It’s the Negative Partisanship, Stupid

I’m trying to decide whether to trust Rachel Bitecofer’s model. She’s a political scientist at Christopher Newport University, a small (enrollment 5,000) liberal arts school in Virginia. In September 2018, while better known analysts, including Nate Silver, were unsure whether the Democrats would win the 23 seats in the House of Representatives necessary to gain a majority, Bitecofer boldly predicted they would win 42 seats. They wound up winning 41.

Impressive debut.

If I understand correctly, Bitecofer’s model flows from the following assumptions:

  1. The national electorate is not static. Each election brings out different voters.
  2. Voters are primarily motivated by “negative partisanship,” i.e. an urgent desire to defeat the other party.
  3. Election results therefore are determined not, as widely supposed, by the switch of swing voters from one party to the other, but by “the entrance (and exit) of partisan voters who are activated or deactivated by negative partisanship.”
  4. Negative partisanship helps the party out of power more than the party in power. Members of the party out of power become more motivated to vote; members of the party in power become complacent and less motivated to vote.

Democrats and Democratic-leaning independents were so motivated to beat Trump and the Republicans in 2018 that they turned out in the overwhelming numbers needed to produce that 41-seat gain despite a surge in positive partisanship among Trump voters.

Last July, the confident Bitecofer published her forecast for the 2020 election.

Are you ready for this?

Her model sees Democrats and Democratic-leaning independents still driven to beat Trump and turning out en masse yet again. The Democratic nominee for president, no matter who he or she is, will hold the states Hillary Clinton won in 2016 and take Pennsylvania, Michigan, and Wisconsin, the three states Trump won by tiny margins. The additional electoral votes from those states will give the Democrat 278 overall, clinching victory.

What’s more, Bitecofer’s model rates four states Trump won in 2016 —Arizona, Florida, Iowa, and North Carolina — as toss-ups. And she has the Democrats likely to add to their House majority, picking up as many as nine seats just in Texas, and perhaps taking over the Senate, too.

So you can see why I really want to trust Bitecofer’s model. Not only does it say Trump will lose, but that the identity of his opponent doesn’t matter — meaning I don’t have to sweat so hard over who to vote for in the upcoming California presidential primary. (More on that in the next post.)

There are caveats. Bitecofer cautions that her model can be upset by developments such as “a ground war with Iran, an economic recession, or a terrorist attack on the scale of 9/11.” It can also be disrupted by the emergence of a third-party candidate. Bitecofer maintains that Clinton lost Pennsylvania, Michigan, and Wisconsin not to Trump but to the Libertarian Gary Johnson and the Green Jill Stein, who siphoned more voters from Clinton than Trump in each of those states.

Although Bitecofer’s model implies that the argument about whether Democrats should appeal to their left-wing base or to moderate swing voters is irrelevant, she does note that the Dems will enhance their turnout “in really important places” if “the ticket has a woman, a person of color or a Latino, or a female who is also a person of color” as either the presidential or vice-presidential nominee.

And one more thing. If Bitecofer’s forecast proves correct, legislatively the Democrats will need to move fast, because 2022 will be a big year for Republicans — their turn to be energized by negative partisanship.

Bitecofer was way ahead of the pack in 2018. But on a national level she’s been right only once, which isn’t enough to make me confident she’s truly got election prognostication down to a science. So while I hope she’s right again in 2020, for now I’m not going to relax.

Former Risk Manager at UC Berkeley, author of four books, ectomorphic introvert.