Article body

The organizational research agenda in the USA and EU countries has been dedicating increasing attention to issues of managing a diverse workforce. For some of the proponents, the arguments for diversity are restricted to questions of social demography and the need to accept that today’s workforce is no longer the homogenous workforce of the management training literature. For others, concern with diversity is also derived from concern with equality in the workplace and the need to move on from previously proposed remedies – usually around affirmative action - for dealing with issues of inequality and discrimination. Important here has been the longitudinal work by Dobbin and his colleagues (see Dobbin, 2009) in the USA which has shown how government action itself towards federal contractors and a new generation of personnel professionals combined to socially construct new patterns of practices, manuals and procedures, which in their turn became guidelines for court decisions.

How to seek a balance between diversity, inequality and discrimination tends to divide contemporary management and organizational scholars. The mainstream is occupied by those for whom diversity management theory and practice represent a positive development of the previous literature on inequality in the workplace; correcting mistakes whilst remaining in tune with affirmative action (Zanoni et al.; 2010; Tatli, 2011; Özbilgin; Tatli, 2011). A more critical approach can be found amongst those who argue that mainstream diversity management discourses and suggested practices actually represent a threat to the conquest of a workplace free of discrimination and inequalities, since their principle pillars are those of individual differences and the business environment performance, essentially putting aside issues of structural inequalities that by certain degree inform organizational practices (see for example, Noon, 2010).

These questions are key to the contemporary Brazilian organizational scene and we place ourselves in this debate for very practical reasons. Brazil has still a long way to go in terms of racial equality both for black people in general and black women in particular, as a recent special article in the Economist confirmed (28th, Jan 2012). A recent study by Wood, Carvalho and Horta (2010) has shown clearly how, despite general improvement in life expectancy at birth for the population as a whole over the last fifty years, the difference within this as expressed by child mortality statistics has remained constant for white and black children. Unfortunately, Brazil’s booming management and organizational science scholars are still way behind their USA and European colleagues in the discussion of such persistent and durable inequalities (Tilly, 1999) and to the study of inequality regimes in work settings (Acker, 2006; Healy; Bradley; Forson, 2011). Our interest in the comparison with the USA experience has a number of reasons. Brazil has currently the largest African origin population of any nation other than Nigeria. Like the USA, its afro-descendants were not voluntary migrants but slaves (more than 3 million in the seventeenth and eighteenth century and still amounting for over 50% of the current population). This was some ten times the number that were shipped to North America where the current black population is approximately 12%, yet the actions taken by government, academics and by business professionals in the USA to overcome inequalities based on race has been more vigorous. Like the US, Brazil has a federal structure and a number of its institutions are modeled directly or indirectly on this approach to governance. Brazil is home to a number of USA companies and its management practices are very much inspired on USA management manuals. Yet, in Brazil, the process of adopting diversity or affirmative action practices has unfolded in very different ways from that of the US. In Brazil, and still with generalized reluctance, it is the diversity discourse that, if anything, first attracted the attention of the more progressive business community and it is the failure to achieve results that has led institutional actors, in this case the Federal agency of the Public Prosecutor’s Office for Labor, to pressure for an affirmative action approach to policy implementation.

As we say in the title of our paper: which foot first? Can diversity management programs make it possible to overcome structural inequalities or is it necessary to have government imposing measures and sanctions – within an affirmative action framework – in order for organizational policies and practices to change? What is more, how does this play out in other societies where some groups are historically treated in an unfair way both in terms of recognition and redistribution (Fraser, 1997)? Bringing the Brazilian case to the table provides, we feel, an important setting in which to test these questions for, as we later comment, the social and political dynamics of racial inequality are certainly enduring and demonstrating considerable resistance to different discourses and strategies of change.

From this point on this paper is divided into more five parts. The first section briefly reviews diversity management literature, given special attention to the change in the tendency from affirmative action towards diversity management. The following section discusses our methodology. In the third section some aspect of Brazilian history and culture, especially those that are credited with having determined relations between blacks and whites in the labor market, are briefly presented. In the fourth section, we will set out the factors at the three levels of analysis that contribute to the way Brazil-based companies approach diversity management and race relations. The fifth section will discuss the data presented in the previous sections, and the conclusions.

Diversity Management in debate

The concept of diversity management was first employed in the USA during the late 1980s, signaling a shift away from affirmative action as a lever for change in the human resources and management arena (Tomlinson; Schwabenland, 2010; Özbilgin, 2005; Wrench, 2005; Humphries; Grice, 1995). Affirmative action, linked to the forceful and more immediate action frame of rights had emerged in the post second world war period as a consequence of the civil rights movements fight for social justice, and was based on such moral values as equality and fairness (Tomlinson; Schwabenland, 2010, Wrench, 2005). Even if the Wagner Act of 1935 had already signalized a new direction for relationship of the public sector and private organizations, since it established some mechanisms to protect trade union workers from being dismissed, affirmative action as it is discussed today was a result of title VII of the 1964 Civil Rights Act. This prompted employers to experiment with a number of antidiscrimination approaches, including numerical employment quotas for disadvantaged groups and making it illegal for employers with 100 or more employees to discriminate on the basis of race, color, national origin, sex, and religion (Dobbin et al., 1993; Dobbin, 2009). Nevertheless, despite such forceful arguments, affirmative action began to loose popularity amongst the criticism that it had led to a lowering of educational standards – for example in college access –, and that, by threatening the principle of equality it was meant to produce, had exacerbated rather than relieved racial tension (Dworkin, 2002, p. 387; Chanlat; Dameron , 2009).

