FIRST EXPERIENCES OF AFFIRMATIVE ACTION FOR GRADUATE COURSES

In 1999, the first part of the quota proposal presented at the Universidade de Brasília (UnB) acknowledged that access to graduate education was even more difficult for black students and presented data on the low number of black professors in public universities such as the Universidade de São Paulo (USP) and UnB. It is noteworthy that, since the late 1990s, there was already an incipient debate about diversity in regards to access to graduate education.

The Universidade do Estado da Bahia (Uneb) was the first public higher education institution to establish, in 2002, an affirmative action policy aimed at the entry of blacks and indigenous people into graduate courses (^{CARVALHO, 2006}). This episode is very significant, seeing as, that very same year, Uneb and the Universidade do Estado do Rio de Janeiro (UERJ) became the first universities in the country to adopt affirmative action with racial criteria for access into undergraduate courses, reinforcing the hypothesis that measures for the graduate studies were born at the same time as those at the undergraduate level, although their development has taken very different paths in terms of entering the government agenda and disseminating it through the country’s institutions.

The creation of affirmative action in graduate courses was encouraged by the Fundação Carlos Chagas (FCC) and the Fundação Ford (FF) through two programs: the Programa Internacional de Bolsas de Pós-Graduação da Fundação Ford, also known as International Fellowship Program (IFP), created in 2001; and the Programa de Dotações para Mestrado em Direitos Humanos no Brasil, created in 2003. Both are considered the first diversity initiatives aimed at under-represented candidates in graduate-level courses (^{ROSEMBERG, 2013}; ^{SANTOS, 2010}; ^{UNBEHAUM; LEÃO; CARVALHO, 2014}).

However, only from 2012 onwards do more affirmative action proposals begin to emerge at this educational level. As we shall see later, the Programa de Pós-Graduação em Antropologia Social do Museu Nacional (PPGAS-MN) of the Universidade Federal do Rio de Janeiro (UFRJ) became a reference in the debate about affirmative action for graduate studies; they began discussing a policy in 2007 and approved one in 2012 (^{VENTURINI, 2017}). It is also worth noting that, to this date, only one state has laws determining that all universities maintained by the federative body must institute the system of quotas for admission to graduate courses,⁷ Rio de Janeiro.⁸

In 2015, when this study began, few programs offered any kind of affirmative action whatsoever. However, the data collected up to January 2018 indicate a total of 737 academic graduate programs with some type of affirmative action,⁹ which represents 26.4% of all programs. In other words, in four years there has been a dizzying degree of growth in these initiatives. Some policies stemmed from the decisions of the programs themselves, while others were created to comply with state laws or University Council resolutions valid for all graduate courses at a given university.

The growing diffusion of this type of policy stems mainly from the fact that many universities have approved resolutions in their university councils, thus determining that all graduate programs create affirmative action measures in their selection processes.

The data in Graph 1 indicate that between 2015 and 2018 the number of policies resulting from university resolutions increased fivefold. In 2015, there were only three universities with a resolution on affirmative action at the graduate level: Uneb (approved in 2002), Universidade Federal de Goiás (UFG) and Universidade Federal do Piauí (UFPI) (both from 2015). In 2016, two other universities passed resolutions (Universidade Federal da Paraíba (UFPB) and Universidade Federal do Amazonas (Ufam) and, in 2017, another 14 universities did so. Regarding the policies adopted by decision of the programs themselves, it is noted that there was an increase from 45 initiatives in 2017 to 134 in January 2018. The concentration of initiatives in 2017 points to the leading role of Normative Order MEC No. 13 (BRASIL, 2016), which determined that all federal institutions of higher education should send proposals for the inclusion of blacks, browns, indigenous and people with disabilities in their graduate programs.

Source: Own elaboration with research data.

GRAPH 1 NUMBER OF GRADUATE PROGRAMS ACCORDING TO THE FORMS OF INSTITUTION OF AFFIRMATIVE ACTION (2015-2018) 

TYPE OF AFFIRMATIVE ACTION

One of the main points of discussion about affirmative action is the policies’ denomination as “quotas”. This type of policy, also known as a system of seat reservation, corresponds to the allocation of a portion of the seats to students belonging to certain social and/or ethnic-racial groups. However, not all graduate programs adopt quotas.

