An Empirical Study of the Effects of Ethnicity on Consumption Patterns in a Bi-Cultural Environment

ABSTRACT - This investigation focused on the relationship between English-French Canadian ethnicity and consumption patterns. The study employed a communication pattern based ethnic classification scheme which accounts for the varying degrees of acculturation toward either end of the English-French ethnicity continuum. Four groups of varying degrees of English-French ethnicity were identified and then examined in terms of consumption differences for food and personal product items. Results indicated between group consumption differences for several product classes. However, the assumption of monotonicity between acculturation and consumption patterns was largely unsupported.

Citation:

Chankon Kim, Michel Laroche, and Annamma Joy (1990) ,"An Empirical Study of the Effects of Ethnicity on Consumption Patterns in a Bi-Cultural Environment", in NA - Advances in Consumer Research Volume 17, eds. Marvin E. Goldberg, Gerald Gorn, and Richard W. Pollay, Provo, UT : Association for Consumer Research, Pages: 839-846.

Advances in Consumer Research Volume 17, 1990      Pages 839-846

AN EMPIRICAL STUDY OF THE EFFECTS OF ETHNICITY ON CONSUMPTION PATTERNS IN A BI-CULTURAL ENVIRONMENT

Chankon Kim, Concordia University

Michel Laroche, Concordia University

Annamma Joy, Concordia University

[The authors would like to thank the Committee of Aid to Scholarly Activities, Concordia University for providing the research funds.]

ABSTRACT -

This investigation focused on the relationship between English-French Canadian ethnicity and consumption patterns. The study employed a communication pattern based ethnic classification scheme which accounts for the varying degrees of acculturation toward either end of the English-French ethnicity continuum. Four groups of varying degrees of English-French ethnicity were identified and then examined in terms of consumption differences for food and personal product items. Results indicated between group consumption differences for several product classes. However, the assumption of monotonicity between acculturation and consumption patterns was largely unsupported.

That consumption is a thoroughly cultural phenomenon is well established in the literature on consumer behavior (Belk, Sherry & Wallendorf 1988; Hirschman 1981; Mick 1986; McCracken 1986.) Yet the research on ethnicity and subcultural differences in consumption has not really explored the meaning of consumption for members of these groups. The underlying premise of the research on ethnicity has been that subcultures, while sharing the values and norms of the dominant culture, express certain significant differences of their own which may warrant differential marketing efforts (Nicosia & Mayer 1976; Zaltman & Bagozzi 1976). In these earlier studies, ethnic groups were identified using objective measures such as place of residence, last name, or language spoken at home and then compared with those of the dominant culture with respect to significant marketing dimensions. Clearly the validity of these studies depends on the soundness of the ethnic classification scheme used (Bergier 1986).

More recent measures use emic or subjective measures of ethnicity (Hirschman 1981). Although the term "degree of ethnic identification" is not equivalent to acculturation, its measurement enables researchers to indirectly assess the degree of acculturation. The term "acculturation" refers to the degree that values and norms of an individual or group correspond to those of the dominant group (Gordon 1964; Yinger 1985). Most studies of ethnicity look at immigrant adaptations to the host society (otherwise known as the assimilation hypothesis) and not vice versa (Glazer & Moynihan 1965). However. as Padilla (1980) notes, this is dependent on the relative bargaining power of the groups in question and is determined within the context of relevant historical and political forces. More recent research in consumer behavior reflects this complexity in conceptualizing the myriad faces of culture (Deshpande, Hoyer & Donthu 1986; Guinn & Faber 1985; Valencia 1985; Wallendorf & Reilly 1983).

ACCULTURATION

Some support for the assimilationist hypothesis has been provided by Valencia (1985) and Schaninger, Bourgeois and Buss (1985), who found a monotonic relationship between acculturation and several consumption dimensions. On the other hand, the pattern of findings by Wallendorf and Reilly (1983) indicate that immigrant adaptation is neither like the culture of origin nor of the culture of residence but lies somewhere between the two.

