Trust in the Affirmative: Examining Relationships of Social Trust, Race, and Affirmative Action in Brazil


Brazil’s origin story began with roots of miscegenation and developed into what scholar Gilberto Freyre calls the “Racial Democracy” where, rather than a plurality of races composing the country, the only race is “Brazilian.”  Race has nevertheless played a crucial role in Brazil, stratifying the population and disproportionally marginalizing those of darker skin to the outskirts of political, economic, and social activity. At the root of these interactions is social trust, or one’s ability to trust in other people in one’s community or society. The capacity of one who is societally marginalized to trust would probably be lower than that of someone who is not marginalized. It is within this conundrum that I focus my research. To what extent is social trust affected by racial marginalization in Brazil and how are levels of social trust affected by corrective legislation regarding racial marginalization (Affirmative Action programs)? Using three ordered logit regression tests, I examine relationships between racial marginalization, social trust, and Affirmative Action. The results show that racism has a strong impact on general social trust. This enhances Putnam’s work on social trust and diversity, offering racism as an explanation of this negative relationship. The tests also reflect that Affirmative Action may not be the necessary solver of problems. This may be the result of elite hegemonic ideology as well as vestiges of distrust from the previous authoritarian regime.

Table of Contents: 


    The 21st century has brought a new “post-racial” lens to the United States that attempts to mark the end of racism in the country. This idea, however, is not unique to the United States. It has long been the status quo for many Latin American countries, particularly Brazil.  Brazil’s origin story has its roots in miscegenation and developed into what scholar Gilberto Freyre calls the “Racial Democracy” where, rather than a plurality of races comprising the country, the only race is ‘Brazilian.”  Race has played a crucial role in Brazil, stratifying the population and disproportionally marginalizing those of darker skin to the outskirts of political, economic, and social activity. The Afro-Brazilian movement, through trans-national conferences and pressuring the state, was successful in implementing a variety of affirmative action programs in the country to begin addressing and correcting racial imbalances.

    Resting beneath these race dynamics, however, are notions of democracy and citizenship. When citizens can be unequally treated and restricted from their social and political rights, neither “Racial Democracy” nor “democracy” has much meaning. Implicit in how a democratic political system functions are interpersonal relationships. Social capital, as defined by scholar Robert Putnam, is composed of networks and active groups that strengthen ties to each other and to the country. At the root of these interactions is social trust, or one’s ability to trust in other people in their community or society. Intuitively, we understand that social capital cannot function efficiently without workable levels of social trust. Furthermore, the capacity of one who is societally marginalized to trust would probably be lower than that of someone who is not marginalized.

    It is within this conundrum that I focus my research. To what extent is social trust affected by racial marginalization in Brazil and how are levels of social trust affected by corrective legislation regarding racial marginalization (affirmative action programs)?

    These questions seem particularly pertinent in contemporary Brazilian politics. Affirmative action programs themselves are a recent phenomenon in Brazil, having been implemented in 2001. Discussions of how race operates and the extent to which racism affects Brazil are up and coming; perhaps even timelier are questions of the role of government in addressing social welfare. Recent protests just last month, called by various names including “The V for Vinegar Movement,” the “Salad Revolt,” and “Brazilian Spring,” indicate civil unrest and general societal mistrust. The protests, while sparked by an increase in bus fares, speak more towards the government’s handling of state resources and the lack of attention to those at the lower ends of society economically. This seems particularly true of those racially marginalized in the protest, who were demanding more government assistance for various socio-economic disparities. Clearly more investigation of issues of politics, economics, and society are needed as they pertain to social trust, race, and democracy. Without more knowledge, getting to the root problems of these protests would be impossible and they could lead to more dire consequences.

    This paper will first examine pertinent literature and history as they relate to these questions in order to allow for a more nuanced view of how race, social trust, and affirmative action have been studied generally, but also in the context of Brazil. It will then move to examine similar case studies and what their results were as they applied social trust and race. The paper will then provide a clear theory in how these concepts might function when tested, as well as a design in which to actually test their relationship. Finally, we will examine the results of the test and draw conclusions about the relationship, if any, of social trust and racial marginalization.

    Literature Review

    Brazilian Racial Marginalization

    Much of the literature points to a societal norm or recognition of a national Brazilian identity that overrides notions of race. This is based on the notion of racial democracy that insists upon a unified Brazilian identity that is composed of a mixture of peoples from African origin, European origin, and indigenous Americans (Freyre 1986). Hence, the dominant ideology held by both many white and nonwhite Brazilians is that racism does not exist, particularly since there was no political history of deliberate discrimination.