There is no doubt that the drift away from a more hard-line rights perspective was influenced by the shift from liberal to more conservative policies in the USA (seen in various court rulings), but it is also the case that diversity management came along at a time when the business in USA as well as in EU countries was facing a very different labor climate characterized by a growth in the number of ethnic minorities in society as well as the increasing presence of white women in senior positions (Chanlat; Dameron, 2009; Humphries; Grice, 1995; Cox; Blake, 1991). Additionally, international trade was at the front of the management discussion with new trading blocks, new forms of business partnerships and joint ventures that crossed cultural boundaries, reinforcing the demand for greater flexibility in organizational performance and a more micro level approach to differences that was possibly also influenced by the difficulty business had with the more direct and explicit affirmative action approach (Gilbert; Stead; Ivancevich, 1999; Özbilgin, 2005; Wrench, 2007).

The relatively rapid acceptance of the “diversity management” approach should not however be seen as a rejection of the principles and values of “affirmative action”, since amongst practitioners there are those who see the former enabling the latter to be accomplished in practice (Kirton; Greene; Dean, 2007). Indeed, as Tatli (2011) has suggested, most of the time there is a decoupling between the diversity management discourse and practice, with the latter remaining strongly based on affirmative action values present often in a very material way in manuals and procedures (Dobbin, 2009). Nevertheless, at a visible and explicit level the diversity management discussion is about the challenge of leading heterogeneous workgroups and the effective use of a variety of resources to accomplish this task rather than the ethical challenge of protecting rights of minority groups. However the order is clear, within the USA the diversity discourse is – at least for the moment – in the spotlight and affirmative action has been pushed into the shadows.

In European Union countries, the history is partly similar in the sense that a number of countries had also experienced affirmative action policies before the arrival of the diversity management discussion on problems of discrimination at the labor market. The difference is that here, much of the early legislation was concerned with gender issues. At the broader level of the growing European Community as a whole, there was concern since 1961 with pay parity between women and men, and in 1975, through Directive 117, the first legal steps toward harmonizing legislation on the prevention of gender inequality in the labor market were taken. Affirmative action came in 1984, with the publication of Recommendation 635 on the promotion of positive action for women in the labor market (Cappellin, 2000).

It was in the 1990s that EU countries, as a community, began to face up to discrimination based on ethnic and racial origin (individual countries had already been working on these questions for some time). During this period non-governmental and immigrant organizations published studies showing that immigrants lack of language fluency and work abilities were not the reasons behind the growing inequalities in the EU labor market. A new directive came into force in 2000 that, amongst others, prohibited discrimination on the basis of racial and ethnic origin in the labor market and urged each national state to integrate the directive into their national legal systems by 2003 (Wrench, 2002). The directive and the subsequent campaigns for organizations to embrace anti-discrimination policies coincided in Europe with a growing awareness of USA discourse and practice of diversity management and, as in the USA, the diversity management discourse prevailed over that of affirmative action (Wrench, 2002). However in both regions, there remained significant groups of academics and practitioners who continue to defend the affirmative action perspective.

As we commented in the introduction, contemporary research literature on both sides of the Atlantic north can be classified into two major blocks (Zanoni et al., 2010; Tatli, 2011; Özbilgin; Tatli, 2011). On one side are those studies that fall into what has been called the mainstream approach, making the business case for diversity management and the benefits an organization can yield from a diversified workforce. Here the concern is often with performance, and thus justifying diversity practices by their potential of improving both business operations and outputs as well as financial results (for example: Chanlat; Dameron 2009; Cox; Blake; 1991; Frimousse; Paretti, 2007; Gilbert; Stead; Ivancevich, 1999; Leroux et al, 2008; Tatli; Özbilgin, 2009).

A second block argues that the mainstream studies tend to draw from the neoliberal well of ideas, assuming social action to be totally explicable through theories of maximizing self-interest. Despite their multiple perspectives, researchers here take a more critical approach, showing concern about the moral case for equality and the dangers that the loss of emphasis on moral values can have for social justice. Here, affirmative action is seen as the most effective route for organizations to achieve a diversified workforce (for example: Greene; Kirton; Wrench, 2005; Humphries; Grice, 1995; Mor Barak; Findler; Wind, 2003; Noon, 2007; Perriton, 2009; Sinclair, 2000; Wrench, 2005; 2007).

