A popular conjecture is that too many homes were built during the boom and part of our current economic difficulties are centered around finding new employment for construction workers that are no longer needed. I wanted to take a crack at the really long term data.
It was bit tricky getting what seemed like good recent data on total housing but I have housing starts going back to 1959. I compared that against the increase in population since 1959.
The graph below is the increase in the population since 1959 versus the number of new homes built since 1959.
There is a slight bulge in total homes built relative to the population in the mid 2000s but it looks as if virtually all of that has been undone in the later half. Here is a closer up look at the last two decades
Another way to look at it is the ratio between total population increase and total housing increase.
From the 70s through the late 80s there was an increase of just over 1.5 people per new home. This doesn’t count the number of homes that were bulldozed so the population per home is likely higher.
However, today we are well above that and indeed above the cumulative ratio of the 1990s. This first cut at the data doesn’t suggest that there are too many houses in the United States. Though there are obviously many other factors to consider, such as household composition and the fraction of people living in apartment buildings.

8 comments
Comments feed for this article
Saturday ~ March 19th, 2011 at 8:00 pm
rhmurphy
Homeownership rate shot up at the same time the Case-Shiller price index went crazy in 1997. It was extremely steady before that. So either we built more houses because of the implicit subsidies or people coincidentally en masse wanted to own homes now. One of those is more plausible than the other.
http://en.wikipedia.org/wiki/Homeownership_in_the_United_States#Historical
Thursday ~ November 1st, 2012 at 8:31 am
Bahadır Uçar
2012 information in there for 50.000.000 $ Houses
http://howmanyarethere.net/how-many-houses-are-there-in-usa-sold-over-50-million-dollars/
People’s are realy rich i think :/ .
Sunday ~ March 20th, 2011 at 1:09 pm
geaugailluminati
Nearly 20% of Florida homes are vacant http://finance.yahoo.com/news/Nearly-20-of-Florida-homes-cnnm-2507768369.html
Census Bureau datareleased in September showed that the number of multifamily households jumped 11.7 percent from 2008 to 2010, reaching 15.5 million, or 13.2 percent of all households. It is the highest proportion since at least 1968, accounting for 54 million people.
http://www.nytimes.com/2010/12/29/us/29families.html
Number of Vacant Homes in U.S. Hits 19 Million
http://www.huffingtonpost.com/2010/04/26/number-of-vacant-homes-in_n_552274.html
Sunday ~ March 20th, 2011 at 1:13 pm
geaugailluminati
another point: in the 50s & 60s families were much larger, so population isnt a good metric..
Monday ~ March 21st, 2011 at 1:53 pm
OGT
Right. Households is the best metric. Household formation out did population growth throughout the ’70′s and ’80′s, but slowed in the last fifteen twenty years as household size has largely stabilized.
Here’ Calc Risk’s Bill McBride:
http://www.calculatedriskblog.com/2010/04/housing-impact-of-changes-in-household.html
Monday ~ March 28th, 2011 at 4:01 pm
Osborne Ink » Blog Archive » Housing Bubble Finally Losing Air?
[...] would be easy to say the United States has too many houses, but that would be wrong. There is some evidence that younger workers are refusing the suburban commute and adding to urban [...]
Wednesday ~ April 6th, 2011 at 9:26 am
Are There Too Many Homes in America, Ctd « Modeled Behavior
[...] have previously argued that the data suggests that the “overbuilding” of the late 2000s was mild at worst in comparison to the growth in population. Indeed, it is reasonable to argue that there was no [...]
Tuesday ~ April 24th, 2012 at 3:54 am
Skepticlawyer » About Austrian economics
[...] result of very specific forms of perverse incentives, not indicative of some general phenomenon, even in housing construction (even if you add in various [...]