us:prosecutor ocd-division ids

[kstohr]

Hi there,

How are prosecutor ocd_id's handled? I ask because I have tracked prosecutor races in the US for the last three years and find it very difficult to work with data that uses standard ocd-id's because they are sometimes innacurate. i.e they associate a prosecutor with a county rather than the prosecutorial district, judicial district or circuit.

I'd be happy to contribute, but given that these are neither local nor state level divisions I am not sure how you would normally include them.

Our list is based on looking up state law in all 50 states and confirming it with ballot information. So, we're confident our definitions are correct and up-to-date... including the 2018 NC redistricting.

Would love to get them added to the ocd-id repo, but not sure how to go about it.

[rsimoes]

I don't think it'd be a problem to add a prosecutorial district type

[kstohr]

@rsimoes That'd be awesome. How do you want us to go about this? Should we create a branch and make a pull request?

[rsimoes]

@kstohr Yes, please do!

[kstohr]

@rsimoes I can make a PR but can you first give me some guidance as to where (what file) in the codebase to add these to (or do you want me to upload a discreet file?) and how you would like to handle the following:

• Some prosecutor districts are co-extensive with counties. How do you want to label these districts? Use the existing ocd-id county labels or create a new prosecutorial-district label and append the county name?
• Some prosecutor districts align with existing judicial or circuit districts. How to handle?
• Finally, some prosecutor districts are purely prosecutorial districts and are not co-extensive with any other type of divisional id.

If you give me the label three level deep for the above scenarios I can generate the ids.

ocd-division/country:us/state:ca/??/
See sample data below.
d4nn6od1ppmo8k.xlsx

[rsimoes]

These are good and interesting questions, and I'm not sure there is consensus on all of them.

The convention we follow at CTCL has been the following:

• If the district is county-dependent by state law, treat it as identical to the county and do not mint a new ocd-id. If it instead has no legal relationship with a county except shared boundaries, mint a new ocd-id hierarchically subordinated to the state.
• Treat the prosecutorial district as identical to the judicial or circuit district and do not mint a new ocd-id.
• Mint a new ocd-id hierarchically subordinated to the state.

Additional/alternative feedback from @opencivicdata/division-id-curators?

[kstohr]

@rsimoes Ok, can you provide a little more color on what you mean by "county-dependent by state law." Is the primary determinant that the position is paid by the county? That the county manages the elections? What is the determinant factor here? I have the information on the above, would just need some guidance on how to make the call.

Texas: Texas is a different kind of beast. It has a few different organizing patterns for prosecutorial districts. I can either mint an OCD-id for all prosecutor districts state-wide or try to follow the above conventions which may end up being less useful.

Finally, don't forget to tell me which file/s to update in the codebase. Ok, to update /country-us/ dir with a file similar to the [us_state_courts.csv](https://github.com/opencivicdata/ocd-division-ids/blob/master/identifiers/country-us/us_state_courts.csv) file, ie us_prosecutor_districts.csv. Note: This file will include only districts for prosecutors responsible for trying felonies (whether appointed or elected). There are county attorneys whose jurisdictions overlap in some cases who are not authorized to try felonies, but those are generally county or municipal level roles.