INTRODUCTION
Research on rap music, and Hip Hop culture more generally, has accumulated over the decades, as scholars engage in spirited conversations regarding this lifestyle and genre of music.Footnote 1 The rap literature is both large and multidisciplinary, so we focus here on its politics-specific processes and implications. Examples of politicians exploiting the power of rap music for voter-mobilization and publicity purposes abound in social media,Footnote 2 and many of the subjects and situations discussed by rappers are inherently political (Bonnette Reference Bonnette2015; Kitwana Reference Kitwana2002; Perry Reference Perry2004; Rose Reference Rose1994, Reference Rose2008; Spence Reference Spence2011). In addition to being an important component of campaigns, rap’s political commentary reports upon and reinterprets the social problems that are customarily addressed through policy agendas, representing a voice for those whose voices are typically marginalized or unheard. These social issues, including racial, educational, economical and criminal injustices, are addressed within rap music and ultimately are instrumental to the political choices citizens make (Dawson Reference Dawson and Lamont1999; Collins Reference Collins2006; Harris-Lacewell Reference Harris-Lacewell2004). This is what rapper Chuck D meant when he famously remarked that rap music is Black America’s CNN (Ridenhour and Jah, Reference Ridenhour and Jah1997). Because it brokers information, rap music works like a socializing agent: generating cultural ties that connect its listeners to the topics exemplified through lyrics (Chang Reference Chang2005; Dyson Reference Dyson2007; Kelley Reference Kelley1996; Keyes Reference Keyes2002; Morgan Reference Morgan1999; Peterson Reference Peterson and Frail Marcos2007; Watkins Reference Watkins2001, Reference Watkins2005), and providing a forum through which rap-activists (or “raptivists”) worldwide can interact, organize, and agitate (Maddex Reference Maddex2014; Saad Reference Saad2016).Footnote 3
Rap’s sociopolitical importance is eclipsed only by its controversial reputation. Rap is reviled by many critics because it is often characterized by racy imagery and destructive themes. Debates over the music’s negative reputation are particularly vociferous within Black communities. Defenders admire the music as a unique and racially-authentic form of artistic expression, while others worry about rap leading Black youth astray (for a summary of this debate, see Rose Reference Rose2008). This begs the question: are all Black youth equally vulnerable to the negative influences present in some rap music? We believe the concerns about rap are exaggerated for two reasons. First, not all African Americans listen to rap, and this diversity in consumption patterns generates important variation in individuals’ exposure to the music’s purported problem areas. Second, and more importantly, many Blacks are capable of enjoying rap while rejecting its problematic subject matter. Rather than absorb it unthinkingly, people can be skeptical consumers and appraise their music of choice. The ability to critique rap, therefore, denotes a person’s ability to identify negative rap messages, and, once identified, express a preference against them.
Beyond demonstrating the extent to which rap fans judge their music’s content, critique is also an inherent component of Hip Hop culture. Questioning the status quo, challenging authority, and fighting against injustice are essential practices that are common to all forms of Black music (Blanchard Reference Blanchard1999; Eshun Reference Eshun2015). Accordingly, and as Tony Mitchell (Reference Mitchell2001) notes, radical social criticism is part of Hip Hop’s DNA (see also Jay-Z 2010; Keyes Reference Keyes2002).Footnote 4 Through rap critique, people signal to their peers how well they “represent” Hip Hop culture (Forman Reference Forman2000; Sachs Reference Sachs2009). Insofar as the willingness or capacity to critique society in general (and aspects of rap music in particular) reveals how closely a person is connected to Hip Hop—and, by logical extension, how likely it is that a person’s music and culture shape their political actions (Dyson 1993; Martinez 1997)—studying rap critique is a necessary step toward understanding an individual’s “rap consciousness.” Here, we define rap consciousness as a person’s ability to strike a healthy balance between their rap consumption and rap critique: they can enjoy the music and not be led astray, and, more importantly, they can demand more of rap by “calling it out” when necessary. In short, conscious rap consumers can engage the music in a prosocial manner: maximizing the music’s positive effects internally while minimizing any negative effects. Readers can draw similarities between our use of the term “consciousness” and what people refer to in current vernacular as being “woke” (Koren Reference Koren2016).
