With Democratic primary turnout hitting record highs, campaigns face a challenge: identifying which voters are actually likely to participate in increasingly competitive and unpredictable primaries.

To meet this moment, DSPolitical launched an exclusive targeting functionality powered by Catalist’s 2026 Democratic Primary Vote Propensity Model. (See more in The Hill.)

By leveraging a deep pool of demographic data and commercial insights alongside traditional voter files, this model allows campaigns to identify exactly who is likely to cast a ballot during primaries. In an environment where every dollar counts,  having the ability to prioritize high-intent voters with precision is the ultimate competitive advantage.

Strategic planning for primaries starts now. Our 2026 Primary Resource Package, which includes an in-depth eBook and a deep-dive fireside chat, provides a roadmap for modern campaign planning and high-impact digital strategy.

To help you better understand how this model can optimize your outreach, we have compiled answers to the most common technical and strategic questions below.

Frequently Asked Questions

What exactly is the 2026 Democratic Primary Vote Propensity Model?

It is a sophisticated scoring system that ranks registered voters on a scale of 0–100 based on their likelihood of participating in the 2026 Democratic primaries. Built by Catalist, the model is customized by state to account for local primary differences. Compared with raw vote history alone, the model gives campaigns stronger accuracy and more flexibility when building target universes.

How is this more effective than just looking at vote history?

Traditional vote history is often a blunt instrument. In backtesting, this model achieved 80% accuracy at a 70+ score threshold. By comparison, targeting based solely on 2018 Democratic primary history yielded only 52% accuracy. This 28-point accuracy lift ensures campaigns aren’t wasting budget on voters unlikely to show up.

Does a score of “80” mean an 80% chance of voting?

No. Scores should be considered a relative ranking, not a probability. For example, a voter with a propensity score of 80 is significantly more likely to vote than someone with a 40. Campaigns should use these scores to rank and segment their audience rather than predict exact turnout numbers.

Which states currently have these scores available?

The model is currently live in 13 key states with May and June primaries:

• California, Colorado, Georgia, Iowa, Maine, Nebraska, Nevada, New Jersey, New Mexico, New York, Ohio, Pennsylvania, and Virginia.
• Additional states with August and September primaries will be released as data becomes available.

Is this used for persuasion or candidate support?

This model is not used for persuasion or candidate support as it predicts participation. To build a complete strategy, campaigns should layer this model with other tools to determine which of these likely voters are also the most persuadable for their specific candidate.

Can I use this to find new registrants?

Because the model relies heavily on historical and commercial data, new registrants often lack the footprint to receive a high score. If your campaign strategy focuses specifically on mobilizing brand-new voters, we recommend building a separate target universe alongside this model.

How do I activate this model for my campaign?

Access is available exclusively through DSPolitical. We can help you set specific score thresholds—choosing a higher score (70+) for a small, hyper-accurate audience or a lower score (30+) for broader reach. These audiences are custom builds that can be activated through managed service or via Deploy with no additional data costs.

Ready to Build Your Winning Audience?

Don’t leave your primary strategy to chance. Whether you have technical questions about the data or are ready to see the specific voter counts in your district, our team is here to help.

Contact us below.