1 Several empirical studies have revealed the existence of a certain malaise among the middle classes, both in France and in other developed societies. Social fragmentation, another way to describe what has been called the decline of the middle classes, as analyzed by sociologists, has been significantly reflected in the programs of the political parties of several countries from 2000–2010 (particularly France, Germany, and the U.S. [1]). Unemployment, which private sector employees have been particularly affected by, is often cited as the main cause of this anxiety, which has reached the center of the social structure. Added to this are the slow down or even the end of growth in real income for salaried employees, the decline in upward social mobility, and the fear of downward mobility which is associated with it—for individuals themselves or for their children—as well as the difficulties of integration into employment for young people and new generations entering the labor market. In North America, rising inequalities, and specifically, the concentration of wealth and regular income in the highest quintile, have attracted the greatest attention, rather than the decline of the middle class. The theme of decline, which emerged in the 1980s, was subsequently expanded to include the study of inequalities.

2 Reports of the crisis affecting the middle classes and the analysis of inequalities in developed societies are often made in relation to the macro-economic (globalization and the transfer of jobs to emerging developed economies) or macro-sociological (generational shifts in employment patterns and ageing) context. In this article, we will take a different approach to analyzing the situation of the middle classes (and, by extension, that of other socio-economic classes) in two developed countries, Canada and France, by focusing on their inclusion in contemporary consumer society from a longitudinal perspective. Studying the endogenous situation of the socio-economic classes from a consumption perspective may shed new light on questions of social fragmentation and social cohesion in addition to the structural factors or macro contexts often cited, such as the increase in unemployment, generational conflict, or even the rigidity of social institutions. In order to do this, we will use the classical sociological approach set out by Maurice Halbwachs in his work on consumption and the social classes.

3 We shall study the malaise affecting the middle classes by examining the patterns of their budgetary expenses, as well as exploring the aspirations and expectations that household members have developed with regard to consumption. Just as early surveys of household budgets revealed the misery of the workers during the industrial revolution, [2] we suggest that analyzing the actual consumption of households, as well as their consumption aspirations, will help to explain the malaise described above. We will adopt a comparative approach, looking at two different societies on either side of the Atlantic, as well as a longitudinal approach. Household budget surveys carried out by the national statistical bodies of these two countries will be used as data sources. As they are well established and reliable, we will spare the reader the technical details behind this data, which can easily be found in other publications or on the Institut national de la statistique et des études économiques (INSEE) (National Institute of Statistics and Economic Studies) or Statistics Canada websites. We will first describe the sociological significance of the expenditure categories that are found in these household budget surveys, and provide a working definition of the middle classes, before presenting the pattern of needs revealed by the surveys in these two countries. This will be followed by an analysis of the convergence of consumption patterns among the middle classes, in particular the study of middle class consumption over time in both countries, and finally, the study of social representations and aspirations.

Comfort, Pleasure, and Distinction

4 From the point of view of social actors, commodity consumption pertains to three kinds of motivational logic, i.e. three overall reasons for action that have been well theorized by various authors: pursuit of comfort, pursuit of pleasure, and pursuit of distinction. A great number of empirical studies have revealed the pertinence of these three motivations for action in research on different aspects of commodity consumption, which we will very briefly review. This review is necessary because these three motivations for action come together in a specific way among the middle classes, providing an endogenous explanation for the malaise that they so strongly feel.

5 For Maurice Halbwachs (1927), commodity consumption was at the heart of contemporary lifestyles, offering a guarantee of day-to-day comfort and well-being. This theory is even more relevant for the enlarged consumer society which was established in the second half of the twentieth century. Working outside the home requires the consumption of goods and services (clothing, transport, eating out, childcare costs, house cleaning, etc.) and rest periods are dominated by the consumption of cultural products, from meals in restaurants, to goods and services (cable, pay TV, etc.) for leisure activities at home, as well as holidays away from the home. Socializing is not exempt from this, because it also involves the consumption of goods and services (travel costs, telephone and communications expenses, outings with friends, presents for parents and friends, babysitters, etc.). [3] In other words, households consume by necessity in their daily lives and they seek a comfortable lifestyle. Indeed, it was precisely the comfort of English households that struck H. Taine on his famous visit to England in the nineteenth century, which at that time was ahead of other countries with regard to the material ease of working-class families.

