In the last few weeks we have been debating the statistics of homeless people, but as today’s column from Marcos Breton demonstrates, maybe we can’t believe any of it, because the data collection methods are flawed and inadequate.
The debate started over local supportive housing at places like Paul’s Place. Many believe you need to get people off the streets and into housing before you can begin to treat other problems. But still others oppose such housing, believing that the homeless need mental health treatment before they can move into stable housing.
However, Paul Thornton of the LA Times wrote a few weeks ago that “most individuals who experience homelessness don’t end up that way because of mental health issues or addiction.”
But not everyone or even every survey agrees with the data that the LA Times Editorial Board came up with, that only about one-third of the homeless suffer from serious mental illness.
Some cite a HUD (Housing and Urban Development) study that found, “An estimated 26% of homeless adults staying in shelters live with serious mental illness and an estimated 46% live with severe mental illness and/or substance use disorders.”
But the HUD study, based on data from January 2010, not only may be dated but like many is known as a Point-in-Time Count and it found that on a given night about 26 percent of sheltered homeless adults had a severe mental illness and 34.7 percent suffered substance abuse.
A big point that I would argue – it seems unlikely that the rate of mental illness in the overall population has changed drastically since 2010. And yet we know that the number of homeless has vastly increased – that is the number we need to understand better to see what’s driving the homeless problem.
As we look at the Marcos Breton column, we can start to see the problem with measuring the homeless problem.
Marco Breton points out, “You can’t believe the numbers fully. You certainly can’t take them at face value.”
He writes: “When you do take the latest federally-mandated homeless count numbers at face value, they seem shocking. One wonders what the devil Sacramento is doing about homelessness when Sacramento is committing millions to the problem.”
The current count finds that homelessness has jumped by 52 percent in Sacramento County over the past two years. I think we have reason to believe that the number has gone up – but by that much?
Mr. Breton demonstrates why we should be skeptical not only about the current numbers but about data in general – especially, point-in-time counts.
A big factor here is how you count the homeless. And the big change is that they had 600 more volunteers counting the homeless this year as opposed to two years ago.
In 2017 it was 300 people, now it’s 900 people.
As he points out, “there is a pretty good chance all those extra counters are gong to find extra homeless people to count.” Indeed, “This year, unlike past years, counters also canvassed more parts of the county. And they relied more on statistical estimates to gather numbers from areas in the county they did not canvass.”
He points out that this was based on a “snapshot” of one 24-hour period in January. Mr. Breton writes that “this count is seen as the closest to accurate that Sacramento has ever done, which means that past counts have been far less reliable.”
Given the more accurate count this time, the 52 percent jump is probably an artifact, at least in part, of the better survey method.
But there are more than just flaws in the numbers, as Mr. Breton pointed out, as “what’s even more flawed is the amount of critical information that we don’t know and is not disseminated in the current count.”
A bigger problem is getting a statistical profile of the homeless.
Mr. Breton argues that “if you ask experts on homelessness for a detailed statistical profile of homeless people, the answer is that we just don’t have enough critical data.”
The report in Sacramento, for instance, finds “only nine percent of homeless surveyed said drugs or alcohol prevent them from keeping a job and stable housing.”
He asks, “Does that sound right? It doesn’t to me.”
I tend to agree with that.
Further: “About 21 percent reported having a severe psychiatric condition, which also seems like an under count.”
He adds, “About 45 percent of homeless respondents said Sacramento needed more affordable housing. OK, but that’s a point that most everyone probably believes.”
Mr. Breton further writes: “Is there hard data on how many Sacramento homeless are homeless because they couldn’t afford their rent? Not really. This is not to disparage the 2019 homeless count or the people who worked on it. But this count is too often confused as definitive when it’s not.”
He concludes: “Without the data, we’re flying partially blind when it comes to confronting the crisis of getting people under roofs and off the streets. How can you cure a disease if you can’t adequately diagnose it?”
On this point, I completely agree. We’re throwing around survey data, whether it comes from HUD or the LA Times, that might not be telling us as much as we hoped.
One piece of data does seem to be more reliable. Mr. Breton writes that “93 percent of the homeless people surveyed in January said they were from this county. This debunks a myth sometimes floated by some, including me, that Sacramento County is a magnet for homeless people.”
He adds, “Given that the current count of homeless is the most robust ever done, the 93 percent of homegrown homeless should speak to us all that this is our problem to solve. These are county residents. We can’t just leave them on the street or put them on buses to Roseville.”
I come back to this point: I believe the number of homeless people has increased in recent years. I don’t believe that a rise in mental illness or addiction explains that increase. And thus I believe affordable housing is a factor.
Beyond that, I don’t think we really know as much as we should about the homeless problem or how to fix it.
—David M. Greenwald reporting