Wednesday, September 01, 2010

Assuming Your Conclusions

A recent Washington Post news story describing a speech by Christina Romer, the outgoing chairman of the Council of Economic Advisors, provides a nice example of how to implicitly assume your conclusion. The author writes:
At week's end, Romer will leave the council chairmanship after what surely has been the most dismal tenure anybody in that post has had: a loss of nearly 4 million jobs in a year and a half. That's not Romer's fault; the financial collapse occurred before she, and Obama, took office. But she was the president's top economist during a time when the administration consistently underestimated the depth of the economy's troubles - miscalculations that have caused Americans to lose faith in the president and the Democrats.
The implication, reinforced elsewhere in the piece, is that the Democrats did the right thing, just not enough of it. Their only mistake was not making the stimulus even bigger. If only they had spent even more and taxed even less, the economy wouldn't be in such miserable shape.

That is one possible interpretation of what happened. Another is that Romer is like a medieval physician explaining to the grieving relatives that if only he had bled the patient a few more times, he would have recovered. One way to judge a scientific theory is by comparing what it predicts to what happens. One of Romer's predictions was that unemployment would peak at 8%. It didn't. I do not know whether Romer has considered the possibility that the reason her prediction was wrong was that her theory was wrong, but pretty clearly the author of the piece has not.

He merely takes for granted the Keynesian orthodoxy of fifty years ago, according to which the right way of dealing with a recession is to run a deficit, and the worse the recession the bigger the deficit should be. The administration followed that prescription, things didn't get better, so obviously the deficit was not big enough. The alternative possibility, that the reason things didn't get better might be that they were following the wrong policy, simply didn't occur to him.

19 Comments:

At 9:18 PM, September 01, 2010, Anonymous Anonymous said...

It's almost as if you aren't at all aware of how Sweden dealt with it's economic crisis during the 1990s.

 
At 9:54 PM, September 01, 2010, Blogger David Friedman said...

I am unaware of how Sweden dealt with its past economic crises--or quite a lot of other countries.

And I am guessing that you are unaware of how the U.S. dealt with its economic crisis at the beginning of the 1920's.

 
At 10:11 PM, September 01, 2010, Anonymous Anonymous said...

Before spending significant effort in critiquing modern governmental responses to economic crises, allow me to suggest you at least familiarize yourself with how modern governments respond to economic crises, as well as the outcomes of such responses.

 
At 10:48 PM, September 01, 2010, Anonymous Anonymous said...

Consider Japan Mr. Intellectual Anonymous who might know less than he/she thinks-- over the last 20 years since it's 1989-90 economic crisis 20 years of multiple mega stimulus programs and heavy government 'action' a la Obama. Voila: A huge government debt to show for it (2x GDP) and persistent economic stagnation. Since this blog is about assumptions, you probably assume that when private sector funds are commandeered by government somehow economic growth is assured, or that central bank money-printing actually creates wealth ...happy hunting for the truth!

 
At 11:03 PM, September 01, 2010, Anonymous Anonymous said...

One of the primary problems with Japan was that the government was propping up insolvent banks (in comparison to the Swedish approach). This froze capital in an insolvency trap as all capital had to be retained instead of lent out. But you're certainly correct, their's is a lesson to be learned.

 
At 12:27 AM, September 02, 2010, OpenID gargunzola said...

Seems to me that stimulus always does work to bail out an economy, however it creates another bubble that is sure to burst. Which in turn would require more stimulus. Stimulus/Keynesian policies are popular with politicians because it legitimises government in the eyes of the public. It also gives government a blank cheque to do whatever is necessary - always temporary - to expand its power in the name of fixing the economy. By contrast, refusing stimulus and bailouts will not avoid economic disasters but it will not exasperate them. It's like going through a hangover and never drinking again (the Austrian approach) instead of carrying on drinking more every day as you build up resistance (the Keynesian approach).

 
At 1:54 AM, September 02, 2010, Anonymous Kid said...

The logic of politics predict that governments will tend to follow policies like the ones gargunzola suggest: create another bubble to "solve" the current one bursting.

What I don't understand is why the market hasn't caught on. If this keeps happening, shouldn't the market adjust to it? Are they forced to play along by some mechanism?

 
At 2:17 AM, September 02, 2010, Anonymous Anonymous said...

The problem I have with the "Keynesian" analysis is that it assumes the economy can and should return to its pre-crash state...if only the right government policy is implemented.

The "Austrians" (though I don't fully agree with them) correctly emphasize that unsustainable booms cause real damage that can't be undone. The economy can't instantly reorganize itself, so elevated unemployment is unavoidable. It's not a technical glitch of the banking system or money supply or whatever, it's a fundamental problem.

 
At 5:51 AM, September 02, 2010, Blogger sconzey said...

Coming at this from the other side, I'm terrified that school reform over here will fail -- the Coalition government have said they want to introduce Swedish-style free schools, but have ruled out profit-making schools, vouchers, and still want a good deal of centralised control over teaching styles and curricula.

Then if it fails, I suppose I'd sound just like Romer, saying it "Wasn't free enough!"

 
At 6:24 AM, September 02, 2010, Anonymous The Money Demand Blog said...

Christina Romer's previous work has shown that the monetary policy is the most powerful tool for macroeconomic stabilization. Her research shows that monetary policy was the main cause for recovery that started in 1933. It is very unlikely that Romer's undersanding of our problems should be compared to medieval quackery.
A quick look at the chart of nominal GDP should convince you that monetary policy has not delivered sufficient aggregate demand. In early 2009 Bernanke supported fiscal stimulus because he knew that he could not push through sufficient monetary stimulus.
There are very serious issues with fiscal policy via higher spending (crowding out, public choice etc.).
However I don't get why lower taxes are not the part of the solution until aggregate demand has returned to the proper level.

