Food Consumption Behavior

ABSTRACT - A taxonomy of determinants of food consumption behavior is proposed. It distinguishes between three types of factors, viz. properties of the food, factors related to the person, and environmental factors. The effects of each category of factors on food consumption behavior is discussed, drawing on a number of science and social science disciplines. However, any comprehensive analysis of food consumption behavior should consider all three types of determinants and their interactions. A number of interdiscplinary research issues are identified that may be of special interest to consumer behavior researchers.

Citation:

Jan-Benedict E.M. Steenkamp (1993) ,"Food Consumption Behavior", in E - European Advances in Consumer Research Volume 1, eds. W. Fred Van Raaij and Gary J. Bamossy, Provo, UT : Association for Consumer Research, Pages: 401-409.

European Advances in Consumer Research Volume 1, 1993      Pages 401-409

FOOD CONSUMPTION BEHAVIOR

Jan-Benedict E.M. Steenkamp, Agricultural University, Wageningen, The Netherlands

ABSTRACT -

A taxonomy of determinants of food consumption behavior is proposed. It distinguishes between three types of factors, viz. properties of the food, factors related to the person, and environmental factors. The effects of each category of factors on food consumption behavior is discussed, drawing on a number of science and social science disciplines. However, any comprehensive analysis of food consumption behavior should consider all three types of determinants and their interactions. A number of interdiscplinary research issues are identified that may be of special interest to consumer behavior researchers.

INTRODUCTION

Food plays a central role in the life of consumers. It is the source of nutrition and hedonic experiences, it serves a social and cultural function, and has considerable economic significance since a major proportion of the household budget is allocated to purchasing food. Yet food consumption behavior has not attracted much systematic attention by consumer behavior researchers. At least part of the difficulty in conducting consumer behavior research in this important area lies in the complexity and the diversity of the influences at work in food choice and consumption, and in the fact that such research requires knowledge of the concepts of and insights from a wide range of science and social science disciplines, including food science, nutrition, medicine, psychology, physiology, psychophysics, sociology, economics, marketing, and anthropology.

The purpose of this paper is to present an interdisciplinary overview of the literature on the determinants of food consumption behavior in order to assist consumer behavior researchers in their own research efforts. No claim can be made that this overview is complete, but a number of key findings from various disciplines will be reviewed. Some research issues are identified for which consumer behavior researchers could make a particularly valuable contribution.

A TAXONOMY OF DETERMINANTS OF FOOD CONSUMPTION BEHAVIOR

Several models of determinants of food consumption behavior have been proposed in the literature. One of the earliest and most influential models was proposed by Pilgrim (1957). In his model, food consumption is dependent on perception. [Pilgrim discussed food acceptance rather than food consumption. However, he acknowledged that the operational definition of food acceptance is food consumption.] Food perception is a function of three factors: 1) physiological effects of the food, 2) perception of sensory attributes, and 3) influences from the environment. Pilgrim hypothesized that these determinants will interact in influencing food consumption, but he did not explore these interrelations. The model also incorporates the time factor, with external influences being either recent or long established, and with some physiological influences being relatively stable for an individual, while other influences will vary over short periods with ingestion of foods (e.g., hunger).

Pilgrim's model served as point of departure for many subsequent models of determinants of food consumption behavior (see Shepherd 1990 for an overview). Although some differences can be observed between these models, they generally distinguish between three types of determinants: 1) properties of the food, 2) factors related to the person engaged in food consumption, and 3) environmental factors.

Figure 1 presents a taxonomy of determinants of food consumption behavior which adopts this distinction between the three types of determinants. It is acknowledged that the boundaries between them are fuzzy, and that mutual influences occur. Thus, any comprehensive analysis of food consumption behavior must consider all three types of determinants. The taxonomy also incorporates a number of specific factors relating to each type of determinant. Properties of the food include physical and chemical properties, and nutrient content, such as physical form, proportions of macronutrients, amount of fiber, energy value, and amounts of specific substances (sugar, salt, seasoning, etc.). These food properties affect food consumption behavior through their physiological (e.g., hunger, satiation, appetite) and sensory effects. Factors related to the person include biological factors (e.g., age, sex, body weight), psychological factors, and personality. Environmental factors include sociocultural, economic, and marketing factors.

Although the three types of determinants will interact, this paper will discuss each type of determinant separately to keep the presentation manageable and because most research evidence concerns the effects of a single type of determinants in isolation. Future research should take a more comprehensive approach, as will be detailed in the last section of the paper.

PROPERTIES OF THE FOOD

Physiological effects

Eating of food reduces hunger, leads to satiation, and causes people to stop eating. The satiation power of foods has considerable significance for consumers in the Western world, given the prevalence of dietary concerns. Food properties affecting the degree of satiation include physical form, proportions of macronutrients and fiber, and energy value (Blundell, Hill, and Rogers 1988). Solid foods have a greater satiation effect than liquid foods. The satiation effect of protein is larger than the equivalent caloric value of fat or carbohydrate. Alcohol does hardly have a satiation effect, while the effect of fiber is substantial.

