We draw on a three-dimensional panel dataset with information shelf prices and available
coupons for 25 RTE breakfast cereal products in up to 65 cities for every quarter from the beginning
of 1989 until the end of 1992.For example,in a given quarter we might observe some cities in
which there was a coupon distributed for a particular brand and other cities in which a coupon
was not distributed for that brand.We compare shelf prices under the two scenarios.The static
monopoly price discrimination theory predicts a (weakly) positive correlation under a broad set of
assumptions.Since coupon and price decisions are made simultaneously it is necessary to control
for common shocks to demand or costs that may affect both decisions.We do so by exploiting
the panel nature of the dataset