ALICOM 99/16

Conference on International Food Trade
Beyond 2000: Science-Based Decisions, Harmonization, Equivalence
and Mutual Recognition
Melbourne, Australia, 11-15 October 1999

Prospects for the Future: Nutritional, Environmental and Sustainable Food Production Considerations - Changes in Cultural and Consumer Habits


W. Bruce Traill, Professor of Food Management and Marketing, University of Reading, UK

Table of Contents

I. Introduction

1. The stage of a country's economic development is of fundamental importance in determining what types of food its population can afford to choose and therefore the type of product supplied, its packaging, promotion and distribution by farmers, manufacturers, wholesalers and retailers. At low income levels where the emphasis is on survival and the food marketing infrastructure is poorly developed, there is little demand for or supply of ready meals, organic food and animal welfare friendly products. Rises in income translate into shifts in food consumption from starchy staples towards `luxury' goods such as sugar and meat. In a series of articles on the geography of world food consumption, Grigg (1992, 1993a, 1993b,1995, 1996) shows a strong positive relationship between GNP per capita and animal protein consumption (80% of inter-country variation explained), but a negative relationship between GNP and plant protein consumption; he also observes a strongly negative relationship between the proportion of calories coming from starchy-staples and per capita GNP at low incomes, but the proportion stabilises at around 30% at about $US15,000 per capita. By contrast within wealthy western economies, changes in income translate into a demand for `quality' (which, like beauty, is in the eye of the beholder), convenience, eating out etc.

2. In this paper, most attention is focused on the developed country consumers. Partly this is because it is the area that the author knows best (particularly with respect to western Europe, from which perhaps too many of the examples are consequently drawn), but also because the changes in consumption at very low income levels are relatively straightforward and well understood. By way of contrast, the driving forces for change in food consumption in the modern developed country, where an individual retail outlet may stock 20,000 different products and introduce 3,000 new ones every year, are more complex. They also represent the future, however long term, for the developing nations.

3. The paper describes the food choice decision and the factors that influence food consumption. It attempts to relate these to current and future food consumption tendencies.

II. An Organising Framework for Analysing Food Consumption

4. There have been many models of food consumption and they have tended to depend on the interests of the researcher and the purpose of the research: for example, economists are interested mainly in the effects of prices and incomes at the market level, often with a view to assessing the effects of government policies (e.g. taxes and subsidies) or defining suitable market intervention strategies (e.g. for price stabilisation). Psychologists are interested in modelling the individual food choice decision, for example linking consumers' needs recognition, information search and choice mechanisms. Their models are used by firms in new product development and by governments interested in promoting healthier eating. Sociologists, anthropologists and geographers have all attempted to analyse the importance of culture and the spatial pattern of food consumption. Market researchers search for groups of consumers with similar food-related behaviour and attempt to design, distribute and promote products targeted to the peculiarities of specific market segments.

5. It is impossible to do full justice to all of these disciplines and their interests and the author is in any case not qualified to do so: he is an economist who has strayed into marketing, product development and nutrition policy, but never into sociology or anthropology. Nevertheless, it is possible to provide a framework for analysing food consumption within which all of these individual disciplines find a place. It is then possible to speculate, using this framework, on some of the future developments in eating and drinking.

6. The organising framework is presented in Figure 1-though it should be recognised from the outset that some of the boundaries between the differing groups of influencing factors are rather more fuzzy that the diagram suggests. The central box shows the individual food choice decision process leading to the choice of individual foods and ultimately to an entire diet. Feeding into the decision process are a number of influences. First of all the physiological and sensory properties (or the perceived properties) of foods affect an individual's choice (what are its nutritional properties? Is it perceived to be safe? does it smell nice, look attractive?). Individual, person-related factors have an important bearing on why people choose different foods even if they perceive their physiological and sensory properties in the same way. Someone overweight and on a diet chooses different foods than an athlete; young people have different needs than the elderly; psychological concerns such as the ethics of meat consumption, attitudes to technology or the importance attached to food in an individual's life all influence the food that person consumes (and where?, when?, how much? and with whom?).