Eating high-calorie products reduces hunger more than eating low-calorie products. People learn to consume more energy rich foods when hungry than when sated (Booth 1982). However, humans' appetite control system appears to be highly responsive to an under-supply of energy, but tolerant of over-supply with the consequence that there is relatively little physiological resistance to overeating (Rogers and Blundell 1990). Although such mechanisms make much sense when food supply is uncertain, they tend to encourage obesity in the Western world where food availability is rarely, if ever, a problem.

The issue of the satiation capacity is particularly relevant to the development of low-calorie products. As energy value is a major source of satiating capacity, the question therefore arises whether the reduction in the caloric content of foods will lead to compensatory behavior that reduces the beneficial effects of consuming low-calorie products. Research indicates that the consumption of low-fat products is probably more effective than the consumption of the equivalent caloric value of low-sugar products. Subjects consuming low-sugar products compensated the decrease in energy intake for about 50% by eating more of other foods while such compensatory behavior did hardly occur for low-fat products (DeGraaf 1992).

FIGURE 1

A TAXONOMY OF THE DETERMINANTS OF FOOD CONSUMPTION BEHAVIOR

Hunger reduction and satiation may be considered normal physiological consequences of eating food. Food consumption can incidentally cause more extreme physiological reactions such as nausea, vomiting, and gastrointestinal upset. When food consumption is followed by these negative physiological reactions, this may lead to a strong aversion to the food. This learning can occur with delays of hours between ingestion and illness. Food aversions can be "irrational" in that people can acquire aversions to foods even when they "know" that the food was not responsible for the illness. Up to 65% of consumers report at least one strong food aversion, with most aversions apprarently being acquired during childhood, a period in which many foods are tried for the first time and when the frequency of sickness with gastrointestinal symptoms may be particularly high (Rogers and Blundell 1990).

Sensory perception

[The classification of sensory perception in the taxonomy of food behavior determinants is somewhat ambiguous. Shepherd (1985) categorized it as a person-related factor, while Randar and Sanjur (1981) placed it under food properties, and Pilgrim(1957) assigned it a separate category in his model. As sensory perception appears to be rather homogeneous (e.g., people tend to perceive the same sweetness, although their liking of the amount of sweetness may differ), the properties of the food are the predominant factor in sensory perception, and therefore we categorize it as a food-related factor.]

How people form sensory perceptions based on the properties of the food product has been investigated primarily in psychophysics. [Psychophysics studies the relationship between sensory perception and physical characteristics of the product. Another stream of research deals with the effects of cognitive cues (e.g., brand names, price) on sensory perceptions (see Steenkamp 1989 for a review and meta-analyses). This research tends to be rather ad hoc in nature and no unequivocal conclusion can be drawn yet.] The complexity of psychophysical relations has led researchers to study different senses typically in isolation. The simplest situation, called "classical" psychophysics, involves sensory perception of a single physical characteristic:

y_i = f_ij (x_j) (1)

where y_i is perception of the stimulus with respect to sensory attribute i (i=1,...,m), x_j denotes the value of the food product on physical characteristic j (j=1,...,k), and f_ij is the psychophysical function relating y_i and x_j. The two best-known specifications of f_ij are Fechner's logarithmic law of perception and the power law (Stevens 1975).

Classical psychophysics is of limited validity for studying sensory perceptions of foods because multiple physical characteristics can affect the perceptions on the same sensory attribute, and interactions among physical characteristics are not taken into account (Frank and Archambo 1986, DeGraaf and Frijters 1989). Equation (2) extends classical psychophysics by modeling the overall perception toward a multivariate stimulus (x1,...,x_k):

yi = g_i [f_i1(x₁), f_i2(x₂), .... , f_ik(x_k)] (2)

where gi denotes a multivariate psychophysical transformation function that allows for interactions in the sensory perception process. Steenkamp and Van Trijp (1989) used this model to quantify the psychophysical relation between physical characteristics and sensory perceptions for meat cuts. They assumed g_i to be additive and f_ij to be linear. Equation (2) is also used in conjoint-type of studies in which the food products are experimentally manipulated on several characteristics simultaneously. In these studies, g_i is typically specified as either additive or configural, and f_ij is free. One example is the study by Huber, Holbrook, and Schiffman (1982) which investigated the effects of three physical characteristics of citrus beverage on refreshment perception.

TABLE 1

A TAXONOMY OF PSYCHOLOGICAL FACTORS INFLUENCING FOOD CHOICE

The most general psychophysical model allows multiple characteristics to affect the perceptions on the same sensory attribute and any single characteristic to affect the perceptions on several attributes. This requires the researcher to study the total profile of relations between characteristics and perceptions, for which the following model may be used:

zq(y₁, y₂, ..., y_m)=g_q[f_q1(x₁), f_q2(x₂),...., f_qk(x_k)] (3)

where z_q and g_q are multivariate transformation functions of sensory attribute perceptions and physical characteristics, respectively, with q = 1,...r, and r _ min (m,k). Thus, more than one set of transformation functions z_q and g_q may be needed to model the psychophysical relations between physical specifications and attribute perceptions. When it is assumed that z_q and g_q are additive and f_qj is linear, Equation (3) can be estimated with techniques such as canonical correlation analysis (CCA) or partial least squares regression analysis (PLS). Martens and Martens (1986) illustrate the use of both techniques for foods.

