A guest contribution from Arndt Leininger,
Research Fellow at the Chairs for Public Policy and Empirical Political Science from the
Johannes Gutenberg University Mainz.

Most political scientists, just like most pundits and pollsters, failed to see Trump’s victory in the 2016 US presidential election coming. This has led to a range of criticisms leveled at political science, from a rejection of forecasting to wider debates on the direction of the discipline. I am less pessimistic about past, present or future achievements of the discipline. Partly, because the fact that a small part of the community did not predict a single event correctly should not interpreted to be more than that. Forecasting is neither the only nor the primary goal of political science. And lest we not forget, some political scientists did foresee a very close race and even a Trump win months before the election (this is what the first suggested reading is about).

In this blog post I want to highlight that ‚regular day-to-day political science research‘ has a lot to contribute to our understanding of the 2016 US presidential election. This why this reading list is limited to published scholarly work (if you want to read up-to-date analyses and commentary on the election from political scientists I suggest you check out The Monkey Cage). I link to the original article as well as, if available, an ungated version. For the sake of brevity I limit the selection to only half a dozen articles which appeared in peer-reviewed journals. Obviously, such a reading list is neither comprehensive nor is it a selection of the six most relevant pieces. I invite colleagues to suggest further readings in the comments sections.

In focusing on forecasting, polling, sources of the anti-establishment vote, gender, populism and countering hate speech online these articles provide insights that are applicable beyond this year’s US presidential election. For all the sneering at the US, Europe has seen its fair share of electoral successes of right-wing populists and is anxiously looking ahead to the presidential election in France.

1. Political Science models that saw this coming

The forecasts which received most attention in the run-up to the election were poll-based forecasting models, partly because they were updated with each new poll coming in. The first reading is on a different set of forecasting models which focus on so called fundamentals such as presidential approval or economic growth. Simple econometric models based on such fundamentals forecast US presidential elections surprisingly well. This election seems to be no exception as this overview of fundamentals based models suggests. The author of the first reading in this list was ridiculed on social media for predicting a Trump victory – not anymore.

Norpoth, Helmut (2016) „Primary Model Predicts Trump Victory“ PS: Political Science & Politics 49, 655–658. (article, ungated)

However unpredictable the ascent of Donald Trump onto the stage of presidential politics may have been, one forecast model has been highly confident for months that he would win the election on November 8, 2016. The Primary Model predicted on March 7, 2016 that Trump would defeat Hillary Clinton with 87 percent certainty. […] There is nothing to add to or subtract from the March forecast here. It was unconditional, final, and not subject to updating. Just in case Hillary Clinton would not be the Democratic nominee, the Primary Model gave the nod to Trump over Bernie Sanders with 99% certainty[.] What are the ingredients of this forecast model? [… T]he Primary Model relies on presidential primaries as a predictor of the vote in the general election; it also makes use of a swing of the electoral pendulum that is useful for forecasting (http://primarymodel.com/). For the record, the Primary Model, with slight modifications, has correctly predicted the winner of the popular vote in all five presidential elections since it was introduced in 1996[.] In recent elections the forecast has been issued as early as January of the election year. (articles has no abstract, this is an edit of the first three paragraphs)

2. Why are polls so volatile when votes seem so predictable?

If elections are so predictable as Norpoth argues why then is polling through the campaigns campaigns so volatile? The authors of the following piece provide an answer to this question. It seems that the primary role of campaigns seems to be informing voters about fundamentals which is why results are so predictable while polls are so variable.

