Consumer Responses to Unit Pricing, Open Dating, and Nutrient Labeling

Citation:

Monroe Peter Friedman (1972) ,"Consumer Responses to Unit Pricing, Open Dating, and Nutrient Labeling", in SV - Proceedings of the Third Annual Conference of the Association for Consumer Research, eds. M. Venkatesan, Chicago, IL : Association for Consumer Research, Pages: 361-369.

Proceedings of the Third Annual Conference of the Association for Consumer Research, 1972      Pages 361-369

CONSUMER RESPONSES TO UNIT PRICING, OPEN DATING, AND NUTRIENT LABELING

Monroe Peter Friedman, Eastern Michigan University

[The author is professor of psychology and director of the Center for the Study of Contemporary Issues at Eastern Michigan University.]

This paper reviews recent research relating to three new informational displays which have been proposed to assist consumers with their shopping decisions. These innovations in labeling practices are unit pricing, open dating, and nutrient labeling. The literature reviews will be followed by a discussion of the relevance of the research findings to public policy decisions relating to the adoption of these proposed aids.

UNIT PRICING

A number of terms (unit pricing, price-per-measure, dual pricing, and value pricing) have recently been employed to describe the practice of providing price information to consumers by such standard measures as the ounce, pound, pint, and quart. Unit pricing, as the practice is now widely called, is not a new development, having been used by supermarkets for many years to display price information for such store-packaged, variable-size products as meats and cheeses. What is new however is the effort to extend this practice to many thousands of Pre-Packaged retail Products.

Unit-pricing advocates point to a host of packaging practices (packaging-to-price, cents-off labels, slack fill, and fractional quantity units) which they feel impair the ability of the consumer to compare prices effectively. They further contend that the tremendous increase in items carried by supermarkets (from about 1,000 in 1930 to 8,000 in 1970) has exacerbated the price-comparison problems of the consumer to an unacceptable level. These advocates believe that unit Pricing will alleviate these problems by shifting the major share of the burden of price comParisons from the consumer to the retailer.

Research on unit pricing has been primarily concerned with three questions:

1. Is there a need for unit-price information?

2. How do consumers react to unit-pricing programs?

3. What are the costs of such programs?

Let us examine these questions in turn.

The Need for Unit-price Information

Studies addressed to this question have examined the ability of consumers to make price comparisons without the aid of unit-price labels. Four such studies have been reported (Friedman, 1966; Friedman, in press; Gatewood and Perloff, in press; and Houston, 1972), each of which instructed subjects to perform hypothetical price-comparison problems in a laboratory or field setting. All four studies revealed significant departures from perfect performance. In three of the four (Friedman, in press; Gatewood and Perloff, in press; and Houston, 1972), a second experimental condition was introduced to examine the value of unit-price information in solving the problems; all three found a substantial improvement in performance under the unit-price condition.

Taken together, these several studies reveal that without additional information, consumers are often unable to make correct price comparisons. The findings suggest that the use of unit-price labels may substantially reduce purchase errors for economy-minded consumers at a savings in time (Gatewood and Perloff, in press) and money (Friedman, in press; and Houston, 1972).

Consumer Reaction to Unit-pricing Programs

Monroe and LaPlaca (1972) have recently reviewed studies of consumer reaction to unit-pricing programs which were introduced in five food chains. Except for the common finding that reported usage of unit-price labels was correlated positively with education, the results of the five studies yield a most inconsistent picture. For example, two of the chains detected no shift in aggregate warehouse movements to products with lower unit prices, while a third found increased sales for the usually lower-priced private label brands. This result is in conflict with consumer survey responses in four of the chains which revealed changes in brands and/or sizes purchased by users of unit-price information. Reported usage of unit-price labels also varied considerably for the respondents in the five chains, with a high of 65 percent of the respondents in one chain and a low of 7 percent in another.

A number of methodological differences among the several studies may well be responsible for the conflicting findings. Likely candidates here are the unit of measure used to express unit-price information, the manner in which the information was displayed, the orientation program used to acquaint store customers with unit pricing, the proportion of store products which were unit-priced, and the time interval which elapsed before customer surveys were conducted.

