A European Regional Analysis of Selected Food Consumption Statements

ABSTRACT - This paper examines different European regions of food culture. It does so by analyzing statements about food attitudes and food consumption from a general, commercial life style survey. The data were collected in 1989 through 20.000 interviews in 15 European countries broken down into 80 regions. The purpose of the paper is to analyze what the variables included in the life style survey can reveal with regards to differences among Europeans' food cultures. This is done through a series of univariate analyses as well as through factor and cluster analyses. However, since the data are not yet ready for more advanced statistical analysis, this paper should definitely be regarded as a work in progress paper, incomplete as it is.

Citation:

Soren Askegaard (1993) ,"A European Regional Analysis of Selected Food Consumption Statements", 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: 410-415.

European Advances in Consumer Research Volume 1, 1993      Pages 410-415

A EUROPEAN REGIONAL ANALYSIS OF SELECTED FOOD CONSUMPTION STATEMENTS

Soren Askegaard, Odense University, Denmark

ABSTRACT -

This paper examines different European regions of food culture. It does so by analyzing statements about food attitudes and food consumption from a general, commercial life style survey. The data were collected in 1989 through 20.000 interviews in 15 European countries broken down into 80 regions. The purpose of the paper is to analyze what the variables included in the life style survey can reveal with regards to differences among Europeans' food cultures. This is done through a series of univariate analyses as well as through factor and cluster analyses. However, since the data are not yet ready for more advanced statistical analysis, this paper should definitely be regarded as a work in progress paper, incomplete as it is.

INTRODUCTION

Nobody with the slighest idea of European cultures would fall into the trap of expecting an overnight European integration to take place following the creation of the European Single Market. There is, however, no doubt that a new kind of borders will appear more clearly in addition to the established national borders. These borders will be of crucial importance to international marketing within Europe. The new borders in question here are the cultural ones, not necessarily following the traditional borders of European nations.

If the growing integration of the world markets into free trade zones with mutual agreements among them is showing the will to eliminate the "physical" barriers to the international exchange of goods, there is no sign that mental barriers to the exchange of goods will not persist in the future. These barriers, expressing the lack of a specific position of certain consumer goods from other cultures within a given cultural universe, have their roots in the cultural differences among nations and regions. In spite of some signs of food products and services with global success such as Coca-Cola or McDonald, this globalization is a rather superficial one. The fact that one finds hamburger restaurants everywhere in the world does not mean, that this kind of consumption is the same everywhere. On the contrary there are large differences in the ways the hamburger concept in inscribed in different (food) cultural universes, how important its status is, if businesspeople can eat there, etc..

The analysis in this paper draws on a commercial database established through a 1989 European life style survey based on 20.000 interviews in a total of 15 European countries. Thus, the paper should be seen in the light of the possibilities created by cooperations of various kinds between science and commercial marketing and management (Zaltman & Moorman 1985).

One chapter in the life style questionnaire was dealing with food consumption, its motivations and actual consumer behavior. The general idea of the following analysis is to detect patterns of similarity and difference between the included European countries and regions in the responses to these variables. Thus, it is by no means a complete analysis but just an attempt to find out what could be learned from a specific source of secondary information.

THE CULTURAL DIMENSIONS OF FOOD CONSUMPTION

That eating behavior is deeply embedded in culture both physically (what we eat, how we eat, ...) and mentally (our cognitive representations and emotions about food and eating) is not new. This has been analyzed several times and from several angles. A variety of social scientific approaches has been used, e.g. a semiotic point of view (Barthes 1961), a sociological point of view (Fishler 1990), a psychological point of view (Lyman 1989), and, of course an anthropological point of view (e.g. Levi-Strauss 1965, Douglas 1982). Needless to say, this list is by no means exhaustive but serves only to illustrate examples of the different aspects studied.

The recognition of the cultural dimensions of food consumption has, of course also left its traces within the consumer behavior literature, from which field one of the most splendid examples is provided by Levy (1981). More recently, the work of Suzanne Gruner has witnessed the interest of the social psychological approach to the study of food consumption (e.g. Grunert 1990).