Gelman, Andrew and King, Gary (1993) „Why Are American Presidential Election Campaign Polls So Variable When Votes Are So Predictable?“, British Journal of Political Science 23, 409–451. (article, ungated)

As most political scientists know, the outcome of the American presidential election can be predicted within a few percentage points (in the popular vote), based on information available months before the election. Thus, the general campaign for president seems irrelevant to the outcome (except in very close elections), despite all the media coverage of campaign strategy. However, it is also well known that the pre-election opinion polls can vary wildly over the campaign, and this variation is generally attributed to events in the campaign. How can campaign events affect people’s opinions on whom they plan to vote for, and yet not affect the outcome of the election? For that matter, why do voters consistently increase their support for a candidate during his nominating convention, even though the conventions are almost entirely predictable events whose effects can be rationally forecast? In this exploratory study, we consider several intuitively appealing, but ultimately wrong. resolutions to this puzzle and discuss our current understanding of what causes opinion polls to fluctuate while reaching a predictable outcome. Our evidence is based on graphical presentation and analysis of over 67,000 individual-level responses from forty-nine commercial polls during the 1988 campaign and many other aggregate poll results from the 1952-92 campaigns. We show that responses to pollsters during the campaign are not generally informed or even, in a sense we describe, ‚rational‘. In contrast, voters decide, based on their enlightened preferences, as formed by the information they have learned during the campaign, as well! as basic political cues such as ideology and party identification, which candidate to support eventually. We cannot prove this conclusion, but we do show that it is consistent with the aggregate forecasts and individual-level opinion poll responses. Based on the enlightened preferences hypothesis, we conclude that the news media have an important effect on the outcome of presidential elections – not through misleading advertisements. sound bites, or spin doctors, but rather by conveying candidates‘ positions on important issues.

3. Why does the rural working class vote for a New York billionaire?

Trump won the election by winning traditional blue states such as Wisconsin, Michigan or Pennsylvania. These states‘ still strong working class populations traditionally tended to vote Democrat. Why then did rural working class people now vote for a New York billionaire who’s policies are detrimental to low-income workers? This article provides an interesting explanation. I guess it is worth pointing out that this now highly topical appeared in the discipline’s premier journal years before the election.

Cramer, Katherine J. (2012) „Putting Inequality in Its Place: Rural Consciousness and the Power of Perspective“ American Political Science Review 106, 517–532. (article, ungated; view article online, ungated)

Why do people vote against their interests? Previous explanations miss something fundamental because they do not consider the work of group consciousness. Based on participant observation of conversations from May 2007 to May 2011 among 37 regularly occurring groups in 27 communities sampled across Wisconsin, this study shows that in some places, people have a class- and place-based identity that is intertwined with a perception of deprivation. The rural consciousness revealed here shows people attributing rural deprivation to the decision making of (urban) political elites, who disregard and disrespect rural residents and rural lifestyles. Thus these rural residents favor limited government, even though such a stance might seem contradictory to their economic self-interests. The results encourage us to consider the role of group consciousness-based perspectives rather than pitting interests against values as explanations for preferences. Also, the study suggests that public opinion research more seriously include listening to the public.

4. What’s sex(ism) got to do with it?

The 2016 election is all the more surprising because many observers saw it as pitting the worst qualified candidate ever against the best qualified candidate ever. In the end the man won leaving many to ask whether sexism is to blame. For those who want to understand the role a candidate’s gender plays in elections this article is a good starting point. Its results suggest that Clinton’s competence advantage was partly undone by her being a woman.

Fulton, Sarah A. (2014) „When Gender Matters: Macro-dynamics and Micro-mechanisms“ Political Behavior 36, 605–630. (article, gated)

Does candidate sex matter to general election outcomes? And if so, under what conditions does sex exert an effect? Research conducted over the past 40 years has asserted an absence of a sex effect, consistently finding that women fare as well as men when they run. Nevertheless, this scholarship neglects sex-based differences in candidate valence, or non-policy characteristics such as competence and integrity that voters intrinsically value in their elected officials. If women candidates hold greater valence than men, and if women’s electoral success stems from this valence advantage, then women candidates would be penalized if they lacked the upper hand on valence. Recent research at the macro-level reports a 3 % vote disadvantage for women candidates when valence is held constant (Fulton, Political Res Q 65(2):303–314, 2012), but is based on only one general election year. The present study replicates Fulton’s (Political Res Q 65(2):303–314, 2012) research using new data from a more recent general election and finds a consistent 3 % vote deficit for women candidates. In addition, this paper extends these findings theoretically and empirically to the micro-level: examining who responds to variations in candidate sex and valence. Male independent voters, who often swing general elections, are equally supportive of women candidates when they have a valence advantage. Absent a relative abundance of valence, male independents are significantly less likely to endorse female candidates. If correct, the gender affinity effect is asymmetrical: male independent voters are more likely to support men candidates, and less likely to support women, but female independents fail to similarly discriminate.

