The other day I read an interesting paper that gave me a thought on why betting markets might become biased even if participants act rationally to maximize their utility. In the paper suspense and surprise (J Ely, A Frankel, E Kamenica, 2015), the authors create an economic model for suspense and surprise (sounds cool right?!) with which they examine different sports, plots in novels and races for political offices. From this I started reading some more on the literature and this quote from A Caplin and J Leahy (2001) caught my eye:

[w]e define suspense as the pleasure experienced immediately prior to the anticipated resolution of uncertainty, and posit that it is positively related (up to a point) to the amount that is at stake on the outcome of an event. This provides a simple reason for agents to bet that their emotional favorite will win in a sporting event. By betting on their favorite, agents increase their stake in the outcome, thereby heightening feelings of suspense

This makes intuitive sense of course, when people bet on their favorite team the game becomes more exciting and thereby increasing their utility. From the public choice literature on rational ignorance we know that since any individual voter has a very low probability of changing the outcome of an election, people don’t have the incentive to consume political news in order to make a good decision on election day. The reason instead to why people inform themselves on political issues is non-instrumental – it’s entertainment such as watching sports or reading novels. And therefore they don’t have any strong reason not to be biased in choosing what party to cheer on.

The answer to the question why people vote is probably similar to why people participate in doing a wave or cheer in a football stadium to show support to their favorite team. It’s not that sport supporters think that their participation in some way will affect the athletes to perform better. But we are social beings and like doing things in group.

This gives an explanation to why prediction markets might be biased when people with strong partisanship bet on political events. But I say might because even if it’s the case that some people act this way, it will not affect the price on the prediction market as long as the marginal trade in the market is motivated by rational actors. As J Wolfers and E Zitzewitz (2004) writes this about why irrational betting participants won’t distort the signals on a betting market, but so is true for the people betting for suspense. So even if a lot of democrats and a lot of republicans bet in order to enhance their feelings of suspense, if there are some people in the middle that act in order to win money (and there is high liquidity on the market), the predictions that derive from betting markets will be good forecasts.

references:

Caplin, Andrew, and John Leahy. ”Psychological expected utility theory and anticipatory feelings.” Quarterly Journal of economics (2001): 55-79.

Ely, Jeffrey, Alexander Frankel, and Emir Kamenica. ”Suspense and surprise.” Journal of Political Economy 123, no. 1 (2015): 215-260.

Wolfers, Justin, and Eric Zitzewitz. ”Prediction markets.” The Journal of Economic Perspectives 18, no. 2 (2004): 107-126.