Consumers' use of nutritional labels while food shopping and at home.
The Nutritional Labeling Education Act (NLEA) was signed into law on November 8, 1990. The objective of this legislation was to provide consumers with information that would assist them in maintaining healthy dietary practices. In part, this involved designing a consistent, understandable, and usable nutrition label. The NLEA instituted sweeping changes to replace the voluntary system of labeling established by the Food and Drug Administration (FDA) regulations in 1973. The new food labeling legislation mandated nutrition labeling on most processed foods under the jurisdiction of the FDA, established reference Daily Values for certain nutrients, defined serving sizes, and limited health claims. It also established guidelines for voluntary labeling of raw fruits, vegetables, and seafood.
In order to establish or maintain healthy dietary practices, consumers must have the necessary information available and actively seek that information. The effectiveness of nutrition information programs, therefore, is contingent upon consumers' use of nutritional labels. It is possible that many individual diets fall short of the Dietary Guidelines because they do not use or care about nutritional information provided on food packages to help them in their food buying decisions (Nayga 1996). It is, therefore, important to know the factors affecting consumer's use of nutritional labels.
Past research on the utilization of nutrition information has attempted to answer several key questions. Do consumers want nutrition information? Will they acquire nutrition information when making food purchase decisions? Having acquired the information, will they adequately understand and use it? Survey results showed that consumers do want nutrition information, would use it, and are willing to pay extra for it (Bender and Derby 1992; Jacoby, Chestnut, and Silberman 1977). Additional research has been done to identify sources of nutrition information and the factors that affect use of this information (Feick, Herrmann, and Warland 1986). A few studies have also been conducted to determine the factors that affect point-of-purchase use of nutrition labels as well as reasons for nonuse (Guthrie et al. 1995; Klopp and MacDonald 1981).
Scant information, however, is available concerning label use by place of usage. For example, do consumers read the labels while shopping or at home? What type of consumers use labels while shopping and what type of consumers use labels at home? Do consumers use labels to compare brands of the same type of foods while shopping? These are the questions relevant to the present study. The objective of this study is to evaluate the impact of socioeconomic, demographic, and health/nutrition related factors on label use while food shopping, at home, and when comparing brands. No other known study has examined this topic. The results of this study can be used for developing label education messages and materials. Knowledge of the relationship between sociodemographic characteristics and label use is useful in the design and implementation of nutrition information programs. For instance, sociodemographic factors can be used to tailor health interventions and marketing campaigns to specific population subgroups. Lin (1995) suggested that targeted consumer education is more efficient than generic programs in improving the general health of the nation.
The economic model of information search, introduced by Stigler (1961), is used as a framework to model nutrition label use. This framework assumes that the utilization of product information is an active process that involves searching out information. When making product choices, the consumer also must be able to evaluate the information obtained in the search and make decisions based on that evaluation (Senauer 1991).
In this study, the use of nutrition labels was considered an act of consumer information search. The economics of information predicts that the consumer will continue to acquire and process information so long as the costs of additional acquisition and processing do not outweigh the additional benefits. The amount of search may vary among individuals due to differences in either their expenditure level on a commodity, the costs of search, or the perceived benefits.
The main cost to the consumer of an information search, or in this case nutrition label use, is the time spent reading nutrition labels. This time could be reflected in either the number of minutes spent while grocery shopping or the number of minutes spent reading nutrition labels at home. The cost of using nutrition labels may be represented as lost wages or sacrificed leisure activities due to the allocation of time.
The benefits of nutrition label use may be reflected in part by better food choices which, in turn, may yield a more nutritious diet. As research continues to establish and explain the diet-health relationships, the benefits of nutrition label use may also be perceived as the reduced risk of chronic diseases, such as coronary heart disease, osteoporosis, or cancer. This reduced risk of disease could be measured in terms of reduced health care costs: doctor visits, medication, hospitalization, etc. and reduced loss of wages from missing work. Moreover, it can be argued that labels could benefit specific "at risk" populations who have the need to follow a specific diet (i.e., low sodium or low fat).