6 Households do more than just consume in order to ensure that their members have a comfortable life however. Household members also have aspirations and expectations that are linked to the objects consumed, as Albert O. Hirschman (1984) demonstrated when he talked about induced consumption. To own a car or a house in the suburbs induces an entire lifestyle “which goes with the possessed object,” or even, for a couple, giving birth to a child entails a (long!) list of expenses that become inevitable. Aspirations to consume are also influenced by the connections that individuals have to others. Therefore, to give just one example, owning a car means keeping an object which stimulates further consumption (e.g. trips to the country) but also an object which serves as a marker of social status, which leads us to the second motivation for action: the pursuit of distinction, which is well known by sociologists. This motivation for action was set out in France by Edmond de Goblot (1927) and before him in Germany, by G. Simmel (1902). However, Simmel’s work was not translated into French or English until later, which limited his influence on the development of this analytical approach to household consumption. The notion of distinction is also present in Thorstein Veblen’s theory of conspicuous consumption (Theory of the Leisure Class 1899). Following on from this earlier work, the question of distinction was developed at great length by Pierre Bourdieu in La Distinction (1979). He suggests that consumption reflects the meanings produced by the class system and that existing differentiation in the pre-existing system of social relations is reflected by consumer choices. The contemporary perspective on distinction is different (less “final,” as Bourdieu’s critics would say), characterized as the symbolic appropriation by social actors of the consumed objects, transformed into cultural objects by man. As Goblot pointed out, “The distinction of the classes is a question of value judgmen” (1967 [1925], 15).

7 More broadly, distinction is no longer "the driver of the creation of needs," as Jean Baudrillard gave us to understand, but one aspect among others, and this aspect may even be relegated by a third aspect to which we now attach a great deal of importance: the pursuit of pleasure, or hedonism.

8 In his essay on fashion, Gilles Lipovetsky proposes that the pursuit of pleasure is the main driver of consumption for individuals who have “become indifferent to the judgment of others.”

9

The goal pursued through objects is not so much legitimacy and social superiority as private satisfaction, which is increasingly indifferent to the judgments of others. Essentially, consumption is no longer an activity regulated by the search for social recognition, it is deployed to obtain well-being, functionality, and pleasure for oneself.

10 This somewhat overstates the case, as the search for distinction is still pertinent, but it should be considered as one dimension among others. Moreover, the pursuit of pleasure (or even hedonism) is of growing and central importance in contemporary consumption. In his evocatively titled book, The Romantic Ethic and the Spirit of Modern Consumerism (1987), Colin Campbell maintains that consumption is linked to the insatiable nature of need, the endless desire based on the experience of pleasure felt by individuals in association with the consumption of objects. For Campbell, the hedonistic approach was characteristic of the emerging middle classes of England in the eighteenth century that were the subject of his research, and this is an important idea for our purposes. The Protestant middle classes of this era expressed their tastes through the consumption of beautiful objects, thus revealing their moral distinction from the poorer classes, who were trapped in the world of need.

11 The effects of these three motivations for action come together in a way that can lead to frustration among the middle classes, even more so than among the poor or very wealthy classes. Why? To answer this question, we must turn to the paradox of relative satisfaction outlined by Richard Easterlin in his classic and frequently cited 1973 article, which took up an idea that was already present in the work of Montaigne, Alexis de Tocqueville, and Émile Durkheim. To quote Montaigne (1580): “When it is far away, the object of our desires seems to be more important than anything else. Once it is in our possession, we desire other things: so great is our thirst.”

12 Easterlin showed that the average level of satisfaction among the populations of various countries with regard to their standard of living remained stable during the period of significant increase in living standards of the thirty golden years (1945–1975), a paradox verified time and again in different national contexts (sometimes called into question, but this is not the place to settle this debate).

13 Entering the world of consumption entails entering the world of aspirations, in the sense given to this term by S. Moscovivi (1960). But the aspiration to consume more and the sense of relative privation can be found most acutely within the middle classes, and they are the endogenous source of the malaise mentioned previously. Poor households and the disadvantaged socio-economic classes are effectively trapped in a world of basic needs, and they have less objective possibilities of projecting into the future and developing a sense of frustration. The poorer classes are preoccupied with daily survival. “It’s not for us” is a common response in surveys asking household members about their needs. They also express strong feelings of being deprived of objects considered essential. They have very little room for maneuver to enter the world of aspirations. Well-off households on the other hand, already have sufficient resources to satisfy their needs as well as their aspirations and they express few or no feelings of privation or frustration. It is completely different among the middle classes, who have experienced extended consumption, but without necessarily having the financial means to satisfy all their aspirations, which grow faster than the available monetary resources. This is a brief overview of the context for growing feelings of frustration or the typical malaise that can be found among the middle classes of several contemporary developed societies.

