Quote of the Day: Milgrom and Roberts (1987) AER paper

|Peter Boettke|

"[N]o mere fact ever was a match in economics for a consistent theory."

HT: Nicola Giocoli

Comments

[Daniel Kuehn]

Sure - but who ever uses a "mere" fact? :)

Empiricism is about corroboration and robustness! That means lots and lots of facts!

So out of curiosity... did Milgrom and Roberts not get the results they wanted to see?

[Pietro M.]

I have a doubt. Has monetary economics succeeded in proving empirically that money is neutral or not neutral?

It appears that the most fundamental proposition of monetary economics cannot be tested empirically...

As far as I remember:

1. The original "proof" (as reported in a paper by Christiano) of non-neutrality were vector autoregressions (VAR).

2. King and Plosser 1984 replicated the same VAR results in a neutral-money framework, proving what was obvious: VAR reports correlation, not causation.

3. Mankiw noticed there was a kink in the movement of the price level with the money supply in that paper (irrelevant argument: change a functional equation, add a parameter, and you can fit everything).

4. Prescott (in "Real facts and a monetary myth") proved that time series data really behaved in the way King and Plosser predicted.

5. Gali, in its recent textbook on new-keynesian economics, tried a new methodology, which may be called falsification by amnesia (looks like AE in the '40s): he cited the empirical problems of money-in-the-utility models, but forgot about King and Plosser's paper.

At the end, it appears we don't know whether money is neutral or not, on a posteriori grounds.

Modern macroeconomics looks to me like a gigantic exercise in data fitting.

[Mario Rizzo]

Any one know of a good -- and recent -- study of the impact of empirical work on (macro)economic theory?

[azmyth]

"Facts are the enemy of truth."
Miguel de Cervantes

[AndrewG]

The statement is obviously true. If the fact was a match, then the theory would not be consistent :)

[Lee Kelly]

Although it seems like a paradox, a fact is called "mere" when its truth contradicts more theories and assumptions, not less. That is, the more one's worldview would need to be revised to consistently account for a fact, the more likely the fact will be dismissed as an error of observation or measurement, or a lie or misrepresentation, i.e. a mere "fact."

This is not altogether irrational. Sometimes accepting a proposed fact as true would create more problems than it would solve. Contrary to many naive empiricists, the facts do not "speak for themselves," but must be interpreted in the light of auxiliary assumptions. Logic alone cannot decide the truth of such propositions, which is why extra-logical rules of method are often sought.

What matters is not the consistency of a theory as such, but the degree to which it offers a satisfying explanation of events in conjunction with our countless other assumptions, prejudices, and biases. Without such strong ties to our overall worldview, "mere" facts are chosen above consistent theories all the time--these are the usually the theories of our intellectual opponents.

[Jerry O'Driscoll]

Some years ago, J. R. T. Hughes (now deceased) wrote a paper surveying the theories of the Great Depression. He concluded that none came close to accounting for the severity of that episode.

I remember one line early in his presentation: "It is important to have the right theory. It is also important to know what you're talking about."

[Rafe Champion]

Yes Jerry, the most awkward question for a lot of young researchers is "Why are you doing that?"

[Barkley Rosser]

Indeed, when someone puts "mere" in front of "fact" they are clearly trying to diminish its import or significance. And certainly there are many facts that are essentially trivial and hence worthy of being designated to be merely "mere."

OTOH, this can also be an excuse or a diversion to try to get people to ignore or downplay something really is an important fact. In the end, economics is a science, if not a hard one, and facts are the bottom line, although some are merely noise. But especially when a fact is accompanied by other facts that point in a particular direction that is uncomfortable for a theory, it (and they) should not be dismissed as being merely mere.

[Juicy Couture Handbags]

I am especially glad to observed this weblog. It was fantastic to read your perspectives in this post! I actually get pleasure from all that you just fill your weblog with. Could possibly the joy and happiness around you right now and generally.

[b makowsky fashion]

Wow! Thank you! I always wanted to write in my site something like that. Can I take part of your post to my blog?