Choosing brands: fresh produce versus other products.

We assume consumers are heterogeneous by their WTP for brands. As we discussed in the data section, we estimate the empirical density function of the underlying WTP based on the perceived WTP ranges using the maximum entropy method. Assuming that the estimated density function of WTP for brands of product category k is [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], where [[bar.W].sub.k] and [[W.bar].sub.k] are the upper and lower bound of [W.sup.*.sub.ik]. We assume a monopolistic competitive market in a sense that brands are substitutes for generic products to some extent. A firm produces a brand product with an extra marginal cost c and charges an extra percentage [p.sub.k] relative to the generic product. An individual consumer will buy the brand if and only if [p.sub.k] [less than or equal to] [W.sup.*.sub.ik]. Hence, the market share of this brand is

(12) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII].

This monopolistic firm will choose the optimal premium to maximize profits,

(13) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

where N is the total number of consumers. The optimal premium is achieved when the marginal revenue equals the marginal cost c:

(14) [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Equation (14) shows that a one-unit increase in [p.sub.k] will increase the revenue by [p.sub.k], but at the marginal loss, [MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII], resulting from a decrease in the demand. Solving equation (14) yields the optimal price premium [p.sup.*.sub.k] Substituting [p.sup.*.sub.k] into equation (12) yields the corresponding market share thereafter.

The simulation results show that the optimal price premiums for brands in fresh produce actually are higher than brands of electronics, clothing, and packaged food; however, the market shares of fresh produce brands are much smaller. For example, when the extra cost of brand products is 10%, the price premium for electronics, clothing, processed food, and fresh fruits and vegetables are 29%, 37%, 39%, and 44%, respectively; and 39%, 48%, 49%, and 59% when the extra marginal cost is 20%. However, the optimal market shares for brands of fresh produce are much smaller than in other product categories. For example, when the extra marginal cost is 10%, only 18% of the population will buy brand products of fresh produce in contrast to 23% for packaged food, 30% for clothing, and 50% for electronics. When the extra cost is 20%, 12% of the population will buy brand products of fresh produce, 16% for packaged food, 20% for clothing, and 30% for electronics. The small market share can partly explain fewer brands of fresh fruits and vegetables.

Once the optimal price premiums are established, we can identify whether an individual consumer will buy brands of a certain product and assess whether people are consistent with brand preferences across product categories. This assessment will provide insight to store organization and prediction of percentage of the population who will shop in each of these stores. Assuming that the extra marginal cost of brands is 10% relative to the generic one, almost half of the respondents will always buy generic products and another 9.27% will only buy brand products. In total, at least 58.94% of the potential consumers are consistent in terms of their brand preferences for electronics, clothing, packaged food, and fresh produce. We can at least identify three types of stores: (a) discount stores that sell generic products targeting half of the potential consumers, (b) elite stores that sell only brand items and attract 9.27 % of the potential consumers, and (c) supermarkets that sell everything. Our analysis suggests that elite stores with brand products are attractive to the consumer segment sharing consistent high brand preferences as well as WTP across product categories. Harry and David is one example of a company that takes advantage of the upper end of the distribution (http://www.harryanddavid.com/).

Overall, the simulation results suggest two things: (a) if we market brands of fruits and vegetables, we have to target a small market segment and charge a high price; and (b) since we find that people who are willing to pay for brands of fruits and vegetables are also willing to pay for brands in other categories, our results suggest targeting outlets that focus on brands of products in all categories, for example, high-end malls, like sky malls, or brand-focused retailers, like Macy's, Nordstrom, or Neiman Marcus.

Conclusions

The basic premise of this article is that brand value comes from its contribution to reduction in quality risk, the prestige it confers, and from superior design. The features for brand products are likely to vary in value among products and be appreciated differently by individuals with different sociodemographic characteristics. Based on this premise, we hypothesize that the relative value of brands in fresh produce is much smaller than in electronics, clothing, and packaged food. We also investigate the roles of sociodemographic factors on the WTP for brands.