Costs of Unit-pricing Programs

Several attempts have been made to measure the retailer costs of installing and operating unit-pricing programs. After reviewing six such studies, Monroe and LaPlaca (1972) conclude that the costs remain relatively constant per store, regardless of store size and sales volume. Thus small independent stores with low sales volume must expend a relatively high percentage of their annual sales to support a unit-pricing program. The estimates calculated in one study (McCullough and Padberg, 1971) of the direct costs of unit pricing as a percentage of total sales ranged from 4.15 percent for low-volume stores (annual sales under $100,000) in a small distribution network (20 stores), to .095 percent for high-volume stores (annual sales over $2,0()0,000) in a large distribution network (90 stores). It is important to note that these are direct costs, and do not consider the procedural efficiencies which may well result for retailers as a byproduct of instituting unit-pricing programs. In particular, three chains which have adopted programs report that they provide "tighter inventory control, better space management, and fewer price-marking errors." (Monroe and LaPlaca, 1972, p. 22).

OPEN DATING

Open dating refers to the practice of printing dates on packaged food products for the purpose of informing consumers about the freshness of the products. Advocates of the practice contend that consumers have a right to know whether the packaged food which they purchase is fresh. They further contend that open dating provides the necessary information to make this determination and thus should be adopted by the food industry to assist consumers with their shopping decisions. That the efforts of these advocates have received substantial public support is evident from the findings of a recent consumer survey study which was reported at the 1971 meeting of the National Association of Food Chains (Ringler and Berner, 1971). 89 percent of an enormous nationwide sample of 250,000 shoppers reported that they were in favor of easy-to-understand date codes to provide assurance of product freshness.

The food industry response to open-dating proposals has focused on a number of technical arguments which question the usefulness of dates as indicators of product freshness. These arguments are based, in large part, on the results of a recently completed two-volume study of food stability which was conducted by the Department of Food Science at Rutgers University (Food Stability Survey, 1971). The Rutgers' report presents extensive findings and recommendations as well as specific data for 18 consumer products. Perhaps its most significant general finding is that for most foods, temperature is a far more important determinant of product quality than elapsed time from the date of packaging. For example, the report finds that a frozen food product may deteriorate as much in two months at 10 F as it will in twelve months at 0 F. Thus freshness dates based simply on temporal considerations are not likely to be accurate unless temperature is carefully controlled at all stages of the food distribution cycle.

Advocates of open dating acknowledge the significance of possible deviations from proper temperature controls and handling procedures. They insist however that these problems can be overcome and point to the current widespread use of open dating by American food chains as evidence of its feasibility.

Among the many questions raised by the open-dating issue is the type of date information to be presented in open-dating programs. Some food processors have argued for a pack date since not only is it a known date, but it could be relatively easily provided at the time of packaging. Some retailers, on the other hand, have expressed a preference for a pull date, so that they would know the last acceptable date of sale. Moreover, some consumer advocates have suggested that a quality assurance date be adopted so that consumers would know the last date a product could be used at peak quality. Other consumer advocates have recommended an expiration date which would inform consumers of the last date on which safe usage of a product can be expected.

Of the more than 60 food chains which have recently adopted open dating, almost all have opted for pull dates. This choice recognizes the fact that proper temperature controls and handling procedures are more likely to be assured in the food processor-to-retailer stages of the food distribution cycle than in its later retailer-to-home-consumption stages. The choice also recognizes the value to the retailer of having simple last-day-of sale indicators to facilitate the removal of non-fresh products from store shelves and to help store clerks with stock rotation. Many retailers also see pull-date benefits for the consumer, believing, as they do, that this indicator of product freshness offers her the best combination of useable and reliable information.

