Census errors and demographic analysis

[SGB]

Here's the scenario:

Today I discovered that the U.S Census Bureau erred with the geographic placement of a population of prison inmates. The prison is in Town A, and the population was placed with the data for Town B. The Bureau today admitted their geocoding error, but it's too late to change the "official figures."

This error creates a significant change in the population growth figure, as illustrated below:
Before After
Town A -26.26% +7.35%
Town B +20.25% +5.75%

[Edit: my kingdom for a table editor!]

The actual number involved is small: 242 persons. But, given the small populations of the towns involved, this error makes a significant difference when evaluating population growth rates.

We are working on our regional comprehensive plan right now, and are discussion how to deal with this error in our demographic analysis. No significant population threshold for municipal finances is involved here, thankfully.

I beseach the Throbbing Braintm for its communal guidance. How could we best address this error in our comprehensive plan?

 

[Cardinal]

We had a similar issue here. The local campus straddles a county line. The dorms are in J____ County, but the housing office is in a building in W____ County. In the most recent census, somebody figured it would be OK, and less work for them, to simply give all of the dorm residents a single address, which was the address of the housing office. Fortunately, we caught the error and changed it in time. All of which does nothing to help with your problem.

I would correct the data internally rather than use information that you know to be wrong. You should definitely note the problem and solution in the write-up.

 

[GeogPlanner]

i agree. this is a chronic problem here in tech valley where a major university straddles a municipal boundary and two of the dorm quads are always misappropriated. it sparked an interesting battle this past year where the city billed the neighboring town for fire and EMS service it had been providing to the entire campus including the town part. needless to say, now the town and city split services on the campus. anyway, we know to watch for the population error these days and it's probably the first error checking we do when the numbers come our. my work at the regional level...we would make changes when the census count was in err and make a note of it in the report or publication. don't see why you can't do that in your case...it would seem legit as long as you make a footnote or something about it.

 

[giff57]

I agree do what Cardinal said. We had the same thing happen with a nursing home, the allocated them to the county and not the city. I was watching for the error, because they did the same thing in 90. Got it corrected inside the window for geographic errors, but I don't think they fixed it in the official count. It did make a difference at the time, because the state use to share revenues based on population. They pulled that rug last year so it doesn't matter as much.