On Wed, 25 Aug 1999, CTBrown wrote:

> There is the all the lemons fell down line of thinking also. What percentage
> of buildings are still around after 100 years. Why are they still here?

Some years ago, I did a crude little study of how the survival rate of
buildings varied by region.  Essentially, I took the number of Dwelling
Units by county from the 1940 U.S. Census and compared that number with
the number of "housing units built 1939 or earlier" in the 1980 census
(the oldest housing category).

Sure, sure, there are lots of potential issues with this data: census
respondents in 1980 could be wrong about when a building was built, single
family houses could have been converted to multiple and vice versa, or to
nonresidential and vice versa, census data inherently has errors, the 1980
data on year-built are based on a sample, etc., etc.

On the other hand, most of the country's housing units are single family,
and the highest concentration of weird exceptions and conversions are
packed into a handful of metropolitan counties, out of the 3,000-plus
counties in the U.S.

I mapped these to counties, using red for "25% or less retention" and
green for "75% or more retention".  New England was blanketed in green,
and there were a thick scattering of green counties across upstate New
York, some of Pennsylvania, and the upper Midwest, particularly Wisconsin.
These were areas where a substantial majority of the housing stock that
existed in 1940 was still in use forty years later.

Much of the lower Mississipi Valley was blanketed in red, and red counties
dotted the entire South.  In these counties, most of the 1940 housing
units had been abandoned or replaced by 1980.

In general, counties which had undergone large increases or decreases in
population tended to have lower retention rates -- but this effect was
secondary to the very strong regional difference.

One might expect that the areas with low retention had older housing units
in 1940, nearing the end of "useful life" -- but this was not the case.
Indeed (though I did not do the statistical analysis) I would guess that
the older the housing stock in a given county was in 1940, the more likely
it was to survive: it already had a track record, so to speak.

Based on this data, it appears that houses built to withstand Northern
winters tend to last longer.

---
Lawrence Kestenbaum, [log in to unmask]
The Political Graveyard, http://politicalgraveyard.com