Monday, March 6, 2017

J. W. Mason — Saving and Borrowing: A Response to Klein


Yes, we are still arguing over terms like "saving" and "borrowing." JW Mason clears some of it up.

J. W. Mason's Blog
Saving and Borrowing: A Response to Klein
JW Mason | Assistant Professor of Economics, John Jay College, City University of New York

3 comments:

Neil Wilson said...

It doesn't clear much up.

The underlying problem seems to be digging down from the aggregate. So why keep doing that?

JW Mason said...

I don't think there's an alternative if you want to describe what actually happened. The data is aggregate by its nature.

Anyway, the whole economics profession is obsessed with building everything up from the behavior at the household level. If one or two of us wants to start from aggregates instead, isn't that ok?

Neil Wilson said...

"If one or two of us wants to start from aggregates instead, isn't that ok?"

Of course it is ok, and I enjoy reading it. But where is it going?

The 'whole economics profession' builds things up from mathematical formula that have no bearing whatsoever on the real world. Their search of 'deep parameters' is going to be a dry well. That has nothing to do with the way I would propose doing it, which would be to model the actual entities doing what they actually do and let the behaviour aggregate and emerge as it should.

The problem is that the job of economics is to advise policy and advise the construction of the control system for the economy. And policy impacts upon the parts, not the aggregate - which is the essence of the Lucas Critique if I understand it correctly. Trying to tease apart the aggregate is a bit like trying to get eggs and flour out of a cake. It can't advise policy. It's also aggregate data collected under the influence of a policy and so is tainted by it. It can't advise other policy environment - since the data collected there would likely be materially different.

I would suggest that composing the aggregate from real world analogues - people, businesses, etc which you can check against the real world for accuracy is a more fruitful way of going forward. Then you can get them to do all the things you mention and see what arises.

I would not try and work out how a business operates by trying to tease apart its accounts. I would build something, show it to the people in the business and ask if it is a reasonable approximation.

We have the computer power and decades of Information Systems experience to create a virtual world we can actually run experiments on.

That's what we need. Something we can let people play with and see what happens.