5. Populism

Donald Trump is frequently called a populist. But what exactly is populism? Cas Mudde is the author of the probably most-cited definition of populism in recent empirical work. He defines populism as a thin-centered ideology (it can be left or right) that divides society into two homogeneous and antagonistic groups, the good people and the corrupt elite, and claims that politics should be an expression of the general will of ‚the‘ people. In this article he and his co-authors develop a survey instrument to measure populism in the population and apply it to a survey of Dutch citizens. I’m counting on seeing results its application in the US and other countries soon.

Akkerman, Agnes, Mudde, Cas, Zaslove, Andrej (2014) „How Populist Are the People? Measuring Populist Attitudes in Voters“ Comparative Political Studies 47, 1324–1353. (article, gated)

The sudden and perhaps unexpected appearance of populist parties in the 1990s shows no sign of immediately vanishing. The lion’s share of the research on populism has focused on defining populism, on the causes for its rise and continued success, and more recently on its influence on government and on public policy. Less research has, however, been conducted on measuring populist attitudes among voters. In this article, we seek to fill this gap by measuring populist attitudes and to investigate whether these attitudes can be linked with party preferences. We distinguish three political attitudes: (1) populist attitudes, (2) pluralist attitudes, and (3) elitist attitudes. We devise a measurement of these attitudes and explore their validity by way of using a principal component analysis on a representative Dutch data set (N = 600). We indeed find three statistically separate scales of political attitudes. We further validated the scales by testing whether they are linked to party preferences and find that voters who score high on the populist scale have a significantly higher preference for the Dutch populist parties, the Party for Freedom, and the Socialist Party.

6. How to trump hate?

These days political science is blamed for inaccurate predictions. Most of the time however, it is blamed for being too focused on post-hoc explanations and problem diagnosis at the expense of prescriptive focus and solution orientation. Hence, my last choice in this reading list is an article which arguably possesses the latter qualities. One of the most worrying immediate effects of the forthcoming Trump presidency is that there seems to be an up-tick in racist and sexist conduct in the wake of the election, both on- and offline. The final article in this reading list presents and tests a strategy to combat racism on social media.

Munger, Kevin (2016) „Tweetment Effects on the Tweeted: Experimentally Reducing Racist Harassment“ Political Behavior 1–21. (article, gated; final draft, ungated)

I conduct an experiment which examines the impact of group norm promotion and social sanctioning on racist online harassment. Racist online harassment de-mobilizes the minorities it targets, and the open, unopposed expression of racism in a public forum can legitimize racist viewpoints and prime ethnocentrism. I employ an intervention designed to reduce the use of anti-black racist slurs by white men on Twitter. I collect a sample of Twitter users who have harassed other users and use accounts I control (“bots”) to sanction the harassers. By varying the identity of the bots between in-group (white man) and out-group (black man) and by varying the number of Twitter followers each bot has, I find that subjects who were sanctioned by a high-follower white male significantly reduced their use of a racist slur. This paper extends findings from lab experiments to a naturalistic setting using an objective, behavioral outcome measure and a continuous 2-month data collection period. This represents an advance in the study of prejudiced behavior.

If you want to read more, you may want to check out this mock college syllabus on understanding Trump.

A tip of the hat to Tarik Abou-Chadi for recommending the articles by Sarah Fulton and Katherine Cramer, to Zoltan Fazekas for recommending the article by Kevin Munger and to Ilyas Saliba for pointing me to the mock college syllabus.