 
At 9:29 AM, September 02, 2010, Blogger David Friedman said...

"Christina Romer's previous work has shown that the monetary policy is the most powerful tool for macroeconomic stabilization. "

I'm sure she has evidence for that thesis, but "has shown that" is, I think, too strong a claim for most theories, including theories in macro. Her quoted comments in the article make clear that things did not happen as she expected and she doesn't have any clear idea of why—which is some reason to think that her views might be mistaken.

I've commented before on the Great Depression that didn't happen—the one at the beginning of the twenties. Harding did what Hoover is popularly (and mistakenly) thought to have done—responded to a sharp drop in prices and rise in unemployment, comparable to the figures at the beginning of the Great Depression, by sharply cutting government expenditure. In a year or two it was all over, with employment back at a normal level.

http://daviddfriedman.blogspot.com/2009/05/two-great-depressions.html

One possible explanation of the Great Depression that did happen—I'm not nearly expert enough to offer an opinion on whether it is the correct one—is that FDR greatly prolonged it by making lots of big changes in what government did, introducing a high level of uncertainty, which made it hard for private actors to adjust. One could view the past two years as a similar pattern on a smaller scale—in which case Obama's policies, including bailouts (continued from the previous administration), "stimulus," Health Care changes, proposed carbon tax, ... are the cause, not the cure, of present problems.

If you don't know what the rules of the game are going to be next year, there's much to be said for waiting to make investments—or, for that matter, long term employment decisions—until you do.

 
At 9:51 AM, September 02, 2010, Blogger Garg the Unzola said...

If 'recovery' started in 1933 then when did it end?

 
At 11:36 AM, September 02, 2010, Anonymous Anonymous said...

"I'm sure she has evidence for that thesis, but "has shown that" is, I think, too strong a claim for most theories, including theories in macro."

I suspect you might have a better understanding of her research if you actually read it instead of picking semantic nits.

 
At 1:38 PM, September 02, 2010, Anonymous The Money Demand Blog said...

"I'm sure she has evidence for that thesis, but "has shown that" is, I think, too strong a claim for most theories, including theories in macro. Her quoted comments in the article make clear that things did not happen as she expected and she doesn't have any clear idea of why—which is some reason to think that her views might be mistaken."
Some macro theories should be uncontroversial. Economic impact of unexpected AD shock should generate no controversy since the time we got the theories of Lucas and Phelps.

The output of macro model should be a probability distribution of outcomes, not a point forecast, and Romer's political naivety here gives us no useful information about the degree to which her macro theory views might be mistaken.

"I've commented before on the Great Depression that didn't happen—the one at the beginning of the twenties. Harding did what Hoover is popularly (and mistakenly) thought to have done—responded to a sharp drop in prices and rise in unemployment, comparable to the figures at the beginning of the Great Depression, by sharply cutting government expenditure. In a year or two it was all over, with employment back at a normal level."

Harding's policy should be studied by Greece, Ireland and Portugal if they would like to avoid a lost decade. It is less applicable to those large countries where both resource utilization and inflation gaps still remain:
1. Harding had gold standard, and fiscal cuts have reduced the risk of gold outflows, whereas there is no problem with sovereign risk premiums in the USA or in France.
2. The economy is much less flexible now because of the prevalence of inflexible debt contracts that were signed before the crisis under the assumption that the Fed continue to achieve the inflation targets.
3. Monetary stimulus could restore the normal level of AD and this is the easiest way the efficiency of labor markets could be improved.

FDR's policy has increased uncertainty (risk of socialism) and reduced the flexibility of markets (industrial and labor market regulations etc.). These policies are responsible for 20% of Great Depression, 80% was caused by bad monetary policy.

Obama. The current level of investment is consistent with the current level of resource utilization. Bad policies you mentioned will certainly reduce growth going forward, but so far they are invisible in the data.

 
At 3:34 PM, September 02, 2010, Anonymous Anonymous said...

Another way to test your hypothesis would be to examine those years in which the government ran big deficits and see if they were coincident with high unemployment or sub-par returns in the market. I think you may be surprised.

 
At 2:28 PM, September 05, 2010, Blogger Jonathan said...

"Economics is the astrology of our time"—Kim Stanley Robinson, recently, requoted in Ansible 278.

 
At 10:14 PM, September 05, 2010, OpenID ndrosen said...

Have you considered the Georgist model of booms and busts as driven primarily by land speculation? Or perhaps I should say neo-Georgist, since people like Professor Mason Gaffney have made advances over Henry George. Recent experience is certainly compatible with real estate speculation being important in the boom-bust cycle, and it is worth noting that Dr. Fred Foldvary used his Austrian-Georgist synthesis to predict the current recession years in advance, and he got the timing pretty well right.

 
At 11:17 PM, September 06, 2010, Blogger PlanetaryJim said...

Didn't unemployment peak at 22%? Or, anyway, that's using the old algorithm and the government's current figures. John Williams over at shadowstats dot com has the particulars if anyone cares.

So, that's close to 8%. So what if millions of people are unemployed? Scum like Romer always land on their feet and sleaze with taxpayer money.

 
At 3:02 PM, September 08, 2010, Anonymous Andrew said...

I found a better example of "assuming your conclusions":

http://daviddfriedman.blogspot.com/2010/08/king-tut-statistics-and-pet-peeve.html

:)

 

Post a Comment

Links to this post:

Create a Link

<< Home