Be afraid, be very afraid…again. From Joseph Lawler
… I take yesterday’s CBO report as affirmation both that … the number of jobs created by the stimulus cannot be determined without making judgment calls about the underlying economic model …
As far as I can tell, the logical conclusion of Mr. Lawler’s article is that since all models forecast poorly, we should not use any models to make any policy conclusions, regarding output, regarding employment.
After reading the CBO report, here is Mr. Lawler’s most telling indictment:
the CBO’s methods include some sweeping assumptions (which are typical in standard Keynesian models) about the economy, and anyone trumpeting the CBO’s findings should include qualifiers about those assumptions.
Here’s how the CBO comes up with the employment numbers. Instead of trying to count the reported jobs and extrapolate from those numbers, they look at the total amounts spent by the government so far, and then estimate the effects of those expenditures using evidence from similar policies in the past and sophisticated macroeconomic models.
At this juncture, I laughed and laughed and laughed. Aside from the fact that Mr. Lawler neglected tax cuts and rebates in the stimulus package, the second paragraph describes how all economic analyses of the stimulus — RBC, New Keynesian DSGEs, monetarist, and Keynesian — would proceed.
By the way, as far as I can recall, nonstructural VARs generally forecast output quite well. That would not necessarily mean that nonstructural VARs are useful for conducting policy experiments.
Politico might as well be a tabloid; I don’t think anyone can take their reporting on the economic situation seriously.
That is all fine,could we expand the spectrum to beware:
The economists talking their employers books.
The economists talking politics.
The women of small virtues willing to serve the mess wine.
Menzie wrote:
As far as I can tell, the logical conclusion of Mr. Lawler’s article is that since all models forecast poorly, we should not use any models to make any policy conclusions, regarding output, regarding employment.
Menzie, this is great. Are you serious? Has the light gone off? Do you finally understand that policy based on “models [that] forecast poorly” result in poor policy? (Tongue firmly planted in cheek)
Is this conculsion really rocket science.
You imply that the only options are poor policies based on poor models or nothing. Surprise, the market has been resolving these problems for centuries. If the government had stayed out we would have been, at worst, in the depression of 1921 and it would all be over by now.
But because government hubris rules, we are facing the depression of 1929. Remember 1933 was 4 years after 1929, 2012 will be 4 years after 2008.
Just for the record unemployment in 1930, one year after the stock market crash, was 8.7%. That is equal to U6 unemployment of 17.5%, but imagine even the altered U2 is higher than 1930 unemployment at 10.2%.
While my first impression was, “here we go again”, I’m actually really glad you picked this windmill to tilt at.
That quote is EXACTLY the sentiment I’m seeing across the blogosphere. It’s an indictment of macroeconomics in general, not the CBO report in particular.
I guess the question I have is, how much of the economic deterioration that occured between the passing of the stimulus until now was a rational reaction to the stimulus itself and other Obamanomic policies? Pretty much by definition, no macroeconomic model is going to be able to capture that kind of… human feedback into the system.
It’s kind of like the reaction to the Lehman bankruptcy. We had an extremely mild recession going up to that point and then a step change to a very severe recession. No macroeconomic model done the day before the Lehman bankruptcy was going to forsee that change.
Ah, seems RicardoZ has found a soulmate. His starting point is that markets are right, policy wrong. Lawler finds a half-a…, sorry, half-baked way to argue that policy is ill-founded, so RicardoZ declares victory.
OK, if the point is that basing policy on models that may not predict well can lead to trouble, I understand the point. Does RicardoZ not see, though, that arguing for a policy based on no evidence is just an extreme case of arguing based on flimsy evidence? Because this is where RicardoZ leaves us. Markets would have fixed everything. No evidence for this statement, but that’s the funny part. (Maybe Menzie laughed and laughed again.)
OK, I know we’ve all read this before, but RicardoZ seems to have been staring out the window in Econ 101 that day – there is no long history of free-market successes because we don’t generally get to try pure capitalism. We only see the results of mixed economies. Some mixes work better than others. In the absence of any evidence of the wonders of free-market capitalism, we can safely assume anyone who crows about the superiority of free markets in determining economic outcomes is happy operating evidence-free.
kharris,
You would do well to understand the depression of 1921.
There is a huge amount of evidence that the market does adjust the errors of government, and dare I say economists. Where there is no evidence is that “stimulus” has ever worked. The best evidence against “stimulus” working is happening right now.
For Menzie’s main point- Yeah, why would we try to use science and numbers to forecast what might happen. Because, you know, things might change, so why try to figure it out? Yeesh, but then again, it is Politico, not a source or readership that cares to use fact much.
Kharris- We do have Ricky Z’s free-market utopias in the real world. We call it Saipan or Somalia. I expect Ricky to be quitting his current dead-end job and be heading there shortly for their freedom.
