Luxury Possessions and Practices: an Empirical Scale

Citation:

Bernard Dubois and Gilles Laurent (1995) ,"Luxury Possessions and Practices: an Empirical Scale", in E - European Advances in Consumer Research Volume 2, eds. Flemming Hansen, Provo, UT : Association for Consumer Research, Pages: 69-77.

European Advances in Consumer Research Volume 2, 1995      Pages 69-77

LUXURY POSSESSIONS AND PRACTICES: AN EMPIRICAL SCALE

Bernard Dubois, Groupe HEC

Gilles Laurent, Groupe HEC

ABSTRACT -

The market for luxury goods has experienced recently a considerable growth. To a large extent, it can be attributed to a powerful "democratization" trend. One used to segment the population in two groups : the "affluent" who had a permanent access to luxury, and the "excluded", who had no access. We argue that a major role on this market is now played by "excursionists" who have an intermittent access to luxury. Accordingly, access to luxury is no longer a dichotomous state, but a matter of degree. In this paper, we develop and test an empirical scale that measures to which degree a person is "immersed" into luxury.

INTRODUCTION

Considered over the last ten years, the market for luxury goods has experienced a considerable growth. The Comite Colbert, which includes most prestigious French luxury brands such as Chanel, Christian Dior, Guerlain, etc. (In 1989, the McKinsey company established that French brand names enjoyed a 50% market share of the world luxury market, Joly, 1991) reported a global turnover of less than FF 15 billion in 1985. The corresponding figure for 1993 was in excess of FF 30 billion (Comite Colbert, 1994). For specific companies, the growth is even more impressive. While in 1977 the Louis Vuitton company was a family business with sales under $ 20 million, its 1993 sales were above $ 1 billion.

To a large extent, such a spectacular growth can be attributed to a powerful "democratization" trend according to which products and brands which used to be rather exclusive are now widely consumed by the public. In 1993, Dubois and Duquesne reported that, over the last two or three years, more than one European out of two had acquired at least one luxury product, chosen from a list of fifteen items. Similarly, Dubois and Laurent (1993) indicated that each European consumer had acquired over the last two years an average of two brands, out of a list containing thirty internationally famous luxury brand names.

Of course, such an evolution has had considerable impact upon the structure of the demand for luxury products. Until recently (see Figure 1), it was possible to identify two major consumer segments in this market: 1) the Excluded, who, in most countries, comprised a vast majority of the population, without access to luxury, and: 2) the "Afflluent" (Stanley, 1988), who themselves could be subsegmented into two clusters: "Old money" (Hirschman, 1988) and the "Nouveaux Riches" (LaBarbera, 1988). It would however appear that nowadays, a major part of the market consists of "Excursionists" (Dubois, Enel, and Laurent, 1994), a third group of consumers who, in certain product categories such as perfumes, could account for more than three purchases out of four (see Figure 2). In opposition to the Excluded, for whom the world of luxury is, at best, a dream, Excursionists do have access to luxury items. But in contrast to the Affluent, for whom Luxury is an "art de vivre," their acquisition and consumption of luxury items is intermittent, often linked to exceptional situations or circurnstances (Dubois and Laurent, 1994). Although their number and weight are difficult to estimate, a recent survey indicates that in the major luxury markets (Europe, USA, and Japan), they comprise between 25% (Spain) and 38% (Italy) of the population (Dubois and Paternault, 1994).

Such an evolution of the luxury demand structure calls for new marketing research instruments. If, in the past, affluent consumers were reasonably easy to identify, for example on the basis of their economic resources (Stanley, 1989) or conspicuous behavior (Mason, 1981), excursionists are much more difficult to profile. Given their emerging predominance, consumer involvement in the luxury market is no longer a dichotomous state but becomes a matter of degree. To measure and quantify a given level of "immersion" into luxury, a scale, rather than a simple yes/no instrument is now needed. The objective of this paper is to report on the development of such a scale.

