Second, the makeup of the hater vote likely differs from that of 2016. With Trump’s higher favorability among Republicans, fewer Republican-leaning voters are likely to fall into this group; instead, there’s probably a larger share of voters who lean Democratic than in 2016. For instance, it’s possible that a disproportionate share of left-leaning voters who backed a Biden rival in the Democratic presidential primary, such as Sen. Bernie Sanders, falls in this group. The same could be said for independents who don’t like either candidate. The question is whether these haters will disproportionately vote for one candidate, as they did in 2016.

It’s possible that these voters will break for Biden, as the NBC News/Wall Street Journal and Morning Consult polls suggest, because they view the election as a referendum on Trump. Yet it’s also possible that these voters will be more divided on whom they back by Election Day — 41 percent said they were still undecided or had no opinion in the Morning Consult poll, and 30 percent didn’t back Biden or Trump in the NBC News/Wall Street Journal poll.

The share of haters in the electorate could grow or shrink in the coming months, too.