The process by which consumers make decisions involves an internal and external search of information. Internal search is sometimes referred to as memory scanning. External search is the process of seeking information prior to purchase. Most research in the area of consumer behavior focuses on external search and attempts to answer the following questions: How much search is undertaken? What is known about the reasons for variations in the amount of search? What do the findings mean?
Based on the framework discussed above, Moore and Lehmann proposed a categorization of the determinants of the extent of information search which includes both internal and external factors (Moore and Lehmann 1980). The categorization developed was an attempt to clarify and expand previous classifications (Bettman 1979; Newman 1977). Working within this classification system, nutrition label use can be modeled as a function of several major categories of variables including (a) individual characteristics, such as sex, age, race, education, household size, special diet status, and primary sources of nutrition information; (b) situational variables including time and financial constraints, such as employment status, average number of minutes spent shopping, and income; (c) product importance, that is, the importance to the consumer of a product or a product attribute, such as price, nutrition, taste, and ease of preparation; and (d) prior knowledge of the product or product attribute which can be measured in terms of nutrition knowledge and importance of following certain dietary guidelines.
DATA AND METHODS
A survey of consumers was conducted during the first and second quarter of 1996. A sample of 200 consumers was obtained from a combination of four supermarkets targeted in different socioeconomic areas of Middlesex County, New Jersey.(1) One supermarket was chosen to represent a low-income population. Two supermarkets were chosen to represent a middle-income population, and the last supermarket was selected to represent an upper-income population. Fifty surveys were completed at each of the four supermarkets, and the first 100 participants received a complimentary copy of the Nutritional International Pocket Guide to Nutritional Labeling (Nutrition International 1993).
After permission was obtained from the supermarket managers, trained student interviewers visited each supermarket, setting up tables outside the main entrance when space permitted.(2) The interviewers approached consumers at random and asked them to participate in the survey. Consumers were given the option of having the survey read to them and filled out by the interviewer or completing it themselves. In several instances where the consumer agreed to participate in the survey, the questionnaire was translated in Spanish. The interviewer recorded the consumer's responses (the questionnaire is available from the authors).
Survey respondents were asked a series of questions regarding the use of nutrition labels. Three questions were modeled separately in this study: (1) How often do you read the nutrition information on food labels while food shopping? (2) How often do you read the nutrition information on food labels while preparing food at home? and (3) While grocery shopping, how often do you compare nutrients such as fat, cholesterol, and sodium for different brands of the same foods? Possible responses were always, often, not often, or never.
For each of the measures of label use, individuals who did not respond to the question were deleted from the sample. Likewise, individuals who failed to report pertinent socioeconomic and demographic information were deleted from the analysis. Consequently, the final number of individuals analyzed in the study was 159.
Table 1 provides the description and means of the variables used in the analysis. A potential self-selection problem may arise from deleting data for individuals who failed to report pertinent socioeconomic and demographic information. Comparing the means of the variables used in the analysis with and without the deleted data provides one way to assess the degree to which this problem may exist. The means of the variables with the deleted data matched with the means of the variables without the deleted data. The sample is also representative of both New Jersey and Middlesex County except for the underrepresentation of whites and overrepresentation of individuals aged 18 to 55 years (Slater and Hall 1996).
The distributions of responses on the three label use questions are exhibited in Table 2. About 70, 62, and 52 percent of the respondents answered either often or always to the questions, respectively. About [TABULAR DATA FOR TABLE 1 OMITTED] 25 percent of the individuals always use labels while shopping, while only 18 and 11 percent of individuals always use labels at home and when comparing brands, respectively.
