This is sad — maybe for a different reason than you might think. There are very few ways to measure whether or not a change has an effect. The most consistent one is to randomly provide the change to some and not others. What I’ve seen over and over, is to folks responsible for implementing programs to reduce suffering, it feels cruel that some would be randomly selected to receive help while others would not. So they choose to give help to everyone. And that makes the impact of offering that help exceedingly difficult to measure. With that approach, the program needs to be wildly successful in order to markedly change some outcome like total number homeless people. You might say, “YEAH that’s what it’s supposed to change!” But reality is messy — a complex outcome like whether someone remains homeless or not is usually explained by a lot of smaller factors combined together. What would it take for YOU to decide to live your life in a radically different way, even the ways that others might consider self-destructive? Why would someone who is homeless be meaningfully different in this regard? It is much more likely that homelessness is solved by systematically addressing many smaller factors, and only once enough of those factors are addressed in concert will the complex reality of homelessness change. But we won’t know how to address factors in concert unless we can measure if a program changes a factor. And again, it is exceedingly difficult to measure this unless some are helped and others are not. And so it is kinder in my mind to provide help in such a way that the impact of the help can be measured. This problem is concentrated where public funds are used, because the public expects proof that the funds are used wisely. If it’s not, the public will insist the funds be used for something else. There is no option to say, “We believe this is the best use for this dollar” — why should they trust you?