Money whitens... but does it?


(…but does it?)

Foreword to this translation

Brazilians find it difficult to explain to non-Brazilians how racial relations function in Brazil. And especially difficult to explain it to North Americans. It is easy, in such case, to fall into one of two opposite temptations: to either try to normalise Brazilian racial relations as very similar to the ones in the United States, or to fancy them as very different, to the point of maintaining that there is no racism in Brazil. But in reality, even the terms used for racial categories in Brazil aren’t directly translateable into American English (“preto” is not “Black”, “pardo” is not “Brown”, etc), and yet Brazilian society is still extremely racist, though such racism manifests itself in very different ways than in the United States.

So, translating a paper originally written in Portuguese into English is a difficult endeavour. When Brazilian researchers say that they are, for statistical purposes, considering “pretos” and “pardos” as one single category, “negros”, how do we explain that to people to whom this doesn’t seem to imply any different categories at all? Of course, this would require a whole different paper, or perhaps more, which is not the intention here. A good read about Brazilian racial categories, written by a non-Brazilian would be, for instance, Edward Telles Race in Another America : the significance of skin color in Brazil, which is a whole book, and yet still can’t be considered complete in explaining how Brazilians reason about “race”. Brazilian views of the same subject might include Petrucelli’s A Cor Denominada, and Simon Schwarzman’s Fora de foco: diversidade e identidades étnicas no Brasil. But the non-Brazilian reader must be aware of some of the difficulties in translating a work like this without misleading those who read into imagining something that does not correspond to reality.

The text below translates “branco” as “White”. As in the United States, Whites in Brazil surely enjoy “White privilege”, and are exempt from racial prejudice. But, as opposed to the United States, they are not a clear majority of the population; as of 2010 they are a plurality, and, perhaps, a bare majority. Most of Brazilian “Whites”, moreover, would probably not be considered “Whites” in the United States (this, combined with Brazilians’ strong tendency to self-deprecation, results in myths such “the Brazilian categories for race are wrong”, “Brazilian ‘Whites’ are not actually White”, etc.), even before they attempted to speak, denouncing their Latino accents.

And this is the word that is easy to translate.

Brazil was not settled by Portuguese families with a protestant ethics regarding sex. It was settled by Portugues male adventurers, who, with few “White” women in the colony, tended to miscigenate, first with Amerindian women, later with enslaved African women. Of course, no “one drop rule” could arise in such conditions.

This did not result in a uniform mixed-race population, however: due to the phenomenon of “assorted mating”, it resulted in a population that is light-skinned in the upper layers, and darker in the lower layers, even though probably having not so much different genotypes when it comes to other hereditary characteristics that are not visibly linked to “race”. To put it in a simplistic way, “White” Brazilians have “White” skins (and hair, and nose), but they may well have the guts of an African. The converse is equally true.

The intense miscigenation resulted in a population where the intermediate phenotypes are very widespread; people whose looks are not “White” enough, but are also obviously not purely “African” are either a quite huge minority, or perhaps a plurality. As opposed to the United States, and their one-drop rule, where those people would be considered “Black”, Brazilians tend to classify them as mixed-racial; there are several words that denote this, and no one is a “good” word for all the involved. For the census, those people are “pardos” – a word that literally translates into “brown”. Hitler’s “Brown Shirts”, for instance, are called “camisas pardas” in Portuguese. I have avoided translating “pardo” in this paper, rather keeping it in Portuguese, and between quotes, because I fear “Brown” has different conotations for the English speaker. In Brazil, it is mostly a term for those who have a phenotype that is partially “White” and partially “Black”. But it is also, quite confusingly, used for people of predominantly Amerindian phenotypes.

Many other words are used to refer to mixed-racial people; two of them are used in this text, “mulato” (which means a mix between White and Black, and explicitly excludes Amerindians, but unhappily is a complicated word, either because of the sexist implicatons of its feminine form, or because of its infortunate etymology, that relates, or is believed to relate, to the offspring of the interbreeding of mares and jackasses), and “moreno”, which alone requires a whole paper of itself. I have kept both words in the original Portuguese. “Moreno” is a troublesome word; unlike “preto”, “negro”, “pardo”, “mulato”, and even “branco”, however, that’s not because of any negative connotations. On the contrary, “moreno” is something that most Brazilians would apply, or like to apply, to themselves. More strictly, it means people with dark hair, or with an olive complexion; but it was so much used in sarcastic or euphemistic ways for non-Whites, that it became popular among them – in part, perhaps, because it is used by very White people when they want to emphasise their sun tanned skin.

