The empirical work in this study serves as a test of the principle tha
t redevelopment of a property will occur when the value of the existin
g bundle, plus demolition costs, is less than or equal to the price of
vacant land. The probability of a commercial or industrial property b
eing redeveloped is predicted using the difference in the estimated va
lue of a parcel in its redeveloped slate and its current use, correcti
ng for sample selection bias. Tbe empirical results support the hypoth
esis that land is redeveloped when its value in the redeveloped state
is high relative to its value in the current use. (C) 1996 Academic Pr
ess, Inc.