Dem
51
Ties
1
GOP
48
Political polling used to be about informing Americans how candidates were doing and which demographic groups liked which candidate on which issues. Now the polling is influencing how candidates act. When Mandela Barnes dropped out of the race for governor of Wisconsin, he said: "It has become very clear who our nominee is going to be." He meant Francesca Hong—on account of the polls showing her 15-20 points ahead—but the polls were wrong and she lost. If the polls had been accurate, he might not have dropped out. Thus, incorrect polls are now not just giving voters the wrong impression, they are causing politicians to make decisions that might not have made in the absence of the poll. Incorrect polls can thus have important real world effects other than confusing voters who follow them.
Also increasingly important are how donors respond to polls. When donors see a landslide on either side, that is a signal not to donate and to move on. When polls are off by 15 or more points, as they were in the Michigan Senate race and Wisconsin gubernatorial race, viable candidates may be starved of funds in the home stretch because donors believe its all over but the shoutin'—even when it really is not.
Given how politicians and donors respond to polls, there is more motivation than ever for malign actors to publish fake polls. For example, a candidate who has enough money already could arrange for an ally to announce a fake poll showing him 20 points ahead to convince donors on both sides (especially the other side) to stop donating.
Candidates and political operatives understand this. Lis Smith, the former paramour of disgraced ex-governor Eliot Spitzer and a Democratic strategist who worked for Pete Buttigieg in 2020, said: "Every candidate in 2028 will have their own cooked poll factory. Polls now are not just a way of measuring the horse race. It's a way for candidates and their allies to influence the horse race."
Prediction markets are now introducing a new problem: Fake polling to make money. Since it is now possible to bet on elections, and bettors use polls to guide them, there is potential for a fake poll to influence markets, almost like insider trading. For example, someone could buy a contract saying that candidate [X] will win some nomination, then release a fake poll showing [X] way ahead. This will cause more people to place a bet on [X] and make the price go up. The "pollster" can then sell the contract at a higher price and get out before another poll brings the price back to earth. Alternatively, buy a contract that [X] will lose, then release a fake poll showing [X] 20 points behind. When the price of that contract goes up, sell it. This is effectively shorting the candidate. The prediction markets offer a way to monetize fake polls. This is not a good incentive and requires media companies and aggregators to be very careful about who they count as real. (V)