    Literature points towards Latin America being a pigmentocracy, or an ordering of people based on phenotype and cultural differences. Studies suggest that there is a preference as well as inequality when it comes to skin color in Latin American countries, where those of lighter skin benefit more than those with darker skin (Bonilla-Silva and Dietrich 2009; Telles and Steele 2012). This dynamic exists just the same for Brazil specifically. Statistics reveal various race-based inequalities and scholars elucidate on both the social and political exclusion of Afro-Brazilians (Hanchard 1999; Telles 2009). This institutionalized exclusion ranges from lacking representation in universities and elite programs like medicine or law with only two percent of students being Black, as well as harsh racially discriminatory processes suffered by Black people (Beato 2004; Hernández 2013). The literature is missing assessments of the levels of trust among these marginalized groups and if race is a specific factor in levels of political participation. This makes sense, given the hegemonic ideology that reverberates ignorance of racial marginalization.

    Social Trust
    Many scholars examine the effect of social capital, or citizen engagement in community affairs, on the performance of government and other social institutions. The facets of social capital include networks, norms, and trust. The theory behind social capital assumes a stronger relationship with people within one’s community via networks, norms, and trust. That means they will be more involved in the welfare of each other and hence will participate more politically. Indeed, even controlling for other variables, levels of social capital are correlated with levels of group membership (Putnam 1993, 1995). Parsing out social capital and looking specifically at social trust, trust also results in high levels of political participation, particularly for racial minorities. One is more likely to participate in civic life if the community as a whole is more trusting of people regardless of their differences (Uslaner 2004). High levels of economic inequality contribute to lower levels of trust, which decrease political support and participation, both of which would be components necessary to help alleviate these problems (Rothstein and Uslaner 2005).

    Even the most critical theories regarding the relationship do not deny that higher levels of social capital and social trust create higher levels of political confidence and participation. Newton (2001) attempted to disprove the connection and only showed that 1) the relationship is true on the aggregate levels and does not translate to individuals, and that 2) while there is a relationship, high levels of social capital do not necessarily infer high levels of political capital. Furthermore, assuming that repression would decrease levels of trust, we also find that repression and political participation are related insofar as reducing repression opens up opportunities for and generally increases citizen political participation (Booth and Richard 1996). This means that repression of any form would decrease trust and hence participation.

    Social Trust and Race/Ethnicity
    Often, areas with high levels of ethnic diversity have lower levels of social trust (Putnam 2007). In making this claim, Putnam conflates ethnic diversity with immigration trends, noting that “diversity and immigration are not identical, and in our subsequent, more detailed analyses we will need to make that distinction more explicit and rigorous” (Putnam 2007, pg. 140). Diversity is treated as if it exists in a vacuum and that societal racial treatment does not play a role in how diversity and trust are interrelated. Putnam lists the many benefits of diversity in the long run such as stimulating creativity and rapid economic growth, but never questions why, with such benefits, trust is inhibited by diversity. The answer, it seems to me, is that racism still plays a role in social, political, and institutional interactions that would naturally foster mistrust of different people despite being in constant contact with them.

    Using modes of policy and associational life, however, can begin to correct the issues of racial marginalization and mistrust (Andrews 2008). Policy decisions and programming can begin building foundations of social trust and racial heterogeneity in correcting the racist institutions that govern our everyday interactions.

    Affirmative Action

    Many studies regarding racial justice and affirmative action are often reduced to issues of diversity. The point of affirmative action programs is not merely to create a more diverse environment, however, but rather to begin restorative justice. Diversity itself is not a discrete category where identity groups often intersect. Affirmative action, in fact, begins “from the implied premise that there is an injustice or an inequality that needs to be remedied, such as sexism, racism, homophobia, or disablism: ‘diversity’ obscures the issue of inequality which is at the heart of the matter” (Thornton 2001; pg. 91).

    A study found that Affirmative Action students in the United States perform slightly worse than regular students in terms of GPA and graduation than non-Affirmative Action admitted students (Rose 2005). This statistic further shows the ramifications of restorative justice that starts at higher education rather than primary education, or perhaps lacking institutional support for these students outside of simply affirmative action. Nevertheless, the point of the programming is to begin correcting injustices and while diversity is a benefit, it should not be the primary benefit or focus. This also helps to explain why support for affirmative action programs is often cited to be higher than what individuals actually feel, as despite the corrective justice it performs, it does seem as though more assistance is given to these students than others (Boven 2000). Affirmative action does operate on an unbalanced field, but rightfully so, due to previous historical imbalances. Furthermore, no study has been done that tests the relationship of affirmative action on social trust. If positive, the relationship could further legitimize affirmative action programming.