14 To demonstrate this, we will analyze changes in French and Canadian household expenses over a long period and we will examine the results of surveys on the consumption categories that are the object of aspirations and privations, in order to compare three broad groupings of socio-economic classes (disadvantaged classes, middle classes, and well-off classes). However, prior to this, we must provide further details on the definition of the middle classes and changes to their standard of living.

Defining the Middle Classes

15 Maurice Halbwachs was one of the first to note the great heterogeneity of the middle classes, which he emphasized time and again in the courses on the middle classes that he taught at the Sorbonne in the 1930s, published by Presses Universitaires de France in 2008. His definition includes a collection of new professions and emerging trades from the first third of the twentieth century, particularly salaried workers. But Halbwachs does not limit himself to defining social classes by their profession or trade, as he also emphasizes the study of income, and above all, expenses, in his empirical studies. Therefore, in his critical examination of the notion of class, Halbwachs criticizes Marx for having defined the opposition of exploiters and exploited in overly political terms, saying that “it is rather a distinction between the richest and the poorest that must be introduce” (Halbwach 2008, 61). According to him, such a distinction would allow a link to be made between the productive activity of individuals and the distribution of income received by workers and employees, in particular.

16

As wealth, evaluated in monetary terms, has degrees that can be precisely measured, an income statistic would also give us a way of measuring the level of the different social classes and the intervals that separate them. […] But where would the cut-off point be? For this purpose, we must not only consider wealth, but also expenditure.

17 There is another reason for choosing revenue as a priority indicator for studying the middle classes: the fact that its members are made up of salaried workers. In his classic study, White Collar: the American Middle Classes (1951), C. W. Mills also emphasizes the diversity of the social composition of the middle classes, which includes civil servants, the range of non-manual salaried workers, teachers, and members of new professions in a rapidly expanding urban world. For Mills, two common features provide a certain unity to an American middle class in full growth in the middle of the twentieth century: belonging to a bureaucracy and being a salaried employee. Mills makes the observation that new employees have effectively replaced artisans, small proprietors, and the self-employed in number among the middle classes, noting that earning a salary gives them access to greater consumption of goods and services, allowing them to construct a new lifestyle. We should also note that the growth of the middle classes was fueled by the establishment of the welfare state, state regulation of labor, and the development of unions for workers and salaried employees.

18 Serge Bosc published an exhaustive inventory of sociological studies of the middle classes in a synthesis study, Sociologie des classes moyennes (2008), to which we refer the reader. The author characterizes the contemporary middle classes as “a composite grouping, with multiple divisions and contradictions” (Bosc 2008, 27), and he observes in conclusion that they “can be seen as a grouping that is both hierarchical and polarized in multiple ways” (Bosc 2008, 108), adding a further detail that is important for our purposes:

19

“One of the main contributions of this corpus has been to show that, at the edges of the so-called middle classes, similarities in the standard of living did not erode the social and cultural distances between them and members of the working classes or the established bourgeoisie”

20 For him, several social worlds (in his words) overlap within this nebulous notion and the individuals who belong to this complex center are not so different in reality to the disadvantaged classes or the well-off classes, according to different criteria such as employment, qualifications, and income. [5]

21 At the very end of his book, Bosc concludes with a question we hope to answer empirically: “Are we therefore talking about a grouping with many forms and without unifying forces, subject to the opposite attractions of the upper and lower classes?” Studying the pattern of household expenses enables us to respond to this question in a precise way, by comparing the needs revealed by the expenses of different groups within the middle classes, in relation to the least and the most well-off sections of it.

Decline of the Center?

24 Let us first of all consider the middle classes as a single grouping in order to define them empirically within the wider socio-economic structure.

25 France is one of the few countries where evaluating the size of the middle class using the method adopted here shows a positive change, which is not the case in other comparable countries such as Germany, the U.S., or Great Britain (Pressman 2007; Bigot 2009). It is not possible to conclude that there has been a decline at the center in France in the last third of the twentieth century or the beginning of the twenty-first century. According to the indicator “disposable income by unit,” (between 75% and 150% of the median, i.e. lines 3 and 4 of Table 1) the middle classes made up 47.9% of all French households in 1979, compared to 50.7% in 2006. A certain decrease in this proportion can, however, be seen from 2000 to 2006 (data not published here).

