Wednesday, September 14, 2016

Paul Romer — The Trouble with Macroeconomics

Paul Romer flushes DSGE.

The Trouble with Macroeconomics
Paul Romer | Chief Economist, World Bank
ht Brad DeLong at Grasping Reality


Matt Franko said...

Tom aren't they missing the basic distinction between a rationalist vs a empirical approach?

Like from a basic philosophy perspective?

He seems to making a case for empirical only approaches but it's not clear to me that he realizes this...

Ignacio said...

"Someday, medicine might reach a point where any of us would be willing to let a dynamic stochastic simultaneous equations equilibrium model of the human body determine which course of treatment we follow for a family member faced with a life-threatening disease. It is not at this point now. As far as I can tell, no one in medicine sees work on such a model as a practical path toward better treatment of disease.
Perhaps this time, macroeconomists should admit that the wreckage runs so deep that they should
abandon the quest for the sacred simultaneous equation model. It might be wiser to adopt the messy methods that medical researchers have used to make discoveries that were implemented and actually improved health."

The problem lies in the lack of accountability of economists. IS the same for 'string theorists', they are not accountable.

Some lack of accountability is due in science, because we have to allow for failure, but when the whole discipline is following on a path and getting more and more extravagant and out of touch with reality empiricists have to raise their critiques. This is what has happened in particle physics the last years. This is not what is happening in economics.

The problem is that by foundation there is a serious lack of empiricists in economics, and particularly in macro. You can't call your discipline a science if it does not work with reality, is just mental games.

Economics is yet proto-scientific, economists DO NOT understand the scientific method, repeat: they do not understand the scientific method. There is a serious lack of training and discussion of methodology in economics, at least Romer here is steeping up on this front.

Ignacio said...

I lost my train of thought: if there is a lack of accountability you cannot pretend to be an 'expert' or 'pundit' and policy be derived from your 'knowledge'. The most harm string theorist do is a few millions wasted on funding, worst case scenario. Economists? Lives ruined, directly and indirectly.

Then the SWL of the world won't be able to come out and say: "hey, we knew all along about it and... we were just kidding", or "there is a lack of theory in those models.". Hello moron: if your theory is wrong it doesn't matter that you have a theory, check the freaking facts first.

Tom Hickey said...

Matt, the great advantage of scientific method is that it combines the rational (formal) and empirical (observable). Romer is not saying to abandon the rational-formal for the empirical-observable, but rather to be scientific and correlate them properly.

For example, conventional macro assumes a real rate of interest "r" that optimizes growth, employment and price level at an equilibrium that market economy tends toward. There is no observable correlation of the model with reality. But economic policy is formulated based on this assumption. It's magical thinking.