Still arguing about the stimulus

The stimulus of 2009 is still a point of argument on this blog. I've argued that the stimulus helped stop the unemployment spiral and that parts of the stimulus, such as support for state workers, were quite helpful in softening the downturn.

The general argument against the stimulus is that it doesn't do any good for the economy, increases the national debt, sucks money out of the private sector, and the economy would correct itself in the same timeframe without one.

Evidence for this viewpoint isn't generally provided by those making the argument. It's a hypothetical, not even close to proven. Part of the problem is that economies don't let you do controlled experiments, where you apply stimulus and measure the result, and then reset, don't apply stimulus, and look for the difference. This problem equally affects both sides of the argument.

So what evidence can there be for one side or another?

If the economy returns to normal soon without stimulus, that would be good evidence that stimulus wasn't necessary. We could also look at how dismal metrics like unemployment go, how long the job slump last, and other factors.

Here's a graph of estimated or measured unemployment since 1890:

Image: wikimedia.org

The Panic of 1893, which caused a spike in unemployment lasting until 1900, stands out, as does the Great Depression. Unemployment spikes since 1945 have been much milder, so we need a different scale for them. Here's a chart of the more recent recessions:

Image: calculatedriskblog.com

I thought it would be helpful to see the first derivative, or the rate of change, of unemployment. Here it is, the 4-week moving average of new unemployment claims:

Image: calculatedriskblog.com

Here's what I see in these graphs:

• The middle graph shows that the peak of the 2009 unemployment was blunted--not as high or sharp as one might expect. It's impossible to be sure what shape the peak should have, but if it was blunted, that means that unemployment may not have peaked because of the stimulus.
• The bottom graph shows that the rate of job losses were as severe as the 1982 recession. The recessions between 1982 and 2008 were considerably milder.
• Since 1982, the period it takes for employment to recover is getting longer.

A commenter arguing against the 2009 stimulus said these things, which I believe are refuted:

• It is entirely normal for the "leveling off" of employment to occur 24 months after the fall, the time period shown in your graph. My reply: It's not a slam-dunk that employment will stop falling in that short a period, as the 1893 and 1929 downturns show.
• The fact that other sectors were not (net) creating jobs while the government was sucking an extra $800 million out of the economy is not evidence that that $800 million wouldn't have created at least as many jobs were it spent/invested in the private sector. My reply: This point ignores how much the private sector was contracting and how private businesses that had cash were sitting on it, not reinvesting it. [Not refuted in the charts, but in this post.]

Commenters who are against the stimulus often claim or imply that the economy would do better without the stimulus, yet they offer no evidence for that assertion. Again, clear evidence is difficult to find because you can't run controlled experiments on an economy. But it's been hard to glean whether proponents have any evidence for this view. It's a belief without evidence--meaning that it's a dogma. On the other hand, it would be particularly hard to find evidence against stimulus because the government has run stimulus programs during recession after recession, so we don't have good comparison cases without stimulus. 

There will probably not be definitive evidence for or against stimulus packages. It's not possible in the nature of economies and how governments respond. However, the dogma against stimulus goes further than healthy skepticism. It replaces a belief in stimulus (with marginal evidence) with a bias against stimulus based on ... what, exactly? That's the problem. It's turned into an article of faith despite having no foundation. 

Perhaps the animus against stimulus is part of the rejection and vilification of all things related to Obama. The idea of stimulus now carries Obama-taint. That's bad because it's hard to make a case against stimulus and the damage it can do (increased deficits, increased budget baselines forever after) if you're just reflexively against it, and you can't put together a coherent argument. 

It's not as though there isn't an argument to be made. Opponents of stimulus can point out that a stimulus may not be necessary in the short term since the economy can correct on its own. Also the deficit spending to support a stimulus may be harmful in the long term.

I suspect there's a good reason that the federal government rarely votes against a stimulus during a downturn. In harder economic times, those in power don't want to argue that they are ineffectual (unable to do anything), so "doing something" has the edge. "Doing something" is also supported by the people who will benefit from the stimulus.

The biggest problem with the Obama stimulus is not that it happened, or that it's unclear whether it was helpful, but that the spending is staying so high for so long. The following chart shows that GDP started improving in 2009 Q2 and returned to 2007 levels in 2010 Q3:

Image: businessinsider.com

Yet the expanded federal spending, which should have been tapered starting in 2010, continued to grow and became the defacto new baseline:

Image: downsizinggovernment.org

The spending reduction in 2012 occurred because the Tea Party pushed budget cuts--it wasn't part of the stimulus plan that spending levels dropped back down. In fact, the supporters of the stimulus, Obama and the Dems in Congress, resisted budget cuts.

So if the government is reluctant to back off increased spending as recovery strengthens, what should we do? Should we forfeit the possible benefits of stimulus due to the drawbacks? I'd like to see a lighter hand on stimulus with definite cool-down features enacted with the stimulus itself. There should be no long-term stimulus, which also means that any tax cuts enacted as stimulus must have expiration dates no more than a couple years away. I'm thinking in particular of the Bush tax cuts that were enacted as a stimulus, but were scheduled to last 10 years.

The severe recession starting in 2008 definitely warranted stimulus measures though many of those measures weren't targeted in effective ways. The stimulus was too large, lasted too long, included too much spending for favored political groups, and was an excuse for higher spending for the foreseeable future. However, complaints about the stimulus are scattershot and politically motivated rather than being analytical.

Extras. Long analysis of stimulus measures from a Koch-supported think tank. Lots of data and good questions. See especially the three hurdles on p. 54. Apply the questions raised to all stimulus programs, not just the Obama stimulus.

Update 1/12/18. I found a graph that shows the severity and depth of the 2007-10 recession with astounding clarity.

Image: calculatedriskblog.com