Wednesday 13 April 2011

The end of macro?

Over at the EconLog blog Arnold Kling writes,
I was born at the wrong time. I attended graduate school near the peak of ultramasculine economics (my term). The math was almost too much for me (in fact, five years later it would have been too much for me). I have always doubted the value of the theorem-proving approach to economics. I note that these days many people are sharing my doubts about the mathematical macro that emerged during what Paul Krugman calls the "Dark Age." However, unlike Krugman, I do not think that we can simply go back to old Keynesianism. I think that there were many problems with Keynesianism circa 1970 that were not solved by coming up with mathematical solutions for the Lucas critique. The Leamer "con" of econometrics problem and the "unit root" problem are what caused me to "lose my religion" regarding macroeconometric models.
Here's a thought. May be the problem is not with the maths or stats of macro but with the actual idea of macro itself. May be aggregate economics just doesn't work, we lose too much valuable information in the process. May be the costs of aggregation are just too high, we need to look at the micro level for solutions to so-called macro problems. For example unlike the macro-level data, micro-level data provides little evidence in support of Solow's productivity paradox to do with the effects of computers in the economy. While the macro level data can tell us that something has changed, as it did mid-90s in the U.S., it can not tell us what changed and why. The productivity data is the aggregated result of changes at the micro level, in this case at the level of the firm. Such changes require a microeconomic explanation. Would it not be better if we were to go back to thinking about issues like unemployment and monetary theory as microeconomic issues. We could see unemployment as a problem to do with labour markets and interest rates as relative prices, monetary policy as having to do with the supply and demand of money etc. Do away with teaching students AD/AS analysis from day one and just teach them about markets and consumers and producers instead.

3 comments:

Matt Nolan said...

"We could see unemployment as a problem to do with labour markets and interest rates as relative prices, monetary policy as having to do with the supply and demand of money"

Since this is exactly what macroeconomics does - and has done for nearly 30 years - I'm not quite sure what you are saying.

Paul Walker said...

No it doesn't see it that way since if it did it would be macro it would be micro. In many cases there is only one aggregate labour market in macro with unemployment being analysed in terms of aggregate demand and supply not as the result of supply and demand changes in multiple (micro-level)markets.

Matt Nolan said...

"No it doesn't see it that way since if it did it would be macro it would be micro"

That is ridiculous.

Microeconomics is the analytical foundation for all economics - to be honest, I love micro. Macroeconomics is the use of the economic method to discuss certain questions - it is an applied field of micro, but at the same time due to specialisation it is unlikely we would see a microeconomist truly replicate macro at the same time (although someone like Huw Dixon has done awesome work along these lines!)

If you wanted to define micro so broadly that "macro is micro" then everything is micro - and the distinction would be non-nonsensical.

As a result, if you want a clear distinction, and you want to say that macro should have micro foundations, then I would note that macroeconomists have know that for about 40 years - hell I was reading a book on inflation from 1978 the other day, and a lot of the discussion about relative prices and expectations sounded surprisingly modern!

By saying that "macro is only AD-AS" you are attacking a straw man version of macro that does not exist.

"In many cases there is only one aggregate labour market in macro with unemployment being analysed in terms of aggregate demand and supply not as the result of supply and demand changes in multiple (micro-level)markets."

Sorry, how much macro-literature have you read? I haven't read all that much as I'm the lowest dregs of the economics discipline - I am a business economist. But even so, I've seen plenty of literature discussing heterogenity in product and labour markets and the computational difficulties (and data difficulties) with analysing them.

I have issues with the representative agent models floating around - but the fact that I see SO MUCH analysis of how to make their assumptions more realistic, and SO MUCH discussion of improving the available data sets for them, gives me faith that macroeconomic analysis will improve in its usefulness to answer the questions put forward to it.