Albeit all miscegenation, there remains a significant minority of Brazilians that are, or look, definitely more “Black” than “White”. Those are called “pretos” (which literally translates as “black”; a black cat is a “gato preto” in Portuguese) by the census. This is another problematic word. During the slave regime, it was often used to denote “well-behaved” slaves or ex-slaves, as opposed to more rebellious “negros”. It is not difficult to understand why such word isn’t appealing to many Black Brazilians, who would prefer “negro”. And indeed, such word is quite widespread, probably more than “preto”. But being a word prefered by those in the Black movement, it is often intended to encompass not only the minority we are discussing, but also the whole “non-White” population. This is far from unanimous, though, and there are indications that to most people “preto” and “negro” are synonimous, or at most differ in cerimonial ways, the later being more polite.

To complicate things, if the boundary between “branco” and “pardo” is somewhat porous, the boundary between “pardo” and “preto” is much more so. A “preto” is a person who looks definitely more Black than White… in the opinion of people who haven’t been acquainted to the mass arrival of African individuals for more than a century and half. And while there is a clear correlation between the Brazilian elite – and upper middle class – and “White” skin colour, the Brazilian poor may be of any colour, “brancos”, “pardos” or “pretos”, and, so, there is no significant sociological difference between the two later categories. For this reason, I have also decided to avoid translating “preto”. On the other hand, I have translated “negro” as “Black”, not because it perfectly corresponds to the American category of Black people, but because in all the instances in which “negro” is used in this text, and in those it quotes, it encompasses both “pretos” and “pardos”, which is to say, the vast majority of the “non-White” population.

Two other racial terms need explanation, though neither is of actual importance to the subject to be discussed here. One is “amarelo”, which literally means “yellow” (the colour of the diamond in the Brazilian flag, for instance, is “amarelo”) and, in the context of racial issues, is applied to people of asian descent. An easy option would perhaps be to translate it as “Asian”. The word I have chosen, however, is a bit more provocative; I have used the literal translation into “Yellow”. This may be seen as a reminder that Brazilians have a different relation to racial words, compared to Americans. While practically anyone may take exception with words like “preto” or “mulato”, there is no Brazilian equivalent to the absolute taboo that surrounds the word “nigger” – Brazilian racism is rather a matter of discrimination and social exclusion than one of name calling. And so, being referred as “amarelos” is of little concern for them; indeed, on the contrary, it is possible that being referred as “asiáticos” would be worse – nationality plays a bigger role in Brazilian national ideologies than in the United States. Or it can be seen to make the point that those people do not face racial prejudice in Brazil, on the contrary of the United States; anthropologically, they are very much like “honorary Whites”; sociologically, they are similar to Whites, being predominantly members of the upper middle class. The last term to explain is “indígena”, which I have translated into “Amerindian”. It should be noted that, to most Brazilians, this is a cultural term, not a racial one; assimilated “native Brazilians” would consider themselves rather “pardo” than “indígena”.

Finally, I think it is necessary to admit that I have had some trouble translating the Portuguese word “escolaridade”. It means the time one has spent in school, or perhaps rather how far one has been able to advance in school grades. I have used either “educational level” or “schooling”, and at times “years of school”, all of which I deem somewhat imperfect.


(…but does it?)

“Money whitens” is a common statement in Brazil. According to popular wisdom, some elements taken in account in racial evalutions by Brazilians would be wealth and income, or some other factors more or less directly related – prestige, schooling, ostensive displays of wealth such as automobiles, dresses, etc(1).

