Chair asks IBO how it conducted the analysis

Thank you both.

Can you explain how you did your analysis?

Absolutely.

So IBO has, through our charter mandated powers to receive data from city agencies,

has data on every state in DOC custody from fiscal year 2008 through 2026.

We use that data to identify people who were sort of ever in the data set, either flagged for a Brad H.

Class member, so someone who is likely receiving mental health care in city jails,

and or had a discharge to state hospital for competency restoration.

So that's how we identified people with a mental illness.

And our numbers may differ from some other reported numbers because we are looking across multiple stays, given that many people do have very short stays in custody.

It's not uncommon to see stays of five days.

days or fewer.

And so people may not make it through all of the processes to be identified as someone with a mental illness in those short stays.

As it relates to people with a history of homelessness, we did a couple of different things.

We used the address data in the DOC data that we have available to identify keywords.

So sometimes you'll just see in that address data homeless or shelter.

So we flagged those keywords as well as matched to shelter addresses that IBO also is aware of.

And again, we looked across people's histories to identify people in case they weren't flagged when they had those very short stays.