Tag Archives: presidential campaign

A Darker Side of Politics

Regular readers of this blog will know of my interest in the political decisions—often referred to as Richard Nixon’s “Southern strategy”—that have resulted in an association between racism and membership in the Republican party. During their political campaigns, Republicans (and sometimes Democrats) use “dog whistle politics”—racially coded appeals that automatically activate the negative stereotypes of their increasingly prejudiced audience.

There is now a fairly extensive literature in social psychology demonstrating that white people respond more negatively to images of dark-skinned African-Americans than those with lighter skin. For example, one experiment found that participants assigned more negative traits and fewer positive traits to dark-skinned blacks than to light-skinned blacks. Another study showed that, among blacks convicted of murder, those with darker skins were more likely to receive the death penalty.

There are persistent rumors that Barack Obama’s skin tone has been manipulated in campaign advertisements. For example, in 2008, Hillary Clinton’s campaign was accused of doctoring images of Obama to make him appear blacker, although it’s not clear whether this was deliberate. A new set of studies by Solomon Messig and his colleagues analyzes images of Obama from the 2008 presidential campaign against John McCain.

Working from a complete library of television commercials aired by both candidates, the researchers electronically measured the brightness of the faces in all 534 still images, 259 of Obama and 275 of McCain. The advertisements were independently coded for content by judges who were unaware of the purpose of the study. The researchers looked at whether each image appeared in an attack ad, and whether the ad tried to associate the candidate with criminal activity. Two differences emerged. Obama’s skin tone was darker in commercials linking him with criminal activity—see example below—than in all other images of Obama.

https://www.youtube.com/watch?v=ONfJ7YSXE5w%20

In fact, 86% of the photos in these ads were among the darkest 25% of all Obama photos. Secondly, in attack ads produced by the McCain campaign, images of Obama grew darker toward the end of the campaign, even as their own images of McCain grew lighter.

The authors did two followup studies to determine whether darker images of Obama activated more negative reactions to black people than lighter images of Obama. They wanted to show that darkening the skin of a familiar black man, whom they refer to as “counterstereotypical,” would have the same effect as the darker faces of the unknown persons used in previous studies. In one experiment, participants viewed one of the Obama images below and completed a stereotype activation task in which they were asked to fill in the blanks of incomplete words such as “L A _ _” and “_ _ O R.” The darker image of Obama on the right elicited more stereotypical completions—“lazy” and “poor,” in these cases—than the lighter image.

The second study was more complicated, involving subliminal priming, but it too found that a variety of darker images of Obama yield more negative reactions than lighter images of Obama.

It’s not clear from these studies what the McCain campaign actually did in 2008. Did they deliberately darken some images of Obama, or did they merely select darker images? If the latter, did they select images because of their darkness, or were they merely trying to choose images than made him “look bad,” without thinking about why. The fact that these darker images appeared in ads attempting to link Obama with criminality, however, suggests that whatever they did was not accidental.

These campaign ads appeared on television seven years ago. The pace of social psychological research—including the publication lag—is often quite slow. The two followup studies probably accounted for most of the delay. Although they allowed the authors to tie up some loose ends, it could be argued that they were unnecessary, since they largely replicated previous studies. The delay was unfortunate, since the analysis of the ads didn’t appear in print until Obama was no longer running for office and the corporate media could treat it as old news. Sometimes postponing the release of information is almost as effective as completely suppressing it.

Of course, there will be other black candidates and many more opportunities for dog whistle politics.

You may also be interested in reading:

Guarding the Hen House

The World According to the Donald

Another Dog Whistle

The World According to the Donald

Would I approve waterboarding? You bet your ass I would—in a heartbeat. And I would approve more than that. Don’t kid yourself, folks. It works, okay? It works. Only a stupid person would say it doesn’t work. . . . Believe me, it works. And you know what? If it doesn’t work, they deserve it anyway, for what they’re doing.

