[Talk-us] Alaska CPD boundaries

[Paul Norman]

> From: Greg Troxel [mailto:gdt at ir.bbn.com]
> Sent: Monday, November 26, 2012 4:48 AM
> Subject: Re: [Talk-us] Alaska CPD boundaries
> 
> 
>   The census bureau divided the unorganized borough into 11 census
> areas.
>   These have no legal significance but serve to sub-divided the state
> into
>   convenient parts. In spite of this they are in many ways like
> counties. I've
>   tagged them the same as counties (admin_level=6) but I'm not convinced
> that
>   this is the best tagging.
> 
> If they're just a census division, it seems wrong to call them boroughs
> (treating them like counties with a different name).  Boroughs as
> counties seem right - it's the way the state divides things less than a
> state and more than a town.  The point of admin_level is government, and
> census boundaries (for the sole convenience of the census bureau,
> presumably) are not in any way governments.

I must admit I've flipped back and forth on my views on this. I initially
wasn't sure if they should be in but I've convinced myself that they should
be and admin_level=6 is probably the best.

In neither case are we tagging boroughs or CPDs as counties as counties.
We're tagging them as admin_level=6 which is what counties in the other 49
states are tagged as. admin_level=6 isn't automatically associated with
counties.

boundary=administrative is for administrative boundaries, not just
government boundaries. Normally the government boundaries are the most
meaningful ones but here there are no government boundaries. From a data
consumer's perspective I think any analysis done is likely to treat the CPDs
as equivalent to the boroughs. They have their own FIPS codes too.

Now, if someone local were to come and say "the CPDs are meaningless and
irrelevant" or "I say I'm from a CPD the same way someone would say they're
from a borough" I'd be happy and we'd have a good answer but I don't believe
anyone who's commented on the list is a local.