Thursday, June 28, 2007

population again

So far there is some coverage of the population estimates that were released today, but so far it's not too bombastic. What just came out are sub-county (i.e. municipal) level population estimates for 2006. Everything else, county-level, MSA-level and state-level population-estimates for 2006 have been out for some time. The Trib did have a top of the fold coverage, but overall it seems to not be as big a story as the last round of population estimates that were put out.

Down in Cincinnati they seem to be convinced that the census has undercounted them. I would caution those who want to make the same argument here. I can think of as many reasons for why the Census overcounts City of Pittsburgh population as there could be reasons for an undercount. Why that could be is too long a topic, but one example: I don't believe the estimates take into account the depth of housing unit vacancy in the city. The methodology they use would not quickly capture increasing vacancy rates in the city, and would essentially assume the same percentage of houses are occupied this year as last. There are some reasons to think city population has not declined as much as the estimates show. Student population in dorms could be increasing and not readily captured in the most recent data. But overall, the issue is not just a city of Pittsburgh issue. There are declines throughout the county and region. In fact, if you rank municipal populations by the percentage decline over the last 6 years (per this data that is) Pittsburgh (the city) ranks near the middle of municipalities across Allegheny County.

In a sense there is actually little new here in terms of information. County-level population estimates came out months ago. All that was released today are estimates of growth within counties based almost entirely on the pattern of new building permits. There has been no new census and certainly no enumeration of people in every municipality in the last year. These are just what they say they are: estimates. One way to interpret the data today is just as an indicator of net residential housing development more than anything else. In most places that correlates pretty closely with population movement.

Here is the a map of the data released today. By and large it looks much like the pattern within the county over the last decade, and the decade before that, and the decade before that.

Estimated Population Change 2005-2006
Allegheny County Municipalities


Anonymous Anonymous said...

Friendly suggestion: I'm red/green colorblind, unfortunately. A blue-to-yellow shade palette is much more accessible to the general population. Thanks.

Thursday, June 28, 2007 3:06:00 PM  
Anonymous Anonymous said...

The curious thing about the Cincinnati case is that the Census estimates state Cincinnati's county of Hamilton has had one of the highest rates of decline of any large county in the country. It's lost over 40k since Census 2000 and 6k in the past year (down to 822k). When the Census revises their estimates for Cincinnati to show an "illusion" of growth... it appears they just shift some population from the suburban municipalities to the city... since the estimate for Hamilton County is not revised. It seems to be a strange and inconsistent mix of methodologies and procedures.

Thursday, June 28, 2007 3:56:00 PM  
Blogger C. Briem said...

For these numbers, the county population estimates are fixed. Before or after revision estimates of the municipalities will add up to the county population. So yes, if they revise the city of Cincinnati up, they will take away people from elsewhere within the county so it adds up. I don't know about strange, but it is a different methodology for the county level estimates and the sub-county estimates. Some of that is data-driven (there isnt enough data to measure births deaths and migration at a municipal level) but there are other reasons you need different techniques for different levels of geography.

Thursday, June 28, 2007 5:42:00 PM  

Post a Comment

<< Home