I think research around homelessness generally has demonstrated that there is a complexity in that prediction as well. Being able to identify who is most at risk is not a straightforward process. There may be indicators of certain things related to precarious employment and the insecurity of housing that may precipitate things, but there's also a parallel issue related to a sense of disengagement and disenfranchisement so that some people may actually elect to pursue a state of homelessness rather than fall into homelessness. It is not a homogeneous group.
On November 17th, 2016. See this statement in context.