Impact of Changing Consumer Preferences

On Willingness-to-Pay for Beef Steaks

In Alternative Retail Packaging*

by

John D. SchmitzAssistant Professor

Department of Agricultural EconomicsUniversity of Wyoming

Dale J. MenkhausProfessor

Department of Agricultural EconomicsUniversity of Wyoming

Glen D. WhippleProfessor

Department of Agricultural EconomicsUniversity of Wyoming

Elizabeth HoffmanProfessor

Department of EconomicsUniversity of Arizona

Ray A. FieldProfessor

Department of Animal ScienceUniversity of Wyoming

Abstract and perceived changes in purchases of substitutemeats affect willingness-to-pay for beef rib-eye

The purpose of this study was to identify steaks in the traditional overwrapped styrofoamhow consumer perceptions of selected attributes of tray and vacuum skh packages. A laboratorybeef steaks, individual consumer demographics auction was used to obtain willingness-to-pay data.

-s research was partially fundedby the Beef Industry Council of the NationalLive Stock and Meat Board andthe Wyoming Beef Council, and is a contributionto the Western RegionalCooperativeResearch Project W-177,Domestic and InternationalMarketing Strategiesfor U.S. Beef. WyomingAgriculturalExperiment StationJournal Article No. JA 1696.

Journal of Food Distribution Research September93/page 23

The results suggest that health related factors,particularly the concern regarding cholesterol,reduced the willingness-to-pay for beef rib-eyesteaks, regardless of package type. For thevacuum skin package to be successful, informationabout the package is necessary, along with provid-ing a consistent and quality product, particularlywith respect to trim.

Key Words

Consumer attributes, Beef steaks, Willingness topay, Alternative packaging

Introduction

Changes in consumer preferences and theeffects of these changes on consumer choice arewell recognized by processors and marketers offood products. The vast array of differentiatedfood products available to consumers is evidence’of the reaction to these changes. Food productsare available which target both health consciousand convenience oriented consumers. While retailbeef products now have less external fat and lessbone, they have not changed appreciably toaccommodate recent changes in consumer prefer-ences. In comparison, methods of merchandisingchicken are entirely different than they weretwenty years ago (Reynolds, 1992).

Consumer concerns about convenience,food safety, diet, and health provide an oppor-tunity for the U.S. food industry to expand busi-ness (Harper, 1992). Convenience remains one ofthe top food issues for modern consumers, whileprice, packaging and freshness continue to have amajor impact on consumer food spending patterns(Barr, 1992). If the beef industry is going todirect efforts toward merchandising and productdevelopment, it is important to understand howindividual consumer characteristics and consumerperceptions of beef attributes affect the individualconsumer’s demand for beef.

The demand for beef has trended downwardduring the past decade (Purcell, 1989). Thedecline in demand can be attributed to changes inrelative prices between beef and other meats andchanges in the lifestyles of consumers, includinghealth and convenience related factors (Chavas,

1989; Barkemn and Drabenstott, 1990). Market-ing efforts of competing meats, particularly poul-try, like]y have contributed to the reduction inmarket share for beef (Allen and Pierson, 1986;Reynolds, 1992).

Centrally prepackaged or case-ready beefhas been considered by the beef industry as ameans to react to the challenge from competingmeats (Allen, 1989; Field and Merdchaus, 1990).Specifically, vacuum skin packaging (VSP), analternative to the traditional overwrapped styro-foam tray (OST) packaging, has potential forlarge-scale utilization in the beef packing industry.The VSP offers many attributes which may bepreferred by consumers including extended shelflife, leak-proof package, visibility of the meat,and longer and more convenient storage at homein the refrigerator or freezer (Pelzer et al., 1991).

The VSP has several advantages from amarketing perspective. A more consistent level ofmeat quality can be obtained due to the centraliza-tion of the packaging process. The VSP providesbetter opportunities for branded products. Addi-tionally, this package type facilitates the use ofinformation on the package, such as nutritionlables and cooking instructions, allowing retailersto emphasize consumer services (Linsen, 1988).

