/ About / Archive / Other media
Nobody is on Twitter, revisited - 11/19/18
A few years ago, Matt Bruenig pointed out that Nobody Is On Twitter. Since then, research (Pew in particular) has given us much better data than we had at the time, giving us a much better picture of just how tiny Twitter's audience actually is. Since I was poking through some of it today, I thought I might as well post a few of the more interesting numbers I came across. First, on how often people use Twitter:

These numbers are based on Pew's latest survey on social media use, as well as reporting on dead accounts by PC Mag. The baseline number Matt mentioned in 2015 - 23% - is pretty close to the 24% we see here, but once we break it down it becomes clear that this number is significantly inflated.
  • Most US adults with Twitter accounts (24%) aren't online every day (13%), and in fact, 44% of that first number may never use their account at all. That leaves about 13% of US adults who are Active Posters.
  • People who are online every day - call them True Posters - account for only 11% of all US adults.
  • Extremely Online Posters - the people who, intuitively, prove most of our content - only account for 6% of the US adult population.
Even these numbers, however, overstate things: not everyone, it turns out, uses Twitter to talk about the news:
  • One recent survey shows that only about 11% of the population ever gets their news from Twitter.
  • Among respondents who ever use social media to get the news there is also significant variation in consumption: for instance, 30% of that group say that they "hardly ever" get their news from social media. If these trends holds for Twitter, then it is probably a significant source of news for something closer to 3-8% of US adults.
This, I think, is a much more realistic assessment of Twitter's news reach: bring together all of the blue check journalists and unverified posters, the sinister operatives and the doe-eyed normies, the Pepe teens and the Gritty teens, the #TCOT grandpas and rose emojis - put them all together, and you are reaching a single-digit percentage of US adults, somewhere between 3-8%.
Democrats improved their performance among Republican constituencies in 2018 - 11/11/18
I've published a new piece in Jacobin, Vox Is Wrong About the Midterm Elections, discussing how Democrats performed in 2018 among various demographic constituencies that have historically voted for Republicans.

This month's election has placed liberal pundits in an awkward situation. On one hand, they remain implicitly committed to the notion that Democrats cannot win over certain socioeconomic demographics - like "rural voters" - without betraying their core values. But on the other hand, by making inroads among those voters, Democrats just won major victories all over the country, and struck a powerful blow against Donald Trump.

The consistent thing would be to insist that these victories were ill-gotten, and that Democrats should have done worse with low-educated voters or white Southerners. The temptation, however, is to pretend that they did do worse - and that's what seems to be happening in liberal pundit world. In unison, writers like Zack Beauchamp have doubled down on their demographic arguments from 2016, insisting that low-educated voters and white Southern voters are unreachable even though the evidence is clear that 2018 Democrats just reached them.

That's the evidence I lay out in Jacobin: Democrats improved their performance in this election among just about every demographic you can name. I largely rely on exit poll measures of voter preference because that's what Beauchamp uses; you can come up with more sophisticated measures if you like (accounting for turnout and whatnot), but I don't see any evidence that they overturn the underlying trend. How so many pundits have arrived at an understanding of the election flatly contradicted by a basic review of the polls should raise some serious questions that I probably don't need to spell out.
Left entryists have fought to take over the Democratic Party. 2018 is their test. - 11/5/18
Three notes about Tuesday's election:

1. This is a test for entryism.

Wherever one comes down on Democratic entryism - the only way forward, necessary but insufficient, a suboptimal political investment, a waste of time, or counterproductive - it seems clear to me that tomorrow has to be regarded as a great empirical test. Nearly three years ago, Bernie Sanders laid out his theory of political revolution, arguing that
by building a movement among average Americans, he'll be able to win elections, defeat special interests, push liberal reforms into law, and build an economy that works for everyone.
Even in defeat, the Sanders campaign became the catalyst for a massive mobilization of left activism; and ever since then, a whole field of left organizations and activists have worked to build campaign infrastructure, recruit candidates, and compete in primaries and in the general. This is by several orders of magnitude the largest entryist movement in my lifetime, and probably of the past half century.

2. The bar has to be set high.

Previous elections guaranteed that Senate Democrats would face nearly insurmountable odds in this one, defending 26 seats to the GOP's nine. On the other hand, previous elections have also dramatically overextended Republican control of both the House and governorships around the country. And the contrarian rhythm of midterm elections - inclining Americans to vote against whoever happens to be occupying the White House - ensured on November 8, 2016 that today's Democrats would have the wind at their backs.

So it really is not enough to point to a "blue wave" of Democratic victories and take this as proof that entryism worked. In fact, even if Democrats exceed the most optimistic expectations - say by taking control of the Senate or winning some dramatic upsets - leftists still shouldn't take that as vindication of entryism. Good news, perhaps, but not evidence that the entryist strategy actually worked.

For entryism to work, a significant number of seats have to be won by left-flank challengers who defeated the Republican and the liberal Democrat. What qualifies as "a significant number of seats" and who qualifies as "left-flank challengers" will of course be up for debate, but these are the terms on which the debate over entryism needs to proceed.

3. The left will face significant pressure to lower expectations.

After Trump's victory and two years of Republican rule, the liberal-left badly wants a win. US liberals will also continue to equate "unity against Trump" with "submission to the Democratic agenda," a form of ideological passive-aggression to which the left remains regrettably vulnerable. For these reasons, leftists will probably not be content to celebrate Democratic successes as successes against Republicans; they will be tempted to celebrate any win with a D on it as a win for left-flank entryism. Attempts to hold entryists to even minimal standards of progress will likely be dismissed as contrarian, pedantic, and divisive.

But this is precisely what the left has to do if we want to hold the entryists accountable. I am not entirely sure how we go about doing this, because as mentioned, the threshold for success is up for debate. But I am sure about two things: at some point there needs to be a reckoning for entryism - and the forces of co-option will do everything they can to resist it.
The voter turnout problem they don't talk about - 11/2/18
Every election year guarantees a round of takes on how the kids are ruining our elections, and 2018 has proven no exception. In just the past few days we've seen multiple articles from the same NBC News / Genforward survey:

The piece that best captured this genre of take, however, was NYMag's 12 Young People on Why They Probably Won’t Vote - a brilliant one-two punch of inflammatory quotes and no data. If you need someone to blame for the US's catastrophically low turnout rates - a genuine and humiliating anomaly by international standards - look no further than some kid who says that "mailing me anxiety".

Anyway, here are the Census numbers on average voter turnout over the past decade, by age:

And here are the same numbers, broken down by income:

Income's effect on voter turnout is at least comparable to age - arguably, it's even more consequential. So why don't we get articles like "12 Poor People On Why They Probably Won't Vote"? Why doesn't NBC bother to survey people with a family income of less than $30,000 a year? I have two theories:
1. Corporate media outlets, their advertisers, and their target audiences generally do not care about the poor. They have little interest in what the poor think about our politics or our candidates, and certainly not enough to assign an article, commission a poll, conduct interviews, or read about it. 
2. Turnout problems among the poor don't lend themselves as easily to "blame the voter" narratives - they cry out for more systematic explanations, and leave open the possibility that our democracy just isn't giving the poor compelling candidates. Conversely, turnout problems among millennials work quite naturally with popular narratives about how young people are irresponsible, selfish, frivolous, and so on - that is, about how the voters only have themselves to blame.
Incidentally, turnout problems among youth and turnout problems among the poor are not unrelated - but the important thing is that we avoid, at all costs, talking about class. Remember that the next time the media tries to blame the problems of American democracy on some kid with anxiety issues.