Critical diversity management research emerged in the mid-1990s in response to the quickly rise in studies using the mainstream approach to diversity. Critical scholars put a spotlight on the issues of structural inequalities, discrimination in the workplace, and the way one is related to the other. They contest that it is not concern about performance, respect to individual differences, and positive organizational outcomes that must be addressed in discussing diversity management. On the contrary, they argue, it is the perpetuation and reproduction of inequalities in marketplace that should come in first place.

These scholars show considerable discontentment with mainstream approach. Amongst reasons given are that it: obscures unequal power relations in organizations and hampers the possibility of action (Zanoni, et al., 2011); is based on an instrumental and individual view of differences (Noon, 2010; Özbilgin, Tatli, 2011; Zanoni, et al., 2011); is a tool to undermine identity-based collectivism in organizations and to obscure group-based systematic inequalities in access to power and resources (Özbilgin, Tatli, 2011; Tatli, 2011; Zanoni, et al., 2011); rejects the need for structural intervention, adopting a (neo)liberal approach of self regulation through assuming that market forces would provide the best way to achieve equality (Noon, 2010); and is used to prioritize soft rather than hard practices to enhance diversity or equal opportunity practices (Wrench, 2005).

Some critical empirical studies on diversity management sought to bridge the gap by suggesting that harder diversity practices – here understood as those that promote changes in the organizational structures – generate a more diversified workforce than softer ones. Kalev and Dobbin (2006), for example, showed that policy interventions that stimulate change in organizational routines – such as those derived from the compliance review established by chapter VII of the USA Civil Rights Act of 1964 – appear to have more significant and lasting effects on workforce diversity than those that just create disincentives to discriminate, such as lawsuits. Following this same direction, Kalev, Dobbin and Kelly (2006) showed that, for private organizations, diversity programs that aim at altering organizational structures by establishing responsibilities, such as affirmative action plans, diversity committees, and diversity staff positions, are followed by significant increases in managerial diversity. That is, through effectively allowing the beneficiaries – in the USA case, white and black women and black men – to reach managerial positions. Softer programs, such as diversity training and diversity evaluation, are not followed by increases in diversity in the workforce, whilst, programs that address social isolation amongst women and minorities – such as networking and mentoring programs – are followed by modest changes.

To be fair, as Tomlinson and Schuwabenland (2010) highlight, there are also those in both the mainstream and critical groupings that will accept some coexistence and argue that business pragmatics can extend themselves to social justice and that diversity management programs can promote social equality. Equally, some scholars on the mainstream side will contend that diversity management policies and practices should be seen as a step towards affirmative action, given that in a number of cases they themselves are a result of the fight against discrimination (Frimousse; Peretti, 2007; Leroux; Frimousse; Peretti, 2008); certainly the reaction to HIV/aids discrimination at work would back this up. However there are others, especially those from the critical side, who take the opposite path and contend that equal rights legislation and affirmative action are prerequisites for the development of diversity management since they create the social, legal and organizational environment and incentives which administrative initiatives require (Greene; Kirton; Wrench, 2005; Wrench, 2007).

Theoretical framework and Method

Given our concern to look more carefully at some aspects of the Brazilian case and to do so in a setting in which both affirmative action and diversity management discourses and practices are present, we chose to structure our investigation by using “relational framework” approach of Syed and Özbilgin (2009) (See similar comments by Jonsen; Maznesviki; Scheneider, 2011 and Tatli, 2011). This opens useful directions for analysis by proposing to bridge the gap between macro(national), meso (organizational) and micro (individual) level policies and actions. As they argue, the single level mainstream conceptualization of diversity management, which considers either the legal or the organizational domain of the policy, fails to grasp the relational interplay of structural and agentic-level concerns of equality (Syed & Özbilgin 2009, p. 2435). The authors’ theory is based on the ideas of such social theorists as Bourdieu and Layder who assume reality as a multi layered phenomenon, governed by hidden and underlying structures not easily grasped by more surface level observations (Reed, 1997; Syed; Özbilgin, 2009; Tatli, 2011).

The agency/structure debate brings important questions about the nature of social reality, the way in which people conceive it and the theoretical means available to explaining the relationship between its constituent parts, raising questions about the nature and the link between human activity and its social context. As Reed has argued, (1997), the way in which we conceive the nature of and relationship between social action and structural constraint will shape our visions of organizations. Simplifying, what has been called the critical realist perspective, argues that society, and its institutions and organizations, is a socially constructed phenomenon, being something outside the mind and relationship of individuals. They also agree that it is not impartially built. On the contrary, for these scholars, society is made up of a variety of social groupings, which are constantly fighting to shape its institutions and rules according to their interests. The “social reality” of society is by no means a static affair and change is very much part of its complexity.

For the purpose of this paper, that is to analyze a process of social change in face of durable inequalities, taking the Syed & Özbilgin model within a hard diversity perspective (see Özbilgin; Tatli, 2011) provides a powerful theoretical framework from which to attempt to unpack the Brazilian case. Developed for the discussion of the international transfer of diversity practices, the model encourages comparisons between the approaches adopted in different countries in managing a culturally diverse workforce (see for example Syed; Kramar, 2010 on Australia). Brought more forcefully within a critical perspective, it may help us to understand what the possibilities for action may be in a setting where these are very much needed.