METHOD

In developing our "immersion" scale, we identified several specifications that such a scale should respect if it was to be later routinely used by interested researchers and luxury companies. First, such a scale should assess in unambiguous and quantitative terms, i. e. with at least interval properties (Stevens, 1946), the degree to which an individual participates in (or, on the other hand, is excluded from) the world of luxury. Second, we wanted our scale to be general, that is, relatively independent of any particular brand or product/service category as well as from consumer personal characteristics such as for example gender or national affiliation (so as to allow for international comparisons). Third, our scale should be reasonably short, simple to administer and objective; it should be easy to include in a standard market survey. For all these reasons, we came to the conclusion that our scale should be based on actual behavior rather than cognitions, opinions, or judgments. Validity and reliability were of course essential properties but would have to be assessed from empirical data (Churchill, 1979).

We then turned our attention to scales which had been recently introduced in Europe in order to harmonize the measurement of economic status, in particular the Esomar scale developed in the late 80's and early 90's by an ad-hoc research group (Esomar Working Group, 1984; Marbeau, 1991; Quatresooz and Vancraeynest, 1992). We discovered however that such scales are exclusively based on possessions, a necessary but not sufficient condition for examining behavior related to luxury. Preliminary qualitative research results, obtained from an analysis of sixteen in-depth interviews with consumers conducted and interpreted by a professional psychologist, had convinced us that to understand consumers' rapport to luxury, we had to go beyond ownership of luxury items. When asked to comment upon their behavior in the domain of luxury, respondents did not talk primarily about what they had but rather about what they did, for example in their leisure time. It then became clear that we had to investigate practices and not only possessions (Holbrook and Hirschman, 1982). In addition, considering practices allows us to take into account services, as well as products. Too often, research on luxury has limited itself to products, while one can find obvious examples of luxury services (e.g., restaurants and hotels). Based on the products and practices spontaneously mentioned by consumers during the interviews but also on the basis of the existing literature, we decided to explore a total of 25 product categories and 19 practices (see Table 1). Included in the list were a few products (such as a soccer ball and a bicycle) and a few practices (such as watching TV question and answer games and betting money on horses) we thought would not be related to luxury behavior, so as to allow for an analysis of the criterion and discriminant external validity of our scale.

As part of a broader survey aimed at understanding attitudes and behavior towards eight specific luxury and upscale products, a sample of 440 French Consumers were asked to report ownership for each of the 25 product categories and to report whether or not they had practiced any one of the 19 activities over the last twelve months. Our sample has a - ]ready been described elsewere (Dubois and Laurent, forthcoming) and therefore will not be discussed in detail here. Suffice it to say that although not randomly drawn, the sample was chosen according to quota set in terms of sex, age, and geographical location. Given the nature of the topic under investigation, it was also decided to underrepresent lower income categories. All interviews were conducted on a face-to-face basis and the ieldwork was carried out by one of the major professional market research companies operating in France.

FIGURES 1 AND 2

RESULTS

In order to build our scale, and in accordance with the specifications mentioned above, we first crosstabbed ownership or practice with gender and removed all the product and practices which were too strongly associated with it (see Table 2). This eliminated obvious products and practices associated with either women (visiting a beauty parlor, owning a perfume, a silk scarf, a necklace of pearls or a diamond ring) or men (drinking a bordeaux "grand cru", owning a soccer ball, going camping, betting on horses), but also sonic less obvious ones (owning rock music records, having read more than ten books, owning a credit card). We then removed, if t had not been already done, all the products (3-4 star hotels, suit or dress made to measure, credit cards, perfume, bordeaux wine, diamond ring, silk scarf ) which were the focus of the larger survey, deciding that for the purpose of the present research, we would use them as criteria for measuring the external validity of our scale.

