It is ironic that the author of one of the most unhinged rants of the entire sad Kavanaugh spectacle nonetheless stumbled on a central insight that has been missing from the overwhelming majority of media discussions of Kavanaugh. In Alexis Grenell’s extended tantrum in the New York Times, directed against white women, she acknowledged what the data show clearly: The anti-Kavanaugh campaign was not really about men versus women.
Despite the disproportionate number of white women who displayed their anger for the cameras, most white women were in fact not Kavanaugh opponents. Indeed, race is a more fundamental cleavage in understanding support for Kavanaugh than gender is, and it is just one of many categories far more predictive than gender in understanding whether a voter supported Kavanaugh’s confirmation.
This can be illustrated using Reuters polling explorer, which compiles data from surveys of thousands of people on various issues and allows one to slice and dice survey results by various demographics including sex, income, religion, employment, political affiliation and ideology, etc. Because of the large overall number of observations (more than 7,800 in the period studied), it is possible to get statistically significant samples even of smaller population subgroups, something not possible with most other polls.
Using Reuters, I have analyzed polling data on the Kavanaugh nomination since September 14, just two days after the first allegations against Justice Kavanaugh surfaced. All data presented are for likely general-election voters and feature a minimum sample of 350 voters (usually much higher).
For reference: During the full period sampled, the Kavanaugh nomination had among likely voters a negative approval rating, 41.4 percent approving and 44.9 percent disapproving.
There was a significant gender gap. Men supported Kavanaugh, 48.2 percent to 42.6 percent, while women opposed him ,34.2 percent to 47.4 percent. The gap between men and women was 9.5 points. But this is a smaller gap than in the exit polls of the 2016 presidential election (twelve points) and virtually identical to the nine-point gender gap faced by Romney in 2012. So despite the outrage theater of the Democratic party and its media auxiliaries, there appears to be in the reaction to Kavanaugh little special gender resonance beyond the general voting preferences of men and women.
Overall, ideology was far more important than gender in terms of how likely voters saw Kavanaugh. Women of all races who at least lean conservative supported Kavanaugh, 68.1 percent to 20 percent, while those women who leaned liberal opposed him, 78.6 percent to 11.4 percent
In general, the demographic core of the Kavanaugh opposition consists of groups who were once on the political margins of American society and have more recently become part of the political mainstream. Gays opposed Kavanaugh, 15 percent to 64 percent. Those of “no religion” opposed him 68.7 percent to 21.6 percent, while African Americans opposed him 73 percent to 14 percent. Given the steady presence of all of these groups in the Democratic coalition, it appears that the politics of the Kavanaugh nomination were simply politics as usual.
But amid all of these data there is one strong outlier. Perhaps the most important factor, dwarfing the importance of gender, is marital status, and it is a factor that is particularly critical for the ways it contradicts the media’s simplistic women-versus-men narrative on Kavanaugh. Even compared with the large marriage gap found in national elections, the Kavanaugh marriage gap was remarkable.
Married people supported Kavanaugh, 49.5 percent to 37.4 percent, with married women of all races supporting him 43.1 percent to 39 percent. Married men supported him 55.2 percent to 36 percent.
Married white men supported Kavanaugh, 61.9 percent to 29.9 percent, while married white women supported him 47.4 percent to 34.8 percent, a stark contrast to the impression delivered by an endless parade of liberal white women on TV bemoaning the stark sexism of the Kavanaugh hearings and screaming at senators in elevators.
Meanwhile, single white women supported Ford, 58 percent to 23 percent, resulting in an incredible 24-point “marriage gap.” Among all registered voters, single white men were split 44 percent and 44 percent (a 16-point gap with their married brethren.) Overall, the marriage gap between married and single white men dwarfed the overall gap between men and women, and the gap between married and single whites as a whole was larger than the gap between whites and all non-white groups (excluding African Americans).
And the marriage gap was hardly limited to white voters. Among both sexes, voters who identified as Asian, Hispanic, or of races other than white opposed Kavanaugh just 40.5 percent to 46.4 percent, and if these married voters followed a religion, they favored him 47 percent to 40 percent.
Single women as a whole opposed Kavanaugh, 65.3 percent to 16.1 percent (a staggering gap of more than 26 points, which dwarfs even the 16-point marriage gap that Hillary Clinton enjoyed with single women) while single men opposed him 36.8 percent to 51.9 percent, a 15-percentage point margin that dwarfed the 2 percentage points that Hillary Clinton carried this group by. Divorced women opposed Kavanaugh 28.6 percent to 51.7 percent. The views of Kavanaugh among divorced women diverged from those of married women to a greater extent than did the views of men from those of women overall. Clearly something about marital status had a dramatic effect on how voters viewed Kavanaugh. But what was it?
The marriage gap, particularly for women, has long dramatically outstripped the gender gap in explaining voting behavior, but because it doesn’t flatter Democrats sensibilities (marriage still being generally considered a good thing in our society), it receives only a fraction of the former’s media attention. And the full reasons for it are still unclear.
Scholars Christopher Stout, Kelsy Kretschmer, and Leah Ruppanner, who have studied the issue, claim that data show that single women experience “gender-linked fate” — meaning that they view what happens to any woman as affecting their well-being — while married women are less likely to generalize about the fates of women as a whole, preferring to feel solidarity with their husbands and families rather than with women they do not know. This hypothesis does have the effect of explaining why Democratic appeals to female solidarity are so effective to single women but not to married ones.
Other factors play a role as well. For example, single women have lower incomes, which tends to push them into the Democratic camp.
While plausible in some respects, that hypothesis does not explain why single men were overwhelmingly more likely to oppose Kavanaugh. The marriage gap among men was massive, more than 16 percent. Surely if we buy some version of a gender-solidarity theory, for men still potentially on the lookout for mates, to see a fellow male pilloried for teenage behavior could be seen as a threat to their own status and behavior with women. (Whereas for married men an attack on the sexual behavior of single men or teenage boys would not affect the status of their own marriage. Yet single men were heavily supportive of Ford. While any answers are strictly speculative, perhaps married couples simply have more-realistic expectations of the behavior of men and women, and particularly of teenage boys and girls, than do their single brethren. Or perhaps they are less likely to believe claims absent corroborating evidence.
It is difficult to say whether, with respect to Kavanaugh, the media’s narrative on gender versus marriage is a result of ignorance or dishonesty. But it is understandable why the media, in their desperation to elect Democrats, would wish to paint the war against Kavanaugh as a war for all women rather than a war against the views of married couples.
To note that the core opposition to Kavanaugh consisted of atheists, African Americans, gays and lesbians, and, most important, the unmarried or divorced of all ethnicities would describe the anti-Kavanaugh grouping far more accurately than does the “gender war” narrative — but it is much less politically useful to the Democrats.
It would have been much harder for the media to portray Kavanaugh as the evil embodiment of the sexually assaulting patriarchy if they had to acknowledge Kavanaugh’s support among derived not just from white men, but also married women. Nor would the Democrats’ narrative be improved by acknowledging his support from non-atheists, and married parents with children along with roughly even backing from married, non-African-American minorities
But unlike the liberals’ “just so” stories about Kavanaugh and gender, those characterizations have the advantage of being true.