NEW! CalculatedRisk Newsletter on Real Estate (Ad Free) Read it here.

Thursday, March 24, 2011

Census 2010: Housing Occupancy State data released

by Calculated Risk on 3/24/2011 05:15:00 PM

The Census Bureau has released housing occupancy and vacancy data for all 50 states and D.C. (and Puerto Rico).

Here is a spreadsheet for the 50 states (and D.C.) including the 2000 and 1990 Census data. Note: Adjusted data is using this analysis (ht Tom Lawler)

The highest vacancy rates were for Maine (22.8%) and Vermont (20.5%), but we have to be careful using this data - those states always have a high vacancy rate on April 1st because of all the vacation homes. So we need to compare to previous Census data to try to estimate the excess number of vacant houses.

In future releases, the Census Bureau will release more housing data: "Later reports from the 2010 Census will show the vacancy status for all vacant housing units and will also show whether occupied units were owned or rented." That data will help estimate the number of excess vacant housing units.

The table below is a summary (and raised questions about the data). It will be a few years before analysis is published on the accuracy of the Census housing data, but Census 2010 was probably more complete than the earlier data (that has been the trend).

Housing Stock, as reported 
Housing Stock, Adjusted (1) 
Increase (adjusted) 13,341,04915,118,272
Units Completed   
1 to 4 units 11,122,00012,700,500
5+ Units 2,237,4002,723,600
Manufactured Homes 2,854,1001,313,700
Total Added 16,213,50016,737,800
Demolitions per year (calculated) 287,245161,953

Using the adjusted Census data for 1990 and 2000, we can compare the increase in housing units to the number of housing units completed. The difference is an estimate of demolitions (or otherwise destroyed units).

We can compare the increase from 2000 to 2010 of 15.1 million total units to the 16.7 million total completions (assuming no adjustment to 2010). That suggests about 167 thousand demolitions per year. But the Census probably undercounted some - so the demolitions would even be less. That is lower than most estimates of demolitions. Oh well ... this is a start.

For some fun from the Census Bureau:
The U.S. mean center of population, as of April 1, 2010, is near Plato, Mo., an incorporated village in Texas County. The U.S. Census Bureau calculated this point as the place where an imaginary, flat, weightless and rigid map of the United States would balance perfectly if all 308,745,538 residents counted in the 2010 Census were of identical weight.

Ever since Chestertown, Md., was determined to be the center of population after the first census was conducted in 1790, the center of population has told the story of America, illustrating how we’ve grown as a nation. It follows a trail across the country ─ across Maryland, Virginia, West Virginia, Ohio, Kentucky, Indiana and Missouri ─ that reflects our history of settling the frontier, manifest destiny, waves of immigration and regional migration.
Note: This is 23.4 miles from Edgar Springs, Mo., the 2000 mean center of population. Here is the data and maps on Centers of Population for the 2010 Census