We then performed a Multiple Correspondence Analysis (Lebart et al., 1984) of the 31 remaining possessions or practices. A Burt table was created by crosstabing answers (62 lines and columns, since each possession or practice is coded as a binary variable), and input into a Correspondence Analysis program (CORRESP in S A S). The results produced a strong first factor, linked to almost all possessions and practices that we had hypothesized to be of a luxury nature. Six products and one practice had been deliberately included in the list ecause we thought they would

not be related to luxury behavior: a color TV set, a R.V., a video recorder, a soccer ball, a bicycle, a rock music record, watching V games. Indeed, they had a low partial contribution to the first factor, and were therefore excluded. Out of the 24 remaining practices and possessions, we eliminated golf equipment and playing golf, because there were very rare in our sample (3 %). Finally, we removed six more possessions and practices which, contrary to our expectations, had lower partial contributions to the inertia: a fur coat or vest, a dog or cat with a pedigree, a micro-wave oven, playing a musical instrument, selling or buying shares, traveling first class by train. After this long process of analysis and deliberation, we therefore finally decided to keep a list of eight practices and eight possessions (Table 3).

In order to further assess the scalability of such a list, we performed a multiple correspondence analysis limited to these 16 possessions and practices. Table 4 presents the results obtained on the first dimension. They indicate that the list of possessions and practices is indeed scalable and lead to a consistent scoring theme. The fact of owning items which are seldom possessed, or practicing an activity which is seldom practiced, receives a high score and vice versa. It corresponds to a higher level of "immersion" into luxury.

The high principal inertia (24.4 %) associated with the first dimension indicates well-patternd scores. Taken together ith the smaller values obtained for the second (9.5 %), third (7.1 %), etc. factors confirm the unidimensionality of the scale. It thus appears that, for each consumer, an immersion score can be calculated. Following Guttman (1944), one could simply assign scores on the basis of the number of possessions and practices. Based on correspondence analysis, one could also use the scores directly obtained on the first factor. Figures and 4 illustrate the distribution of values obtained with each method. They are, of course, very similar, and , as expected, strongly correlated (R=0.995 ). The Excluded are perhaps easier to identify with the second method, while the really Affluent , who clearly emerge in both cases, correspond to roughly 10% of our sample (those with a Guttman score higher than 12 or a factorial score higher than 8). For the sake of simplicity, it was therefore decided to use the first score (Guttman sum) in the remainder of the present analysis.

TABLE 1

LIST OF PRODUCTS AND PRACTICES

To have a better understanding of which products and practices are representative of high level of immersion, Figures 5 and 6 were prepared. They indicate, for each set of products and practices, tile proportion of positive responses among five groups of consumers defined on the basis of the immersion scale score. Such figures clearly reveal the exclusive roles played by the ownership of a luxury car (a car with more than 12 HP) and by the acquisition of an antique. On the other hand, listening to music or possessing a tennis racquet are much more common behavior. Such figures also confirm the unidimensionality of the scale by displaying lines which, with a few exceptions, do not cross each other. Most of those who, for example, own a big car already owned most of the other items. Similarly, purchasing an antique is only done by those who have already practiced many other activities on the list. Table 5 displays the percentages observed for each category.

Finally, to assess the convergent and discriminant validity of 1959), a series of logistic regressions were run. We hypothesized that a high score on the immersion scale would adequately predict ownership of products or practices clearly associated with luxury behavior, while they would lack any predictive power for more "mundane" products. We decided to select two specimens of each kind for each category (products and practices) i. e. a total of eight items. An effort to contrast items in terms of monetary unit value and level of ownership (or practice) resulted in the following list: a bottle of Bordeaux "grand cru", a silk scarf, staying at a 3-4 star hotel, and using a credit card abroad (for luxury products and practices); owning a soccer ball, owning a TV set, betting on horses and watching TV games (for the second group). Results are presented in Table 6 and clearly establish the discriminant and convergent validity of our scale.

