When taxes raise the full price of a good above that in nearby jurisdi
ctions consumers have an incentive to cross into the lower-price juris
diction to make purchases. Using a simple microeconomic model of the c
onsumer's border-crossing decision, we derive an econometric model to
test the significance of border crossing and estimate the magnitude of
the resulting sales. Examining cigarette sales in the continental U.S
. over the period 1960 to 1986, we find strong evidence that border cr
ossing is a significant factor in explaining sales differentials betwe
en states. Implications for demand estimation and excise tax policy ar
e discussed.