your concerns are present, and yet, the main result still goes through

% 1. Defining variables

var y c k i l y_l w r z;
varexo e;

parameters beta psi delta alpha rho gamma sigma epsilon;

% 2. Calibration

alpha = 0.33;
beta = 0.99;
delta = 0.023;
psi = 1.75;
rho = 0.95;
sigma = (0.007/(1-alpha));
epsilon = 10;

% 3. Model

(1/c) = beta*(1/c(+1))*(1+r(+1)-delta);
psi*c/(1-l) = w;
c+i = y;
y = (k(-1)^alpha)*(exp(z)*l)^(1-alpha);
w = y*((epsilon-1)/epsilon)*(1-alpha)/l;
r = y*((epsilon-1)/epsilon)*alpha/k(-1);
i = k-(1-delta)*k(-1);
y_l = y/l;
z = rho*z(-1)+e;

% 4. Computation

k = 9;
c = 0.76;
l = 0.3;
w = 2.07;
r = 0.03;
z = 0;
e = 0;

var e = sigma^2;


Sandwichman said...

"Me again" said,

"let's stick to the model"

Your point, then, is that what Ohanian says in ordinary prose doesn't count? Here, again is what he said, "because Hoover's wage maintenance and work-sharing program reduces steady state hours and capital stocks." What part of "because" am I misinterpreting?

Model... shmodel. If it is true that the model doesn't show what Ohanian claims it shows... then the model is irrelevant. Null and void. The model itself may be a piece of cake. Has nothing to do, though, with Ohanian's argument and the press release trumpeting his conclusions.

You say that objecting to the model's assumptions as unrealistic, just plain wrong and silly and to the model as not the right one to think about the issues are all valid criticisms. Thank you very much. That is what I thought I was doing in the post. I didn't say that "capital stock only fails because of the footnote." Professor Ohanian said that (see the two extended quotes above). Go argue with him.

But, please, please, please, stick around for part two. I want to see how you riddle your way out of this conundrum:

"...little empirical work has been done to measure the consequences of work sharing." -- Lanoie, Raymond and Shearer, "Work sharing and productivity: Evidence
from firm level data." 2001.

"...a vast mass of [empirical] research has been undertaken into how the relationship among effort, efficiency and time manifests itself within the production process." - Chris Nyland, Reduced worktime and the management of production. 1989.

The Big Mook said...

Question for "just some code"?

There are at least 3 versions of Dynare: Gauss, Scilab, and Matlab.

since you did not specify the version you used, I wonder if using one of the other versions, the result will be steady shinola or the variable shinola squared to the omega factor of pi?

But to digress back to Sandwichman's post: It appears that Ohanian cited evidence to support his theory that clearly does not.

Factors that were also suspect to result in lowered productivity, were cited by Lanoie, et al to be the increased work day: 8 to 9 hours, a lack of preparation for the schedule change for either workers or managers.

Ohanian cited "evidence" that does not support his conclusion. Without real evidence or a plausible explanation, I conclude that Ohanian was simply offering his unsubstantiated opinion and masquerading it as fact.

Tom, are you implying that "a vast mass" can not be made to mean the same thing as "little"

All a professor has to do is apply a little* "E.I.T." ( Enhanced Interpretation Techniques )

Arguably more (or less) humane than the other EIT (Enhanced Interrogation Techniques)
we only torture language.

*or a vast mass of EIT.

callitwhatitisAnonymous said...

Lets stick to the model. I love it. Lets stick to my model and not get involved in all those pesky facts. You will soon see that my model supports my model if you use my math. Don't go back to the original data though that always leads to confusion. This guys hilarious.

Jack said...

What concerns me more so than Ohanian's apparently flawed analysis and resulting erroneous conclusions is that he can hide his peculiar professional product flaws behind a weighty set of titles. He holds a full professorship at UCLA and is also the "Director" of a privately funded sub-chapter of UCLA's Economics Dept,the Robert Ettinger
Family Program in MacroEconomic Research. Pretty impressive, sounding that is, and how is the public and the press to know any better? This is something akin to taking legal advise from John Yoo regarding the interrogation of prisoners because he holds a professorial position at Berkley. How does it happen?

I would usually suggest that one follows the money when it comes to peculiar research findings and analysis. The same holds true in regards to faculty positions or "scholar/associate" positions at a so-called think tank. ChicagoBooth is a good example of money in academia. Check out the Council on Chicago Booth. It reads like a who's who of corporate America with a significant investment banker representation. The express purpose of such a Council, as stated on their site is to "advise Booth's leadership in their efforts to maintain faculty excellence." The Ettinger Family Program in Macroeconomic Research is not so illuminating in regards to influence on faculty selection. said...

The Ohanian paper has a number of serious problems with it, starting with its discussion of historical facts. Yes, Hoover made some of the moves described, but they had little impact. Nominal wages fell, if not as much as did prices, but then prices fell much further in the supposedly wonderful 1920-21 recession. Unions were almost non-existent and virtually powerless, until after FDR's Wagner Act in 1935. He is conjuring a ludicrous boogeyman.

Furthermore, if this sort of thing was the source of the problems, why did the economy turn around when FDR came in who gave much more power to unions and engaged in far more '"monopolistic" practices than did Hoover? This whole shell game of trying to "Rooseveltize" Hoover is hilarious.

Regarding the model for Mr. Better Informed, or whatever he is calling himself, its biggest problem is assuming equilibrium, indeed, rational expectations equilibrium, in the midst of a massive disequilibrium situation. He also does not model that nominal debts were fixed, meaning their real burden rose as nominal wages fell, something even Milton Friedman pointed out as being a problem of the period.

All in all, the paper is a piece of trash that has gotten way too much tub thumping by various people who have not thought it through very carefully or do not know what they are talking about.

Jack said...

So it seems to be generally agreed to by a group of well versed economists that there is, in their profession, a source of disinformation so egregious as to generate suspiscion regarding the intent of the author. Ohanian is certainly not the only source of such disinformation, but his effort seems to be outstanding in its lack of scientific rigor and his simple ignorance of basic historical facts.

What's going on? This guy is not a quack. He holds a PhD from a respectable university, Rochester.
He's listed as a consultant on the Minneapolis Federal Reserve Bank web site. He's a professor at UCLA, no podunk itself. So where's the filter mechanism for Economics as a field of science? How do I, a layman with an interest in the results, know woh is worth listening to? Better yet,
how does the media know how to judge if they had an intention to judge the quality of what they're being told? Frankly, given the influence of the field on our economy, and as a result on our lives, this seems unacceptable. The greatest danger to our democracy is the lack of candor in our media. With "professionals" like Ohanian available for comment any lie can be established as a reasonable assumption, nearly a fact. Really an unacceptable circumstance.