TABLE 2 Distribution of Responses (Percent) to the Label Use Questions Question Never Not Often Often Always Label use while shopping 6.3 23.3 45.9 24.5 Label use at home 19.1 18.5 44.6 17.8 Label use to compare brands 17.0 30.8 40.9 11.3
While searching for information, consumers may use a broad array of health and nutrition information sources. Depending on the source, consumers may acquire varying amounts of health and nutrition information due to differences in the type, quality, and perceived credibility of the information. Nutrition labels are one possible source of nutrition information. Two additional sources may be categorized as a people source and a media source (Capps 1992). Consumers may obtain information from people, such as health professionals, nutritionists, dieticians, or home economists or from the media which includes both broadcast and print sources, such as radio, television, newspapers, magazines, books, health organization publications, or food company publications. Survey respondents were asked to indicate their primary source of nutrition information. Fifteen percent of the individuals indicated physicians, dieticians, or nutritionists; 10 percent indicated health organization publications; 33 percent indicated books, magazines, or newspapers; 13 percent indicated radio or television; 21 percent indicated food packages or labels; and 7 percent indicated other sources. The majority of individuals surveyed (56 percent) indicated that some form of media source was the primary source of nutrition information. For purposes of this study, individuals who indicated either physician, dietician, nutritionist, or health organization publications were grouped together. Similarly those who indicated either books, magazines, newspapers or radio, television, or other were grouped together. The grouping was motivated in part by the three categorizations of nutrition information sources discussed earlier as well as the possible perceived differences in the quality and credibility of the information.
Nutrition label use was modeled as not only a function of socioeconomic and demographic factors but also nutrition knowledge and attitudes. Two indices were created from questions in the survey to measure nutrition knowledge and the importance of following certain dietary guidelines. A total of eight questions were used to measure nutrition knowledge and four questions pertained to dietary guidelines.
When given a choice of two foods, most consumers were able to correctly identify which had more cholesterol and which had more fat. However, when asked two, more specific questions, most consumers were not able to correctly respond. Consumers were asked to indicate the percentage of daily caloric intake that should be from fat (30 percent or less) and the limit for total daily intake of salt/sodium (2400 mg).
For each question a correct answer was assigned a one and an incorrect answer was assigned a zero. A total score was obtained for each individual. Scores ranged from 1 to 7. Nine percent of the individuals scored 7 points and 12 percent had scores of 3 points or less. The majority of individuals (65 percent) had a score of 5 or 6 points.
The "Dietary Guidelines for Americans" were designed to help answer the following question: What should Americans eat to stay healthy? These guidelines provide advice about food choices that promote health and prevent diseases (U.S. Department of Agriculture 1995). Consumers in the survey were asked how important the following advice was to them personally: avoid too much salt/sodium, avoid too much saturated fats, choose a diet low in cholesterol, and eat a variety of foods. Possible responses ranged from not at all important (1), neutral (3), to very important (5). Slightly more than half (55 percent) responded that it was important or very important to avoid too much salt/sodium. The percentages were higher when asked about saturated fats and cholesterol (73 percent and 67 percent respectively). Similarly a high percentage of individuals (69 percent) said eating a variety of foods was important or very important.
For each question, responses of important or very important were assigned a two, responses of neutral were assigned a one, and responses of less important and not at all important were assigned a zero. Again a total score was obtained for each individual. Scores ranged from 0 to 8. Forty percent of the individuals had a score of 8 and 14 percent had a score of 3 or less. The mean score was 6.13 for the sample.
The dependent variables are measured on a scale that is discrete and ordinal. Hence, ordered logit models are employed in the analyses. The most suitable technique of estimation when using logit is maximum likelihood. Although this technique requires the use of iterative algorithm, this procedure assumes the large-sample properties of consistency and asymptotic normality of the parameter estimates so that conventional tests of significance are applicable.
The three models are specified as follows:
[Prob.sub.j] = [[Beta].sub.0] + [[Beta].sub.1] Male + [[Beta].sub.2] Age28 + [[Beta].sub.3] Age40 + [[Beta].sub.4] Age50
+ [[Beta].sub.5] Nonwhite + [[Beta].sub.6] Income1 + [[Beta].sub.7] Income2 + [[Beta].sub.8] College
+ [[Beta].sub.9] Graduate + [[Beta].sub.10] Unemployed + [[Beta].sub.11] Hsize + [[Beta].sub.12] ShopMin
+ [[Beta].sub.13] People + [[Beta].sub.14] Media + [[Beta].sub.15] Specdiet + [[Beta].sub.16] Nutrknw
+ [[Beta].sub.17] DGimp + [[Beta].sub.18] Price + [[Beta].sub.19] Nutr + [[Beta].sub.20] Taste
+ [[Beta].sub.21] Eprep + [[Beta].sub.22] error
where [Prob.sub.j] is an ordered variable that range from 1 = never to 4 = always concerning extent of label use by the individual under circumstance (j). A description of the variables are in Table 1. For estimation purposes one classification was eliminated from each group of binary variables to avoid the problem of perfect collinearity. The base group is those individuals who satisfy the following description: female, age 56 years and older, white, combined family income greater than or equal to US$50,000, high school education or less, employed, and food packages or labels as primary source of nutrition information.(3)
The maximum likelihood estimates of the three models are exhibited in Table 3. Due to the complexity of factors influencing individual label use and the cross-sectional nature of the data, the McFadden [R.sup.2] statistics, as expected, are relatively low but reasonable. No degrading multicollinearity problems were detected from the data based on the collinearity diagnostic tests conducted (Belsley, Kuh, and Welsch 1980).