Among the 737 graduate programs with affirmative policies analyzed, 63.9% apply the quota system exclusively, in which a percentage of the available positions is reserved for certain groups, as in the example below.

THE SEATS

A total of 08 seats are offered, with 01 seat open to people with special needs, to be filled according to the approval criteria. (UNIVERSIDADE DO ESTADO DO PARÁ - UEPA, 2017, own translation)

PPI Quotas

25% of the seats offered will be reserved for black, brown and indigenous candidates (PPI quotas), according to the order of classification of candidates who declare interest in competing for such seats. (UNIVERSIDADE FEDERAL DO ESPÍRITO SANTO - UFES, 2017, own translation)

Seats

18 (eighteen) seats are open for the Master’s in Communication. 6 (four) of these places are reserved for the ethnic-racial quota (black, brown and indigenous). If these are not filled, the reserved seats will be allocated to the remaining candidates. (UNIVERSIDADE FEDERAL FLUMINENSE - UFF, 2017, own translation)

NUMBER OF SEATS

2.1 Number of seats for candidates residing in the country (see Annex 6 for the number of seats per Research Line):

2.1.1 Doctorate: 16 (sixteen) seats;

2.1.2 Academic Masters: 28 (twenty-eight) seats.

2.2 Specific seats for foreign candidates:

2.2.1 Doctorate: 2 (two);

2.2.2 Academic Master’s Degree: 2 (two).

2.3. In accordance with a resolution approved at a meeting of October 14, 2016 by PPGHIS-UnB to establish an Affirmative Action policy, 20% of the seats will be allocated to the approved black candidates who choose to apply under the Affirmative Action policy.

2.3.1. From the total number of seats described in items 2.1 and 2.2, 4 (four) doctoral positions and 6 (six) seats of the Academic Master’s degree will be allocated to self-declared black candidates, who opt to compete for affirmative action policy of PPGHIS-UnB, according to the definitions presented in item 3.10.

(UNB, 2017, own translation)

In addition to this modality of quota, other programs have chosen to create additional seats. which are reserved for candidates who meet the necessary requirements. ¹⁰ Although they do not receive the name “quota”, it is a form of reserved seats, the only difference being that the number of seats open to all candidates is unaffected, being a milder variant of the quota system. We also note the adoption of mixed formats, in which there are quotas for certain groups and additional seats destined for other group(s) of beneficiaries.¹¹

Furthermore, in the majority of programs, all candidates, regardless of the reserved seats, are submitted to a selection process regulated by a single call for applicants, with all the admission rules and criteria. Other programs, in turn, have chosen to submit the candidates for the reserved seats to separate selection processes and regulated by different calls and with rules that differ from the ones that apply to all candidates.¹²

Finally, only one graduate program created quotas for certain groups while also offering bonuses for these candidates in certain stages of the selection process. ¹³ This shows that the bonus modality adopted previously in undergraduate courses was ignored by graduate policies, not without reason, since studies indicate that bonuses promote low rates of racial and social diversty in relation to the quotas (^{VENTURINI, 2015}, 2016; VENTURINI; FERES JÚNIOR, 2018).

Thus, it was possible to identify six types affirmative action policies in graduate programs:

Source: Own elaboration with research data.

GRAPH 2 PROPORTION OF TYPE OF AFFIRMATIVE ACTION ADOPTED (N=737) 

BENEFICIARIES

In the affirmative action policies for undergraduate courses, the main beneficiaries are students graduating from the public school system, followed by low-income candidates and ethnic-racial groups (blacks, browns and indigenous), especially in federal universities, in compliance with Law No. 12,711 (BRASIL, 2012).¹⁴

This does not occur in graduate courses with affirmative action, in which black students are the main beneficiaries; they are the target of 687 initiatives, which corresponds to 92% of the policies analyzed. Calls were identified in which only candidates that self-declare as black can request participation through affirmative action,¹⁵ with no benefits being extended to browns. These are exceptional cases, since in most programs that have ethnic-racial benefits there is an explicit reference to blacks and browns (682 initiatives). Other groups that greatly benefit from affirmative actions are the indigenous (675 initiatives) and people with disabilities (583 initiatives). To a lesser extent, there are also policies in favor of students of low socioeconomic status,¹⁶ quilombolas,¹⁷ transsexuals and transvestites¹⁸ and holders of humanitarian visas (refugees),¹⁹ as shown in Graph 3.

