Is the deepening animosity between Democrats and Republicans based on genuine differences over policy and ideology or is it a form of tribal warfare rooted in an atavistic us-versus-them mentality?
Is American political conflict relatively content-free — emotionally motivated electoral competition — or is it primarily a war of ideas, a matter of feuding visions both of what America is and what it should become?
Jonathan Rauch, a senior fellow at Brookings, recently put the issue this way in an essay at the National Affairs website: “Here we reach an interesting, if somewhat surreal, question. What if, to some significant extent, the increase in partisanship is not really about anything?”
This debate has both strategic and substantive consequences. If left and right are split mainly because of differences over policy, the chances of achieving compromise and overcoming gridlock are higher than if the two sides believe that their values, their freedom, their right to express themselves, their very identity, are all at stake. It’s easier to bend on principle than to give up a piece of yourself.
The 2020 presidential contest has taken on the attributes, in the words of Michael Anton, a senior fellow at the Claremont Institute, of a “Flight 93 Election,” an election with potentially devastating consequence for the loser.
If the opposition wins, Anton wrote about the 2016 election — in a view that holds even more true for the 2020 contest — the assault on one’s own values and principles will be worse than any “of us have yet imagined in our darkest moments. Nor is even that the worst. It will be coupled with a level of vindictive persecution against resistance and dissent hitherto unseen in the supposedly liberal West.” Anton was writing from the right, but the same apocalyptic fear of the consequences of defeat applies to the left.
The dispute over the nature and origins of partisanship is a major issue within contemporary academic political science, with enormous practical consequences.
“This is probably one of the most debated questions among people who study American political behavior,” Steven Webster, a political scientist at the Washington University in St. Louis, told me.
Lilliana Mason, a political scientist at the University of Maryland, is a leading scholar of the us-versus-them school, which has come to be known as “affective partisanship.” She sets out her argument in “Uncivil Agreement: How Politics Became Our Identity.”
“Group victory is a powerful prize,” Mason writes, “and American partisans have increasingly seen that as more important than the practical matter of governing a nation.”
She goes on:
American partisans today are prone to stereotyping, prejudice, and emotional volatility. Rather than simply disagreeing over policy outcomes, we are increasingly blind to our commonalities, seeing each other only as two teams fighting for a trophy.
Shanto Iyengar, a political scientist at Stanford and a pioneer in the study of affective partisanship, put it this way in response to my email:
There is a growing body of work showing that policy preferences are driven more by partisans’ eagerness to support their party rather than considered analysis of the pros and cons of opposing positions on any given issue.
Iyengar cited a paper by James N. Druckman, Erik Peterson and Rune Slothuus, political scientists at Northwestern, Texas A&M and Aarhus Universities, “How Elite Partisan Polarization Affects Public Opinion Formation,” that found
stark evidence that polarized environments fundamentally change how citizens make decisions. Specifically, polarization intensifies the impact of party endorsements on opinions, decreases the impact of substantive information and, perhaps ironically, stimulates greater confidence in those — less substantively grounded — opinions.
Instead of voters making reasoned policy choices, “party endorsements carry the day,” the authors note, adding that “elite polarization fundamentally changes the manner in which citizens make decisions.”
I asked Druckman by email about the basis for partisanship, and he replied,
the evidence is clear that, over the past twenty years, partisan emotions have splintered such that people feel more attached to their party and more animus toward the other party. A likely effect is that when partisan elites are also separated, policy substance becomes less relevant. So yes, I think it is clear emotions have increased and this has the potential to undermine substance.
Alan Abramowitz, a political scientist at Emory, is a leader of the opposing camp. Instead of partisanship propelling ideological and policy decisions, Abramowitz argues that
Policy and ideological differences are the primary drivers of polarization. Democratic and Republican voters today hold far more distinctive views across a wide range of issues than they did in the past. And it is among those Democrats and Republicans who hold views typical for their party, that is liberal Democrats and conservative Republicans, that dislike of the opposing party is strongest.
Webster, writing with Abramowitz, argued in a 2017 paper, “The Ideological Foundations of Affective Polarization in the U.S.” that ideology plays the central role in partisanship. Citing American National Election Studies data, Webster and Abramowitz contend that
Opinions on social welfare issues have become increasingly consistent and divided along party lines and social welfare ideology is now strongly related to feelings about the opposing party and its leaders.” In addition, Webster and Abramowitz tell us, survey experiments demonstrate “that ideological distance strongly influences feelings toward opposing party candidates and the party as a whole.
In an email, Peterson cited further evidence supporting the importance of policy and ideological conviction. “There is a substantial component of partisanship that reflects sincerely held issue opinions,” Peterson wrote, pointing out that
when I’ve asked partisans to evaluate hypothetical co-partisan politicians who do not toe the party line on highly salient issues (e.g., Obamacare repeal, assault weapon bans), they are substantially less supportive of candidates who they disagree with on these core issues.
In other words, partisans’ beliefs trumped their loyalty to party. Similarly, Peterson wrote,
on salient issues (e.g., abortion, the appropriate threshold for the minimum wage), partisans are also more than willing to disapprove of policy proposals that come from a co-partisan when they disagree with the content of the policy.
These patterns, Peterson stressed, “are not something that would be expected if partisanship swamped all other considerations.”
At the same time, however, there is intriguing evidence pointing to the malleability of voters’ views on key issues in response to partisan pressure — evidence, in other words, that voters are willing to change their stance in order to conform to the views of fellow partisans.
