Negative partisanship may be the most toxic form of polarization

This week, we are celebrating the fifth anniversary of the Mischiefs of Faction political science blog, which I was incredibly lucky to join in the spring of 2014. I’d like to take this opportunity to reflect on how my views on party polarization have changed over the past year.

Gradually since the 1970s, and accelerating since the ’90s, the American party system has become more polarized in several ways. First, it has become much more ideologically sorted. Among politicians, activists, and the mass public, liberals are much more likely to be Democrats and conservatives to be Republicans than they were decades ago.

Second, Democrats and Republicans simply like each other much less. You can see this in the mass public, where people are more likely to say they feel “cold” toward members of the other party and to hope a family member doesn’t marry someone from the other party. And it is evident among politicians, who have more trouble making compromises to pass bipartisan bills, as well as interact, and simply like each other less. A recent extreme example of this was the physical threats exchanged last weekend between members of the Texas state legislature.

It is less clear whether the politicians and members of the public are actually taking more extreme policy positions than they did in the mid-20th century. Comparing the extremism of issue positions in very different political eras is complicated and can involve debates over advanced statistical methods. Also, the answer almost certainly varies by issue area. But even setting this question aside, it is clear that the American party system has become better ideologically sorted and that there is more animosity between the parties than in the Cold War era.

Is this a problem? The typical political science answer five years ago was that a democracy could accommodate extremely polarized parties as long as it had the right institutions. Polarization may be causing problems in the US, but that is only because we have a Madisonian system that only works when politicians are willing to work together. Power is divided between Congress, the presidency, and the courts, which are often controlled by different parties. Supermajority rules in the Senate increase the need for the parties to work together if they hope to get anything done.

By this logic, our problems are caused by presidentialism, the Senate’s rules, and perhaps too strong judicial review. We could accommodate more polarized parties if we had a unicameral parliamentary system, in which the parliament elected a prime minister and Cabinet to rule until the next election. (This would presumably solve our problems, whether legislators continued to be selected in single-member districts, as in the UK, Canada and Australia, or by voters choosing among party lists, as in Italy or Israel.)

While coalition governments are necessary in this system when no party wins a majority of seats, evidence from around the world seemed to suggest that these negotiations before governments are formed were less likely to lead to the entire political system collapsing than systems where different branches are in constant conflict. This was roughly my view before 2016.

My views have changed. I still think that presidential systems produce their own “perils,” but I no longer think a system with fewer veto points can solve our problems. Specifically, the election of Donald Trump has led me to conclude that, regardless of our political rules, negative partisanship among politicians and the mass public is a serious danger.

How did the Republican Party and the United States end up saddled with a president like Donald Trump? He reflects some existing trends in the Republican Party but in many ways breaks with them. Prior to Trump, there was an internal party fight about immigration, with one faction supporting moderate reforms of the immigration system and a more friendly tone toward Hispanic and Islamic immigrants. No less than the Republicans’ last president, George W. Bush, supported this approach.

Trump has decisively endorsed the anti-immigration and anti-Islamic wing if the party. Yet in areas like civil rights, voting rights, health care policy, reproductive rights, etc., he has allied himself with the conservative wing of the party, but his positions reflect the dominant thinking in the party. One could imagine a nominee like Ted Cruz or Marco Rubio putting forth similar policies.

However, Trump brings a host of politically toxic behaviors that are unique to him. He appears to plan to use his office to accept bribes from domestic and foreign sources through his global business empire. He endangers American alliances and leadership in Europe and Asia. He insults allies often for no strategic purpose whatsoever. And he has such limited cognitive capacity that he can only absorb one-page foreign policy memos, without any nuance, and needs briefers to constantly mention his name to keep him interested. There is a substantial probability that his presidency will continue to be consumed by scandal, most likely bribery, obstruction of justice, or whatever else is going on with Russia.

None of these types of behaviors were part of Republican or conservative ideology prior to Trump. These things distract politicians in Washington and depress the president’s approval rating, making it harder for Republicans to enact conservative policy change that the Trump administration ostensibly supports. It is likely that if Cruz or Rubio were president, he would be able to move policy further to the right.

This is why hardly any Republican politicians endorsed Trump in the primaries. He is a poor vessel for their ideas. Why didn’t his weakness as a Republican standard-bearer hurt him more in the general election?