Source: Own elaboration with research data.

GRAPH 3 PROPORTION OF INITIATIVES ACCORDING TO THE TYPE OF BENEFICIARY (N=737) 

It is noticeable, therefore, that the main targets of graduate policies are black, brown and indigenous students, which demonstrates a change from undergraduate courses, in which social quotas have long been the majority (^{FERES JÚNIOR; DAFLON; CAMPOS, 2011}). The UFG, for example, was the second higher education institution to pass a resolution stating that all programs should create seat reservation programs. The first initiative for the adoption of ethnic-racial quotas at UFG came from the Programa de Pós-Graduação em Antropologia Social, which created its own program in August 2014. A year later, the University Council expanded the measure to the entire university (^{DINIZ FILHO et al., 2016}).²⁰ Regarding the UFG’s policy and the choice of ethno-racial criteria, the participants in the group that drafted the resolution describe that:

To that circumstance, there was no legal framework in the UFG on the subject, not in the exact sense in which it proposed the PPGAS Call for Applicants. As previously reported, it is concluded that the affirmative action policy has consistently integrated the institutional practice in the UFG for some time. Even so, two innovations were urged by the Call: the adoption of an affirmative policy in the scope of the stricto sensu Graduate Programs and the use of the ethnic-racial criterion, unrelated to other social markers, such as income level. (^{DINIZ FILHO et al., 2016}, own translation)

A preference for ethnic-racial policies is clear not only by the prevalence of these groups among the beneficiaries, but also by the justifications presented by the programs for the creation of affirmative action, since - as we shall see - many have used arguments related to Brazil’s existing racial inequalities, the low representation of these groups in graduate courses, as well as the importance of the ethnic and cultural diversity of the student body. In addition, this preference may demonstrate a greater acceptance of racial affirmative action, as well as recognition by some graduate programs that Brazil is not a “racial democracy” (^{GUIMARÃES, 2012}), that is, that there are barriers for black and brown people to reach prestigious positions in the social structure.

In some cases, the benefits for ethnic-racial groups are associated to a socioeconomic need clause, as in the policy established by State Law No. 6,914 (RIO DE JANEIRO, 2014), which regulates all state universities in Rio de Janeiro. Under the law, a maximum of 30% of the total number of seats in graduate courses should be reserved and distributed to low-income students according to the following criteria: 12% for black and indigenous graduates, 12% for graduates of the public and private higher education, and 6% for people with disabilities, as well as children of civilian and military police officers, military firefighters and prison security inspectors, who were killed or disabled due to service. However, economic need is a precondition for candidates to benefit from quotas. In other words, black and indigenous students can only benefit if they prove their dire financial situation. According to the manual prepared by the Sub-Reitoria de Pós-Graduação e Pesquisa da UERJ,²¹ students who have a monthly gross per capita income equal to or less than R$ 1,086.00 (one thousand and eighty-six reais)²² are considered financially underprivileged. It is worth nothing that UERJ considers in the calculation of per capita family income not only salaries, pensions and participation in company profits, but also the values of scholarships received by the candidates, so that students of scientific initiation and masters with a scholarship may not be considered financially underprivileged and, therefore, man not be eligible to enter via quotas in selection processes for doctorates at the university.

Another aspect that draws attention is the fact that the measures directed to graduate studies have included, among their beneficiaries, groups that had not been the target of policies for undergraduate courses, such as transsexuals and transvestites and refugees. In early 2017, UFBA approved an affirmative policy for all its graduate programs and became one of the first universities to worry about the insertion of transgender people at this educational level. In addition, the graduate programs in History of UFRGS and in Performing Arts of UnB also have measures in favor of trans people.

FORMS OF INSTITUTION AND POLICY APPROVAL

Analyzing the way affirmative policies are instituted by the graduate programs is relevant since it allows us to identify the institutions that created these policies by their own initiative and those that were forced to do so. The origin of the initiative indicates whether there was a real desire for change and transformation within the institution, as well as being directly related to the type of institutional change promoted by decision makers.

The data collected allowed us to identify four ways in which policies were instituted and approved:

As mentioned above, the data collected indicate that, until January 2018, only 18.3% of the programs created measures by their own initiative,²³ while the vast majority did so as a result of university council resolutions.

