Race and gender in the primaries

Let's look at some data as opposed to throwing numbers and accusations around. Exit polls ask if the gender and race of the candidate was important in who you voted for. We'll have to make a couple of assumptions.

1.    The responses are an accurate reflection of the person responding to the question.

2.    As in all polls, we'll have to assume the results reflect the voting population.

3.    And we'll also have to assume that voters using gender as a reason to vote for Clinton aren't sexist (since in this campaign its generally understood that if you aren't voting for Clinton specifically because she's a women that represents a sexist approach) and that voters using race as a reason to vote for Obama aren't racist (likewise, if you're not voting for Obama specifically because he's black that is construed as a racist approach).

More in the extended entry.

One of the potential drawbacks to this approach is that we may overestimate the number of people that are truly sexist or racist since people stating that gender or race was important in deciding who to vote for may have other reasons for selecting a particular candidate. However, if you responded to these questions in the affirmative, it had to have a non-trivial influence on your vote. And I would argue that if the gender of the candidate was so important that it influenced who you voted for, that is a pretty good indication of someone who most people would consider sexist.

Let's first look at the last two elections, Oregon and Kentucky. In Kentucky, 16% of the voters said gender was important and of those people, 19% were Obama voters. That calculates out to 3.0% of the total voters that could be considered exhibiting some sexist bias in their vote. 21% of the voters said race was important and of those people, 81% were Clinton voters. That calculates out to 17.0% of the total voters exhibiting some racial bias. Alternatively, if you want to break it down by percentages of each candidate's voters, then 10.1% of Obama voters could be considered somewhat sexist and 26.3% of Clinton voters somewhat racist.

Oregon numbers were a little different. 17% said gender was important and doing the same calculations as above, that results in 7.5% of all voters using sexist reasoning for their votes (13.1% of Obama voters). 10% said race was important, resulting in 4.8% of all voters being using racist reasoning (11.6% of Clinton voters).

I didn't look at all of the states but I did the same calculations for Indiana and Pennsylvania. Sexist voting tendencies could be attributed to 13.6% of Obama voters in IN and 12.6% in PA. Racist voting tendencies could be attributed to 16.7% of Clinton voters in IN and 20.6% in PA.

At least in the four states I looked at, "sexist" voters were between 3 and 8% of the total voting population (between 10 and 14% of Obama voters) while "racist" voters varied between 5 and 17% of the total voting population (between 12 and 26% of Clinton voters). Based on this limited analysis, it would appear that race had a bigger negative impact on the elections than did gender. Unfortunately, all of the numbers were greater than zero.



Display:


Re: Race and gender in the primaries (2.00 / 1)

Very good analysis. Surprised I haven't seen anyone run these numbers before.


by Barrett Brown on Wed May 21, 2008 at 11:16:55 PM EST

Tough to buy those assumptions (none / 0)

Sexism and racism are almost certainly areas where people will under-report. It's got to be VERY hard to get anything like an accurate count of who's going to let race and gender bias their vote.

Also, these numbers don't hold positive and negative bias. That is, is a woman who wants to see the first woman elected president "sexist"? Is an African American who wants to see the first AA elected prez "racist"?  It's tough to assign those labels, imo.


by PhilFR on Wed May 21, 2008 at 11:22:08 PM EST

Re: Tough to buy those assumptions (2.00 / 0)

No, he accounts for that, unless I'm misreading it; he's not deeming those who vote for Clinton and then cite gender as an important issue to be sexist. He's only counting Obama voters who cite gender as sexist, so the woman in your example above wouldn't count, and neither would men who voted for Clinton because she's a woman, incidentally.


by Barrett Brown on Wed May 21, 2008 at 11:27:27 PM EST
[ Parent ]

Re: Tough to buy those assumptions (none / 0)

I worked under the assumption that someone voting FOR Clinton because of her gender wasn't sexist. Likewise, someone voting FOR Obama because of his race wasn't racist. It's only when someone votes against a candidate because of gender or race do I assume that they had some sexist or racist basis for their vote.


by kjblair2 on Wed May 21, 2008 at 11:34:34 PM EST
[ Parent ]

Re: Tough to buy those assumptions (none / 0)

Missed that. sorry!


by PhilFR on Wed May 21, 2008 at 11:39:59 PM EST
[ Parent ]

Re: Tough to buy those assumptions (none / 0)

And I agree with the potential for under reporting since most exit polls have some human interaction, even if its just handing the poll back to the person.


by kjblair2 on Wed May 21, 2008 at 11:36:12 PM EST
[ Parent ]

Re: Tough to buy those assumptions (none / 0)

I suspect the under reporting isn't as bad as you might guess. The exit poll questions seem fairly neutrally worded, so I suspect that people are reasonably likely to answer honestly. Saying that race entered into your decision or gender entered into your decision isn't the same as being asked if you are a racist, or if you think women shouldn't be allowed to be president.

Hard to say which (anti-woman or anti-black bias) would be more likely to be accurately self-reported.


by letterc on Thu May 22, 2008 at 12:49:14 AM EST
[ Parent ]

Re: Race and gender in the primaries (none / 0)

thanks but,

no surprise...unfortunately.


by citizendave on Wed May 21, 2008 at 11:28:41 PM EST

Re: Race and gender in the primaries (none / 0)

I think the thing that this wouldn't account for is the degree to which sexism is possibly more deeply subliminal. Obama looks presidential, even though he is black, but Clinton doesn't look presidential, in part because she is a woman. Part of the range of appropriate presentation for a presidential candidate clashes with seeming appropriately feminine. There isn't a clearly established wardrobe for a woman running for president. Etc.

All of these factors influence whether someone feels favorable towards Clinton as a candidate, but they don't register as something that would cause someone to answer "yes, gender played a role in my decision."

I think there are a lot more (white) people who feel that no black person can be trusted to be President than think no woman can be trusted to be president, but there are a lot more people (men and women) who will think that any specific woman just doesn't seem quite right for the presidency.

I'm sure this primary campaign is going to be fodder for a lot of dissertations on race and gender.


by letterc on Thu May 22, 2008 at 01:02:32 AM EST


You are not logged in.

In order to post a comment, you must be logged in. If you have a member account, please log in to comment.

If not, you can make an account right here. It's quick and free.