The results were announced surprisingly early on Tuesday night: Voters in Kansas had rejected a proposed constitutional amendment that would have endangered the availability of abortion in the state. Or, to put it more clearly, Kansans voted in support of protecting access to abortion.
Yes, that Kansas. So, in short order, observers began extrapolating outward to November. The results in this red state, some argued, showed that abortion access was an unqualified political winner for the left. Kansas, it seemed, may have validated predictions that the Supreme Court’s overturning of Roe v. Wade could backfire on Republicans.
What happened in Kansas was remarkable and unexpected. What it means about what happens in November, though, is less clear.
Let’s consider those two things separately.
What happened in Kansas?
More than 900,000 people cast ballots on the proposed constitutional amendment. That’s over 25 percent more votes than were cast in the state’s Republican and Democratic Senate primaries. And it’s about two-thirds of the number of votes cast in the 2020 presidential election. It’s a staggering level of voting for a measure put on the ballot in August.
Of particular note is that in every single county, the results on the amendment question were to the left of the 2020 results. (That’s reflected on the graph below, where every county is to the left of the diagonal line. The line indicates an equivalent margin in the 2020 presidential contest and in the amendment vote.) There were 14 counties — home to nearly a third of the state’s population — that backed Trump in 2020 and voted against the proposed amendment this week. Those counties backed Trump by 20 points and “no” by 13 points.
In counties that voted for the proposed amendment, “no” got about 15 percent more raw votes than Joe Biden did in 2020. “Yes” got fewer than half the votes that Trump did. Some quick back-of-the-envelope math by MSNBC’s Steve Kornacki figures that about a fifth of the “no” vote came from Republicans.
An overwhelming victory for the “no” side, certainly — but not one that should be seen entirely as a bolt of out the blue. As Natalie Jackson of PRRI noted, 2018 polling showed that support for keeping abortion legal in the state was about evenly split. The biennial General Social Survey shows broad agreement among Americans on both the left and the right that some access to abortion should be protected. The overturning of Roe transformed the political debatefrom one of figuring out how readily available abortion should be to one over whether it should be available at all, shifting the political response.
In fact, that the result in Kansas so uniformly shifted to the left is a reminder that the constitutional amendment vote may be the exception, not the rule.
What does this tell us about November?
In the immediate aftermath of the Supreme Court decision upending Roe, there was no obvious benefit to Democrats in averages of congressional polling. Yes, some polls showed Democrats moving into better position, but polling averages, a better measure, showed only modest shifts. Predictions that Roe would spark a pro-Democrat backlash weren’t clearly manifested.
One argument offered in response was that polls weren’t capturing a shift in enthusiasm. It’s a valid criticism; one factor that probably contributed to the polling miss in 2016 was that a flurry of Trump voters turned out to vote who hadn’t been included in pollsters’ turnout estimates. That’s a slightly apples-to-oranges example, but you get the point: If a ton of infrequent Democratic voters come out to vote in congressional races, that would be tricky to capture in polling. (We will, for now, set aside the trickiness of polling people on the likelihood that they will vote or on how news events affect that likelihood.)
But this brings us back to the cautionary vote above. Certainly a substantial part of the reason that the proposed amendment failed was that it was largely separate from partisan politics and political candidates. There is a big difference between asking people to weigh in on an issue and asking them to weigh in on a candidate who embodies a range of issue positions. When you are voting on an amendment that would erode the availability of abortion, that’s the only consideration. When you are voting for a governor or a senator, you’re voting on their support for or opposition to abortion — and a galaxy of other things. For senators and members of the House, there’s an added layer of calculus: how well your political team is positioned in Congress.
Voters in Kansas and elsewhere will not go to the polls in November to vote for representatives who will vote only on abortion. And that will complicate how they vote.
Even among Democrats, polling conducted by YouGov last month found that abortion isn’t the issue that’s likely to be most important in the upcoming election. Instead, it is among the top issues. Some voters will absolutely turn out to shift elected leadership toward protection of abortion — but many will be motivated to turn out for other reasons. Are voters going to be energized to turn out to vote for a random Democratic House candidate simply to backstop abortion rights? It’s hard to say and it’s hard to measure.
For many on the left, the results in Kansas were a reminder of precisely that point: Turnout matters. But electoral politics are rarely downstream from views on one single issue.
There is a demonstrated exception. In 2018, Democrats saw huge success in large part because Trump was so unpopular. This, however, is not an exception that bolsters the idea that Democrats will also see huge success this year.