Friday, June 30, 2017

"Modern Sexism" in the 2016 Presidential Elections.


Introduction


A poll was carried out right after the 2016 presidential general election by the Diane D. Blair Center of Southern Politics and Society at the University of Arkansas.  The poll sampled 3,668 individuals, and we are told that the sample is representative.

Angie Maxwell and Todd Shields used the data from that poll to study the effect of "modern sexism" on the presidential election results, both those from the primaries and from the general election.  The short findings from that study:

...sexism absolutely did matter. Trump’s voters were more sexist than Clinton’s (and Ted Cruz voters were even more sexist than Trump voters). Republicans were far more sexist than Democrats. White respondents were more sexist than black Americans and Latinos. Female respondents, not to be outdone, were also quite sexist! And Bernie primary voters who didn’t vote for Clinton in the general were more sexist than those who did.


What Is Modern Sexism?



Before we look more closely at those findings, it's necessary to understand what this study means by "modern sexism:"

Most people who have sexist or racist beliefs will not answer poll questions about those honestly, for obvious reasons.  Researchers have tried to get around that problem by using proxy questions or assertions, the kinds which correlate with negative beliefs about people of color and/or about women (1).  For the sexism part, the Blair Center poll used the following assertions, asking, for each of them, whether a respondent agreed or disagreed with them and how strongly (2):

  • Many women are actually seeking special favors, such as hiring policies that favor them over men, under the guise of asking for “equality.”
  • Most women interpret innocent remarks or acts as being sexist.
  • Feminists are seeking for women to have more power than men.
  • When women lose to men in a fair competition, they typically complain about being discriminated against.
  • Discrimination against women is no longer a problem in the United States.

In what sense could those assertions be seen as sexist?  Note that the first, the second and the fourth contain those little words "many," "most" and "typically."  Taken together, those three assertions spell out a dismal view of women, especially of women in the labor force, in education and in the public sphere.  The third turns feminism into a search for a matriarchy, not for equality,  and the last assertion argues that women already are equal in the United States, which makes any further feminist activism an attempt to dominate over men.

My first reaction to that list of assertions was to notice how much it shared with the MRA sites where women ("all" women or "many" women) are "typically" seen as vile creatures not deserving of any kind of equality and where feminism is certainly viewed as a plot for enslaving men.

It's only the last assertion you won't often find on those sites, because at least the more vicious sites don't see anything wrong with women having fewer rights, given that women are viewed as lesser human beings.  It's not really possible to discriminate against women when women deserve less than men.

My second reaction was to remember that the famous Alt Right site Breitbart.com (the home of our Dear Leader's companion, Stephen Bannon) often publishes stories with those very messages about the perfidy of women in general and of feminists in particular.

Finally, my third reaction was to recall all the biased conservative articles I have read (and dissected here) which argue that, say, the earnings gap between men and women is a totally imaginary one, that women earn less because they choose to earn less and so on.




The Results:  Modern Sexism Levels For Various Voter Groups in the 2016 Elections 


Given my reactions, the actual findings of the study came as a bit of a shock: Though the plurality of the 3,668 respondents gave, on average,  nonsexist answers to those five assertions (47.1%), more than one third (36.2%) had answers which gave them an average sexist score.

The following three tables summarize the results about the degree of modern sexism in the poll. The first shows them for everyone, the second for men and the third for women.  The orange color refers to the percentage of nonsexist answers, the green to the percentage of sexist answers and the yellow to neutral answers:










A few comments are worth making about the tables:

First, the majority of women in the poll (52.5%) gave, on average, nonsexist answers, while only 41.2% of the men in the poll did, and women scored somewhat higher on the nonsexist scale than men in all the demographic, regional and political groupings. 

Second, Democrats (65.2%) were much more likely to come across as nonsexist than Independents (38.4%) or, especially, Republicans (30.8%).

Third, African-Americans (both men and women) (59.0%) gave less sexist answers than Whites (46.1%) or Latinx (42.3%).

Fourth, the most sexist demographic group among men consists of Whites (44.2%), whereas the most sexist demographic group among women consists of Latinas (36.3%) (3).

Fifth, the highest percentages of sexist answer averages came from those who identified as Republican, both among men (56.3%) and among women (49.9%).

