Kamala Harris can't meme her way to victory. Or can she?

The rules may be different when it’s a sprint and not a marathon.

I know I’m doing a lot of media for the book — which comes out on Tuesday! — but I suspect newsletter readers will get a lot out of this interview with Tim Miller at The Bulwark, which hits on some of the themes in this column along with book stuff.

---

Two weeks ago, I almost wrote a newsletter along these lines of this typically well-argued column by Freddie deBoer, which critiqued the nascent Kamala Harris campaign for catering too much to the tastes of Very Online progressives instead of swing state voters, saying that it reminded him of Democrats’ failed strategy in 2016:

And many seem intent on remaking a core 2016 mistake: acting as though the Democratic candidate’s job is to become the President of Online rather than the President of the United States, begging Harris to devote her campaign to memes and social media, playing to people like them instead of the middle class white retirees in Wisconsin and Arizona who will actually determine this election. 

I’ll push back on some elements of the Harris-Clinton comparison in this column — but to be fair, my conclusion was going to be pretty similar to deBoer’s. Sometimes you get a little lucky with the stories you don’t publish, because I wonder if we’d be at the point where I was taking a little shit for it. Since the column I never wrote — and despite a meme-ified, Very Online strategy — Harris has only moved up in our forecast, going from a 38 percent chance of winning the Electoral College at model relaunch to a 54 percent chance now. (That doesn’t yet account for a trio of strong New York Times/Siena swing state polls for Harris, which will be incorporated in our model run later today.)

The vibes these days.

Meanwhile, the mood among Democrats online has become exuberant to the point of almost being rabid, particularly any time that JD Vance opens his mouth. It’s possible that this is some sort of sugar-high period. Democrats — especially younger Democrats — self-evidently had a lot of pent-up energy after three cycles (Clinton, Biden and Biden again) of older, boring candidates that they were told to fall in line behind. Harris — though technically a Boomer by a few months — is a comparative breath of fresh air.

It’s also possible that this burst of Democratic enthusiasm is causing some degree of partisan non-response bias in polls and inflating Harris’s numbers. I’m not trying to be a downer to our Democratic readers: there’s an obvious bull case for Harris based on her momentum — and the fact that even the Trump campaign is acting as though it’s behind. But there are reasons why the forecast is basically 50/50 and not 70/30 Harris or something. Some degree of mean-reversion is usually the safe bet. And deBoer and I still have at least three good reasons to be concerned about a Way Too Online campaign:

• Online subcommunities vastly overrate how representative their views are of the larger electorate;

• Memes can giveth but also taketh away: online focal points shift quickly;

• There are some things you can’t meme your way out of. Voters could see with their own eyes that Joe Biden was too old or that there was a ton of inflation early in Biden’s term. Polling suggested that when other Democrats tried to downplay age-related concerns about Biden, it not only didn’t work but backfired, making them seem like liars.

And yet there’s a part of me — just a part — that would buy a ticket to ride out the memementum. That thinks this is so crazy that it just might work. That thinks it was probably a mistake for Harris to pick Tim Walz instead of Josh Shapiro, but that not wanting to disrupt the campaign’s favorable trajectory with a higher-risk pick was a valid consideration. The case rests on three factors that are somewhat unusual to the circumstances of this campaign:

1. It’s a sprint, not a marathon

From Biden’s withdrawal on July 21 to Election Day on Nov. 5 is 107 days. That’s a long time by international standards — our friends in the United Kingdom just had an entire election in six weeks — but short as compared to modern American elections, which can last for almost two years. (Barack Obama was de facto running for the 2008 election by December 2006, for instance.)

I’m very much a believer in playing the long game, and that’s one reason I hate any strategy that involves spreading bullshit and hoping you don’t get called on it. As Mike Pesca points out, for instance, partisan media critics that scolded the press for covering obviously valid questions about Biden’s age wound up embarrassing themselves — and for that matter didn’t really do any favors to Democrats, either. Biden just wasn’t going to be able to duck-and-weave his way through an entire campaign by working the refs. 

But Harris? Well, Donald Trump is in the midst of a terrible news cycle. Then the DNC starts in 9 days. Parties typically get a boost in their polling following their conventions, which makes it easier to sustain the good vibes. And then it will be Labor Day, the traditional stretch run of the campaign. While there will probably be a difficult news cycle or two for Harris — that’s why the Trump campaign is asking for more debates, it wants more opportunities to trip her up — it’s possible she just speed-runs her way through 11/5.

2. All candidates except Trump are poorly-defined, meaning memes can make a bigger difference

Have you ever heard of a brand of cheese named Moonvale? Well, hopefully you haven’t, because I just made it up — or actually, ChatGPT made it up. (I chose from among its list of suggestions for appealing-sounding cheese brands.) I think it’s a pretty good name. And if I worked for a big advertising agency, I’d love to be assigned to the product because it’s such a blank canvas — much easier to change perceptions of “Moonvale” than Kraft or Velveeta or Kerrygold.

That’s essentially where Democrats are right now with Tim Walz. Look, I cover politics for a living, but if you’d asked me three weeks ago what I knew about Tim Walz, I’d have told you that he’s the governor of Minnesota and I could probably have remembered that he used to be a Congressman, but that’s about it. The Walz brand starts out with vaguely positive connotations, but otherwise he’s a blank slate. There’s a race to define him, and early definitions can be sticky. And Democrats are probably winning the race — branding Walz as some sort of likable but occasionally salty Sitcom Dad — in part because the Trump campaign was caught oddly flat-footed by Democrats changing their candidate.