7. Socio-demographics such as whether an individual lives in a town or city or the country, their level of education, family composition etc. are also important. Finally, there are environmental factors, economic, marketing and cultural. The first two are related: a wealthy country with a well-developed food marketing system offers abundant choice and the means to pay for `value-added' (i.e. expensive) products. However, it is worth maintaining marketing and economics as separate variables in recognition that different marketing systems may exist at the same level of economic development and this influences food availability and choice. For example, the United Kingdom and Italy are economically comparable, but the former has a highly concentrated retailing and food manufacturing system whereas in the latter it is highly fragmented. Thus in Italy, the range of products available and the rate of new product development is much lower than in the UK-so consumers in the UK have more choice. Some might argue that UK consumers would be better off with less choice and more quality!

Figure 1: Conceptual model for consumer behaviour with respect to food

Undisplayed Graphic


Needs Recognition

8. A so-called need arises when there is a discrepancy between a desired state and an actual state. Some might call this hunger, but hunger is not the only need that food satisfies and depleted food reserves in the body is not the only way in which the desired and actual state may diverge (Steenkamp, 1996). For example, a need may be recognised for a long-lasting energy supply before running a marathon, or for fruit, vegetables, olive oil and red wine to ward off cancer and heart disease. Dissatisfaction with an existing product (resulting from a gap between expectations and reality) may cause a gap between the actual and desired state.

9. The desired state may be influenced by culture, lifestyle, demographics (migrate to a city; have a child), by product experience (the popularity in Northern Europe of Mediterranean food following tourism experiences), or by marketing/advertising.

Search for Information

10. Information search-looking for alternative solutions to eliminate a need--is relatively unimportant for food compared to more complex and expensive items such as automobiles, computers, TVs, homes (Engel et. al., 1995). Quality differences are typically small, innovations relatively minor, price changes from one shopping experience to another are not generally significant and the time available to purchase a week's shopping before other duties call or boredom sets in is strictly limited. Consequently, for the vast bulk of their purchases, shoppers rely upon past experience with products. This is not to say, however, that consumers don't demand information, nor that information, such as nutritional labelling, doesn't influence food consumption decisions. In fact, there is a clamour in Western European countries for the labelling of foods containing genetically modified products. Manufacturers and retailers have long recognised the benefits to their sales of selected product information (often, where regulation allows, related to medicinal and/or health claims-`this product may reduce the risk of coronary heart disease'-or, where such claims are not permitted, related to nutritional claims or implicit health claims-`high in oat bran'; `full of energy' (NCC, 1997).

11. The fact that consumers are not prepared to spend a lot of time searching for information on individual foods explains their demands for simple labelling that they can trust. For example, a uniformly presented nutritional label that can be trusted reduces the time needed to interpret the label and act on the information. A universally recognised symbol to indicate that food has been irradiated or contains genetically modified ingredients saves the concerned consumer having to read every label to find this information.

12. Problems arise when consumers in one country demand different information than consumers in another, or where governments adopt differing ways of presenting the information. American consumers seem to be unconcerned about genetic modification and so are not interested in the labelling of such products, whereas European consumers are concerned and so do demand to be informed. Is the outcome (different systems in each country) a constraint to free trade or a recognition of diverse consumer demands for information?

13. Problems also arise when legislators in one country adopt a system (e.g. of nutritional labelling) that to them seems `best' (and perhaps is for their own consumers), but is different from the systems adopted in other countries (which those other countries also consider `best'). Given that consumers want a uniform labelling system, laissez-faire is not a satisfactory solution, but resulting attempts to require imported products to conform to domestic labelling legislation may be construed (sometimes reasonably) as a trade barrier.

14. That consumers are not prepared to spend a lot of time searching for product information also partly explains their demands for someone else (usually governments, sometimes retailers) to ensure that food is safe. Why, for example, should every individual consumer feel obliged to learn all about safe levels of a huge range of chemical contaminants in food when the task can be delegated to government legislators? Consumers expect that the food they purchase is safe (sometimes they expect an unreasonable level of completely risk free safety), so when sometimes they are let down (e.g. salmonella in eggs, BSE-both in the UK) they over-react and demand further legislative control-again, in a form that could act as a barrier to trade. That consumers in different countries may have different tolerances to safety issues is also a problem. The French seem prepared to accept the higher risk of listeria associated with eating soft unpasteurised cheeses, the Americans do not. Is labelling the solution, or international agreement?

