Of five false conspiracy theories that Kennedy has promoted all five generated far more agreement among Trump voters than among Harris voters: that
COVID-19 vaccines are more harmful than the virus itself (55% vs. 8%); that climate change is being used as a
pretext for imposing totalitarian controls on society (68% vs. 7%); that Prozac and other antidepressants have led to a rise in school shootings (35% vs. 12%); that vaccines cause autism (25% vs. 5%); and that chemicals in the water supply could turn children transgender (8% vs. 4%).
So it’s no surprise that when poll respondents are asked how they would vote both with and without Kennedy on the ballot, about 55% of Kennedy’s remaining voters break for Trump, on
average, and about 45% break for Harris.
That split means Kennedy's exit probably won’t affect the election in a dramatic way. If Trump were to immediately gain 55% of Kennedy’s remaining voters, explains
data journalist Nate Silver, Harris’s average national lead would shrink from 2.5% to 2.1%.
In reality, the effect might be even smaller. Disaffected by definition, Republican-leaning Kennedy voters have had ample opportunity to support Trump in the past; instead of swinging his way, they could back a different third-party candidate, like Libertarian Chase Oliver. Or they could stay home in November.
Still, small margins in key swing states have decided both of Trump’s previous presidential elections. That could happen again — and if it does, Kennedy’s decision this week could make a (minor) difference.