The unusual appearance of the meat is alimitation of VSP. Because the cut is not exposedto oxygen, the beef in this package is not thefamiliar bright red color but has a natural purplishcolor. However, when the meat is reexposed tothe oxygen contained in the air, the meat returnsto the familiar bright red color in about 15 min-utes. Information about the unusual color and itscause may be important to induce initial purchaseof beef in VSP (Lynch, Kastner and Kropf, 1986).Moreover, adoption of the VSP by the industryrequires, in part, information about consumerattitudes toward this package type.

Objectives

The objectives of this study are two-fold,The first objective is to determine how consumercharacteristics and perceived concerns about theattributes of beef affect consumer willingness-to-psy for beef steaks. This information is necessary

September931page24 Journal of Food Distribution Research

if specific consumer groups are to be targeted withpromotional information, and it also can guide thecontent of the promotional material. Additionally,it may provide information as to which attributesor perceived attributes of beef may be more dam-aging to retail demand and give direction to prod-uct development activities. The second objectiveis to determine to what extent an alternative pack-age type, namely the VSP, may enhance the desir-ability of retail beef to consumers. The results ofthe analysis should be useful in developing mer-chandising strategies and new products for beefand in developing marketing strategies forvacuum-packaged retail beef.

Study Design

The data used for the analysis were col-lected from two areas, Denver and Los Angeles,during May and August, 1989. A representativesample of consumers in each location wasrecruited by a marketing research firm. Studyparticipants were required to be primary purchas-ers of meat in their household and to have pur-chased beef within the month prior to the study.Respondents were paid $35 for participating in thestudy. The total number of observations includedin the analysis (after eliminating incomplete ques-tionnaires) was approximately 630.

Each individual participating in the studywas asked to complete a questionnaire. A sectionof this questionnaire surveyed respondents regard-ing their concerns or dissatisfactions with selectedcharacteristics of beef steaks. These characteris-tics represented a broad cross-section of concernsregarding convenience, health, and quality ofbeef. These questions identified attributes of beefas perceived by the respondent, such as highcholesterol content, high in calories, at-homestorage, minimize waste, and expense. Studyparticipants also were asked about their meatpurchase patterns in the three years prior to thestudy. Specifically, respondents were asked iftheir family’s usage of poultry, pork and fish hadchanged in the previous three years (less often, nochange, or more often). Demographic characteris-tics also were obtained from each respondent.

Willingness-to-pay data used in the analysiswere obtained via a laboratory experimental auc-

tion. The auctions were conducted after the par-ticipants completed the questiomaires. In a labora-tory experimental auction, participants submit bidsfor a product and the winner(s) of the auctionpurchase(s) the product at a price determined bythe auction rules (Coppinger, Smith and Titus,1980; Coursey and Smith, 1984; Cox, Robersonand Smith, 1982; Cox, Smith and Walker, 1985;Forsythe and Isaac, 1982; Kagel, Harstad andLevin, 1987). This procedure reveals how muchparticipants are willing to pay for the product. Amultiple unit Vickrey (1961) auction was selectedfor this study; specifically a fifth price auction.This type of auction is theoretically a demandrevealing allocation or price setting mechanism inthat bidding one’s true value is a dominant stra-tegy against any strategy used by other bidders(Cox, Roberson and Smith, 1982; Forsythe andIsaac, 1982; Vickrey, 1961). However, thebehavioral properties for revealing demand in aVickrey auction have been challenged (Courseyand Smith, 1984). These authors found thatsubjects consistent y underbid relative to their truereservation prices. This result suggests that thebids in the Vickrey auction should not be inter-preted as true representations of a subject’s maxi-mum willingness-to-pay. Thus, for purposes ofthis study, the bids are deemed to reflect a partici-pant’s behavior resulting from their attitudestoward beef and package type. Caution should beexercised in interpreting the actual magnitude ofthe bid in response to an attitude.

During the auction phase of the study,participants in groups of eight, were asked to bidfor the right to select steaks from a display case(48 auction sessions were conducted at each studysite--Denver and Los Angeles.) These steakswere choice, boneless, lip-off rib-eye steaks inOST and VS packages. Each package containedtwo eight-ounce steaks. Individuals participatingin the study were given several opportunitiesduring the auctions to closely examine the steaks.To assure that consumers were bidding for ahomogeneous product and to maintain the incen-tive properties of the auction, a representativesteak was chosen from the display case and usedin the auctions. Participants with winning bidswere allowed to select steaks from the displaycase atler the auctions were completed.