Source: Own elaboration with research data.

GRAPH 4 PROPORTION OF GRADUATE PROGRAMS WITH AFFIRMATIVE ACTION ACCORDING TO THE FORM OF INSTITUTION (N=737)²⁴  

Among the policies instituted at the initiative of the programs, there are also cases in which an institute composed of several graduate programs approved affirmative action, but each program had the autonomy to define the type of policy to be adopted, the beneficiaries and the adjustments to the selection calls. This is the case, for example, of the Instituto de Filosofia e Ciências Humanas (IFCH) of Unicamp, which approved the principle of affirmative action at a collegiate meeting held on March 11, 2015 (^{SILVA, 2016}), but the specification of the policy and its execution was within the scope of autonomy of each graduate program.²⁵

Until the cut-off date of the present research (January 2018), quota policies were identified in graduate studies approved by the university councils of 19 public universities:

UFG was the second institution to approve a resolution determining that all programs should create this type of policy, and that was more than a decade after the approval of the quota program for graduate students at Uneb. According to an article published by actors involved in the policy formulation process, the first initiative to adopt ethnic-racial quotas in the UFG started from the Programa de Pós-Graduação em Antropologia Social in August 2014, which was later extended to the university as a whole and approved in 2015 (^{DINIZ FILHO et al., 2016}).⁴⁴

Although it does not make affirmative policies in graduate courses compulsory, it should be noted that the Normative Order 13 issued by Ministério da Educação (MEC) (BRASIL, 2016) has been mentioned as an incentive for several programs and has contributed to the creation of such measures by several programs. It is a fact that the number of programs with seat reservations increased significantly in the last two years.

It is also possible that the final report of the working group instituted by Capes and MEC in 2015 to discuss the topic of inclusion in graduate studies had the effect of inducing programs to adopt measures, as it recommends that Capes use pro-diversity measures as criteria in its four-year evaluation of graduate programs.

CAPES GRADE

One of the main aspects in the debate about the creation of affirmative action policies for graduate courses is a concern with academic excellence. At universities such as USP and Unicamp, during the discussions about the institution of these measures, objections were raised regarding potential decrease in quality and excellence due to the creation of quotas (^{DE FIORI et al., 2017}; ^{SILVA, 2016}).

Although it is not possible to say with certainty that the change of criteria in admissions would result in a reduction of excellence, there are concerns about impacts on the programs’ quality and Capes evaluation, due to the supposed need to extend the deadlines for defending theses and dissertations, student achievement and the impact of their publications, among others (^{DE FIORI et al., 2017}).

Therefore, the programs with the best evaluations could present greater resistance to the creation of affirmative action policies due to the concern with their quality and the evaluation made by Capes. A study done on the state of affirmative action policies for undergraduate courses before Federal Law Nº 12,711 (BRASIL, 2012) came into operation shows that the universities with the best performances in the Índice Geral de Cursos (IGC)⁴⁵ were those that promoted less diversity (^{FERES JÚNIOR et al., 2013}). Thus, it is important to analyze whether policies are also being instituted by well-evaluated graduate programs.

To this end, we crossed affirmative policies with the grade obtained by the graduate programs in the last Capes four-year evaluation (Graph 5). Most of the programs that created affirmative action have grades 3 (30%) and 4 (41%), while only 11.5% of the initiatives implemented by programs with grades 6 (7.6%) and 7 (3.9%), those with the highest standard of excellence according to Capes. In other words, the trend of greater resistance to these policies by the programs with better evaluations was confirmed in our study.

Source: Own elaboration with research data.

GRAPH 5 DISTRIBUTION OF AFFIRMATIVE ACTION PROGRAMS ACCORDING TO CAPES EVALUATION GRADE (N=737)  

If we analyze the distribution of the initiatives taking into account the proportion of programs with affirmative action in relation to the total number of programs in each Capes score, we obtain a result very similar to the previous graph.⁴⁶ That is, totaling the proportions by category (100%), we control for the number of programs in each one (Graph 6).

Source: Own elaboration with research data.

