SOCIOECONOMIC MOBILITY, EXPECTATIONS AND ATTITUDES TOWARDS INEQUALITY IN BRAZIL

Abstract After the end of the political and economic cycle that gave rise to the phenomenon that became known as the Brazilian new middle class, in this article we argue that this reading was not limited to verifying the increase in income and consume power of thousands of families, but also framed it within a broader narrative that has hindered the sociological investigation of the phenomenon. Thus, our first objective is to develop, with the support of empirical data from the National Household Sample Survey (PNAD-IBGE), an alternative interpretation of the changes observed in Brazilian society in recent years. Our second objective is to reflect, with the help of data from the Survey on the Middle Class (CESOP-UNICAMP-2008), and also from qualitative research previously prepared by the authors, on the possible impacts of these changes on the expectations of individuals and, thus, on their perceptions and attitudes towards the enormous inequalities still present in the country.

tion, the representation of these movements as the emergence of a new middle class seemed to have triumphed. This interpretation was endorsed not only by more academic works, but also by a large portion of the mass media 2 and by official discourse. 3 Already in that period, studies emerged problematizing the thesis of the new middle class; and it was in this context that the authors of the present text published an article (Scalon & Salata, 2012) containing data and reflections that throw this interpretation into doubt. Now, given the conclusion of the political and economic cycle that gave rise to that phenomenon, we are better placed to interpret the movements that occurred in Brazilian society during that period and discuss some of their possible consequences.
Middle class is an expression laden with meanings that, even if unintentionally, have become absorbed into the thesis of the Brazilian new middle class -despite the choice of the term being based only on statistical criteria. 648 sociol. antropol. | rio de janeiro, v.10.02: 647 -676 , may. -aug., 2020 socioeconomic mobility, expectations and attitudes towards inequality in brazil In 1883, for example, the abolitionist Joaquim Nabuco (quoted in Owensby, 1999, p. 3) complained that "the middle class, the driving force of the nations, cannot be found anywhere in Brazil" and sought to explain the country's 'backwardness' on this basis. Later, in 1900, the French sociologist Jacques Lambert (quoted in Owensby, 1994, p. 3) asserted that "it was above all in relation to the upward mobility of the middle classes that the new Brazil would distinguish itself from the old". This idea has a long history (Wahrman, 1995), stemming from the British experience in the nineteenth century when the rise of the middle class coincided with the political, economic and cultural pre-eminence of Britain and its empire. This era gave birth to the narrative of a middle class that was universal, modernizing and a symbol of progress. From this period also emerged the notion that a nation's prospects could be inferred from the size of its middle class.
In this reading, the middle class is the main bearer of the attitudes and values responsible for the modernization of society, such as individualism, meritocracy, forward thinking, and so on.
In interpreting the changes experienced by Brazilian society at the beginning of the twentieth century as the emergence and/or expansion of a middle class, this perspective, explicitly or not, alluded to the aforementioned narrative of the virtuous middle class. In fact, such an approach was not limited to observing the economic rise of a broad contingent of families: it also suggested that this phenomenon would be accompanied by changes to behaviours and attitudes. 4 Consequently, it was not only an interpretation of the recent changes experienced by Brazilian society, but also, based on this interpretation, it brought along a vague and imprecise projection of what their consequences would be (Souza & Lamounier, 2010).
As a result, then, of the baggage of meanings associated with the term, we believe that the approach based on the idea of a new middle class hindered comprehension of the changes recently occurred in Brazilian society and their potential repercussions. We argue that these changes may have had important impacts on the behaviour of individuals -in terms of their expectations, perceptions and attitudes towards inequalities -but that the latter could only be adequately explored through an alternative approach to the thesis of the new middle class.
Consequently, the present articles aims not only to update the analysis already made by ourselves a few years ago, but also approach questions that became more evident with the end of the cycle responsible for driving the economic and social upward mobility of this group. Our main focus here will be not so much on the operational and conceptual aspect of the definition of the middle class, already examined by us previously (Salata, 2016;Salata & Scalon, 2013Scalon & Salata, 2012), but on the establishment of an alter-article | andré salata and celi scalon native interpretation that allows us to advance the debate on the repercussions of the observed phenomenon, focusing principally on the expectations and perceptions of the individuals concerning the inequalities in the country.
