Journal of Real Estate Finance and Economics, 18:1, 63±87 (1999)
# 1999 Kluwer Academic Publishers, Boston. Manufactured in The Netherlands.
Where, When, and by How Much Does Abnormal
Weather Affect Housing Construction?
JAMES T. FERGUS
Division of Research and Statistics, Federal Reserve Board, Washington, DC 20551
This article investigates whether departures from normal in precipitation or temperature have a signi®cant
contemporaneous effect on housing starts in each month of the year, for the nation as a whole and in each of the
four Census regions. It also evaluates the extent to which these immediate effects are reversed in later months.
The results indicate that atypical weather has statistically signi®cant effects on the change in housing starts that
are concentrated in the months of the ®rst quarter and that the magnitude of these effects is quite substantial.
However, such effects also are found in some other months as well. Signi®cant lagged effects are found that tend
to offset the contemporaneous effects of weather deviations.
Key Words: weather, housing starts, home construction, housing supply, construction cost
In interpreting data on housing construction, economic policymakers and decision makers
in the private sector share a common problem: how to distinguish the underlying trend of
construction activity from distortions owing to peculiar weather conditions. Journalists, as
well as the economists they quote, often lean heavily on the weather as an explanation for
unusual changes in housing starts, especially in winter months. Consider the two following
examples from the Wall Street Journal:
It's the season to believeÐexcept when it comes to housing starts. The eye-popping
9.2% rise in November starts masks what is really just a solid housing market. . . . Builders
may have simply delayed their digging this fall because of weather problems and were
playing catch-up in November (December 18, 1997, p. A2)
Mild February weather kept backhoe operators digging foundations, pushing up housing
starts a stunning 12.2%. . . . Because the weather has been so unusual in recent monthsÐ
from ¯ooding in the West to mild conditions in the EastÐit's dif®cult to spot a trend in
housing starts. ``The weather's just fouled up everything,'' [one economist said.] ``What's
really going on out there? Who knows?'' (March 19, 1997, p. A2)
Furthermore, conditions in the ®nancial markets reportedly depend in part on the
interpretation placed on weather developments (Washington Post, March 1997, p. C10)
Starts of new U.S. houses unexpectedly advanced in February to the highest level in
almost three years, helped by mild weather and rising consumer con®dence in the