    Brazilian Affirmative Action

    While studies of affirmative action have been done, they are mostly centered in studies of the United States. Brazilian affirmative action is different in a number of ways: it uses distinct quotas for certain groups (Blacks, public school students, etc.), University admittance is based on an entrance exam, and distinguishing between who is Black and who is not is a legitimate process versus one predetermined by self-identification The programs mandated are based on the “constitutional principle of quality, which mandates compensation for the past discrimination that created current racial inequalities” (Da Silva Martins et. al 2004). This fact helps explicate why quotas are more supported in Brazil than in the United States, with 70 percent of the population agreeing with their implementation in universities (Smith 2010). The identification of race is particularly difficult in that, since race operates as more of a category than a distinct group identity per say, explicit identification becomes exceedingly difficult (Tavalaro 2008). Hence, the programs themselves create race (and subsequent racism) as a “political fact” rather than a belief or foreign ideology. As a political fact in this regard, issues of trust come into play both with an individual’s relationship to the state in recognizing and fixing this political fact as well as among individuals to cope with this political fact. Despite knowing levels of support for affirmative action, no literature exists that examines the relationship between affirmative action and successfully diminishing perceptions of racial marginalization nor its effect on general social trust, or even social capital, especially in the Brazilian context. How is trust affected by a system that both fosters restorative racial justice while also creating a system of imbalance among races competing for slots in a university?

    Case Studies/Impetus for research and examination

    To better assess how to successfully examine social trust and race marginalization, it is imperative to look at other examples of countries where social trust, racial marginalization, and democracy all clash. Two countries come to mind: 1) The United States and 2) South Africa. Both of these nations have had distinct issues of racial marginalization, notably in the forms of Jim Crow segregation in the United States and Apartheid in South Africa. Hence, examining these cases will illuminate if there have been previous connections of social trust and racial marginalization and what exactly that relationship is.

    Negative racial attitudes towards those of darker skin have existed throughout the history of this country, from its earliest colonization.  The institution of slavery began long before the United States of America even existed; therefore the country was born with racial animus. Even as slavery ended, social negativity towards all racial minorities but particularly African-Americans persisted. The doctrine of Separate but Equal began in 1896 with the Plessey v. Ferguson case, allowing for Jim Crow segregation laws to rule legally until 1955 (and illegally in many states until the 1970s). Despite various glass ceiling breakages, jarring disproportionate treatment of African Americans still exists in the United States even today, notably with the peculiar mass incarceration of Black bodies that heavily imbalance other races despite similar tendencies to commit crimes among all races.

    Looking past specific historical events that would explain why there would reasonably be lower rates of social trust among racial minorities, Eric Uslaner has previously conducted empirical studies testing the question of social trust, political participation, and race (Uslaner 2004). He found that states with higher levels of trust had higher levels of political participation of African Americans in terms of writing letters to the editor and to Members of Congress, as well as joining political associations and making public speeches. He also found that higher levels of trust reduced negative disproportionate treatment, finding that higher levels of general social trust in states had lower levels of minority suspension rates in schools, lower levels of emotional disturbances among African Americans, and even lower poverty rates. Not surprisingly, the racial history of the United States creates a distrustful environment. Higher levels of trust, however, seem to reduce disproportionate treatment and spur higher levels of political participation.

    South Africa has a similarly interesting history of racial disenfranchisement. Colonization efforts by European powers, including the Portuguese and Dutch began in 1652, but increased in focus at the beginning of the 19th century. This occupation already connotes an imbalanced relationship ethnically, where European powers have a more dominating relationship over the South Africans. Their system of racial segregation was called Apartheid, which lasted from 1948 to 1994, equally reasonably spurring massive distrust with not just ethnic differences but also racial differences.

    Similar tests of attitude and trust have been done regarding South Africa as well. John Duckitt and Thobi Mphuthing noted the massive socioeconomic inequalities with white South Africans earning on average 10 times as much as black South Africans (Duckitt and Mphuthing 1998). Looking specifically at trust, they found that before the election of Nelson Mandela to the presidency, an election that marked both the end of Apartheid and the end of white minority rule, levels of hostility and distrust were high among black South Africans towards white South Africans. Testing after the election, they found socioeconomic positioning and levels of trust among black South Africans much higher than they had been previously, based on surveys. They note that in such a short period, it is very unlikely that real changes occurred so drastically, but they note that it involves the massive change in attitude from the dynamic paradigm shift. Once more in this case, levels of social trust and racial marginalization seem to be related and have a greater impact on political and economic changes nationally.

    Slavery had a pernicious effect in Brazil that has resulted in a lasting form of racial marginalization that operates more socially and spatially than politically. While Brazil has not explicitly condemned those of darker skin through legislation, the toxic attitudes regarding Afro-Brazilians nevertheless have critical ramifications politically, economically, and socially. Neither of the other cases have Brazil’s lack of legal precedent for racial oppression, nor is there such a vehement denial of racism in the cases. Furthermore, Brazil has high levels of racial mixing from past miscegenation, so it falls directly in the middle of the spectrum of race marginalization based on the percentage of the population. By that I mean that the United States racially oppressed with a white majority ruling while South Africa racially oppressed with a white minority ruling, even though both were punitive against those of the African origin. Another interesting aspect to consider is that Brazil already has incredibly low general social trust and I wonder how possible it is to discern trust regarding marginalization from the other factors that may cause distrust.