In all three models, although the signs of the coefficient of the male variable are negative, the results indicate that males are not significantly less likely to use nutrition labels. Past research has revealed that men are significantly less likely to use nutrition labels than do women (Nayga 1996). The finding in the present study is intriguing because it could reflect the changing role of men in the household. Because of the increasing labor force participation rate of women, [TABULAR DATA FOR TABLE 3 OMITTED] men may have started sharing some of the responsibilities of food shopping and preparation.
In terms of the other demographic variables, individuals who have an annual family income of less than $30,000 are less likely to use labels to compare brands while shopping than those with annual family income of at least $50,000. Education level was found to be significant only in the model of label use at home. Individuals with some college education, a bachelor's degree, and a graduate or professional degree are more likely to use nutrition labels than individuals who do not have any college education. Similar studies involving point-of-purchase use of labels also found education to be a significant factor. When using nutrition labels it is sometimes necessary to make a few calculations. Those with more education may be more capable of interpreting the information provided on nutrition labels.
The employment variable is statistically significant in all three models. Specifically, unemployed individuals are more likely to use nutritional labels while shopping and at home as well as to compare brands. Employment status was hypothesized to represent the time constraints involved with label use.(4) It appears that unemployed individuals are more able to allocate time to the use of nutrition labels. For employed individuals, nutrition educators may need to place additional emphasis on the benefits associated with label use or design ways so as to minimize the time needed to obtain and evaluate the information.
Results also suggest that the more time an individual spends on grocery shopping per visit, the more likely the individual will use the labels while shopping. Interestingly, individuals who indicated that their primary source of nutritional information is books, magazines, radio, TV, and newspapers are less likely to use labels while shopping and when comparing brands than those who use labels as their primary source of nutrition information. The reason for this result is not clear. However, it is possible that consumers face confusing messages from these information sources. Further research is recommended to determine consumers' attitudes and perceptions toward the type, quality, and perceived credibility of the information available from each source considered.
Special diet, which is used to represent the individual's current health status, is statistically significant on label use while shopping model. Specifically, those who are on a special diet are more likely to use labels while shopping. Those who have current health concerns appear to perceive greater benefits from label use, assuming they are able to use the information to make healthy food choices. For those who may not have health concerns, it may be possible to design a nutrition education program that emphasizes the role that a healthy diet plays on a person's or other household members' future health which could encourage use of nutrition labels.
Interestingly, nutrition knowledge has no significant effect on all three models. Hence, it is not clear if nutritional knowledge adequately measures the consumer's ability to evaluate the information found on nutrition labels. As noted, consumers had a general food-nutrient association knowledge but were unable to respond to more specific questions. These specific questions are important when trying to determine if consumers can understand the information on labels and incorporate that into choosing foods for a healthy diet. This issue is particularly relevant since Mazis and Raymond (1997) found that increased consumer knowledge of nutrition information may reduce consumer misperceptions.
The dietary guidelines variable, on the other hand, has a positive effect on label use in all three models. Individuals who place greater importance on following the dietary guidelines are more likely to use nutrition labels. The guidelines provide general advice about food choices in terms of nutrients, for example, salt/sodium, fat/saturated fat, and cholesterol. For this group of individuals it would be important to determine what particular components of the nutrition label are being used. If, for instance, individuals who place greater emphasis on choosing a diet low in salt/sodium only use the nutrition label to obtain the sodium content of a product, then they may be choosing a product high or low in other key nutrients. There is a trade off involved with using only certain components of the nutrition label. Healthy food choices depend, in part, on the consumer's ability to use and evaluate all the information provided on the label. Moreover, it is also important to know how individuals trade off higher levels of nutrients in one product with lower levels in others. These could be important areas of future research.