Evaluation of Alternatives

15. There is an enormous literature on the evaluation criteria used by consumers. Most straightforward are the various models that presume consumers view foods as consisting of a bundle of attributes. For example, different beers may be characterised by their alcohol content, their bitterness (and presumably other flavour characteristics), their colour, packaging, promotion, place of sale and price. Different consumers attach different importance to the attributes and therefore make different choices. A demand for variety explains why an individual does not always make the same choice.

16. Certain attributes are product specific, such as those suggested for beer above or the fatness of meat, the perceived freshness of fruit and vegetables. Other attributes are more generic and suggest items that guide consumers in all of their consumption behaviour. Table 1 shows the result of a survey of consumers in 7 European countries on the importance of a range of attributes to their product choice (not just food). Although it is common to find a discrepancy between what consumers say is important and how they behave in the market place, the results are nevertheless informative.

17. That nebulous concept, quality, comes out on top, followed closely by price. Brand names are a form of guarantee to consumers, but in the food area it is apparent that other guarantees of product quality are becoming important, notably labels of geographic origin or other quality guarantees, such as farm-assured meat.

Table 1. Relative Importance of a Number of Product Attributes in Product Choice










Product quality


















Brand name/reputation













































Choice of several sizes









Ease of use









Possibility of making selections


















Practical packaging









Good state of packaging









Variety of brands


















Source: AGB/Europanel 1992.

18. Consumers have perceptions about the quality of particular products from particular geographic sources. For example, European wine has long been geographically controlled through the appellation d'origine controllee and its equivalents, as have cheeses. Recent European legislation has enabled a wider range of products to claim geographic origin `exclusive rights' to a name such as `Ardennes pate' or `Parma ham' (though not to products that have become generic such as cheddar cheese). A monopoly on a name is a valuable marketing tool, but can also act as a barrier to competition, domestic and international. The battles in Europe over Feta cheese were illuminating-there is no doubt that if Greece could have re-appropriated the name, consumers everywhere would have had to pay more. It is doubtful whether the objective quality of the product would be much different (though arguably it would remove from the market the risk of purchasing low-grade product which appears identical and can only be detected as inferior when eaten). But consumers may gain utility from the knowledge that the product they consume is `authentic'.

19. The liberal's solution to the dilemma would be to offer consumers choice through country (or region) of origin labeling, but in an age where an ever-growing share of food consumption is outside the home (and unlabelled), this solution may be seen as inadequate. Legislators face a difficult task of balancing the demands/rights of domestic and foreign producers and consumers, but are inclined, perhaps naturally to give most weight to domestic producers. Only international fora are able to give an objective assessment of all the actors' needs. Naturally there will be conflict between those countries that have internationally recognisable product-place associations (mainly Europeans, but also Argentine beef, Kenyan coffee, Indian tea, Australian wine) and those that don't, but wish to break into a new product area-often the developing countries.

20. Farm assurance schemes are becoming popular in Europe, most notably for beef following the BSE fiasco. Generally these involve agreements between retailers, manufacturers and farmers and cover a range of `credence characteristics'-attributes of food that are valued by consumers, but not objectively recognisable in the product (e.g. whether an animal has been treated kindly or whether environment friendly farming methods have been employed). Agreements may cover stocking density, feeding regimes, transport of live animals, slaughtering procedures etc. Usually the products are sold as retailers' own- (private-) label where they contribute to the overall store image and enable stores to offer their customers a range of price/quality options. Consumers are effectively guaranteed that the product they buy contains attributes that they value, even though they are unable personally to detect them.

21. Valued attributes with a high income elasticity of demand (or associated for other reasons with income) include organic food, lite (low calorie) products, safety, low fat, functional foods, convenience, freshness (e.g. chilled fruit and vegetables), ecological products (including recycling), animal welfare and additive (including GM ingredient) free. These are the types of products that developing countries will need to set up systems to produce and sell if they are to compete in developed country food markets. Their own consumers, particularly the better off, will also increasingly demand these products.