Any atentive observer of Brazilian society knows that this is quite an exaggeration. First, because darker individual are rarely considered White, whatever their personal wealth, so that such statement only makes sense about those who are considered “pardos” in IBGE’s censuses’ terminology. Second, because, of course, it should be referred to some parameter to which the “whiteness” of individuals couls be compared; for instance, if it is said that individuals who become wealthier begin to be considered White by others, then such “whitening” could be compared to the opinion of such individuals concerning themselves, and conversely. In the way it is usually made, though, the assertion would need an external standard, a neutral vision, capable of establishing an “objective” racial criterium. And evidently neither such neutral vision nor such objective criterium exists, or even can exist: all human individual express, in their opinions, the way they were socialised, and human “races” are not biological realities, that could be be measured in an neutral way.

It seems clear, then, that the supposed “whitening” of individuals, as a function of money, can only exist in terms of an external convention, contrasted to the intimate opinion of observers. In other words, it is not that an individual becomes White because he or she became rich, but that the opinion that such individual is Black, or “pardo”, can no longer be expressed, as a result of such enrichment – the individual does not become White, but starts to be treated as such, in contradiction with physical appearance.

A broader way to explain the problem would be the following: in Brazilian society, “race” is a taboo subject, and should not be mentioned. Since racism is directed against Blacks, not against Whites, it is of more importance not to mention the race of others when such is Black (and, secondarily, not to mention one own’s race when that is White); and, as Brazilian society values personal wealth, it is more important to respect the taboo concerning the rich than concerning the poor.

However, another myth about racial relations in Brazil has that it is the individual’s own opinion that is most affected by social ascension, which is to say, more than being viewed by others as lighter than he or she “really is”, it is the individual that begins to see him or herself as such(2). A famous anecdote attributed to the painter Santa Rosa is commonly quoted in support of such idea (comforting a friend, a beginner artist, who felt sidelined due to racial prejudice, Santa Rosa would have said “I too have been Black, so I know how those things are)(3). If this is true, then the hypothesis we raised above, concerning the public expression of racial classifications as taboo, cannot hold. It would be important, then, that we could contrast self-classification and alter-classification of individuals diversely placed in society concerning income, wealth, and prestige, to verify if there is indeed some “whitening” effect related to income, and if such “whitening” occurs more, or less, in self-classification or in alter-classification.

The text we have quoted, by Rafael Guerreiro Osório, presents three researchs that compare self-classification and alter-classification. Below, the data as rearranged by Osório:

Alter-attributed skin colour vs Self-attributed skin colour – São Paulo – 1986 (in %)
Skin colour, self-attributed
White “morena” “mulata” “preta” Other “oriental” Total
Skin colour, alter-attributed White 56.2 9.3 3.5 0.9 0.9 0.5 71.1
“preta” 0.9 0.2 1.2 2.3 0.2 0.0 4.7
Yellow 0 0 0 0 0 2.4 2.4
“parda” 2.4 6.6 8.2 3.3 1.0 0 21.6
Total 59.5 16.1 12.9 6.4 2.1 3.0 100

Source: Idesp, Iuperj, Gallup, Research “As eleições de 1986 em São Paulo”. In: Valle Silva (1999a, p. 119). Rearranged data.

Alter-attributed skin colour vs Self-attributed skin colour – Brazil – 1995 (in %)
Skin colour, self-attributed
White “parda” “preta” Other(4) Total
Skin colour, alter-attributed White 44 5 0 3 52
“parda”/”mulata” 6 20 5 4 35
“preta” 0 3 7 1 11
Other(5) 0 1 0 1 2
Total 50 29 12 9 100
Source: Datafolha, 1995. In: Turra e Venturi (1995, p. 89).

Alter-attributed skin colour vs Self-attributed skin colour – Brazil – 1996 (in %)
Skin colour, self-attributed
White “parda” “preta” Yellow Amerindian Total
Skin colour, alter-attributed White 39.1 4.9 0 0.1 0 44.1
“parda” 3.5 46.2 0.9 0.1 0 50.6
“preta” 0 1.8 3.1 0 … 4.9
Yellow 0 0 … 0.3 … 0.3
Amerindian … 0 … … 0 0
Total 42.7 52.9 4.0 0.4 0 100
Source: Bemfam, Pesquisa Nacional de Demografia e Saúde (National Research on Demography and Health), 1996, in microdata. Elaboration: Disoc/Ipea.

