When I last taught econometrics – some 10 years ago – there was a practical reason for why Bayesian methods played little or no role in the textbooks. Even though classical methods asked the wrong questions* and forced people to wade through a myriad of complicated and contradictory ways of answering them, it was at least possible to extract point estimates from the data. A practitioner who wanted to address problems more complicated than the
linear normal regression model would find that Bayesian methods had
very little to offer in the way of concrete advice, so she could be forgiven for concluding that spending time on
them was pretty much pointless.
But that’s not the case anymore. The development of Markov chain Monte Carlo (MCMC) techniques means that there aren’t any questions that classical econometricians can tackle more easily than their Bayesian colleagues, and there are quite a few once-intractable models – stochastic volatility, multinomial probit – where MCMC has made estimation routine. But you wouldn’t know it from the current generation of textbooks. (There are of course several texts that provide an excellent grounding in modern Bayesian econometrics, but they are very much the exception.)
Bayesian methods are best adapted to the questions of most interest, and are easier to use. Unfortunately for the next generation of economists, this perspective has yet to significantly infiltrate how econometrics is taught.
*Ask yourself which is of more interest:
a) Probability statements about an observable statistic, conditional on an unobserved feature of interest.
b) Probability statements about an unobserved feature of interest, conditional on an observable statistic.
If you answered a), then you are God – or perhaps one of the lesser
deities who is bored with just knowing The Truth, and is looking to
make some extra cash by betting on what statisticians will conclude
from what Nature draws out of its urn of coloured balls. But if your fate is to live among mortals, knowing only what mortals
can learn, then b) is the only question that is worth spending time
thinking about.
Worthwhile Canadian Initiative 
I’m rapidly becoming a fan of yours. Shame their aren’t more.
I’m sorry your last post was so misinterpreted (something to do with Labour Day I think). But no publicity is bad publicity. I bet your hit count has gone up heaps.
By the way, I couldn’t agree with you more. Makes me sort of wish I was retired and had time to do some economic research again. Doing it on your own is feasible now with powerful PCs and the internet.
This is late, but why does the market fail? If the Bayesians can ask/answer better questions, shouldn’t they start earning excess profits in terms of citations, etc, inducing a shift toward Bayesian econometrics?
The problem is twofold:
– Researchers who have spent their entire careers using classical methods and who have never had Bayesian methods explained to them would have to take time out to retool. The returns from such an investment are decreasing with age, so established researchers won’t bother learning it, and they certainly won’t teach it.
– Theoreticians will resist the move, because most of the ‘problems’ that they’ve spent so much time trying to solve turn out to be either not a problem or trivial in the Bayesian paradigm. If they adopt Bayesian methods, what would they do with their time?