In our case study, we examine the factors that concurrently shape the way Brazil-based companies deal with diversity management and, more specifically, racial inequality at the three levels of analysis proposed. At the macro-national level, we will examine Brazil’s laws and public actions on equality of treatment, especially those aiming at promoting equality in the workplace. At the meso-organizational level, we will discuss the configuration of the workforce in the labor market and the way Brazilian-based business organizations approach the problems of equality and diversity; and at the micro-individual level, the focus of the investigation will be the Brazilian black people, their historical role in the labor market and the factors that contribute to their subordinate participation. Our focus on Brazilian Banks is both opportune and representative. Firstly, the Banks through the national Federation of Brazilian Banks (FEBRABAN) were the objects of direct pressure from the Federal Public Prosecutor’s Office for Labor. Secondly, the Banks are also large and very visible employers and in many respects highly modernized. Over recent years – following similar trends elsewhere – the financial services sector has seen a number of significant mergers and acquisitions and a small number of public and private high street banks control a large amount of the market. They may not be at the leading edge of personnel practices but they are certainly within the top quartile.

We adopt a multidisciplinary approach (Kamenou, 2007) informed not only by the diversity management literature but also by the discussion on inequality and Brazilian racial relations. The paper draws on a multiple source of secondary data, published until 2011, from both Brazilian official research centers (Brazilian Institute of Geography and Statistics – IBGE and Institute for Applied Economic Research – IPEA), and specialist research centers for black studies (Laboratory for Economic, Historical, Social and Statistical Analysis of Race Relations, Federal University of Rio de Janeiro – LAESER-UFRJ), as well as on publically available surveys and qualitative studies carried out by non-profit organizations (Ethos Institute) and employee organizations, including social responsibility data from the FEBRABAN. We have also included evidence from nineteen in-depth interviews that were carried out between 2010 and 2011. Five of these were with representatives of Brazilian black social organizations that were involved in the Bank case. Two other interviews were carried out with the then general prosecutor from the Federal Public Prosecutor’s Office for Labor at different points in time, one in 2010 and the other more recently (August, 2011) and an earlier interview with the prosecutor’s key advisor in 2010. Finally, we have also been helped by interviews with two personnel managers from leading São Paulo companies and eight students involved in a program to place black students as trainee interns, developed by a São Paulo based black students college. A further interview was carried with the College President.

Black people in Brazilian society and its labor market

Brazil is internationally known as a culturally diverse country and this is also the image that can be found on official posters and photographs. However what is less known is that this diversity consists basically of two sets of peoples: the whites, by and large European and Asian descendants whose arrival in the country was broadly voluntarily, and the Afro-descendants whose forefathers were brought to Brazil in slavery. Figures from the last official statistical survey accomplished in 2010 by the IBGE, show that approximately 50.74% of the Brazilian population self-categorize themselves as black or mulatto (semi-black) and that the whites accounts for 47.73%. The remainder was made up of people classifying themselves as indigenous or Asiatic in some way (Japanese, Chinese and Korean, Instituto Ethos, 2010). In general, Asian descendents tend to be seen as white.

Even if Brazil has until recently had a fame of being a “racial democracy”, the relationship between the two major groups (whites and blacks or “negros” as they are described) has never been egalitarian. Again simplifying, whites have always been at the top level of society, dominating the governmental, political, economical, professional and social arenas. Figures from de last census (2010), for example, show that black people earn 45% less than whites. Whilst a few afro-descendant have recently reached more visible social positions, they are until now disproportionally underrepresented in all the important arenas, including the labor market. Given that they form, in number, the majority, the question that inevitably arises is why and how?

Some answers can be found in the history of ideas, especially in the late XIX century and the period surrounding the abolition of slavery in 1888. Accordingly to Schwarcz (1993), it was in the XIX century that the polygenic perspective of humanity began to take over from previous monogenic ideas. The polygenic followers believed that humankind’s intellectual capability as well as its moral character was determined by its genetic inheritance. They also believed that humanity could be divided into races, with the Negro being the less endowed. Following this line, some Brazilian intellectuals argued that the miscegenation between black and white would lead to total degeneration. Others argued the opposite: that miscegenation was the only solution, since, in terms of social identity, it would lead to all differences (including blacks and indigenous people) “disappearing” (Schwarcz, 1993; Skidmore, 1976). The force of the arguments, for and against, strengthened the polygenic perspective and their echo could be found amongst the country’s political leaders and its citizens and present in reports of racial discrimination following abolition.

