Prediction and its discontents: guidance for Australia from the debate over social science forecasting

The well-documented failure of many experts to predict many events of strategic importance-from 9/11 to the global financial crisis to the Arab Spring-has led to a fashionable belief amongst some academics and popular commentators that prediction is a 'fool's errand'. Problems inherent in forecasting political events, such as cognitive biases, strategic interaction, complexity effects and various others, make this an attractive position. However, although a proper degree of humility is warranted about the ability to predict strategic events, this article argues that not making predictions at all is not an option. No genuinely workable policy advice could flow to Australian policy makers unless analysts are able, however roughly, to try to forecast the future. Moreover, with some hard, determined work, social scientists have recently produced work which promises to be able to foresee important events such as state failures and civil wars with reasonable accuracy. Experts should be encouraged to make, and should routinely be evaluated on, predictions about their area of expertise. Even failed predictions can be useful in pointing up gaps in knowledge and understanding which would otherwise have remained opaque.