Social expectations 5 are a fundamental element in comprehending inequalities and individual attitudes towards them, since they tend to adapt to the opportunities unequally distributed by a given society, facilitating it's reproduction and legitimization (MacLeod, 2009). At the same time, contexts of economic and social changes can alter the relationship between expectations and opportunities (Abeles, 1976). Hence, it is possible that the moment traversed by Brazilian society over recent years has had relevant social consequences in this area, which we shall look to discuss in the second part of this article.
The work is organized as follows: in the next section we present some data on the relationship between occupational classes, income and consume power in Brazil over recent years. This information will help us to establish, in the ensuing section, an alternative interpretation to the thesis of the Brazilian new middle class, approaching the possible repercussions of the phenomenon in terms of individual expectations and, in related fashion, their perceptions and attitudes towards inequalities. Finally, in the conclusion, we return to the main questions raised in this introduction, connecting them to the data and arguments discussed in the other sections.
In the first section we utilize data from the National Household Sample Survey (PNADs-IBGE), already widely known in the country. In the second part of the text, we make use of diverse sources of data, both qualitative and quantitative, previously collected and/or used by us in earlier texts. For the quantitative data, the main source is the "Survey on the New Middle Class" (CESOP-UNICAMP), conducted in 141 cities of Brazil in September 2008. 6 Meanwhile the qualitative data comes from two different sources. The first is the work of Salata (2016), where, in a study of class identities and perceptions, he conducted semi-structured interviews with individuals from the traditional middle class and the new middle class living in Rio de Janeiro. The second is the result of a focus group study (Scalon & Oliveira, 2012), carried out in 2008 with middle-and working-class youths on their opinions and values concerning inequality and social justice.
Finally, the reader should be advised from the outset that some of the arguments presented here, especially in the second part of the text, are speculative in kind. Despite seeking support in empirical data of a quantitative and qualitative character whenever possible, the article's main objective is to construct a sociological reflection via the data, not to formalize a model or provide an empirical test of hypotheses. Our analysis will focus principally on the period between 2002 and 2014, during which the phenomenon analysed here emerged, and when much of the debate on the same unfolded. 7 650 sociol. antropol. | rio de janeiro, v.10.02: 647 -676 , may. -aug., 2020 socioeconomic mobility, expectations and attitudes towards inequality in brazil From the new middle claSS to the Socioeconomic mobility oF the working claSS The research carried out by Neri (2008), which initiated the reading of that phenomenon as the emergence of a new middle class in Brazil, defined classes as income intervals, 8 and observed that the intermediary intervals -especially the then famous 'C' income interval -had expanded, corresponding to more than half of the Brazilian population at the time. In Neri's reading, this meant that thousands of families had left the lower income bands, defined as a level of poverty and vulnerability, and joined the middle class.
In this article we use alternative income intervals to those employed by Neri, with five layers rather than just four. Figure 1 shows the distribution of Brazilian households between 2002 and 2014 among these five bands of total household per capita total income: up to R$250; from R$250 to R$500; from R$500 to R$750; from R$750 to R$1,500; and over R$1,500. These cut-off values were defined on the basis that, in 2014, they corresponded to approximately the 10th, 30th, 50th and 80th percentiles. Like the intervals established by Neri  article | andré salata and celi scalon As expected, we can identify a clear trend towards reduction of the two lower bands and expansion of the others, characterizing a process of upward economic mobility of households in the country. In 2002, 56% of households (26.8 million) had a per capita income of up to R$500; in 2014 this figure had fallen to 30.9% (19.6 million). On the other hand, the percentage of households with more than R$500 per capita leapt from 43.5% to 69% in the period, or from 20.6 to 43.8 million.
As Figure 1 also reveals, a large proportion of this increase took place not in the higher income band, but in the intermediary bands, closer to the median of the distribution. This indicates that many low-income households began to generate incomes closer to the intermediary level. For this reason, Neri (2008) interpreted this movement as an expansion of the middle classdefined by him in statistical terms.
