Thursday, July 29, 2010

Blinder & Zandi: Methodology & Impact

As promised, I took the time to review Blinder and Zandi's full report, paying particular attention to methodology outlined in Appendix B. It's all pretty straight forward and nothing jumped out at me as particularly specious.

One reaction to my post yesterday and the report was that government spending (G) doesn't result in increased demand. This is a patently false statement because the aggregate demand equation in the short-run is essentially the GDP equation restated.

  AD = C + I + G + (X-M)  \

If G becomes larger, then AD becomes larger. I would also restate I have not seen any evidence of government spending crowding-out private investment.

On a slightly divergent note, the stimulus bill has been criticized because it has not created jobs, but if you look at where the stimulus bill put money, the largest part that has currently paid is in tax cuts ($188 billion). I would argue that President Obama was wrong to seek some sort of commen ground with Republicans in the Senate that had him approving a bill that didn't put enough people back to work. Instead, the bill was meant to entice private industry to hire workers with a payroll tax holiday. This is an indirect process that has paid very limited dividends.

Finally, I'll close on this point. In the New York Times article John B. Taylor of the Hoover Institution is quoted as saying that the stimulus program had, "very little impact." Little impact, as he observes it, is still an impact. In this situation, when private investment had dried up and the gears of the economy had threatened to halt entirely policy measures were taken that averted a crisis.

It's fine to discuss the levels of impact of these measures, but I haven't heard any serious economist say they had absolutely zero positive impact on our economic situation.

8 comments:

Colin said...

One reaction to my post yesterday and the report was that government spending (G) doesn't result in increased demand.

Allow me to amend my response to that I don't it has an impact in the long-run. Yes, stimulus can produce a short-term affect similar to a sugar rush, but then the impact wears off and you are back where you started, only with a higher debt level. We saw this in Japan durng the 1990s and there is evidence of a slowdown now as the stimulus money peters out.

This is because money spent by the government with the goal of simply shoveling it out the door as fast as possible probably going to have as great an impact further down the road as other forms of private spending. For example, R&D does not factor into GDP:

http://www.creditwritedowns.com/2010/07/gdp-higher-if-research-and-development-treated-as-investment.html

Yet plainly it is easy to see a scenario in which the long-term health of an economy benefits more from R&D spending than, say, giving $150 million to the Smithsonian (as the stimulus did). Common sense tells us that not every dollar spent will have the same impact, with some more useful than others, even though GDP makes no such distinction.

More problems with GDP highlighted here:

http://biggovernment.com/vderugy/2009/10/30/the-economy-is-growing-right-and-i-dont-have-a-french-accent/print/

Keynesian economic theory essentially views the economy as a giant flywheel, with the notion that once some initial force is applied that the flywheel will begin to spin on its own with only minimal continued effort. Stimulus is thought to provide that initial push. But this is incorrect. Long-term economic health is based on money flowing to productive ends, something that government is quite bad at doing.

On a slightly divergent note, the stimulus bill has been criticized because it has not created jobs, but if you look at where the stimulus bill put money, the largest part that has currently paid is in tax cuts ($188 billion).

There weren't any tax cuts in the stimulus bill, only tax breaks. Not one single tax rate was reduced. As Bruce Bartlett notes, these are best understood as government spending by another means:

http://www.forbes.com/2010/05/27/finance-economy-tax-code-opinions-columnists-bruce-bartlett.html

There is nothing about these tax breaks that Republicans or anyone else should embrace.

Little impact, as he observes it, is still an impact.

But at what cost? Shouldn't there be some kind of cost-benefit analysis? We're digging ourselves deeper and deeper in the hole, and for what? A 1% bump in GDP?

Jason said...

On GDP & :
Maybe we should include R&D into the I. And I saw that BEA study, but you're talking about a question of accounting. The report has no relevance to the model Blinder & Zandi use.

As to whether or not $150mil is better spent on R&D or the Smithsonian institution, I would ask if you would support the government directing R&D funding choices? The stimulus bill was all government money, and I would think you'd be happy the government didn't play kingmaker in R&D investment.

On the Long-term:
I completely agree that long-term economic growth is based on the intelligent allocation of resources. The government is not great at this, but as the study points out without the short-run policy choices the government enacted the long-run prospects would have been more hampered then they are right now.

On the impact:
Blinder and Zandi tell us the policy measures taken had a substantial impact on GDP and unemployment. I haven't seen Taylor's report, if it exists.

Colin said...

I would ask if you would support the government directing R&D funding choices?

No, I don't at all, I think the money should be left in the private sector.

The government is not great at this, but as the study points out without the short-run policy choices the government enacted the long-run prospects would have been more hampered then they are right now.

On the impact:
Blinder and Zandi tell us the policy measures taken had a substantial impact on GDP and unemployment. I haven't seen Taylor's report, if it exists.


This all comes back to whether this model is at all useful. Let's remember what this study is not. The study is *not* an evaluation of the real world impact of the stimulus. The authors did not perform an actual examination of what occurred, looking at where money was spent and its results, counting jobs or anything of the kind.

What the study is, is running a model. That's it. Two Keynesian economists ran a Keynesian model to examine a Keynesian policy measure and arrived at a favorable result.

Now, I am not trained in econometrics and am not really positioned to evaluate the robustness of their model. But then again, I also think that modeling something as complicated as the US economy and producing results with any kind of certainty or exactness is largely a fool's errand. One prime example of this is the modeling the White House produced which compared unemployment numbers with and without the stimulus prior to its passage. We all know how that one went.

I'll leave additional commentary to someone far more qualified than me:

http://econlog.econlib.org/archives/2010/07/how_the_blinder.html

Colin said...

BTW, I would also recommend giving this a read:

http://www.nationalaffairs.com/publications/detail/crisis-economics

Colin said...

Another critique of the Blinder-Zandi paper:

http://johnbtaylorsblog.blogspot.com/2010/07/more-on-blinder-zandi-working-paper-on.html

Ben said...

Isn't John Taylor the guy from the Hoover institute quoted in the NY Times piece?

Ben said...

Follow the link and answer your own question: "Yes"

Scot said...

Another thing to keep in mind with the stimulus is how it is being offset by a reduction in state and local spending. Something like 48 states have produced what are essentially 'anti-stimulus' acts that have countered the impact of federal stimulus and constrained job growth.