Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
This paper analyzes decisions regarding the location of headquarters in the U.S. for the period 1996-2001. Using a unique [fi]rm-level database of about 30,000 U.S. headquarters, we study the [fi]rm- and location-speci[fi]c characteristics of headquarters that relocated over that period. Headquarters are concentrated, increasingly so in medium-sized service-oriented metropolitan areas, and the rate of relocation is signi[fi]cant (5% a year). Larger (in terms of sales) and younger headquarters tend to relocate more often, as well as larger (in terms of the number of headquarters) and foreign [fi]rms, and [fi]rms that are the outcome of a merger. Headquarters relocate to metropolitan areas with good airport facilities--with a dramatic impact, low corporate taxes, low average wages, high level of business services, same industry specialization, and agglomeration of headquarters in the same sector of activity--with all agglomeration variables having an important and signi[fi]cant impact.