The present analysis represents a first attempt to outline regional cultural borders within Europe concerning certain specific food consumption patterns. It is part of a larger research project concerning the cultural dimensions of food consumption launched by the end of 1991. That project will seek to establish theoretical models of the cultural aspects that are the most important in the formation of specific eating and drinking patterns in different societies. This way, we hope to be able to develop a scheme for the systematization and analysis of information about food consumption and food attitudes. For this purpose, a larger number of interviews are scheduled for the end of 1992.

However, it proved to be possible to scan an existing database concerning food consumption for interesting patterns that would serve as a starting point for the further analysis. Thus, the present article is merely a sort of pilot work - an exploration of a set of given data to give hints about differences and similarities in the food cultures of Europe.

THE LIFE STYLE SURVEY

The data base originates from a life style study carried out in 1989 in 15 European countries, namely the twelwe EC countries plus the four EFTA countries of Norway, Sweden, Austria, and Switzerland (Luxemburg counting as part of a region in Belgium). The survey was organized by the CCA (Centre de Communication AvancT) life style research agency in Paris in collaboration with the Europanel network of market research agencies. The questionnaire and the analysis were made by the CCA while the field work. sampling, interviewing, etc. were done by the different Europanel agencies in each country.

The explicit purpose of the survey was to generate a pan-Europran life style typology according to the same principles as the ones used previously by the CCA for life style studies in France, the West Indies and United States since the 1970'es. This rather specific approach to life style analysis is described in depth in Cathelat (1990).

These life style studies, together with all other commercial life style analyses, have been very much criticized for several reasons, most notably for the typologies' minimal ability to predict consumer behavior (Kapferer & Laurent 1981, Valette-Florence & Jolibert 1985), for the reliability and validity of the results obtained and for the reluctance of exposing the methods to scientific tests. (Valette-Florence 1989).

However, these problems of the validity of the life style typologies need not concern us here, as the data used in this connection are not broken down into a life style typology but analyzed by countries and regions.

TABLE 1

NUMBER OF REGIONS AND SAMPLES IN EACH COUNTRY

Since the aim of the survey was to generate a common European life style typology, the geographic and national cultural differences were deliberately of less interest to the CCA. However, the goal was also to make the survey representative down to a regional level in order to detect not only national differences but also regional differences in the weight and profiles of the life style segments as well as in the response patterns.

Therefore, the individual countries were divided into several regions according to the standard practices of each national research agency. The number of regions in the survey reached a total of 80, but with some variation between the countries.

SAMPLING

The sampling was as already mentioned carried out by the Europanel market research agencies in the 15 countries. Representative samples were drawn from existing panels of individuals. Thus, one should consider the risk of some degree of "panel effect", even if it's exact importance is hard to examine. The samples in the 15 countries were constructed as shown in Table 1.

Of course, the disproportionate representation of the different countries makes it difficult to create reliable average scores for Europe. The more interesting point is therefore, and in accordance with the general theme of the research project, to look at the differences in the scores of each of the variables in each of the different countries and regions.

The regional divisions used in the survey were the regions normally used by the Europanel agencies for subdivision of the respective countries (a list is available upon request). Herein lies another validity problem, namely that of regional subdivisions less suited to the examination of food cultures in that important borders between food cultures may exist within the region. The possibility of regional analysis is, however, already a progress compared to most international analyses, which operate on a national level only.

A very satisfying general response rate of 75 % was obtained with no significant differences among the countries or regions.

THE VARIABLES

The origin of the data of course means that there has been no scientific control over the variables included in the survey. The variables were selected to satisfy the subscribers to the results of the study, and thus have a commercial rather than sicentific character. The data accesible for this research project were, with some exceptions, the questions in the "food chapter" of the questionnaire. The exceptiones concerned a total of three questions that could not be included due to a subscriber's private ownership of the information.

Thus, the number of questions or groups of questions left for inclusion in our study was 11, comprising a total of almost 400 items of information or variables. The different questions dealt with a variety of aspects of food consumption attitudes and behavior and was easy to understand for the respondents. The percentage of "no answer" was generally about 1 or 2 % and in no cases higher than 4 %.