One approach that has recently gained renewed attention is the importance of communication processes in general and the use of language in particular (Shibutani & Kwan 1965). In the study of consumer behavior, O'Guinn & Faber (1985) develop a measure based on communication patterns. The underlying rationale for this measurement approach is that communication is the most fundamental means by which individuals develop their understanding of a new culture (Kim 1985). Communication involves interaction with the environment. As strangers interact with people in the host environment, they learn and acquire the acculturative capacities in their cognitive, affective, and behavioral processes (Kapoor & Williams 1979; Pedone 1980; Shibutani & Kwan 1965; Tzu 1984; Wen 1976).

Of particular relevance to this study are two forms of communication- interpersonal and mass communication. A person's interpersonal communication can be observed through the degree of his/her participation in interpersonal relationships with members of the host society (Kim 1985), whereas individuals participate in mass communication processes through such media as radio, television, newspapers, magazines, movies, theatre, museums, lectures and posters (Tzu 1984).

RESEARCH OBJECTIVES

The focus of this research is on the relationship between English-French Canadian ethnicity and consumption patterns. While much of Canada is rooted in the English culture, French and English cultures co-exist as the two dominant forces in Quebec, exerting influences on each other. Conscious political efforts by the Government of Quebec to preserve and expand the influence of French culture imposes an assimilation force on the English perhaps to a similar extent to which the English impose on the French. In recognition of this two-way process, English-French Canadian ethnicity in this study is viewed as a continuous concept with the two polar extremes corresponding to "strong English Canadian identity" and "strong French Canadian identity." The middle portion of this continuum would then represent various degrees of acculturation either French Canadians or English Canadians have experienced toward the other side.

French-English Canadian Consumption Differences

While there is a large volume of research investigating English-French Canadian consumption and lifestyle differences (Schaninger, Bourgeois & Buss 1985; Saint-Jacques and Mallen 1981; Tigert 1973; Vickers & Benson 1972), nearly all these studies use a dichotomous classification scheme based on a single indicator and, furthermore, many of them are post hoc in design.

Most of these studies indicate that the French are more oriented toward cooking and baking, less reliant on instant or convenience food, prefer soft drinks, sweet beverages, and consume larger quantities of alcoholic beverages (Barnes & Bourgeois 1977; Clifford 1979; Mallen 1973; Saint-Jacques 1981) than English Canadians do. In terms of life-style profiles, French Canadians seem more concerned with personal appearance and fashion, and are more oriented towards their home and cleanliness and are more likely to own washers, dryers, and other home appliances than their English counterparts are.

METHODOLOGY

Data

Data used in this study come from a survey of residents in various districts of the Greater Montreal Area performed in 1986. A quasi-probability sampling procedure was chosen for the reasons of cost and speed. Fourteen census tracts with a high concentration of population of English and French ethnic origin were included in the first sampling stage. Within each census tract, streets were chosen randomly. Interviewers knocked at the door and asked for the lady or the man of the house. Then, after the introduction, a filter question screened those respondents who did not overtly identify themselves as either French Canadians or English Canadians.

Qualified respondents were asked their preference of a French or English questionnaire. The self-administered questionnaire was left with the consenting respondents to be picked up at a later time. In total, 600 completed responses were obtained, of which 200 were English and 400 were French. The non-respondents did not differ significantly from the respondents in terms of demographic characteristics.

Measurement

In the beginning of the questionnaire, respondents were asked to indicate the degree of preference for the language of the questionnaire on a scale ranging from 1="I always prefer French questionnaires" to 9="I always prefer English questionnaires". This particular question is also considered as a communication variable, and will be used in conjunction with others in the ethnic classification scheme. Measurement of communication patterns in interpersonal and mass media use situations required the subjects to estimate the percentages of the times they use French, English and other languages (adding up to 100) in the following contexts: 1. with spouse; 2. with children; 3. with relatives; 4. at work; 5. when watching television; 6. when listening to radio; 7. when reading newspapers; 8. when reading magazines or books; 9. when shopping; 10. with close friends; 11. when in school.

The next section of the questionnaire contained questions for consumption frequencies of 35 personal and food items. Also included in the questionnaire were various life style questions using 10-point likert scales as well as those requesting the respondent's demographic information.

