There is a substantial literature predicting that a franchisee-owned store will generate higher profits than a franchisor-owned store, all else equal. However, attempts to estimate the effect of franchising on store performance are hampered by an important selection issue: the franchisor may choose to assign the least desirable locations to franchisees. I overcome this issue by using a 2007 corporate sale that resulted in all franchisor-owned Applebee’s stores in Texas being sold to franchisees as a source of exogenous variation. While I do not observe store profits, Texas makes store-level alcohol sales data available for all bars and restaurants that have a liquor license; I use alcohol revenues as a proxy for store performance. I first find evidence that both observable and unobservable location-level factors were important in Applebee’s decision to own or franchise a store prior to the corporate sale. I next create a structural demand model which uses consumer and store locations to predict alcohol sales for all liquor-selling bars and restaurants in Texas over a 10 year period. Using this model, I find that franchising an Applebee’s store increases its alcohol revenues by 7 percent. I also find that franchising a store produces a consumer utility gain comparable to a 2.8 mile reduction in distance from the individual’s home to the store.
Industrial Organization; Applied Microeconomics