Loading migration data.
Preparing county corridors and profiles…
Preparing county corridors and profiles…
Net migration is movers in minus movers out over the stated period, per 1,000 residents on July 1 at the end of the period (Census Population Estimates). IRS counts tax returns and the people on them (N2); PopEst estimates all residents.
IRS full county headers cover 3,087 counties and Connecticut planning regions; Census estimates cover 3,144.
The map and county dossiers show net migration from complete county headers, not only from corridors that can be published pair by pair. Two sources answer the same question in different ways: IRS counts tax filers and the people on their returns; Census Population Estimates model all residents.
Address on the return filed in calendar year 2022 vs the one filed in 2023. IRS 2022 series (enhanced address matching from 2022–23); not comparable with earlier years without an overlap design. People are IRS N2 counts: exemptions through 2018–19, individuals from 2019–20. Census periods run July 1 to July 1. Residents are Census estimates for July 1 at the end of each period. Where IRS withholds part of a county's header, no IRS value is shown; missing is not zero.
Ranked by IRS net people and by Census net domestic migration, counties agree with Spearman 0.850 across 3,087 counties for IRS filing years 2022→2023 (Census V2025); across 13 computed years it ranges from 0.585 to 0.851. Published direct corridors alone agree far less (Spearman 0.687 vs 0.865 for full headers, Census V2023; the sign differs in 578 counties), which is why the map no longer uses them.
Connecticut · nine planning regions
IRS and Census report Connecticut as nine planning regions from 2022. Values are not allocated to the eight legacy counties, which show no Movement value on the map.
The same regions in the corridor view · 2022→2023
Moves are tax returns between two places that IRS publishes; smaller flows are left out. Each region's 2020 vote is the sum of its towns in Connecticut's official Statement of Vote. The regions carry no metro label for these years, because the metro definitions then in force predate them. Connecticut's two official sources differ on Trumbull's Republican vote; this uses the Statement of Vote, and Greater Bridgeport is safe D under either figure.
This view keeps only exact published direct county-to-county corridors and labels each county by its 2020 presidential result. 6,470,553 tax returns are classified, covering 99.82% of comparable published returns. Gross corridor counts are larger from Democratic-won counties to Republican-won counties. Normalize by distinct origin-county population and the direction reverses. Both facts belong together.
74.4% of D-won→R-won and 75.6% of R-won→D-won published returns are intrastate.
Within the same metro alone: 503,715 D-won→R-won returns and 361,847 in the reverse direction. This is consistent with ordinary regional housing mobility, not proof of a political cause.
IRS filing years 2022→2023: the address on the return filed in calendar year 2022 vs the one filed in 2023. 26.76% of domestic county-header returns are in IRS residual "other flows" buckets and cannot be assigned to any corridor or context; every ratio here describes published direct corridors only.