Table 1:

Population distribution by disposable income per consumption unit (relative to the median), Canada and France, per year

	Canada	France
Relative to the median	1978	1996	2008	1979	2006
< 60 %	17.5	16.5	19.6	17.7	15.8
60-75 %	12.2	12.0	11.6	14.1	12.2
75-100 %	20.3	21.5	18.7	21.0	21.9
100-150 %	28.6	27.6	26.3	26.9	28.8
> 150 %	21.4	22.3	23.7	20.3	21.2
Total	100	100	100	100	100

Population distribution by disposable income per consumption unit (relative to the median), Canada and France, per year

26 The situation is very different in Canada. The data in Table 1, in comparison with the French data, shows the clearly visible decline of the center towards the end of the twentieth century and in the first decade of the twenty-first century, i.e. a roughly equal number of households moving downwards and upwards in the distribution of income per unit. Several factors explain the Canadian situation. Very high levels of immigration (a quarter of a million people enter Canada each year) feed the growth in the number of low-income households, while inequality has increased over twenty years (strong growth in the number of well-off households, particularly those with two high incomes).

27 Furthermore, several studies show that the median standard of living, measured in constant euros, has not fallen in France in the last third of the twentieth century or in the first decade of the twenty-first century. The middle classes are not pauperized, although the relative position of the middle class changed and deteriorated from the middle of the 1990s, as indicated by measuring the relationship between median and mean household income. The closer this measure gets to 100%, the more equal income distribution is (because high incomes bring the mean above the median). From 1970 to 1995, this measure moved from 84% to 90%, showing that the standard of living of the middle classes progressed more quickly than for the French population overall (Bigot 2010). The relative situation of the middle classes therefore improved over this long period. However, the opposite occurred after 1995, as median income fell back to 85% in 2006: the income of the French middle classes has progressed more slowly over the course of the last ten years, according to available data.

28 The trend that has characterized French society since the end of the twentieth century could be seen several years earlier in other countries (Germany, the U.S., and Great Britain), where the very well-off classes experienced a more significant rise in disposable income, separating them in some ways from the rest of society (see the OECD report of 2008). In France, the gap between income levels did not grow as strikingly as in the U.S., although it is possible that social representations of the situation in France, as described above, were influenced by comparisons with the widely reported U.S. situation. The situation for Canadian households shows a similar change to the one seen in France from 2000–2010, i.e. a growing gap between median and mean income levels, as a result of the growth in inequality mentioned previously.

29 Moreover, inflation appears to have affected the middle classes more than the wealthiest classes. In fact, an INSEE study showed that the prices of goods and services consumed by households in the fourth to the seventh decile increased more between 1996 and 2006 (brackets of between 17% and 16.7%, depending on the decile) than the average prices paid in the higher deciles (increase of around 15%). High-income households effectively consume more electronic products and telecoms and other communication services, which saw a price drop in real terms. Likewise, the increase in the cost of accommodation and household bills had a greater impact on the middle classes (and even more so on poorer households, let us not forget). The same applies in Canada. Studying household expenditure patterns is also important in understanding the situation of the middle classes. We begin by studying changes in consumption patterns in the two countries before addressing the question of convergence between the socio-economic classes.

Consumption Patterns in Canada and France

30 We grouped household expenses into eleven consumption categories for France and ten for Canada. For technical explanations of the content of these expenses categories, see INSEE (2009) and Langlois (2003). The categories created for each country are not strictly comparable but they are very close, enough to be able to identify similarities and differences in changes to them over time. However, we must clarify some details on the contents of the different consumption categories. Food includes expenses for the home, but also food eaten outside the home, such as meals eaten during working hours (at lunchtime), or in the canteen. Accommodation expenses in Canada include the repayment of mortgage capital (effectively a loan), as this can be considered as a monthly expense that has an impact on family budgets. This makes sense if we wish to measure the budgetary output of households in each of the categories for any given year. Capital repayments are not included in the expenses of French households, but we made some estimations for rent payments. Leisure includes expenditure on culture, outings, and holidays. The “health” category is intended in the broadest sense and it includes money spent on personal care, such as beauty products. In Canada, protection includes all kinds of insurance (life, unemployment, and health) apart from car and home insurance.

