Preventing homelessness: Algorithm can identify LA County residents most likely to lose housing

Spotting the homeless is not a difficult task in Southern California, but spotting people before they become homeless is. But now, Los Angeles County officials have a new tool that will help them identify people most likely to end up homeless.

Researchers at UCLA's California Policy Lab and the Poverty Lab at the University of Chicago have come up with a new algorithm that will help them do just that.

The predictive modeling program analyzes data of L.A. County residents who receive county assistance and can predict which of those people have the highest odds of losing their housing.

"We built a model that can analyze hundreds of thousands if not millions of data points that would be very difficult for a human to analyze," said Janey Rountree, the Executive Director of the California Policy Lab.

That information will allow county workers to offer emergency assistance to those residents in order to keep them from slipping into homelessness. Rountree says relatively small amounts of assistance early on can prevent expensive stays in emergency shelters.

"If we can reach them in that time period with a relatively modest intervention, we avoid all of the trauma and costs that comes with being homeless and living on the street," said Rountree.

The number of homeless in the region continues to skyrocket. Every day, County officials say they place 133 people into temporary housing while another150 fall into homelessness. Rountree says the new predictive modeling should put a big dent in the number of people headed for the streets.

Researchers are providing the model, which is expected to keep nearly 7,000 people off the streets a year, to the county free of cost.