August 27, 2007
Contingency and Counterfactual
Dani Rodrik, in the closing of a post on historical determinism and development:
This may seem discouraging if you are interested not only in understanding the world, but also in changing it. On closer look, though, [Acemoglu, Johnson, and Robinson]’s historical determinism leaves plenty of room for human agency and policy choices to make a difference, as I have argued here. Statistically, plenty is left unexplained by historical factors.
Ha. Neat. I sort of like that: We get to be the error term.
Related: My train reading these days is Virtual History, a collection of counterfactuals edited by Niall Ferguson. Fun discovery: To spin an even mildly convincing counterfactual, you have to make sure the fundamental facts leading up to your branch-point are really solid. So oddly it’s in the fake-history book that I’m learning about all these real events (a lot of World War II stuff, etc.) in more detail than I ever have before. I think Ferguson and other fans of counterfactual would say yes, that’s the point.
Just discovered: Philip Tetlock, the terrific Berkeley researcher I saw give a Long Now talk on experts and forecasting earlier this year, also has a book of counterfactuals! Why was I not told of this earlier??
Psst: Any favorite what-if scenarios?