    Hypothesis1: A decrease in racial marginalization will lead to an increase in social trust.

    Hypothesis2: Corrective legislation [notably affirmative action programs to decrease racial marginalization] will increase social trust.


    Racial marginalization creates a toxic environment for one to live in effectively. This type of marginalization in Brazil restricts those of darker skin spatially to poorer zones or regions, creates a large socioeconomic gap that disproportionally affects those of darker skin, and denies those of darker skin proper political representation or acknowledgement of their political, economic and social needs. This is made worse by the vehement refusal to note that there is racism occurring within the boundaries of Brazil. This is in sharp contrast to other countries where racism was supported through oppressive legislation as well as oppressive judicial and executive rulings. The United States with its long history of slavery and Jim Crow laws and South Africa with punitive colonization and Apartheid come to mind in these regards. Brazil, in comparison, while abolishing slavery much later than most other countries (in 1888 via the Golden Law), never passed any laws to explicitly target Afro Brazilians. Racism persisted because of the lack of corrective legislation and acknowledgement of past errors, though, and no foundation or reconstruction was set to adjust the socioeconomic standing of former slaves.

    Naturally, it seems easy to envision how this type of marginalization would affect levels of social trust in the country. High levels of economic inequality contribute to lower levels of trust, which decrease political support and participation, both of which would be components necessary to help alleviate these problems. Repression tactics decrease levels of trust and inhibit proper interaction among communities with each other as well as with the state. Therefore, I posit that decreasing (or eliminating) racial marginalization will increase levels of social trust. This can be done through usage of institutions and programs that begin to address the negative vestiges of slavery. These restorative measures can decrease socioeconomic gaps as well as give faith to marginalized groups that the state recognizes their concerns, resulting in mutual recognition of legitimacy, both of which increase levels of trust. This assumes that the programs are successful in decreasing racial marginalization, however. But if they are, then this logic should follow.

    Examining the component of social trust is crucial in that cases have shown how social trust plays a part generally in social capital, but leads to a more inclusive community. Intuitively, if one trusts someone, they will be more willing to work with that person. The notion of community is integral in that this is how political participation is promoted as well as notions of inclusion in citizenship and the legitimacy of democracy. A democracy is built upon each individual having an active role to play in shaping the politics of their nation, and this cannot happen when a group of people are systematically excluded from the political process and are socially marginalized. Furthermore, all of the benefits of citizenship cannot be attained with racial marginalization, including access to governmental resources or accurate political representation. While this case focuses on Brazil, it may nevertheless provide another example of the effects of social trust more broadly.

    Research Design

    Spatial and Temporal Domains

    Rather than generalize about race, I focus on one country in particular because race tends to operate uniquely in countries due to the level of its entrenchment in both history and society. This study specifically examines racial marginalization as it occurs in Brazil, beginning from the implementation of affirmative action quotas in public universities from 2000 to 2012. Brazil is ironically simultaneously self-perceived as a racial democracy due to the multiracialism of the population, but also as a country that reproduces a style of racism that is more akin to colorism or pigmentism. In other words, it is a country that denies the existence of racism yet still faces its harsh consequences. This country is important to look at in whole for this reason, rather than regionally, because a diversity of races or people of color exist all over the country, even though large clusters of Afro-Brazilians tend to live in the Northeastern part of the country. Furthermore, the temporal domain is crucial because it includes information from immediately before the period when affirmative action quotas were put into place until none of their usage was maintained. My unit of analysis is the impact of race marginalization and affirmative action on general social trust.

    Data Source

    World Values Survey is a large collection of surveys asking a variety of questions pertaining to social values. This dataset seemed particularly fruitful to use in regards to questions of race, trust, and the role of government since these are all essentially value-based criteria. WVS was also the only dataset that included all of this data together, including each of the control variables. There were a total of 4431 survey results from 1981 to 2008 for Brazil, a good sample size to test from to more accurately test each variable.

    The survey dataset LAPOP does deal more directly with matters of race and skin color in its questions, and it also has questions related to trust.  I chose not to use LAPOP, however, to gauge a longer time period, gather better proxy questions that involve social trust, race, and government programming, and include all of the control variables I sought to test. Furthermore, direct engagement with the questions may not necessarily lead to better results since race has often been an elusive category to quantify in the first place, so a larger sample size may prove more useful than a more specific question.

    Dependent Variable

    General Social Trust. The study aims to measure the impact of racial marginalization on social trust and the effect of affirmative action programs that work to decrease racial disparities. As such, my dependent variable is general social trust. I look at general social trust as opposed to social trust of particular races or groups of people because it speaks more to the immense diversity of Brazil in addition to relationships among the population overall. To test general social trust, I operationalize the WVS survey responses to question V24, asking respondents, “Generally speaking, would you say that most people can be trusted or that you need to be very explicitly and generally asking respondents their view of the trustworthiness of their fellow citizens?”