In the sociological literature, however, the middle class tends to be identified by the market and labour situations involving professionals, administrators and white collar workers, whose main assets are their qualifications (Goldthorpe, 2000;Mills, 1951;Ehrenreich & Ehrenreich, 1978;Wright, 1976). 9 In other words, they concern the aggregates of individuals/occupations that share class situations located somewhere between the big employers and the manual workers -that is, between those whose main asset is ownership and those whose sole or main resource is their labour force -like the professional workers, managers and administrators. 10 Meanwhile, the working class is taken to encompass not just skilled manual workers but also unskilled workers, a large proportion of the self-employed and non-manual routine workers, as well as rural workers. In effect, these are the categories of individuals whose market opportunities depend primarily on their labour force. 11 To operationalize the concept of classes, in this article we utilize the schema of nine categories employed by Scalon (1999) and originally developed by Nelson do Valle Silva. We define the class situation of the families through the information provided on the key persons from the households -that is, the person responsible for the household unit, or considered as such by other members. 12 Figure 2 shows the percentage of households in each of the income bands, according to the occupational category of the household heads, for the years from 2002 to 2014. 652 sociol. antropol. | rio de janeiro, v.10.02: 647 -676 , may. -aug., 2020 socioeconomic mobility, expectations and attitudes towards inequality in brazil   The income inequality between some of the socio-occupational categories becomes clear. It is possible to perceive that we have a first group whose income is concentrated in the higher bands, formed by Owner-Employers, Professionals, and Administrators and Managers. The other group is formed by categories with income concentrated more in the lower bands with greater difficulty in attaining the higher income limits: Self-Employed Workers, Non-Manual Routine Employees, Skilled Manual Workers, Unskilled Manual Workers and Rural Workers. These results manifest the income gap -but also translate a distance in terms of other dimensions of social stratification -which separates the (upper) middle class from the working class in the country. 13 What most interests us here, though, is the variation in the income composition between the years for each category. 14 The most striking results concern the categories of workers (rural, unskilled, skilled, routine non-manual and also self-employed), since among these we find a sharp reduction in the proportion of households in the two lowest bands -up to R$500 per capita -and a correlated expansion of households in the other bands.
Among the households whose heads were Unskilled Manual Workers, for example, 68.3% were in the two lower bands in 2002, 28,8% were in the intermediary bands, and 2.8% in the upper band. By 2014, the percentage of these households in the two lowest bands had fallen to 37.6%, the proportion in the two intermediary bands raised to 53.3%, and the percentage in the upper band had jumped to 8.9%. In effect, we find a huge contingent of working-class house-article | andré salata and celi scalon holds that have left behind the lower income strata in recent years and entered the higher income levels.
In order to better visualize the effects of these movements in the income composition of the working class, Figure 3 displays aggregated information on households whose heads were non-manual routine, self-employed, skilled manual or unskilled manual worker, 15 which will be combined here in the category of "urban workers." 16 Once again, we utilize the five household per capita income bands cited previously. socioeconomic mobility, expectations and attitudes towards inequality in brazil As we can see, there has been a significant change over recent years in the participation in these income bands among working class families. We can highlight the drop in the percentage of families in the lower band (up to R$250) from 25.3% to 7.6% and a correlate increase in the number of households in the fourth income band (between R$750 and R$1,500), which rose from 18.3% to 33.7%. It is important to emphasize that there was also an increase in the participation in the highest band (over R$1,500), rising from 8.1% to 16.8% between 2002 and 2014. Thus, we can affirm that this class ceased to be mostly composed by families in the lower income bands, since a large contingent of these entered a higher income stratum.
But the increase in income was not the only factor involved in the upward mobility of this category. Accompanying this increase in income -or, in part, as a result of this increase, combined with the expansion of credit and other factorsthese working class families were also able to access a series of consumer goods and services during this period that had previously been more restricted to the middle and upper classes. In fact, beyond rising income levels, it was the access to these consumer goods and services that most attracted the attention of the public debate in Brazil on the changes that society was experiencing. Figure 4 illustrates this movement (next page).