In all of them, it is notable that self-atribution of White colour is always lower than alter-attribution. In the first two, moreover, self-atribution of Black colour is higher than alter-attribution. So, the data seem to contradict immediately the notion that the opinon of the individual in question is more affected by social ascension than the opinion of others. If money whitens, it must then whiten more from the point of view of external observers than from the point of view of the individual him or herself. And indeed, Osório goes on, from this data, to support IBGE’s methodology, arguing that self-classification is a better method than alter-classification.

There would be a way to circumvent this objection, though, which would be to argue that no only money whitens, but poverty also darkens. In such way, it would be possible to save both commonplaces: alter-classification would be more objective, and self-classification would vary more intensely with income – but the predominant effect would be self-darkening, due to the majority of the population belonging to the lower income strata. It is not possible to go further from the data of these researches, in the form they are presented. It would be necessary to have data desagregated not only by self- and alter-attributed race, but also by income stratum.

But at least one research exists, that may help to clarify this point, though it has not been used to such end. Its data may be found in Miranda-Ribeiro, Paula, and Caetano, André Junqueira. Como eu me vejo e como ela me vê: um estudo exploratório sobre a consistência das declarações de raça/cor entre as mulheres de 15 a 59 anos no Recife, 2002(6) (How I see myself and how she sees me: an exploratory study on the consistency of race/colour declarations among women between 15 and 59 years old in Recife, 2002. We reproduce its results below, rearranged and with modifyed labels so they can be directly compared to the researches shown above(7):

TABLE 4: Alter-attributed skin colour vs Self-attributed skin colour – Recife – 2002 (in %)
Skin colour, self-attributed
White “parda” “preta” Yellow Amerindian Total
Skin colour, alter-attributed White 20.20 3.58 0.15 0 0.92 24.86
“parda” 15.53 26.91 5.7 0.34 2.51 51.00
“preta” 1.76 9.68 8.39 0 1.26 21.09
Yellow 0 0 0 0 0 0
Amerindian 1.26 0.96 0.09 0 0.71 3.02
Total 38.75 41.13 14.33 0.34 5.41 99.97
Source: SRSR Research, UFMG/Cedeplar, 2002.

This is a research focused exclusively in women, which, in principle, doesn’t look too relevant on the subject that concern us here, there being no indication that either self- or alter-classification are significantly influenced by sex. On the other hand, the scope of this research is the city of Recife, not the whole country – which certainly influences significantly the results. To start with, this research presents opposite results compared to the three former ones: self-attribution of White colour is considerably higher than alter-attribution, and the opposite is the case with the Black and “parda” colours, in which alter-attribution is quite higher than self-attribution. With the due caution, due to such differences, we shall observe the results to which the study arrives, when it analyses the data closelier. First place, it is necessary to point out that the study, for analytical ends, disconsiders the small number of Yellow and Amerindian women, and groups Black and “parda” women in one single category, “negras”. This way, the study is able to establish four categories concerning the relation between self- and alter-declaration of skin colour:

Consistently White women;
Consistently Black women;
Women who “darken themselves”; and
Women who “whiten themselves”.

We now may consider what should happen according to the hypothesis we are criticising. If money “whitens” and self-classification is more subjective that alter-classification, data should show women who “whiten themselves” (according to the hypothesis, women who are in fact “negras”, but see themselves as White, probably due to their socio-economic level) with a higher socio-economic level than consistently Black women and than women who “darken themselves”, and those with a lower socio-economic level than consistently White women.

However, when we compare such predictions with the data from the research, what we see is quite different. Among consistently White women, some 31% have 12 or more years of school, against 11% among the consistently Black – as we would expect. But, from women who “darken themselves”, almost 25% have 12 or more years of schooling, and, from women who “whiten themselves”, only 9% have 12 or more years of schooling. So, women who “whiten themselves”, on contrary of what we would expect, have the lowest level of school attendance among the four groups, not only lower than consistently White women – which would be reasonable – but also lower than consistently Black women and than women who “darken themselves”. And women “who darken themselves” have the second highest levels of formal education, lower than consistently White women and above consistently Black ones – but much closer to the former.

We know that, although educational levels and income strata are correlated, they have different – and sometimes opposite – effects on political and social positions of individuals. Unhappily, the study does not present the data in a format that might allow us to compare the several groups according to variables less problematic in their relation to income, which limits the possibility of understanding how the complex correlation between educational levels and income is reflected in the results of this research.

