The Second Journey: America's Refugee Map Does Not End at the Airport
Most refugee route maps end at baggage claim. A line arcs from a camp in Kenya or a border town in Jordan to a mid-sized American city, a pin drops, and the story closes with a family unlocking an apartment door. It is a useful image and a misleading one. In the United States, the initial placement city is a starting coordinate, not a destination.
The phenomenon has a name inside the resettlement system: secondary migration. It is the movement of refugees away from the community where a resettlement agency first placed them, toward another city, suburb, or rural county, sometimes within weeks and sometimes after several years. It is well known to practitioners, thinly documented in public data, and almost entirely absent from the maps that shape how journalists, philanthropies, and local officials understand the geography of displacement. Mapping it properly is one of the more consequential data tasks available to researchers in 2026.
The line that stops at the airport
Initial placement in the United States is administered through the Reception and Placement program, which is funded federally and carried out by a small number of national resettlement agencies working through hundreds of local affiliates. Caseworkers weigh housing availability, local cost of living, the presence of an ethnic community, and the capacity of the affiliate itself when deciding where a newly arrived family will begin.
That decision is made under constraint, often quickly, and with limited information about what a family will want a year later. When relatives in another state offer a spare room, when a warehouse job opens in a meatpacking town, when rent in the original city climbs past what a first job can cover, families move. The resettlement system does not stop them, but it also does not track them with any precision.
What the federal data can and cannot show
The Office of Refugee Resettlement publishes state-level tables on secondary migration in its annual reporting to Congress. These are genuinely valuable and genuinely limited: they lag by a year or more, they aggregate to the state rather than the metro or neighborhood, and they depend on voluntary reporting through state refugee coordinators whose own capacity varies widely.
The result is a dataset that can tell you that Minnesota, Ohio, Texas, and Washington tend to gain population while several first-placement states tend to lose it. It cannot tell you which Census tract a family landed in, how long the move took, whether it was a single jump or a series of short hops, or whether the household later moved again. Public data does not support a reliable national count of movers in any given year, and any figure presented as definitive should be treated with suspicion.
What the aggregate tables do establish is directionality. Secondary migration is not a rounding error in the resettlement story. It is a structural feature of how displaced families rebuild lives in a country where labor markets and housing costs vary enormously across a three-hour drive.
Where the second map leads
Several destination clusters recur across agency reporting, community organizations, and state-level data: the Twin Cities, Columbus and Cleveland, Des Moines, Fargo, Utica and Buffalo, Lewiston in Maine, Clarkston and the Atlanta suburbs, Amarillo and Houston, the Seattle metro and Kent, San Diego, and Sacramento. These are not random. They combine affordable rental stock, entry-level employment, established religious institutions and grocery economies, and kinship networks thick enough to absorb a newly arriving household.
Policy climate now belongs on that list of pull factors. Families making a second move increasingly weigh state-level signals about benefits eligibility, school language programs, and immigration enforcement posture, alongside the older calculations about rent and work. When those signals shift, the map shifts with them, often faster than the agencies that fund services can respond.
Why an unmapped move becomes a local budget problem
Nearly every funding stream tied to refugee services is allocated on the basis of where someone was originally placed, not where they currently live. School district English learner funding, county health department outreach, workforce board contracts, and housing navigation grants all inherit that lag. A district that receives forty new students in a single semester because a cousin in another state sent word about an apartment complex has no line item that anticipates them.
The practical consequences are unglamorous and cumulative. English classes fill before enrollment closes. A clinic loses its interpreter. A food pantry absorbs a month of demand it did not budget for. None of this appears in a placement roster, and none of it is visible on the maps most commonly circulated.
Housing is the tightest constraint
Secondary migration frequently delivers families into metropolitan areas already contending with low rental vacancy and rising rents. The coping strategies that follow — doubling up with relatives, rotating sublets, accepting overcrowded units — are difficult to measure and easy to misread as stability. Households that appear housed in administrative records may be one lease termination away from shelter. Mapping that reality requires granularity that most public datasets cannot supply, and it requires a care with personal information that any competent mapper should take seriously.
A modest research agenda
The gap is closable without surveillance. Four steps would materially improve the picture:
- Publish existing federal secondary migration tables in machine-readable form, with explicit methodological caveats attached rather than buried.
- Build standing relationships with state refugee coordinators and ethnic community-based organizations, which often learn about arrivals before any agency does.
- Use de-identified, district-level school enrollment data on English learner populations as a proxy for household movement, aggregated to protect students.
- Treat uncertainty as content. A map that shows a range is more honest, and more useful to a county planner, than a pin that implies false precision.
What the second map argues
Every map makes an argument. One that ends at the airport argues that resettlement is an event: a plane, a case file, a key. One that follows the second journey argues something closer to the truth — that resettlement is a process that continues for years, that its geography is decided partly by families themselves, and that the places receiving them deserve notice, resources, and advance warning rather than surprise.
Refugee Maps exists to draw routes that are usually left off the page. In the American context, the most important of those routes is not international at all. It runs between two exits on an interstate, and almost nobody is charting it.
