—CONTENT—
Early voting numbers from the May 20 primaries are offering the first real data point of the 2026 cycle — and the map is already looking different than the models predicted.
The Numbers That Surprised Everyone
When turnout models for the 2026 midterms were built last fall, they pointed to predictable patterns: high-interest Republican primaries, historically low Democratic enthusiasm, and battleground states behaving like battleground states. The early voting data that came in from the May 20 primaries is complicating that story in ways that are forcing both campaigns to rethink their ground games.
According to state-by-state returns reported Tuesday, early voting in Nevada, Arizona, and Georgia collectively exceeded 2024 first-round primary early vote totals by a margin that surprised even the most optimistic electoral forecasters. Nevada reported 847,000 ballots cast before Election Day — a 12 percent jump over the 2024 midterms at the same point. Arizona’s mail-in program pulled in over 340,000 ballots by close of business Monday, a 9 percent increase over the 2024 cycle.
What the Partisan Breakdown Shows
The partisan composition of those early returns is where things get interesting — and where the models need recalibration. Republicans in Nevada’s early returns outperformed their 2024 primary-to-primary share by roughly 3 points. But independents — a bloc both parties have been fighting over — turned out at a rate 4 points higher than their modeled share, and their stated issue priorities are not aligning neatly with either party’s playbook.
In Arizona, the early voting picture is complicated by the state’s top-two primary structure, which makes partisan cross-over voting harder to read. But the raw ballot returns suggest that the independent voter surge isn’t limited to one state, and that neither party has a lock on which issues will drive those voters to the polls in November.
“We’re seeing something we didn’t model for in three of our top five targeting states,” said a senior Republican National Committee data strategist who requested anonymity to discuss internal modeling. “The independent voter turnout is real, it’s happening earlier than we expected, and we don’t have a clean story for why yet.”
The Democratic side has its own recalibration problem. Internal models projected a significant enthusiasm gap favoring Republican primary voters heading into the summer. The early voting data from Georgia partially contradicts that assumption — Democratic turnout there outpaced 2024 primary comparisons by nearly 5 points, driven heavily by voters under 40.
The Redistricting Variable
Any analysis of the early voting data has to account for the redistricting rulings that went into effect in three battleground states just before the May 20 primaries. Two federal courts issued rulings in late April that restored district boundaries in Pennsylvania and North Carolina that had been redrawn after 2020. Those rulings — and their effect on competitive House seats — are adding roughly 2.3 million eligible voters to districts that are now considered competitive.
The early voting data from those restored districts is sparse — primaries in those areas are being held under modified schedules — but the registration data tells a partial story. The Pennsylvania 7th Congressional District, restored to its 2020 lines after a court reversal, has seen a net gain of 31,000 registered voters since April, with a partisan lean that leans 4 points more Democratic than the old boundaries.
What It Means for November
Early voting data from a single primary day is a single data point — and electoral modeling is full of cases where a surprising primary turnout figure turned out to be an artifact rather than a trend. The 2022 midterms had several early-voting anomalies that looked ominous for Republicans in June and resolved into comfortable margins by November. Nobody should be rewriting their November models based on one Tuesday’s worth of data.
But the direction of the surprises matters. Both parties entered 2026 with assumptions about enthusiasm, turnout composition, and which voters would show up first. The data from May 20 is cutting against some of those assumptions simultaneously across multiple states. That’s not a noise — that’s a signal that the ground game calculus is shifting before the general election is even fully underway.
The states to watch next are Wisconsin, Michigan, and North Carolina, all of which have primaries scheduled over the next six weeks. The early voting windows in those states open by mid-June. By the time those numbers come in, both parties will have a much clearer picture of whether the May 20 data was a genuine shift or a statistical outlier. Right now, the honest answer is: not sure yet. But it’s worth paying attention to.