One example is the almost universal prediction that Clinton would win. This (unfortunately) has the effect of suppressing opponent-voter turnout, as too many v

One example is the almost universal prediction that Clinton would win. This (unfortunately) has the effect of suppressing opponent-voter turnout, as too many voters care about being on the "winning" side. This was well-demonstrated in the 1994 election. As news of the Republican sweep (also not widely predicted) spread, Democrat turnout in Western time zones was reduced, favoring downballot GOP candidates. This was so pronounced in California, the GOP actually took the majority in the Assembly (a situation otherwise unheard-of in modern California). District 28 was decided by only 389 votes.