The years between 1930 and 1950 witnessed the emergence of a new school of thought, arguing that Brazilian racial diversity instead of being a disaster, as it was commonly thought, was in fact an advantage. Gilberto Freyre, one of the most prominent thinkers of this period, and his contemporaries proposed that Brazil’s weakness was not its black people and their mixed descendants, but instead its culture. Brazilians were used to attaching too much importance to foreign cultures rather than to their own (Dávila, 2006, p. 30). The cultural proposal did not change convictions about black inferiority, and even Freyre would argue that black people would gradually “disappear” and all Brazilian people would become white (Jaccoud, 2008). The cultural argument did however change the racial discourse, bringing in the idea of “racial democracy” (Guimarães 2008, p. 180). That is, that, all Brazilian, no matter their race, are first and above all Brazilian and therefore all Brazil’s inhabitants were able to live together without distinction (Jaccoud, 2008, p. 52).

The fame of “racial democracy” was such that after the Second World War, a Unesco study team was sent to investigate how the country had managed to build such a harmonious and well-integrated society. Expecting to find a model that could be imported to other racially diverse Western countries, what they found was disappointment. The research results revealed that the Brazilian social fabric was completely undermined by inequalities and that even after over fifty years of abolition, black people remained in almost like slave conditions. They were discriminated against by whites and had no chance for upward mobility and advancement (Andrew, 1992; Munanga, 2001. The result was the emergence of a third and more critical school of thought about race relations in Brazil, whose representatives claimed that its interracial relations were still based on the values of slavery (Munanga, 2001). By the 1980s, with the advent of racial relations research programs, it became clearer that white prejudice against blacks was based not only on the former slave relations but also on social inequalities that are reproduced continuously every day by the people themselves and the country social and political institutions, the two come together in a complex manner, creating a dialectical systems of discrimination, which relies upon both social class and ethnic-racial relations (Dávila, 2006; Henriques, 2001; Munanga, 2001).

Focusing the labor market, Theodoro (2008) argues that abolition resulted in former slaves being excluded from the most dynamic regions of the country and from its more prominent economic sectors. Many of them ended up working in unpaid jobs as a mean to guarantee minimum needs. Government agents, following the polygenic perspective and concerned to purify Brazilian society, set up a program through which white immigrants from Europe were encouraged, and many times monetarily subsided, to enter the country. Arriving here, even without the necessary technical training and experience, they were promptly hired by the recent opened industrial plants.

Today, more than 120 years since the 1888 abolition of slavery, the condition of Brazil’s black population has hardly changed since that of the first half of the XX century. Afro-descendants are extremely underrepresented in the more desired professional and social positions and extremely overrepresented in poverty settings, low-paid jobs, jails and criminal gangs. With relation to the labor market, black people with the same or superior educational level as white people (men or women) occupy positions of lower status, lower pay and less influence. They are rarely promoted, are more often dismissed and remain unemployed for a longer time (Castro; Guimarães, 1993; Henriques, 2001; Myers, 2003; Sá Barreto 1988). Discrimination against black women in the labor market is even greater and they are by far the worst off (Andrews, 1992; Castro; Sá Barreto, 1988; Instituto Ethos, 2010; Paixão; Carvano, 2008).

Whilst the Brazilian black social movement had always raised its voice in favor of the complete integration of the afro-descendants in society, it was from the1980s onwards that the country witnessed an inflexion in their agenda from a claim for the integration of the afro-descendants in society toward demands for social justice and policies aimed to promote equal opportunities (Guimarães, 1999: 110). According to some of the interviews, black social movements in the country began to discuss affirmative action in the first half of the 1990s, seeing it as a mean to guarantee blacks the opportunity to enter university and maybe the labor market. In terms of university entrance, a number of public universities have moved towards quotas – but not without considerable controversy (Penha-Lopes, 2008).

In terms of the labor market, in November 2003 a group of black social organizations presented petitions to each of the 26 states and the federal district agencies of the Federal agency of the Public Prosecutor’s Office for Labor (MTP - Ministerio Público do Trabalho) against Inequality in the Labor Market. The goal was to denounce the differences in treatment practiced in the labor market and need for action to correct such imbalances under penalty of “threat to the core principles of the legal democratic state” (Lopes, 2006). This was to prove an important trigger for future events when in 2005 the MPT began to develop its own program for promoting equal opportunity for blacks and white women in the labor market that would lead to the events that are discussed in the next section.

Diversity management in Brazil: the Syed and Özbilgin model in practice

We begin the analysis by concentrating on the macro-level organizational context, the institutional framework and the institutions themselves involved with the equality and diversity legislation. In the second subsection we will examine how business organizations in the banking sector have responded to diversity management concept and practices. Finally, in the micro-individual level subsection we will discuss data on the educational backgrounds of black people in Brazil and their readiness for the labor market.

Macro-national level factors

Given that until recently the ideology of racial democracy has been quite successful in maintaining the denial of social inequalities based on race – including amongst blacks themselves – Brazilian society has never had to engage with the creation of mechanisms to overcome racial inequality. At the federal level, the first law on racial discrimination was passed in the year of 1951. It prescribed racial discrimination against foreign blacks as a misdemeanor, and was motivated by complaints from some influential foreign black visitors who faced racial discrimination whilst travelling around the country. However it said nothing about Brazilian’s own black people (Jaccoud et al, 2009).