    Independent Variables

    Racial marginalization. Racial marginalization is how I define the phenomenon of institutional racism via color preference in Brazil. As a result of the social racial hierarchy, those of darker skin have been forced to the periphery politically, economically, and socially, with increased barriers to political participation, economic mobility, and social interactions. Quantifying this phenomenon, however, is a difficult task since this often occurs without being recognized, as evidenced by the concept of Racial Democracy. A pertinent question from the survey to test race marginalization is question V37, asking to indicate “People of a different race” as those whom they “would not like to have as a neighbor.” Because race marginalization often occurs implicitly rather than explicitly, it is difficult to achieve a truly accurate response that measures total effects or sentiments of racism. This question comes closest, however, in establishing the social dynamic of racism. It makes intuitive sense for one to not want to be in the same neighborhood or community with people considered to be undesirable. Hence, this question determines negative racial sentiments in a manner that does not directly conjure explicit racial disdain. Still, as previously noted, notions of race are still cloudy and this question may not translate well to a society that does not completely acknowledge race as a concept yet.

    Affirmative action implementation. Ample data on the effects of affirmative action over time has not been collected yet, especially in the case of Brazil. There are indeed studies of approval ratings of Brazilian affirmative action, which will be used in the analysis, but this data is limited in scope (Smith 2010). Rather, I test this variable through questions of governmental programs aimed to provide for the general welfare of the state as a proxy. The question that best addresses this is V98, asking on a 1 to 10 scale how they place their views, with low values indicating, “Government should take more responsibility to ensure that everyone is provided for,” and high values indicating, “People should take more responsibility to provide for themselves.” This question clearly has flaws in the sense of neither directly addressing affirmative action nor welfare programs in general. It is also contingent on the ideology of respondents rather than the effectiveness of the governmental programs. The question does relate to opinions of government, however, and may be able to gauge, to a degree, favor in affirmative action over time if it betters general social trust. The issue is that since this question is so generalized, there may be an averaging effect of governmental competency in that bad governance could decrease sentiments of government even if good programming is implemented.

    Control Variables

    A number of other factors could affect how social trust is generated in the context of Brazil in regards to race. Notably, I include age (age of respondents), income (by income bracket of respondents), class (by determination of class bracket by respondents), gender (as marked by respondents), education level (via primary, secondary, higher, or none), and ethnicity (as marked by respondents). All of these variables were collected again from the WVS. Each of these factors must be kept constant to best measure the relationship between the dependent and main independent variables. The controls equally have roles in social trust, race marginalization, and affirmative action and are standard in the literature.


    I execute an ordinal logistical regression test on my data to first examine the relationship of social trust and racial marginalization for my first hypothesis that posits social trust will decrease as racial marginalization increases. For my second hypothesis I run two ordinal logistical regression tests on my data, one to test the relationship between racial marginalization and affirmative action, and the next to test the relationship between social trust and affirmative action. The use of ordinal logistical regression tests is pertinent in that many of the variables being tested allow for more than binary responses. The tests predict the likely outcome of each variable via odds ratios. The tests yield results that mark the likely relationship among the variables, the direction of those relationships, and if they are significant relationships. For any of my hypotheses to be correct, they must result in a negative relationship with high significance levels.


    Table 1 reports the odds ratios for the ordered logistical regression test on the impact of race marginalization on general social trust. This regression directly tests the legitimacy of my initial hypothesis:

    Hypothesis 1: A decrease in racial marginalization will lead to an increase in social trust. Race marginalization was measured by a proxy question that noted racial animus in communities and tested directly against general social trust, with six other control variables. No odds ratio reveals a positive relationship and many bordered around the value of one, meaning that the odds were nearly 1:1 in most cases, noting that the likelihood of occurrence is slim. In other words, most of the data shows that the odds of the variables affecting social trust are slim. Race marginalization, however, has the most extreme odds ratio of all (0.399:1), suggesting that the relationship between high race marginalization and low social trust is strong.

    The ratios of occurrence make little sense without noting if these relationships are significant. Four relationships in particular prove to be significant as they affect general social trust. Age proves to be the most significant, followed by race marginalization, and education, each at a 99 percent confidence interval. Income is also highly significant at a 95 percent confidence interval. Combining the significance with the odds ratio information, clearly race marginalization plays a huge role in determining trust. In this regard, the null hypothesis for the first test can be rejected that there is no relationship between racial marginalization and general social trust. Hence, the results strongly support Hypothesis 1, as the results show that the more racial marginalization there is, the less social trust there will be.

    The subsequent tests assess first if affirmative action can reduce racial marginalization and if it has any impact on general social trust. Table 2 reports the ordered logit results of affirmative action’s impact on racial marginalization, with affirmative action measured by opinions of the government’s responsibility to the welfare state as a proxy. Again, many of the results hover around the 1:1 odds ratio, meaning that their odds of impact are slim. This includes the affirmative action variable. For this test, the only majorly different odds ratio is also the most significant variable. Education has an odds ratio of 0.637:1 and is significant at the 95 percent level. The results suggest accepting the null hypothesis and concluding that there is not significant support of affirmative action having an impact on racial marginalization.

