Monday, June 21, 2010

did the city grow? Drumroll........................

For those who read story today talking about some population projections for the county...  the story was mostly an update to a story written 3 years ago and is worth (re)reading for the perspective.

What close readers caught in the story today was that we expect to see the latest (2009) population estimate for the CITY of Pittsburgh tomorrow. We have already learned what the estimated population is for the state, MSA and county are for 2009...  but folks want to know what is the comparable number for the city proper.  Should be the last big thing we care about before we learn the 2010 census population counts are released.

Should always know your data and municipal level population estimates are a big case in point.  Nobody is out there counting the number of people who live in the city each year.  The census comes up with an estimate based on various levels of modeling...  A far cry from actually counting people and no doubt there will be surprises when we learn better numbers from the 2010 census when they come out next year.

So the city's population, along with the population of all other individual municipalities in the state are really slices of the county population estimates we already know.  How is the county population divided up?  A bit simplified, but basically it mostly comes down to building permits issued.  Why?  Well, folks have to live somewhere to begin with.  There is also a big issue of working with what data you have and the data on buildng permits is one of the few consistent metrics at a municipal level that can be used to impute population trends.

So what to expect.  The big news of late is that Allegheny County's population in 2009 was a whopping 434 people.. something just over 0.03%.  Will the city's estimated population go up as well and appear to stop what has likely been a 60 year or more decline?  Below is a summary of what the building permits data are for Allegheny County and the City proper annually since 2000.   These are the total estimated housing units covered by issued building permits.  No guarantee they were all built, but it's a start. 

I see at least 3 different stories in that one table.  One story is just how low the total number of new houing units expected in either city or county in any year.  These are not big numbers compared to places that have seen growith in recent years or decades.  Thus one of the reasons that parts of the region have some of the oldest housing stocks in the nation.

2nd story is the trend, which reflects the housing crisis in part, but a downward trend for the county since 2006. 

Finally the building permits in the city remain virtually anemic as they have been for a long long time.  We hear a lot about some new developments... but for a city our size the number of new housing units is low and low in proportion ot the share of the county's population.  City is roughly a quarter of the county's population, but as you see from the table, typically accounts for less than 10% of all new housing units to be built..  When you account for demolitions, let alone de-facto abandonment, we must have a net decline in functional housing units in the city proper.

Soo..  if you follow the logic, the county's gain in population does not necessarily mean there will be much growth in the city..  The low % city housing units means more of the growth in the county will be elsewhere within the county.... but we will see in the morning.  The story may be that the city's population is still declining (per the estimates to come out) but it is likely at a slower rate of decline than in recent years.
Lots of other things going on even...  we really won't have a great idea what the populations are of local municipalities until next spring. So take this all with a grain of salt.  In the city and region there are other things that will impact current and future population.  Lots more college students living in the city of late.  College students impact population forecasts very differently than the household population.  One way or another households are really paying the taxes and the household population of the city of Pittsburgh is clearly as low as it has been in a long long time. 

ok, ok.. some data.

Total Housing Units in Annual New Privately-Owned Residential Building Permits

as % 


Anonymous BrianTH said...

So say a young couple with a child buys a house being vacated by a single retiree (moving to Florida, or the great retirement community in the sky, or so on). Net population gain is plus-two.

Building permit? Or no?

More generally, what if the City and the County were on different tracks as far as average people per household is concerned (relatively to their numbers as of the 2000 Census). Would the Census imputation methodology get the right result?

Monday, June 21, 2010 5:10:00 PM  
Blogger Jim Russell said...

Woods & Poole is all over the place. What's going on with its model?

Monday, June 21, 2010 5:17:00 PM  
Anonymous BrianTH said...

Found my own answer (which I was vaguely recalling). Courtesy of the Census:

"The Census Bureau multiplies the occupancy rate and average persons per household (PPH) from the latest census at the subcounty level by the estimate of housing units. The estimate obtained from this method is then controlled to the final county population estimate. The assumption implicit in this method is that changes in the occupancy rate and/or the PPH are measured by the updated county population estimate and that the rate of change in the occupancy rate and/or PPH is uniform within counties."

So if it happened to be the case that there was a significant difference between the City and the County in terms of rates of change in the occupancy rate and/or PPH, it would throw off the City's population estimate.

I wonder if that might be true in this case.

Monday, June 21, 2010 5:21:00 PM  
Blogger C. Briem said...

W&P's long range forecasts used to match ours pretty closely... and it does in the near term still. longer term there is some divergence and I am not sure why right now.

for the city.. you answered your question... but there are other problems. We know pretty well there is a large and growing housing vacancy problem in the city which is also not going to be captured in the annual estimates data. Housing that just goes unused.. coupled with city's backlog of demolitions would be a bias that would make the reported estimates data higher than may be true on the ground. That may be one of the biggest things we look at at the neighborhood level when the census stuff comes out.

Monday, June 21, 2010 9:33:00 PM  

Post a Comment

<< Home