Monday, October 10, 2011

Economists' political bias and model choice

One can count on Gilles Saint-Paul for innovative research topics. During his career, he has addressed and impressive array of topics that range far beyond Economics strictu sensu. For this reason, I have reported several times about his latest research.

His latest opus is an introspection in our profession and how our political biases influence our modelling choices. He claims that an economist with conservative inclinations will favor a model with smaller fiscal multipliers. While the ethical thing to do would be to be driven by empirical evidence, this may just be a subconscious choice. But at least economists strive to be logically consistent, and if one choose a large multiplier, then then must also claim that demand shocks are substantial, as models with large multipliers rely on this. Looking at evidence from the Survey of Professional Forecasters, Saint-Paul finds that forecasters who believe that expansions are more inflationary also adhere to the belief that public expenses are less expansionary.

Saint-Paul goes further, though. His claim is that we live in a self-confirming equilibrium. We devise theories to understand our surrounding and take decisions, and those decisions then shape the economic environment. Theories can thus survive even if they deviate from the true structure as long as the decisions make it conform. This is a statement about a lack of uniqueness of the path to the rational expectations equilibrium. In a sense, this is not too disturbing, as long as decisions are still optimal and outcomes do not differ too much from the rational expectations first best. And if this true, we will never know what the rational expectations first best is. Of broader implications would be if the political agenda of an economist would lead an economy on an different path, on a different self-confirming equilibrium. Is this why Europe and the United States are different? Were Keynes and von Hayek that influential?

4 comments:

NormanB said...

Here's my take on political differences in economists, very non-technical.

Liberal economists when their policy prescriptions don't work propose that they were only wrong in the degree of their solutions and the way to remedy that is to do more of what they orginally proposed. We see that with Paul Krugman who wants more stimulus. If you can call the War on Poverty an economic policy then my theory holds true there, also. As the socio-economic status of African-Americans declined in the face of huge government policies the Democratic response, then and even now, is to do more of it. Which brings me to Conservatives.

Conservatives look at outcomes and change policies. The ill-named Welfare Reform of the 1990's brought to Clinton's desk three times by Gingrich before it was finally signed (so I'm laughing at Clinton's recent speech taking credit for this policy) has had very good effect on A-As.

Conservatives look at the school system where the Liberals only want to keep spending more money even though test results haven't improved and want to do things a different way.

Its ironic that the Conservatives are much more flexible than the Liberals for their titles would lead you to think the opposite. The title, Progressive, is an even bigger joke.

Kansan said...

While St-Paul's story can make sense, it is impossible to prove. What if an economist has a particular political inclination because he saw the results of his models? Observationally, this is equivalent as we have no information when the political inclination was formed.

Devin M said...

@NormanB
Wow. Just, wow. Conservatives have been arguing for 30 years now that all that's necessary to grow the economy is more tax cuts. The result has been the 30 years of the slowest economic growth since GDP numbers began being collected 80 years ago. The best performance of the last 30 years (ironically enough, under the reviled Clinton), still ranks as below average compared to the previous 50. And let's not forget this performance followed a much-derided tax hike.

And yet Conservatives continue to assert that we know tax increases destroy jobs and the economy, while tax cuts do the opposite. 30 years of evidence says there is no significant negative impact, and if there is a relation between tax increases and economic growth, it's actually positive. Yet Conservative economists haven't changed their tune one bit.

Actually, on second thought, I will concede one area where conservative economists have been extremely flexible. When conservative politicians need a conservative economist to claim deficits and debt are destroying American, they'll quickly find one. But a few years earlier, when they needed a conservative economist to claim deficits/debt didn't matter, this was also no problem. Indeed, on any number of issues, conservative economists (along with conservative scientists...if that's not an oxymoron) seem to have unimaginable flexibility and the ability to change their views 180 degrees when it suits the political strategy of conservative politicians. Such flexibility is hardly admirable, IMO.

Anonymous said...

This subject has been treated before, albeit less rigorously. I would be astonished if you don't enjoy:

http://www.nakedcapitalism.com/2011/04/blacklisted-economics-professor-found-dead-nc-publishes-his-last-letter.html

Google "academic choice economics" for other choice morsels.