GRAPH 6 PROPORTION OF PROGRAMS BY AFFIRMATIVE ACTION POLICIES AND CAPES EVALUATION GRADE (N=2,763)  

In the distribution of the forms of institution in each of the evaluation grades (Graph 7), it is possible to verify that within each highest grade, there is a greater proportion of programs that have created affirmative action by their own initiative. That is, 31.6% of programs graded 6 and 24.1% of programs graded 7 that created affirmative action did so by their own initiatives. In any case, the data indicate that in all grades, most programs created the measures by determination of university councils or state laws.

Source: Own elaboration with research data.

GRAPH 7 PROPORTION OF PROGRAMS BY CAPES EVALUATION GRADE AND FORM OF INSTITUTION (N=737)  

UNIVERSITY PROFILES

With regard to the profiles of the universities that adopted affirmative action, there is a predominance of federal universities with 85.9% of the identified policies, followed by state universities with 14.1%. There were no municipal university graduate programs with affirmative action policies.

If we analyze the distribution of programs within each type of public university according to whether or not affirmative action policies are created (Graph 8), the predominance of federal universities in relation to state universities is even clearer, since they have more than twice the number of programs with these measures. The greater adhesion of the federal universities is directly linked to the edition of Normative Order No. 13 (BRASIL, 2016) and to the fact that several universities have approved resolutions in their university councils.

Source: Own elaboration with research data.

GRAPH 8 PROPORTION OF PROGRAMS ACCORDING TO TYPE OF PUBLIC UNIVERSITY AND CREATION OF AFFIRMATIVE ACTION (N = 2,763)  

AREAS OF KNOWLEDGE

Studies show that although access to the university has been democratized, and there has been a significant increase in the number of blacks and browns in higher education, this varied greatly for different careers: areas less valued by the labor market, such as the Humanities and Applied Social Sciences, become more diverse than the so-called “hard” areas (^{RIBEIRO; SCHLEGEL, 2015}). Thus, we analyzed the distribution of affirmative policies to graduate courses in the different areas of knowledge, according to the way Capes defines them: Exact and Earth Sciences,⁴⁷ Biological Sciences,⁴⁸ Engineering,⁴⁹ Health Sciences,⁵⁰ Agricultural Sciences,⁵¹ Applied Social Sciences,⁵² Humanities,⁵³ Linguistics, Literature and Arts⁵⁴ and Multidisciplinary Studies.⁵⁵

Among the 737 programs that created affirmative action measures in the analyzed period, we found that more than one fifth of the programs are in the field of Human Sciences (Graph 9). Next come the areas of Multidisciplinary Studies, Health Sciences and Applied Social Sciences.

Source: Own elaboration with research data.

GRAPH 9 PROPORTION OF PROGRAMS THAT CREATED AFFIRMATIVE ACTION DISTRIBUTED BY AREA OF KNOWLEDGE (N=737)  

The graph with absolute numbers does not allow us to compare the different areas because the number of programs in each area is different. This is resolved in the chart below, where we can compare proportions. When analyzing the proportion of graduate programs in each area of knowledge that have affirmative action policies (Graph 10), regardless of their form of institution, we can see that the areas of Human Sciences and Applied Social Sciences are those that have the greatest number of programs with affirmative action. The Engineering area has the lowest proportion of programs with affirmative action measures.

Source: Own elaboration with research data.

GRAPH 10 PROPORTION OF PROGRAMS BY AREA OF KNOWLEDGE AND CREATION OF AFFIRMATIVE ACTION POLICIES (N=2,763) 

With regard to the programs that instituted affirmative action by their own initiative, disregarding those that created their policies by legal determination or university council resolutions (Graph 11), the areas of Human Sciences and Applied Social Sciences continue at the forefront (50% and 20.1%, respectively).

Source: Own elaboration with research data.

GRAPH 11 PROPORTION OF PROGRAMS THAT CREATED AFFIRMATIVE ACTION POLICIES BY THEIR OWN INITIATIVE BY THE AREA OF KNOWLEDGE (N=134)  

In the distribution of the forms of institution of affirmative action within each area of knowledge (Graph 12), it is possible to verify that the Engineering programs only created these measures as a consequence of resolutions of university councils (16.4%), and of state laws (83.6%); that is, not a single program in this area created affirmative action of their own accord. They only did so when required by their universities or by law. This is probably due to a culture of meritocratic conservatism present in Engineering that results in greater resistance to changing selection processes and to the creation of affirmative action policies.