PERSON-RELATED FACTORS

Biological factors

Age and body weight have been identified as major biological factors affecting food consumption behavior (DeGraaf 1992). There are certain taste preferences apparently present at birth which can bias food choice. In a seminal study, Steiner (1979) found that newborns like sweet stimuli and reject bitter stimuli. Studies of bottle-fed infants showed that sugar enhances intake and it is added to infant foods in many cultures, presumably for this reason (Rozin, Pelchat, and Fallon 1986). Such innate biases can be highly beneficial to the chances of survival as sweet foods normally contain energy while bitter tastes tend to be ecologically correlated with the presence of toxins (e.g., alkaloids in plants) (Rozin and Vollmecke 1986).

Similarly, infants can reliably detect, differentiate, and respond to a range of odors only a few hours after birth. Odor preferences have been demonstrated in very young infants, which confers distinct biological advantages. Foods that smell pleasant are probably suitable for ingestion, in that they are not spoiled and do not contain dangerous levels of pathogenic micro-organisms (Van Toller and Kedal-Reed 1990).

Despite these innate biases, the taste and smell senses are highly adaptable throughout life. Thus, food preference acquisition is probably a life-long process. People learn to appreciate slightly bitter products such as coffee, beer, and grapefruits. Food preferences also change due to aging processes. For example, decline in olfactory sensitivity with age is one of the reasons why many elderly perceive food to be less tasty. See also Schiffman (1979) and Cowart (1981).

There is a positive relation between food consumption and body weight. A partial explanation for this phenomenon can be found in the physiological reaction to foods. In human beings, the smell and sight of foods are sufficiently potent to trigger cephalic phase insulin responses, and the response in obese people is about four times greater than in lean people. The food-induced cephalic phase responses are accompanied by an increase in the desire to eat. Consequently, obese people may have greater appetite than lean people, thus perpetuating their overweight (Blundell, Hill, and Rogers 1988).

Psychological factors

Rozin, Pelchat, and Fallon (1986) proposed a taxonomy consisting of three basic types of psychological reasons for acceptance or rejection of foods: sensory affective, anticipated consequences, and ideational, and four types of rejection/acceptance: distaste/good taste, danger/beneficial, inappropriate/appropriate, and disgust/transvalued. Each of these reasons in one form motivates one or more types of acceptance, and in the opposite form motivates one or more types of rejection (see Table 1).

Sensory affective reactions may be biologically based (such as liking of sweet tastes) or acquired. Foods that generate positive sensory affect for any individual are usually acceptable food in this person's culture. Given the key role of sensory aspects in food acceptance (Bonner and Nelson 1985), individual differences in sensory affective responses probably account for most of the variation in food preferences within a culture (Rozin et al. 1986). Food items eliciting a negative sensory affective reaction are placed into the category of "distaste" or even "disgust," while items eliciting a positive sensory affective reaction are categorized as "good tasting" or "transvalued."

Some foods are accepted or rejected as food primarily because of the anticipated consequences of consumption. These consequences can reveal themselves rapidly (e.g., through nausea) or after some delay (e.g., low-fat products and other health-foods). Further, the consequences can be either physiological (such as nausea, pleasant feeling of satiation), psychological (e.g., feeling guilty) or social (unbefitting to one's lifestyle or social class). Anticipated consequences lead to rejection of the food when they imply danger (e.g., cramps, social rejection) and to acceptance of foods when they are beneficial (e.g., social acceptance, getting slimmer).

Some foods are accepted or rejected primarily because of our idea or knowledge of what they are or where they come from. For example, in the movie The War of the Roses, Michael Douglas was enjoying the meat prepared by Kathleen Turner (positive sensory affect) until he learned that it was his dog, which led to instantaneous vomiting. Ideational reasons also lead to less extreme reactions, pertaining to the appropriateness of foods. For example, vegetarians do not consider it appropriate to eat animals. Ideational reasons may become increasingly important in the Western world as objections to food production processes are rising. For example, Steenkamp and Oude Ophuis (1987) found that animal welfare was the most important motivation for eating free range pork. Sensory affective reasons (better tasting) and anticipated consequences (healthier) were of minor importance.

One intruiging category of the Rozin et al. taxonomy is acceptance of food items based on them being "transvalued." Ideational reasons appear to be the key underlying reason, with positive sensory affective reactions being ascribed to transvalued items. A famous historical example is the 14th century's Saint Catherine of Siena who savored the pus from sick people because she loved them so much (and their plight represented a metaphore of Christ's suffering).

The Rozin et al. framework considers acceptance or rejection of generic food products but is less useful to analyze choice of food items or brands within those categories. A substantial body of literature investigates this issue by looking at the role of various psychological factors such as evaluative criteria, perception of brands, perceived risk, risk reduction strategies, the role of nutrition information, advertising and other information sources, country-of-origin, branding, and price-perceived quality tradeoffs. To summarize, it appears that the extent of search effort is limited, few people comprehend and process nutritional information, and the amount of perceived risk for foods vis-a-vis other products is rather low. Many consumers are willing to pay more for better quality, while quality perceptions are predominantly influenced by the perception of the food product on sensory and health-related attributes. Brand name and appearance of the physical product are two key cues in evaluating food quality. See Solms and Hall (1981), Steenkamp (1989), and Thomson (1988) for reviews.

