Tuesday, March 29, 2016

Making Sense of the Trump Phenomenon

I've had a few friends ask me for an explanation of the Trump phenomenon and what I think will happen this year (or maybe I'm just making that up to justify this exercise in self-indulgence...). In doing so, my goal is not so much to provide a definitive answer (insert broad statement about the social sciences, etc) but rather to give a sense of the different perspectives on Trump support and what is likely to happen.

Who could possibly explain this?

Defining Trump
A lot of people have debated whether or not Trump is a 'fascist.' The answer is no, though there are obvious overlaps. There is a basic reason why Trump isn't a fascist: Trump, despite his many flaws, accepts democracy as the basic rule of the game; leaders like Mussolini and Hitler sought explicitly to overthrow democracy.

The better comparison is with the European parties of the radical right (as in parties such as the Front National in France or the UK Independence Party in Britain). Here, we do see how Trump fits this comparison. Like the European radical right, Trump advocates strongly nationalist policies, 'tough' approaches to domestic and foreign problems, and positions himself as the only candidate who listens to the 'ordinary people' and has the ability to solve their problems.

Who Votes for Trump?
It has become commonplace to refer to Trump's voters as 'working-class.' But what does that mean these days? The definition of 'working-class' has become muddled in the 21st century as the traditional working class has shrunk.

A few years ago, there was a big debate about this after the publication of Thomas Frank's What's the Matter with Kansas?, in which he argued that the Republican Party had successfully gained the votes of many working-class whites by appealing to cultural issues that 'distracted' from economic concerns. Rebuttals by Bartels and Gelman based on systematic evidence challenged Frank's claims that working-class whites were voting Republican. In short, lower-income white voters do not lean Republican. That is Bartels' and Gelman's point; income and Republican vote propensity are positively correlated. However, using education as a measure gives us a different picture: white voters without a 4-year college degree trend significantly to the Republican Party.

Do they look 'working-class' to you?

But are these voters 'working class'? They mostly fall into the middle-income group. But many of the white voters in the lower third, which has remained more Democrat-leaning, aren't actually working: this includes retirees, the disabled, the long-term unemployed, etc. So, if the problem with the Frank argument is that Republicans aren't actually working class, then the problem with the Bartels/Gelman argument is that many lower-income Democrats aren't really working class. The broader point, though, is that this characterization of white working-class Republicans is misleading, in large parts because present-day economic realities no longer match with ideas of class rooted in the 20th century.

Does education explain Trump support?

But what about Trump supporters? It is unarguably true that they come closer to characterization of working-class Republicans; they are less educated and lower income on average. However, and this is the source of a second conceptual problem, this is only true as long as our reference category is limited to Republican primary voters. If we compare Trump supporters to the electorate at large, then it is not at all clear that they are particularly 'working class.' Instead, your prototypical Trump supporter is someone without a college degree in the middle income ranges--in other words, somebody a bit like 'Joe the Plumber.'

Why Do People Vote for Trump?
So why are the 'Joe the Plumbers' of the world voting for Trump (and UKIP, Front National, etc)? Broadly conceived, most arguments break down into two camps.

The first argument is economic, and tied into that 'working class' notion above. This argument suggests that Trump supporters--like those of the radical right in Europe--are the 'losers' of globalization and technological advances. In this argument, older white working-class men have lost ground economically as a result of 'outsourcing' (greater competition from overseas labor), immigration (greater domestic labor competition), technology (which has made them redundant), and  policy changes that have benefited holders of capital while hurting labor (e.g., the rollback of labor unions, changes in the tax code). Moreover, these policy changes have generally had mainstream bipartisan support--Democrats and Republicans have both contributed to these trends since the 1990s (think of Clinton's support for NAFTA or his liberalization of Wall Street regulations). As a result, these voters have been 'left behind' by the mainstreams of both parties--leaving them open to a candidate like Trump.

There are problems with this as a monocausal explanation. First, the truest losers of globalization don't actually support Trump. These would be the workers living in poor, heavily-minority urban areas (think the South Side of Chicago or Flint, MI) or people in manufacturing- or mining-oriented smaller towns. For the most part, those individuals are largely on benefits or in precarious service-sector jobs. Most don't vote. Those that do tend to vote Democrat (or for left-wing parties in Europe). Trump voters are clearly from a step or two higher up the class ladder. Second, it doesn't explain the particular antipathy of Trump/radical right voters towards immigrants--as opposed to, say, a Bernie Sanders style radical left-wing candidate.

