Comment Before Kansas voters weighed in last month, there was only one poll gauging support for the state’s proposed constitutional amendment allowing new restrictions on abortion. That poll showed a close result, maybe a handful of points. And then Kansas voters rejected it by nearly 20. The vote came just two months after the Supreme Court ruling Dobbs v. Jackson Women’s Health Organizationthe decision that struck down abortion protections enacted by Roe v. Wade. It wasn’t hard to draw a line: Voters are angry Dobbs helped bury the constitutional amendment in Kansas; There was even little data to support the idea that women in particular led to the amendment’s defeat: New voter registrations among women increased after the decision. Figures provided to the Washington Post show similar increases in at least two other states. The story here makes sense. The pieces click. But a new question immediately emerges: Is this a story about empowered women voting in Kansas, or is it a story about midterm elections in general? Subscribe to How To Read This Chart, a weekly data newsletter from Philip Bump Let’s start with a relatively careful assessment. Tom Bonier, CEO of Democratic data firm TargetSmart, began pointing to an increase in the density of women registering to vote after Dobbs a few weeks after the announcement of the decision. It’s worth noting that the Washington Post attempted to match its results without being able to do so universally (which may be a function of missing data on The Post’s part). Bonier’s assessment, offered in a New York Times essay over the weekend, is that there is a “clear pattern” showing increases in the percentage of women among new enrollments in some states, though none match Kansas. The fact that Kansas had a special election focused specifically on abortion is, of course, an important factor in this whole debate: The state was voting expressly in what Dobbs addresses. Our data shows a murky picture: increases in states like Pennsylvania, but not in places like New Mexico. The increase in Pennsylvania, meanwhile, didn’t just happen after theDobbs. There was a period in February, for example, when women made up an equally disproportionate share of new registrants. Maine is another state in which many point out that women have seen an explosion among new enrollees recently. Post can confirm this post-Dobbs increase, but the state has also seen similar increases of women making up as much as 65 percent of new enrollees over the past two years, well before the ruling. There’s also no doubt that Democrats have seen improved polls since then Dobbs. YouGov tracks responses to a general ballot question — asking people whether they plan to vote for Democrats or Republicans on their House ballot — and has seen Democratic support rise several points since early July. (Due to a methodological problem, YouGov does not have general polling during the period between a draft opinion being leaked to Dobbs and the final decision.) This is due to the increase in support from both women and men in recent weeks. (The lines in the chart above show the average of the three most recent polls.) Notice, however, that Democrats had about the same level of support from women earlier this year as they do now. If we pick three months—January, April, and July—we can see that average support among women fell in April before rebounding. Does that mean that Dobbs (and/or leaked opinion) reversed a slide? That spring was an aberration? That there is still something missing from the vote? If we look at the general ballot margin — that is, the gap between Democrats and Republicans — the advantage among Democrats among women immediately after Dobbs it was actually lower than at the beginning of the year. (The latest YouGov survey shows a wide gap, but this is a poll.) Interestingly, while the number of people expressing uncertainty about their vote has fallen (which tends to happen as elections approach), the decline has been much sharper among men. In other words, men are more likely to report more certainty about their vote recently. One could read this as Dobbs it does not have a strong effect on women. Or we could read it another way: that polls don’t capture excitement, just as this poll in Kansas was unexpected. After all, consider that the imbalance in voter registration among women in Kansas shifted the electorate from 52.2 percent women to 52.4 percent by mid-July — a change of 0.2 percent. That not enough to result in a loss of 18 points for the constitutional amendment. But even if it’s not causal, it can be an indicator of excitement, which can be harder to measure. In the YouGov poll, enthusiasm for voting between men and women was about the same immediately before Dobbs as it is now, with men showing more interest in voting. But perhaps polls are simply not capturing the views of women newly motivated to vote, just as polls have repeatedly grossly underestimated Republican voters in 2016 and 2020 in particular. These polling shortfalls have been subjected to an enormous amount of analysis (almost all of which is somewhat vague), but it is not entirely improbable that women who have recently become involved in political post-Dobbs may escape the attention of pollsters. There are other assumptions we could challenge, of course. The gap between men and women on the legality of abortion has long been much narrower than that between Democrats and Republicans, for example, which means there may be some new energy among liberal men, too. So much of it is disappointingly uninteresting. Especially for staunch defenders of abortion access, it seems obvious that American women will be outraged and flood the polls in November. They can. In this case, as the old warnings say, the most useful assessment of its effect Dobbs may be reduced in participation. Lenny Bronner and Emily Guskin contributed to this report.