Much of the research on choice between specific food items has employed the Fishbein and Ajzen (1975) framework. Illustrative is the study of Termorshuizen, Meulenberg, and Wierenga (1986) concerning different types of milk (buttermilk, whole milk, etc.) and related drinks (coffee, juice, etc.). In these studies, it is typically found that attitude toward the food item is more important than social influences in predicting food choice (e.g., Termorshuizen et al. 1986, Shepherd and Farleigh 1986), the main exception being when the food item is consumed primarily in a social context (e.g., Birch 1986).

Personality

In this paper, two personality variables will be discussed, viz. variety seeking and quality-consciousness. These variables were selected because for both variables measurement instruments have been developed within the context of foods, they have been validated in different countries, and empirical evidence suggest that each variable has considerable explanatory power.

Consumers offered a selection of foods tend to eat from several of them rather than just the single most favored item. The preference for variety is so powerful that the attractiveness of a favorite dish declines rapidly if it is presented at each meal. Consumer behavior researchers have related variety seeking in general to Optimum Stimulation Level (OSL; see Steenkamp and Baumgartner 1992 for a review). Additional insight into the causes of variety seeking with respect to foods can be found in considering psycho-biological factors as incorporated in Rozin's (1976) "Omnivore's Paradox" (see Figure 2). Omnivores, including human beings, have freedom of choice, but this also means a basic constraint in that a minimal variety in the diet is needed to obtain all necessary nutrients. However, some of these foods may be toxic. The omnivore's paradox arises because on the one hand, diversity in food intake is a necessity for survival, yet on the other hand, the omnivore should protect itself against poisoning by being very conservative in its choices. The tradeoff between the fear and resistance to change ("neophobia") and the need for change, variety, and novelty ("neophilia") creates anxiety. Consumers with high OSL will evaluate anxiety more positively than consumers with low OSL and hence will engage in more variety seeking.

Van Trijp and Steenkamp (1992) developed and validated an 8-item instrument, called VARSEEK, to measure consumers' variety seeking tendency with respect to foods. VARSEEK scores were found to be systematically related to OSL, and more importantly, to variation in food consumption in the Netherlands and Finland (Van Trijp, Lahteenmaki, and Tuorila 1992, Van Trijp and Steenkamp 1992).

Steenkamp (1990) proposed quality-consciousness as a personality factor affecting the quality perception process and the role of perceived quality in consumer behavior. It is defined as a mental predisposition to respond in a consistent way to quality-related aspects which is organized through learning and influences behavior. Steenkamp (1989) developed and validated a measurement instrument for consumers' quality-consciousness with respect to foods, QC-F. QC-F has been used successfully in research on the formation of quality perceptions, store choice behavior, and consumers' willingness to pay for quality, among others. Recently, the present author has tested QC-F in a comparative study involving 10,000 homemakers in nine EC countries. It was found that the functional, conceptual, verbal, measurement, and scalar equivalence (Douglas and Craig 1983) of QC-F was high. Significant differences in quality-consciousness with respect to foods could be observed between countries. The Netherlands and Great Britain rated on average lower than the other countries while Belgium and Germany rated highest. The demographic profile of quality-conscious buyers was quite similar across countries. They were on the average older, with fewer, and fewer young, children (Steenkamp 1992).

FIGURE 2

THE OMNIVOR'S PARADOX

ENVIRONMENTAL FACTORS

Sociocultural factors

What we eat, how it is prepared, the rules and meanings which permeate every aspect of food consumption practices, etc., are all sociocultural matters, irrespective of their biological, psychological, or economic dimensions which they clearly possess (Fischler 1988). Rozin et al. (1986) argued that if one were interested in determining as much as possible about an adult's food preferences, and one could only ask him or her one question, the question would be: what is your culture or ethnic group? In the process of enculturation, exposure to particular foods is controlled and food values and attitudes are conveyed.

By early adulthood, every individual in every culture has adopted a culturally based set of beliefs and attitudes about objects in the world, with respect to their edibility. Which products are considered food or non-food is largely culture-dependent (see Table 2 for some examples). Anthropologists have tried to explain classifications of food and non-food by considering the degree of closeness of the animal to ourselves and to our homes (Murcott 1986). Animals which are either too close (e.g., pets) or too remote (e.g., lions) will not be considered food. Horses are more or less regarded as a sort of pet in the U.S. and Britain and are therefore non-food, while they are considered edible in France and Belgium because they are presumably more remote. The same might be applicable to dogs in Western Europe and the U.S. versus Korea and China. However, this explanation does not appear to be comprehensive. Are snails or frogs closer (or more remote) in Britain or the U.S. than in France? The concept of "too remote" seems particularly ambiguous. For example, ostrich steak is a delicacy in the Netherlands although this animal is very remote from the Dutch culture.

