Better data would dramatically improve this situation: - Statistical data on who is homeless and why - Data on how a given person's homeless life tends to proc

Better data would dramatically improve this situation: - Statistical data on who is homeless and why - Data on how a given person's homeless life tends to proceed over time - Data from homeless people on how they view their lives + their "satisfaction" (whether they are mentally ill or not" - Data that attempts to predict health / satisfaction in the future from past circumstances I'm sure some exists... but I don't hear much about it. The article points to the value of data, by indicating that compiling a consistent history (across service providers) of a homeless person's history can make it very clear when someone is in the (extreme) case that they are not going to take good care of themselves in the future. The (wonderful) discussion on this post goes through all the complexity of balancing respecting autonomy with promoting well-being. But even the best comments are anecdotes. I don't know if the greater evil (today) would be a) allowing people to destroy themselves or b) imprisoning them in a bureaucracy. Understanding with data (top down overall + longitudinally) could give a sense for where the common cases are, and what cases we consider are very rare. In criminal justice, we imprison innocent people. It sucks, and we will try to minimize it, attempting to "let 10 guilty people go free vs. convicting one innocent" But we don't go so far as to imprison almost no one. By analogy, with better data, I would feel more empowered to have an opinion of what direction to push in here.