Growing Inaccuracies in Official Counts Jeopardize LA Homelessness Wins

Louis Abramson, Jason M. Ward, Sarah B. Hunter, Rick Garvey

ResearchPublished Oct 15, 2025

An unhoused person sits on a sidewalk at night, in Los Angeles, California, July 24, 2025

Photo by Qian Weizhong/VCG via Reuters

Official homeless counts increasingly underestimate the number of people living on the street in three key LA neighborhoods, according to the latest Los Angeles Longitudinal Enumeration and Demographic Survey (LA LEADS) data.[1] These shortfalls coincide with a rise in people living without the protection of a vehicle or tent (known as rough sleepers), who are also among the most challenging to count. This means that the places with the highest needs are becoming the very places where the county’s official count is most underestimated.

Without a correction, the Los Angeles Homeless Services Authority’s (LAHSA’s) undercount could divert resources from the communities that need them most, eroding recent gains on a top policy priority.[2]

Every year, LAHSA conducts the Point-in-Time (PIT) Count—LA County’s homelessness census.[3] Its core component is a volunteer-led visual survey of over 2,800 census tracts to determine the number of people living unsheltered on the street. This annual snapshot informs both public perception and policymaking around this crisis, and the reversal of its long upward march over the past two years may indicate that new strategies are working.[4]

The PIT is also used to determine the flow of resources. It enabled the distribution of $220 million in federal funds in 2024 and directed another roughly $100 million in regional funds in 2025 via Measure A—a county sales tax.[5]

With so much riding on the PIT, its accuracy is more critical than ever. Unfortunately, new LA LEADS data raise three concerns about the PIT as a tool for resource allocation.[6]

First, the PIT is increasingly an undercount. Figure 1 shows the total people and dwellings in the 2022–2025 PIT (diamonds) and LA LEADS (circles) data in Hollywood, Venice, and Skid Row—epicenters of homelessness in LA. The counts covered the same census tracts using the same methodologies, but LA LEADS used professional counters and internal consistency checks. Although the PIT was within 5 percent of LA LEADS in 2022–2023, this strong agreement eroded sharply in 2024, when a 26 percent PIT shortfall emerged. That gap widened to 32 percent in 2025.

Figure 1. LA LEADS Study Area

This line chart compares two measures of unsheltered persons or dwellings over time from January 2022 to January 2026.

PIT Total Unsheltered

The PIT data shows four point-in-time measurements:

  • January 2022: 2,459 persons or dwellings
  • January 2023: 2,633 persons or dwellings
  • January 2024: 1,911 persons or dwellings (26% undercount)
  • January 2025: 1,694 persons or dwellings (32% undercount)

SOURCES: Features data from Abramson et al., Annual Trends Among the Unsheltered in Three Los Angeles Neighborhoods; LAHSA, “2022 Greater Los Angeles Homeless Count Data”; LAHSA, “2023 Greater Los Angeles Homeless Count Data”; LAHSA, “2025 Greater Los Angeles Homeless Count Data”; LAHSA, “HC2023 CensusSubTracts Updated SchemaAdj 20230123.”

An extrapolation of the 2025 undercount in three neighborhoods to the city as a whole shows that up to 7,900 persons and dwellings may be missing from LA’s most recent official homelessness count. This deficit—which is more than Orange County’s entire PIT Count—could affect the region’s ability to sustain progress in reducing homelessness.[7]

Second, the size of the PIT’s undercount depends on location. Table 1 shows the 2025 PIT official counts as a fraction of LA LEADS enumerations. Both counts were collected on or near the same date and tracked the same modes of unsheltered living in each neighborhood. While Hollywood’s PIT total was 81 percent of the LA LEADS estimate, Skid Row’s was just 61 percent. Venice fell in between.

Table 1. 2025 PIT Enumeration Totals as Fractions of LA LEADS Totals

Hollywood Venice Skid Row All
Total 0.81 (NL=556) 0.76 (NL=449) 0.61 (NL=1,480) 0.68 (NL=2,485)
Rough sleepers 0.82 (360) 0.89 (178) 0.58 (730) 0.69 (1,268)
Tents + makeshifts 1.76 (45) 0.91 (35) 0.69 (599) 0.76 (679)
Cars + vans + RVs 0.49 (151) 0.65 (236) 0.45 (151) 0.56 (538)

SOURCES: Features data from Abramson et al., Annual Trends Among the Unsheltered in Three Los Angeles Neighborhoods; LAHSA, “2025 Greater Los Angeles Homeless Count Data”; LAHSA, “HC2023 CensusSubTracts Updated SchemaAdj 20230123.”

NOTE: Data reflect raw enumerations and are not adjusted for assumptions about the average occupancies of tents and vehicles. LA LEADS data were taken within two weeks of the PIT. Median one-standard-deviation errors are ±0.05; NL denotes the LA LEADS value (denominator) for each entry.

If this neighborhood-level unevenness persists across different municipalities—which seems likely—some cities could receive 30 percent more Measure A Local Solutions Fund dollars per unsheltered person or dwelling than others, meaningfully skewing funding away from true needs.