At a constitutional level, although former Brazilian constitutions have also formally considered the principle of equality, it is only in the current 1988 charter that equality of opportunity and treatment is required to be substantive for all its citizens and the state is required to be proactive in order to attain this objective. The constitution establishes racism as a crime subject to prison and for which bail is not allowed (Brasil, 2007). It also requires equal remuneration and a quota for disabled workers in the public service (Brasil, 2007; Sarmento, 2008; Silva, 2008).

The underlying assumption in the 1988 Constitution that Brazilian is an unfair society opened the door to a number of anti-racial discrimination laws and other measures (Jaccoud, 2009; Sarmento, 2008). Largely as a result of the black social movement, the last twenty years have witnessed not only the passage of laws on racial issues at both the federal and states levels but also the implementation of official programs. Unfortunately, many of them were discontinued in part for lack of experience and those that continued have been poorly implemented and controlled (Conceição, 2010; Jaccoud, 2009).

Although Brazil ratified two ILO conventions on discrimination in the labor market and the workplace (Convention 100 and 111), its governments have never encouraged public and private organizations to adopt any form of affirmative action for black people, even after approval of the 1988 Constitution. More recently from the late 1990s onwards, a number of public and trade union research organizations published statistical studies with clear evidence of inequality, confirming that racial discrimination is a characteristic of the labor market. Despite this, neither the executive nor the legislative branch of the state has created any explicit policy to guarantee equal opportunity in the labor market for black people (Conceição, 2010; Jaccoud 2009). In contrast, the Brazilian parliament passed a bill in 1999 to require companies with more than a hundred employees to adopt affirmative action in favor of the disabled (Bahia, 2006).

In response to the State’s lack of action to benefit other discriminated groups as well as to respond to the black movement’s demands for fair treatment in the labor market, the Federal Public Prosecutor’s Office for Labor (Ministério Público do Trabalho- MPT) decided in 2005 to develop a program for promoting equal opportunity for blacks and white women in the labor market. The program was to be implanted in all major areas of the Brazilian economy beginning with the financial services sector and the banks (Conceição, 2010; Varella, 2009) (Brazil’s macro-level institutional structure included the classical three powers, legislative, executive and judiciary, as well as the independent public prosecutor’s office, which is seen in the constitution as the guardian of the democratic state. A specific branch of the agency deals with labor matters.)

The plan was for a Program for the Promotion of Equal Opportunity for All (Programa de Promoção da Igualdade de Oportunidade para Todos -PPIOT) to be implemented in two stages. First the MPT would contact the most prominent companies of a given economic sector and required them to make available information on workforce composition and earnings with respect to race and gender. In the second stage the MPT would assess the information provided by the companies and, if this showed an overall underrepresentation of black women and men amongst the work force as a whole and underrepresentation of black men, and black and white women in middle to senior management, the companies would be offered the option of signing a collective consent agreement (Termo de Ajustamento de Conduta, hereafter referred to as TAC) rather than face prosecution. In the TAC, the companies would commit themselves to adopt measures to increase the number of black people amongst their work force as a whole and also to increase the number of black men and white and black women in middle and senior management positions. If the companies did not meet the terms of the agreement, the MPT would move to prosecution (Conceição, 2010). The next subsection will present data on the program outcomes so far at the meso-organizational level of analysis.

Meso-organizational level factors

Brazil-based business organizations, as with many others around the world, are always prepared to argue that their actions are based on the principles of rationality, equality and merit: the dominant managerial discourse. But this is far from the case. A number of studies have shown that racial inequality is commonplace in Brazilian companies and that they have done and continue to do almost nothing to change the situation (FEBRABAN, 2011; Castro; Guimarães, 1993; Fleury, 2000; Henriques, 2001; Instituto Ethos, 2007; 2010; Jaime, 2010; Myers, 2003; Sá Barreto 1988; Varella, 2010). On the subject of diversity management, a more recent focus for research, evidence tends to show there is more acceptance of the requirement to employ the disabled. There have also been some moves to promote equal opportunity for women (Fleury, 2000). Unfortunately no data is available on the gender and racial distribution of disabled employees.

Recent data from the Ethos Institute (a leading NGO for business ethics) shows that amongst the higher level managers of the 500 best companies in Brazil only 5.3% are Afro-descendants (with varying skin colors) and just 0.2% is completely black. This same study showed that only 3% to 6% of these companies have policies to promote racial diversity amongst their workforce (Instituto Ethos, 2010). The Brazilian commercial bank association (FEBRABAN) publishes annually a consolidated social balance sheet for its affiliates, the biggest commercial banks. The 2009 report shows that African descendants make up to 15% of the total workforce. These figures are hardly different from those of 2005 (14%), the year in which PPIOT was implemented (Febraban, 2006; 2010).