22. One guarantee that is becoming internationally recognised, is the HACCP system. This is mainly a food safety guarantee used by retailers (or further manufacturers) to ensure that their suppliers enable them to meet their consumers' expectations and governments' regulatory requirements. Developing country suppliers will need to recognise this requirement for trading with the developed world.

23. Attribute-based models of food choice are still the most common, but two others deserve mention, means-end chains (mec) and lifestyle models. The former argues that consumers don't purchase foods because of their attributes per se, but for the consequence of those attributes and what they contribute to consumers' ultimate values. For example, a consequence of consuming food with the attribute that it is ready prepared is that it saves time; a (perceived) consequence of consuming organic food may be that it is safer and/or tastes better and/or contributes to the environment; a (perceived) consequence of eating farm-assured beef may be that it tastes better. According to means-end chain theory, consequences are important if they contribute to the ultimate values of consumers such as security, family values, fun and enjoyment, social recognition. These relatively stable values, which vary by consumer and by culture, are what guide consumption. Large companies (e.g. Unilever) have adopted mec to guide product development, target groups of consumers with similar values (and who perceive similar links between attributes, consequences and values), and to guide their promotion messages (food and drink advertising in Western Europe is increasingly life-style oriented, showing how a product will contribute to ultimate consumer values such as a happy family life, sexual prowess or fame and wealth).

24. Life-style modeling is very similar to mec, and assumes that food is chosen to fit in with one's chosen lifestyle, which of course is once again influenced by values (Grunert, et al, 1993). Shopping, eating out and food preparation are all a part of life-style. Table 2 suggests 5 lifestyle types from a European survey, and their characteristics. Grouping consumers according to their life-styles is deemed more useful than traditional socio-demographic segmentation techniques. For example, the market intelligence agency Mintel suggests that consumers in London, Paris and Milan have more in common with one another than with their fellow countrymen. Companies are attempting to uncover such cross-country groupings of consumers with similar life-styles that can be targeted with similar products and marketing strategies.

25. What remains unproven with all of these models is a strong ability to predict actual food consumption behaviour. Thus, while one may form hypotheses such as that `fast fun lovers' eat out more than `traditionalists' and perhaps even eat less meat, it remains to be shown that an old-fashioned model of consumption based on prices, incomes and demographic characteristics would predict any less well. It is possible that life-style and value-based models are more useful for fine-tuning the promotion, distribution and packaging parts of the marketing mix than for predicting consumption of different food products.

26. In fact, the link between food evaluation by consumers and their choice of food is blurred by a number of factors, not least habit (e.g although most foods are now available all year round in the developed world, consumption of strawberries is much higher during the local growing season than at other times of the year (Traill and Righelato, 1996) and turkey consumption is low except at Xmas despite turkey being available, cheap and nutritionally valuable throughout the year (Steenkamp, 1996)). Variety seeking is another reason that consumers do not choose their favourite food all the time (van Trijp, 1996).

Table 2: Description of Five Pan-European Lifestyle Segments


Fast fun lovers

Controlled elitists











Young or middle-aged; urban; average income

35-64 yrs old; good education; high income

Young couples with average income

45-64 yrs old; middle class

Young people; high level of studies


Have money to spend it

Civil rights; morality

The family; children; comfort

Order; the family; moral values

A better world and more social justice


Individualists and innovators





Priority expenditure

Look and pleasure

Up-market products [financial investments, etc.]

Comfort [the home, etc.]

Basic products

Culture and leisure; luxury products

Major interests




The home and the family



Fast; innovatory; exotic

Gourmet food; high quality food

Modern/ traditional

Traditional; structured; home-made

Quality; variety; facility


Alcoholic drinks; mineral water; soft drinks

Good wines

Limit oneself

Not to excess

Wide choice of alcoholic drinks


The most practical [supermarkets, shopping centres]

Small specialist stores

Modern stores; customer assistance

Supermarkets; local stores

Quality, facility


Choice; novelty; attractive packaging

The brand; quality

Confidence in brands

Finding a bargain

Choice; good quality/price relationship


Fun, fashion

Traditional broadcasting

Dreams and utility

TV and entertainment

News; entertainment


Cultural stereo-types [American dream, etc.]