Very few Republicans, and essentially no Republican national politicians, endorsed Hillary Clinton because of their reservations about Trump. As comparison, when their party nominated George McGovern in 1972, prominent Democrats such as former Texas Gov. John Connally, Virginia Gov. Mills E. Godwin, Jr., and Boston Mayor John F. Collins endorsed Nixon. As core a Democratic interest group as the AFL-CIO refused to endorse McGovern and remained neutral. In this year’s French presidential runoff between pro-EU globalist Emmanuel Macron and far-right candidate Marine Le Pen, the major center-right party, the Republicans, and their presidential nominee, François Fillon, endorsed Macron.

The major reason the 2016 US presidential election didn’t go that way was negative partisanship. This is the tendency to vote for a party not mainly because you like it, but because you are repulsed by the other major party. Democrats vote for their party more because they want to prevent Republicans from winning than because of affection for their party’s candidates, and vice versa. The logic of negative partisanship is unaffected when your party nominates a weak candidate.

Trump was a very unusual Republican candidate. He tacked on a lot of unconventional positions and scandals on top of some standard conservative positions. And his history of wildly inconsistent positions made it uncertain if he would even stick through his whole term with the conservative positions he did support.

But on the other side, the Democrats nominated a person who was tightly associated with the party and liberalism. Clinton has been a national figure since her husband ran for president in 1992. Even beyond that, she has been practically a Zelig of modern liberal politics. She was a staff member on the congressional Watergate committee. She originally kept her maiden name (as many feminists did in the 1970s), but changed it after Bill Clinton lost his gubernatorial reelection race in 1980, which some in Arkansas blamed on Hillary’s feminism.

She was controversial for her liberalism and feminism in the 1992 presidential campaign. The presidential first lady cookie-baking contest, which we now see as a staple of presidential campaign, actually began in 1992 when Family Circle magazine wanted to put Clinton in her place after she was quoted saying, “I could have just stayed home and baked cookies and had teas, but what I decide to do was to fulfill my profession. ...” She was first lady with a Democratic president, a prominent Democratic senator, and secretary of state in a Democratic administration. The Citizens United Supreme Court decision, which has a significant issue in 2016, was a case about campaign spending attacking her.

The big question in the 2016 election was whether Trump would underperform a traditional Republican. Historically, it is hard for an American party to win a third presidential term. It has only happened once (1988) since World War II. Had the Republicans thrown away a great opportunity by nominating Trump? It turns out the answer was no.

The election turned out approximately as one would expect based on moderate economic growth in 2016 and a party running for its third term in the White House. Republicans didn’t suffer an electoral penalty for nominating Trump. The biggest reason is that voters who usually support Republicans almost all “came home” to favoring Trump on Election Day. A similar phenomenon happened at the elite level. Most national Republican politicians (Paul Ryan, Reince Priebus, Mitch McConnell, most senators and House members) endorsed Trump. Those few Republicans who didn’t endorse Trump (John Kasich, Jeb and George W. Bush) didn’t endorse Clinton.

Negative partisanship swayed Republicans at the mass and elite level. Many Republicans voted for their party’s nominee primarily in order to avoid a Clinton presidency. Clinton, with her high visibility and close connection with liberalism, is almost ideally suited to activating Republicans’ traditional partisan and ideological loyalties.

The country would be substantially better off if the electorate penalized parties for nominating inexperienced, uniformed, impulsive, corrupt candidates for president. Whether you are a conservative or a liberal, you would be better off if the Republicans in 2016 had nominated and elected Marco Rubio, Ted Cruz, or even Mike Pence. One of them would implement many of the same policies, but without the massive corruption, the degradation of American political institutions, the danger of starting a major military conflict by accident or incompetence rather than ideology, or the many other Trump specific pathologies.

It could be that Clinton was unusual in her ability to spark negative partisanship in Republicans. A 2016 Democratic nominee who was newer to the national stage might not have triggered so much negative partisanship. But the pattern is troubling. We would all be better off, regardless of our ideology, if parties had the incentive to nominate high-quality candidates for president. If they were punished, they would work hard to stop these candidates from winning their nominations, even possibly reforming nomination systems to try to make it less likely. That might have happened on the Republican side if Trump had lost. But as things stand now, parties have less incentive to block low-quality candidates.

Changing the electoral system (or supermajority rules) won’t stop this problem. Many parliamentary parties already have rules that prevent outsiders from leading the party. (For an exception, see the new rules the British Labour Party instituted that allowed Jeremy Corbyn to become leader.) Regardless of the system, those with influence in parties have less incentive to choose smart, temperamentally fit leaders when they know it won’t affect their vote total. Their voters are primarily motivated by the opposing candidate. To address the pathologies of polarization, we need to think harder about developing cures for negative polarization.