Monday, November 09, 2020

In the 2020 Elections, Partisan Stability Defeats Expectations of Change

Imagine someone who set off into the Siberian wilderness right after the 2018 midterms, didn't experience any of the 2020 campaign, and returned to civilization today to see the outcome of last week's election. This person would come back to the following results: 

• A close national presidential contest decided by crucial victories in Wisconsin, Michigan, and Pennsylvania.

• A set of Senate races that almost perfectly mirrored the presidential results in each state, with the lone exception of Republican moderate Susan Collins outrunning the rest of her party's candidates in Maine.

• A mostly incumbent-friendly set of House elections, with a few of the 2018 Democratic wave's biggest upset winners—Kendra Horn of Oklahoma, Joe Cunningham of South Carolina, Xochitl Torres Small of New Mexico—unable to defend their Republican-leaning seats when facing a presidential-year electorate.

• A continuation of the ongoing pro-Democratic shift in the nation's largest metropolitan areas, pushing former "red states" Arizona and Georgia into full partisan competitiveness and reducing the size of the Republican statewide advantage in Texas.

• A partially countervailing solidification of Republican electoral strength in small towns and rural areas nearly everywhere in the country except New England, keeping key states like Wisconsin and Pennsylvania highly competitive while shifting Ohio and Iowa further toward the GOP.


If this person were a Democrat, he or she might be perfectly content with an election in which Donald Trump was defeated and Nancy Pelosi's House majority remained intact. Or maybe there would be some dismay that the popular repudiation of Trump was not strong enough to deliver Democrats an outright national landslide that handed the party clear control of the Senate. Likewise, a Republican supporter might mourn Trump's defeat—or conclude that, under the circumstances, the outcome could have been even worse. 

But there would be no reason for our returning wanderer to be surprised about any of the major results of the 2020 election. The outcome represents a doggedly consistent continuation of the basic electoral fundamentals of the last decade or more: closely-matched popular support for each party, severe and growing geographic polarization, rampant straight-ticket voting, and important Republican structural advantages in the electoral college and congressional apportionment, especially in the Senate.

Those advantages meant that Biden's victory over Donald Trump was in fact very narrow; a shift of 1 percent of the vote in Georgia, Wisconsin, and Arizona would have re-elected Trump despite Biden's comfortable lead in the national popular vote. But they also increase the magnitude of Biden's achievement: the defeat of an elected incumbent president for only the fourth time in the last century, and the ejection of a party from control of the White House after just four years in power for only the second time since 1896. And the pro-Republican tilt of our electoral institutions helps to explain why Democrats were unable to grow their margin in the House of Representatives or win a majority in the Senate, which wound up tempering some party members' delight in the election results.

That disappointment was mostly a testament to the power of pre-election polling to sway expectations. Thanks to overly rosy survey results, optimistic Democrats had visions of revolutionary success dancing in their heads: Senate victories in Montana, Kansas, and even South Carolina; breakthroughs in House races in Indiana and Missouri; even Texas turning "blue" for the first time in 44 years. This wasn't just partisan daydreaming; Republican operatives told journalists that their own private polls were as potentially devastating to their party as the publicly-released data.

There's a lesson here beyond the obvious need to re-evaluate the trustworthiness of polling methodology and forecasting models. For all the ways in which American politics seems to have entered a period of rapid and disorienting change, the partisan preferences of the electorate have only become more and more entrenched over time. Events and developments that might seem inevitably transformative, from Trump's election to the COVID epidemic to piles of small-dollar donations raised by congressional candidates from Maine to Alaska, have repeatedly proven to have only minor effects on the voting choices of citizens loyally committed to their existing partisan teams. If we simply assume that this electoral stability will prevail until proven otherwise, we are much less likely to be surprised by what we see.