But anyway, the study seems to give us elements to show that the two hypothesis we are discussing – that “money whitens” and that racial self-classification is less objective than alter-classification – are mutually excludent. Indeed, if alter-classification is more objective, then not only money does not “whiten”, but, on the contrary, it darkens – and it is poverty that “whitens”. This is the line of reasoning that the study takes, suggesting that women that “darken themselves” do it exactly in function of their higher schooling:

In theory, women with higher schooling would tend to have more information, and, maybe, a higher consciousness of social issues, racial ones included. So, it is plausible to imagine that, albeit looking White in the eyes of interviewers, women with more years of school and who do have Black origins, ancestry, and/or identity have no problems in affirming them, and, quite probably, are proud to belong to such group.

And, coherently,

Women who see themselves as lighter than the world sees them are those with low educational levels, and, quite probably, with a lower socioeconomic level. Within this line of reasoning, those who declare themselves lighter are those whose pheontype, in some way, is somewhat closer to those of Whites, but not enough to be perceived as Whites by third parts also. We cannot forget that the lower socioeconomic leve of these women means that they do not have access to exterior signs of wealth. According to this logic, they would understand whitening as a form of social ascension.

The alternative, that the study unhappily does not explore, is that it is alter-classification that is strongly more subjective than self-classification. In this case, it would still possible to maintain that “money whitens” – but it would be necessary to make it quite clear that what is meant by that is that personal wealth – or, more probably, status and prestige associated with it – modifies, quite intensely, the perception that society has of individuals, rather than the perception that those have about themselves: money whitens in the eyes of others, not in the eyes of the rich or enrichned individual.

Both alternatives – that money darkens, instead of “whitening”, or that self-perception is more objective than alter-perception – seem to contradict not only well-established traditions in Brazilian socio-anthropology, but also common sense and logic. The interest of each individual hin his or her own social situation – and the perception of “race”, in a society characterised by racial prejudice, as is the case of Brazilian society, is an important element of “class” perception – is, of course, much more intense than the interest of third parties. Moreover, in the case of the research under discussion, the third parties are not representative of society in general, but qualified observers – interviewers selected and trained by University staff, for the specific task of classifying individuals according to skin colour. On the other hand, the link between social status and skin colour is also very intense, to the point that a negative correlation between both, as posited by the study, has to seem a forced one.

And so we might be tempted to decide between those alternatives through the criterion of absurd – to decide which of those ruptures with tradition and common sense is less nonsensical, and chose it as the most probable. To demove the reader of taking such a move, we shall now argue to show that, tough both alternatives conflict with common sense, neither is indeed absurd, and, in fact, both make considerable sense, if we don’t allow ourselves to be influenced by the social (and, to a certain measure, academic) prejudices that underlie beneath common sense.

The idea that self-identifying as “White” can have a compensatory value for individuals with lower acquisitive power and social status may be strange at a first look but it is hardly absurd. Indeed, for the individual of higher status, whose position is assured, skin colour may well be irrelevant, or even play a role of individual self-valorisation throught the contrast between skin colour and a discordant social status (not only am I rich, but rich and Black, and so rich by my own personal effort, not through inheritance), which is perfectly compatible with the extreme individualism predominant in Brazilian society. On the contrary, to the poor individual, who owns nothing, White colour may well be the only remaining asset. Indeed, examples of racism among the lower social strata abound worldwide, and are an important component of popular mobilisation by the Right. And so, we should not discard the possibility, even though it it completely destroys an idea so traditionally accepted as “money whitens”.

In the same vein, the idea that “racial” self-classification is more, not less, objective than alter classification seems quite defensable. Indeed, it is the idea that it is in the best interest of the interviewee to present him or herself as “Whiter” than he or she really is that might be somewhat fantastic. If it was some kind of classification to be inscribed in official documents, which might have some kind of impact upon each citizen’s life, maybe; but we are dealing with anonymous research, devoid of any practical effect on the material existence of people. On the contrary, life experience will have taught each one which is the “race” in which they are framed, and those those have always been consistently considered non-White by relatives, neighbours, friends, classmates, co-workers, customers, etc., know that and know that denying such reality bores no fruit. As the popular saying goes, each one knows “where their shoes hurt”, which is not the case of the supposedly “neutral” interviewer; and so there is at least as good reason to beleive that self-classification is more objective than to believe the opposite.