We explore the mobilizing power of rap consciousness among the music’s most avid consumer base: young African Americans (Cohen Reference Cohen2010, 74; Tanner et al., Reference Tanner, Asbridge and Worley2009) because, as Bakari Kitwana (Reference Kitwana2002) notes, “rap music more than anything else has shaped the new Black youth culture” (p. 9). Research linking rap music to mass-level political behavior is hardly new (see, e.g., Bynoe Reference Bynoe2004; Hemphill Reference Hemphill2012). However, we are, to our knowledge, the first to explore two uncommon determinants of sociopolitical participation in African American communities: rap music consumption and rap critique. Drawing insights from several social science and humanities disciplines, we argue that rap music represents a mechanism through which young Blacks receive information about social issues that is also relevant to politics, and that rap consumption, when combined with critique, can spur our youth into political action.
We begin by building upon the literature on Hip Hop’s “positive” and “negative” effects to describe rap critique as a tool for navigating the music’s potentially negative points; and we explore the interplay of rap critique and rap consumption. Next, we develop a novel hypothesis regarding the possibility that “imagining a better world” (measured as the degree to which individuals are critical of the music they enjoy) can correlate with the desire to “build a better world” (operationalized as a person’s voter turnout and non-voter participation). We test this claim using the Black Youth Project (BYP)’s Youth Culture Survey (Cohen Reference Cohen2005), analyzing the impact of rap on participation as a selection problem: African Americans sort themselves non-randomly into varying levels of rap consumption, and this “selection bias” influences the degree to which Blacks translate their music attitudes into political action. We discuss the research findings, focusing on the conditional impact of rap critique as a predictor of civic activism, and we offer some concluding remarks regarding the contributions this study makes and the implications it carries for the scholarship on rap music and political behavior.
RESEARCH ON RAP MUSIC’S SOCIOPOLITICAL IMPACT
The literature on rap music is as vast as it is eclectic. We find it useful to organize rap research into broad perspectives. Lester Spence (Reference Spence2011) applies the economic concepts of “supply and demand” to the study of rap music and politics, and we adopt the author’s ideas here.Footnote 5 As such, in our use of these concepts, one perspective in the literature emphasizes the supply side by linking politics to the “production” of rap (i.e., the artists, art form, and agents who corporatize the music). These studies focus on rap’s responsibility to convey certain messages. For example, supply side research analyzes the degree to which lyrics contain political references (Decker Reference Decker1993; Kurbin 2005; McDonnell Reference McDonnell1992; Stapleton Reference Stapleton1998), rappers serve as political activists (see Asante Reference Asante2008; Boyd Reference Boyd1994, Reference Boyd2002; Clay Reference Clay2012; Wire Tap Staff 2003), and the culture embraces gender (Collins Reference Collins2006; Morgan Reference Morgan1999; Pough Reference Pough2004, Reference Pough2007) and sexual (Boykin Reference Boykin2005; Dean Reference Dean2008) inclusion. Other supply side studies celebrate the music’s aesthetic components (Keys 2002; Krims Reference Krims2000; Salaam Reference Salaam1995; Shusterman Reference Shusterman1991; Slovenz Reference Slovenz1988) and regional nature (Forman Reference Forman2000; Keyes Reference Keyes2002; Kubrin Reference Kubrin2005), chronicle the trends of “mainstream” vs. “underground” rap(pers) (Cleary Reference Cleary1993; Dyson Reference Dyson2007; George Reference George2005; Leach Reference Leach2008; Lena Reference Lena2006; Nelson Reference Nelson1999; Ogbar 2007), and expose how the entertainment industry profits from Hip Hop (Quinn Reference Quinn2004; Watkins Reference Watkins1998, Reference Watkins2005).
Until recently, fewer studies explored the demand side, which, as we detail below, draws connections between the consumption of rap music and a host of outcomes. Our exploration of rap’s “positive” and “negative” consequences fits within (and contributes to) the demand side literature. There is a tendency in this line of research to focus not on mass-level participation but on attitudinal variables like policy preference, Black Nationalist ideology, and agreement with feminist viewpoints, etc. (see Bonnette Reference Bonnette2015; Dawson Reference Dawson and Lamont1999, Reference Dawson2001; Harris-Lacewell Reference Harris-Lacewell2004). We deviate from this trend by emphasizing the music’s ability to stimulate political activism. In the pages that follow, we: 1) theorize about rap consciousness as a mobilizing resource; 2) operationalize consumption and critique (our proxy variables for rap consciousness) as intersecting pathways to activism; and 3) conduct in-depth empirical tests of the connection between exposure to popular music and political involvement. In so doing, we introduce an alternative vision of Black politics, and more specifically, a set of variables and theoretical considerations traditionally left out of political science research.
