It looks like many Australian localities have a field called "qs:la_lc" which contains a Australian Bureau of Statistics ID in it.
This ID (Starts with SSC for localities) is extremely useful as it allows you to easily fetch a wide range of Census demographic information for that area, and it's also used in other datasets as a concordance such as WikiData which means it's also useful for linking between datasets.
Unfortunately it looks like "qs:la_lc" is derived from the 2011 Census dataset.
The ABS releases a new set of IDs with each Census, the IDs don't carry across from Census to Census.
E.g. Acacia Gardens has IDs 2011: SSC10010 and 2016: SSC10014.
I can see utility in keeping both IDs and other projects do store both, making a note which Census dataset the ID belongs to.
E.g. https://www.wikidata.org/wiki/Q7966919 has both 2011 and 2016 SSC IDs listed.
Can I propose that we deprecate and remove the "qs:la_lc" property?
And instead store these IDs in the concordances property as "ausstat:2011" and "ausstat:2016"?
This would be very clear, allows us to store both at the same time and is future proof for the 2021 census which will again use different IDs.
Thanks for the issue, @NickStallman. We try to track each admin issue in the main whosonfirst-data repo - I'll close this an replace it with a new one in that repo.
Hi Guys
It looks like many Australian localities have a field called "qs:la_lc" which contains a Australian Bureau of Statistics ID in it.
This ID (Starts with SSC for localities) is extremely useful as it allows you to easily fetch a wide range of Census demographic information for that area, and it's also used in other datasets as a concordance such as WikiData which means it's also useful for linking between datasets.
Unfortunately it looks like "qs:la_lc" is derived from the 2011 Census dataset. The ABS releases a new set of IDs with each Census, the IDs don't carry across from Census to Census.
E.g. Acacia Gardens has IDs 2011: SSC10010 and 2016: SSC10014.
I can see utility in keeping both IDs and other projects do store both, making a note which Census dataset the ID belongs to. E.g. https://www.wikidata.org/wiki/Q7966919 has both 2011 and 2016 SSC IDs listed.
Can I propose that we deprecate and remove the "qs:la_lc" property? And instead store these IDs in the concordances property as "ausstat:2011" and "ausstat:2016"? This would be very clear, allows us to store both at the same time and is future proof for the 2021 census which will again use different IDs.