[lostinquebec]

I think the articles problem is the lack of a definition of "crisis", and that leads to a bad mismatch between the headline and what is a reasonable look at the data.

My issues with the article are:

It starts from a too high aggregation level - almost no one is homeless in California, they are homeless usually in a very narrow area e.g. within 500 metres of bridge X.

Percentage change in raw numbers are not a good way to measure "crisis", and especially not at a state or city aggregated level. It seems more like traffic to me. Even in peak hour traffic, many roads are free of traffic, and the worst affected roads are those where the level of cars exceeds the roads ability to cope. That happens in a thin range of total cars per minute for specific roads, not due to X% increase overall. The same is true of homelessness in a city vs ... let's call them "hotspots".

Perhaps even more important than numbers in specific areas is the actual conditions homelessness creates in those areas. "Bad conditions" could be everything from human faeces on the street increasing, to murders, over doses, disease and unsanitary conditions, and it is possible for "bad conditions" to decrease and homeless numbers go up, or visa versa. That is harder to measure for sure, but probably closer to what most people mean by crisis.