31 Changes in the pattern of spending on consumption is well documented in both countries and has been analyzed in several publications already mentioned. Let us briefly review the main trends of these changes, to provide some context for the question of longitudinal and temporal convergence between the socio-economic classes, which is the next topic. The two countries have several features in common, as indicated in the last column of Tables 2 to 5. There has been an overall noticeable reduction in the share of the budget spent on food in the household budgets of both countries, which can be seen as part of a long-term trend found in developed societies. In Canada, food moved to third place in household spending patterns in 2008 and the accommodation category has been in first place for at least a quarter of a century. This trend is relatively similar in France, with food having gone from 27% in 1979 to 18% in 2006, and accommodation taking second place within spending patterns. The relative amount spent on transport has grown significantly in the spending patterns of both countries, demonstrating the emergence of a new omnipresent need in the contemporary lifestyles of households. The “leisure” category is next, in fourth place, showing the growing importance that it has in modern lifestyles and the commercialization of this type of activity. The culture industries have developed considerably over the last twenty-five years, along with the leisure “industry,” their services feeding the satisfaction of needs that are reflected in household expenses.

Table 2:

Household expenditure patterns for the median standard of living, Canada, 1978

Expenditure category	< 60 % of median	60 % - 75 % of median	75 % - 100 % of median	100 % - 150 % of median	> 150 % of median	All
Food	28.9	25.6	23.9	21.5	19.2	22.1
Accommodation	27.4	24.8	23.1	22.8	22.2	23.2
Transport	10.6	14.7	15.1	15.4	15.7	14.9
Leisure	6.7	7.0	7.9	8.6	9.2	8.4
Clothing	6.9	7.1	7.6	8.1	7.8	7.7
Equipment	7.0	6.8	6.9	7.0	7.4	7.1
Protection	2.3	3.8	5.3	6.0	7.2	5.8
Health	3.8	3.7	3.5	3.4	3.1	3.4
Education	0.9	1.1	1.3	1.3	0.9	1.1
Other	5.5	5.5	5.5	5.9	7.3	6.2
Total	100	100	100	100	100	100

Household expenditure patterns for the median standard of living, Canada, 1978

Table 3:

Household expenditure patterns for the median standard of living, Canada, 2008

Expenditure category	< 60 % of median	60 % - 75 % of median	75 % - 100 % of median	100 % - 150 % of median	> 150 % of median	All
Food	19.0	17.2	16.1	14.8	12.8	14.9
Accommodation	32.8	31.2	29.2	27.3	27.4	28.5
Transport	12.9	16.4	16.6	17.7	16.7	16.5
Leisure	7.9	8.0	9.0	9.1	9.8	9.1
Clothing	5.0	4.1	4.8	4.6	5.1	4.8
Equipment	6.8	6.0	6.0	6.0	6.2	6.2
Protection	3.5	5.5	6.8	8.6	9.2	7.7
Health	5.6	5.7	5.3	4.7	4.1	4.7
Education	2.9	2.0	2.0	2.1	1.9	2.1
Other	3.5	4.0	4.3	5.1	6.9	5.4
Total	100	100	100	100	100	100

Household expenditure patterns for the median standard of living, Canada, 2008

Table 4:

Household expenditure patterns for the median standard of living, France, 1979

Expenditure category	< 60 % of median	60-75 % of median	75-100 % of median	100-150 % of median	> 150 % of median	All
Food	31	36	31	27	20	27
Clothing	9	8	9	9	9	9
Accommodation	15	14	14	13	13	14
Household equipment	8	7	8	8	9	8
Transport	12	12	13	15	16	14
Health	7	6	6	6	5	6
Communication	2	1	1	1	1	1
Leisure and culture	7	7	7	8	9	8
Education	0	0	0	0	0	0
Hotels and restaurants	3	3	4	5	6	5
Other	6	7	7	8	9	8
Total	100	100	100	100	100	100

Household expenditure patterns for the median standard of living, France, 1979

Table 5:

Household expenditure patterns for the median standard of living, France, 2006

Expenditure category	< 60 % of median	60-75 % of median	75-100 % of median	100-150 % of median	> 150 % of median	All
Food	20	20	19	18	13	18
Clothing	8	7	7	8	8	8
Accommodation	25	22	18	15	11	16
Household equipment	5	6	7	7	10	7
Transport	10	13	15	16	15	15
Health	4	4	4	3	4	4
Communication	5	4	4	4	2	4
Leisure and culture	8	8	10	11	15	11
Education	1	1	1	1	1	1
Hotels and restaurants	3	4	4	5	7	5
Other	12	12	12	12	13	12
Total	100	100	100	100	100	100

