First, I just want to agree wholeheartedly with Jay Ulfelder’s conclusion on Dart-Throwing Chimp:
Look, these decisions are going to be made whether or not we produce statistical forecasts, and when they are made, they will be informed by many things, of which forecasts—statistical or otherwise—will be only one. That doesn’t relieve the forecaster of ethical responsibility for the potential consequences of his or her work. It just means that the forecaster doesn’t have a unique obligation in this regard. In fact, if anything, I would think we have an ethical obligation to help make those forecasts as accurate as we can in order to reduce as much as we can the uncertainty about this one small piece of the decision process. It’s a policymaker’s job to confront these kinds of decisions, and their choices are going to be informed by expectations about the probability of various alternative futures. Given that fact, wouldn’t we rather those expectations be as well informed as possible?
And I just want to underscore something that Daniel Nexon on the Duck referred to as the peformativity problem, referencing some interesting work in economics. I do agree that an interesting intellectual question is if political scientists ever get good at forecasting–a big if—and those forecasts do generate policy interventions, then the forecasts might become self-defeating or self-fulfilling. The goal is for them to become self-defeating, but one could imagine the opposite: military intervention or aid could destabilize a perilous society or warnings of risk could lead to better data collection which could lead to categorization of a problem as being more serious than in an area without warning. Either self-defeating or self-fulfilling predictions present a host of empirical problems. It’s worth a longer discussion, but I think it’s worth starting the discussion. My hunch is political science is shifting methodologically to a fight between the forecasters (probably bolstered by big data) and the causal inferencers. The issues associated with causal inference have been fleshed out in some detail, but the issues associated with forecasting are still relatively unexplored. Let’s start exploring.