For example, while the average ownership of a bottle of Bordeaux Grand Cru is 44% in the whole population, it jumps to 93% for an individual with a score of 15 and drops to 13% for an individual with a score of zero. Conversely, the probability of owning a soccer ball (32% on average) is much less affected (39% and 29% respectively). While less contrasted, results obtained for the other products were of a very similar nature. Our scale offered significant convergent validity for products such as a fur coat, a diamond ring, a golf equipment or a necklace of pearls, while exhibiting hardly any discrimination for a bicycle, a RV, or a pet. In terms of practices, very significant results were obtained for playing a musical instrument, playing golf, selling/buying shares, visiting a beauty parlor, or riding the train first class. As expected, results were much less impressive for camping, or even reading books. We also checked that our scale almost always outperformed income, a sociodemographic variable traditionally used in the analysis of the affluent. As an example, Table 7 exhibits the results obtained for the four luxury products and practices selected previously.

TABLE 2

POSSESSIONS, PRACTICES, AND GENDER

TABLE 3

FINAL LIST OF POSSESSIONS AND PRACTICES

TABLE 4

CORRESPONDENCE ANALYSIS SCORES FOR THE FINAL LIST OF POSSESSIONS AND PRACTICES (FIRST FACTOR)

FIGURE 3

IMMERSION SCORES BASED ON GUTTMAN SCORING SYSTEM

FIGURE 4

IMMERSION SCORE BASED ON CORRESPONDENCE ANALYSIS

FIGURE 5

POSSESSION OF EACH GOOD AS A FUNCTION OF TOTAL SCORE

CONCLUSION

The objective of this paper has been to introduce a scale designed to measure the extent to which an individual is "immersed" into luxury. The spectacular evolution of the market for luxury goods implies that the tradional distinction between the Affluent and the Excluded now represents a gross oversimplication of the structure of demand. With the emerging predominance of "excursionists," participation in the luxury market is now a matter of degree and calls for new measurement instruments. In fact, the introduction of a clear-cut, intermediary segment of "excursionists" (Figure 2) between the "affluent" and the "excluded" may be an oversimplification. A better view of the market, suggested by figures 3 and 4, could be that of a continuum from the most affluent to the totally excluded, with no clear cut-off point that would allow for an obvious separation between the three segments of Figure 2.

FIGURE 6

FREQUENCY OF EACH PRACTICE AS A FUNCTION OF TOTAL SCORE

TABLE 5

FREQUENCY OF EACH PRACTICE AND POSSESSION AS A FUNCTION OF TOTAL SCORE

TABLE 6

CONVERGENT AND DIVERGENT VALIDITY OF THE IMMERSION SCALE

TABLE 7

COMPARISON OF THE PREDICTIVE POWER OF OUR IMMERSION SCALE AND OF INCOME

Because our qualitative research had indicated that ownership of luxury products was not a sufficient indicator, we decided to include in our research both practices and possessions, which tap on different aspects of the luxury world (their shared variance was found to be less than one third in this study). Empirical evidence has shown that answers provided by respondents to our list of sixteen products and practices possesses good scalability properties. Our list can be used therefore as a measurement scale. Not only is it reasonably easy to implement, objective, and unambigous, but, more important, such a scale exhibits unidimensionality as well as discriminant and convergent validity. We hope our scale will help future researchers in their efforts to understand a fascinating but much under-researched market.

Finally, an obvious limitation of our scale is that it has been developed in France. Some possessions or practices, viewed in France as belonging to the domain of luxury, may be perceived differently in other countries. For example, holding a conversation in a foreign language may not be perceived as a luxury practice in Scandinavia. We are looking forward to an international validation of our scale, from that point of view, in multiple countries.

REFERENCES

Campbell, Donald T. and Donald W. Fiske (1959), "Convergent and Discriminant Validation by the Multitrait-Multimethod Matrix", Psychological Bulletin, Vol. 56, pp. 81-105.

