The New York City Department of Homeless Services publishes a Daily Report. It is one of the most operationally important public datasets in the city: a daily, system-by-system census of how many people are sheltered, how many are placed, how many are turning to the system that night.
The April 2026 single-adult shelter population, per the DHS Daily Report, stood at 16,680. That figure is the lowest monthly average in three years.
For comparison, the population peaked at 19,538 in July 2024. The trajectory from that peak has been consistent: October 2024 at 19,355; January 2025 at 19,029; June 2025 at 19,219; January 2026 at 17,567; April 2026 at 16,680.
The decline is approximately 15 percent over 21 months.
For a population that drove news cycles in 2023 and 2024 — the asylum-seeker arrivals, the right-to-shelter litigation, the cascade of emergency facility openings — the consistent downward direction over 10 months is the most significant single trend in the city's homelessness data right now.
An immediate caveat is necessary. The shelter population is not the same as the homeless population. The DHS Daily Report measures who is in the shelter system on any given night. It does not measure who is unhoused but outside the system: people sleeping on the street, in cars, in subways, doubled up with relatives, couch-surfing, or in irregular short-term arrangements. A 15 percent decline in the shelter population does not mean a 15 percent decline in homelessness. It means a 15 percent decline in shelter system occupancy, with the broader picture more difficult to measure directly.
This distinction is fundamental to interpreting the data, and is returned to throughout this analysis.
What the data actually says
The DHS Daily Report tracks several distinct populations. The single-adult shelter system, the largest, is described above. Other categories include adult families (couples without children), families with children, and people turned away or temporarily placed.
For families with children, the April 2026 number was approximately 35 families requesting temporary housing on that specific day, with zero families placed in overnight accommodations that evening — though the longer-term family shelter population is tracked separately and varies considerably.
Reading the trend over multiple years:
Pre-asylum-seeker baseline (FY 2022): Single-adult population averaged 8,400 to 9,000.
Asylum-seeker surge (FY 2023 to mid FY 2024): Population doubled, reaching the July 2024 peak of 19,538.
Stabilization (mid FY 2024 to early FY 2025): Population plateaued in the 19,000-19,500 range.
Sustained decline (mid FY 2025 to April 2026): Population dropped progressively to 16,680.
The 10-month decline through April 2026 represents the first such sustained drop since the surge began in 2022.
What is driving the change
Multiple factors operate simultaneously. Disentangling their individual contributions requires more granular data than the daily report alone provides. Several plausible contributors:
Reduced asylum-seeker arrivals. Federal and state policy shifts since late 2024 reduced the rate at which new arrivals enter the NYC shelter system. The slowing inflow is documented in successive Daily Reports.
Reuniting with sponsors. Many asylum seekers transitioned from shelter to housing with relatives or community sponsors. This represents successful exit from the shelter system without necessarily requiring a city housing voucher.
Housing voucher placements. CityFHEPS and other voucher programs continued to issue vouchers throughout the period, supporting transitions out of shelter for eligible residents. According to coverage by THE CITY in May 2026, CityFHEPS placements remained active under the administration's continued program funding.
Turn-away policies. The Adams administration's policies, continued in part under the Mamdani administration, requiring some asylum-seeker arrivals to re-apply for shelter on a periodic basis, contributed to the rate at which people exited the system, sometimes without permanent housing.
Improved exits. The administration has described faster processing of voucher applications, expanded relocation assistance, and other measures targeting shelter exit rates.
Reduced new entries. Some population reduction reflects fewer new entries, not just more exits. The Daily Report shows declining "single adults requesting temporary housing" volumes in 2026 compared to 2024.
The relative contribution of these factors is not directly observable in the Daily Report alone. Annual reports from DHS, the Department of Social Services, and the Mayor's Management Report typically address the decomposition with more granular numbers, but those reports are released months after the period they cover.
What the data does not show
The shelter system numbers are an operational measurement of who is inside the shelter system. They are not a measurement of who is homeless in NYC. The two populations overlap but are not identical.
The DHS Daily Report does not track:
People at the margin of homelessness. People couch-surfing, sleeping in cars, doubled up with relatives, in irregular short-term housing — none of these populations are captured in shelter data. They may be larger than the sheltered population.
Unsheltered street homeless. People sleeping on the street, in subways, or in other public spaces are not in the shelter system. The HUD Continuum of Care Point-in-Time count, conducted annually each January, attempts to enumerate this population separately, with significant methodological limitations.
Outcomes after shelter exit. When someone exits the shelter system, the Daily Report does not track whether they entered permanent housing, returned to home-country relatives, became unhoused on the street, or moved into another irregular arrangement.
