Based on these tenderness categories, we created four qualities of beef: A, B, C, and V.
Quality A was very tender, quality B tender, quality C less tender, and quality V uncategorized generic
beef. In the experiments, we emphasized that quality V beef could be very tender, tender, or less
tender and, furthermore, that the distribution of tenderness in quality V reflected the variation in the
total beef sample, i.e., 25% very tender, 50% tender, and 25% less tender. Figure 1 was used to
explain the tenderness variation of the four qualities. The participants were also shown samples to
demonstrate that the various qualities did not differ in their visible attributes.
Several studies have used uniform fourth- or fifth-price auction to elicit WTP (e.g., Coursey,
Hovis, and Schulze, 1988; Umberger and Feuz, 2004). Compared with the frequently used second-
price auction the uniform fourth- (and fifth-) price auction has several benefits. First, a fourth-price
auction will have a smaller difference between the average participant’s valuation of the product and
the price. Therefore, a bid that differs from a participant’s WTP is more likely to have real economic
consequences. Second, with multiple winners it is not as exclusive to win the auction and any auction
winning utilities not associated with the product are reduced. Third, extreme outliers are less likely to
affect the price information that the participants receive during the experiment. As the second-price
sealed-bid auction, the uniform fourth-price sealed-bid auction is an incentive compatible mechanism
for eliciting WTP (Vickrey, 1961). Bidding your WTP is then a weakly dominant strategy. Bidding
more than your WTP gives a positive probability of winning when the price is higher than your
valuation of the product. Bidding less than your WTP decreases the probability of winning when the
price is less than the valuation (Shogren et al., 1994b).
We used a uniform fourth-price sealed-bid auction. In this type of auction, the participants
submit sealed bids, the fourth-highest bid is the price, and the participants bidding higher than that
price are the buyers. The participants were asked to bid simultaneously on the four qualities of beef.
We started by explaining to the participants the auction procedure. Next, we conducted two
hypothetical auction trials to familiarize the participants with the principles of the fourth-price sealed-
bid auction. In these hypothetical trials, no real purchases were made. The participants were told
explicitly that these trials were training trials that were to be followed by real trials. We then
conducted four trials of non-hypothetical fourth-price sealed-bid auctions. We used multiple trials to
make it possible to conduct within sample comparison of bids before and after introducing real
economic incentives, before and after tasting the different qualities, and to allow the participants to
refine their bids to more accurately reflect their own valuation. After each trial, the price for each
quality and the identification number of the winning bidders were written on a whiteboard. After the
four trials were completed, we randomly chose one binding trial and one binding quality. The
participants with higher bids than the price for the binding quality in the binding trial were the buyers.
After two hypothetical and two real trials of bidding, the participants were allowed to taste
samples of the different qualities. We informed the participants that we had randomly drawn one loin
from each quality and that these loins were not necessarily representative of the respective qualities,
but rather were examples on beef within the different qualities. We randomly drew one loin from each
quality, sliced the loins into steaks, and grilled the steaks on both sides until the core temperature
reached 60°C. Each steak was cut into small wedge shaped pieces, placed on white plastic dishes, and
labeled with one of the letters A, B, C, or V depending on the quality. The same four loins were used
in all four sessions. The loin drawn to represent V had a WB score categorizing it as very tender.3
Risk
The four qualities, A, B, C, and V, may be looked upon as lotteries, LA, LB, LC, and LV, where the
outcome of each lottery is the tenderness of the beef. By segmenting the beef into tenderness
categories, we have reduced but not eliminated the variability in tenderness for quality A, B, and C.
The participants assign subjective probabilities for various tenderness outcomes for each of the four
qualities. These subjective probabilities vary between the participants, and are based on the individual
participant’s experience and judgment. We have no information about the probabilities that the
participants assign to the various tenderness outcomes within each quality, and neither do we have any
information on their valuation on the various tenderness outcomes within each quality.
Each of the qualities A, B, and C may be viewed as a simple lottery, Lk, with k = A, B, C and
where the outcome of each lottery is the tenderness of the beef, t = 1,...,T. Each simple lottery is a