US Migration Flow Atlas: the metro network behind IRS moves
Which US metros gained or lost tax-return households, how much of that movement maps directly between RentMarker metros, and which counterpart market leads each direction?
National totals can hide the network carrying movement into and out of a metro. This atlas keeps the IRS all-US totals as the benchmark, then exposes only directly matched inter-metro county-pair flows.
- Complete metro network
- 696 all-US totals plus direct inbound and outbound pairs
- Directional pairs
- 16,928 exact inter-metro routes
- Mapped tax returns
- 3,449,254 counted once on the inbound side
- Median net rate
- +0.2 / 1k IRS households per 1,000 ACS residents
A net total can hide two large, differently distributed directions
The complete network contains 696 metros and 16,928 directional market pairs. Those pairs carry 3,449,254 directly mapped tax-return households, each counted once through its destination. The all-US inbound and outbound totals remain the denominator, so movement outside a current RentMarker metro never gets reassigned to a convenient nearby market.
Jefferson, GA has the highest included net inflow rate at 21.4 tax-return households per 1,000 residents. At the other end, Edwards, CO records -6.8 per 1,000. The rate controls for market size; the table keeps the underlying counts visible.
IRS migration describes filing-unit movement. It does not identify renters, motives, household formation or the housing demand created by a move.
The largest direct corridor runs from Los Angeles, CA to Riverside, CA with 41,500 tax-return households. Open the migration atlas to change direction, inspect coverage and see whether one corridor dominates or the flow is spread across many counterparts.
Net tax-return household movement, population-adjusted
Bars preserve direction around zero; the table and atlas retain the underlying counts and direct corridor coverage.
The first 24 rows, with a path to the underlying market
| Rank | Metro | Net households | Per 1,000 | Mapped in | Mapped out | Top inbound | Top outbound | Open |
|---|---|---|---|---|---|---|---|---|
| 1 | Jefferson, GA4,506 in · 2,692 out | 1,814 | 21.4 | 67.5%5 counterparts | 54.3%3 counterparts | Atlanta, GA1,952 returns | Atlanta, GA712 returns | Map flows |
| 2 | The Villages, FL8,009 in · 4,998 out | 3,011 | 21.0 | 52.3%43 counterparts | 50.9%18 counterparts | Orlando, FL1,066 returns | Orlando, FL899 returns | Map flows |
| 3 | Spearfish, SD1,427 in · 986 out | 441 | 16.2 | 19.1%3 counterparts | 29.6%2 counterparts | Rapid City, SD212 returns | Rapid City, SD264 returns | Map flows |
| 4 | Myrtle Beach, SC15,027 in · 9,094 out | 5,933 | 15.5 | 60.8%91 counterparts | 43.1%55 counterparts | New York, NY1,538 returns | Charlotte, NC428 returns | Map flows |
| 5 | Ocala, FL15,821 in · 9,892 out | 5,929 | 14.8 | 68.2%68 counterparts | 57.0%37 counterparts | Orlando, FL2,039 returns | Orlando, FL1,148 returns | Map flows |
| 6 | Homosassa Springs, FL6,689 in · 4,382 out | 2,307 | 14.2 | 54.4%21 counterparts | 45.7%13 counterparts | Tampa, FL1,484 returns | Tampa, FL672 returns | Map flows |
| 7 | Wilmington, NC16,734 in · 11,089 out | 5,645 | 12.5 | 46.7%52 counterparts | 40.4%39 counterparts | Raleigh, NC1,321 returns | Raleigh, NC871 returns | Map flows |
| 8 | Lewisburg, TN1,324 in · 902 out | 422 | 11.7 | 56.6%2 counterparts | 48.0%2 counterparts | Nashville, TN677 returns | Nashville, TN367 returns | Map flows |
| 9 | Lakeland, FL30,845 in · 21,697 out | 9,148 | 11.6 | 81.6%109 counterparts | 71.1%87 counterparts | Orlando, FL8,908 returns | Orlando, FL5,458 returns | Map flows |
| 10 | Port St. Lucie, FL19,797 in · 14,250 out | 5,547 | 10.6 | 70.6%46 counterparts | 57.7%43 counterparts | Miami, FL8,105 returns | Miami, FL2,775 returns | Map flows |
| 11 | Panama City, FL8,111 in · 5,865 out | 2,246 | 10.6 | 38.6%47 counterparts | 35.8%34 counterparts | Crestview, FL536 returns | Crestview, FL358 returns | Map flows |
| 12 | Punta Gorda, FL9,405 in · 7,526 out | 1,879 | 9.3 | 55.8%43 counterparts | 48.1%31 counterparts | North Port, FL1,585 returns | North Port, FL1,433 returns | Map flows |