Demographic detail. Daily aggregate counts do not include age, household composition beyond family/adult distinctions, primary language, or country of origin.
Cause of decline by individual. Whether someone exited shelter because they got a voucher, because they reunited with relatives, because they returned to a country of origin, or because they aged out of eligibility is not separately tracked in the Daily Report.
For a complete picture of homelessness in NYC, the shelter system numbers are only one input. A meaningful decline in shelter occupancy could be accompanied by a stable or rising population of people experiencing other forms of housing instability. The data presented here describes one specific subset of the larger landscape; it does not characterize the landscape as a whole.
What the policy implications are
The shelter population decline is occurring against a backdrop of ongoing policy debates:
The CityFHEPS expansion legal challenge continues at the New York State Court of Appeals, with the Mamdani administration's March 24 appeal still pending. The Council's 2023 expansion laws would extend voucher eligibility to people facing eviction before shelter entry, and would reduce barriers to access for people already in shelter.
The shelter population data does not directly resolve the CityFHEPS expansion debate. Two interpretations are possible:
Interpretation A: The shelter population is declining substantially under the existing CityFHEPS framework. Therefore the case for expanding the program is weakened — the system is delivering meaningful exits without the expansion.
Interpretation B: The decline is largely driven by reduced asylum-seeker arrivals, not by improved voucher access. The underlying demand for vouchers among eligible NYC residents remains. The expansion would help a different population (people facing eviction before shelter entry, and people in shelter with incomes above current eligibility thresholds), not the asylum-seeker arrivals who drove the recent surge.
Both interpretations have factual support. The administration's position aligns with Interpretation A; advocate positions emphasize Interpretation B.
And a structural caveat applies to both interpretations: the shelter data does not measure unhoused populations outside the system. If broader homelessness — including street homelessness, couch-surfing, and irregular housing arrangements — remained stable or worsened during the shelter decline, the operational improvement in shelter occupancy does not translate to an improvement in housing security across NYC.
What this means for current shelter residents and applicants
For shelter residents currently in the system or applying:
The shelter system continues to operate. Daily intake at PATH (Path Application Center) and the various shelter system entry points remains active. The system is not closing or shrinking faster than population decline would suggest.
CityFHEPS vouchers continue to be issued at the current eligibility thresholds. People in shelter who meet existing criteria can apply.
Anti-eviction legal services through Legal Aid Society, Legal Services NYC, and other providers remain available for tenants facing eviction outside the shelter system.
State Housing Access Voucher Program funding ($50 million in FY26 state budget) is available, though the state program is smaller than what the CityFHEPS expansion would provide and currently locks out tenants whose federal Emergency Housing Vouchers are expiring.
What is verifiable now
The DHS Daily Report is published every business day at nyc.gov/site/dhs/dashboard/dailyreport.page. Public data tracks the prior-day shelter population by category. Historical data going back years is available in the PDF release format.
The DHS Annual Report provides additional decomposition — by entry pathway, by length of stay, by exit destination — but is published with significant lag.
The Mayor's Management Report (MMR), released annually in September, provides DHS performance metrics including shelter population data, average stay duration, and exit outcomes.
The HUD Continuum of Care Point-in-Time count, conducted in late January annually, attempts to enumerate unsheltered individuals.
For New Yorkers tracking homelessness data, these four sources provide cross-referenced views of a population that is, by its nature, difficult to count completely.
The longer view
A 15 percent decline in single-adult shelter population over 21 months is meaningful but not a solution. The April 2026 baseline of 16,680 is still nearly double the FY 2022 pre-surge baseline. NYC remains the city with the largest sheltered homeless population in the United States by an order of magnitude.
The trajectory matters for what it suggests about whether the surge of 2022-2024 was the peak of a structural change, or a temporary spike that the system is gradually working through. The data through April 2026 favors the second interpretation as it pertains to the sheltered population — but the broader question of homelessness in NYC, including the populations the shelter data does not capture, remains a separate and less directly measurable question.
The data is public. The numbers are checked daily. The picture they describe is one part of a larger one that other datasets, with their own limitations, attempt to address.
Sources: NYC Department of Homeless Services Daily Report at nyc.gov/site/dhs/dashboard/dailyreport.page, public dataset covering May 2026 through historical data; DHS Daily Report PDF release of May 13, 2026 (containing data through May 12, 2026); historical DHS Daily Report data from FY 2022 through April 2026 as included in the monthly summary tables of the May 13 release; NYC Mayor's Management Report annual publications; HUD Continuum of Care Point-in-Time count methodology; THE CITY coverage of CityFHEPS placements (May 2026); NYC Council CityFHEPS expansion legislation (2023); state Housing Access Voucher Program FY26 budget.
