At this point in the 1976 U.S. presidential race, Jimmy Carter had an 11-point lead on Gerald Ford, down from his 33-point lead in July. His lead would further narrow such that the election was a cliffhanger. Carter won by the narrow margin of 2 percentage points. Although the vote swing in the Carter-Ford race was unusually high, large swings among voters were not uncommon in earlier times, marking also the 1980, 1988 and 1992 campaigns.
Ever since, the candidates’ poll standings in early September have varied only marginally as the campaign has worked its way toward the general election. In early September of 2012, for instance, Barack Obama had a 2-point lead, which stayed in that range through much of October before ending with a 4-point Obama victory.
The notion that the 2020 campaign is a referendum on Donald Trump’s presidency has been a persistent storyline of this year’s campaign coverage. And properly so. A strong correlation exists between Americans’ approval of Trump’s presidency and whether they plan to vote for him.
But there’s a second-level storyline that’s often missing in the analysis, and it extends back two decades, the time at which wide swings in voter support disappeared during presidential general elections. It’s the story of party polarization, which began in the 1980s and was firmly in place by the early 2000s. It was marked by a widening of the partisan gap but also a strengthening of partisanship and antipathy toward the opposing party.
The hostility that many partisans have for the other party is a larger driver of the vote than might be assumed. Party identification was once the best predictor of how people would vote on election day — Democrats lining up behind their party’s candidate and Republicans backing their party’s nominee. But party identification no longer has that distinction. When Alan Abramowitz and Steven Webster examined post-1990s elections, they found that “ratings of the opposing party were by far the strongest predictor” of vote choice. “The greatest concern of party supporters,” they write, “is preventing the opposing party from gaining power.”
Today’s presidential nominees face a double hurdle in trying to woo voters. They need not only convince them that they’re the better choice but also convince them that the party they represent is an acceptable choice. To be sure, most undecided voters are weak partisans, so their loyalty to the party they typically prefer is not insurmountable. What complicates the effort to convert them is that they tend to hold negative views of both parties — a tendency that is relatively new and coincides with the advent of heightened and divisive partisanship among political elites.
The power of the partisan divide is exemplified in Trump’s unwavering support from his party’s supporters and constant disapproval among Democrats. His approval ratings from nearly his first days in office have varied within a narrow range. Some pundits have found it remarkable but it’s actually part of a larger trend. It is no surprise, of course, that Republicans and Democrats would judge a president differently. That’s been true in every Gallup poll since the question was first asked in 1945. But the gap is wider now than at any time in the past. Through the 1970s, those who identified with the president’s party were about 30 percentage points more likely to approve of his performance. The gap steadily widened thereafter, climbing to an average of 70 points during Obama’s presidency, when 85% of Democrats but only 15% of Republicans approved of the job he was doing. From the day a president takes office, Americans of the other party have decided that nothing the president can do will be acceptable or, for those of the president’s party, unacceptable.
So how many undecided voters are there at the moment and how will they respond to Trump and Biden in the campaign’s remaining weeks? They number less than 10% of self-described likely voters, down even from recent campaigns and a mere third of the number of presidential campaigns of a few decades ago. It’s unlikely that either candidate will capture a super-majority of the undecided. Typically, the split is within a narrow range, enough to tip a razor-thin election but otherwise a footnote in post-campaign analyses.
A disruptive event could upend that expectation. In the closing days of the 2016 presidential campaign, FBI director Jim Comey’s announcement that he was reopening the investigation of Hillary Clinton’s emails cost her the election. Trump had the edge with late-deciding voters. Such events are unpredictable in their timing and also in which party they help, although Trump is trying to change that by getting the FDA to fast track a coronavirus vaccine, hoping that its availability — or even the announcement of its availability — will sway the late deciders.
Thomas E. Patterson is Bradlee Professor of Government & the Press at Harvard’s Kennedy School and author of the recently published Is the Republican Party Destroying Itself? Journalist’s Resource plans to post a new installment of his Election Beat 2020 series every week leading up to the 2020 U.S. election. Patterson can be contacted at firstname.lastname@example.org.
Alan I. Abramowitz and Steven Webster, “The Rise of Negative Partisanship and the Nationalization of U.S. Elections in the 21st Century,” Electoral Studies 41 (2016): 16, 21.
Ezra Klein, Why We’re Polarized Simon & Schuster, 2020.
Wiilam Mayer, ed., The Swing Voter in American Politics, Brookings Institution Press, 2008.
Geoffrey Skelley, “Just How Many Swing Voters Are There?” FiveThirtyEight, September 19, 2019.