A political stock market (PSM) clearly beat the polls in predicting the out
come of a Swedish referendum on whether or not Sweden should join the Europ
ean Union. In fact, polls were unable to make such predictions since the nu
mber of undecided respondents always far exceeded the observed YES/NO margi
n. However, an obstacle to PSMs serving as a superior forecasting instrumen
t is that they can be sensitive to price distortions - by interest groups t
hat may wish to effectuate, and pay for, such distortions - or forecast com
petitions tied to PSM trade gains, the latter of which was tested here.