Conservatives on Twitter are buzzing about this as a sign that Trumpmania may be overblown. If Trump fans aren’t likely to actually turn out for him once we start counting votes, who cares what his polls say? (Maybe not coincidentally, Rasmussen’s post-debate poll of likely Republican voters was the worst one for Trump this week.) I was surprised at the numbers at first, thinking Trump’s fan base is probably mostly composed of older voters. Older voters tend to skew right, after all, and older voters would be more familiar with him from his decades of fame. And older voters famously turn out in high numbers on election day. Shouldn’t the eligibility numbers among his fans be high, not freakishly low?

But that’s the wrong way to think about this.

Trump finishes at or near the bottom in five of the six social-media demographics tested: His audience is the least likely to be eligible to vote, the least likely to come from key early states, the least likely to include women, Christians, or college students. What gives? I think that’s easily answered, at least in the eligibility category. Namely, most of Trump’s social-media audience was built before he got serious about politics this summer. He has 3.7 million Twitter followers, a gigantic number. To put that in perspective, Marco Rubio’s followers number around 815,000. If you’re following Rubio on Twitter, it’s because you’re into politics. If you’re following Trump, it’s most likely because you’re into “The Apprentice” — or possibly because the guy’s been using his Twitter feed for years to attack other celebrities who’ve criticized him, making his one of the most gawkworthy accounts on the platform. Simply put, Trump’s social-media following contains an ocean of people who like TV but may not care much about politics, even to the point where many aren’t registered to vote. That’s where that 39.4 percent figure is coming from. Importantly, though, that doesn’t mean that Trump’s primary polls are overstating the support he can expect on election day. It may be that the 25 percent he’s pulling in 2016 surveys represents righties who are passionately committed to his cause and will turn out for him. It’s just that those righties are a smallish percentage of the total number of people who follow Trump on social media for a wide variety of reasons. In fact, like Ace says, this gives Trump a unique opportunity to introduce apolitical people into the political process. If he can get some of his “Apprentice” fans who’ve never voted before to register and support him in the Republican primaries, he’d be a deadly serious threat to win.

As for the other results here, I can buy that his following is mostly male and that he’s not a huge draw for those with “strong Christian values.” I can buy that his appeal is more working-class, i.e. a smaller percentage comparatively of college students. I can also buy that the percentage of his total following is unusually small in the early states. That too is an artifact of his TV fame, I think: Trump probably has much larger percentages of followers from populous blue states like New York and California than the rest of the competition thanks to “The Apprentice.” It’s not that he has fewer total followers in Iowa than, say, Marco Rubio — the opposite is true, almost certainly — it’s just that his total followers are likely more evenly distributed nationally thanks to his celebrity than those for Rubio, who probably has big clusters in his home state of Florida and the early presidential states where voters are starting to pay attention to him. Exit question: What does the “center-left” row in the graphic above signify? It’s not explained anywhere at the link, and it makes no sense to me that Trump would lag in that metric. If his fame is still mostly apolitical, he should be the most likely candidate here to attract center-leftists. Liberals watched “The Apprentice” too, after all.