Research on Open Dating

The economic costs of installing and operating an open-dating program have been examined by the U.S. Department of Agriculture and a study report is expected to be released in the near future. Other studies have looked at consumer awareness, understanding, and usage of open-dating information. Zehner (1971), in a survey study of 1,800 Michigan consumers, found that 80 percent of the participants had heard about open dating even though, at the time of the study, only a few Michigan food chains were operating open-dating programs. Zehner also showed future-dated packages to her respondents and asked them what was meant by the dates. The largest proportion (35%) interpreted them to be expiration dates while the second largest (26%) thought they were pull dates. When asked what they would like open dates to represent, 27 percent expressed a preference for expiration dates, 22 percent for a guide to freshness, 19 percent for pack dates, and 12 percent for pull dates. It is especially interesting to note that the least preferred indicator (pull dates) is the one most widely used by food retailers.

In a second study on open-dating, Taylor (1971) interviewed 1,700 female shoppers at 18 stores of a Chicago food chain which had practiced open-dating for several months by printing pull dates on some 150 private brand items. Slightly more than half of the respondents reported that they were aware of the dating information. Awareness was higher in middle-income areas (575) than in low- and high-income areas (each about 50%).

About half (429) of the shoppers who had indicated awareness of the pull dates agreed to be interviewed in greater depth. Almost two-thirds (63%) of these shoppers reported they had used the dating information at least once. However one wonders about the success of their efforts in light of the finding that only 20 percent of the 429 were able to correctly interpret the pull date as the last day a product can be sold. Finally, when asked to cite advantages and disadvantages of the open-dating program, about half of the user group (54%) reported that "product freshness" was the principal advantage and 80 percent reported no disadvantages.

A third investigation of open-dating is being conducted by the Consumer Research Institute and a final report is expected to be published shortly. The report will present findings on a series of three studies. The first of these studies compares consumer reactions to pull dates and pack dates, each of which was introduced in three or more stores of a national food chain. The second study uses national mail and telephone surveys to examine consumer reactions to food quality and freshness. The third study in the series will provide before-and-after data relating to new open-dating ordinances which have recently been passed in New York City and Dade County, Florida. An effort will be made to determine the effects of the legislation on consumer attitudes and usage patterns.

To sum up, the few behavioral studies which have been conducted on open-dating raise more questions than they answer. Apparently, open-dating information is highly desired by American consumers (indeed, according to the Ringler and Berner study, more 90 than a host of other supermarket practices such as discounting, nutritional advice, and unit pricing). Yet expiration dates, which are most preferred by consumers, according to the Zehner study and some preliminary findings of a Consumer Research Institute study (Hoofnagle and Stokes, 1971), would appear to be the most difficult form of open-dating information for the food industry to provide. Other questions concern the confusion relating to the meaning of pull dates by the respondents in Taylor's study and, in light of this confusion, the relatively few disadvantages of the program which were reported by its users. [What may be operating here is something akin to what would seem to be the typical voter's attitude toward regulations requiring public disclosure of campaign contributions by political candidates. Few voters are likely to take the time to make sense of these disclosures but they may nevertheless favor such regulations knowing that the possibility of this information being used by even a small segment of the electorate may well serve to dissuade politicians from accepting unduly large contributions from single sources. If the reader will excuse a bad pun, the disclosure of information, even infrequently used information, may be favored by the public to assure that packaged foods can be safely bought and that political candidates cannot!] A final unresolved question is the likelihood of selective consumer purchasing of late-dated food products. If, as has been suggested, many consumers would sort through the various packages on sale in supermarkets in an effort to buy the very freshest, many perfectly good food products would spoil on the store shelves, with a resultant economic loss to the retailer and the consumer. Although the Rutgers study states that the U.S. and foreign experiences with open-dating has found few instances of such selective purchasing, the authors cite no evidence to support this statement.

NUTRIENT LABELING

Prior to the White House Conference on Food, Nutrition, and Health, which was held in Washington in December of 1969, only modest efforts had been made to evaluate nutrient labeling as an aid to purchasers of packaged foods. The Conference, which was called "a watershed in American social history" by its chairman Jean Mayer (1972), in a report to the scientific community, broke from the virtual inactivity of the past by recommending that the Food and Drug Administration (FDA) consider the development of a system for identifying the nutritional qualities of packaged foods. The Conference further recommended that the food industry be encouraged to provide nutritional information on package labels to enable consumers to follow recommended dietary regimens.

