Gambling, risk and closed decision-making

Betting on the outcome of elections has a very long history, which is well-documented in particular in the case of presidential elections in the US, where it can can be traced to 1868, and further back to the early sixteenth century in the case of papal elections. The difference between these two types of election is that the former is open, involving a diverse and very large number of electors, while the latter is closed, in the sense of a small number of decision-makers whose choices are shrouded, at least before the decision, in a layer of secrecy.

This paper examines the second type of decision-making, using the 2013 papal election and the 2012 US Supreme Court decision on the 2010 Affordable Care Act ('Obamacare') as case studies of the relative efficacy of different methodologies that have been applied to forecast the outcomes. There are three main methods that are conventionally used to forecast these sorts of decisions – the identification and aggregation of expert judgements, inforamtion derived from betting and prediction markets, and statistical methods.

Each of these methods has been applied to forecast closed-door decisions in the past, but the availability of data from in-running betting exchanges and from prediction markets, which are used to aggregate information from a diverse range of observers, is relatively new.

Each of these is examined, compared and contrasted in the context of these closed-door decision-making processes.

A large, novel data set of in-running trades is used to assist the analysis. The results will provide additional evidence on the performance of these different forecasting methodologies in predicting small-group decisions where the decision-making process is shrouded in secrecy.   

---

Back to: Gambling and Other Risky Behaviors

<< Previous Abstract | Next Abstract >>