Household expenditure patterns for the median standard of living, France, 2006

32 The numerous changes to consumption patterns in French and Canadian households that have been briefly summarized here are directly linked to an increase in living standards over twenty-five years but also to changes in the life cycle of household members and transformations in our system of production (such as the reduced costs of telecommunications and electronic products, which encourage an increase in consumption). As their standard of living increases, households can devote fewer resources to food and clothing and spend more on satisfying new emerging needs such as commercialized leisure or transport. Increased life expectancy, in good health, has also led to changes in consumption over the life cycle. Couples whose children have flown the nest have more than twenty-five years of consumption ahead of them, without any children to care for in the home, at a time in their lives when mortgage repayments for their accommodation take up less of their budget (for most property-owning households). Observations made about changes over time can be explained by these three broad reasons and their combined effects (increased wealth over the time period, changes in household composition, and a diverse and extended range of goods and services available) but this is not the place to make such distinctions. Instead, we intend to look at how the situation of the middle classes has changed at different times, in comparison with other socio-economic classes, in order to determine whether there has been temporal and transversal convergence in each of the two countries studied.

Temporal and Transversal Convergence of Consumption: the Situation of the Middle Classes

33 The comparison between the budgetary coefficients of the socio-economic classes aims to define household needs from a transversal perspective. This is the first aspect that we will examine. We will then analyze temporal changes to budgetary coefficients over a long period in order to see if there has been a convergence of needs.

34 An initial general observation must be made: consumption patterns are very different for the two sub-groups identified within the middle classes in the two countries. The typical pattern of commodity consumption of the lower middle classes is similar to that of the disadvantaged classes, while the typical pattern of consumer spending of the upper-middle class is closer to that of the well-off classes. This empirical outcome provides an answer to the question raised by Bosc in the conclusion of his work on the middle classes. Depending on their level of commodity consumption, one section of the middle classes is drawn towards the bottom of the social hierarchy, while another section is more in line with the consumption patterns of the upper classes. This means that fragmentation can be observed at the center of the social structure, which no doubt influences social representations and fuels the feeling of downward mobility. Before we look at this aspect however, we will analyze in greater detail the different consumption categories used.

35 The share of the budget spent on food falls as the socio-economic status of a household increases, for every year of the study carried out in Canada and in France. The data also reveals a temporal convergence of budgetary coefficients for food between income categories in both countries. In the tables, we distinguished between two classes of poor household (less than 60% of the median and between 60% and 75% of the median). The share of expenses on food in poor households in Canada went from 28.9% to 19% in 1978 and from 25.6% to 17.2% in 2008, representing a considerable decrease. The amount spent on this category also diminished for rich households. The French data showed the same trends (Tables 4 and 5). The transversal and temporal changes are in the same direction for the food category, and the classic law highlighted by Ernst Engel in the nineteenth century is confirmed on both counts.

36 The analysis of spending patterns in both countries further indicates that temporal change is greater than transversal change. The reduction over time of the spending coefficient for food was more pronounced for households with a lower standard of living, both in Canada and France, freeing up resources to be spent on other consumption categories. This is an important point. Poorer households may have a certain room for maneuver to satisfy new emerging needs, but they have also experienced a significant increase in the budgetary coefficient for the accommodation category; a similar trend has been seen here in both countries. A decrease in the food coefficient was accompanied by greater pressure on accommodation expenses in their budgets. Overall, these households remain trapped in a world of need, their expenses dominated by basic requirements. This is different for the middle classes who, due to an increase in their real income levels, may enter the world of aspirations as defined above and increase their expectations in terms of commodity consumption. However, we can clearly see that the lower half of the middle class has less room for maneuver than the upper half.

37 The “accommodation” category deserves particular attention because its analysis reveals a divergence between temporal and transversal change. This observation has a number of consequences. In fact, transversally, the amount of the household budget spent on the “accommodation” category falls as the socio-economic situation of households improves, but this share of the budget has increased over time for all households. The amount spent on accommodation has increased even more significantly among the lowest income households, and the gap between the two middle class sub-groups is particularly pronounced. This means that the lower half of the middle class must spend a greater share of their resources on housing, putting a brake on the satisfaction of other discretionary needs (which are the object of powerful aspirations), particularly spending on leisure, as we shall see.

