Should list codes include subnational information in the first component?

[timgdavies]

The handbook for org-id.guide states that:

"Each organisation list code is made up of two parts: a jurisdiction code, and a list name code."

A dash is used to separate these components. The documentation then says "For any list which contains entries only from a given country, the ISO 2-digit country code should be used." but does not say anything about sub-national jurisdiction codes from ISO-3166:2.

Looking at our current data, we see two practices present. For example:

• A number of regional corporate register lists in Canada include the sub-national code before the dash - e.g. CA_PE-PEI-IPE
• Other lists with only subnational coverage (e.g. Scottish Charity Commission in the UK) use just the national ISO code - e.g. GB-SC

We should either:

• (1) Update the guidance so that we explicitly note that jurisdiction code may include subnational information;
• (2) Update the guidance so that we explicitly state that jurisdiction code should be 2-digit, and update the Canada entries accordingly to align them with this;

Comments on this are invited over the next two weeks.

[timgdavies]

My preference is for option (2), as this means systems have a clear predicable set of values before the first dash, and it is easier to be consistent across countries.

[BobHarper1]

It makes sense that there is a consistent approach to using (or not using) subnational codes in the jurisdiction code part of the list code.

My immediate preference is for (2), but with a small reservation. It hinges on whether there is the potential that there could be a type of list that is repeatable across many sub-national divisions: e.g. each of 10 regional jurisdictions in Canada has a corporate register held individually. If the jurisdiction code part were 2-digit only, you would need to differentiate the list name part ten times.

I know that in the Canada example, all those company registers have quite different names. The problem would only be likely where each subnational register had a very similar name.

I appreciate that this differentiation could be done by giving guidance to use the subnational ISO code in the list name part. Would that result in too inconsistent an approach?

[BobHarper1]

As an addendum to this, I think that we can also make changes to the Schema, depending on the option chosen, to update the List Code property with validation keywords (for maxLength and also string pattern)