Analysis

Methodologically, this study employs a cluster analysis based taxonomic procedure which allows for the detection of groupings with varying degrees of acculturation toward one end or the other on the English-French ethnicity continuum. In contrast to the a priori classification scheme used in the past studies (Deshpande, Hoyer, and Donthu 1986; Valencia 1985; Wallendorf and Reilly 1983; Schaninger, Bourgeois, and Buss 1985), the cluster based approach does not assume in advance any particular number or descriptions of groups.

Since the focus of this study is on English-French Canadian ethnicity, the analysis included only those respondents who reported the use of only French and/or English for all of the eleven communication contexts mentioned in the prior section.

The first step in the analysis plan involved cluster analyzing the subjects with the twelve communication variables (including the 9-point scale question on the language preference for the questionnaire). For each of the eleven communication contexts, the percentage of times English used was subtracted from the percentage of times French used, and inputed into the analysis. Responses on the 9-point scale question of questionnaire language preference was transformed to match the ranges of the eleven communication variables (i.e., 1="I always prefer French questionnaires" was recoded to +100, 9="I always prefer English questionnaires" was recoded to -100, and middle scale points accordingly).

Clustering of the respondents was performed with the BMDPKM program (Dixon et al. 1983). Cluster solutions of two to six group partitioning were examined in terms of the average F-ratio (average between group variance divided by average within group variance for the twelve variables). This average F-ratio is an indication of the degree of homogeneity among the subjects within each group in a cluster solution. In all of the five cluster solutions, the average F-ratios were very high 12776, 932, 701, 585, 492 for two to six cluster solutions respectively. An examination of the successive drops in F-ratio (an elbow test) apparently indicated that higher number cluster solutions may still contain quite heterogeneous groups with respect to the twelve input variables. At this point, it was decided that, the four group solution containing two rather unacculturated groups, one French Canadian and one English Canadian, and two groups in-between on the acculturation continuum will be examined for reasons of parsimony in terms of the consumption patterns for the product items mentioned in the previous section.

The next phase of the analysis dealt with the validation of this cluster based ethnic identification with discriminant analysis. With the membership information of the four cluster solution, discriminant functions using the twelve communication variables as independent variables are estimated from a randomly selected half of the total sample (i.e., analysis sample) and are subsequently tested for their accuracy in predicting membership in the remaining half (i.e., validation sample). This process is followed by testing the predictive power of each of the twelve communication variables. This was also accomplished with split sample discriminant analyses incorporating one communication variable at a time as the independent variable and the classification membership as the dependent variable.

The final stage of the analysis involved comparisons among the four clusters with respect to the consumption frequencies for the selected items. The comparisons will utilize analysis of variance and multivariate analysis of variance incorporating several demographic and life style variables as covariates.

RESULTS

The Cluster Based Four Group Taxonomy

Profiles of the four clusters based on their mean values of the twelve communication variables are presented in Figure 1. First, the profile shown on the far left of the figure corresponds to the group which can be labeled as "strong English". They have been educated in English, and English is the predominant language for communications in family, social, and for media use. However, not all of them seem unilingual. On average, members in this group reported a fair amount of use of French at work and when shopping. It appears that some members in this group have acquired some level of competency in French language to facilitate their communication at work as well as in shopping situations.

The next profile to the right pertains to the "moderate English" group. While English appears to be the primary language for these people, the use of French at work, while speaking to relatives and while shopping is considerably more frequent than the strong English group. It is likely that the moderate English group uses French with relatives primarily because of higher incidence of intermarriage. Insofar as the communication patterns reflect the degree of acculturation, these respondents appear to have undergone a substantial extent of acculturation toward the French Canadian culture.

To the right of the moderate English group is the "moderate French" cluster. These respondents communicate mostly in French in the family, work, and social settings. On the other hand, they tend to watch English TV programs, listen to English radio channels, read English magazines more often than those of the French language, and read French newspapers slightly more frequently than English newspapers. Their profile is apparently quite distinct from that of the moderate English group in many aspects. This casts some doubt in the homogeneity of the "bilinguals" category employed by Schaninger, Bourgeois, and Buss (1985). Finally, the profile of the "strong French" group is presented at the far right of Figure 1. Similarly as the strong-English group, French is the dominant language of communication at home, with friends, and for their formal schooling. However, they reported a sizeable extent of English media usage. In particular, nearly one- fourth of their TV viewing is for English programs. Also noticeable is their fair amount of use of English in work places.