Concerning the perceived importance of certain product attributes, the results indicate that individuals who place a greater emphasis on price while shopping are less likely to use nutrition labels at home. The individual's perceived importance of nutrition while shopping is statistically significant and positive in all three models. Therefore, the higher the individual's perceived importance of nutrition while shopping, the higher is the likelihood the individual will use nutritional labels while shopping, at home, or when comparing brands. It would be interesting to know if indeed the food choices made by this group of consumers yield a more nutritious diet (Nayga 1997). Individuals who place greater importance on taste while shopping are also more likely to use labels while shopping.
The results of this study generally indicate that socioeconomic, demographic, and health/nutrition-related factors have varying effects on label use while shopping, at home, and when comparing brands. No other known study on this topic has been published. The results generally suggest that unemployed individuals and those who place greater importance on nutrition while shopping and on following the dietary guidelines are more likely to use nutritional labels while shopping, at home, or when comparing brands while shopping. Findings also indicate that education has a positive impact on the likelihood of using labels at home. Individuals who use the media (e.g., books, magazines, radio, TV, and newspapers) as primary sources of information are less likely to use labels while shopping and when comparing brands than individuals who use the labels as their primary source of nutritional information. The differences in the empirical results between the models further justifies the separate analysis conducted on the three models. In addition, these results provide additional evidence that supports previous conclusions made about nutrition education programs to tailor these programs for different groups of people.
While this exploratory study provides some valuable information on three types of label usage, several questions remain unanswered. For instance, if consumers use nutrition labels, a key question still remains as to whether consumers understand the information obtained. As noted, the use of product information is a three-step process. It involves not only the search, but also the evaluation and decision making phase. A study involving reported label use and actual food purchases may indicate to what extent consumers understand the information obtained from labels. If, however, consumers use nutrition labels and are able to adequately evaluate the information but still do not make healthy food choices, then a search for other factors affecting food choices is necessary. Is it economic factors, such as price or income, that affect food choices or is it advertising? Is it the importance of such attributes as taste or ease of preparation? The benefits of the NLEA depend on the assumption that consumers would change their food selection and consumption practices in response to changes in the food labels. However, label use itself is part of a much larger problem, that of the determinants that affect food choices. Knowing the characteristics of individuals who use nutrition labels is only one part of the analysis. Therefore, future research should extend this study and determine if reading labels affects the choice of food and diet.
Other questions that could be examined in the future include: Do consumers interpret different brands of the same food as such or as different products? How may frequency of eating away from home influence label use? What does label use at home suggest--is it altered serving sizes, different food combinations to ensure adequate diet, etc.? How does repeated purchases of similar products affect label use? Is it affected by the fact that if the consumer repeatedly selects a core group of products, he/she need not read the labels once becoming familiar with them? Is label use important in achieving a higher score on a Healthy Eating Index? These are not easy questions to answer but are critical to our comprehensive understanding of the impacts of label use.
1 The survey was designed to represent the population of Middlesex County, New Jersey. The four supermarkets are from different chains. A total of about 554 shoppers were approached to participate in the survey, for a response rate of 36 percent.
2 Unfortunately, no records were kept concerning the time of day or the day of week the interviews were conducted. However, efforts were made to have equal distribution of interviews by time of the day and by day of the week (e.g., weekday versus weekend). No records were kept as well on whether the respondents were approached before or after the shopping occasion.
3 The location where the respondent (store location) was interviewed, originally included in the models, was statistically insignificant. Therefore, store effects were not included in the models.
4 The "unemployed" category may include individuals whose time is fully committed to home productive activities or are retired, as well as those who are seeking employment. A referee suggested that although both groups may have fewer time constraints in terms of the allocation of their time, those who are not seeking work in the labor force are technically not unemployed.
Belsley, D. A., E. Kuh, and R. E. Welsch (1980), Regression Diagnostics, Identifying Influential Data and Sources of Collinearity, Wiley.