Third, the PIT undercount has worsened as rough sleeping has grown. Figure 2 shows the PIT undercount (y-axis) compared with the fraction of rough sleepers in each year’s LA LEADS data (x-axis). The thick black horizontal line is where the points would lie if the PIT totals equaled the LA LEADS totals. Instead, the PIT’s accuracy has gotten worse even as rough sleeping has become more prevalent (diagonal trend line).[8]

Figure 2. PIT Undercount Versus Rough Sleeper Presence

This scatter plot shows the relationship between rough sleeper fraction (x-axis, ranging from 0.40 to 0.55) and PIT fraction of LA LEADS total count (y-axis, ranging from 0.6 to 1.1).

Data Points by Year

  • 2023: Rough sleeper fraction of approximately 0.42, PIT fraction of LA LEADS total count of approximately 1.05
  • 2022: Rough sleeper fraction of approximately 0.42, PIT fraction of LA LEADS total count of approximately 0.97
  • 2024: Rough sleeper fraction of approximately 0.47, PIT fraction of LA LEADS total count of approximately 0.745
  • 2025: Rough sleeper fraction of approximately 0.51, PIT fraction of LA LEADS total count of approximately 0.685

SOURCES: Features data from Abramson et al., Annual Trends Among the Unsheltered in Three Los Angeles Neighborhoods; LAHSA, “2022 Greater Los Angeles Homeless Count Data”; LAHSA, “2023 Greater Los Angeles Homeless Count Data”; LAHSA, “2025 Greater Los Angeles Homeless Count Data”; LAHSA, “HC2023 CensusSubTracts Updated SchemaAdj 20230123.”

NOTE: The PIT increasingly underestimates the unsheltered population of Hollywood, Venice, and Skid Row as the proportion of rough sleepers in those areas grows.

Rough sleepers have the highest social and clinical needs and are therefore arguably the most important group to count accurately. If under-identification trends continue, PIT inaccuracies may actively push resources away from high-need areas.[9]

The above issues seriously challenge the understanding of homelessness and the ability to support and evaluate strategies to combat it. Ironically, these issues may also partially stem from the success of such initiatives as LA City’s Inside Safe.[10] By offering motel rooms to people living in tent encampments, such policies contributed to halving the number of tents identified by LA LEADS since late 2021. However, as the initiatives drove a real decline in unsheltered homelessness, they also removed the easiest-to-count unsheltered subpopulation, likely contributing to the PIT’s growing inaccuracy.

Whatever the undercount’s cause, the LA LEADS results suggest a simple corrective: PIT organizers should rely more heavily on professional field teams to independently cross-check volunteer counts during the PIT. This solution is relatively inexpensive, requires little managerial overhead, may be fundable through Measure A, and would yield a more transparent, accurate PIT.[11] The return on this investment in data quality and policy confidence would be substantial, to say nothing of the material gains from a better census. In a policy area in which easy wins are rare, this is one of them.

LA has made meaningful progress in reducing homelessness. As the PIT becomes a larger factor in sustaining that progress, the PIT’s inaccuracies can no longer be ignored. There is a straightforward fix that policymakers could adopt.

Notes

  1. Abramson et al., Annual Trends Among the Unsheltered in Three Los Angeles Neighborhoods.Return to content ⤴
  2. LAHSA, homepage. Return to content ⤴
  3. LAHSA, “2025 Greater Los Angeles Homeless Count Data.” Return to content ⤴
  4. LAHSA, “Declining Homelessness Is Now a Trend in Los Angeles County.” Return to content ⤴
  5. LAHSA, “HUD Awards over $220 Million to Address Homelessness in Los Angeles”; County of Los Angeles Homeless Initiative, “Measure A.” Return to content ⤴
  6. Abramson et al., Annual Trends Among the Unsheltered in Three Los Angeles Neighborhoods. Return to content ⤴
  7. USAFacts, “Which US Cities Have the Largest Homeless Populations?” Return to content ⤴
  8. Abramson, “A Concerning Rise in Rough Sleeping Threatens Recent Progress on Unsheltered Homelessness in Los Angeles.” Return to content ⤴
  9. Technically, vehicles show the largest PIT undercount compared with LA LEADS (Table 1). However, since rough sleepers outnumber vehicles 3:1 in the LA LEADS study area and are the highest-need subgroup, this fact matters less when it comes to ensuring that resources track needs, geographically. Return to content ⤴
  10. Abramson et al., Annual Trends Among the Unsheltered in Three Los Angeles Neighborhoods; City of Los Angeles, “Inside Safe.” Return to content ⤴
  11. County of Los Angeles Homeless Initiative, “FY 2025–26 Draft Funding Recommendations.” Return to content ⤴

References

Topics

Document Details

Citation

Chicago Manual of Style

Abramson, Louis, Jason M. Ward, Sarah B. Hunter, and Rick Garvey, Growing Inaccuracies in Official Counts Jeopardize LA Homelessness Wins. Santa Monica, CA: RAND Corporation, 2025. https://www.rand.org/pubs/research_reports/RRA4438-1.html.
BibTeX RIS

Research conducted by

This research was supported by the Lowy Family, whose generous gift supported the establishment of the RAND Center on Housing and Homelessness.

This publication is part of the RAND research report series. Research reports present research findings and objective analysis that address the challenges facing the public and private sectors. All RAND research reports undergo rigorous peer review to ensure high standards for research quality and objectivity.

This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit www.rand.org/pubs/permissions.

RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.