In reaction to the PPIOT, each of the five leading commercial banks to which the program was first applied refused to sign the TAC proposed by the MPT. They argued that as law abiding Brazilian companies they did not discriminate against blacks. In the resulting prosecutions, the MPT – facing a largely conservative judiciary – lost the battle in the lower courts. However, under pressure by the Prosecutor’s Office, FEBRABAN agreed to undertake a full census of the workforce. Afterwards, the higher courts followed the lower courts and also ruled against the MPT (Varella, 2009), but by this time the census had been taken showing the almost complete absence of afro-descendents, especially in the highest level positions.

The result of the survey was released in July 2009. Its conclusions confirmed what was already suspected: (i) white women earned 78% of white men’s wages and reported more obstacles to career opportunities in the banks hierarchy, (ii) only 20.6% of employees in the financial system were black (black or brown), (iii) this group earned on average 84.1% of the salary of whites, (vi) the discrimination was still higher to black women, since they amounted to a total of only 8% of the population employed in the banking sector, and (v) whilst only 60% of all bank employees had been working in the banks for at the most ten years, for the blacks workers, 66% had been there for less than three years (Contraf-Cut, 2009; FEBRABAN, 2011). The results also showed that whilst the proportion of white man and women with postgradaute education was the same (approximately 50%), only 19% of those in higher positions surveyed were women (nearly always white). In the same way, although blacks are a little behind whites in respect to professional training, they accounted for only 4,8% of the executives in higher positions.

As the aide of the chief prosecuting attorney reported when interviewed, as a result of these figures, FEBRABAN (on behalf of the bank´s involved) agreed to: (i) increase the number of blacks, especially black women, on the bottom line; (ii) expand the mechanisms of career advancement for white women; (iii) create a diversity committee for sensitize former employees, especially those on the top, on the diversity values (iv) encourage suppliers to take part in the process by creating mechanisms to achieve the same goals in their own business; and (v) create mechanisms to oversee whether the goals were attained, not only by the banks themselves but also by the companies in their supply chain.

By this time, advocacy organizations in the disability arena had also managed to include the rights of the disabled on a separate and parallel MPT agenda. In this case, however, the MPT was able to convince FEBRABAN to ratify a collective consent agreement (TAC) with relation to the inclusion of disabled (October, 2008). These were later to be included in the broader agreement on human resource policy.

On its side, the MPT temporarily suspended the lawsuit as recognition of the changes in place. Additionally, from the interviews carried out for this study, the then chief prosecuting attorney and creator of the PPIOT program, saw the chances of the MPT winning the legal battle as remote, arguing that Brazilian judges were still over inclined to deny racism, even when evidence proves the contrary. The situation today continues very much the same. A close examination of the FEBRABAN site and published documents reveals very little, if any, action on the inclusion of black employees. Data on the census and the accompanying concerns is present but there is no evidence of plans. In contrast there are at least two special programs devoted to social inclusion through education and professional qualification for people with disabilities. The 2010 consolidated social balance sheet from FEBRABAN (published in August, 2011) did not show any data about black employees – discontinuing a practice that had been in place since 2002. The report was restricted to figures about gender, age and educational level of the new entrants, as well as the resulting composition of the workforce. The only mention of black men and women is a reference to a statement that one of FEBRABAN’s priorities is to encourage a greater participation within the banking systems of women, blacks, young trainees, and person who are disabled, and that FEBRABAN has signed a partnership with some governmental secretariats to strengthen and create policies directed to women and the black population, as well as to other vulnerable groups, with the purpose of increasing their number in the labor market (FEBRABAN, 2011).

Micro-individual level factors

Historically there is an educational gap between black people in Brazil when compared with their white fellow citizens (IPEA 2007), making it more difficult to find well-paid and higher-status jobs. A longitudinal study by Paixão and Carvano (2008) compared blacks and whites in the educational system in Brazilian from 1995 to 2006. The result shows that even if black people are gradually overcoming the educational deficit at elementary and high school levels as well as at university, there are still great differences especially in the higher levels and in grade performance.

In relation to undergraduate degree programs, Paixão and Carvano reported a significant growth rate (415%) in black students. However the greater part are enrolled in the lower rated fee paying schools and universities and those that are in the top rated public and private universities are majoring in easier access courses such as humanities and social service and are extremely underrepresented in medicine, engineering and other higher status courses (Penha-Lopes, 2008).

At the level of a simple non-causal observation, the data describes part of the lack of success of black people in the labor market, since academic performance is highly correlated with both opportunity and payment level in the labor market (IPEA 2007: 282-290). This is also the argument – lack of academic qualification – that employers use regularly to justify the nearly complete absence of blacks in higher-paid positions and the absence of diversity programs (Fleury, 2000; Myers, 2003). As a causal implication, the argument is flawed as it disregards the institutional and structural racism of the country and the literature and data on durable inequalities (Tilly, 1999). It is worth recalling that the IPEA annual report on Brazil’s social policy performance has always demonstrated that discrimination against blacks is an important variable in the explaining of inequalities in the labor market. Its 2007 edition brought together a range of statistical data on educational level and labor market wage for blacks and whites in the period from 1994 to 2005. This compared blacks and whites of similar educational levels and showed that blacks with the same educational level as that of whites always earn less. This is the case throughout the educational spectrum: from those who don’t finish elementary school, through those who did high school on to those who went to university and beyond to postgraduate training. Black people as a whole enter the labor market in disadvantage to whites (Paixão and Carvano, 2008) and in situations when this disadvantage is removed and they are educationally equal, they earn less. Thus, as the IPEA study concluded, the unequal treatment given to black people in the labor market cannot be explained on the behavioral grounds of a lack of education but needs to be seen in terms of racial discrimination both in the market and in relation to access to education (IPEA, 2007, p. 282-290).

