Trump Supporters vs Feminists

Attributes that Predict Trump Support

My first question was this: what attributes of a person’s worldview and/or personality predict positive attitudes towards Donald Trump?

Because this was more of a fishing expedition than a hypothesis-driven scientific study, I tested the correlation of each potential predictor variable in the data set with attitudes towards Trump, and used a cutoff of p<0.01 (i.e. less than 1% chance of finding a “significant” correlation due to noise) with the Bonferroni correction for multiple comparisons.

I used the 27 survey questions with scale 1–5 or 1–10 answers and excluded questions specifically about SSC. You can read the survey questions here. For simplicity’s sake, I did not perform any multivariate analyses. At the individual-comparison level, that’s p<0.0004 (i.e. 0.04% chance of any given correlation being noise) and R>0.05. This means that some of the correlations I report here are consistent enough to be interesting, but small enough to not be taken too seriously.

After breaking down the data this way, here were the top predictors of having a favorable view of Donald Trump:

Unsurprisingly, Trump supporters are more right wing, more anti-feminist, more anti-immigration, not concerned about global warming, anti-gay marriage, open to scientific racism (these days branded as human biodiversity), opposed to basic income, and opposed to minimum wage. All fairly standard right-wing stuff up to this point.

[EDIT: Some commenters have rightfully pointed out that people who gave a high favorability rating to the concept of “human biodiversity” weren’t necessarily racists, but instead may have just been very excited about the associated advances in biological science. While using biological differences to invoke racist tropes (i.e. scientific racism) is one possible application of these ideas, it does not necessarily represent the views of those who ranked highly on this trait. I apologize to any survey respondents who were unfairly represented in this analysis, and I encourage any new readers to bear in mind that what I termed “scientific racism” in this essay may have a lot more to do with science than with racism. The term used in the original survey was “human biodiversity”.]

Where it gets more interesting is that Trump supporters were less trusting towards other people, less concerned about bio-terrorism, more likely to lift weights, less concerned about AI risk, and more likely to believe in the importance of innate talent.

Based on these data, I would stereotype Trump supporters as “cynical, right-wing masculinists with minimal concern about the potential harms of technology.”

Attributes that Predict Feminism

Aside from partisan identity, the strongest predictor of someone’s attitudes towards Trump were their attitudes towards feminism. This seems to jive pretty well with the biggest anti-Trump marches happening in real life being the women’s marches. I therefore repeated the same correlation analysis (still using p<0.01 with Bonferroni correction) with Feminism as the dependent variable.

Here are the top predictors of having a positive view of feminism:

Like a mirror image of Trump supporters, feminists are more left wing, anti-Trump, concerned about global warming, pro-immigration, pro-gay marriage, anti-racist, supportive of increasing the minimum wage, and supportive of basic income. Interestingly, feminism seems to be even more left wing than Trump support is right wing, although the correlations aren’t all that different. Aside from partisan identity, attitudes towards feminism were best predicted by attitudes towards Trump.

Aside from their explicitly political attributes, feminists are more trusting of others, believe less in the importance of innate talent, believe in personal growth (growth mindset), are less likely to lift weights, more satisfied with life, and more likely to appreciate puns. They’re also more likely to be concerned about bio-terrorism risk.

Based on these data, I would stereotype feminists as “cheerful, left-wing, and effeminate, with concerns about the potential harms of technology.”

Attributes that Predict Partisan Identity

The strongest predictor of both Trump support and feminism is partisan identity on a graded left-right scale. Partisan identity also shows up as a small but significant factor for a lot of other variables I was curious about, even those that are not explicitly political (e.g. life satisfaction, weight lifting), so I decided to take a look at what attributes were most highly correlated with partisan identity.

Here’s what I found:

Based on this data, the most polarizing political issues (at least among SSC readers) are global warming, feminism, minimum wage, Trump, gay marriage, basic income, and immigration, in that order. Additionally, Republicans are generally less concerned with the potential harms of technology (i.e. bio-terrorism, global warming, AI risk) than are Democrats (simplifying right vs left wing to Republicans vs Democrats to make terminology easier).

