everything has real world consequences. do you have any useful predictions?
Sure, an economic analysis of different global warming policies will predict whether how good of a job they do of avoiding global warming costs (say cap-and-trade vs. some particular geoengineering policy). A theoretical understanding of different monetary policy regimes will help you predict which ones will lead to fewer recessions. A theoretical understanding of different tariff policies will help you predict what policies will lead to higher welfare.
I've long opposed discussing politics on Less Wrong. Elsewhere, however, I have been known to gaze into the abyss; and so it came to be that I wrote a handful of blog posts of the Oxford Libertarian Society Blog. I had the deliberate intention of bring a little bit of rationality into politics - and so of course ended up writing in something like Eliezer's style.
I wanted to establish some theory first, so the initial posts were about The Conservation of Expected Evidence and Reductionism, and then one particular Death-Spiral.
As you'll probably notice, one of my defences against the little-death has been to err on the side of attacking Libertarian positions; I provided an account of Traditional Socialist Values so we remember that our enemies aren't inherently evil, and then analysed an abuse of The Law of Comparative Advantage, showing cases where it didn't apply.
I can't promise I'll update at all regularly.
Post inspired by Will Newsome and prompted by Vladimir Nesov.
http://oxlib.blogspot.com/