The corporate media find Donald Trump to be by far the most newsworthy candidate of the 2016 presidential campaign. According to the Tyndall Report, as of November 30, he accounted for more than a quarter of the campaign coverage on the nightly ABC, CBS and NBC newscasts, more than the all the Democrats combined. Both the New York Times and the Washington Post have made preliminary attempts to describe his rhetoric.

Patrick Healy and Maggie Haberman of the Times “analyzed every public utterance by Mr. Trump over the past week from rallies, speeches, interviews and news conferences”—95,000 words, we are told. Four days later, Paul Schwartzman and Jenna Johnson of the Post did a “review of the businessman’s speeches, interviews and thousands of tweets and retweets over the past six months.” The Times article focuses more on the content of the speeches while the Post identifies what they call campaign strategy. Unfortunately, neither article describes the process by which the analyses were done or provides any meaningful data. I assume they simply read the speeches and recorded their impressions.

According to the Times, Trumps’s speeches are characterized by “constant repetition of divisive phrases, harsh words and violent imagery.” They identify  several overlapping themes.

https://www.youtube.com/watch?v=LznMz3YC5Vk%20

  • Us vs. them. Trump takes advantage of the human tendency to categorize people into ingroups and outgroups, and to show ingroup favoritism, but he carries this to an extreme by characterizing the outgroup as inherently evil. In this exchange with a 12-year-old girl, he describes terrorists as “animals,” then makes a promise.

You know what, darling? You’re not going to be scared any more. They’re going to be scared. . . .We never went after them. We never did anything. We have to attack much stronger. . . . We have to be much smarter, or it’s never, ever going to end.

Trump identifies the grievances of his audience, and attributes these problems to disliked groups, as when economic problems are blamed on Mexican immigrants. According to the scapegoat theory of prejudice, prejudice, discrimination and violence toward minority groups increase during times of economic hardship.

  • Ad hominem attacks. Trump frequently attacks the person rather than his or her ideas. As Ted Rall points out, we have Trump to thank for introducing the word “stupid” into campaign discourse. According to the Times, he used the word “at least 30 times.” (Unfortunately, this number is meaningless without something to compare it to. How often do other candidates use such negative descriptors?) Other favorite adjectives are “horrible” (14 times) and “weak” (13 times). No target is out of bounds, including mocking a reporter with a disability.
  • Violent imagery. ISIS is described as “chopping off heads,” and Trump is going to “bomb the hell out of” our enemies. “Attack” is a favorite word. At one rally, Trump appeared to endorse the roughing up of a “Black lives matter” protester in the audience.
  • Creating mistrust. Trump tries to create suspicion about scientific facts and other data provided by the government and the news media. His audience is told that “nobody knows” the number of illegal immigrants or the rate of increase of health care premiums, when in fact reasonably accurate estimates are available.
  • Ingratiation. If you’ve seen films of Adolf Hitler, or American demogogues such as Joseph McCarthy and George Wallace, you known that they are not attractive or charismatic speakers. Trump, however, is a practiced entertainer. He is relaxed and informal (a favorite word is “guy”). He flatters his audience. While other candidates are stupid, he claims that no one is smarter than the American voter.

The Post adds two comments about Trump’s campaign strategy.

  • Message testing. Trump takes an experimental approach to constructing his stump speech. He tries out various lines, using audience response as the criterion of success. The article describes a joke about Bernie Sanders’ hernia operation that was tested, revised, tested again and eventually abandoned when it did not get laughs.
  • Consistent presentation. Trump repeats the same words and lines in almost every speech. Of course, all candidates have a standard stump speech. The difference may be that Trump appears to be ad libbing, but is not.

I found two other reports which compare Trump’s rhetoric to that of other candidates.