In the subsequent step, using the discriminant analysis, a split-half validation of the four group taxonomy was performed. Results showed that the classification functions developed from the analysis sample predicted the membership of the cases in the validation sample with a 93.2 percent accuracy. With respect to the predictive power of each of the communication variables, results of the twelve split-half discriminant analyses using one variable at a time as the independent variable are summarized in Table 1. Discriminating power of the individual variables, measured in the percentage of the correct assignments made-in the validation sample, ranges from 62.5 percent for the language used at work to 80.4 percent for the language of the newspaper. Apparently, no single item matches closely the cluster procedure in its classification results. These findings indicate that the often used single language items (e.g., mother tongue or language most often spoken at home) in the French and English Canadian cross-cultural studies are likely of questionable validity as ethnic classification variables.

Between Group Differences: Food and Personal Product Consumption

The investigation of the between group differences in consumption patterns for food and personal items controlled for the influences of several covariates. As has been noted, there may be differences in various life styles between English and French Canadians which could lead to different consumption behaviors. Five life style dimensions were included in MANCOVA as covariates: traditionalism, financial conservatism, extravagance, home orientation, and health consciousness. Also included as covariates were a group of demographic variables: age, household income, household size, working status, education, and sex of the respondent.

FIGURE 1

FOUR GROUP CLUSTER SOLUTION

TABLE 1

PREDICTIVE ACCURACY OF EACH COMMUNICATION VARIABLE

Results of MANCOVA on 35 personal and food items after removing the effects of the life style and demographic covariates are contained in Table 2. As shown at the bottom, multivariate tests of difference among the four groups produced a significant result (p<.001). A further examination of the univariate analysis results revealed ten items for which a significant between-group difference existed. Among these significant items, the strong English groups showed a highest consumption frequency for four (canned baked beans, popping corn, frozen vegetables, and headache/pain relievers), the moderate English group in two (lamb and paper plates), the moderate French group in one (pasta), and the strong French group in three (maple syrup, frozen pies/cake, and electric shaver).

Partially supported by these results is the French Canadian's liking for sweets. The most frequent consumption of maple syrup and frozen pies/cakes reported by the strong French group confirms the past findings. However, the group showing the second highest average consumption frequency for both of these two items was the moderate English group, closely followed by the moderate French, and then by the strong English. The data in this study does not show a monotonic consumption trend among the four groups in this instance. For the other sweet items included in the analysis, chocolate/candy bars and cola, no significant difference was found in the consumption frequencies of the four groups.

The results also provide some clues to detecting differences among the groups in convenience food consumption. On the basis of the existing knowledge, the strong English are expected to be most convenience oriented in food consumption. The results, again, provides only a limited support for this expectation. The strong English reported the highest incidence of consuming frozen vegetables and canned baked beans, but no significant between-group differences were found for other convenience food items such as frozen TV dinners, canned soups, and ready-mix cakes. Frequency patterns among the four groups for frozen vegetable and canned baked bean consumption do show increasing consumption trends from the strong French to the strong English.

Another finding of interest is the strong English group's highest consumption frequency for headache/pain relievers, followed by the moderate English,- and then by the two French groups. This is contrary to the earlier finding by an industry study by J. Walter Thompson Co. Ltd. (1969). However, the study is 20 years old, and some consumption patterns of both English and French Canadians may easily have changed. What may be reflected in the present finding is the stronger emphasis by French Canadians on Joie de vivre. The relatively more pleasure oriented and relaxed life style of French Canadians may be associated with a less need for these remedies.

SUMMARY AND CONCLUSION

Several outcomes of this paper are worth noting. In particular, the cluster analysis based procedure using interpersonal and mass media communication variables proved to be effective in identifying four groups of varying degrees of English-French ethnicity (or acculturation). It provided a basis for a more rigorous examination of the effects of varying levels of acculturation on consumption behavior in the bicultural Quebec environment. It was also shown that no single communication variable was capable of producing classification results comparable to those obtained from the twelve item cluster procedure.