Journal of Food Distribution Research September93Jpage25

Participants were asked to take part in 10auctions. After the instructions which explainedthe specifics of the auction were read to the par-ticipants, two trial auctions for each package typewere conducted. These were designed to acquaintparticipants with the auction procedures. Follow-ing these four trial auctions, six purchase auctionswere conducted, three for each package type.After bids were submitted by each of the eightparticipants involved in a particular auction, thebids were ranked privately, The auction monitorreported only the fifth-highest bid price to theparticipants. The “winners” were those whosubmitted the top four bids. In the purchaseauctions, each winner paid the fifth-highest bid ofthe auction they won for the package of steaks.

All information other than the market pricewas kept private, and participants were instructedto count themselves as winners if they had bidmore than the reported price. Ties were decidedby the roll of a die and any discrepancies betweenmonitor records and participant records weredeferred until the end of the experiment. Duringthe trial auctions, the monitor checked each partic-ipant’s record sheet to make sure he/she under-stood how to determine when his/her bid was awinning bid and what was to be paid for a pack-age of steaks. At the conclusion of the series ofauctions, participants paid for steaks bought in anyof the purchase auctions. The individuals werethen given slips of paper which entitled them topick up steaks they had purchased from the dis-play they had viewed prior to the auction,

The auction was conducted under threeinformation treatments with respect to VSP: noinformation, verbal information, and informationplus a demonstration. Under the information andinformation plus a demonstration treatments, thefollowing benefits and characteristics of VSP werepresented to the study participants,

. What is good about this package?

- You can see both sides of the meat,- It preserves freshness while allowing the

natural aging process to continue.- It is tamperproof, assuring freshness.- It extends refrigerator shelf-life, meat is

good for 30 days after packaging,

The vacuum skin makes an excellent freezerpackage and is microwave defrostable.

● Why is the meat’s color’different?

- The meat is vacuum sealed immediately aftercutting, preserving the rich natural burgundycolor of quality beef.

- Once the package is opened and the meat isexposed to oxygen, the meat will bloom thefamiliar bright red color.

. Why is the juice in the package?

- This juice is naturally present in fresh beefand its presence is an indication of freshness.The amount of juice present varies horn cutto cut and is dependent on temperatures.The colder the meat is kept, the less juicethere will be.

In the demonstration treatment, the VSPwas opened and participants observed the beefsteak change color when exposed to oxygen inthe air.

Model

The analytical model relates the partici-pant’s willingness-to-pay for steaks to a set offactors which are categorized as selected beefsteak concerns; individual demographic character-istics; perceived changes in usage of poultry, porkand fish; and the type information given to thestudy participants about the VSP. Willingness-to-psy was measured by the simple average of eachindividual’s three bids for each package type.Bids for each package type were analyzed sepa-rately so as to determine the impact of the factorsidentified above on each package independently,as well as to determine whether the impact ofthese factors display significant differencesdepending on the package type.

The model takes into account that consum-ers make decisions based on perceptions of prod-uct characteristics and that preferences varyamong consumers due to demographic characteris-tics. Each of these considerations is importantfrom the standpoint of marketing, e.g., market

September931page26 Journal of Food Distribution Research

segmentation, product merchandising and newproduct development and in developing marketingstrategies in general.

The theoretical framework used in thisstudy follows that summarized by Capps andSchmitz (1991). The utility function, originallyspecified by Basmann (1956) in his work on vari-able preferences, is expressed as

u, = U(qt; O(cYJ)

where utility (UJ is dependent on the commodityvector (q~ and the perception of quality (O(q)) ofa good by a consumer in time t. Maximization ofU, with respect to q,, given cr,, subject to anincome constraint yields the demand relationship

where p is a vector of prices and y is income. Inthis study, CXtis defined as a vector of perceivedcharacteristics or attributes of a good in q!. Con-sideration of perceived attributes in the empiricalanalysis is important because nutritional elements,for example, may be perceived to be differentthan actual levels (Schmitz and Nayga, 1991).