Churchill, Gilbert A. (1979) Jr, "A Paradigm for Developing Better Measures of Marketing Constructs", Journal of Marketing Research, February, vol. 16, pp. 64-73

Comite Colbert (1994), Rapport 1993 et Perspectives 1994, Paris: Comite Colbert

Dubois, Bernard and Claire Paternault (1994), "Le Luxe en Europe: Demande Comparable mais Visions Contrastdes", Communication Business News, Decembre 19, n*374, pp. 8485.

Dubois, Bernard and Gilles Laurent (1993), "is There a Euroconsumer for Luxury Goods?", in W. Fred van Raaij and Gary J. Bamossy (Eds), European Advances in Consumer Research, Vol. 1, Provo, UT, Association for Consumer Research, pp.58-69.

Dubois, Bernard and Gilles Laurent (1994), "Product Upscale Proneness: A Situational Approach", HEC Working paper.

Dubois, Bernard and Gilles Laurent (1994), "Attitudes towards the Concept of Luxury: An Exploratory Analysis", in Siew Leng Leong and Joe Cole (Eds), Asia-Pacific Advances in Consumer Research, Vol. 1. Provo, UT, Association for Consumer Research, pp. 273-278.

Dubois, Bernard and Patrick Duquesne (1993), "The Market for Luxury Goods: Income versus Culture", European Journal of Marketing, Volume 27, n' 1, pp.35-44.

Dubois, Bernard, Francoise Enel, and Gilles Laurent (1994), "La Face Cachce du Luxe: Ruses et Excursionnisme", HEC Working paper.

Esomar Working Group (1984), "A Step Forward in International Research: Harmonization of Demographics, for Easier International Comparison", European Research, October, pp. 182-189.

Guttman, Louis (1944), "A Basis for Scaling Qualitative Data", American Sociological Review, Vol. 9, pp. 139-150.

Hirschman, Elizabeth (1988), "Upper Class Wasps as Consumers: A Humanistic Inquiry", in Elizabeth Hirschman (Ed), Research in Marketing, Vol. 3, JAI Press Inc., pp. 115-147.

Holbrook, Morris B.and Elizabeth C. Hirschman, "The Experiential Aspects of Consumption: Consumer Fantasies, Feelings, and Fun", Journal of Consumer Research, September, Vol. 9, n'2, pp. 132-140.

Joly, Hubert (1991), "Industrie du Luxe: "Rebondir" sur la Crise: Leviers pour le Succes dans les Ann6es 1990", Revue Franqaise du Marketing, n' 132-133, pp. 97-109.

LaBarbera, Priscilla A. (1988), "The Nouveaux Riches: Conspicuous Consumption and the Issue of Self-fulfillment", in Elizabeth Hirschman (Ed), Research in Marketing, Vol. 3, JAI Press Inc., pp. 179-210.

Lebart, Ludovic, Alain Morineau, et K. L. Warwick (1984), Multivariate Descriptive Statistical Analysis: Correspondence Analysis and Related Techniques. New York: NY: J. Wiley and Sons

Marbeau, Yves (1992), "Harmonization of Demographics in Europe 199 1: The State of the Art", Marketing and Research Today, March, pp. 33-40.

Mason, Roger (198 1), Conspicuous Consumption, New York: N. Y.: St Martin's Press.

Quatresooz, Jean and Didier Vancraeynest (1992), "Using the Esomar Harmonized Demographics: External and Internal Validation of tile Results of the Eurobarometer Test", Marketing and Research 'Today, March, pp. 41-48.

Stanley, Thomas J. (1988), Marketing to the Affluent, Homewood, Ill.: Irwin.

Stanley, Thomas J. (1989), Selling to the Affluent, Homewood, Ill.: Irwin.

Stevens, S. S. (1946), "Oil the Theory of Scales of Measurement", Science, Vol. 103, pp. 677-680

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