Comments
iAUDIT! –
“If you want a friend you can trust, get a dog”.
-Harry S Truman
Sacramento just released the results of its most recent Point in Time (PIT) count, claiming a 29 percent decrease in homelessness and a 41 percent decrease in unsheltered homelessness. A few weeks ago, San Francisco reported the results of its count showed overall homelessness increased by seven percent but unsheltered homelessness decreased by 16 percent. LAHSA has yet to release the results of the Point in Time (PIT) count completed in January 2024, but there are indications the numbers may be released by the end of June. As previously reported, L.A.’s 2023 count showed sharp increases in overall homelessness, as well as chronic and unsheltered homelessness.
To understand what these numbers mean, and what the differences may tell us, we have to start with understanding the nature and purpose of PIT counts. The US Department of Housing and Urban Development, (HUD), which controls the flow of federal dollars for homelessness and housing programs, mandates a count of the homeless population be conducted biannually (every two years). Some local governments, like Los Angeles, choose to perform annual counts. The counts are managed by designated Continuum of Care (CoC) agencies charged with the administration and coordination of regional homelessness programs. For most of LA county, the CoC agency is LAHSA. LAHSA funds the PIT count and publishes the results.
As is true of most federal regulations, HUD’s rules for the PIT count are long, full of jargon, and difficult to comprehend. However, a key sentence in the rules state “CoCs must ensure that their count estimate accurately reflects what they believe to be the entire sheltered and unsheltered population for the CoC’s entire geographic area”. The PIT count’s results are meant to measure the progress CoC’s make in reducing homelessness, to ensure local governments are using HUD money according to funding rules, and to assist in planning program improvements. To keep the counts in sync with the federal budget process, HUD requires PIT counts to be done in January so results can be used in preparing the fiscal year budget. HUD recognizes PIT counts are estimates, because it is nearly impossible to count every homeless person in a few days, and the unhoused population by its nature is dynamic, with people constantly moving in and out of homelessness. Nevertheless, HUD expects CofC’s to use generally accepted professional statistical methods to perform the count to ensure it as accurate as possible.
In Los Angeles, the PIT count is used as a planning tool to allocate resources by region. LAHSA divides the County into eight Service Planning Areas (SPA’s). Most service contracts are written to address the needs of a specific SPA’s unhoused population. Outreach teams are allocated based on the number of unsheltered people in each SPA. LAHSA tracks increases and decreases in homeless populations by SPA as a way of measuring program performance. For example, after the 2022 PIT count showed a significant decrease in the number of homeless people in the SPA that includes West L.A., both LAHSA’s leaders and Councilmember Bonin boasted about the success of their No Barrier Housing First policies.
As we now know, LAHSA’s 2022 PIT count was deeply flawed. In some areas, hundreds of unhoused people in encampments were missed. A survey by the RAND corporation found significant differences in its sample areas compared to LAHSA’s count. The survey also found rates of chronic homelessness were much higher than LAHSA reported. The embarrassment caused by the false claims of reductions on the Westside forced LAHSA to stop publishing count data by Council district. The 2022 count was not an isolated incident of procedural errors. Previous counts had been plagued by questionable results. In 2022, LAHSA partnered with USC to bring some statistical legitimacy to the count, but the count was still deeply flawed—even after a significant delay as managers tried to clean the data.
The 2023 count fared little better. Even as it was being conducted, volunteer counters were complaining that LAHSA’s new counting app was buggy and didn’t always upload to the central counting system. Again, LAHSA said it would overhaul its procedures for 2024, but almost immediately, inconsistencies in counting polices and an even worse experience with and “new and improved” counting app. The count was so flawed, the County Board of Supervisors ordered an audit of the selection process for LAHSA’s new counting software. It is now June, and we still have no results from the January count.
Inaccurate and fluctuating numbers are nothing new for homelessness programs. As detailed in this article, the City’s, County’s and LAHSA’s shelter and housing numbers include repeat and lost clients. Numbers shared by jurisdictions rarely match, a sign of poor coordination and communication among government agencies. Local agencies have so much trouble producing meaningful numbers that the federal judge overseeing the independent audit of city programs held a hearing on just how poor data practices are.
What is the effect of all these questionable numbers? Why should we care if we don’t know how many people ae unhoused or if they receive no services? We should care because homeless numbers and money are inextricably linked. The City has budgeted at least $1.3 billion for homeless interventions. The County contributes $3 billion more. LAHSA’s sole reason for existence, and its $800 million budget, are meant to alleviate homelessness. Because PIT counts are so flawed, we really don’t know how many homeless people live on our streets. You can’t solve a problem if you don’t know its dimensions. When the City claims it has sheltered or housed 21,000 people, we don’t know how many of them have been recycled through the system at least twice. We don’t know the effect of the hundreds of millions spent on mental health and substance abuse services because the County can’t keep its service numbers straight.
Sloppy data management usually indicates equally troublesome practices in other areas. One of the reasons LAHSA pays providers by contact hour or by the number of people “engaged”, and why the City pays for safe parking by the space instead of more meaningful measures is because they don’t have service numbers they can rely on. They choose the easiest, (and often costliest) way to pay providers because they simply do not have the capability to measure performance by any other method. Of course, when you place no value on meaningful data, you make it almost impossible to hold program managers accountable for results. It makes it easy for the City to spend $800 million on acquiring 2,750 hotel units for homelessness housing, and leaving 1,200—47 percent, representing a value of more than $349 million—vacant for months or years. It invites fraud, like that alleged against Shangri-La developers and Step Up Los Angeles, one of the area’s largest service providers.
Poor data management and its consequences erode the public’s trust in local government’s ability to address the homelessness crisis. In San Francisco, the Mayor’s office tried to emphasize the decrease in unsheltered homeless while ignoring a 37 percent jump in people living in vehicles. In Sacramento, the supposed decreases in homelessness have been met with cynicism by providers in the field, who say they’ve seen no reduction in people asking for help. In L.A., officials claim there is nowhere to put our homeless when 1,200 units sit vacant while City departments point fingers at one another.
Like words, numbers have meaning and power. Every time LAHSA or the City claims they have sheltered someone, there should be a real, unique person who has benefitted from their programs. When they provide services to the same people over and over, they exclude others who need help. When they pay a provider for nonexistent services, they rob others of the funding that could lift someone out of homelessness. When leaders refuse to recognize the reality of their failures by skewing or hiding numbers, they deny people on the street a chance to put homelessness behind them. Numbers are far more than just figures on a webpage. They represent the consequences of decades of failing our most vulnerable residents.
(Tim Campbell is a resident of Westchester who spent a career in the public service and managed a municipal performance audit program. He focuses on outcomes instead of process.)