| 13 | Athens, TN2,188 in · 1,585 out | 603 | 8.8 | 34.4%3 counterparts | 36.7%3 counterparts | Cleveland, TN334 returns | Cleveland, TN308 returns | Map flows |
| 14 | Deltona, FL24,395 in · 18,167 out | 6,228 | 8.8 | 65.9%83 counterparts | 58.7%74 counterparts | Orlando, FL5,884 returns | Orlando, FL3,739 returns | Map flows |
| 15 | Clewiston, FL1,912 in · 1,435 out | 477 | 8.7 | 65.7%5 counterparts | 47.2%6 counterparts | Miami, FL633 returns | Cape Coral, FL268 returns | Map flows |
| 16 | Cambridge, MD970 in · 686 out | 284 | 8.7 | 35.3%3 counterparts | 22.9%2 counterparts | Baltimore, MD151 returns | Easton, MD114 returns | Map flows |
| 17 | Williston, ND1,543 in · 1,202 out | 341 | 8.6 | 11.7%6 counterparts | 13.0%4 counterparts | Minot, ND52 returns | Minot, ND59 returns | Map flows |
| 18 | Spartanburg, SC10,511 in · 7,321 out | 3,190 | 8.5 | 63.9%47 counterparts | 58.8%21 counterparts | Greenville, SC3,374 returns | Greenville, SC2,451 returns | Map flows |
| 19 | Albemarle, NC1,818 in · 1,271 out | 547 | 8.5 | 60.3%3 counterparts | 52.2%5 counterparts | Charlotte, NC1,048 returns | Charlotte, NC548 returns | Map flows |
| 20 | Crossville, TN1,987 in · 1,466 out | 521 | 8.2 | 24.7%6 counterparts | 26.5%4 counterparts | Cookeville, TN199 returns | Cookeville, TN198 returns | Map flows |
| 21 | Greeley, CO13,725 in · 10,885 out | 2,840 | 8.1 | 75.3%36 counterparts | 65.3%37 counterparts | Denver, CO4,223 returns | Denver, CO2,527 returns | Map flows |
| 22 | Daphne, AL7,537 in · 5,555 out | 1,982 | 8.0 | 47.1%38 counterparts | 44.2%25 counterparts | Mobile, AL1,002 returns | Mobile, AL842 returns | Map flows |
| 23 | Lake Havasu City, AZ7,036 in · 5,267 out | 1,769 | 8.0 | 60.8%26 counterparts | 47.9%21 counterparts | Riverside, CA1,127 returns | Phoenix, AZ565 returns | Map flows |
| 24 | Palm Bay, FL20,098 in · 15,118 out | 4,980 | 7.9 | 71.2%107 counterparts | 62.1%75 counterparts | Orlando, FL2,482 returns | Orlando, FL1,961 returns | Map flows |
The visible table is intentionally limited for reading. The CSV contains every row that passes the printed inclusion rule.
Inclusion first, interpretation second.
Include a current RentMarker metro only when it has positive ACS population, published IRS inbound and outbound tax-return household totals, at least one directly mapped inbound and outbound inter-metro pair, and identifiable leading counterparts in both directions.
Methodology version: v19Method
- All-US inbound and outbound counts are IRS county flows aggregated through exact current CBSA county membership after same-market moves are removed from both totals.
- Each direct county pair is read once from the IRS inflow-side record, maps both county ends independently to exact CBSA membership and excludes pairs whose ends resolve to the same metro.
- Net migration per 1,000 divides inbound minus outbound tax-return households by ACS population and multiplies by 1,000.
- Mapped coverage divides direct inter-metro households by the corresponding all-US total; leading-counterpart concentration divides the largest direct flow by all mapped flows in that direction.
Counter-signals
- A large net count can be modest relative to population, while a high per-resident rate can come from a small metro.
- The leading corridor may represent a small share of all movement, so the coverage and concentration measures should be read with the count.
- Mover adjusted gross income is not renter income, disposable income or proof of housing demand.
Limitations
- IRS migration omits non-filers and measures tax-return household moves, not every resident, renter, lease or reason for moving.
- Direct mapped coverage excludes county pairs whose other end falls outside a current RentMarker metro; missing movement is not assigned to a nearby market.
- IRS 2022-2023 flows, ACS population and current OMB metro membership cover different release periods, and the atlas is descriptive rather than a forecast or recommendation.
Every input release used by this study
Release and retrieval dates stay visible so a reader can tell whether two measures describe the same moment.