Prestige and information

Model family

Simple; personal account

Brand; creativity

Source: AGB/Europanel 1992


27. Physiological factors (e.g. hunger) and sensory properties (e.g. taste and appearance) are important determinants of an individual's food consumption. Protein has a higher satiation effect than the equivalent calorific value of carbohydrate or fat and the satiation effect of alcohol is minimal. In general, animals (including humans) are more tolerant of over-supply of energy than under-supply (Rogers and Blundell, 1990), an evolutionarily sound strategy, but one that leads to obesity in modern western states where food and alcohol are easily affordable. To the extent that the temptation to consume too much is related to the tantalising array of foods available to developed-country consumers, there is little doubt that economic development elsewhere will result in the spread of obesity, albeit that lessons learned in the West (the value of exercise in weight control, the ability to reverse trends in CHD by education, exercise and diet) will not have to be re-learned from scratch in the developing world.


28. We have already discussed many of the individual person-related factors that influence individual food consumption behaviour, such as lifestyle, value system, quality and environmental consciousness. Some of these are influenced by culture and we will therefore return to them. It is worth additionally mentioning socio-demographic factors such as age, gender, employment status (of husband and wife), education and family size. As households become smaller and wives as well as husbands work, time saving is a major influence on food consumption behaviour, resulting in a seemingly endless trend towards convenience foods (prepared and semi-prepared meals), take-away meals and consumption outside the home. It is not known to what extent food consumed outside the home is nutritionally or otherwise different from food consumed at home. It is somewhat ironic that, at a time when consumers are demanding more and more control over the food they eat by requiring more information (through food labels) on nutritional composition, environmental and animal friendliness, absence of GM (and other) ingredients or brands which give a guarantee of safety and quality, they are consuming more foods in restaurants, canteens etc. over which they have very little knowledge and control.


29. Having discussed the food choice process and modern ideas about how and why certain foods are chosen according to values and lifestyles, it remains true that an enormous amount of food consumption can be explained by economic factors (incomes primarily) and culture (country or region of birth). The two cannot be totally divorced: for example, a culture that values freedom and individuality is more likely to have a liberal economic system that tolerates large food manufacturers and retailers (that are an important influence on what we eat) than one which values collective responsibility; a high level of income leads to foreign travel that leads to an awareness of and interest in consuming the foods of other cultures and a breakdown (or reduction in strength) of one's own food culture.

Economic Factors

30. Economic factors are all pervasive in influencing food consumption. They affect what you can afford to buy, what life-style you can afford to live, where you shop (without a car it is difficult to shop in an out of town superstore), what home appliances-- micro-waves, freezers etc--you own (and therefore what convenience foods you buy), your level of education (which is linked to attitudes to the environment, organic food, biotechnology etc.).

31. The most well known relationship is between income and the proportion of expenditure on food, Engel's Law stating a negative relationship (Figure 2).

Figure 2:
Undisplayed Graphic

Source: Steenkamp, 1996.

32. The law holds within a country (the poor spend a higher proportion of their income on food than the rich in the same society), and at the aggregate level (poor countries spend more of their GNP on food than wealthy ones-in the United States the proportion is around 10%, but can be over 50% in very poor countries).

33. Although the share of expenditure on food decreases with income, total expenditure per capita on food does rise as income per capita rises as people trade up to higher value products. Figure 3 shows trigger levels employed by Unilever to adjust its product portfolio to a country's income level.

Figure 3: Trigger levels of GDP per capita for different types of foods

34. To say that the proportion of income spent on food falls as income rises is the same as to say that the aggregate income elasticity of demand for food is less than one or inelastic. It is also true that as incomes rise, people become less price sensitive-the price elasticity of demand becomes closer to zero. Table 3 shows that price is less elastic in higher income UK than lower income Spain for a range of food products.

Table 3: Compensated price elasticities in Spain and the UK, 1996






















35. This has the implication that price reductions brought about by reform of agricultural support policies in Western economies to comply with GATT\WTO agreements nowadays have relatively little impact on consumption of food, at least at the aggregate product level-it may still cause substitution at a less aggregate level, for example among individual meats where price and cross-price elasticities are high.