Though limited, the following data illustrate that thousands of working class Brazilian families, over the last decade, were able to renew or purchase their first mobile phones, refrigerators, televisions, washing machines, computers, cars and so on, to the extent that the consumption of some of these goods ended up symbolizing the upward economic mobility of this stratum of the population. Numerous reports appeared in the mass media about these new consumers originating from the working classes. Frequently too, these reports were accompanied by images of packed stores and supermarkets, showing the emergence of this new consumer, who also began to circulate in spaces previously confined to the (upper) middle class, like shopping malls and airports (Bomeny, 2011).
Besides access to consumer goods, Figure 4 also shows the percentage of households with a density (residents per bedroom) higher than two, which serves as an indicator of housing conditions. As can be seen, this percentage falls from 28.2 to 15.9, denoting a very perceptible improvement during the period. The reduction in unemployment, the rise in income, and, finally, the increased access to credit, combined with public policies introduced for this purpose, enabled many families from this category to acquire properties and/or enlarge their existing homes, improving their housing conditions. It is important to stress that the upward socioeconomic mobility discussed here took place not only in terms of income and consumption, but also in relation to schooling, especially among the younger population. As earlier studies have already observed (Salata, 2018b;Picanço, 2015;Mont'Alvão, 2015), along with the changes observed in the income pattern of these families, there was also a notable increase in the presence of their children in universities. The changes observed are notable. While, in 2002, 44% of working-class children had completed secondary education, in 2014 this percentage had risen to 63%. Meanwhile, the percentage of those who had entered higher education jumped from 12% to 27% over the same period. Thus, and despite the enormous inequality in educational opportunities still present in the country, we can note a substantial increase in the completion rates of secondary education, as well as a perceptible reduction in the barriers impeding access to higher education. If we consider only those children who had completed secondary education, the percentage who entered higher education rose from 27% to 42% among these families during the period.
The data presented in this section endorse our reading, already published previously, that instead of a growth of the middle class, what happened was that working class families achieved slightly better economic conditions, with higher incomes, consume power and schooling for their children. What is notable during this period, therefore, is a substantial socioeconomic improvement for working class families. In this sense, we could refer to a process of socioeconomic mobility for part of this class and a slight approximation to the middle class. However, a more complete understanding of this phenomenon and its possible repercussions must look beyond a narrowly economic dimension. and attitudes towards inequalities (Lipset & Bendix, 1991;Goldthorpe et al., 1971).
Despite some important divergences between the various approaches to this topic, one point shared by many of them is that an individual's perception of his situation is always relative, and that his expectation should be taken as a reference point from which the objective situation can be evaluated (Merton, 1970;Runciman, 1966;Gur, 1971;Hirschman, 1975). For this reason, and apparently paradoxically, it may be that in contexts of objective improvement, greater individual dissatisfaction is observed, or, conversely, that in situations of great inequality there is much less disquiet.
Setting out from this logic, Santos (2006), seeking to explain the relative stability of a social order as unequal as Brazil's, argued that the very high level of disparities present in the country limits the expectations of those in the lower strata. Given the enormous distance separating them from the higher strata, entering the latter rarely forms part of their future aims. Because their expectations are so low, therefore, the level of dissatisfaction of the lower strata of Brazilian society is, Santos proposes, far less than what an uninformed investigator might expect.
The world of the (upper) middle classes would appear unattainable and, therefore, beyond their horizon of expectations -focused simply on avoiding the most precarious and insecure socioeconomic situations. The comparison with the middle class is not seen to nourish feelings of relative deprivation among workers, since, due to the enormous gulf observed, the middle class does not serve as the group of reference for their judgements (Cardoso, 2010). socioeconomic mobility, expectations and attitudes towards inequality in brazil income, lessening their gap to the middle class, but also to consumer goods, spaces and services once mostly reserved to the wealthier classes. Once again, this is a fact that should not be observed just from an economic angle, since some of its most significant consequences reside in the social sphere (Ridgeway, 2013), such as the increased presence of working classes individuals in shopping malls, hotels, airports and the like. Furthermore, as we have seen, their presence has also expanded in the universities, which may indeed boost their chances for upward social mobility.
