In this report we introduce a model-based framework for estimating the number of people who experience homelessness in Los Angeles over the course of a year based on point-in-time data. We refer to this as the annualized estimated of the population size. Unlike capture-recapture methods, this model is based on a single set of anonymous point-in-time data. This makes it viable for nearly all jurisdictions that conduct annual counts of their homeless population and gather basic information about the duration of homelessness within this population. In addition, it allows us to estimate the breakdown of the annualized population by duration of homelessness. A practical application of this is to quantify the distribution of homeless services needed to provide progressively more intensive levels of intervention for individuals with longer durations of homelessness. We incorporate measures of uncertainty in our final estimates, reflecting that demographic data is derived from a sample of the target population. As such, our model provides more complete information about the annualized population than the model developed by Burt and Wilkins (2005) that is currently in use in Los Angeles.
Estimating the Annual Size of the Homeless Population in Los Angeles Using Point in-Time Data
Economic Roundtable
Year: 2018