Even if an item is considered food in a culture, some items may be always inappropriate (e.g., pork for Jews and Muslims, beef for Hindus), or forbidden in certain circumstances (e.g., eating during daytime during Ramadan for Muslims). Different explanations have been forwarded to account for these cultural taboos, including functionalist and structuralist views (Fischler 1988). The functionalist view maintains that any cultural trait serves a specific function, and that this function can be explained only by extra-cultural (biological, physical, economic, etc.) factors. For example, the Jewish and Muslim taboo on pork is explained by the contention that in the Middle East, undercooked pork was a source of trichinosis. The structuralist view holds that cultural traits can and must be explained by other cultural traits. A structuralist explanation of the pork taboo is that pork did not fit in the food taxonomies of the Israelites and Arabs, and was therefore rejected as impure. These peoples were pastoralists, herding cattle, sheep, goats, and camels. Cloven-hoofed and cud-chewing, these animals provide a model of proper kinds of food which does not include pork because pigs do not conform to these criteria (Murcott 1986).

The French sociologist Bourdieu argues that differences in food consumption practices within a society are largely based on social class (Bourdieu 1985). He maintaines that food consumption is a vehicle for social stratification, an embodiment of class inequality, and of the stratification of knowledge, esthetic sensibility, and values. Tastes in food, and the distinctions which they signal, are part of a cultural system which enacts the relationships between social classes, and which expresses the values of their members. One example is the consumption of fatty foods which decreases with social class. An underlying reason appears to be that for lower-class people, food choice is predominantly based on sensory affective reasons, while for higher-class people, health-related reasons are also of great importance (DeGraaf and Stafleu 1992). In line with this, Rabier (1990) found in a study involving all EC countries, that prevalence of obesity of both men and women decreased with social class while attention for nutritional information of fat increased with social class.

One aspect common to all cultures relates to food preparation: cooking of food is a universal human practice. Anthropologists have argued that humans cook foods to reassure themselves that they are human, not animal. Leach (1970) argued that when people eat, they establish a direct relation between themselves (Culture) and their food (Nature). Cooking is a means by which Nature is transformed into Culture.

Economic factors

When considering the influence of economic factors on food consumption, two factors are of paramount importance, viz. incomes and prices (Deaton and Muellbauer 1980). [It could be argues that income is a personal factor (cf. Randall and Sanjur 1981). However, for most people, individual incomes are determined by general economic conditions and therefore, it is discussed as environmental factor.] Both factors have been found to affect quantities and types of food bought by consumers. In the 19th century, Engel formulated a "law," stating that "the poorer a family is, the greater the proportion of total expenditure which it must use to procure food. The wealthier a people, the smaller the share of expenditure on food in total expenditure" (Burk 1962). The validity of Engel's law has been suported in many countries and at different times.

TABLE 2

EXAMPLES OF DIFFERENT CLASSIFICATIONS OF ANIMAL SPECIES (FOOD VERSUS NON-FOOD)

The present author tested Engel's law on a cross-section of country aggregates instead of on different income groups within one country. Share of food in total expenditure was regressed on the index of real GDP/capita (i.e., adjusted for differences in purchasing power). [The use of purchasing-power parities is necessary to render income data internationally more comparable.] Data for 85 countries were obtained from The Economist Book of Vital World Statistics (1990). The scatter plot of the data and the regression equation are shown in Figure 3. Real GDP/capita accounted for 62.7% of the variance in food share of expenditure (F(1,83) = 139.4, p < .001). This provides strong support for Engel's law at the level of country aggregates. The regression coefficient indicates that one point increase in the real GDP/capita index leads to a reduction in food share of expenditure of .394%.

Engel's law deals with aggregate food expenditure. His law cannot be directly applied to quantities consumed of individual foods as the income elasticity of food consumption varies considerably between different foods. The consumption of some foods, called "inferior goods," decreases with increasing income. Examples are basic foodstuffs like potatoes and cereals. The consumption of more luxurious items, such as fresh fruits, meat, and cheese, increase with rising income. However, the income elasticity of most food products is below unity (Ritson 1977), which is consistent with Engel's law at the aggregate level.

In most cases, price elasticity of demand is negative, implying that consumption decreases with price increases (after adjustment for changes in the general price level). Cross-price elasticity may be positive or negative, dependent on whether the two goods are substitutes or complements. In general, food consumption has been found to be rather price inelastic, which is plausible considering that food is both a necessity and a saturation good (Tangermann 1986).

Three exceptions to the general rule that price elasticities are negative are the "Giffen" case, the "Veblen" effect, and the "perceived quality" effect. The Giffen case, which was first observed in the 19th century by Sir Robert Giffen for potatoes in Ireland and bread in England, occurs when a certain basic food item which accounts for a high share of total expenditure, increases in price. This reduces the purchasing power of (poor) families, so that they can no longer afford to consume "luxury" items (i.e., everything else) and have to resort to even higher consumption of the basic food item whose price has increased. The Giffen case has lost much of its relevance for developed countries but may still apply to some income groups in developing countries. The Veblen effect is caused by the phenomenon of conspicuous consumption. For example, Chivas Regal whisky or Beluga caviar may lose much of their appeal when the price would be substantially lowered. The third case of non-negative price elasticities concerns the "perceived quality" effect. Empirical evidence (e.g., Monroe 1973) suggests that when prices are very low, demand falls because the quality of the product becomes suspect. The latter two phenomena are more likely to occur for individual brands than for generic food categories.