As a consequence of the changes previously observed, therefore, the mid- Likewise, while only 8% of interviewees saw themselves in positions 7-8, 42% believed that their children would reach them. There seems to be, therefore, a fairly high expectation in relation to the future of these families, which also became clear when they were asked whether they believed that household income would be lower, higher or equal to the present. As can be seen in Figure   7, almost 60% of interviewees believed that household income would be higher the following year. 20 From both the intragenerational and intergenerational viewpoints, then, expectations were high at that moment in time. The data allows us to ascertain that positions 7 and 8 on the scale roughly seem to signify entrance into the middle class. Among interviewees who perceived themselves to be occupying positions 7-8, the majority (46%) identified themselves as middle class, and 9% as upper middle class. Meanwhile, among those who see themselves in positions 4-6, most identified (when stimulated) as working class, comprising 28% of respondents. It would be feasible to suppose, therefore, that these interviewees expected their children to join the middle class.
As another potential illustration of this hypothesis, we provide an excerpt from a dialogue with an interviewee 21 in a study undertaken in Rio de Janeiro Despite not recognizing himself as middle class (Salata & Scalon, 2015), there is a perception of improvement in recent years and, also, an approximation with the middle class, which, despite still appearing very difficult, now seems to be within the realm of the possible. On this point, it is essential to emphasize that in the interviewees' responses there is a clear signalling of the differences that mark the place of the middle class, which the interviewees from the so-called 'new middle class' recognized themselves as not occupying (Salata & Scalon, 2015). Even so, there was a prospect of improved life conditions at that moment, and joining the middle class already appeared as a future possibility, as can be seen in the excerpt below as well: 23 The observation that the interviewees from the C income band did not These higher expectations might be imagined to have led to a more critical stance in relation to a social order that is still extremely unequal, stimulating the desire to deepen the observed changes. Here again we refer to the already well-known thesis of growing expectations, which end up exceeding objective opportunities, leading to a feeling of dissatisfaction and increasing demands (Abeles, 1976 There is, however, an open field of possibilities in relation to the repercussions of this phenomenon whose crucial aspect, we believe, is whether people's failures and successes in matching the supposed new expectations, will be -or already are being -read as the result of processes understood as just and, therefore, as legitimate, or not. 24 As already demonstrated by Scalon (2004aScalon ( , 2007, inequalities in Brazil are mostly perceived to result from differentials of endeavour, education and intelligence, which ends up legitimizing them: In the case of Brazil […] the acceptance of inequality seems to be related to higher levels of adaptation to the forms and criteria of stratification.
[…] When upward social mobility and obtaining a better income are seen to result from individual merits and characteristics, inequality is more widely accepted (Scalon, 2004b, p. 31 Fortunately, the questionnaire used in the "Survey on the Middle Class" (CESOP-UNICAMP) contained some questions on this theme, whose responses -again just for the household headed by urban workers -are shown in Figures   8 and 9. 662 sociol. antropol. | rio de janeiro, v.10.02: 647 -676 , may. -aug., 2020 socioeconomic mobility, expectations and attitudes towards inequality in brazil  to be remembered that the typical image of a poor family refers to extremely precarious conditions -with the lack of food heavily marked -with which those interviewees with the profile of the new middle class did not identify (Salata, 2016). Hence, the perception of a society in which poor people are unlikely to overcome their situation does not contradict the representation of a social order where opportunities are seen to be relatively open to hard-working and talented individuals, whatever their social origin.
As expected, this representation aligns with the image of a society where education, intelligence and endeavour are understood as very important criteria of stratification, as can be ascertained in Figure 9.  (Goldthorpe & Jackson, 2008) seem to be attributed more importance than social origin as conditioning life chances.

663
article | andré salata and celi scalon  sociol. antropol. | rio de janeiro, v.10.02: 647 -676 , may. -aug., 2020 socioeconomic mobility, expectations and attitudes towards inequality in brazil True enough, the fact that 88.5% of respondents stated that knowing the right people is very important could raise some doubts on this point. As there is a strong positive correlation between this one and the first three items, however, we suspect that, for the interviewees, "knowing the right people" would also be, to some extent, an achievement dependent on individual skill and effort. Nevertheless, the data do not allow us to pursue this idea in more depth.