Fear of a “Rap Planet”: The Negative Impact of Consumption
Justified or not, there is a widespread suspicion of rap music. Those who study this suspicion write about what Travis Dixon and Takeshia Brooks (2002) nicknamed rap’s “negative effects.” This line of research examines the music’s deleterious influence on society—particularly, on Black youth.Footnote 6 Inspired by a phrase appearing on the cover of the rap group, Public Enemy’s, third album, scholars (e.g., Rose Reference Rose1991) and bloggers (see Deterline and Jones, Reference Deterline and Jones1994) characterize this suspicion as the “fear of a rap planet.” Prominent individuals expressing this fear include Reverend Calvin Butts, civil rights activist C. Delores Tucker, social issues advocate Tipper Gore, ex-congressman Bob Dole, former senator and Vice President Dan Quayle, and Supreme Court nominee Robert Bork (see Ogbar Reference Ogbar1999 for details). Within African American communities, rap’s negative reputation carries with it a generational connotation. Many older Blacks believe that rap corrupts their children, promotes teachings that impede racial progress, and encourages everything from lawbreaking to sexual exploitation to excess (Cohen Reference Cohen2010).Footnote 7 Some elders even criticize members of the Hip Hop community directly, blaming them for Blacks’ poor living conditions and circumstances. Old-guard critics, such as Bill Cosby and numerous others, continue to fuel considerable debate by pointing out—sometimes quite publicly—the potential shortcomings of Black culture (see, e.g., Cosby and Pousssaint, 2007; Dyson Reference Dyson2005).Footnote 8
To borrow statistical terminology, underlying the anxiety that the music engenders are two implicit assumptions about the “distributions” of rap critique and rap consumption. First, the negative-effects literature presumes that consumption is consistently heavy among young African Americans. Put differently, the belief is that most young African Americans listen to enough rap that the music can pervert their attitudes and actions. If consumption is a measurable concept in which higher values signal greater music exposure, then the average consumption level for young Blacks should be high, and the variability around that average should be low. Similarly, the second assumption is that critique is consistently light: young people’s ability to call out their music’s objectionable substance is distributed such that it has a low mean and low variance. In the next section, we question these assumptions about the distribution of rap critique and rap consumption, for we are not convinced that the negative consequences of the music are automatic. Concerns about rap are overstated because they either neglect or do not fully appreciate the heterogeneity that young Blacks exhibit in their consumption and critique levels. Exposure can fluctuate considerably in the sense that not all young African Americans are rap fans. Critique can vary as well: some accept the music’s negativity, while others disagree with its troublesome features. We explain below how Black youth can consume their music without being led astray.
Overcoming the Negativity: The Importance of Rap Critique
The literature on negative effects is not the only perspective on rap. Research on the music’s “positive effects” challenges the distributional assumptions mentioned above. This field of study is particularly well-tilled by counseling (see Elligan Reference Elligan2004; Hadley and Yancy, Reference Hadley and Yancy2012; Tyson Reference Tyson2002) and education (e.g., Giroux Reference Giroux1994; Ladson-Billings Reference Ladson-Billings1995; Martinez Reference Martinez1994; Powell Reference Powell1991) scholars, as practitioners in these areas discover that the music enables them to connect more effectively with clients of color. Positive effect studies remind us that the displeasing themes in rap do not offset the beneficial information the music provides its listeners. Bringing the logic of the positive-effects perspective into the realm of politics, Cathy Cohen (Reference Cohen2010), Emery Petchauer (Reference Petchauer2011), and Rachel Sullivan (Reference Sullivan2003) employ qualitative and mixed-methods approaches that empower rap fans to say for themselves how the music shapes their lives. Each of these authors finds ample evidence of its popularity among young African Americans, lending credence to concerns about rap’s pervasiveness (i.e., the assumption that consumption has a “large mean”). However, the youth of Black America have, by and large, not been polluted by rap because, as the authors explain, individuals are capable of launching advanced and content-oriented critiques of this music. The power to criticize calls into question the idea that consumption has a “small variance” because dissatisfied Blacks can opt-out of the music to some degree. Likewise, the notion that rap critiques cluster tightly around some hypothetical average seems doubtful: beyond the fact that not all rap is negative (and therefore worthy of critique), the assumption trivializes both the degree of control rap fans exert over their consumption practices and the sheer diversity of attitudes these fans express about their music.