    Still, the more important component is the relationship between affirmative action and general social trust. Even if affirmative action does not reduce racial marginalization, if the trust of society can be affected, everyone can benefit in the long run. Table 3 below reports the relationship between affirmative action and general social trust.

    Once more, the odds ratios stay close to the value of 1. The variable for affirmative action is extremely close to one and is not significant at all. While age is highly significant, education seems to reflect a more prominent negative relationship with social trust that is also highly significant when government is involved. Income is significant once more in the relationship with social trust, but at a 95 percent level rather than 99 percent. In both tests, affirmative action proves to be insignificant and has no strong relationship to racial marginalization or general social trust. In both cases, the null hypotheses must be accepted. Hence, the results do not lend enough support to accept H2: Corrective legislation [notably affirmative action programs to decrease racial marginalization] will increase social trust.


    The first test successfully proved that racial marginalization and general social trust are connected in a strong and significant relationship. The odds are low that a Brazilian citizen will distrust someone of a different race while maintaining a high level of general social trust. This makes sense for both a dominant racial group and those of darker skin who are more marginalized. If the dominant group maintains prejudice against other races, general trust cannot be achieved. Equally, if marginalized groups feel as though they cannot trust the dominant group due to a history of prejudice and racism, general social trust still cannot be achieved. Recall that scholars like Robert Putnam regard trust as a crucial building block to both social capital and democratic society at large insofar as trust enhances political participation and notions of national citizenship (Putnam 1993). Clearly, these results illuminate a great need to further examine how race relations affect greater political dynamics in a country.

    These results also help to clarify Putnam’s more recent findings on the relationship between diversity and social trust. Based on a similar measure of opinions of race in neighborhoods, Putnam found, “In more diverse communities, people trust their neighbors less” (Putnam 2007, pg. 148). On one hand, my study expands this finding from a U.S. context to a Brazilian context. My study, however, delves further into the matter of diversity, explaining that it is not just the variety of people in a community that decreases social trust, but racial marginalization and the historical racism that penetrates society and its institutions. Hence, more research needs to be conducted to further examine the effects of racial marginalization on social trust, social capital, and democracy in general. Furthermore, better data may be collected to better assess this claim. As noted earlier, racism in Brazil is much more than subjective feelings of one race to another, but the greater political, economic, and social marginalization of those who have darker skin color. Therefore, more substantive data must be collected to better illustrate this dynamic and note the extent to which racism has political ramifications.

    The ability of affirmative action to reduce racial marginalization and increase general social trust was proven to be lacking with the last two ordered logistical regression tests. In many ways, the results make sense. The data spanned a 27-year period where only six years had affirmative action quotas and programs actually in place in Brazil. In addition, the question used to measure affirmative action was a highly generalized proxy based on notions of government responsibility. The question is highly ideologically biased and does not confirm actual success. Last, the way affirmative action operates, it attempts to correct historical racial marginalization and does so for privileged students of color who have ascended through a society that has already marginalized them for 18 years. Hence, affirmative action inherently conducts an ex post facto type of correction that benefits a privileged few and is only measurable on a long-term basis. This is not to delegitimize the potential or necessity of affirmative action, but to say that it may not be enough to solve problems of race in a society.

    Ultimately, the test revealed no correlation between affirmative action and social trust. My findings are strikingly similar to the findings of scholar Amy Erica Smith in her study of citizen support for affirmative action in Brazil. She found that the majority of Brazilians, approximately 70 percent, support affirmative action in Brazil (Smith 2010). She found level of education, race, and income level of the respondents to her survey the most significant factors affecting opinions. Specifically, those who were white or lighter skin colored, those with more education, and those of higher income levels were more likely to disapprove of affirmative action quotas and programs. Given this data, it makes sense that affirmative action would not affect social trust because a high majority of citizens accept the programs.

    Perhaps the most interesting find from the data was the consistent significance of three control variables in the tests against social trust: age, education, and income. These figures were equally consistent with Smith’s findings and most seem reasonable to accept. Both age and income make sense in the data; the older a person is, the less likely he or she is to have general social trust, and the more one earns, the less likely he or she is to trust. Given Brazil’s recent emergence from an authoritarian regime, those who are older and earn more have previous experience with corrupt and abusive rule and may naturally have had their trust weathered. In addition, the recent emergence of affirmative action and attention to race issues marks a dynamic paradigm shift from previous notions of Racial Democracy. These older and more privileged people would naturally be more disadvantaged with racial justice in the short term because it disrupts their ideological foundation and their societal privileges, as these people are also more likely to be homogenous in race and less marginalized.

