The Latest Literature

One benefit of formal modeling is that you can later test how reasonable the assumptions are that lead to the conclusion, as well as the see if the other implications of your model are true.

I agree that the research shouldn't stop here, though.

You are twisting points 1 and 2 in a rather absurd way. Particularly the recent financial crisis is an important example in which

1. Government non-intervention lead to a dramatic and obvious market failure (e.g. the collapse of Lehman and subsequently of the interbank market).

2. Government intervention help make these markets working again.

3. Researchers are trying to design goverment regulations and intervention optimally in order to ensure that chances of market failures reduce and interventions are less distortionary.

This is a meta-model that is very different from the one you describe, and to me it seems much closer to reality.

I think they're starting from a faulty premise. It sounds more like this:

1. We know that government intervention causes market failure.
2. Therefore, when we observe a government-caused market failure, our task as economists is to justify further government intervention.

litehouse,

Why do you claim that it was "Government non-intervention" that "lead to a dramatic and obvious market failure"?

If you mean that the government intervened in the wrong ways prior to the current recession, I can understand. Definitely, the government has been and still is too lax on predatory borrowing. But no sane person can believe that the government was had a policy of non-intervention in financial markets at any time in recent history.

Touche' Arnold

I had the same thought when I got my copy.

It didn't start with this issue. It's in the Peter Diamond school of economic modeling. There's something pyschological about the academic establishment in this regard. A Schumpeterian would say they adopt and promote ideas which increase their own importance in society. Alternatively, there is self selection going on. Lastly, there is a political selection process (but then how did that get started).

There has been lots of empirical examination of Akerloff's original model in markets and every such market has myriad institutions and practices which mitigate adverse selection or similar perceived market defects. What many modeling economists don't want to deal with is that market inefficiencies tend to begat changes in practices, institutions, organizations, etc. which shift the playing field over time. Government intervention gives up such dynamic efficiencies and can only change when a crisis or collapse occurs.

AJ

Charlie,

Formal modeling would be even more valuable if the results and not just the assumptions could be tested against reality. That doesn't seem possible in economics which is why I am becoming increasingly suspicious of formal modeling.

@ James

I was referring to a specific event (the government could have bailed out Lehman Brothers, but chose not to) that lead to a collaps of the interbank market. I still disagree with Arnold on point 1 and 2. Just have a look at the papers he refers to: They start with the premise that moral hazard and informational asymetries are important features of some markets. They then show that this can lead to a breakdown or malfunction of the market (which we have observed), and they discuss what happens if the government intervenes in these cases.