36. Naturally it follows from what has been said already that although the income and price elasticities for food per se may be low in wealthy countries, the elasticities with respect to the value-adding activities associated with food are much higher. In other words, consumers are willing to pay more for the packaging, processing and service elements of food consumption as they get richer. This is one reason for the relatively low variation in retail prices in response to large proportional changes in farm-gate food prices.

Cultural factors

37. Connor 1991, p2 argues that `all consumers are basically alike' in the sense that consumers `with the same incomes and socio-demographic characteristics, facing the same relative prices, and holding the same information, will tend to choose the same basket or array of goods'. Thus, as incomes, prices and demographic factors in Europe catch up with developments in North America, so food consumption patterns will converge. Connor shows that for a range of processed foods, European consumption correlates strongly with American consumption 5 and 10 years earlier, but not with consumption in the same time-period. Traill (1996) shows that the European nations' food consumption patterns are also converging. Hermann and Röder (1995) and Gil et al (1995) apply different statistical methodologies to, respectively, OECD and EU food consumption data, in both cases concluding that convergence is occurring, though in the latter case concluding that the speed of convergence is diminishing.

38. However, Gil et. al. (1995) also demonstrate that significant dietary differences remain: using cluster analysis, they identify 7 Western European country groupings: Portugal and Spain; Greece and Italy; Benelux, France, Ireland and the UK; Austria, Germany and the Netherlands; Finland; Denmark; and Norway and Sweden. These groupings are barely related to income. One reason that food consumption patterns should not be expected completely to converge among countries even if socio-economic and demographic factors do, is that culture is an important influence on behaviour and cultural diversity has proved resistant to the pressures from foreign travel, global media and telecommunications. `We have a biological need to consume a certain amount of calories and quantity of liquid; this need, however, does not define what to eat and with what, how to cook it, when to eat it and in what social circumstances' (Askegaard, 1995). Sociologists have studied relationships between, for example, social class and food culture (Mennell, 1992) . Anthropologists (e.g. Douglas 1982) emphasise the role of food as a central part of social rituals. They define food culture as a culinary order whose traits are prevalent among a certain group of people defined from the micro-level (family) to the macro level (countries, regions, social classes: Askegaard, 1995). Indeed, religion is an important determinant of meat consumption, Jews and Muslims abstaining from pork, Hindus from beef. Such countries have lower meat consumption than would be predicted by GNP alone (which, however, explains 66% of cross country variation in meat consumption: Grigg, 1995). Food products are important cultural modes of expression used for communication purposes (Douglas, 1982).

39. Indeed, there are close relationships between personality and nationality that are presumably explained by culture. For example, Hofstede (1984), uses four dimensions to categorise personality, (i) individualism vs collectivism (importance attached to individual freedoms vs society) (ii) power distance (tolerance of inequality in wealth and power/centralisation of authority) (iii) risk (the extent to which risk is avoided through laws, rules, religion) (iv) masculinity vs femininity (emphasis placed on masculine values of performance, aggression, visible achievement). In a study based on questionnaires with 116,000 IBM employees world-wide, the responses were used to group similar countries. Some of the groups are shown in Table 4: The Anglo-American, group consists of the U.S., UK Australia, New Zealand, Ireland and South Africa; the Nordic group has all the Scandinavian countries; the Germanic group includes Germany, Austria and Switzerland; the Latin-European contains France, Spain, Italy, Portugal, Belgium and Latin America. These country groupings are intuitively more logical than those obtained by grouping countries according to economic and demographic variables (e.g. Krause, 1995), which puts the U.S., Japan and Sweden in the same cluster. Roth (1995) shows that cultural power distance, cultural individualism as well as regional socio-economics affect the performance of companies' brand image strategies.

Table 4: Characteristics of Country Groupings


Power Distance

Uncertainty Avoidance


















Latin European










Far Eastern





LL=Very Low, L=Low, M=Medium, H=High, HH=Very high.