In the area of Human Sciences, in turn, 52% of affirmative action policies were created in compliance with resolutions of university councils, 39.2% by their own initiative, and 8.8% as a result of state law. Thus, this area presents the highest proportions of autonomously created measures, both in terms of distribution of programs with this form of institution (50%) and within its field (39.2%). The predominance of humanities (Humanities and Applied Social Sciences) seems to be related to their objects of study and to the proximity of issues related to vulnerable groups, such as social stratification, the study of indigenous languages, the distribution of groups in the city, rates of violence and vulnerability, among others.

Source: Own elaboration with research data.

GRAPH 12 PROPORTION OF PROGRAMS BY AREA OF KNOWLEDGE AND FORM OF INSTITUTION (N=747)  

REGIONAL DISTRIBUTION

With regard to the distribution of policies across the national territory, if we analyze only the 737 programs that created affirmative actions, the Southeast and Northeast regions present the largest proportion of initiatives, which is related to the fact that more than 60% of universities with resolution on the issue are found in these regions. The South region has the lowest proportion, with only 8.3% of affirmative action policies.

Source: Own elaboration with research data

GRAPH 13 PROPORTION OF PROGRAMS THAT CREATED AFFIRMATIVE ACTION POLICIES BY REGION OF THE COUNTRY (N=737) 

When we control for the number of programs, using percentages, the North region appears as the one with the highest proportion of programs with affirmative action initiative (52.8%), followed by the Midwest region (39.5%). The region with the fewest initiatives is the South, where only 11.4% of the programs have created affirmative action policies. It is worth noting that the Southeast - the largest region in terms of number of graduate programs - has twice as many initiatives as the South (23.7%) and only half as many as the North (52.8%).

Source: Own elaboration with research data.

GRAPH 14 PROPORTION OF PROGRAMS BY AFFIRMATIVE ACTION POLICIES AND REGION OF THE COUNTRY (N=2,763)  

If we consider only the programs that instituted affirmative action measure by their own initiative, disregarding those that were compelled to do so by legal determination or university councils, the Southeast region also concentrates the largest number of initiatives (37.3%), diametrically opposed to what happened with the quotas in the undergraduate courses until at least 2012, since the Southeast was the region showing most resistance then (^{FERES JÚNIOR; DAFLON; CAMPOS, 2011}).

Source: Own elaboration with research data.

GRAPH 15 PROPORTION OF PROGRAMS THAT CREATED AFFIRMATIVE ACTION BY THEIR OWN INITIATIVE BY REGION OF THE COUNTRY (N=134)  

In the distribution of initiatives by states, there are those in which all graduate programs in the state have adopted affirmative action, such as Acre, Amapá, Sergipe and Tocantins. However, in the case of Acre and Amapá, these are states with a very small number of graduate programs - 9 and 4, respectively.

On the other hand, in Maranhão there is no program with this type of policy. Then, the states with the highest proportion of affirmative action programs are Amazonas (81.1%), Goiás (86.8%), Piauí (94.1%) and Roraima (75%).

In numerical terms, the states with the greatest number of policies are Minas Gerais (151 programs), Bahia (77 programs), and Rio de Janeiro (100 programs). The smallest numbers are in Pernambuco (2 programs) and in Santa Catarina (3 programs).

Source: Own elaboration with research data.

MAP 1 PROPORTION OF AFFIRMATIVE ACTION PROGRAMS BY STATE (N=2,763)  

The state of São Paulo concentrates one of the largest numbers of affirmative policies created by the graduate programs themselves in the Southeast region, with 20 initiatives, surpassed only by Rio de Janeiro, with 22 initiatives.

Source: Own elaboration with research data.

GRAPH 16 PROPORTION OF PROGRAMS THAT CREATED AFFIRMATIVE ACTION BY THEIR OWN INITIATIVE BY STATE (N=134)  

Paradoxically, to this day, São Paulo is the state that has most resisted the implementation of racial quotas in undergraduate education. Just consider the inefficiency of the bonus programs implemented by Unicamp (^{VENTURINI, 2016}; VENTURINI; FERES JÚNIOR, 2018) and USP (VENTURINI, 2015), or the enormous delay with which Unesp, Unicamp and USP adopted racial quotas. Furthermore, when we examine the proportion of programs with affirmative action policies, as in Map 1, we find that São Paulo does indeed have a small percentage, 3.6%, a similar quantity to the states of Paraná (3.2%) and Santa Catarina (3.2%).