Once we make such considerations, we remain with the realisation that the two myths are incompatible and cannot both correspond to reality. If one is true, the other is necessarily false, though we shouldn’t discard the possibility that they are both false. To go further requires either more precise or comprehensive data or new hypotheses capable of surpass this deadlock. In lack of the former, we are going to advance a hypothesis that we deem able to give an account of the facts as we know them and symultaneously demystify both the ideas that “money whitens” and that “racial” self-classification is more subjective than alter-classification.

Let’s contrast a supposition of Miranda-Ribeiro and Caetano’s study, that we left uncriticised, with a statement we made here in support of the idea that racial self-classification is more subjective. Miranda-Ribeiro and Caetano argue that women that see themselves as lighter-skinned than the world sees them are those with a lower educational, and probably socio-economic, level. We have pointed out that in the case of the research under discussion, the third parties are not representative of society in general, but qualified observers – interviewers selected and trained by University staff, for the specific task of classifying individuals according to skin colour. Now, putting those statements side by side, we may realise that “qualified observers” are not “the world”, the latter, on the contrary, being overwhelmingly composed of non-qualified observers. And more, “qualified observers”, from the point-of-view of educational levels, resemble much more the group of observed subjects that has higher schooling experience, than the group of observed subjects with lower educational levels. And consequently, we can expect the opinion of qualified observers to match the opinions of the subjects with a higher level of schooling than the opinions of the “less educated” subjects. And indeed, if we look at the figures of the four groups in which the study divided the women it discusses, we will see the following:

group %
Consistently White women: 21.97
Consistently Black women: 55.13
Women who “darken themselves”: 4.08
Women who “whiten themselves”: 18.82

Which means, the group of women who “darken themselves” is the smallest, being only 15.64% of the women that the interviewers consider White. The group of women who “whiten themselves” is more than four times bigger, and it makes 25.44% of the women considered non-White by the interviewers. So, the idea that the interviewers have of “race”, whatever it is, seems closer to the idea that the group that they have classified as “White”, which has, on the average, a higher educational level – exactly as the interviewers. On the contrary, the group that the interviewers have classified as non-White – that, in average, has a much lower educational level compared to the interviewers – has a quite different idea of race, compared to that of the interviewers. If we look at the data contrarywise, focusing on whether the women agree or disagree with the interviewers, we see that, among those who agree, 28.5% belong to the groups with higher educational level, while among those who disagree, only 17.8% belong to those with longer schooling years.

The hypothesis we are about to propose, thence, is the following: the concept of “race” of the groups with more schooling is much more restrictive than that of the groups of less schooling. By “restrictive” we here mean that, for the groups with a higher educational level, the set of phenotypes that classify as “White” is much smaller than the set of phenotypes that classify as “White” according to the groups with a lower educational level. We can’t know, at this point, whether such difference is due to educational levels themselves, or to the higher socio-economic level that underlies higher schooling. Our impression is that the difference has more to do with income strata than with schooling. Such impression is strengthened by the fact that the classificatory scheme of the interviewers seems intermediate between those of the more educated subjects, on one hand, and the less educated ones, on the other. If schooling was the direct cause of the more restrictive classificatory system, it would seem to us that the position of the interviewers would not be intermediate, but, on the contrary, more extreme than that of the more educated subjects (if for no other reason, because it was directly influenced, through training, by the classificatory scheme of the academic environment, where certainly average schooling is considerably higher). If, on the contrary, the restrictive classificatory system is a consequence of belonging to a higher social stratum, the intermediate position of the interviewers would be easier to explain – after all, they are middle class people, but still depend on ocasional jobs, characteristic of lower social strata.

How does such a hypothesis impact the common sense assumptions, that “money whitens” and that racial self-classification is less objective than alter-classification?