Anyone who thinks corporates fix markets for the common good is too naive to be taken seriously. Ricky neglects to mention that 1919-1921 was a 3-year depression that only ended due to a ridiculous asset bubble and a massive growth in inequality in the ’20s (sound familiar?). And pray tell, how did that great bubble of prosperity end for us? Oh, that’s right, the market caused a huge increase in U.S. demand from 1930-1933, right?
Then again, Ricky’s been adamant that the actions of the previous years have nothing to do with the state of the current economy or fiscal policies. With the exceptions of 1982 and 1994-2000 of course.
Dr. Chinn:
Response here: http://bit.ly/6aBLoJ
Ppcm identifies the worst prevaricators on the planet : economists talking their employers books. Further I am sick to death of hearing about The Beard and his expertise on the Great Depression. How about some analysis from someone who does not trumpet their mastery of that miserable period and portend great things gleaned from it. Did the Fed back then pump liquidity into the top 19 when we are told that the Fed was miserly in the extreme with liquidity in general ? This is obviously a specialist site and I am out of my depth but at least there seems to be real intellectual turmoil here and very little beard preening.
J. Miller wrote:
Anyone who thinks corporates fix markets for the common good is too naive to be taken seriously.
I totally agree. When Hoover/FDR adopted Italian corporatism they gave us the Great Depression. When Nixon adopted similar Fascism by taking the US off of a currency standard and instituting wage and price controls other government interventions he gave us the Great Inflation. Now the Bush/Obama tag team has taken us into the Great Stagnation which will last longer than any stimulus- based Keynesian can ever imagine.
Our problem is the economists have been led away from what real economics should be about. Economics is about the interaction of two traders who exchange for the benefit of both of them. Economists of today have been seduced into believing that favoring one trader over another improves the trade. In fact it taints and diminishes the trade for both and both must expend effort and capital to protect themselves from this favoritism.
My call is for a return to freedom and trade where effort and resources are used to the betterment of mankind. Hopefully you will understand this and join in the only real means of recovery, the reestablishment of freedom, personal and economic.
Thomas
May I suggest the reading of these studies as they outline the big difference between today s crisis and 1929-1934
In summary the USA were net exporter of capital,oil,and manufacturing was a major component of the GDP.
It seems that the G7 have borrowed quite considerably from long term growth at the expenses of fundamental macro balances.
The collusion between the financial world,the central banks,states governments has found the perpetual movement.
Whilst reading the hereunder studies no mention is made of the financial sphere.
S Kuznets
http://papers.nber.org/chapters/c2257.pdf
National Bureau
of Economic Research
BULLETIN 50
APRIL 18, 1934
Production in Depression and Recovery
Copyright 1935. National Bureau of Economic Research, Inc. A. BLISS
A NON.PROFIT MEMBERSHiP CORPORATION FOR IMPARTIAL STUDiES IN ECONOMIC AND SOCIAL
BROADWAY, NEW YORK
Recent Corporate Profits in the United States
Copyright 1934, National Bureau of Economic Re.rcarch, Inc.
SOLOMON FABRICANT
Joseph Lawler: Thanks. Those are structural VARs recounted in the historical section. There are a variety of ways in which to impose structure — restrictions on the matrices in the structural representation of the system, which either restrict short run (impact) effects or in the long run. Faust et al. critique both types of restrictions, if you are interested. In addition, restriction can be imposed by way of a two step procedure in which case one of the variables involved is purged of some sort of endogenous component. In any case, nonstructural VARs are the ones which I asserted were usually quite successful in forecasting — but much less so for policy analysis, exactly because it is only under some usually implausible cases that the Cholesky decomposition yields residuals equivalent to structural errors.
Regarding your over-arching critique of the CBO approach, perhaps an analogy would be helpful in illuminating which approaches make sense, and which don’t. I give a patient with a fever aspirin. However, the fever continues to rise. I could conclude that aspirin caused the fever to rise relative to what would have occurred otherwise. Or I could use information regarding the effect of aspirin on fevers, obtained from previous experiments/experiences, and use that information to infer what the fever would have been in the absence of dosing with aspirin. How would you conduct inference in this case?
Dr. Chinn,
Thanks for the clarification regarding structural vs. nonstructural VARS. That had confused me as I’d noticed that the studies the CBO referenced used structural VARs.
Your analogy is well taken.
But I think you are still missing my point. I don’t have a critique of the CBO’s approach. As far as I can tell, it is doing a great job. Nor do I think that the stimulus has made the economic situation worse, by the way.
Maybe extending your analogy could illustrate the point I’ve been trying to make.
Imagine that a second doctor warns that the first doctor that the evidence available is actually inconclusive, and that the prescription of aspirin won’t cure the fever. After the dose, the fever continues to rise. But the first doctor advertises to the world, on the basis of the same earlier studies, that he has successfully treated the patient, without bothering to mention that he has no new evidence to support that claim. I would not think that the second doctor would have any reason to be newly persuaded that the aspirin worked in this case or in general.
Note that the first doctor in this analogy is the administration and left-leaning groups, not the CBO. Also, this is a separate criticism from the one I leveled at the administration for touting the recovery.gov numbers.