Archive for September 1st, 2011
Today’s post was written by Alliance research associate Pete Witte.
Last week, the Department of Housing and Urban Development (HUD) released the Homelessness Pulse Report: First Quarter, 2011, which attempts to provide a timely sense on how sheltered homelessness is changing in a number of communities. The major findings show that sheltered homelessness increased in three-quarters of the CoCs included in the report, but that the newly sheltered homeless population decreased in greater than half of the CoCs.
But, as is so often the case with data reports, in HUD’s latest homelessness report the devil’s in the details.
First, a little background.
HUD views the purpose of the Pulse Report as twofold: to disseminate data more frequently than the AHAR and to monitor progress against the Federal Strategic Plan to Prevent and End Homelessness.
To this aim, the Pulse Report presents quarterly trends of homelessness data with a view on two data points: (1) estimates of sheltered point-in-time counts (i.e. “Are we ending homelessness?”) and (2) estimates of newly sheltered homeless (i.e. “Are we preventing homelessness?”). These data points are analyzed at quarterly intervals ending with January, 2011 and are also broken into two subpopulations, members of families and individuals (though there are also minimal data on newly sheltered unaccompanied youth). The report provides data from a total of 23 CoCs (representing 361 counties and 422 cities). It’s also worth noting that the participating CoCs are selected because they have a history of quality data.
Now, the details.
When looking at trends for changes over time among any data, two periods within a timeframe are used for gaining a better sense for what’s going on within the data: the prior 3-month period (usually, the prior annual quarter) and the same 3-month period one year ago. With that in mind, the current report does not shed much light on trends in homelessness.
Again, the major finding in the point-in-time counts of sheltered homelessness is that the population increased in 75 percent of CoCs (between October, 2010 to January, 2011). However, this is probably mostly explained by the fact that shelter populations usually trend upward during this period due to cold weather. The other trend period analyzed is sheltered PIT counts at the 10-month interval, which shows that 75 percent of CoCs experienced an increase in sheltered homelessness (shelter PIT count dates in the Pulse are being realigned to match the January PIT counts so that 12-month intervals will be used in future Pulse Reports beginning with the Fourth Quarter, 2011). But the 10-month trend’s usefulness is minimized by the fact that only four CoCs had data available to analyze for the report.
Among the newly sheltered homeless population findings, there was a decrease in 55 percent of CoCs (when comparing October to December, 2010 with January to March, 2011). The 12-month trend of the newly sheltered homeless population showed that 50 percent of CoCs experienced a decline in population. But, as with the sheltered PIT counts, the longest trend period had analyzed data for only four CoCs.
“So,” you ask, “then what does it all mean?”
“Well,” I finally get to the point and respond, “it means not much.”
At least, as it relates to gaining a greater sense as to the “pulse” of the nation’s homelessness problem, it means not much just yet. The only thing of import that we can definitively say about HUD’s most recently released homeless data report is that sheltered homelessness increases—both overall and among newly homeless populations—during periods of cold weather. But we already knew this to be the case.
Well, we can also definitively say that future Homelessness Pulse Reports will be more informative beginning with the Fourth Quarter 2011 report. The alignment of the shelter PIT counts with the annual January PIT counts and the inclusion of a larger sample of CoCs in this Pulse Report are steps in the right direction. These changes to the Pulse Report will help to shed greater light on how well we are doing with ending and preventing homelessness. But all this is down the road when we are able to look at trends that are not affected by weather or whose impact is minimized by small sample sizes.