This basic theoretical formulation can beexpanded to include the k demographic character-istics of individual i (&J, to account for variationin preference due to these factors. Given the typeof data available and circumstances relevant to thestudy design and investigation, the basic theoreti-cal model can be modified further. For example,a cross-sectional survey, while appropriate forobtaining information regarding perceived productcharacteristics, generally will not yield informa-tion on quantity consumed of a good. Thus, inthis study, willingness-to-pay is used to representpreference and behavior in the market by theindividual consumer. Moreover, in the case ofcross-sectional survey data, prices, e.g., prices ofcompeting products, are assumed to be constantacross individuals. Perceived changes inpurchases of competing products are included inthe model to account for changes in willingness-to-pay for beef steaks due to effects associatedwith competing meat products for beef--poultry,pork and fish. Finally, the model includes vari-ables which pellain to the specific manner in

which the study was designed and conducted-Iocation, information treatment and interactionsbetween selected product attributes and the infor-mation which was presented to the study partici-pants.

The specific variables used in the model areidentified and described in Table 1. Perceivedattributes of beef steaks are categorized intohealth, convenience, appeal and merchandisingcharacteristics. These variables measure a respon-dent’s concern or dissatisfaction (O - no concern;1- concern) with beef steaks. Expected relation-ships also are identified in Table 1. Some ofthese expected relationships are identified recog-nizing the properties of the study product-welltrimmed rib-eye steak, e.g., trim level; somerelationships cannot be determined a priori.Means for the variables also are reported in Table1.

Some of the study participants bid zero ineach of the three auctions for a given packagetype. These participants represent approximate] yfive percent of the total number of observationsfor each package type. This bidding strategyprovides for a censoring of the dependent variableat the zero level. Censoring of the dependentvariable can lead to biased and inconsistent ordi-nary least squares (OLS) estimates (Krnenta,1985). Heckman (1979) proposes a two-stepprocedure for estimating such models. The firststep of this procedure uses a probit model todetermine the probability that a specific observa-tion will be non-censored. This probability, theratio of the standard normal density functiondivided by the cumulative standard normal densityfunction, is then included as a factor in a secondstage OLS estimation. This ratio is the inverse ofwhat is commonly referred to as the Mills ratio.The second stage estimation omits all of the cen-sored observations. Heteroskedaticity, a correla-tion between the magnitudes of the error term andthe explanatory variables in the model, is inherentto the Heckman procedure. A variety of tests forthis correlation were conducted using the diagnos-tic procedures provided in SHAZAM, Version 6.2These tests indicated that heterskedaticity waspresent in the data. Therefore, the second stageestimates were calculated using Weighted LeastSquares (WLS). Tests and weights were based on

Journal of Food Distribution Research September931page27

iII

iI1

1

Go“

.11 0 iiI

iII

!

September 93/page 28 Journal of Food Distribution Research

●

0 o“ 0“

1

I

so“

Journal of Food Distribution Research

i

a)

7EI

0

“r ow

o

+

+

00

0

4

0

I

I

September 93fpage 29

the Breusch-Pagan test (see Kmenta, 1985, p.294-295).

Outlying data observations are a concernusing data such as that used in this study whenleast-squares estimation techniques are employed.In the case of an outlying observation, one obser-vation (one respondent in this case) may exertundue influence on the overall model results. Thepresence of such observations can be detected bythe use of influence diagnostics (Belsley, Kuh andWelsch, 1980). Observations which are identifiedas exerting undue influence were removed fromthe estimation procedure, These data representabout five percent of all non-zero observations.

One important goal of this study is to deter-mine whether package type statistically affects themagnitude of the estimated model coeftlcients. Inorder to test whether these magnitudes areaffected, it is necessary to isolate coeftlcients forboth the OST and the VSP within one modelformulation. This is accomplished through theuse of an intercept shifter and slope shifters foreach variable in the model. An intercept shifter isaccomplished by the inclusion of a dummy vari-able for the package type (O for OST and 1 forVSP). Slope shifters are developed by multiply-ing the dummy variable by each of the othervariables in the model. This multiplication resultsin new variables which equal zero for the OSTand equal the value of the factor for the VSP,

Results from such a model indicate themagnitudes of the coefficients for both packagetypes as well as provide a test of significantchange across package type, Coefficient estimatesfor the intercept and the traditional variablesprovide the coefficients for the OST directly,Coefficient estimates for the VSP are obtained bythe summation of the variable (intercept) and itsrespective slope (intercept) shifter. T-tests for theVSP are conducted using standard approaches forthe summation of two coefficients. Coefficientsobtained by this method are the same as would beobtained if both models were estimated indepen-dently. However, the intercept and slope shiftervariables provide a direct test of whether thecoefficient is significantly different across packagetypes. Significant t-values on these shifters indi-cate that a significant change in the coetllcient

occurred. Probability (P) values associated withthese t-test are used to determine the probabilitythat the coeftlcients are different.

