Critique of Study by Moscovitch

With the benefit of some time to think, I want to point a few potential shortcomings to Edward Moscovitch’s study of housing supply in US metro areas.

Moscovitch published a report for the Massachusetts Housing Partnership that correlated high levels of housing production with high levels of job growth.  He looked at 242 metro areas from 2000 to 2006.  He found that high shares of manufacturing employment were also likely to be found in regions that had low job growth.

First, I wonder how this study would change if it were compiled over a longer period of time.  The years from 2000 to 2006 were one of low interest rates, a relatively booming economy, and short of 9/11, constant prosperity.   Actually, now that we are in a real recession, the decline after 9/11 hardly feels that severe at all.  More like waiting for the party to start back up again.  A lot of times it is worth tracking a theory over a full economic cycle, from boom to bust to boom again.  Maybe if the study began in 1980, it would tell us something different.

s well, there is the extent to which housing supply and job growth are somewhat interrelated.  If you have a lot of homes being built, then you have a lot of construction jobs.  The impact is immediate.  There is spillover as well, as each construction dollar certainly contributes to demand for more local goods and services.  A BLS study found that $1 billion in construction spending creates 24,000 jobs.

That is also why I doubt that the finding on manufacturing is replicable outside of this time frame.  In today’s world, manufacturing is changing, but communities that do not add to their job mix beyond manufacturing are laggards.  Today, we see how well Pittsburgh is doing because it moved beyond manufacturing and into education and health care.  Youngstown or Detroit didn’t not fare as well.

I like his conclusions for policy makers about zoning and taxes.  He suggests that large-lot zoning creates burdens on growth, presumably because it creates both public costs (infrastructure, services) and for individuals (travel time, travel expenditures) that are not efficienct.  He also rebukes the idea that low taxes are correltated with job growth.

Then there are factors that are not a product of manufacturing or of housing supply. Moscovitch definitely acknowledges this point.  He tracks where retirees choose to live.  That makes sense, because those are individuals with fewer restrictions on their location choices and no restriction based upon job choice (his dependent variable).  He finds that some movement is just intrinsic to the community.  Weather matters, for instance.  People like living in warm areas.  The Southeast and the West are full of places with pleasant climates.  It sounds “not academic,” but nonetheless, its true, like the way that water is wet.  You can’t deny the attraction to good weather.  It creates a supply of labor, and perhaps there is the chance that available labor can in some instances also lead to job production.