A close election drove more turnout than did political newspaper advertisements. The tightness of races was more significant than the weight voters put on the issues addressed by the various referenda.
It’s difficult to know which election Swiss voters might have believed would be close before 1998, but vote analysis from the 1981–98 period shows a weak relationship between voter turnout and the closeness of the actual vote. Once poll data were introduced in Swiss newspapers, election closeness and turnout were strongly linked.
The researchers hypothesized that if poll closeness spurred voting, they would see a bigger impact on turnout in politically unrepresentative municipalities—places where one party or political ideology dominates. (Perhaps these areas are like echo chambers, wherein sameness lessens voting urgency.) Prior to polling, unrepresentative municipalities had no relationship between turnout and close election results, while politically representative municipalities did, suggesting more-mixed areas supplied informal voting-expectation news. Once polling was introduced, close polls drove higher turnout in politically unrepresentative areas.
“And, in the era with polls, politically unrepresentative municipalities’ relationship between closeness and turnout became statistically indistinguishable from that of politically representative municipalities,” the researchers write.
They add that while this is admittedly speculative, the findings suggest that the closeness of polls could have had an effect on Donald Trump’s victory over Hillary Clinton. Though major polls all suggested Clinton would win, those conducted by more-right-leaning news organizations had the race much closer than those that lean left. It’s possible that Trump supporters saw an opportunity, while some potential Clinton voters were overconfident in a win and stayed home.