This study examines the effect of university host community size on st
ate growth. It is argued that if some positive spillovers from univers
ities are localized, needing a host community for capture, or if unive
rsities share agglomeration economies with their host communities, and
ii these effects are large, one may find a significant effect on stat
e growth. Using pooled data from eight U.S. censuses-primarily state-l
evel and university county-level variables-a significant positive effe
ct of university community size is found on state aggregate personal i
ncome growth. Weaker effects are found on state employment growth (pos
itive) and state population growth (negative). This suggests that stat
es and perhaps less developed countries could leverage more economic b
enefits from siting university programs in larger urban areas. A sensi
tivity analysis reveals no upper bound to the effect on income but a d
eceleration of the effects on employment growth and population growth
when cities become very large. (C) 1998 Elsevier Science Ltd. All righ
ts reserved.