LAST UPDATED 26 August
On Friday, August 23, independent candidate Robert F. Kennedy Jr. announced he was suspending his presidential campaign and endorsed former President Donald Trump. Kennedy's support, which we project to be 4 percent of the vote, could provide a decisive boost to Trump in swing states. Our model predicted a 4-in-10 chance that Kennedy's share of the vote in November will be larger than that of Trump or Vice President Kamala Harris in at least one pivotal state. Both candidates would likely like to see that probability drop to zero.
However, it is difficult to quickly analyze RFK's influence on the race. On the one hand, his influence on our polling average is small, and Kennedy's voters otherwise mostly look like strong partisans. But it is not so easy for a forecasting model to react to such a sharp change in the race, mainly because it needs data (from the polls) to do so. While you and I know that Kennedy supported Trump, which should theoretically help him a little, the model does not have access to that information until it sees polling data showing such a change. If we were to program the model differently, we would have to speculate, largely with ad hoc rationalizations, about the future course of the race. We prefer to leave such forecasts to the model. Readers should expect our model to react to this news over the course of the next week if Kennedy's support for Trump does indeed change the polls.
538's forecast is based on a combination of polls and campaign “fundamentals” such as the economy, party affiliation and incumbency. It is not intended to “declare” a winner, but rather to give you a sense of how likely each candidate is to win. Check out our methodology to learn exactly how we calculate these probabilities.
—G. Elliott Morris