    Most fascinating of all is the education value, shown to be significant in all three tests. U.S social trust literature always points to education having a positive impact on social trust (Putnam 1995, Newton 2001). Both in Smith’s findings and my results, however, education has a negative relationship with social trust and racial marginalization: those with higher education are more likely to distrust generally as well distrust governmental programs and other races. Smith notes that those more educated are unsupportive of affirmative action because they have already benefited from higher education and see the concept of quotas as unfair (Smith 2010). Actual opinions versus reported opinions may also differ in attempting to be more politically correct or respectable (Boven 2000). The vestiges of the dictatorial regime may also have an impact on who receives education and what kind of education one receives in Brazil. Lastly, the most educated also represent the most elite in the society; they are predominately lighter skinned, so racial animus and notions of Racial Democracy may also play a role in these relationships. In any case, more research should be conducted to review how education, trust, and racism interplay in Brazil.


    Discussions of race and attempts to correct the historical race marginalization are quickly emerging in Brazil. As a fast growing economic power, more eyes from abroad are looking at Brazil and noticing various types of injustice that occur within the country’s boundaries. This study examined the relationship between racial marginalization and social trust as a jumping off point for further examination of how racism operates in Brazilian society and the greater effects it has politically. Affirmative action, as a recent proclamation of racial disparity in Brazil and the first major step to correcting these problems, proves to be a critical place to begin analysis on how government can take action to correct historical racial injustice. It is also a strategy heavily adopted and currently in contention in the United States, further meriting its examination abroad.

    Given the complexity of each component in my study, more nuanced data would be appropriate to further affirm or expand on my results. More recent data that looks explicitly at affirmative action and its relationship with solving racial marginalization as well as its impact on social trust may yield more fruitful results than the tests conducted in this project. Nevertheless, my data is telling of a serious problem of racism and how it affects Brazilian society at large. Furthermore, a closer examination of the previous Brazilian dictatorial  regime and how it may have impacted social trust and race relations may be necessary to help explain why the results came out in the fashion that they did.

    Many scholars have studied the effects of social capital, but few have looked specifically at social trust and even fewer on how race plays a role in general social trust. Most provocatively, scholar Robert Putnam found that diversity has a negative impact on social trust, but never explained why this may be the case (Putnam 2007). This study makes the argument that this dynamic may be explained through racial marginalization and societal racism that would fuel distrust among a diversity of people. Future research should examine this relationship in other contexts of heavy race marginalization and, eventually, on a more general scale. This is vital in that social trust is a critical component of civil society and how a democracy can most effectively be run.

    Particularly in Brazil, social trust is of great importance. Brazil has been touted as a country already with incredibly low amounts of social trust despite its economic rise. With incredible racial marginalization occurring in the underbelly of the country, socio-economic problems on a larger scale are become more glaring. The very recent Salad Uprising protests in June and current outcries over government spending and responsibility are most likely linked to social trust, or the lack thereof. Marginalized groups have reached a tipping point of governmental distrust and are demanding more attention from the government to fix the problems that ail the country rather than increasing standards of living to finance huge expenditures like the upcoming World Cup. Clearly, more study must be focused on Brazilian civil society and marginalization as well as governmental welfare, as most dangerous and disastrous effects may follow from the protests if a solution is not found quickly. This project marks the first step in assessing the complex social relationships among peoples and between people and the government.


