Using The Model To Gauge The Model

John Taylor notes that you cannot seriously expect to use a model to gauge the effectiveness of the same model.  This is exactly what the ARRA folks have done, where they assumed that the “stimulus” had a big effect, and, by plugging the same numbers into the same model, shazam!

There are two ways of falsifying a model:  you either find internal contradictions, or you find external conditions which render invalid the predictions of the model.  To perform the second operation in a legitimate manner, you cannot use the model to drive your analysis of its conclusions.  Taylor has a new paper out arguing this, and applying it to ARRA, where he shows that the “stimulus” was about as “stimulating” as we expected:  not at all.

Expect Taylor’s paper to show up in a future edition of In The Papers, where I’ll discuss it in more detail.


Bad Economic News

Looks like we’re headed for another Recovery Summer.  The economic news this last month has been dreadful:  18,000 jobs created, 9.2% unemployment, 16% underemployment, and in general miserable numbers.  Stefan Karlsson says it’s even worse:  even with all of this, average weekly takehome pay went down and people are dropping out of the job search.

Megan McArdle has a graph which shows a very nice perspective on this:  unemployment through the Obama administration has been miserable.  One point on which I strongly disagree with McArdle is her statement that “Contrary to what his political critics will claim, this is not because Obama has somehow screwed things up–either by passing the stimulus, or not passing enough stimulus.  The fact is, we don’t really know much about how to treat the high and lingering unemployment of a post-financial crisis recession[…]”

Given historical US unemployment rates, there has to be some kind of reason for why they, all of a sudden, look like the bad old days of Europe (Germany pre-Hartz IV, Sweden before its financial reforms, etc.).  What has Obama done besides insinuate the State further into every reach of the American economy?  Obamacare alone is probably worth a point of unemployment, as the costs of compliance will push up the marginal cost of a worker, and firms will act today on impending legislation.  The “stimulus” didn’t stimulate at all, and TARP (both the Bush and Obama versions) created a wonderful incentive for giant, government-friendly mega-corporations to get hundreds of billions of dollars, while the big employers (small businesses) get squeezed.  And it doesn’t help that, under the Obama administration, the housing market has remained a mess.  Instead of fixing the incentive problems that the government has in place and killing government institutions like Fannie Mae and Freddie Mac, the administration has doubled down on failure, like a roulette player who’s sure that this time, it just has to land on Red!  And finally, what is federal spending as a percentage of GDP now?  When the government is wasting that much taxpayer money and putting into place so many structural impediments, of course there will be problems.

So what could a President Obama have done to prevent this?  Cut spending, reduce regulation, cut corporate taxes, kill failed government programs, and generally not bet on or subsidize failures.  The landing might have been harder (not that this has been all that wonderful), but you would have seen results that look much more like the Reagan example and Obama would be sailing toward term #2.  But, because he is a Democrat on the left wing of his party, he could never agree to such terms, even pragmatically.