Use the stepper to walk through the story · ← → keys work too
Source: PA Dept. of State precinct returns & registration files, 2018–2025
Click to step through →
Six elections of results (2018–2025), checked against party registration in all 430 Montgomery County precincts. Step through the story below.
These precincts vote bluer than their registration predicts in every kind of election, most likely a mix of independents crossing over and extra independents on the rolls. Canvassing and persuasion have a proven track record.
Registration is moving Democratic fast (+11 and +8 points since 2018). Registration drives and welcome outreach here ride a trend already in motion.
Performance depends on who shows up. Royersford 2 beats expectations by 7+ in quiet judicial years but is only average in presidential years. GOTV territory.
All 34 Area 4 precincts, ranked. Click a column header to sort. Click a row for that precinct’s full history.
Want to slice it yourself? Download the full data, one row per precinct with per-election registration, vote share, and deviation columns: Area 4 detail (CSV) · all MontCo precincts (CSV)
Data. Precinct-level election returns and voter registration counts from the Pennsylvania Department of State bulk data portal, for the 2018 and 2022 gubernatorial, 2020 and 2024 presidential, and 2023 and 2025 statewide judicial elections. All 34 Area 4 precinct boundaries were verified stable across the full period. A handful of precincts elsewhere in the county that existed for fewer than five of the six elections (mainly congressional-split ballots retired after the 2022 redistricting) are excluded from six-election averages.
Expected vote. For each election, we fit the county-wide relationship between a precinct’s Democratic registration share and its Democratic vote share across all ~430 Montgomery County precincts (unweighted pooled least-squares with election fixed effects; R² = 0.92, which includes those election-year baselines). Vote share is the Democratic share of all votes cast for the office, not the two-party share. Each election is measured against its own contemporaneous registration snapshot, not today’s. “Performance vs. expectations” is the gap between a precinct’s actual Democratic vote share and what the county-wide pattern predicts for a precinct with its registration, in that election.
Races used. President (2020, 2024), Governor (2018, 2022), PA Supreme Court (2023), and PA Commonwealth Court (2025): the highest-profile Democrat-vs-Republican contest on each ballot.
Caveats. Six elections reveal consistent patterns but cannot establish causes. Typical precinct noise is about ±3.5 points in any single election, which is why we emphasize the six-election average and consistency rather than any single result. A precinct’s deviations are correlated across elections, so a 6-for-6 record is common (nine Area 4 precincts are above expectations in all six elections; the six named in the story are the largest). The model predicts from Democratic registration share alone; as a robustness check we re-fit it with independent registration share added, and all six named precincts remain above expectations (five of six in all six elections), though the gaps shrink somewhat because part of the effect is simply having more independents on the rolls. The model specification went through independent review. This page contains only precinct-level aggregates of public records, with no individual voter data.