The eleven question groups included in the data base can be described as follows:

1. A question comprising nine binary choices of "food temper" (I eat quickly vs. I eat slowly, etc.)

2. A question comprising nine statements about changes in behavior patterns (I do X "more than before", "as much as before", "less than before", or "I never did X")

3. A question comprising 33 variables concerning liking or disliking different kinds of tastes and consistencies in food.

4. A question comprising 16 statements about the willingness to change food consumption habits in the future (of the kind "In the future, I am willing to eat...).

5. A question about eating habits at different times of the day on weekdays and on sundays. The question comprised 12 general food items and 6 points in time during each of the two kinds of days (a total of 144 informations).

6. A question about nibbling habits with a total of 33 variables.

7. A question with 11 statements about the quality of food products from 9 European countries (a total of 99 informations)

8. A question about preferences for French wines comprising 5 variables.

9. A question about wine consumption habits (8 variables, 4 four for red and four for white wine).

10. A question concerning diet behavior (a scale with four diet involvement levels).

11. A question concerning drinking habits (frequency of drinking different products) with a total of 31 variables

FIGURE 1

PREFERENCES FOR EATING ONE SOLID DISH VS. A COMPOSIT MEAL, 61 REGIONS OF THE EC

Needless to say, one cannot avoid validity problems concerning the semantic differences in the interpretation of a questionnaire translated into 11 different languages. A typical example is provided by some of the "taste items", where a profound semantic difference should be expected between e.g. the Irish' and the Spanish' conception of what the item "very spicy" means. Such differences should of course be taken into account during the interpretation of the results. However, we have not as yet been able to detect outright translation mistakes in the questionnaire.

Likewise, there will be vast differences in the socio-economic backgrounds of each country and region that cannot be detected directly through this data material. These aspects must instead serve as means of interpretation of the results. It should be remembered that this analysis by no means can be regarded as a complete investigation of food cultures in Europe.

THE DATA

It was not possible to obtain the raw data of individual answers, neither was it possible to get the data in a computer format compatible with the ones used at Odense University. So the data was from the beginning in the format of gross percentages of each of the variables per region and per country obtainable on print only.

With this data format, we have had to treat the variables as if one (hopefully informed) person from each country and each region has answered the question: "How many percent of the population in your region / country ..." This is of course not the optimal kind of data to work with, but it gives the advantage of having all the data in an interval format. And for this exploratory study that is only supposed to give some hints to possible dimensions for further research in the actual research project, we have assumed that these data would be better than no data.

THE ANALYSIS OF SELECTED VARIABLES

Up till now, the only type of analysis we have been able to do is analysis of one variable at a time. This has been done in order to get a broad overview of the differences and similarities in the different parts of Europe. This can be done for all the European countries included in the survey or for some of them, according to the purpose of the specific analysis. The following example is analyzed solely on the basis of data from the EC countries.

Here, one of the more clear-cut cases of such an individual variable is shown in figure 1. This is a mapping of one of the variables from question 1, in this case a binary choice of a preference for either a meal consisting of one proper dish with meat and accompanying vegetables as opposed to a composite meal consisting of many small dishes.

By analyzing the variations in terms of a minimum of 10 % divergence from European average, the border between Latin European and Anglo-Saxon and Germanic food cultures becomes clearly located between Germany and France and running through the southern part of Belgium and northern France. North of this dividing line, there is an overwhelming preference for the solid meal (the white parts of the map), whereas south of the line there is in general a majority of preferences for the composite meal (the black parts of the map). One can also note the possible existence of a slightly more solid "Iberian-Atlantic" food culture but this remains to be confirmed. Finally, a more Latin European preference pattern can be found in certain capitals of northern Europe, Brussels and Copenhagen, possibly a sign of a more "cosmopolitan" food culture in these regions of their respective countries.

Concerning convenience foods, some further specific patterns can be detected. On the question of whether the respondent buys instantly preparable food products more than before (figure 2, from question group 3), this trend seem to be more manifest in "Atlantic Europe: Great Britain and Ireland, France, Spain and Portugal whereas Germany seem to be rather off trend, either because the use is already quite frequent or because of more traditional norms concerning the preparation of meals.