Matt Viser of the Boston Globe transcribed all the speeches in which the candidates announced they were running for the presidency, and analyzed them with the Flesch-Kincaid Readability Test. The test uses word and sentence length to determine how difficult a passage is to understand. The results are expressed as a grade level. Trump’s speech was the simplest. It could be comprehended by a fourth grader. One hypothesis is that candidates try to match the educational level of the audience they are hoping to influence.

21language_graphic_WEB-1547

Angie Drobnic Holan, the editor of PoliFact, a political fact checking website, published an analysis of the ratings of all the statements by 2016 candidates that she has fact-checked since 2007. Trump scores second only to Ben Carson in dishonesty. Of 70 Trump statements, 76% have been found to be false, mostly false, or “pants on fire”—reserved for the worst lies. There seems to be a relationship between honesty and party affiliation.

2015-12-12_11-39-52

PoliFact describes the process by which statements are selected and analyzed. There are two possible sources of bias in these data. They do not analyze a random selection of candidate statements. Maybe front-runners or people who are disliked by reporters are subjected to greater scrutiny. And since the content of these statements varies, there is no uniform method of deciding whether a statement is true or false.

To my knowledge, no one has done a scientific content analysis of Trump’s rhetoric. Such an analysis would require selecting a random sample of statements to be analyzed, operationally defining the speech categories to be counted, and comparing Trump’s totals to those of other candidates. Unfortunately, the time and effort required to do such an analysis makes it unlikely that it will be done until after the campaign is over.

You may also be interested in reading:

Trumping Bernie

Bullshit

Trumping Bernie

Although we are right to be concerned about the growth of campaign advertising, especially when the candidates’ resources are unequal, you could argue that the amount of free time (or space) given to candidates by the news media is more important.  Statements made by and about candidates in the free media are not as likely to be discounted as advertisements. (The discounting principle states that our confidence in a particular explanation of behavior is weakened by the presence of alternative explanations. In an advertisement, our knowledge that this is a paid message intended to persuade us makes us less likely to see the message as truthful.)

Senator Bernie Sanders and Donald Trump have both shown similar increases in support in the presidential polls. According to the Real Clear Politics average of all polls, support for Sanders has grown from 12.7% to 25% from July 1 to the present, while Trump has gone from 6% to 22%. Sanders’ progress might be considered more impressive, however, since he has received less free media coverage.

5440388253_7a8e8c1584_b-1

Fairness and Accuracy in Reporting (FAIR) used the Nexus database to count the number of free mentions of Sanders and Trump on 13 news media sources between July 1 and August 15. In all cases, Trump received more coverage. The chart below gives the percentage of stories mentioning Sanders as a percentage of those mentioning Trump. For example, on CNN, the ratio was 33%, meaning that Sanders was mentioned one-third as often, or that Trump received three times as much coverage.

TrumpSandersChartOn average, Sanders received 36% as much coverage as Trump. However, there were some interesting differences among the 13 outlets.

  • On the three outlets with the largest audience, the broadcast networks ABC, CBS and NBC, Sanders was mentioned only 16% as often as Trump, or in other words, Trump received six times as much publicity.
  • The country’s five major newspapers averaged 34%, very close to the overall average.
  • Public radio and television, which presumably have a mandate to be fair, scored only slightly above the overall average in Sanders coverage.
  • MSNBC, sometimes described as a “progressive” cable network, did have the highest percentage of Sanders coverage, but still mentioned Trump more often.

As Bernie Sanders continues to draw large crowds, the message of the news media seems to be that there’s “nothing to see here.” If we were to ask them why Trump receives so much more publicity than Sanders, my guess is they would claim that his flamboyant personal style makes his activities of greater public interest. However, I suspect that a more important cause is that a central theme of Sanders’ campaign is reducing economic inequality, an unwelcome message to those who own, advertise on, or perform on the news media.

19197893134_9616570321_n

In addition to discounting campaign advertising, these data suggest we should also discount some of the coverage candidates receive in the free media. However, we’re less likely to do that, since their conflicts of interest are less obvious.