That leads to the other common argument, which is cultural: Trump supporters are motivated by a fear of immigration, growing diversity, etc. Seeing their neighborhoods (and America as a whole) become more diverse has led to a nativist backlash. Plenty of studies have shown that Trump supporters are motivated by ethnocentrism, racism, and hostility towards Muslims. There are a few problems with this claim. First, it's a bit circular to argue (for example) that people support a candidate who calls for stopping immigration because they don't like immigrants. Second, and maybe more important, is that it doesn't really explain why these individuals have such uniquely strong motivations. Finally, it doesn't explain the unique concentration of support among less-educated, lower-middle class voters. Is xenophobia uniquely a phenomenon of these social groups?

Needless to say, both perspectives have merit, but neither perfectly explains Trump support.

The Dispositional Argument
To understand why a certain set of people support Trump (or the European radical right), it helps to consider the role of individual disposition (or personality). Recent studies help in this regard. The excellent Authoritarianism and Polarization in American Politics provides a good starting point.

The idea behind the dispositional argument is that individuals have basic, fundamental orientations towards the world, risk and change, and how society should function. For example, some people are naturally more risk-averse while others are more risk-accepting. Some people prefer change and novelty, while others prefer certainty and predictability. These dispositions tend to go together in coherent ways: risk-averse individuals are more likely to prefer certainty and predictability (and vice versa). These dispositions also predict orientations towards society: people who prefer certainty and predictability in their lives will be wary of big political or social changes. They will prefer stability and tradition, and they will worry more about the threat of disorder or decay. Quite a few studies find that Republicans (or conservatives) display higher needs for security and certainty (NSC), which manifests in greater intolerance to social change and out-groups and a greater willingness to use harsh or punitive measures against threats to the social order. A few studies suggest that Trump supporters uniquely display these higher NSC, even among Republican voters, though this claim has been challenged. That is the 'demand-side' of the Trump electoral phenomenon, and it pretty typical of what we see in Europe.

But there has to be a 'supply side' to this story as well. Political elites construct party competition. The second part of Hetherington & Weiler's account is that Democrat and Republican politicians made a series of strategic moves starting in the 1960s and continuing into the 21st century that brought us to this stage. On each occasion, Republican party elites adopted the position (opposition to further civil rights legislation, support for the Vietnam War, the 'moral majority' agenda, opposition to same-sex marriage, Iraq/the 'war on terror') that attracted those high-NSC individuals. To varying degrees, the Democrats adopted the opposite position. These 'worldview' issues are somewhat precariously aligned with traditional economic issues on taxation, the welfare state, etc. In short, the current alignment of parties is a recent (and, likely, temporary) phenomenon, but it is also one that feeds the current high levels of party polarization--because the parties are divided on basic 'worldview' questions about security, diversity, and change.

(By contrast, the European radical right has mostly risen as a series of new, or re-engineered, parties that have filled a gap in the existing party mainstream. These 'political enterpreneurs' recognized there were a lot of votes to be gained by appealing to anti-immigrant, anti-EU positions, and that no party was capturing those votes.)

This 'supply side' question is important, though, because it provides insight into what may happen. The Republican Party has fused low-tax, small-government economics with socially conservative and nativist social position (and the Democrats roughly vice versa). This is always a potentially uneasy alliance. Part of what Trump has done is successfully drawn the core of those high-NSC Republican voters, but who are not necessarily as supportive of traditional Republican policies to lower taxes and shrink government, while Cruz (and the other candidates) have captured traditional Republican 'small government' voters.

Can Trump Actually Win?
This is the question that many are most concerned about. I will suggest three scenarios, in what I believe to be order of likelihood.

Scenario 1: Trump has no chance
If we take these dispositional arguments seriously, then they suggest a clear 'ceiling' on Trump support. Perhaps 35-40% of the electorate could vote for Trump, and their support for him will be very firm. While he appeals to those high-NSC voters, his rhetoric also repels most other voters. Past American candidates who have run on these sorts of appeals (George Wallace, Ross Perot, Pat Buchanan) have had only limited appeal. Similarly, European radical right parties have shown the ability to consolidate up to 30% of the electorate, but they haven't shown the ability to gain an electoral majority.

