Researchers have reached few if any definitive conclusions on how political prediction markets fare in comparison to polls. Opinion on the matter tends to fluctuate depending on the performance of particular markets and polls in each election cycle.
A study of the Iowa political market through 2004 by professors Robert Erikson and Christopher Wlezien found that, until the very last days before the election, polls provide more accurate snapshots than prediction market prices. Erikson and Wlezien observed that traders “tend to hold persistent beliefs about the vote division that contradict the polls.” Because these beliefs are often incorrect, traders with “modest knowledge” of how to read polls can “reap handsome profits” in the Iowa markets.
A 2014 survey of Intrade, however, concluded that the site’s predictions “appear to have
consistently beaten individual pollsters.”
A more general 2015 analysis also suggested that prediction markets have advantages over polls, finding consistent empirical evidence that markets are “at least as accurate as
opinion polls, and much quicker to incorporate new information.”
Events since 2016 have produced conflicting opinions. Pundits at the Economist and Financial Times suggested that polls are more accurate than prediction markets after the failure of PredictIt investors in 2016 to forecast Brexit and the nomination of Donald Trump. Prediction markets, they argue, have proven to be “losers” given that the polls, unlike investors, “did an outstanding job of predicting these supposedly unthinkable events.” Others maintain that, even in the cases of the Republican primaries and Brexit, prediction markets were well calibrated; traders were not “fooled,” but rather, acted rationally in considering factors beyond polls.
What these mixed studies suggest is that both political prediction markets and polls provide different types of insights that, with the appropriate analysis, can inform better prediction. In a poll-obsessed culture, organizations that make use of political prediction markets stand to gain an edge for the foreseeable future. As Michael Abramowicz correctly predicted in his 2007 book, Predictocracy: Market Mechanisms for Public and Private Decision Making, “it will be a long time before media reporting of the results of prediction markets will significantly affect public policy.”