Asylum-seekers, migrants, and the homeless shelter system

Between 2022 and 2025, several large North American cities — New York, Chicago, Denver, Boston, Toronto, Montreal — saw substantial arrivals of asylum-seekers, refugees, and other recent migrants who entered the formal shelter system. The political coverage of this generated heat that often obscured the actual dynamics. This article tries to lay out what happened, what it means for the homeless services ecosystem, and what works.

What actually happened

Between 2022 and early 2025, several US states (Texas, Florida) bused asylum-seekers north to Democratic-led cities. Independent migrant journeys, post-pandemic backlogs in the immigration system, and changes in border-processing policy added arrivals through normal pathways. By mid-2024:

• NYC's shelter system was housing roughly 65,000 people in DHS shelters, with another ~65,000 asylum-seekers in a parallel emergency-relief system. The combined caseload was historically unprecedented.
• Chicago opened emergency shelters in police stations and city facilities. At peak, 15,000+ asylum-seekers were in shelter.
• Denver, with a much smaller population, housed ~5,000 asylum-seekers at peak.
• Boston's family shelter system reached its statutory cap, prompting waitlists and motel-based overflow.
• Toronto and Montreal both saw months-long waitlists for shelter and motel-room placements.

By 2025, arrival rates moderated and several cities were closing emergency arrival sites — but the underlying housing pressure remained.

Why asylum-seekers end up in homeless shelters

The path goes roughly like this:

1. A person arrives at the US or Canadian border, often after a journey through Mexico or via a humanitarian-parole program.
2. They are processed and released with a court date months or years away.
3. Federal services for asylum-seekers are limited. In the US, asylum-seekers cannot work legally for at least 180 days after their asylum application is filed. In Canada, applicants can apply for work authorization more quickly but still wait weeks to months.
4. Cities receive them via NGO or government coordination — sometimes formally, sometimes not.
5. With no work authorization and no money, many enter the city shelter system because it's the only existing infrastructure for housing people with zero income.

This is why cities have absorbed the surge, not because they wanted to, but because the alternative (asylum-seekers sleeping outdoors at the train station) was politically and practically worse.

What this strained

The pre-existing US/Canadian homeless services ecosystem was designed around a different population:

• US homeless adults are disproportionately men, single, often with substance-use or mental-health concerns.
• Asylum-seekers are disproportionately families with children, in good physical health, motivated to work, with strong social cohesion.

The shelter infrastructure (large men's congregate shelters, single-adult focus, mental-health-centered case management) is poorly fit to the latter population. Cities have built parallel arrival systems precisely because the existing shelter system didn't work for migrant families.

This parallel-system approach is what created the two separate counts you'll see in some 2024 reporting: NYC reported "shelter population" separately from "asylum-seekers in city care." Other cities used various conventions.

The strain on the pre-existing homeless services has come from competition for limited resources:

• Hotel-room availability — cities renting hotel rooms for asylum-seekers pushed up the rates for non-migrant emergency placements.
• Vouchers and rental subsidies — increased demand at the bottom of the rental market.
• Political bandwidth — cities focused on the migrant surge had less attention for chronic-homelessness programs.

What actually helps

Programs that have worked well across the arrival cities:

• Migrant-specific shelter and casework, not blending arrivals into the existing single-adult homeless system. Families are housed in families; case management is immigration-focused (asylum filings, work-authorization applications, school enrollment).
• Volunteer English-language and work-readiness programs. Asylum-seekers with even basic English and a job lead almost always exit shelter quickly. Without those skills, they stay.
• Catholic Charities, Annunciation House (El Paso), Casa Juan Diego (Houston), and the Hebrew Immigrant Aid Society (HIAS). These are the organizations with decades of experience in this work. They're better at it than cities are.
• Faith-community hosting. Several cities (Chicago, Boston) experimented with congregations hosting families in church basements with city-coordinated logistics. Costs less than commercial hotels and produces better outcomes.
• Rapid work authorization. The 2023 federal Temporary Protected Status (TPS) expansion for Venezuelans materially shortened shelter stays — people who can work, exit. The 180-day waiting period for non-TPS asylum-seekers remains the largest single accelerator-or-not.
• Public school enrollment as a stability anchor. McKinney-Vento applies to homeless children regardless of immigration status. Enrolling kids stabilizes families faster than almost any other single thing.

What hasn't worked

• Treating migrants as a homelessness problem rather than an immigration system overflow. The cities that performed best (NYC's HERRC sites, Chicago's "Landing Zones") explicitly separated the response from the existing homeless services. Cities that tried to absorb arrivals into general shelter saturated and de-stabilized both populations.
• Police-station and underutilized-building sheltering at scale. Chicago's police-station and city-building strategy in 2023-24 was widely seen as a crisis-management failure. Hotel and congregate-but-family-friendly options outperformed.
• Open-ended hotel placements without work-authorization timelines. A family in a hotel for 18 months while waiting for asylum-court hearings is far costlier than a family in transitional housing with work authorization.

A note on the framing

A lot of news coverage frames migrant homelessness as either "asylum-seekers are causing the homelessness crisis" or "asylum-seekers are blameless and the homeless services system should absorb them." Both framings miss the structural picture.

The most accurate framing:

• Cities have absorbed asylum-seeker arrivals into homeless services because federal immigration policy created an arrival without federal housing support. The cities were the receivers of last resort.
• The pre-existing housing-cost crisis means there was very little slack in city shelter capacity to begin with.
• Asylum-seekers in this position are not "the homeless population" in the usual sense — they are a separate population with separate needs that overlapped with shelter infrastructure.
• Solving the issue requires federal-level action (faster work authorization, federal housing assistance for asylum-seekers, expanded shelter capacity funded at the federal level), not just city or state policy.

For student research

If you're writing on this topic:

• Cite the specific city documentation (NYC DHS, Chicago Office of Emergency Management, Denver's reports) for caseload figures. The numbers move month to month.
• Distinguish "asylum-seekers" from "refugees" from "undocumented immigrants" from "TPS holders" — the legal categories matter and have different rights.
• Be aware that this is a politically polarized topic where careful sourcing matters more than usual. AP, Reuters, ProPublica, and city-specific outlets (THE CITY in NYC, Block Club Chicago) tend to be the most rigorous.
• The Migration Policy Institute and the Niskanen Center have published substantive policy analysis.

Related: Resources for families · Homelessness by population · What to do if you become homeless.