These Conference recommendations reflect the increasing number of processed and formulated food products which have recently appeared in the marketplace, most of which are not easily classifiable into standard nutritional categories. As Mayer has stated:

...Consumers regard labels as their primary source of information about the contents, safety, and nutritional value of the food they buy. Given our existing nutrition education programs, it is not difficult for the average citizen to select simple meats, vegetables, and fruits, and produce a balanced diet. But what if the choice involves such new foods as frozen pizza or spinach souffle? How should these products be classified, and how should they be labeled? (1972, p. 240)

The FDA responded to the Conference recommendation on nutrient labeling by proposing criteria for information panels on food packages in the Federal Register of March 30, 1972. These criteria specify the nutrients to be listed on food packages, the units by which they are to be quantified, and the permissible locations on food packages for presenting this information. _ If the criteria are formally adopted by the FDA, all nutritional information which appears on food packages will be required to adhere to these regulations. In the ninety-day period for filing written comments relating to this proposal, over 2,000 were submitted. They are now being processed and incorporated into a revised document which is expected to be published later this year.

Let us look briefly at the contents of the proposal and the research findings from which they derived. Since almost without exception, these findings are from unpublished reports, our review will of necessity be largely limited to the summaries which have been presented in the Federal Register in support of the proposed regulation. The proposed regulation calls for the prominent display of the following information on all food packages which present nutritional information:

1. A definition of serving size

2. Calorie content per serving

3. Number of grams of protein, fat, and carbohydrates per serving, as well as the amount of protein per serving expressed as a percentage of the 65 gram Recommended Daily Allowance (RDA)

4. The amount per serving of seven vitamins and minerals expressed as percentages of their RDAs. The seven are vitamin A, vitamin C, thiamin, riboflavin, niacin, calcium, and iron.

Research on Nutrient Labeling

The research which led to the proposed regulation began with a mail survey by Call and Hayes (1970) of 793 members of the American Institute of Nutrition, which sought their professional opinions on nutrient labeling. The study found that 85 percent of these nutritionists favored more nutrient information on the labels of food packages, with a majority giving highest ratings to the disclosure of calories, protein, fat, vitamins A, C, and D, calcium, iron, and additives and preservatives. While no consensus emerged on presentation modes for communicating nutrient information (e.g., per serving, per package, or per unit of weight), a majority expressed the opinion that vitamin and mineral disclosures should be based on a standard such as the RDA.

In a second study on nutrient information which was published in a trade magazine (Chain Store Age, October 1970), it was found that small shifts in consumer purchase patterns in the direction of nutrient-labeled food had followed their introduction in the marketplace.

Using this information as well as the advice and counsel of various professional, business, and consumer groups, the FDA proceeded to carry out additional research on nutrient labeling. Two series of investigations were initiated, one by the Consumer Research Institute, and the other by David Call and Daniel Padberg of Cornell University.

The first in the series of Consumer Research Institute studies was undertaken to secure information on presentation modes and label formats, as well as an indication of consumer understanding and usage of nutrient label information. This study examined the reactions of 950 educated, middle-class households which shopped by catalog, to the inclusion of nutrient information in the catalog descriptions of various food products. The preliminary findings revealed the following:

1. Purchase patterns shifted toward brands with nutritional advantages.

2. Additional consumer reactions to the catalog labeling program included a more positive attitude toward nutrition as well as greater nutritional knowledge.

3. Three experimental formats for disclosing the amount of RDA for the listed nutrients (numerical percentage, adjective representation, and pictorial representation), yielded no differences in consumer understanding or usage.

4. Differences in consumer reaction were also not found for two modes of displaying information for individual nutrients which are not present in a product. In one mode the nutrient was listed with a zero value, while in the other it was not listed at all.