Source: Adapted from Hofstede (1985)

40 None of this is to say that cultural behaviour is forever fixed: for example, the Mediterranean diet is being eroded in the Mediterranean countries as fat intake levels move towards those of other European nations (Serra-Majem and Helsing, 1993); and immigrants have been shown to adopt, albeit slowly, the food consumption patterns of their hosts (Choe et al, 1993).

41. Given that consumers within countries are not all the same, and that differences among countries are also important, how should one view the convergence issue? The most sensible approach is to understand emerging global consumer markets as `groups of buyers that share the need and desire for a product and the ability to pay for it rather than those who share a common border. Buyers in a segment seek similar benefits from and exhibit similar behaviour in buying a product' (Blackwell et al., 1994, p221). According to this approach, demographic and economic considerations remain important, but so do psychometric, attitudinal, cultural and lifestyle characteristics

III. Conclusions

42. Total apparent food consumption has increased everywhere in the last 30 years (Hallam, 1996), but especially in Asia, and continuing growth in food consumption can be expected in the developing world. More importantly, continuing increase in animal consumption can be expected, particularly in Asia, linked to continuing income growth. However, income elasticities in OECD countries are now so low that little further income-related change can be expected in the composition of those countries' diets. Health-related factors may increase fruit and vegetable consumption in these countries, but so far the only evidence for this is from the United States (Hallam, 1996).

43. It is apparent that as incomes and economic systems world-wide converge, particularly as the newly industrialised countries become wealthy enough for food quality to become the key issue in those countries, world-wide food consumption patterns will also converge. In this they will be helped by a tendency for socio-demographics (numbers of children, ageing population, women working, rural vs urban living) to become more alike as the stage of economic development progresses. It should however be clear from what has been said that although consumption behaviour will become more similar, convergence will be far from complete-quite significant differences will remain, as indeed they do between say Japan and the USA now. Thus although the modern credence characteristics of food quality will become more important everywhere as incomes rise, they will manifest themselves in different ways in different cultures. The very fact that European food cultures and habits remain so different despite the countries being so close, geographically, but also in terms of religion, income and socio-demographics bears witness to the fact that global differences will be very resistant to economic development.

44. It is not the purpose of this paper to address directly the trade implications of these changes, but perhaps a final word is in order. In the past, trade liberalisation in agriculture has been difficult because of countries' desires to protect their farm populations (and their election votes). In future, as quality attributes of food become ever more important to consumers (relative to price), particularly as the attributes of quality are themselves culturally determined, the barriers to trade liberalisation will be countries' desires to protect their consumers (and their election votes). Science-based decision making will be hard to `sell' at a time when science is one of the attributes consumers would most prefer to be absent from their food.

IV. Recommendations

45. The consumer food choice process is complicated and often seems irrational, yet consumer perceptions, rational or not, have important implications for agricultural and food markets and for trade in those products. Consumer education and information (eg labelling) are frequent responses of policy makers, who assume that educated and informed consumers make more sensible choices. This may be true, however, it is also recommended that producers and regulators should be better educated and informed about the nature of consumers' food choice processes. In that way, they may make less costly mistakes and better understand the types of education and information to supply.

46. Furthermore, there is a lot still unknown about consumers' use and reaction to different types of labelling and more research is needed in this area, particularly with respect to how culture influences the demand for, use of and response to information. It seems inevitable that different cultures would prefer different information provision regulations. A major challenge is therefore to find formats that can be internationally accepted. It is suggested (recommended) that a spirit of compromise and sound research are vital ingredients.

47. Producers and regulators must also appreciate that most consumers don't want to spend the time to inform themselves fully about the intricacies of diet/health/environment interactions, but would prefer to entrust producers, regulators and scientists to do the job for them. Unfortunately that trust has been lost to a great extent and consumers are turning increasingly to organic food and local production as a safe haven. This is a serious (but completely legal) barrier to international trade. Rebuilding trust is essential to the development of a vibrant international trading system in agricultural and food products and to the acceptance of continuing technological innovation (particularly, but not exclusively biotechnology related). It is recommended that top priority is accorded to `trust-building' in the international agro-food system


AGB/Europanel (1992). Prospects for after 1993.