Results

The results of the analysis of the influenceof the combined variables in each of the categoriesare presented in Table 2. The results suggest thecombined model variables explained a significantamount of the variation in willingness-to-pay foreach of the package types, Combined health,appeal and demographic variables (excludingincome) significantly influenced bids for rib-eyesteaks in both the OST and VSP. Informationpertaining to the VSP significantly affected thebids for steaks in both package types. None ofthe cross-effects significantly influenced the bidprices for beef in either of the package types. Inaddition to the health, appeal, demographic andinformation variables, the bids for beef steaks inthe VSP were significantly influenced by thecombined income and competing meats variables.Package type significantly affected the influence ofthe combined health and information variables onwillingness-to-pay.

The estimated model is presented in Table3. In general, health related concerns negativelyaffected the willingness-to-pay for rib-eye steaksin both the OST and VSP. For both packagetypes, the cholesterol concern significantlyreduced the average bid price. Individuals identi-fying trim as a concern bid significantly more forbeef in the OST, but bid significantly less forsteaks in the VS package. Under the informationtreatment, a concern with respect to microwave-ability significantly increased the willingness-to-psy for both package types, while a concern aboutthe inconvenience of at-home storage of beefsignificantly reduced the bid for OST packagedbeef. Among the appeal variables, the wasteconcern was significant in reducing the bid pricefor the OST and the variable not tender signifi-cantly decreased the bid for steaks in VSP.Respondents who express a concern that the pack-age wrapping does not protect the meat bid signif-icantly more for the VS packaged product. Otherappeal and merchandising attributes of beefsteaksdid not significantly influence the bids for beefsteaks in the OST.

September931page30 Journal of Food Distribution Research

Table 2. Results of Combined Effects of Variables on Willingness-to-Pay by Package Type, F-Statistics.

Variable Category Overwrapped Vacuum-Skin Change

All Coeftlcients 6.375 “ 8.652 ● 1.610 ●

Health 4.340 ● 3.182 ● 2.760 “

Appeal 2.062 ● 2.724 “ 0.632

Merchandising 1.456 1.411 0.928

Demographic 4.536 * 4.042 ● 0.641

Income 0.857 4.252 ● 1.875

Competing meats 1,161 2.395 ● 0.827

Inform + color 3.146 ● 8.176 ● 3,411 ●---------------------- -—------- ---------- --------- ------------------- ------------ ------------

Cross-effects---------------- -------— ---------------- -- .---— ----------------- ------- ----------------------

Micro x inform 1.489 1.586 0?015

Store x inform 1.482 1.404 0.301

Color x inform 1,681 1.757 0.986

“ Significance level a = 0,10,

Demographic variables played a key role inexplaining variation in the willingness-to-pay forsteaks in each of the package types. Age andincome significantly increased the bids in bothpackages. In general, education was significant indecreasing the bids for beef in the OST, but didnot significantly affect the VSP bids, Individualsreporting increased purchases of poultry or fishbid significantly less for steaks in the VSP.

Information pertaining to the VSP, and beefin the VSP, as well as information plus thedemonstration of opening the VSP, significantlyincreased the willingness-to-pay for steaks in bothpackage types. However, there was no significantdifference between the coefficients associated withthe information and information plus demonstra-tion treatments in the VSP equation. This maysuggest the demonstration is not necessary.