The ethnic classification scheme employed in this study incorporated a realistic assumption that acculturation is a two-way, reciprocal process affecting both cultures in consideration. A look at the profiles of the classified groups revealed that English Canadians were acquiring French communication skills perhaps nearly as much as French Canadians were acquiring English communication skills.

TABLE 2

MANCOVA ON FOOD AND PERSONAL PRODUCT CONSUMPTION FREQUENCY

Substantively, the analysis on the consumption differences among the four groups of different English-French ethnicity indicated significant multivariate results for food and personal product items. First, without the strong and moderate categorization of each group, results produced some supporting evidence for the French Canadian's liking for sweets, highly limited support for the English Canadian's stronger convenience orientation in food consumption. Results also showed a higher dependency on headache and pain relievers by English Canadians. The tests of between group differences made in this study controlled for several demographic and life style variables, thus, are more rigorous than those made in the past.

With regard to the monotonicity in the consumption differences among the four groups on the significant consumption variables, the overall patterns of findings provide weak support. For almost two thirds of the significant items (9 out of 14), mean levels of the moderate ethnicity groups either fell outside the range defined by those of their strong ethnicity counterparts or when in the middle range, did not show the monotonicity between themselves. Thus, these observations are more congruent with those of Wallendorf and Reilly's study (1983) on Mexican Americans' consumption behavior, which did not corroborate the linearity assumption between acculturation and consumption patterns. Yet, while Wallendorf and Reilly observed the Mexican Americans' consumption of several food items exceeding both the Anglo Americans' and Mexicans in Mexico, this was not a noticeable pattern in this study. Also to be noted here is that the test of monotonicity using 4 groups is more stringent than the 3 group classification used by Schaninger, Bourgeois and Buss (1985).

In conclusion, this study has demonstrated that the traditional approach to ethnic identification, especially in the study of English and French Canadian consumption differences, is too simplistic, and that within group differences in each major subcultural group resulting from varying degrees of acculturation are strong and warrant a closer examination

REFERENCES

Barnes, James G. and Jacques C. Bourgeois (1977), "Factors Which Influence Per Capita Consumption of Beverage Alcohol," Report to the Research Bureau, Non-Medical Use of Drugs Directorate, Health and Welfare, Canada, ERD, 77-104.

Belk, Russ, John Sherry Jr. & Melanie Wallendorf (1988), "A Naturalistic Inquiry into Buyer and Seller Behavior at a Swap Meet," Journal of Consumer Research, Vol. 14, pp. 449-470.

Bergier, Michel J. (1986}, "Predictive Validity of Ethnic Identification Measures: An Illustration of the English/French Classification Dilemma in Canada," Journal of the Academy of Marketing Science, 14 (Summer), 3742.

Cervantes, F.J. (1980), "The Forgotten Consumers: The Mexican-Americans," in Proceedings, 1980 American Marketing Association Educators Conference, 180-183.

Deshpande, Rohit, Wayne D. Hoyer and Naveen Donthu (1986), 'The Intensity of Ethnic Affiliation: A Study of the Sociology of Hispanic Consumption," Journal of Consumer Research, 13 (September), 214-220.

Dumazedier, Joffre (1967), Toward the Society of Leisure, New York, Free Press.

Gillet, Peter L. and Robert A. Scott (1974), "Shopping Opinions of Mexican-American Consumers: A Comparative Analysis," in 1974 Combined Proceedings, ed., R.C. Curhan, Chicago: American Marketing Association, 135141.

Hirschman, Elizabeth C. (1981), "American Jewish Ethnicity: Its Relationship to Some Selected Aspects of Consumer Behavior," Journal of Marketing, 45 (Summer), 102-110.

Kapoor, S. and W. Williams (1979), "Acculturation of Foreign Students by Television," Paper presented at the International Communication Association Convention, Philadelphia, PA.

Kim, Young Yun (1977), "Communication Patterns of Foreign Immigrants in the Process of Acculturation," Human Communications Research, 4 (Fall), 66-77.

Kim, Young Yun (1985), "Communication and Acculturation," in Intercultural Communication: A Reader, eds., Sarnovar, Larry A. and Richard E. Porter, Belmont, California: Wadsworth Publishing Company, 379-386.

