Democratic primary turnout is surging, but for many campaigns, that’s where the “good news” ends. In a landscape defined by crowded fields and record-breaking participation, the old rules of voter targeting are no longer just inefficient—they’re a liability.
In our recent fireside chat, DSPolitical President Ryan Fanning and New River Strategies Founder Benjy Messner unpacked why the “New Math” of 2026 requires a fundamental shift from historical binaries to probabilistic distributions.
Download the full recording below to see the complete data breakdown.
The Denominator Problem: Scale vs. Reality
While turnout in states like Texas and North Carolina is breaking records, the participating electorate remains a tiny fraction of the eligible universe. In Texas, Democratic primary turnout rose from 7% in 2018 to 12% in 2026, but that still leaves 88% of registered Democrats who did not cast a ballot.
For campaigns, the takeaway is clear: treating digital like a broadcast medium leads to massive waste. In a race that can be decided by a single percentage point, surgical precision is the only way to protect limited fundraising resources.
The “Coin Flip” Problem: Why History is a Bad Teacher
In the past, campaigns built their “buy” lists simply by looking at who voted in the last few primaries. But in the volatile 2026 environment, relying on old behavior is effectively a gamble with your budget. Ryan shared a striking “validator” from Ohio that proves just how much the ground has shifted:
- The Old Way (Strict History): If you only targeted people who voted in the 2018 primary, your “accuracy” in 2022 was just 52%. In other words, nearly half the people you paid to reach never showed up, and you likely missed a massive wave of new participants.
- The New Math (Propensity Modeling): By using Catalist’s new Democratic Primary Vote Propensity Model, which looks at a wider range of data to rank voters by their likelihood to show up, accuracy jumped to 80%.
The Tactical Playbook for 2026
To win in a fragmented field where voters decide late, Benjy and Ryan suggest shifting your framework:
- Rank Likelihood, Don’t Build Buckets: Instead of “voter vs. non-voter,” use propensity scores to rank-order your entire universe. This allows you to prioritize your spend without accidentally cutting off “net new” participants who don’t have a long vote history yet.
- Meet Irregular Voters Where They Are: Younger and diverse voters—those most likely to sit out without a nudge—aren’t on cable. Reaching them requires a presence where they actually consume media: streaming, digital audio, and gaming consoles.
- Build a “Test-and-Learn” Infrastructure: Use digital to iterate quickly. Start with high-propensity cores, expand as the field clarifies, and ensure your digital insights translate across mail, field, and TV.
The Bottom Line: In 2026, energy creates the opportunity, but strategy determines who captures it. Scale without precision is just noise.
Ready to dive deeper into the data?
Don’t leave your 2026 strategy to chance. Watch the full conversation to see the specific model thresholds and tactical digital applications discussed by Ryan and Benjy.