Sunday, March 22, 2009

Vacant homes graph

From the Congressional Budget Office:
Most analysts believe that the correction in house prices is far from complete; for example, according to the February Blue Chip consensus (which is the average of about 50 forecasts by private-sector economists), the 20-city version of the S&P Case-Shiller index will fall a further 14 percent in 2009.

Housing starts in January plunged to 477,000 (at a seasonally adjusted annual rate), an all-time low since the U.S. Census Bureau started tabulating them in 1959. Although they rebounded somewhat, to 583,000 in February, that number contrasts with the more than 2 million starts at the height of the boom in 2005. Even though the construction of new homes has been at an extremely low rate for more than a year, no progress has been made toward reducing the excess supply of vacant units. ...

After rising for much of last year, mortgage rates—both for conforming loans and for larger, or jumbo, loans—fell late last year, and they have remained low thus far in 2009. Lower mortgage rates have spurred applications for refinancing; nevertheless, the number of applications for loans to finance purchases of homes has fallen this year.


  1. I wish I knew more about where those houses were located. For example, the rust belt has been losing population for 20+ years meaning the vacancy rate forever increases. If the homes are located up there, we (the US) will recover long before that number decreases much.

    Contrast that to the sunbelt where there has been excessive building but also continuing migration. If most of the vacancy was there, it absolutely has to decline before there is any recovery.

  2. I would also like to know where the houses are located. Except for a spike in the late 80s and the recent bubble, this graph is really a straight line. This suggests something like that it's because of the rust belt (and the Great Plains states) losing population.

  3. It also suggests that as household sizes have become smaller, and the family unit less 'extended', there's more "float" in home occupancy. E.g., When extended families sharing a home were more the norm, you could have a subset of the family (son and daughter-in-law) move to a new dwelling, leaving the parents and other family members in the original home ... and never having a "vacancy". Nowadays, with smaller and smaller family units, a housing unit can become more easily vacant. Similarly, the fact that people move more often would also add to the float represented by the numer of vacancies.

  4. 7% of DC residents have AIDS. What kind of family units does that translate to nowadays in DC, hmmmm

  5. 3%, not 7%, and that is more than enough.