Using experimental auctions for marketing applications: A discussion

Journal of Agricultural and Applied Economics, Aug 2003 by Lusk, Jayson L

The present article discusses general issues associated with experimental auctions and their relative advantages and disadvantages over other marketing research techniques. Experimental auctions create an active market environment with feedback where subjects exchange real goods and real money, which is not generally the case with other methods. The article also discusses four experimental design issues associated with experimental auctions: auction mechanism, market feedback and bidder affiliation, demand reduction and wealth effects, and multiple attribute valuation. Each of these experimental design issues, if not properly controlled, have the potential to create serious flaws in marketing recommendations.

Key Words: auctions, experimental economics, marketing, valuation, willingness-to-pay

Agricultural markets have historically been dominated by the production and sales of generic commodities. However, in recent years, a pronounced trend has developed toward a more demand-driven marketplace in which agricultural producers must give considerable thought into consumer demand for specific food and fiber attributes prior to making production decisions. The shift in agricultural marketing and production can be noted by observing the increase in quality differentiated foods and fibers in the marketplace such as non-genetically modified foods, organic foods, organic clothes, "natural" meat, irradiated meat, ready-to-eat meat, etc. This shift has also generated a large number of policy issues, including appropriate labeling polices for genetically modified, organic, and "natural" foods; determining optimal levels of food safety regulation; and assessing the impact of these consumer-driven policies on agricultural producers. In an effort to assist agricultural producers and agribusinesses in determining potential profitability of selling new goods or modifying existing products, researchers are increasingly using market-research techniques. In that regard, experimental auctions (EA) are becoming an increasingly popular market-research technique among agricultural economists (e.g., Buhr et al.; Buzby et al.; Dickinson and Bailey; Fox; Fox et al.; Hayes et al.; Hoffman et al.; Lusk et al., 2001a, 2001b; Melton et al.; Menkhaus et al.; Roosen et al.; Shogren, List, and Hayes; Umberger et al.).1

However, it is interesting to note that using EA for the purpose of making marketing recommendations has been generally limited to the agricultural economics literature. Only a couple of studies have used EA in general marketing literature (e.g., Hoffman et al.; Wertenbroch and Skiera), and the use of EA in the general economics professions has been generally limited to induced-value studies (e.g., Coppinger, Smith, and Titus; Kagel, Harstad, and Levin) or to theory testing (e.g., List and Lucking-Reiley). Thus, as a profession, we appear to be the "bearer of the torch" in using EA for marketing applications. The present article focuses on addressing two primary questions. First, why are we as a profession the only to use EA for marketing applications, and are we justified in doing so? Second, given that EA have a place in making marketing recommendations, what aspects of experimental design should practitioners consider when designing an EA?

Experimental Auctions

Agricultural economists have been using EA with marketing implications since at least the beef marketing studies of Menkhaus et al., Hoffman et al., and the food safety studies of Shogren et al., (1994a,b) and Hayes et al. Using EA to elicit applied valuations grew out of the general experimental economics literature that had, for the most part, primarily focused on induced-value experiments in which subjects were given values by the experimenter rather than the experimenter eliciting a subject's "homegrown" value for a good. Using auctions for the purpose of eliciting values for foods safety or quality was a natural fit for agricultural economists because of our interest in applied issues and our concern with using methods consistent with economic theory. Despite the increased use of EA in the agricultural economics literature, EA are only one of many tools available to carry out marketing research. In the following, the advantages and disadvantages of EA over two other widely used marketing research techniques, contingent valuation (CV) and conjoint analysis (CA), are discussed in an effort to assess the relative merits of EA.2

In a typical CV question, a new product is described, and subjects are asked, hypothetically, either how much they would pay for the new good or whether they would purchase the good at a particular price level (e.g., Lusk). The CV method is extensively used in the environmental valuation literature, primarily because most environmental goods are public and undeliverable (e.g., clean air). The hypothetical nature of CV has some merits. First, in a hypothetical CV, virtually any new product can be described and valued without actually having to develop or deliver the good. Second, in a CV exercise, subjects can be asked how they would behave in a grocery store, for example, whereas, values elicited in an EA are contingent on tastes and preferences at the time and location of the experiment. Despite the advantages of CV, a great deal of research has shown that subjects overstate the amount they are willing to pay for a good in a hypothetical setting as opposed to when actual payment is required (e.g., Cummings, Harrison, and Rutstrom; Fox et al.; List and Shogren 1998). The advantage of EA over CV is that EA are incentive compatible and are conducted in a nonhypothetical context that involves real goods and real money. EA also have advantages over CV methods because EA put subjects in an active market environment where they can incorporate market feedback and be held accountable for their behavior. Furthermore, EA might be preferred over most generally accepted CV methods (such as the dichotomous choice question) because individual willingness-to-pay values are elicited from each individual, whereas only ranges on individuals' willingness to pay are observed with a dichotomous choice CV question.