In a March 2019 study, “White People’s Racial Attitudes Are Changing to Match Partisanship,” Andrew Engelhardt, a political scientist at Brown, explored changing views among whites from 2016 to 2018, based on surveys conducted by American National Election Studies. Tracking a measure of white views of black Americans, he found that there has been:
A profound shift in whites’ evaluations of black Americans in just a two-year period. The modal white Democrat moves from placing at the scale’s midpoint in 2016 to locating at the scale’s minimum (least racially resentful) in 2018. For Republicans, the modal respondent still places at the scale’s maximum (most resentful), but the percentage of white Republicans here increases from 14 percent to 21 percent. While these shifts may seem small given the scale, I show below they represent a rather substantial change on a measure that has otherwise evolved quite slowly since the 1980s.
On a separate American National Election Studies thermometer measure of group favorability, where 0 is very unfavorable and 100 is very favorable, Engelhardt reported that:
In 2018 Democrats rated black people at a 77, up 7 points from 2016. But they rated whites at a 70, a 2 point decrease. Republicans’ feelings about black people improved slightly (64 vs. 69) in these two years but this was far outpaced by increased warmth toward white people (74 vs. 81). While Republicans consistently feel more positively about white people than black people, white Democrats’ attitudes look quite different. White Democrats now feel more warmly toward black people than white people.
Engelhardt’s findings lend support to the views of Alexander Theodoridis, a political scientist at the University of California-Merced, who contended in an email that
For most people, party identity appears to be far more central and salient than particular issue positions. We see increasing evidence of people adjusting their issue positions or priorities to fit their party allegiance, more than the reverse. We are very good at rationalizing away cognitive dissonance. More important than this chicken-or-egg question is the reality that ideology and party have become very highly sorted today. Liberal and Conservative are now tantamount to Democrat and Republican, respectively. That was not always the case. Furthermore, all sorts of descriptive and dispositional features (ranging from religion and race to personality type and worldview) are also more correlated with political party than they were in the past. All this heightens the us-versus-them nature of modern hyperpolarization.
This debate is sometimes framed in either-or terms, but the argument is less a matter of direct conflict and more a matter of emphasis and nuance.
Yphtach Lelkes, a professor of political communication at the University of Pennsylvania, wrote me that “ideology and partisanship are very hard, and likely impossible, to disentangle,” but, he argued, the larger pattern appears to be that
while both seem to be occurring, ideology driving partisanship only seems to be occurring among those that are most aware of politics, while partisanship driving ideology seems to be happening among everyone.
Similarly, Leonie Huddy, a political scientist at SUNY-Stony Brook, wrote me that the debate “is more complicated than simple tribalism versus consistent ideology.”
There is “clear evidence of partisan tribalism,” Huddy observed, “especially when it comes to a potential win or loss on matters such as impeachment, presidential elections, and policy issues central to electoral victory or defeat,” but at the same time
Democrats and Republicans have become increasingly divided on social, moral, and group-linked issues and are less likely to follow the party on these matters.” She pointed out that the tribal loyalty of many Republican voters would be pushed beyond the breaking point if the party abandoned its opposition to abortion, just as it is “difficult to imagine feminist women continuing to support the Democratic Party if it abandoned its pro-choice position on abortion.
While both sides in the debate over “affective” versus “ideological” partisanship marshal reams of survey data in support of their positions — often data from the same surveys — one thing both sides are in full agreement on is that partisan hostility has reached new heights. This is reflected in two recent papers, one by Abramowitz and Webster, “Negative Partisanship: Why Americans Dislike Parties But Behave Like Rabid Partisans” and one by Nathan Kalmoe, a political scientist at Louisiana State, and Lilliana Mason, “Lethal Mass Partisanship.”
Negative partisanship — based on animosity toward the opposition party, not love of your own — turns out to be one of the crucial factors in the outcome of recent elections and it will almost certainly be a key factor going into the next election.
An astute Democratic strategist, who did not have authorization to speak on the record, sent me calculations from the 2016 and 2018 elections showing that the overwhelming majority of voters hold the opposition party in contempt. They are immovable, in his opinion, and impossible to convert: “We’re seeing anti-party sorting — an increasing number of voters are rejecting at least one of the parties, and they are doing so more strenuously,” he said by email. There are, he continued, more voters who
have a very negative opinion of just one party (87 percent) than identify with one of the parties (67 percent). So, negative partisanship explains the behavior of many more voters.
In addition, “negative partisans vote more consistently against the opposite party than partisans vote for their party.”
The remaining “persuadables” — an estimated 13 percent of voters, with little or no partisan commitment — will play a central role in determining the outcome in 2020.
My source cited polling data from a “consortium of Democratic groups” showing that in 2016 the small fraction of the electorate made up of persuadables voted for Trump 41-36, but in 2018 they voted for Democratic House candidates 57-41. At the moment, he said, polling shows that these swing voters currently prefer a generic Democrat to Trump 54-28, with 19 percent undecided.
I asked the strategist how he expects this volatile group to make up its collective mind in 2020. His answer should not provide comfort to anti-Trump partisans:
Your question is among the most urgent ones facing Democratic strategists. There won’t be a single answer — that group is not a monolith, and who the Democratic nominee is will make a difference. That said, we’re not there yet. Unfortunately, too many Democratic strategists with the most money to spend are still using content development practices that don’t match what we know about those voters.
At the moment, he said, “no one — including political commentators — has evidence-based answers to your question of what will move this group (or any other definition of ‘swing’ voters).”
In other words, in an unpredictable world of intensifying partisanship and rancor, Democratic strategists — and Republican strategists too — are pretty much flying blind.
The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips. And here’s our email: firstname.lastname@example.org.