We used an A-B-A design to evaluate the effects of two commonly used promotional activities-price reduction and increased exposure, in combination and separately-on sales and thefts of candy at a grocery store. The combination of activities and the increased exposure condition produced the greatest increases in sales. The combination of activities was also associated with the greatest increase in thefts.DESCRIPTORS: business problems, consumer behavior, theft, stimulus control, side effectsBehavior analysis has been proposed as a useful technology for understanding, predicting, and controlling consumer behavior (Rothschild & Gaidis, 1981). Although behavior analysts have shown that undesirable consumer behavior (i.e., theft) can be reduced with no negative effects upon sales (Carter, Holmstrim, Simpanen, & Melin, 1988;McNees, Egli, Marshall, Schnelle, & Risley, 1976), there seem to be no similar analyses of promotional activities and their effects on sales and theft in retail settings. METHOD: The study was conducted at a grocery store described in previous research (Carter et al., 1988). The store used an optical price scanning system with cash registers connected to a computer that maintained data on sales, prices, and so forth. All target products were endosed in factory packaging containing 13-digit bar codes. The store's antitheft activities were unchanged throughout the study. There were no reports of anyone being apprehended for stealing target products. Prior to the start of the study, the store owner selected 20 candy products as targets. The products, chocolate bars and packages of candy, were of similar price and size and would not be the objects of other promotional activities, although 40 other candy products were the object of some promotional activity during the study. All target products were normally located along one wall of the store, on four-tier wall shelves, and were visible to cashiers.Measurement system. Retail theft was measured with the aid of optical scanning equipment (Carter et al., 1988). The study began with an inventory of target products to determine actual stock and recording of cumulative sales figures. After subsequent biweekly inventories (on Tuesdays and Fridays) conducted prior to store opening, sales and thefts for each target product were calculated. All differences between calculated and actual stock were considered to be theft. During the study, none of the target products were reported to be damaged or returned. Counts of actual stock greater than the figure for calculated stock occasioned a recount. When the number of thefts exceeded three for an individual product, the observer recounted. Miscounts were open to discovery at subsequent inventories. Procedure. The study was conducted over 6.5 weeks. An A-B-A design was used to assess the effects of the intervention. The initial baseline period was 2.5 weeks (five inventories), followed by a 2-week intervention period (four inventories) and a second baseline (2 weeks, four inventories). The intervention period was limit...