Individuals who expressed a concern aboutthe color of beef steaks bid significantly more forthe beef in the VSP. This result is at odds withthe belief by some in the industry that color is amajor deterrent for acceptance of the VSP. There

Journal of Food Distribution Research

are two possible explanations for this result; bothare related to the way the study was conducted.First, it is possible that consumers think colorindicates freshness. Before the auction, in orderto assure participants that the product was nottainted, researchers told the respondents that thesteaks were purchased from a local distributor thatday and were similar to those that could be pur-chased at a local retail outlet. This introductionmight have persuaded participants that the productwas wholesome and fresh, regardless of its colorin the VSP. Second, the question about the con-cern or dissatisfaction with color simply listed“color” and respondents were asked to identify ifcolor was a concern or not a concern. A moredetailed question or prompting with respect to thecolor issue may have been more appropriate.

The location variable was positive andsignificant in each of the two equations. Specifi-cally, average bids were significantly higher in theLos Angeles site for both the OST and VSP, ascompared to Denver. Differences in the generalprice level between the two locations may havecontributed to bid differences by location.

September93/page 31

Table 3. Estimated Parameters and Summary Statistics, Willingness-to-Pay for the OST andVSP

Overwrapped Vacuum-Skin Probability ofEqual Coeffl-

Variable Coeftlcient T-Value Coefficient T-Value cientsl,JLL

0.014-0.254 ●

-0.015-0.0820.194 *0.0370.330 ●

-0.002-0.229-0.393 ●

0.002-0.046-0.201 ●

0.0390.0440.1040,0750.0900.013 ●

0.039-0.443 ●

-0.143-0.251 ●

-0.240 ●

0.0940.1290.312 ●

-0.0690,067

-0.0960.424 ●

0.134 ●

0.231 *-0.1510,091

-0.9170.660 *

4.YY

0.15-3,13-0.19-1.192.590.232.21

-0.02-1.25-1.910.09

-0.60-2.580.530.620.900.951.033,880.49

-2.01-1.12-2.17-1.640.820.801,60

-0.880.81

-1.243.871.291.40

-1,030.71

-1.029.11

Intercept 1 cl- 1.096 “ 0.153ToolargCholCalArtingTrimngMicro 1Micro2Micro3StorelStore2Store3GristWasteNtenderCseeperBdisluPackTooexpAgeGenderEduc2Educ3Educ4Educ5Inc2Inc3Inc4PltyppPkppFhppInform2Inform3RedcollRedco12Redco13MillLoc

-0.030-0.122 *-0.039-0.028-0.180 ●

0.2070.557 ●

0.184-0,223-0.1110.160

-0.044-0.060-0.139 “-0,002-0.0910.190 ●

0.0760,015 *0.093

-0.096-0.018-0.018-0.0410.341 ●

0.621 ●

0.722 *-0.243 ●

0.021-0.102 ●

0.857 ●

0.702 ●

0.419 *0,113

-0.044-0.0520.746 ●

4.Ub

-0.29-1.50-0.45-0.35-2.301.043.621.22

-1.28-0.670.89

-0.49-0.70-1.81-0.02-0.752.080.864.301.12

-0.40-0.14-0.15-0.313.304.562.78

-2.590.23

-1.297.275,942.250.72

-0.28-0.079.65

0.3760.1260.4190.3030.001 ●

0.2540.1450.1840.4900.1440.2870.4940.1120.047 “0.3290.1230.1700.4530.3490.3210.1440.2460.053 ●

0.1580.056 ●

0.010 ●

0.1040,078 ●

0.3560.4800.004 ●

0.001 *0.2260.1100.2550.2280.209---- ------.- -— ----- --.----- --—---- ------- —------ -------- ------- ------- ------- -

R-~qr 0.284 0.404 iM66-57--Adj. R-Sqr 0.240 0.368 0.326 ~1Observa- 630 635 1265 ~1tions

~/ Probability the coefficients between the two equations are the same.~/ Combined model.

● Significance level a = 0.10.

September931page32 Journal of Food Distribution Research

Coeftlcients which exhibited magnitudessignificantly different between the two equatignswere associated with the trim, tenderness, educa-tion, income, perceived changes in purchases ofpoultry and information variables. Informationregarding the VSP significantly increased the bidsfor the VSP, as compared to the OST, relative tothe no information treatment. The response inbids for the two package types was significantlylower for the VSP, as compared to the OST, byindividuals who expressed a concern with respectto trim and tenderness. The difference in thecoeftlcients associated with trim may be due to theincreased visibility of the product in the VSP.Since study participants were given the oppor-tunity to view and handle the packages before theauction, it is possible the perception of the overallquality of beef in the VSP, as compared to beef inthe OST, contributed to the perception that thebeef in the VSP may have been less tender thanthat in the OST. Thus, it is possible the tender-ness property of beef steaks may be realizedthrough how the consumer perceived the overallquality of the product. Finally, respondentsreporting increased purchases of poultry bid sig-nificantly less for steaks in the VSP, as comparedto steaks in the OST.