First, it completely displaces the discussion about racial self- and alter-classification: in the common sense assumption, the significant issue is whether the individuals are classifying themselves or others, not to what social groups those individuals belong. When applied to the available data, such assumption arrives a quite questionable conclusion: that women with more school years “darken themselves” because they have a “higher consciousness of social issues, racial ones included”. In our hypothesis, on the contrary, individuals reason according to their social belongings: it is not necessarily that women with more schooling have a higher consciousness, but that their apply to themselves the same criterium that their social environment applies to all women – probably, of course, because this is the criterium they experience when applied to themselves by their environment: they are considered “negras” by their co-workers, friends, neighbours, etc. If so, it is not necessarily a more progressist vision, or one less tinged by racism.

Conversely, the vision that women with lower educational levels have of themselves would not reflect, in our hypothesis, a smaller “consciousness of racial issues”, but the world vision of women who belong to a social environment in which schooling is lower: a social environment in which people are not so peremptorily classified as “Whites” and “Blacks”, with the consequence that those women are themselves classified by others – neighbours, co-workers, friends – as Black less often; and so it is not, necessarily, a more conservative vision, or one more in line with racist conceptions(8).

Second, our hypothesis dismisses any sense in the idea that “money whitens”. What happens is the opposite: in Brazil, the overwhelming majority of the population with higher educational levels is indeed White – consistently White women figures are five times that of women who “darken themselves”, and the other two groups are those composed by women with considerably lower educational levels. But even those women who “darken themselves” are considered “White” by the interviewers, and, in all probability, by women with lower schooling. Contrarywise, the majority of the population with lower educational levels is indeed non-White – even in their own evaluation: consistently Black women outnumber women who “whiten themselves” by almost three to one. And even those women who “whiten themselves” – lower class Whites, if we may say so – are considered Black by more educated women.


1. Rafael Guerreiro Osório. O SISTEMA CLASSIFICATÓRIO DE “COR OU RAÇA” DO IBGE. p. 13: The biggest issue of identification by self-attribution implies the problema of social variation of colour, since the extensive available literature on the subject, in spite of the different emphasis on class or race, unanimously posits that social ascension social may whiten, with abundant examples of the phenomenon.

2. Rafael Guerreiro Osório. O SISTEMA CLASSIFICATÓRIO DE “COR OU RAÇA” DO IBGE. p. 13: Knowing that, in the light of the current ideal of whiteness, it is to be expected that people who carry less Black traits in the appearance tend to consider themselves White, and that this tendency varies with socio-economic situation, with wealthier people also tending to the choice of White colour, the fact that colour classification is made by self-attribution may seem problematic. If, for instance, the huge difference between the average house income of Blacks (“pretos” and “pardos”) and that of Whites is to be considered, one might ask how much of such difference, in truth, is due to the fact that the vindication of whiteness is bigger among the wealthier and smaller among the poorer.

3. A similar sentence is also attributed to the football (“soccer”) player Róbson.

4. “Other” includes Yellow, Amerindian and others. Data rearranged from the original tables “Cor auto-atribuída segundo cor observada” and “Cor observada segundo cor auto-atribuída”.

5. “Other” includes Yellow, Amerindian and others. Data rearranged from the original tables “Cor auto-atribuída segundo cor observada” and “Cor observada segundo cor auto-atribuída”.

6. Paula Miranda-Ribeiro e André Junqueira Caetano. Como eu me vejo e como ela me vê: um estudo exploratório sobre a consistência das declarações de raça/cor entre as mulheres de 15 a 59 anos no Recife, 2002.

7. There is some inconsistency in totals, probably due to the way the percents were rounded up in the cited text.

8. Observe that this reasoning would put back into question what we have stated above about the relevance of sex in relation to self-classification (in principle, doesn’t look too relevant on the subject that concern us here, there being no indication that either self- or alter-classification are significantly influenced by sex). However, since men are certainly more vulnerable to disagreeable encounters with police – and probably, to the point in which their traditional role as providers may have yet not modified by the development of capitalist social relations, are more frequently in contact with individuals of different social classes (employers, bosses, etc.) – it is possible that sex actually influences self-classification in the direction of a higher adequation of men to the standards of upper income strata. The same could probably be said about women who work out of home, especially in workplaces, and it would be interesting to verify whether the self-classification of housewives (and perhaps housemaids) differs from that of factory workers or employees in commerce and services.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License