Askegaard, S (1995). European food cultures: an exploratory analysis of food related preferences and behaviour in European regions, MAPP Working paper no 26, Aarhus School of Business, Aarhus.

Blackwell, R.D., Ajami, R. and Stephan, K. (1994). Winning the Global Advertising Race: planning Globally and Acting Locally, pp209-232, in Hassan, S.S. and Kaynak, E. (eds). Globalisation of Consumer Markets: Structures and Strategies, International Business Press, New York, 333pp.

Choe, S T, Capella, L M and Arnold, D R (1993). Acculturisation, ethnic consumers and food consumption patterns, Journal of food products marketing, 1(4), 61-79.

Connor, J. (1991). North America as a Precursor of Changes in Western European Food Purchasing Patterns, NC-194 Occasional Paper OP-22, 56pp.

Connor, J (1994). North America as a Precursor of Changes in Western European Food Purchasing Patterns, European Review of Agricultural Economics, pp 155-173.

Douglas, M (1982). Food as a system of communication. London, Routledge and Kegan Paul.

Engel, J F, Blackwell, R D and Miniard, P W (1995). Consumer Behaviour, New York: The Dryden Press, 8th edition.

Gil, J M, Gracia, A and Pérez y Pérez, L (1995). Food Consumption and Economic Development in the European Union, European Review of Agricultural Economics, 22, 385-399.

Grigg, D (1993). International variations in food consumption in the 1980s, Geography, 78, 251-266.

Grigg, D (1994). Income, industrialisation and food consumption, Tijdschrift voor Economische en Sociale Geografie, 85(1), 3-14.

Grigg, D (1995). The pattern of world protein consumption, Geoforum, 26(1), 1-17.

Grigg, D (1996). The starchy staples in world food consumption, Annals of the Association of American Geographers, 86(3), 412-431.

Grunert, K G, Brunso, K and Bisp, S (1993). Food related life-style: development of a cross-culturally valid instrument for market surveillance, MAPP working paper no 12, Aarhus School of Business, Aarhus, 38pp.

Hallam, D (1996). Implications for demand of the main agricultural products of changes in food consumption patterns, OECD AGR/CA/APM/MD/CFS(96)11, 37pp.

Hermann, R and Röder, C (1995). Does Food Consumption Converge Internationally? Measurement, Empirical Tests and Determinants, European Review of Agricultural Economics, 22, 400-414.

Hofstede, G. (1984) Culture's Consequences, Sage Press, London.

Krause, J.H., Wilson, W.W. and Dooley, F.J. (1995). Global Market Segmentation for Value-Added Agricultural Products, Agribusiness, 11(3), 195-206.

Mennell, S, Murcott, A and van Otterloo, A (1992). The sociology of food. Eating, diet and culture. London: Sage.

National Consumer Council (NCC) (1997). Messages on food: Consumers' use and understanding of health claims on food packs. London: Department of Trade and Industry.

Rogers, P J and Blundell, J E (1990). Psychobiological bases of food choice, in Ashwell, M (ed), Proceedings of the 12th British Nutrition Foundation Annual Conference, BNF, London, 31-40.

Roth, (1995). The effects of culture and socio-economics on the performance of global brand image strategies, Journal of Marketing Research, XXXII, 163-175.

Serra-Majem, L and Helsing, E (eds) (1993). Changing patterns of fat intake in Mediterranean countries, European Journal of Clinical Nutrition, 47(1) Supplement 1.

Steenkamp, J-B E M (1996). Dynamics in consumer behaviour with respect to agricultural and food products, in Wierenga, B, Grunert, K G, Steenkamp, J-B E M, Wedel, M and van Tilburg, A (eds), Agricultural marketing and consumer behaviour in a changing world, Proceedings of the 47th Seminar of the European Association of Agricultural Economists, Wageningen, 13-15 March, 15-38.

Traill, W B (1996). Globalisation in the European food industries, European Review of Agricultural Economics, XXXX.

Traill, W B and Righelato, R (1996). Technology Foresight for horticultural products: strawberries, MAFF.

Van Trijp, H, Hoyer, W D and Inman J J (1996). Why switch? Product category explanations of true variety seeking behaviour, Journal of Marketing Research.