Summary and Implications

The results of this study suggest healthrelated factors, particular y the concern regardingcholesterol, reduced the willingness-to-pay forbeef rib-eye steaks, regardless of package type.Overall, these results lend support for the recentlyinitiated Beef Industry Council promotion cam-paign which features endorsements from healthprofessionals.

The development of new beef productscapitalizing on the desirable attributes of compet-ing meats is important. Reducing waste andproviding packaging which better protects theproduct and also has convenient storage propertiesappear to be warranted.

For the VSP to be successful, informationabout the package and the color of the meat in thepackage is necessary, while a demonstration ofbeef returning to a bright red color after being

reexposed to oxygen may not be needed. Theconcern about the color of retail beef may beassociated with a consumer’s perception of thefreshness of the product. Moreover, given thevisibility of the beef to the consumer in the VSP,it is important to provide a consistent and qualityproduct, particularly with regard to trim.

Several of the demographic variables weresignificant in explaining the bids, suggesting thepotential for market segmentation merchandisingstrategies by the industry. For example, age andincome contributed to increased bids for bothpackage types. The VSP appears to be morefavorably accepted among individuals in higherincome categories, as compared to the OST.

Finally, this study was conducted only forbeef rib-eye steaks. It is conceivable that per-ceived characteristics of other beef cuts affectconsumer perception of quality, and thus prefer-ence, in differing manners. This may be especi-al]y true for cuts which have more purge accumu-late in the package than do rib-eye steaks. More-over, demographic characteristics of consumersalso may influence the perceptions of, and prefer-ence, for different cuts of beef differently. Thus,a better understanding of how the factors exam-ined in this study affect the individual consumerpreferences for alternative cuts of beef may pro-vide useful marketing information for the beefindustry.

Three caveats are relevant with respect tothe interpretation of the findings of this study.First, the results may have been affected some-what by university personnel providing informa-tion about the beef and packages rather than aretailer or processor. Participants were told thebeef in both package types was fresh and pur-chased from a local distributor. Respondents hadno reason to doubt the quality and safety of theproduct and may have submitted bids during theauction to reflect this trust. For example, theresult that the demonstration of opening the VSPdid not significantly increase the value of the beefin this package type, as compared to the informa-tion treatment, may have been influenced byuniversity employees providing this information asopposed to employees of private industry. Giventhe manner in which the study was conducted,

Journal of Food Distribution Research September931page33

there was no way to determine the influence ofthis probable effect.

Second, the relative values of the estimatedintercepts for the package type equations do notprovide an indication of the relative bids for thetwo package types, Average bids were higher forthe VSP. However, bids for steaks in the VSPand the OST under the no information treatmentwere not significant y different. There is a poten-tial for the generally higher average bids forsteaks in the VSP, relative to the beef in the OST,to have been due, in part, to the novelty of theVSP. The analysis can neither determine if thiseffect is present or its magnitude.

Third, the study product had very littlepurge. This was because the beef was packagedshortly before the study and because rib-eye steakshave less purge than many other cuts. Purge inthe package accumulates with time associated with’transportation, storage, marketing, and peoplehandling packages in retail counters. The VSP inretail stores often has more purge than was pres-ent in the study product.

References

Allen, J.W. and T.R. Pierson. 1986. “PackagingHas Rising Role in Red Meat’s Turna-round, ” l%e National Provisioner,December 13:7-25.

Allen, J. W, 1989, “Centrally Prepacked FreshRed Meat: Success or Failure?” 7heNational Provisioner, August 5:8-10.

Barkemn, A.D. and M. Drabenstott, 1990. “ACrossroads for the Cattle Industry, ” Eco-nomic Review, Federal Reserve Bank ofKansas City, Nov/Dec:47-66.