Besides the explicitly political survey questions, Republicans tend to believe incrementally more in innate talent where Democrats believe a bit more in personal growth, and Republicans are incrementally more likely to lift weights, are a bit happier, go to bed a bit earlier, and are a bit more satisfied with their lives.

In aggregate, I think these findings pretty much line up with the theory I presented in a previous essay that the left vs right divide tends to be defined by Democrats placing a higher value on compassion (e.g. sticking up for minorities and basic income) vs Republicans placing a higher value on personal discipline (e.g. lifting weights and going to bed earlier).

Trump Supporters vs Feminists: Beyond the Partisan Divide

Although partisan identity was the strongest predictor of both Trump support and feminism, it only explains ~25% of the variance in each trait. In addition, there is an apparent contradiction with respect life satisfaction: it correlates a bit with both feminism and right wing political views.

I therefore decided to correct everything else in the data set for partisan identity, such that all values were regressed towards the political center. I figured this would allow me to look at the essences of Trumpism and feminism beyond the influences of ordinary partisan tribalism.

Here’s what I found for Trump supporters:

And here’s what I found for feminists:

For Trump supporters, social issues remain significant while some economic issues fall away, after correcting for partisan identity. There were relatively few questions in the survey on economic and foreign policy issues, but would I suspect that Trump supporters are less distinctively Republican in these areas, but more right wing on social issues.

Additionally, even after correcting for partisan identity, Trump supporters are a bit still less trusting of others (perhaps helping to cause the more right-wing social views), and still a bit more likely to lift weights. Trump supporters are also a bit less satisfied with their lives.

For feminists, social issues and economic issues (i.e. minimum wage and universal income) both remain significant after correcting for partisan identity. I can think of two reasonably likely explanations for this: 1) feminists are even more left-wing than they realize, and 2) many feminists think of minimum wage and the basic income essentially as social issues, and may be more moderate on issues to which the “sticking up for the disadvantaged” paradigm does not readily apply.

I would suggest that there’s some of both going on, and that the second explanation tracks reasonably well with Hillary Clinton’s politics. After correcting for partisan identity effects, feminists also become a bit happier and a bit more satisfied with life than the average person, as well as a bit more politically engaged, a bit more trusting of others, and a bit more likely to go to bed early.

As a side note, if you correct for the influences of feminism, leftism correlates somewhat with a comparatively cynical worldview. There may a lesson in that dichotomy concerning the current divide within the Democratic party. There seems to be a hint in this data set that the anti-establishment wings of both parties are less satisfied with their lives, which would make some sense.

What about life satisfaction?

After all these political statistics, you may have forgotten my original objective: to take a look at what attributes predict greater life satisfaction. Originally, some right wing political views showed up when I ran this analysis, but from looking at the rest of the data, that may have more to do with Democrats and Republicans living in totally different worlds (not just different political views, but different jobs, different pay grades, different answers to non-political worldview questions, etc) than with right wing political philosophy promoting life satisfaction.

After correcting for partisan differences, here’s what I found:

People who were more satisfied with their lives also tended to be happier, more likely to believe in personal growth, enjoy puns more, more okay with noisy conversations, more feminist, more trusting, and more likely to lift weights.

Looking down the next few items that showed up as statistically significant (e.g. getting up earlier, anti-racism), which I removed for brevity’s sake, the same general themes remain. Those who are more satisfied with their lives tend to be 1) more optimistic/cheerful, 2) more compassionate towards others, and 3) show more signs of self-discipline.

Interestingly, #2 and #3 are pretty much the same attributes that I argued were the key to leaving a meaningful legacy in my previous article. I suspect that if you could find a way to meaningfully quantify “legacy”, life satisfaction and personal legacy would be highly correlated…