FINAL CONSIDERATIONS

The work sought to present an overview of the affirmative action policies created in graduate courses in the period from 2002 to the beginning of 2018. In general, it is possible to observe that in a universe of 2,763 academic graduate programs (master’s and doctorates) in public universities, more than a quarter of the programs (26.4%) had already instituted some kind of affirmative action measure by January 2018. The data collected point to a significant diffusion of this type of policy in the last four years, especially due to the approval of resolutions by higher bodies within the universities determining the creation of these measures by all graduate programs. The diffusion of these policies as of 2017 and the predominance of federal universities point to the inducing role of Normative Order MEC No. 13 (BRASIL, 2016) and its interpretation as mandatory by programs and universities.

In general, we observe great variation in the format of the policies adopted by graduate programs, especially in relation to the criteria adopted in the selection process, the modalities and the social groups benefited.

The mode of reserved seats predominates, whether as a percentage of places open to all candidates or as extra seats, while the bonus modality, previously adopted in undergraduate studies, was ignored in graduate policies. There are also mixed modalities and the division of selection processes into different calls for proposals for certain target groups.

Unlike undergraduate courses, in which the main beneficiaries are students graduating from the public school system, black, brown, indigenous and disabled beneficiaries prevail in graduate policies, in addition to the inclusion of new groups, such as transsexuals and transvestites and humanitarian visa holders (refugees).

With regard to the form of institution of policies, the study identified 19 public universities that passed resolutions on their university councils determining that all graduate programs must adopt affirmative action measures. Only 18.3% of the programs created measures by their own initiative and only one state of the Federation passed a state law regulating the measures (Rio de Janeiro).

Due to concerns about the impact on the quality of programs and the consequent evaluation of Capes, the best evaluated programs could be more resistant to the creation of these policies. However, the data indicate that there are programs with grades 6 and 7 instituting affirmative action, which can be used as an argument in the discussions of programs that are still in the analysis and approval phases. The areas of Human Sciences and Applied Social Sciences are the ones with the largest number of graduate programs with affirmative action policies. The data points to a greater resistance of programs in harder areas, especially Engineering, which only created this type of policy to comply with state law or university resolutions.

Finally, the Southeast region stands out in the regional distribution of programs with affirmative action, especially those that created measures by their own initiative. The proportions in the Southeast region are significant, since the state of São Paulo had been resistant to the implementation of racial quotas for many years. Furthermore, it is interesting to note that there is a state - Maranhão - in which a graduate program has yet to institute this type of policy.

The very expressive and rapid growth of affirmative action in graduate school is, at least in part, a consequence of the success of affirmative action policies in undergraduate courses, created since 2003, that is, practically ten years earlier, despite the extemporaneity of UnB initiative. Certainly, these two waves of diversity initiatives are the product of a greater recognition by Brazilian higher education institutions of the problem of reproducing social inequality via access to higher education. That is, until the creation of affirmative action policies, Brazilian public higher education served primarily the white middle class. This social and racial segregation was particularly acute in the most popular courses. The quota policy transformed the profile of higher education, promoting intense social and racial diversity (ASSOCIAÇÃO NACIONAL DOS DIRIGENTES DAS INSTITUIÇÕES FEDERAIS DE ENSINO SUPERIOR; FÓRUM NACIONAL DE PRÓ-REITORES DE ASSUNTOS ESTUDANTIS, 2019). It is clear that, as we have shown, some areas of higher education, such as engineering, are quite averse to policies of this nature, but even these are often forced to adopt them by higher standards from universities or even state or federal laws.

It was not only public universities that opened up to diversity, as the data from our study show, state legislatures have played an important role in creating these diversity initiatives. They were crucial for the advent of affirmative action in undergraduate courses and have also played a role in initiatives for graduate students. In other words, the recognition of the problem and the willingness to try out solutions goes beyond the walls of the university and penetrates the political system. It remains to be seen whether with the conservative wave that swept the country in the last elections - when political forces historically averse to diversity won the presidency, several state governments and won a majority in the two federal legislative houses - this trend of increasing diversity in Brazilian higher education will persist.