Bender, Mary M. and Brenda M. Derby (1992), "Prevalence of Reading Nutrition Information and Ingredient Information on Food Labels Among Adult Americans: 1982-1988," Journal of Nutrition Education, 24(6): 292-297.
Bettman, James R. (1979), An Information Processing Theory of Consumer Choice, Reading, MA: Addison-Wesley Publishing Company.
Capps, Jr., Oral (1992), "Consumer Response to Changes in Food Labels: Discussion," American Journal of Agricultural Economics, 74(5): 1215-1216.
Feick, Lawrence F., Robert O. Herrmann, and Rex H. Warland (1986), "Search for Nutrition Information: A Probit Analysis of the Use of Different Information Sources," Journal of Consumer Affairs, 20(2): 173-192.
Guthrie, Joanne F., Jonathan J. Fox, Linda E. Cleveland, and Susan Welch (1995), "Who Uses Nutrition Labeling, and What Effect Does Label Use Have on Diet Quality," Journal of Nutrition Education, 27(4): 163-172.
Jacoby, Jacob, Robert W. Chestnut, and William Silberman (1977), "Consumer Use and Comprehension of Nutrition Information," Journal of Consumer Research, 4(2): 119-128.
Klopp, Pamela and Maurice McDonald (1981), "Nutrition Labels: An Exploratory Study of Consumer Reasons for Nonuse," Journal of Consumer Affairs, 15(2): 301-316.
Lin, C. T. J. (1995), "Demographic and Socioeconomic Influences on the Importance of Food Safety in Food Shopping," Agricultural and Resource Economics Review, 24(2, October): 190-198.
Mazis, M. and M. Raymond (1997), "Consumer Perceptions of Health Claims in Advertisements and on Food Labels," Journal of Consumer Affairs, 31(1, Summer): 10-26.
Moore, William L. and Donald R. Lehmann (1980), "Individual Differences in Search Behavior for a Nondurable," Journal of Consumer Research, 7: 296-307.
Nayga Jr., R. M. (1996), "Determinants of Consumers' Use of Nutritional Information on Food Packages," Journal of Agricultural and Applied Economics, 28(2): 303-312.
Nayga Jr., R. M. (1997), "Impact of Sociodemographic Factors on Perceived Importance of Nutrition in Food Shopping," Journal of Consumer Affairs, 31(1): 1-9.
Newman, Joseph W. (1977), "Consumer External Search: Amount and Determinants," in Consumer and Industrial Buying Behavior, A. G. Woodside, J. N. Sheth, and P. D. Bennett (eds.), Amsterdam, The Netherlands: North Holland Publishing Company, 79-94.
Nutrition International (1993), The Nutrition International Pocket Guide to Nutrition Labeling, South Brunswick, NJ: Nutrition International.
Senauer, B., E. Asp, and J. Kinsey (1991), Food Trends and the Changing Consumer, Minneapolis, MN: Eagan Press.
Slater, Courtenay M. and George E. Hall (1996), 1996 County and City Extra; Annual Metro, City and County Data Book, Lanham, MD: Bernan Press.
Stigler, George J. (1961), "The Economics of Information," Journal of Political Economy, 69(3): 213-225.
U.S. Department of Agriculture (1995), Nutrition and Your Health: Dietary Guidelines for Americans, Washington, DC.
Rodolfo M. Nayga, Jr., is Assistant Professor, Department of Agricultural Economics, Texas A&M University; Daria Lipinski is a Ph.D. student, Department of Agricultural, Resource, and Managerial Economics, Cornell University; Nitin Savur is a former undergraduate student, Department of Agricultural Economics and Marketing, Rutgers University. The authors thank Carole Makela and three anonymous journal reviewers for their helpful comments and suggestions.
|Printer friendly Cite/link Email Feedback|
|Author:||Nayga, Rodolfo M., Jr.; Lipinski, Daria; Savur, Nitin|
|Publication:||Journal of Consumer Affairs|
|Date:||Jun 22, 1998|
|Previous Article:||Disabled elders' out-of-pocket home care expenses: examining financial burden.|
|Next Article:||A Bayesian approach for analyzing the services of banking institutions.|