All of that is before we even consider Trump's personal qualities. European radical right parties are led by experienced and serious politicians; George Wallace was a very successful politician as well. Trump's offensive rhetoric against women, critical journalists, and his rivals (and their wives) is off-putting to many middle-class Americans. And, frankly, he is a bit of a bullshit artist; he has no deep policy knowledge, and he has assembled an embarrassingly poor group of foreign policy advisers. Can he really survive 5 months of increasingly critical media coverage, daily attacks by the Democrats, and 3 debates with Hillary Clinton? Rather, Democrats will be mobilized to vote against him, independents will not support him, and a fair number of moderate, suburban Republicans will stay at home too. The result will be a Clinton landslide, with big downballot effects too.

In short, only an extreme scenario (economic crisis, series of terror attacks, Clinton scandal) could see Trump win.

Scenario 2: Polarization as Normal
Whenever there is a competitive Democratic primary, pundits start wondering if the supporters of the losing candidate will fall in and support the winner. In 2008, there was a group of Clinton supporters calling themselves PUMAs ("Party Unity, My Ass") who made a lot of media noise about not supporting Obama in the general election. Obviously, that didn't happen. We're hearing it again now with Sanders supporters.

What happened to all those PUMAs? The general election campaign. No matter how contentious the primary campaign is, most partisans will fall in line with their party's candidates when they are reminded of just how much they dislike the other party.

The same thing may well happen with Republican voters this fall. Despite their misgivings, the specter of a Hillary Clinton presidency may be enough to get most Republicans to turn out and vote for Trump. Add to that the high level of polarization (and, particularly, what Abramowitz & Webster call "negative polarization"), and this scenario becomes plausible.

If this occurs, then the 2016 will be similar to every other presidential election in the 21st century. It will be competitive, decided by the same handful of swing states, and it will likely come down to the state of the economy, voters' satisfaction with Obama, and the sense of whether things are going in the right direction or not. In other words, it will be very close.

Scenario 3: Re-Alignment Scenario?
In this scenario, a Trump candidacy breaks through the current party polarization and generates a re-alignment (in progress) of voters. As some pundits have predicted, large numbers of culturally conservative working-class white voters support Trump, while upper middle-class Republicans defect to Clinton. The electorate begins to look inverted by class, with Democrats getting most support from  more affluent but socially liberal (i.e., low-NSC) voters, while Republicans become the party of less affluent, but culturally conservative (high-NSC) voters.

This scenario probably leads to a Clinton victory, but one considerably closer than in Scenario 1. It could also result in an "Electoral College victory" in which Trump gains enough in 'Rust Belt' states like Ohio, Wisconsin, and Michigan to win them, while Clinton mostly just picks up votes in states that Democrats won in 2012, like Colorado and Virginia. Thus, Trump could win the Electoral College vote while losing the popular vote. Wouldn't that be fun? This scenario is tough to project, but I don't think it's terribly likely.

However, this makes more sense as a long-term, post-2016 scenario for how the two parties evolve, than as something that will unfold in the next 6 months.

Scenario 4: Third-Party Chaos
This one is pretty straightforward. Mainstream Republicans decide to sacrifice this election but save their party from something like Scenario 3 in the long term. They nominate a rival candidate, with more traditional Republican credentials, to run against Trump and Clinton in the general election, knowing that this will divide the Republican vote and give Clinton a victory. Like Woodrow Wilson in 1912, Clinton would end up winning comfortably though perhaps not impressively.

I don't consider this likely for various reasons. First, it's rare that mainstream party leaders make a conscious choice knowing that it will cost them an election. It's far more likely that Republican leaders will convince themselves in the coming months that 'Trump won't be so bad' as the nominee. Second, there is a huge problem of coordination: deciding on a rival candidate, organizing the campaign, fund raising, getting that person on the ballot in every state, etc. Third, who would want to run in this role? It would be a bad idea for any Republican with political ambitions to run this way, potentially dividing the party and picking up the stigma of a big electoral loss.

If you're still reading, go pour yourself a drink...