Focus group studies with young people carried out in 2008 (Scalon & Oliveira, 2012;Scalon & Oliveira, 2019), for example, corroborate the emphasis on individual merit when it comes to assessing the chances of upward mobility. Turning to the results from focus groups held with working class youths between the ages of 16 and 19, there is clear acceptance of the idea that upward social mobility is connected to merit via education and endeavour: "Study is the basis of everything. For us to have something in the future, studying is essential." "Study helps a lot, but effort… You can do a thousand courses, but you can't sit there waiting." "It's about having a strong will." (Scalon & Oliveira, 2012, p.412) Consequently, the data suggests the predominance among working class interviewees of the perception of a relatively open society where individuals, through their own efforts and talent, and via education, are capable of overcoming the barriers of class and social background imposed on them.
We should recall that the promise and hope of individual upward mobility has a long history among the Brazilian working class (Cardoso, 2010). Contributing to this situation are, among other factors, the fragmentation of urban work situations, the enormous contingent of unemployed workers or workers employed in the informal sector of the economy, the perception of the upward mobility of a large proportion of the people who, in the past century, migrated from the rural world to the city, as well as a frequently authoritarian political environment, which disallowed the expression and competition of alternative projects. 26 The set of evidence gathered here, therefore, is the result of a long-term process.
As we have highlighted, though, there is a possible new factor in this scenario: rising expectations and the possibility of upward mobility -at least intergenerational -to the middle class has become seen as an attainable aim for the country's urban working class. Some families will succeed in these expectations, but many others will not, which may lead to a more critical stance in relation to a social order as unequal as Brazil's. On the other hand, as we have seen, these kinds of inequalities tend to be legitimized by the perception that they result from merit and hence it may be that the success and failure to match the -pos- that there has been no growth in the middle class in the sociological sense of the term, it is impossible to ignore the changes that have taken place during the period, which may have a significant impact on Brazilian society. However, given that over recent years there has been a big change in the context that gave rise to the so-called new middle class, now marked by economic stagnation, rising unemployment, growing inequalities and so on (Neri, 2019;Scalon et al., forthcoming), we need to ask ourselves about the interest in a phenomenon that appears to have succumbed. 27 Figure 10 shows the evolution of the composition by bands of household per capita income, as well as its mean and median, for the period from 2015 to 2018. What we see is a situation of stabilization, but not yet one of reversal, since the participation of each of the income bands has altered only slightly, likewise the mean and median values. It is worth underlining that there were indeed changes, as expected, within the established income bands, in the sense of a fall in income among the poorer sectors and an increase among the wealthier families. 28 In any event, the data show us that it would be too hasty simply to decree the end of the phenomenon that became known as the Brazilian new media class.  sociol. antropol. | rio de janeiro, v.10.02: 647 -676 , may. -aug., 2020 socioeconomic mobility, expectations and attitudes towards inequality in brazil Just as its observation was followed by an immense commotion in the public debate, generating a dissemination of meanings that went beyond empirical observations, as we argued in the introduction to this article, the economic and political shift over the last few years has led to the sudden virtual disappearance of any debate on the theme. We believe, however, that it is still necessary to deepen our sociological knowledge of the phenomenon and its consequences.
In this article, we have argued that the expectations of the working class may well have risen, approximating the middle class. What was once perceived as beyond their possibilities, interpreted as distant, has perhaps become, given the process of upward mobility evinced over the last decade, part of a desired future taken to be possible, attainable and realizable. If true, we are faced with the so-called 'tunnel effect' (Hirschman, 1975) when small objective improvements lead individual and collective expectations to rise. In this scenario, the horizon of desires of the working class, previously limited (Santos, 2006), seems to have expanded in recent years.
On one hand, this movement could reinforce individual strategies of upward social mobility, thereby legitimizing a very unequal and unjust social order. On the other hand, frustrated expectations might end up lending support to a more critical attitude in relation to inequalities. In this sense, the interpretative angle offered in our analysis places in perspective two alternative trajectories that basically depend on whether the perception of improvement in living standards achieved in recent years, along with the success or failure in achieving the aspired levels, will be comprehended as individual or collective.
Will the dissatisfactions and/or frustrations in failing to match expectations be attributed solely to singular, individual factors or to a wider social dynamic?
