with Matthew Wall and Kevin Cunningham. Journal of Elections Public Opinion and Parties, Volume 22, Issue 1, pp. 3-26 (2012).
This article focuses on the usefulness of a novel source of data for forecasting party seat shares in the UK’s 2010 legislative elections. Forecasters of this property face what we call the ‘votes-to-seats’ problem, due to the operation of the Single Member Plurality (SMP) electoral system. We review existing approaches to this problem and argue that a new data source – constituency-level estimates of each candidate’s likelihood of winning their seat available from online betting markets – could potentially allow forecasters to elide the votes-to-seats problem. The article investigates methodologies for translating data from constituency-level betting markets in each of the UK’s 650 constituencies into aggregate, national-level predictions of parties’ seat shares. The accuracy and volatility of forecasts produced from constituency-market gambling data from the site Betfair.com proved to be a poor source for predicting the elections, and were systematically biased in several ways. Nonetheless, we argue that future research in this area should compensate for these biases to harness the potential of constituency prediction markets for electoral forecasting.