    • Andrews, Rhys. 2008. “Civic Engagement, Ethnic Heterogeneity, and Social Capital in Urban Areas: Evidence from England.” Urban Affairs Review 44(3): 428-440.
    • Beato, Lucila Bandeira. 2004. “Inequality and Human Rights of African Descendants in Brazil.” Journal of Black Studies 34(6): 766-786.
    • Bonilla-Silva, Eduardo and David R. Dietrich. 2009. “The Latin Americanization of U.S. Race Relations: A New Pigmentocracy.” In Shades of Difference: Why Skin Color Matters. Ed. Evelyn Nakano Glenn. Stanford: Stanford University Press, pp. 40-60.
    • Booth, John A. and Patricia Bayer Richard. 1996. “Repression, Participation and Democratic Norms in Urban Central America.” American Journal of Political Science 40(4): 1205-1232.
    • Boven, Leaf Van. 2000. “Pluralistic Ignorance and Political Correctness: The Case of Affirmative Action.” Political Psychology 21(2): 267-276.
    • Da Silva Martins, Sérgio, Carlos Alberto Medeiros, and Elisa Larkin Nascimento. 2004. “The Road from ‘Racial Democracy’ to Affirmative Action in Brazil.” Journal of Black Studies 34(6): 787-816.
    • Duckitt, John and Thobi Mphuthing. 1998. “Political Power and Race Relations in South Africa: African Attitudes before and after the Transition.” Political Psychology 19(4): 809-832.
    • Freyre, Gilberto. 1986. The Masters and the Slaves: A Study in the Development of Brazilian Civilization. Berkeley: University of California Press, pp. xxvii-xliv.
    • Hanchard, Michael. 1999. “Black Cinderella?: Race and the Public Space in Brazil.” In Racial Politics in Contemporary Brazil. Ed. Michael Hanchard. Durham: Duke University Press, pp. 59-81.
    • Hernández, Tanya Kateri. 2013. “Brazil: At the Forefront of Latin American Race-Based Affirmative Action Policies and Census Racial Data Collection.” Racial Subordination in Latin America: The Role of the State, Customary Law, and the New Civil Rights Response. Cambridge: Cambridge University Press, pp. 148-170.
    • Newton, Kenneth. 2001. “Trust, Social Capital, Civil Society, and Democracy.” International Political Science Review 22(2): 201-214.
    • Putnam, Robert D. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton: Princeton University Press.
    • Putnam, Robert D. 1995. “Tuning In, Tuning Out: The Strange Disappearance of Social Capital in America.” PS: Political Science and Politics 28(2): 664-683.
    • Putnam, Robert D. 2007. “E Pluribus Unum: Diversity and Community in the Twenty-first Century.” Scandinavian Political Studies 30(2): 137-174.
    • Rose, Heather. 2005. “The Effects of Affirmative Action Programs: Evidence from the University of California at San Diego.” Educational Evaluation and Policy Analysis 27(3): 263-289.
    • Rothstein, Bo and Eric M. Uslaner. 2005. “All for All: Equality, Corruption, and Social Trust.” World Politics 58(1): 41-72.
    • Smith, Amy Erica. 2010. “Who Supports Affirmative Action in Brazil?” Americas Barometer Insights (49): 1-8.
    • Tavolaro, Lília G. M. 2008. “Affirmative Action in Contemporary Brazil: Two Institutional Discourses on Race.” International Journal of Politics, Culture, and Society 19(3/4): 145-160.
    • Telles, Edward. 2009. “The Social Consequences of Skin Color in Brazil.” In Shades of Difference: Why Skin Color Matters. Ed. Evelyn Nakano Glenn. Stanford: Stanford University Press, pp. 9-24.
    • Telles, Edward and Liza Steele. 2012. “Pigmentocracy in the Americas: How is Educational Attainment Related to Skin Color?” America’s Barometer Insights.
    • Thornton, Margaret. 2001. “EEO in a Neo-Liberal Climate.” Journal of Interdisciplinary Gender Studies 6(1): 77-104.
    • Uslaner, Eric M. 2004. “Trust and Social Bonds: Faith in Others and Policy Outcomes Reconsidered.” Political Research Quarterly 57(3): 501-507.
    • World Values Survey. 2012. “1981-2008 Integrated Data.” (Accessed June 19, 2013).

    Table 1: Ordered Logit Results: Racial Marginalization impact on General Social Trust

    Variable Odds Ratio (Std. Err.) Z P>|Z|
    Race Marginalization 0.399(0.130)*** -2.81 0.005
    Age 0.981(0.005)*** -3.68 0.000
    Gender 1.297(0.212) 1.59 0.112
    Education 0.736(0.085)*** -2.65 0.008
    Class 0.919(0.085) -0.92 0.360
    Income 0.917(0.035)** -2.28 0.023
    Ethnicity 1.000(0.000) 0.44 0.657

    N=2499, LR χ2= 35.37, Pseudo R2=0.0294
    Significance levels: * : 10%, ** : 5%, *** : 1%

    Table 2: Ordered Logit Results: Affirmative Action’s Impact on Racial Marginalization

    Variable Odds Ratio (Std. Err.) Z P>|Z|
    Race Marginalization 1.008(0.034) 0.24 0.812
    Age 1.005(0.007) 0.66 0.509
    Gender 1.262(0.278) 1.06 0.290
    Education 0.637(0.119)** -2.42 0.016
    Class 1.046(0.120) 0.39 0.698
    Income 0.966(0.051) -0.66 0.512
    Ethnicity 1.000(0.000) -1.36 0.175

    N=2507, LR χ2= 14.81, Pseudo R2=0.0191
    Significance levels: * : 10%, ** : 5%, *** : 1%

    Table 3: Ordered Logit Results: Affirmative Action’s Impact on General Social Trust

    Variable Odds Ratio (Std. Err.) Z P>|Z|
    Race Marginalization 0.991(0.026) -0.34 0.734
    Age 0.980(0.005)*** -3.72 0.000
    Gender 1.281(0.210) 1.51 0.130
    Education 0.730(0.084)*** -2.73 0.006
    Class 0.926(0.086) -0.82 0.411
    Income 0.924(0.035)** -2.07 0.039
    Ethnicity 1.000(0.000) 0.60 0.547

    N=2496, LR χ2= 29.48, Pseudo R2=0.0245
    Significance levels: * : 10%, ** : 5%, *** : 1%