And Democrats are definitely winning the race to define JD Vance, who is the least well-liked VP candidate since at least 1980, according to data compiled by my friend and former colleague Harry Enten. Now, I’d argue that Vance gave Republicans poor material to work with — when I asked ChatGPT for a list of moderately unappealing cheese names, the suggestion I liked the most was “Sourwood Swiss”. It’s not a hopelessly gross name — better than some other ChatGPT suggestions like “Moldy Meadow Manchego” — but something that very much sounds like an acquired taste. One that would require a careful rollout plan, with a lot of message discipline — never a strength of a campaign involving Donald Trump

And Harris wasn’t really all that well-defined either, as evidenced by the fact that her favorable ratings have already undergone some big (and positive) swings. Ironically, it may have helped that she exited the 2020 primary campaign so early — two months before Iowa — before perceptions hardened, because it was an awful campaign.

Conversely, perceptions of Biden and especially Trump were really locked in. Even before the debate, one reason to be pessimistic about Biden’s chances was that absolutely nothing was moving the polls — Trump’s felony conviction may have shifted then by half a point toward Biden, but not by more than that. That may have been why the White House was banking so much on the polls being skewed — it felt almost inevitable that Biden would go into Election Day trailing by a couple of points in Pennsylvania and Wisconsin and Michigan, hoping for a modest polling error to draw an inside straight to exactly 270 electoral votes.

3. The media will probably play along

The mainstream media has several different motivations. It can often have a liberal bias as that term is typically defined and it also can have a bias toward centrist both-sides-ism — this is the Indigo Blob theory. And then there’s the bias it pretty much proudly owns: it “roots for the story”. It likes drama, it likes novelty, and it likes unexpected turnarounds and comebacks.

And it likes underdogs. As Matt Yglesias points out, Harris is a great underdog story. Often treated dismissively even by people in her own party — and derided as a “DEI hire” by Republicans — she seized the nomination in a moment of peril for her party and proved to be a much better campaigner than nearly anybody expected, unburdened by what had been. And she could be the first woman president, but without any of the boring “inevitability” of Hillary Clinton. It’s a really good story arc, probably good enough to sustain the media’s appetite for drama through November.

How the tug-of-war between the partisan and nonpartisan elements of the Indigo Blob will play out on stories like Republican claims about Walz’s military service is harder to say. You’re beginning to see some critiques from thoughtful journalists about Harris’s lack of press availability and lack of policy specifics — and along other lines, critiques about the JD Vance couch jokes, which whether funny or not¹, are misinformation.² But it probably helps Harris that the press was pretty tough on Biden toward the end — deservedly tough, I’d say, but tough — and also that the centrist side of the Indigo Blob sort of won the fight against the partisan side when Biden dropped out.³

Also, again, we don’t have that long to go. Within the Indigo Blob, stories that are against Democratic interests are sort of all-or-nothing. They usually don’t congeal into critical mass, but when they do there can be a real pile-on (see e.g. But Her Emails). The Harris campaign can probably keep a lid on this stuff before a largely sympathetic press.

Maybe none of this should be surprising

In our model, Harris is projected to win the popular vote by 2.5 percentage points. If that sounds familiar, it’s because it’s right in the range of Clinton’s margin against Trump (2.1 points) and not so far from Biden’s margin in 2020 (4.5 points). It also almost exactly duplicates the average popular vote margin in the past six elections dating back to Bush-Gore, which is D +2.6.

So while Harris’s surge undoubtedly feels pretty great for Democrats after contemplating Biden’s near-inevitable defeat, in some ways she’s just performing like a league-average Democrat. Trump isn’t popular, and Democrats usually win the popular vote. The Electoral College is a big problem for Democrats, but our model shows the Electoral College-popular vote gap as being slightly smaller than in 2016 or 2020.

Maybe it’s not the memes, in other words. Polling from Democratic groups suggests that the “Republicans are weird” message is effective with the base, but not particularly with swing voters. Maybe this was the race we were going to get all along when Democrats replaced an 81-year-old candidate who was clearly in no condition to be president for another four years with a normal and competent one.

The Harris campaign is going to need more cushion, because it could very, very easily lose the Electoral College in an election held today — outside of that NYT polling, the race still looks pretty close to a toss-up — and it still faces headwinds like economic volatility, immigration, and Biden’s baggage. Locking down your base is what gets you to a tie, but it’s not enough to put you over the top — Democrats don’t even have a party ID edge any more (about as many voters identify as Republican as Democratic). You need swing voters for that. Let’s give the memes credit for (1) unifying the base around Harris; (2) restoring Democratic enthusiasm; (3) negatively defining Vance and perhaps (4) positively defining Walz⁴. That’s quite a lot, actually. But deBoer’s warning about the potential for the campaign to turn into a replay of 2016 shouldn’t go unheeded. If Harris’s campaign is acting with a lot of confidence lately, the Clinton campaign had plenty of confidence, too.

1. My verdict: reasonably funny when they started in Weird Left Twitter, and not very funny once they spread to Resistance Twitter.

2. I don’t really buy that it’s not misinformation because it’s a joke — not everyone knows it’s a joke, and it’s a joke that’s being weaponized for political purposes. Though I’d rather the whole “misinformation” label be retired.

3. Among other things, the centrist side tends to be more poll- and horse-race driven and to cover politics as a strategic game — and Democrats clearly made the right strategic move when Biden dropped out.

4. It’s too soon to be sure about this last one.