The social media wars are close at hand.
Like it or not, the 2028 election is going to be fought online. Already, would-be contenders are gearing up for the kind of permanent full-court press that would make even a 2020-era strategist shudder. That prospect has had me wondering more about the online world’s actual connection to election outcomes. So I decided to find out.
For the past six months, I have been tracking the follower counts of Democratic candidates in competitive statewide primaries. My criteria were specific:
Senate and gubernatorial primaries in which there was no incumbent,
At least two candidates held notable social media following,
Only one Democrat emerged from the primary,
And more than two candidates performed in the double digits on Election Day.
This eliminated, for example, races in which only one candidate maintained a social media presence, making comparison impossible, or only one candidate. Follower counts were evaluated the evening before or day of an election so as to capture candidates’ standing before results occurred. This left thirty races total, a notable pool from which to conclude.
With primary season now nearly wrapped up (the only race potentially falling into our category is New Hampshire’s Senate primary, a potentially interesting race but one that won’t change the data here), I’m happy to share the results here.
“Court Watch is credited for discovering some of the most high profile indictments of our time that could have very easily been overlooked” – CBS News
Court Watch is a news site covering all the most interesting federal court filings, from search warrants, indictments, to everything in between. Our original reporting has spurred hundreds of national and local media stories. Subscribe for free and get the news before it becomes news.
The leading platforms
The platform most predictive of eventual primary winners was Instagram: in 68% of applicable races, the candidate with the most followers on Instagram won. At the very bottom was Threads — interestingly, Instagram’s Meta sister — where having the most followers correlated with victory only half the time.
There do exist some real differences when races are broken down by category. In gubernatorial primaries, Twitter followings became the most predictive. In Senate races, Instagram and YouTube came out on top. Interestingly, Threads remained the least indicative in both race types.
There is some indication that the outcome of gubernatorial races, which have long been less nationalized than Senate campaigns, is less connected to social media numbers. Out of all thirty races, only three candidates — Hannah Pingree in Maine, David Crowley in Wisconsin, and Xavier Becerra in California — emerged victorious from primaries despite not having the most followers on any platform. All three were running for governor. By contrast, the winner of every Senate primary outperformed the loser on at least one of the measured platforms.
Official vs. campaign platforms: a warning sign
From the beginning, I took note of the difference between official accounts, which are managed by a politician’s government staff and bound by certain ethics rules, and campaign accounts, which are free to function however a candidate pleases. In more than a few instances, the distinction mattered a great deal. This was seen most strongly in the Texas Senate primary.
By Election Day, Rep. Jasmine Crockett’s campaign accounts only barely outpaced James Talarico on Instagram and Facebook (2.2 million vs. 2 million, and 987k vs. 936k, respectively). But Crockett’s political celebrity had created mammoth followings on her official accounts, bringing her an additional 2.8 million followers on Instagram and 1.2 million on Facebook. That underscores an interesting nuance in the numbers. On the one hand, Crockett’s overall presence on the platforms, factoring in both campaign and official accounts, dwarfed Talarico’s, making his victory look even more impressive in one light. On the other hand, Talarico’s victory points to the idea that personal accounts, rather than official ones, are the better metric of candidates’ support.
Other races provide evidence for that theory. In Illinois, Rep. Raja Krishnamoorthi, who spent years preparing for a Senate campaign, had built up a large following across his official accounts — the biggest by far of the three major candidates. But on every platform, his campaign accounts lagged the eventual primary winner, Lt. Gov. Juliana Stratton.
The 2028 implications
The disconnect between official and campaign accounts becomes particularly notable when evaluating the 2028 Democratic contenders. I analyzed the candidates who currently occupy the RealClearPolitics top ten, and, to begin with, one fact is plain: former Vice President Kamala Harris and Rep. Alexandria Ocasio-Cortez are in a completely different league than the other potential candidates when it comes to overall followers. Harris is a particular behemoth — when excluding official accounts, she nearly doubles Ocasio-Cortez.
Among the top 2028 contenders, governors face deep challenges in the social media race. As seen above, three of the four bring up the rear when it comes to total follower count. But beyond top-line numbers, something else caught my eye: governors are, by far, the most dependent on their official accounts.
Among the top ten candidates, the four governors have the highest percentage of followers coming from accounts they can not use for future presidential campaigns. Two of them, Pennsylvania’s Josh Shapiro (70%) and Illinois’s JB Pritzker (55%), derive a majority of their followers from official accounts. Kentucky’s Andy Beshear (43%) and California’s Gavin Newsom (39%) are in less extreme positions. But they are still a far cry from the rest of the field — Pete Buttigieg and Jon Ossoff, for example, land at roughly 10% and 8%, respectively.
This may well become quite relevant after the midterms, when the 2028 primary could pick up in earnest, and candidates will begin to rely on their existing accounts to fundraise and build name recognition. While the line between personal and official accounts has blurred in recent years — both parties now use government-run accounts for blatantly partisan messaging — it’s difficult to imagine candidates getting away with open electioneering or fundraising.
At the same time, as the 2026 primary season has proven, followers do not directly correlate with eventual performance. And right now, there’s evidence that some figures, especially Shapiro, are polling far higher than their social standing.
Still want more? Check out the full, raw data here.
A note on methodology: All data was compiled, first calculated, and checked by me. I then used Claude to check calculations, fact-check numbers, and create two of the charts above.
Worthy reads
Noah Wyhof-Rudnick has an interesting look at the ways prediction markets, which are gobbling up an increasing amount of attention among both amateur and professional politicos, overrate certain candidates.
Politico Magazine’s Jeremy White has a long read diving into the ideological ambiguity of Xavier Becerra, who is likely to lead the largest state in the country after November.
The Atlantic’s Elaine Godfrey has a fascinating profile of Rob Sand, Democrats’ nominee for governor of Iowa (early readers will recall my interview with Sand was TLR’s first-ever edition)
J. Miles Coleman at Sabato’s Crystal Ball analyzes how progressives’ base of support in 2026 primaries looks distinctly different from Bernie Sanders’ 2016 coalition.
In The Argument, Cassie Pritchard makes a compelling, empathetic case for a different approach to achieving transgender equality.
Patrick Ruffini, a conservative pollster, presents a warning that summer polls could be overrating Democrats in key races.
And Nate Silver has a response to Ruffini, arguing the dynamics of the political landscape are fundamentally stable.
Got feedback on today’s newsletter? Email me [email protected]