Donald Trump got a pretty good “bounce” in the polls from his convention in Cleveland. We’ll know by early next week if Hillary Clinton can say the same.
But history offers a cautionary lesson that some convention bounces are like young love in the early spring: They just don’t last. As Harry Enten shows at FiveThirtyEight, presidential candidates’ net favorability ratings often rise or plunge between the conventions and Election Day. And some famously large convention bounces were really misleading when the deal went down.
One such bounce was in fact so chimerical that it’s now puzzling it existed at all. The 1980 Democratic Convention that renominated Jimmy Carter is now remembered as a rolling disaster, which began with an effort to dump the sitting president, continued with a speech by losing primary candidate Ted Kennedy that upstaged the nominee, and then concluded with Carter chasing Kennedy around the stage pursuing in vain the traditional clasped-hands unity gesture.
But guess what? Carter’s net favorability rating rose 24 points between the beginning and end of the two conventions that year. He was unpopular earlier and unpopular on Election Day, but for a while there the sun really shined on the 39th president that year.
A more recent and less dramatic example of this dynamic was in 2008, when John McCain got a net 8-point convention advantage, drawing even with Barack Obama before both Sarah Palin and the U.S. economy imploded.
So yeah, the bounces are important, and we are all in a perfectly appropriate habit of beginning to pay attention to polling once the conventions have ended and we are truly into the general-election season. But you cannot take bounces to the bank, and particularly in a year when the conventions are relatively early, the numbers can turn on a candidate like an old love gone sour.