Marketing factors

Food as such, i.e., the physical material contained in food, is necessarily subject to saturation. This is much less the case for the extent to which value is added to the food product through marketing activities. When incomes rise, consumers prefer frozen products to canned foods, Coca-Cola to Royal Crown Cola, meat from a specialty store to meat from a discount supermarket, etc. The income elasticity of demand for raw product at the farm level is much lower than the income elasticity for "marketing activities" (Tangermann 1986). This implies that with rising incomes, the share of farm-level value in food value at the retail level decreases, while the margin for marketing activities increases. In Germany, for example, the share of farm-level value decreased in 1964-1981 from 50.8% to 43.6% (Tangermann 1986). This explains why food industry is still growing while there are definite limits to the amount of food that can be consumed by a person.

One prominent way to add value to the food product is by introducing a brand or strengthening an existing brand. It has been observed that: "It will be more important to own markets than to own factories. The only way to own markets is to own market-dominant brands." (Aaker 1991, p. ix). Brand equity provides value to consumers by helping them to interpret and process information, by increasing their confidence in the purchase decision, and by increasing the use satisfaction because the product obtains an emotional, expressive dimension. This value is created by improving the quality image of the brand, and by enhancing name awareness, brand loyalty, and positive brand associations (Aaker 1991). All elements of the marketing mix may be employed to create brand equity. Strategies for improving quality image include (but are not limited to) selective distribution, high prices, and consumer-oriented product development. Name awareness can be increased by e.g., distribution and advertising, brand loyalty by promotional programs and after sales service, and positive brand associations by advertising.

FIGURE 3

RELATION BETWEEN REAL GDP/CAPITA (U.S.=100) AND SHARE OF FOOD IN TOTAL EXPENDITURE FOR 85 COUNTRIES

One marketing tool, viz. packaging, deserves special mention as some interesting developments can be observed that may lead to dramatic changes in the use of this tool in the future, at least in Europe. Due to waste disposal and pollution problems, and because of increased environmental consciousness, packaging is rapidly becoming an important issue in public policy. One may expect less packaging in general and increased demand for degradable and recyclable packagings. In Germany, pressure by environmentally conscious consumers has recently led to legislation allowing consumers to leave "unnecessary" packaging behind in the supermarket. Some German supermarket chains have reduced the number of items with double packagings (e.g., canned tuna fish in a cardboard box) by 80%. Packaging will be increasingly evaluated on its functional necessity. Environmentally safe packagings are becoming an increasingly important marketing tool, and may be used to project an environmentally conscious image.

Food consumption trends

Given the increase in tourism, the rapid dissemination of information through mass media, and public attempts to modify food consumption based on new insights from medical science, it may be expected that some parallel trends in food consumption can be uncovered in different countries. Uhl (1992) considered this issue for the two largest food markets in the world, the EC and the U.S. Trends in food consumption over the last two decades were analyzed for 45 food products (poultry, beer, vegetable oils, cheese, etc.). He concluded that for 88% of the products studied, similar trends could be observed, i.e., consumption was either rising, falling, or stable in both markets. Generally, the products that show growth in consumption have either a healthy image or are considered to be lifestyle products (e.g., poultry, pasta, frozen foods). Many of the declining products in both markets are staple commodities with low ego involvement or unhealthy images (e.g., fresh potatoes, whole milk). Uhl's (1992) study suggests that health concerns and lifestyle may be two key determinants of food consumption in the future, at least in the U.S. and the EC.

CONCLUSIONS

This paper developed a taxonomy of determinants of food consumption behavior, based on a review of the food literature. The determinants were categorized as related to the food, the person, and the environment. A large number of disciplines has studied the influence of various determinants of food consumption behavior. This research has yielded important insights into the role of factors as diverse as age, satiety capacity of foods, and cultural taboos, which were reviewed in this paper. However, most of the research tends to be partial, typically concentrating on a single factor or a couple of related factors. Repeatedly, it has been pointed out that food consumption determinants interact, and that a comprehensive approach would add considerably to our understanding of food consumption behavior. However, there is a lack of research that takes a broad approach, aimed at integrating determinants from several categories. Developing such a research program is a major challenge for food consumption behavior research in the future. It is believed that the discipline of consumer behavior can play a key role in this programmatic research, given its interdisciplinary background. Consumer behavior researchers can contribute to at least four interdiciplinary research issues.

First, future research should investigate how physical characteristics of the product and social/cultural factors jointly shape perceptions of foods. This involves an extension of the psychophysical model to include social psychophysical cues as well. Research along these lines is of great importance for new product development as it links marketing, R&D, and food technology, thus increasing the chances of a successful product introduction (cf. Hauser and Simmie 1981).

Second, the interaction between cognitive and cultural factors and physiologically-induced effects needs research attention. It will be important in determining the demand for and effectiveness of low-calorie products, and dieting programs in general. Third, physiological influences may act as situational variables in consumer behavior (cf. Belk 1975), and interact with other determinants in shaping food perceptions. For example, consumers tend to buy more when they are hungry than when they are satiated. Physiological factors also affect psychophysical relations (see Huber, Holbrook, and Schiffman 1982 for one of the few marketing studies). Despite the obvious implications (e.g., retail environment, sport drinks), interactions involving physiological factors have not received much attention by consumer behavior researchers. A fourth research issue concerns cross-cultural research on determinants of food consumption behavior. For rigorous testing of consumer behavior paradigms, and for effective marketing of foods, it is important to have insight into similarities and differences between consumers in different countries.