38 It is the inverse of what has happened for the “clothing” category, which also shows divergence in transversal and temporal change: transversally, the amount spent on clothing has increased with income, while overall spending has decreased over time (this is more obvious in Canada than in France). This is a fundamental point: looking at households with different income levels in any given year is not always a good predictor, without bias, of changes in consumption when real incomes rise over a longer time period.

39 The “transport,” “leisure,” and “protection” categories have grown in importance, both transversally and longitudinally, demonstrating the appearance of new needs to be satisfied, including for low-income households. As we can see in the tables, the budgetary coefficients for these three important consumption categories increase as income rises, but they have also increased for all socio-economic classes from one year to the next, leading to the conclusion that there has been both temporal and transversal convergence for these consumption categories.

40 The “transport” category is currently undergoing an important shift. The amount spent on transport is increasing for all households but particularly in the patterns of need for households with the highest standard of living, even sharing the top position with accommodation expenses for households in the highest income quintile (unpublished data). This is due to the multiple modes of transport that households may possess, such as two cars in the suburbs, and a third car for the young adults (the children of the household) in places with poor public transport. Another factor explains this top ranking for expenditure on transport in the wealthiest French and Canadian households: the purchase of more luxurious cars and higher running costs.

41 This is also the case for budgetary coefficients relating to the protection category (various different kinds of insurance) and for other expenses, which have increased more significantly among the wealthy. Here again, it is interesting to note similarities in observations between France and Canada.

42 Based on this analysis, we can see that transversal and temporal trends are moving in the same direction for eight budget categories in Canada and five budget categories in France. They have increased for five categories in Canada, i.e. transport, leisure, protection, education, and other for Canadian households, while they have fallen from both analytical perspectives for the “food” category. The observations for France are broadly similar: spending on transport, leisure, and “other expenses” (which includes jewelry, personal care, ceremonies, legal and financial services, etc.) shows an increase, both temporally and by living standards at any given time, while less money is spent on food, both transversally and longitudinally.

43 This analysis means that new needs, for transport (particularly cars), leisure, and protection, have rapidly been established, in parallel with the growing wealth of households and the rise in discretionary income. Expenses in these categories have increased over time for all socio-economic classes. We should, however, remember that the two types of increase do not necessarily change at the same rate among the different socio-economic classes. The richest households have increased their budgetary output (their relative expenses have grown) more quickly than poorer households, as we can see for transport, protection, and other expenses. Furthermore, we note that there are clear distinctions between the two middle class sub-groups. The consumption patterns of the sub-group known as the “lower” middle classes is similar to that of the disadvantaged classes, while the consumption of the “upper” half is closer to that of the well-off classes.

44 This analysis allows us to draw another important empirical conclusion, which supports observations already made by Gardes, Gaubert, and Langlois (2000) for Canada, with a different methodology and a shorter time frame. When a consumption category is dynamic over time (its budgetary coefficient increases rapidly as income increases between two time periods), it is generally more socially than temporally differentiated. The examples of transport and protection in Canadian budgets in Table 5 speak volumes on this point. This result means that the temporal change in incomes for the least well-off does not allow them to attain the same standing as the richest households. In other words, rich households increase their consumption of luxury goods and discretionary goods more quickly than households lower down the income scale. Our analysis, which distinguishes two sub-groups within the middle classes, supports this observation and also shows the persistence of different consumption behavior within the two middle class sub-groups. The conclusions drawn from the budgetary surveys analyzed support Bosc’s hypothesis on “the opposite attractions” that characterize the middle classes. This does not mean, however, that this clear differentiation in consumption behavior applies to social representations and aspirations. Nothing could be more uncertain, if we trust the predictions of the Tocqueville-inspired paradox mentioned earlier, which, we believe, can be identified specifically within the middle classes.

The Sense of Privation and Consumer Aspirations

45 Due to the lack of available data to analyze the social representations of Canadian households, we will focus on examining the results of CRÉDOC surveys that show which good and services French households consider themselves to be deprived of, and what they would like to be able to consume if their incomes were to rise at the turn of the year 2000 (the most recent survey available).