A second study in the Consumer Research Institute series further explored consumer reactions to the three experimental formats and two display modes which were examined in the first study. This second study consisted of large-scale mail surveys of a U.S. probability sample of 2,000 consumers as well as a second sample of 2,000 low-income consumers. A third phase of this study used personal interviews to survey a sample of 600 customers who had not completed high school.

The preliminary findings revealed a slight overall advantage for the numerical percentage format over the pictorial and adjective formats - an advantage which was more pronounced for the sample of high school dropouts. Also of significance was the high percentage of respondents who were able to perceive differences as well as make correct nutritional choices using any of the formats (80 percent for the national sample, 70 percent for the low-income sample, and 90 percent for the low-education sample). With regard to the two display modes, a small advantage was found for lists which included only those nutrients present in a product, over lists which included all nutrients. However no attempt was made to determine the possible educational benefits to consumers of the more complete display mode.

Looking next at the Call and Padberg studies, we find that this effort was undertaken to secure a better understanding of consumer use and knowledge of nutrient labeling as well as consumer interest in nutritional information. Also sought was an indication of possible nonuse benefits to consumers. A second study was initiated to evaluate various nutrient label formats and to determine how a good diet may result from using nutrient information.

The preliminary findings which have been reported from the first of these studies are sketchy, representing early results from nutrient labeling experiments initiated by two food chains. The nutrient labeling programs of three additional food chains are also being evaluated by Call and Padberg, but at this writing, no findings have been reported.

To briefly summarize these various findings, it would appear that professional nutritionists agree on the need for more nutrient information of food packages as well as the specific nutrients to be listed. In addition, preliminary findings of three studies of various presentation modes and formats for nutrient information suggest that a variety of modes and formats can be effectively used and understood by consumers in general, and by disadvantaged consumers in particular, with a slight advantage appearing for numerical percentages of RDA over other forms of disclosure.

POLICY IMPLICATIONS OF RESEARCH ON CONSUMER AIDS

Although many significant insights have resulted from the research findings reported herein, their overall contribution to the policy-making process may be rather modest. This is so not because of specific deficiencies in the individual research studies, but rather because of the inherent limitations of the empirical research process as a contributor to the resolution of policy issues.

The questions for which answers are sought by policy makers concern the outcomes which are likely to follow the implementation of various policy proposals. Ideally, these answers would be expressed as a series of quantitative functions, each of which would indicate how much of a particular cost or benefit is likely to result at various points in time, from the date of adoption of a policy proposal to some fairly distant point in the future. This statement of ultimate criteria explicitly recognizes that 1) costs and benefits are multi-dimensional in nature, and 2) the long-term effects of new policies may differ markedly from their short-term effects.

With consumer aids, these two considerations are particularly important. The first emphasizes the necessity of going beyond simple frequency counts of aid users to a consideration of such additional criteria as consumer savings in time and money, nutritional improvements in diets, and both user and nonuser satisfaction with consumer aids. The second consideration suggests that since consumer aids are new and relatively unfamiliar to many shoppers, it may be several years before a true picture of stable usage and benefit patterns is likely to emerge. Indeed, when one considers the increasing attention being devoted to consumer economics courses in the public schools, it may be appropriate to speak of a generation gap with respect to public understanding of consumer aids which may take many years to close.

In light of the uncertainties inherent in these, as well as other developments, it seems clear that the future-oriented informational needs of the policy maker will remain unsatisfied. The methodological crystal ball of the social scientist, while not opaque, is still too cloudy to provide reliable predictions of future usage and benefit patterns for consumer aids.

Given the very real limitations in our forecasting powers, it is well to ask what role can be played by the social scientist in guiding policy decisions on consumer aid proposals. Two possibilities suggest themselves. The first consists of conducting longitudinal studies of shopper reactions to consumer aids, in an effort to establish at least a few critical data reference points from which cost and benefit curves could be constructed and extrapolated into future time periods. Needless to say, there are many problems with this approach, not the least of which is the substantial waiting time before useful information is likely to result. A series of longitudinal studies conducted over a short period, say one to two years, is unlikely to provide a reliable basis for projecting consumer reaction over a longer period; and the myriad, real-world political pressures on policy makers often do not permit them to wait out a longer study period before formulating and acting upon public positions on proposed consumer aids.