REFERENCES

Aaker, D.A. (1991), Managing Brand Equity, New York (NY): The Free Press.

Belk, R.W. (1975), "Situational Variables and Consumer Behavior," Journal of Consumer Research, 2 (December), 157-167.

Birch, L.L. (1986), "The Acquisition of Food Acceptance Patterns in Children," in Eating Habits: Foods, Physiology and Learning Behavior, eds. R.A. Boakes, D.A. Popperwell, and M.J. Burton, Chicester: Wiley, 107-130.

Blundell, J.E., A.J. Hill, and P.J. Rogers (1988), "Hunger and the Satiety Cascade - Their Importance for Food Acceptance in the Late 20th Century," in Food Acceptability, ed. D.M.H. Thomson, London (UK): Elsevier Applied Science, 233-250.

Bonner, P.G. and R. Nelson (1985), "Product Attributes and Perceived Quality," in Perceived Quality, eds. J. Jacoby and J.C. Olson, Lexington (MA): Lexington Books, 65-79.

Booth, D.A. (1982), "Normal Control of Omnivore Intake by Taste and Smell," in Determination of Behavior by Chemical Stimuli, eds. J. Steiner and J. Ganchrow, London (UK): Information Retrieval, 233-243.

Bourdieu, Pierre (1985), Distinction: A Social Critique of the Judgment of Taste, London (UK): Routledge & Kegan Paul.

Burk, M.C. (1962), "Ramifications of the Relationship between Income and Food," Journal of Farm Economics, 44 (1), 115-125.

Cowart, B.J. (1981), "Development of Taste Perception in Humans: Sensitivity and Preference throughout the Lifespan," Psychological Bulletin, 90 (1), 43-73.

Deaton, A. and J. Muellbauer (1980), Economics and Consumer Behavior, Cambridge: Cambridge University Press.

DeGraaf, C. (1992), "Determinanten van Voedselkeuze" (Determinants of Food Choice), in Voeding van Elke Dag (Everyday Food), eds. C. DeGraaf, C.P.G.M. de Groot, W.A. van Staveren, and J.C. Seidell, Houten (Netherlands): Boon (forthcoming).

DeGraaf, C. and A. Stafleu (1992), "Houdingen ten aanzien van Vet in Europees Perspectief" (Attitudes with respect to Fat in European Perspective), in De Europese Consument van Voedingsmiddelen in de Jaren 90 (The European Consumer of Foods in the 90s), ed. J.-B.E.M. Steenkamp, Assen (Netherlands): Van Gorcum, 23-29.

DeGraaf, C. and J.E.R. Frijters (1989), "Interrelationships among Sweetness, Saltiness and Total Taste Intensity of Sucrose, NaCl and Sucrose/NaCl Mixtures," Chemical Senses, 14 (1), 81-102.

Douglas, S.P. and C.S. Craig (1983), International Marketing Research, Englewood Cliffs (NJ): Prentice Hall.

Fischler, C. (1988), "Cuisines and Food Selection," in Food Acceptability, ed. D.M.H. Thomson, London (UK): Elsevier Applied Science, 193-206.

Fishbein, M. and I. Ajzen (1975), Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research, New York (NY): Addison-Wesley.

Frank, R.A. and G. Archambo (1986), "Intensity and Hedonic Judgments of Taste Mixtures: An Information Integration Analysis," Chemical Senses, 11 (3), 427-438.

Hauser, J.R. and P. Simmie (1981), "Profit Maximizing Perceptual Positions: An Integrated Theory for the Selection of Product Features and Price," Management Science, 27 (January), 33-56.

Huber, J., M.B. Holbrook, and S. Schiffman (1982), "Situational Psychophysics and the Vending-Machine Problem," Journal of Retailing, 58 (Spring), 82-94.

Leach, E. (1970), LTvi-Strauss, Glasgow (UK): Collins.

Martens, M. and H. Martens (1986), "Partial Least Squares Regression," in Statistical Procedures in Food Research, ed. J.R. Piggott, London (UK): Elsevier Applied Science, 293-361.

Monroe, K.B. (1973), "Buyers' Subjective Perceptions of Price," Journal of Marketing Research, 10 (February), 70-80.

Murcott, A. (1986), 'You Are What You Eat: Anthropological Factors Influencing Food Choice," in The Food Consumer, eds. C. Ritson, L. Gofton, and J. McKenzie, Chichester (UK): Wiley, 107-125.

Pilgrim, F.J. (1957), "The Components of Food Acceptance and Their Measurement," American Journal of Clinical Nutrition, 5, 171-175.

Rabier, J.R. (1990), "Nutrition and Cancer Prevention: Weightwatching and Cutting Down on Fatty Foods," Proceedings of the First European Conference on Food and Nutrition Policy, Budapest (Hungary): Commission of the European Communities.

Randall, E. and D. Sanjur (1981), "Food Preferences: Their conceptualization and Relationship to Consumption," Ecology of Food and Nutrition, 11, 151-161.

Ritson, C. (1977), Agricultural Economics: Principle and Policy, London (UK): Collins.

