As my editors will tell you, I have a bad habit of leaving stories half-finished. Sometimes it’s because I get bored or busy, but sometimes it’s because the subject gets less and less newsworthy as I’m in the process of finishing a story.
That’s the case with a story I’ve been working on about Beto O’Rourke, who has become less relevant to the Democratic primary now that he’s polling at only about 3 percent. But the story may help to explain why O’Rourke has struggled so much, so I’m going to resurrect it in brief, bulletpoint form.
Here’s the gist of the argument: O’Rourke is probably competing for young voters more than for older ones, for white voters more than nonwhite ones, and for moderate voters more than for very liberal ones. (His voting record in Congress was fairly moderate, although the policy positions he’s staking out now are more of a mixed bag.) There are plenty of young voters, white voters and moderate voters in the Democratic electorate. But there aren’t that many who are young and white and moderate.
According to the Cooperative Congressional Election Study,1 63 percent of voters in the 2016 Democratic primaries were white, 51 percent identified themselves as moderate or conservative, and 56 percent were born in 1965 or afterward, per the Pew Research Center. Multiply those numbers together, and you’d expect:
63% * 51% * 56% = ~18%
…about 18 percent of Democrats to be all three things at once. That’s enough to form a real base when you’re competing for a party nomination, especially when Democratic rules require you to win at least 15 percent of voters in a state or congressional district to secure convention delegates.
But when you actually look at individual-level voter data, you find something different: Only 12 percent of Democratic primary voters are young and white and moderate. That’s far fewer voters to go around, especially when you’re also competing with, say, Pete Buttigieg for the same voters.
What gives? Well, these various characteristics are correlated with one another, so you can’t just multiply the different numbers together to come up with the right number of voters, which would imply that they were independent from one another. And they’re correlated in ways that are not helpful for Beto (or Buttigieg). Younger Democrats tend to be more liberal than older ones. And white voters — not all whites, but the ones who vote in Democratic primaries — are more liberal than minorities. There are some young, white, moderate Democrats, but not as many as you’d expect.
|Share of Democratic primary electorate|
|Group||If age, race and IDEOLOGY were uncorrelated||Actual|
|Young, white, liberal||16.9%||19.2%|
|Old, white, moderate||14.3||16.8|
|Old, white, liberal||13.6||14.2|
|Young, nonwhite, moderate||10.6||13.8|
|Young, white, moderate||17.8||12.4|
|Young, nonwhite, liberal||10.1||10.1|
|Old, nonwhite, moderate||8.5||8.4|
|Old, nonwhite, liberal||8.1||5.1|
Conversely, there are more young, white, liberal voters (i.e. the Bernie Sanders base) than you’d expect in the electorate than if these characteristics were uncorrelated, and also more old, white, moderate voters than you’d expect (i.e. Joe Biden base). These are the “hot spots” in the electorate, so to speak. After that, the next-largest groups in the electorate are old, white liberals (probably a good group for Elizabeth Warren) and young, relatively moderate nonwhite voters (which could be a good one for Kamala Harris, especially if she tacks more to the center).
That doesn’t explain or excuse all of O’Rourke’s problems. But he has less of a natural base of voters than many of the leading candidates do, which doesn’t only hurt his standing in the polls but also means there are fewer people willing to stand up to defend him when things are going poorly, as they have been recently.