Mallen, Bruce (1973), "The Present State of Knowledge and Research in Marketing to the French-Canadian Market," in Canadian Marketing: Problems and Prospects, eds., Thompson, D.N. and D.S. Leighton, Toronto: Wiley of Canada, Ltd., 98-112.

Mallen, Bruce (1977), French Canadian Consumer Behavior: Comparative Lessons From the Published Literature and Private Corporate Marketing Studies, Montreal: Advertising and Sales Executives Club of Montreal.

McCracken, Grant (1986), "Culture and Consumption: A Theoretical Account of the Structure and Meaning of Consumer Goods," Journal of Consumer Research, 13, pp. 71-84.

Mick, David G. (1986), "Consumer Research and Services: Exploring the Morphology of Signs, Symbols and Significance," Journal of Consumer Research, 13, pp. 196-2.

Mendelsohn, H. (1964), "Sociological Perspectives on the Study of Mass Communication," in People, Society, and Mass Communication, eds., Dexter L.A. and D.M. White, New York: Free Press of Glencoe, 29-36.

Nunnally, J.C. (1967), Psychographic Theory, New York: McGraw Hill.

O'Guinn, Thomas C. and Ronald J. Faber (1985), "New Perspectives on Acculturation: The Relationship of General and Role Specific Acculturation with Hispanics' Consumer Attitudes," in Advances in Consumer Research, Vol. 12, eds., Elizabeth C. Hirschman and Morris B. Holbrook, Provo, UT, Association for Consumer Research, 113-117.

Padilla, AN., ed. (1980), Acculturation: Theory, Models and Some New Findings, Boulder, Colorado: Westview.

Pedone, R.J. (1980), The Retention of Minority Language in the United States, Washington, D.C.: National Center for Education Statistics.

Peterson, T., J.W. Jensen, and W.L. Rivers (1965), The Mass Media and Modern Society, New York: Holt, Rinehart and Wiston.

Pierce, Robert C. (1980), "Dimensions of Leisure: Satisfactions," The Journal of Leisure Research, Vol. 12, No.l, 5-19.

Saegert, Joel, Robert J. Hoover and Marye T. Hilger (1985), "Characteristics of Mexican American Consumers," Journal of Consumer Research, 12 (June), 104- 109.

Schaninger, Charles M., Jacques B. Bourgeois and W. Christian Buss (1985), "French-English Canadian Subcultural Consumption Differences," Journal of Marketing, 49 (Spring), 82-92.

Shibutani, T. and K.M. Kwan (1965), Ethnic Stratification: A Comparative Approach. New York: MacMillan Co.

Tamilia, Robert D. (1988), 'The Duality of Canadian Culture: Toward an Understanding of the Quebec and French Canadian Markets", Working Paper, No. 01-88, University of Quebec at Montreal.

Tigert, Douglas J. (1973), "Can a Separate Marketing Strategy for French Canada be Justified: Profiling English-French Markets Through Life Style Analysis," in Canadian Marketing: Problems and Prospects, Donald N. Thompson and David S. Leighton, eds., Toronto: Wiley of Canada, Ltd., 119- 131.

Tzu, Lao (1984), "Strangers' Adaptation to New Cultures," in Communicating with Strangers: An Approach to Intercultural Communication, Gudykunst, William B. and Young Y. Kim, Addison-Wesley Publishing Company, Inc., 205222.

Valencia, Humberto (1985), "Developing an Index to Measure Hispanicness," in Advances in Consumer Research, Vol. 12, eds., Elizabeth C. Hirschman and Morris B. Holbrook, Provo, UT: Association for Consumer Research, 118-121.

Vickers and Benson (1972), "Cherchez la Femme," Marketing, 31 (January), 2.

Wallendorf, Melanie and Michael Reilly (1983), "Ethnic Migration, Assimilation, and Consumption," Journal of Consumer Research, 10 (December), 293-302.

Wen, K. (1976), 'Theories of Migration and Mental Health," Social Science and Medicine, 10, 297-306.

Yinger, Milton J. (1985), "Ethnicity," Annual Review of Sociology, 11, 151-180.

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