Discussion and Conclusions

The purpose of this article was to examine, in the Brazilian case and by extension to other countries where groups have been historically treated unfairly, whether diversity management programs can make it possible to overcome structural inequalities or if it is necessary to have government imposing measures and sanctions – based on affirmative action values – in order for organizational policies and practices to change. To address this question we presented evidence about the way diversity management concepts and practice developed in the USA and EU countries and an empirical study of recent events in Brazil surrounding a very visible part of its economy (its high street banks). Our analysis used the Syed and Özbilgin (2009) model within a critical perspective.

In section two, we looked at the current debate on diversity management and noted the critical theorists arguments that organizational diversity programs based on mainstream diversity management approach could both obscure unequal power relations in organizations by treating differences on an individual base rather than something related to social groups (for example, Noon, 2010; Wrench, 2005). The analysis of the Brazilian case suggests that at least in circumstances such as ours, these concerns are valid. That is, that work organizations in Brazil will not adopt a diversity program aimed at including blacks (even though a number may be USA or EU subsidiaries) unless, similar to the early USA experience, they are forced to do so. There, chapter VII of the Civil Rights Act of 1964 created an environment for organizational compliance and the resulting policy intervention stimulated change in organizational routines that have had significant and lasting effects on workforce diversity (Kalev, Dobbin and Kelly, 2006). In other words, the case seems to be for harder diversity management practices rather than softer ones. Our findings contribute to diversity management literature in the sense that they offer evidence that the way Brazil has chosen to deal with the problem of managing a diverse workforce seems not work to solve racial inequalities in its workplace. Besides, by bringing the Brazilian case to the table, the paper allows the diversity management literature to be enriched by a country specific study.

The Brazilian case demonstrates the usefulness of the multilevel framework for managing diversity in bringing together for comparative purposes overlapping levels of explanation. Rather like the parts of the proverbial onion – where neither is the beginning or the end but each is part of the whole - all are relevant and, in this case, interract unfavorably for the country’s afro-descendants. As we have seen, there is great resistance to affirmative action in the country not just as a post WWII liberal proposition – as in the case of the USA – but set within a long term refusal to face a matrix of ideas that at times are close to racially inscribed social apartheid.

When the country moved out of millitary rule, the transition to democracy was not an easy process. In Brazil, as elsewhere in Latin America, democracy was at first a superficial rather than substantive institutional and action frame. As a number of scholars have commented (O’Donnel, 2001; Carvalho, 2001), the constant ruptures in latin america’s institutional frameworks have produced a setting in which, whilst some social rights had been secured largely as a result of populist governments in the 1940s and 1950s, political rights were frequently under threat and basic civil rights, including the right to have rights, often lagged considerably. Thus the evocative notion of rights can often be found being expressed emotionally in relation to health and housing but without any reaction in relation to judicial access or prison conditions. Perhaps, for these and similar reasons, including the continued pervasiveness of the ideology of “racial democracy”, where movement by personnel officers happens, it tends to follow the soft diversity approach and takes the easier route of discussing other vulnerable groups such as the disabled. But even here, it has been the legislation and the Public Prosecutor’s Office collective consent agreement that has tipped the balance.There is little sign of an effective fight against racial discrimination in the workplace. The creation of the first black college in São Paulo may make it easier to force the issue of internships for black students and those involved are hopeful on this score. But without the hard policies on occupational and workforce composition and opportunities it is not likely to succeed on its own.

The question now becomes – which way forward. Whilst the diversity frame offers space for different versions of managerial good conduct (and being a “good place to work” now gains points in business magazines), the rights based approach is more harsh and direct. As has been commented (Ignatieff, 2001) human rights are difficult to carry forward at the best of times, based as they are on a social agreement about the freedom from degradation and deal with dimensions that may be difficult for courts to assume. Outside the courts, human rights methodology is still very much based on the concept of “naming and shaming”. That is, by making explicit use of the moral high ground to make the position of those who deny rights publically uncomfortable; in our case, by turning the organizations and those who work in them object of public scorn. This might seem excessive and aggressive in the current business context, but it has been used to effect in discussing the working conditions of labor employed by the suppliers of worldwide household name brands and, when employed in the European boycott of South Africa by the many groups in the anti-apartheid movement, it certainly had an effect on both the organizations themselves and who worked in them. It could be the missing step.