46 Three conclusions can be drawn from the CRÉDOC data. First of all, the goods and services that household members consider themselves to be deprived of (Part A of Table 6) are more frequently discretionary goods and services such as leisure, holidays, clothing, household equipment, or car accessories, rather than goods to meet basic needs (e.g. food, medical treatment, and accommodation). The proportion of households that mention items in the latter category corresponds roughly to the percentage of lower income households in French society. In addition, goods and services as the object of aspirations (those that households would like to consume more of if their incomes increased, Part B of Table 6) are always mentioned more frequently than the same goods and services as the cause of privation. This confirms that the aspirations stem from desire. Household members aspire to consume more goods and services, beyond any feelings of privation. Thirdly, the distance between feelings of privation and aspiration is more marked for discretionary items such as leisure and holidays, which indicates the opening up of aspirations once basic needs have been satisfied. Accommodation is exceptional in this respect. Feelings of privation relating to this object of consumption are not very strong (22% overall) but the aspiration to spend more on this consumption category is strikingly high, at one household in two. In fact, the work of INSEE and CRÉDOC shows that French households enjoy considerable comfort in the accommodation category, apart from the least well-off households, but a significant proportion of them would like to improve their homes if their incomes increased. Accommodation is invested with great importance for daily well-being and is the object of heightened aspirations, with the aim of improving it. We note that the aspiration to spend more on food, medical care, and telecommunications is higher among low-income households, which is a good indicator of the non-satisfaction of needs considered to be basic, as we predicted above. In other words, low income households are indeed trapped in the world of need.

Table 6:

Feelings of restriction and aspirations to spend by consumer category and standard of living, France, 2001

	Proportion of individuals who state that they have restricted spending in these categories (%)	Proportion of individuals who would increase spending in these categories if they had more money (%)
	A	B	B-A
Expenditure category	< 75 % of median living standard	75-100 % of median living standard	100-150 % of median living standard	> 150 % of median living standard	< 75 % of median living standard	75-100 % of median living standard	100-150 % of median living standard	> 150 % of median living standard	75-100 % of median living standard	100-150 % of median living standard
Holidays and leisure	62	56	49	31	72	80	81	83	+ 24	+ 32
Clothing	51	43	40	24	57	55	51	42	+ 12	+ 11
Household equipment	48	46	35	20	44	46	40	28	0	+ 5
Telephone	42	41	34	16	25	20	16	9	– 21	– 18
Car	41	35	33	17	39	30	35	28	– 5	+ 2
Accommodation	32	24	18	11	53	53	50	43	+ 29	+ 32
Food	28	19	17	10	43	35	28	16	+ 16	+ 11
Medical care	14	4	6	1	24	21	16	8	+ 17	+ 10

Feelings of restriction and aspirations to spend by consumer category and standard of living, France, 2001

47 But what is even more striking in the study of survey data for France, is the alignment of the aspirations of households situated in the lower half of the French middle classes with the aspirations of the upper half and the wealthy, which is not true at all for their actual expenditure, which is in fact more similar to the consumption patterns of the lower classes. The hypothesis of a greater extent of aspirations among the middle classes is therefore confirmed, at least for France. We may also consider that the same applies to Canada, and an article, now out of date, by Langlois (1982) observed a similar phenomenon at the end of the 1970s. Unfortunately, this study has not been updated.

Conclusion

48 The two groups that make up the middle classes in two developed countries, Canada and France, appear to be clearly differentiated, confirming once again the theory proposed that they are not a homogenous whole, hence the use of the plural form to describe them. The two sub-groups, which were empirically defined by whether their members were above or below the median for disposable income, have different spending patterns. The consumption patterns of the lower middle-class are closer to those found among the poorer classes and the consumption patterns of the upper middle-class are more like those of the well-off classes.

49 However all members of the middle class have developed aspirations to consume that are closer to those of the wealthiest classes, at least in France. As they are no longer trapped in the world of daily need to be satisfied in the first instance, its members would like to consume in greater quantity the discretionary goods and services to which they can legitimately aspire. But this aspiration cannot be fully satisfied, despite the continued rise in monetary resources, because the size and range of the available goods and services they aspire to is increasing even more quickly. This conclusion can be added to observations that have long been made by observers and analysts, from Montaigne to Alexis de Tocqueville and Émile Durkheim, as cited above, not to mention the numerous contemporary authors following up on the work of Richard Easterlin.

50 The hiatus between monetary resources and aspirations, which is typical of the center of the social structure, as defined by Maurice Halbwachs (with an emphasis on expenditure budgets and income), leads us to conclude that the malaise affecting the middle classes is not only due to exogenous factors such as the increase in unemployment, changes to the system of national production, and globalization. It is also the result of an endogenous social process specific to these middle classes, i.e. a notable extension of aspirations that goes well beyond the objective means they have of satisfying them.