For the moment, however, we are not in a position to answer this question, which without doubt merits greater attention from empirical research. 29 For the purposes of this work, we content ourselves with the suggestion that one possible, or probable, consequence of the growth in income, consumer power and access to university among working class families, is the rise in their expectations, such that joining the middle class is now seen as a possibility. In this sense, it is probable, though not certain, that the movements observed in recent years have had an impact on the attitudes of individuals. However, their real impacts still need to be investigated.
What we observe, therefore, in analysing the changes experienced by Brazilian society over recent years is not the emergence of a new middle class, but a working class in a slightly better socioeconomic situation -despite the enormous inequalities still present -whose expectations may have risen closer to the middle class, at least in terms of their future aspirations. This, without doubt, seems a less enthusiastic and positive reading, since it does not evoke the idea of an expanding middle class, always associated, as we have 668 sociol. antropol. | rio de janeiro, v.10.02: 647 -676 , may. -aug., 2020 socioeconomic mobility, expectations and attitudes towards inequality in brazil noteS 1 For more recent analyses of the theme, see, among others, Cardoso & Préteceille (2017) and Kopper & Damo (2018).  (Thurow, 1987;Curry-Stevens, 2008;Ravallion, 2010), this approach takes the intermediary sectors of income distribution as a proxy for the middle class, and presumes that countries with a broad middle class have certain advantages such as greater political stability, less accentuated social conf licts, better quality democracy and so on. 5 Different from aspirations, which concern the desire to obtain a determined result, or to achieve a particular aim, expectations concern the estimates that individuals make of their chances of attaining determined objectives (Ma-cBrayne, 1987). 6 Its sample is representation of the Brazilian population aged 16 or over and contains 2002 cases. The selection criteria were probabilistic in the first two stages (cities and geographic clusters) and, for the selection of interviewees in the third stage, gender, age, schooling and sector of activity quotas were used. The same data was utilized by Souza and Lamounier (2010) and Salata (2016).
7 The year 2014 marked an economic turning point where the favourable context that gave rise to the so-called new middle class began to change rapidly.
9 For a review of the diverse sociological approaches to and definitions of the middle class, see Salata (2016).
10 In the field of the sociolog y of social stratification, the approach that we adopt here has been defined as CARs, an abbreviation for capitals, assets and resources (Savage, Warde & Devine, 2005). This is opposed to the now classic employment aggregate approach (Crompton, 1998) 12 For a debate on this type of procedure, its virtues and limitations, see Goldthorpe (1984).
13 It is important to note that in PNAD, as a sample survey, large proprietors are under-represented. For this reason, the Proprietor-Employer category is primarily composed of small and medium-sized employers. In other words, it is closer to the notion of the middle class than the upper class/rich. 14 In previous works, we have already shown that there is no significant variation in the socio-occupational structure during the period, at least when measured by the NVS schema (Salata, 2018a) or EGP schema (Scalon & Salata, 2012). Using a more detailed classification, however, Cardoso and Préteceille (2017) identify some relevant alterations.
15 As the debate on the new middle class primarily revolves around the urban universe, we have opted to exclude rural workers from these graphs.  sociol. antropol. | rio de janeiro, v.10.02: 647 -676 , may. -aug., 2020 socioeconomic mobility, expectations and attitudes towards inequality in brazil 17 The data presented has limitations that should be taken into consideration. In order to gain access to the sociooccupational class of the family head of these youths, we had to work only with individuals who were in the position of sons and daughters within the household. Consequently, the data must be analysed with caution.
19 As for the PNADs, in the database of CESOP's survey on the middle class, these households represent approximately 68% of cases.
20 Despite the fact that the data used was collected almost eight years ago, which could appear to make it out of date, it is a positive factor that the survey's interviews were conducted in 2008, precisely when the obser ved upward economic movement was occurring. 24 See Crosby (1976). Figures 8 and 9 would be very close were we working with the entire sample rather than just those we classify here as working class.
27 In fact, this does not seem to be a reality specific to Brazil, also being observed in other countries like the United States and the United Kingdom (Jackson & Grusky, 2018). attitudes.