Rogers, P.J. and J.E. Blundell (1990), "Psychobiological Bases of Food Choice," in Proceedings of the 12th British Nutrition Foundation Annual Conference, ed. Margaret Ashwell, London (UK): BNF, 31-40.

Rozin, P. (1976), "The Selection of Foods by Rats, Humans, and Other Animals," in Advances in the Study of Behavior, Vol. 6, eds. J.S. Rosenblatt, R.A. Hinde, E. Shaw, and C. Beer, New York (NY): Academic Press: 21-76.

Rozin, P., M.L. Pelchat, and A.E. Fallon (1986), "Psychological Factors Influencing Food Choice," in The Food Consumer, eds. C. Ritson, L. Gofton, and J. McKenzie, Chichester (UK): Wiley, 85-106.

Rozin, P. and T.A. Vollmecke (1986), "Food Likes and Dislikes," Annual Review of Nutrition, 6, 433-456.

Schiffman, S. (1979), "Changes in Taste and Smell with Age: Psychophysical Aspects," in Sensory Systems and Communication in the Elderly, eds. J.M. Ordy and K. Brizzee, New York (NY): Raven Press, 227-246.

Shepherd, R. (1985), "Dietary Salt Intake," Nutrition and Food Science, 96 (September/October), 10-11.

Shepherd, R. (1990), "Overview of Factors Influencing Food Choice," in Proceedings of the 12th British Nutrition Foundation Annual Conference, ed. M. Ashwell, London (UK): BNF, 12-30.

Shepherd, R. and C.A. Farleigh (1986), "Attitudes and Personality Related to Salt Intake," Appetite, 7, 343-354.

Solms, J. and R.L. Hall (1981), Criteria of Food Acceptance, Zurich (Switzerland): Foster Publishing.

Steenkamp, J.-B.E.M. (1989), Product Quality, Herndon (VA): Books International.

Steenkamp, J.-B.E.M. (1990), "A Conceptual Model of the Quality Perception Process," Journal of Business Research, 21 (4), 309-333.

Steenkamp, J.-B.E.M. (1992), "Cross-Nationale Analyse van het Kwaliteitsbewustzijn van Consumenten met betrekking tot Voedingsmiddelen" (Cross-National Analysis of Quality-Consciousness with respect to Foods), in De Europese Consument van Voedingsmiddelen in de Jaren 90 (The European Consumer of Foods in the 90s), ed. J.-B.E.M. Steenkamp, Assen (Netherlands): Van Gorcum, 127-136.

Steenkamp, J.-B.E.M. and H. Baumgartner (1992), "The Role of Optimum Stimulation Level in Exploratory Consumer Behavior," Journal of Consumer Research, 19 (December) (forthcoming).

Steenkamp, J.-B.E.M. and P.A.M. Oude Ophuis (1987), "Consumer Behavior with respect to Free Range Pork," Netherlands Journal of Nutrition, 48(9), 271-276 (in Dutch with English summary).

Steenkamp, J.-B.E.M. and H.C.M. van Trijp (1989), "Quality Guidance: A Consumer-Based Approach for Product Quality Improvement," in Proceedings of the 18th Annual Conference of the European Marketing Academy, ed. G.J. Avlonitis, Athens (Greece): EMAC, 717-736.

Steiner, J.E. (1979), "Human Facial Expressions in Response to Taste and Smell Stimulation," in Advances in Child Development, Vol. 13, eds. H.W. Reese and L.P. Lipsitt, New York (NY): Academic Press, 257-295.

Stevens, S.S. (1975), Psychophysics: Introduction to its Perceptual, Neural, and Social Prospects, New York (NY): Wiley.

Tangermann, S. (1986), "Economic Factors Influencing Food Choice," in The Food Consumer, eds. C. Ritson, L. Gofton, and J. McKenzie, Chichester (UK): Wiley, 61-83.

Termorshuizen, J.G., M.T.G. Meulenberg, and B. Wierenga (1986), "Consumer Behavior in respect of Milk in The Netherlands," European Review of Agricultural Economics, 13 (1), 1-22.

The Economist Book of Vital World Statistics (1990), London (UK): The Economist.

Thomson, D.M.H. (1988), Food Acceptability, London (UK): Elsevier Applied Science.

Uhl, J.N. (1992), "Comparison of Food Consumers' Behavior in the U.S. and the E.C.: Implications for Research, Policy, and Managment," in De Europese Consument van Voedingsmiddelen in de Jaren 90 (The European Consumer of Foods in the 90s), ed. J.-B.E.M. Steenkamp, Assen (Netherlands): Van Gorcum, 87-109.

Van Toller, S. and M. Keller-Reed (1990), "Infant Response to Food Odors: Does This Implicate Early Preference?," in Proceedings of the 12th British Nutrition Foundation Annual Conference, ed. M. Ashwell, London (UK): BNF, 41-49.

Van Trijp, H.C.M., L. Lahteenmaki, and H. Tuorila (1992), "Variety Seeking in the Consumption of Spread and Cheese," Appetite, 13, 155-164.

Van Trijp, H.C.M. and J.-B.E.M. Steenkamp (1992), "Consumers' Variety Seeking Tendency with respect to Foods: Measurement and Managerial Implications," European Review of Agricultural Economics, 19, 181-195.

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