Empirical results based on the data from College Station and Bryan, Texas, on WTP for brands of four product categories support our hypotheses. Tests on the mean difference of WTP across product categories based on the stated WTP or the estimated WTP distribution using the maximum entropy method suggest that WTP for brands in fresh produce is much smaller than in electronics, clothing, and packaged food. Using the random effect tobit model on the stated point WTP and the ordered probit model on stated range of WTP, we also find similar results that WTP for brands of fresh produce is least among four product categories controlling for relevant demographic variations. The empirical study also shows (a) the nonlinear effects of income and age (income increases WTP at a diminishing rate and age decreases WTP at an increasing rate); (b) females, less educated people, or smaller households are willing to pay more for brands; and (c) white respondents are less willing to pay for brands than other ethnic groups including African Americans, Asians, and Hispanics.

Based on the distribution of WTP for each product, we determined the optimal brand premium and the corresponding market share. The simulations suggest a potential for a small market of brands of fresh produce with a high margin relative to the generic products. Since consumers with strong brand preferences in other product categories tend to have a higher WTP for brands of produce, one strategy is to introduce brands in outlets that emphasize brands across the board. Another is to target market segments with consumers who are more likely to buy brands regardless of product categories, like high-income individuals, young people, and female shoppers.

The empirical results of this article are based on the stated preferences and not actual behavior. In spite of the obvious limitation of the data, it allowed us to compare preferences to brands in several product categories and relate WTP with sociodemographic variables. However, further empirical work with actual purchasing behavior is needed to further study the role and potential of brands of fresh produce.

Second, while we found WTP for brands of food products in general, it is useful to assess WTP for subgroups (e.g., organic versus nonorganic products), as well as different types of foods (vegetables versus fruits). Further research should assess the value of brands of organic food as well as the role of geographic indictor labeling as substitutes and/or complements to brands.

[Received May 2006; accepted July 2007.]

References

Aaker, D., and E. Joachimsthaler. 2002. Brand Leadership: The Next Level of the Brand Revolution. London, U.K.: Simon and Shuster.

Ainslie, A., and P.E. Rossi. 1998. "Similarities in Choice Behavior across Product Categories." Marketing Science 17(2):91-106.

Allenby, G.M., and P.E. Rossi. 1991. "Quality Perceptions and Asymmetric Switching Between Brands." Marketing Science 10(3):185-204.

Auty, S., and R. Elliott. 1999. "Fashion Involvement, Self-Monitoring and the Meaning of Brands." Journal of Product and Brand Management 7(2):109-123.

Caves, R. E., and D. P. Greene. 1996. "Brands' Quality Levels, Prices, and Advertising Outlays: Empirical Evidence on Signals and Information Costs." International Journal of Industrial Organization 14(1):29-52.

Chiang, J. 1991. "A Simultaneous Approach to the Whether, What and How Much to Buy Questions." Marketing Science 10(4):297-315.

Erdem, T., Y. Zhao, and A. Valenzuela. 2004. "Performance of Store Brands: A Cross-Country Analysis of Consumer Store-Brand Preferences, Perceptions, and Risk." Journal of Marketing Research 41(1):86-100.

Fussell, P. 1983. Class: A Guide through the American Status System. New York: Touchstone.

Ghosh, A., and S.L. McLafferty. 1987. Location Strategies for Retail and Service Firms. Lexington, KY: D. C. Heath.

Gupta, S., and P.K. Chintagunta. 1994. "On Using Demographic Variables to Determine Segment Membership in Logit Mixture Models." Journal of Marketing Research 31(1):128-136.

Hanemann, W.M., J. Loomis, and B. Kanninen. 1991. "Statistical Efficiency of Double-Bounded Dichotomous Choice Contingent Valuation." American Journal of Agricultural Economics 73(4):1255-1263.

Hayes, D., and S.H. Lence. 2002. "A New Brand of Agriculture? Farmer-Owned Brands Reward Innovation." AgDM newsletter article. Available at http://www.extension.iastate.edu/agdm/ articles/hayes/HayDec02.htm (Last accessed in July 2007).