Barr, A. 1992. Director of Good HousekeepingInstitute, Seminar statement reported in,Lean Trimmings, Western States MeatAssociation Newsletter, June 15.

Basmann, R.L. 1956. “A Theory of Demandwith Variable Consumer Preferences, ”Econometrics, 24:47-58.

Belsley, D. A., E. Kuh, and R.E. Welsch. 1980.Regression Diagnostics, Identifying Inj7u-ential Data and Sources of Collineari~,Wiley.

Capps, O. Jr. and J.D. Schmitz. 1991. “A Recog-nition of Health and Nutritional Factors inDemand Analysis,” Western Journal ofAgricultural Economics, 16:21-35.

Chavas, J .P. 1989. “On the Structure of Demandfor Meat, ” in I%e Economic of MeatDemand, R.C, Buse, Ed., University ofWisconsin, Madison. pp. 125-135.

Coppinger, V., V. Smith, and J. Titus. 1980.“Incentives and Behavior in English, Dutchand Sealed-Bid Auctions, ” Economiclngui~, 18:1-22.

Coursey, D.L. and V.L. Smith. 1984. “Experi-mental Tests of an Allocation Mechanismfor Public, Private, or Externality Goods,”Scandinavian Journal of Economics,86(4):468-484.

Cox, J. C., B. Roberson, and V.L. Smith. 1982.“Theory and Behavior of Single ObjectAuctions,” Research in Experimental Eco-nomics, JAI Press Inc., 2:1-43.

Cox, J. C., V.L. Smith, and J.M. Walker. 1985.“Expected Revenue in Discriminative andUniform Price Seal-Bid Auctions,” inResearch in ExperimentalEconomics, V.L.Smith, Ed., JAI Press Inc., 3:183-232.

Field, R. and D. Menkhaus. 1990. “ConsumersWill Buy Vacuum-Packed Beef,” i%eNational Provisioner, June 11:16-18.

Forsythe, R. and R.M. Issac. 1982. “Demand-Revealing Mechanisms for Private FoodAuctions, ” in Research in ExperimentalEconomics, V.L. Smith, Ed., JAI PressInc., 2:45-61.

Harper, C, 1992, ConAgra Chairman and CEO,Seminar statement reported in, Lean Trim-mings, Western States Meat AssociationNewsletter, June 15.

September931page34 Journal of Food Distribution Research

Heckman, J.J. 1979. “Sample Selection Bias as aSpecification Error, ” Econometrics,47:153-61. ,

Kagel, J. H., R.M. Harstad, and D. Lewis. 1987.“Information Impact and Allocation Rules -Auctions with A-fflliatedLaboratory Study, ”55(6):1275-1304,

Kmenta, J, 1985. Elements

Private Values: AEconometrics,

of Econometrics,Macmillan Publ., 2nd Edition.

Linsen, M.A. 1988. “Meat-ing the Challenge, ”Progressive Grocer, May:235.

Lynch, N, M., C.L. Kastner, and D,H. Kropf.1986, “Consumer Acceptance of VacuumPackaged Ground Beef as Influenced byProduct Color and Education Material, ”Journal of Food Science, 51:253-255.

Pelzer, P. M.L,, D.J. Menkhaus, G.D, Whipple,R.A. Field, and S.W. Moore. 1991, “Fac-tors Influencing Consumer Rankings ofAlternative Retail Beef Packaging,” Agri-business: An InternationalJournal, 7:253-

PurceH, W.D. 1989. “The Case of BeefDemand, Choices, 2:16-19.

Reynolds, R. 1992. “Beef or Chicken: TheWhats and Whys of the Consumer Prefer-ence Shift, Lean Trimmings,Western StatesMeat Association Newsletter, June 1.

Schmitz, J.D. and R,M, Nayga, Jr. 1991, “FoodNutritional Quality: A Pilot Study on Con-sumer Awareness, ” Journal of Food Distri-bution Research, 22:19-33.

Vickrey, W. 1961. “Counterspeculation, Auc-tions, and Competitive Sealed Tenders, ”Journal of Finance, 16:8-17.

267.

Journal of Food Distribution Research September 931page 35

September 93/page 36 Journal of Food Distribution Research