These results further clarify the finding that Donald Trump's pussy-grabbing comments didn't bother a sufficient number of Republican women for most of them not to vote for him. Erin C. Cassese, who has also studied modern sexism, notes (on the basis of a different data source):

Republican women score significantly higher on modern sexism than both male and female Democrats, though they score lower than male Republicans. This finding is instructive in light of Trump’s alleged “women problem,” in that Republican women may have been less likely than Democrats to situate his comments in terms of a broader systems of discrimination. While modern sexism influences policy attitudes for Republican women, they are just as ideologically extreme as Republican men and just at likely to demonstrate partisan loyalty at the polls.

Indeed, in the Blair Center poll Republican women come across as more sexist than either Democratic or Independent men.  That finding teaches us not to assume that women cannot be sexist against their own gender (4).

So far the results I have addressed apply to the general election.  The results from the Democratic and Republican presidential primaries are also of interest:





Note the very large differences between those scores.  The United States indeed appears to consist of two countries with very different values.

Bernie Sanders' primary voters have the lowest average modern sexism score, but the small minority of Sanders primary voters who went for Trump in the general elections do have a fairly high average sexism score:




The Take-Home Lessons From This Study?

Are there any? 

The large differences in the average measures of modern sexism between Republicans, Democrats and Independents are worth keeping in mind: 

No, the Democrats are not every bit as bad on this issue as the Republicans, and  when participating in debates about the various reasons why a slight majority of white women voted for Trump despite his pussy-grabbing comments it's good to remember that those would be Republican white women and an openly sexist president doesn't look that outrageous to almost half of them.

One caveat about that finding:  This particular poll was carried out right after the general elections where the Democrats ran the (first) female candidate and the Republicans ran the (first) gloatingly sexist male candidate.  The allegiance to one's party may have (subconsciously) affected the answers to the modern sexism assertions.  If the race had been between a Republican woman and a Democratic man the party gap in the modern sexism measures could have been smaller.

It's also worth thinking about how party affiliation and modern sexism (as well as modern sexism) end up correlated with each other.  For example, sexists are more likely to join the Republican Party, because its platform includes the control of women's sexuality and opposition to any measures which might counteract sex discrimination against women in education and labor markets.

But it's equally possible that those who have joined the Republican Party will then become more sexist, given today's political information bubbles.  Anyone who gets his or her news mostly from Fox News (with its Barbie-rules about female broadcasters) or Breitbart.com or Rush Limbaugh or other right-wing sources gets frequent updates on the horrors of feminism, on fake rape claims and on other weaknesses of the female sex.

Finally, lest one lose all belief in humanity, note that the majority of people do come across as nonsexist in that study.



-----

(1)  The modern racism measures are constructed in a parallel way by creating assertions with which the respondents are asked to agree or disagree.  Two examples:

"It's really a matter of some people not trying hard enough; if blacks would only try harder they could be just as well off as whites."

And

"Over the past few years, blacks have gotten less than they deserve.

(2)  The researchers turned the agreement or disagreement levels to these assertions into numbers, added up those numbers and then averaged the result over the five assertions.  The resulting measure is used in the tables I look at later in this post:

Responses were measured on a 5-point Likert scale from “strongly disagree,” “disagree,” “neither/neutral,” “agree,” or “strongly agree.” Depending on their answers, respondents held a cumulative Modern Sexism score ranging from 5 to 25. Giving a “strongly disagree” answer to all five statements resulted in the minimum score of 5, whereas a “strongly agree” answer on all five questions resulted in the maximum score of 25. Answering “neutral” to all five statements resulted in a score of 15. Thus, any score over 15 indicates that the respondent holds some cumulative level of Modern Sexism, while any cumulative score under 15 indicates a general lack of Modern Sexism.
(3)  I have no hypothesis about the reason for that last finding, given that Latinas are more likely to vote for Democrats.  It could have something to do with the impact of the Catholic Church?  Or more recent entry into the country from perhaps a more patriarchal society? 

(4)  The above quote also reminded me of a conversation I had with a very elderly American white working-class Republican-voting woman about a decade ago.  She told me how she had been sexually harassed at work when she was young, and she even mentioned a boss who had tried to rape her, but she did not connect those events to anything wider (e.g. broader systems of discrimination).  They were just "how things are."  She saw no need to change anything, probably, because of that lack of wider consciousness.  Without it, each experience remains purely personal.

(5)  Maxwell and Shields also construct logit equations for predicting how an otherwise average voter of a certain type might vote if his or her level of modern sexism varied.  I'm not discussing that part of the study in this post.  The reasons are statistical and model-specific: I believe too many of the independent variables (party affiliation, modern sexism, modern racism, ideology, biblical beliefs) are correlated with each other for the interpretation of a specific coefficient to be very meaningful.  But your mileage might vary. The full equations are downloadable as the Appendix from the study site.