A second and perhaps more promising approach to the problem of providing future-oriented information to policy makers draws a distinction between the actual and Potential future value of information aids. Advocates of this approach believe that the social scientist can contribute little which bears directly on the former, but a great deal which sheds light on the latter. Thus they suggest that we do not attempt the near-impossible task of predicting the outcomes of policy decisions relating to consumer aids, but devote our efforts instead to determining the potential value of various proposed aids for the consumer.

But how, one may ask, do we define "potential value," and perhaps more importantly, how do we measure it for various consumer aids? A useful beginning may be made by viewing a consumer aid as having potential value if it presents useful information to the consumer in a highly communicable form. The determination of usefulness could be made by surveying various specialists, such as nutritionists and consumer economists, or consumers at large, although in some instances the latter group may lack the technical expertise to respond in a meaningful manner. Communicability, on the other hand, is more appropriately tested in a laboratory or field setting to assure that actual consumers are able to understand and effectively use the information provided by consumer aids.

Although many of the research studies reviewed in this paper have examined various aspects of information usefulness and communicability for consumer aids, they have typically failed to build upon one another to provide a meaningful picture of the potential value of the consumer aids under investigation. In large part this is due to the fact it is not the concepts of say, unit pricing or open dating, which have been researched in these many studies, but various particularities of these concepts. And with few exceptions, the particularizations selected for study have been determined by the exigencies of practical circumstance rather than the logic of a relevant theoretical model. As a result, it is difficult, if not impossible, to fit together what may be pieces from several different jigsaw puzzles into a unified whole.

The one exception to this pattern is the research on nutrient labeling. Due to the efforts of a single government agency, a programmatic approach to researching the potential value of this consumer aid has been undertaken which systematically explores the effects of various design parameters on different segments of the consumer population. While research of this sort will not furnish any ultimate answers to the policy maker, it may well serve the important function of helping him to decide which consumer aid proposals warrant serious consideration and which do not.

REFERENCES

Call, D. L. & Hayes, M. G. Reactions of nutritionists to nutrient labeling of foods. The American Journal of Clinical Nutrition, 23, 1970, 1347-1352.

Food Stability Survey, Volumes I & I; Rutgers University, The State University of New Jersey, Department of Food Science, 1971.

Friedman, M. P. Consumer confusion in the selection of supermarket products. Journal of Applied Psychology, 1966, 50, 529-534.

Friedman, M. P. Consumer price comparisons of retail products: the role of packaging and pricing practices and the implications for consumer legislation. Journal of APPlied Psychology, in press.

Gatewood, R. D. & Perloff, R., An experimental investigation of three methods of providing weight and price information to consumers. Journal of Applied Psychology, in press.

Hoofnagle, W. & Stokes, R. Consumer reactions to food quality in Conference Proceedings: Food Stability and Open Dating. October 21-22, 1971, Rutgers University, New Brunswick, N.J. 24-32.

Houston, M. The effect of unit-pricing on choices of brand and size in economic shopping. Journal of Marketing, 36, 1972, 51-54.

Mayer, J. Toward a national nutrition policy, Science, 176, 1972, 237-241.

McCullough, T. D. & Padberg, D. I. Unit pricing in supermarkets: alternatives, costs and consumer reaction. Search Agriculture, 1, 1971, 1-25.

Monroe, B. & LaPlaca, P. J. What are the benefits of unit pricing? Journal of Marketing, 36, 1972, 16-22.

Ringler, M. & Berner, G. A. Consumer Attitudes Toward the Food Industry, paper presented at the 38th Annual Meeting of the National Association of Food Chains.

Zehner, M. Consumer study on Stability and Open Dating. Conference Proceedings: Food Stability and Open Dating, October 21-22, 1971, Rutgers University, New Brunswick, N.J., 33-39.

----------------------------------------