A depressing finding. The unfortunate reality seems to be the 'market' is pretty small for these hotlines.

We need to solve the underlying issues of loneliness

The title doesn't match the paper, which is about the results of improving the pre-existing hotlines setup, not about implementing a hotline or not.

This incorrect conclusion is directly called out in the paper:

> These estimates should be interpreted with a few limitations in mind. First, because 988 was implemented nationally, the design estimates the effects of greater exposure to 988-related local capacity expansion, not the total national effect of 988 relative to no crisis line.