In figure 3, the frequency of a similar variable, namely the trend towards buying take-away food, is illustrated. It is clearly seen that this trend is most visible in France and French-speaking Belgium as well as in Catalonia and the rest of eastern Spain. The Dutch speaking regions, Ireland, Scotland and Denmark form a "Northern group" of relatively off trend regions.

These are but a few examples of the types of variables and results in the univariate analyses. It should be added, that a breakdown of the respondents into housekeepers and non-housekeepers does not seem to affect the general picture of distribution of frequencies. Thus, to obtain maximum representativeness, the examples here have all been drawn from the gross percentages of all respondents.

This part of the analysis is still going on, and no general conclusions have been reached yet. One of the few conclusions that has been made till this point, is that, in terms of food culture, Belgium is certainly as divided a country as the political cleavages of it indicates. Again and again, the Flemish and French speaking parts of Belgium come out in "black and white", indicating fundamentally different response patterns in the two populations. Centrally located Brussels most often joins the french speaking part of Belgium what its food culture is concerned.

But such partial conclusions of course will need the support of more substantial evidence from a more profound analysis of the data. Thus, more general conclusions will only be reached after the data material has been prepared for and subject to factor and cluster analyses. These analyses will serve both as validation and exploration of the data base.

If none of the 'common sense' differences between European food cultures come out, it is a serious blow to the validity of the data base as a whole. On the other hand, if the data base shows some indications of regional food cultures, these results are interestiong both in themselves and as possible hints about which dimensions are the most discriminating among European food cultures and which regions cluster with which as to form generally different European food culture areas.

However, since the data on June 10 1992 were not yet ready for computerized statistical analysis, these results will only be available in a not too distant future. Thus, as an indication of a truly "work in progress paper", the rest of it will left open.

THE FACTOR ANALYSES

THE CLUSTER ANALYSIS

CONCLUSION

FIGURE 2

PERCENTAGE OF RESPONDENTS STATING THAT THEY BUY INSTANTLY PREPARABLE FOOD MORE THAN BEFORE, 61 REGIONS OF THE EC

FIGURE 3

PERCENTAGE OF RESPONDENTS STATING THAT THEY BUY TAKE-AWAY FOOD MORE THAN BEFORE, 61 REGIONS OF THE EC

REFERENCES

Barthes, R. (1961): "Som man spiser. Udkast til en psyko-sociologisk analyse af nutidens mad", Hug, vol.10, nr.48, 28-36 (orig. Annales, vol. 16, No. 5, Paris 1961).

Cathelat, B. (1990): Socio-Styles-SystFme, Editions d'Organisation, Paris.

Douglas, M. (1982): "Food as a Communication System", in M. Douglas (ed.) In the Active Voice, London.

Fischler, C. (1990): L'homnivore, Editions Odile Jacob, Paris.

Grunert, S. (1990): "Personality Traits as Elements in a Model of Consumer Behavior", in K. Grunert & F. +lander (eds.) Understanding Economic Behavior, Kluwer, Dordrecht, 309-332.

Kapferer, J.-N. & G. Laurent (1981): "Une analyse des relations entre les classifications socio-culturelles et et de styles de vie et l'achat de produits courants", IREP, 205-223.

LTvi-Strauss, C. (1965): Le cru et le cuit. Mythologiques 1, Plon, Paris.

Levy, S. J. (1981): "Interpreting Consumer Mythology: A Structural Approach to Consumer Behavior", in Journal of Marketing, vol. 45, Summer, 49-61.

Lyman, B. (1989): A Psychology of Food, Von Nostrand Reinhold, New York 1989.

Valette-Florence, P. (1989): Les Styles de Vie, Economica, Paris.

Valette-Florence, P. & A. Jolibert (1985): "Un essai empirique de clarification des approches des styles de vie", Actes de l'Association Francaise de Marketing, 1, 133-157.

Zaltman, Z. & C. Moorman (1985) "Sharing Models of Inquiry", in Advances in Consumer Research, vol. 12, 312-14.

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