The idea that many Blacks can prevail over rap’s potentially negative messages—just as they overcome other disruptive forces—is consistent with Richard Allen’s (Reference Allen, Berry and Asamen1993; Reference Allen2001) argument that Blacks can cultivate psychosocial tools that they use to protect themselves from oppression while, at the same time, helping them reassert their cultural identity.Footnote 9 By offering another perspective on the “rap negativity” debate, positive effect scholars, like Dixon and colleagues (2009) acknowledge that people can enjoy the music while still wanting some of its lyrics to improve. The ability to “imagine a better rap world” gives rise to the idea that skepticism can shield people from the music’s adverse consequences. Petchauer elaborates on the importance of rap fans having a “critical consciousness” when he recommends that Black youth be skeptical consumers (see Petchauer Reference Petchauer2011; Irby and Petchauer, Reference Irby, Petchauer, Porfilio and Viola2011). At a minimum, young African Americans who are exposed to (but are critical of) rap may become resistant to its negative impact. This resistance can manifest itself in several ways: holding a skeptical eye to the music can mitigate rap’s detrimental influence, either by blocking unfavorable content before it reaches the mind, or by counteracting the negative messages getting past those defenses with positive ones (Dixon and Brooks [Reference Dixon and Brooks2002] make a comparable argument).
Cohen (Reference Cohen2010) does not argue for rap critique as a form of protection per se, but instead posits that Black youth are expected to be acquainted with the rap and Hip Hop world.
The necessity of black youth being conversant in the culture of hip-hop should not be underestimated; the idea that black youth, because of their critiques of rap, and in particular gangsta rap, should walk away from the art form seems unrealistic. It would be like a stock trader refusing to read the Wall Street Journal because he or she disagreed with the political ideology of the paper (p. 78).
Because of the inevitability of being ensconced in the rap world—and the cultural necessity of living in this world (Kajikawa Reference Kajikawa2015)—rap critique can serve as a protective function: Black youth can develop aspects of their upbringing that can help prevent future bad decisions or situations. As Sullivan (Reference Sullivan2003) mentions, Blacks have the potential to grow and develop rich, full lives just like any other youth; however, their pathways to these lives are more easily sidetracked than those of their White counterparts. Economic and emotional hardships that might be mere “bumps in the road” for Whites could be catastrophic for African Americans (Giroux Reference Giroux1996), and the likelihood of “staying on track” increases if a young Black person has access to (and makes appropriate use of) institutional, communal, familial, and emotional resources that can shelter her from trauma and, when necessary, steer the person in the right direction (Cohen Reference Cohen2010). Being protected can mean having the cultural wherewithal to connect with—and learn from—the music while overcoming the urge to imitate dangerous role models or to internalize harmful ideas.Footnote 10 This is what Cohen and others mean by rap consciousness. Socially conscious artists offer their style of music as an alternative to, if not a rejection of, the negativity that shows up intermittently in mainstream rap. Likewise, conscious consumers are those who crave good rap messages, filter out bad ones, and, most relevant to our research, articulate their preferences via critiques. In short, positive effects scholars explain how individuals can consume rap while engaging with its content in ways that can diminish its potentially negative impacts.
RAP CRITIQUE AND CONSUMPTION AS INTERTWINING CHARACTERISTICS
In statistics, proxy variables serve in the place of unobserved quantities of interest. While not direct measures of a desired quantity, good proxies correlate strongly with an unobserved variable (Lewis-Beck et al., Reference Lewis-Beck, Bryman and Liao2004). In our case, consciousness itself is unmeasured, but we can infer it by gauging a Black person’s rap consumption and music critiques. Consumption and critique serve as an admittedly crude (but arguably useful) substitute measures of rap consciousness. In Table 1, we explore this idea more fully by developing a typology of rap consumer-critics. Specifically, we arrange attitudes about rap music along a continuum anchored by Black youth who are commendatory in their responses on one end and those who express condemnatory statements on the other. For simplicity’s sake, we use the columns in Table 1 to quantify rap skepticism as a categorical variable ranging from “low” to “high” critique.Footnote 11 Rather than presume that all Black youth enjoy the music (or that they partake of it in “equal doses”), the rows in our table account for the heterogeneity of rap consumption. We acknowledge that some consumption is involuntary, for young Blacks do not always control the music around them. Nevertheless, it is safe to conclude that an individual who consumes rap frequently is getting a higher level of exposure than someone who seldom or never partakes of rap.Footnote 12 Consumption, therefore, is our measure of an individual’s underlying preference for rap music.
A Typology of Rap Music Critics
Source: Created by the authors.
Four idealized types of consumer-critics become apparent as we consider how Black youth might distribute themselves along levels of rap consumption and rap critique. Our typology helps us to investigate, for example, young African Americans in the top left cell of the table: they consume rap infrequently, if at all, and because they are less familiar with the lyrics and imagery, take little exception to the music. Compare these critics to the ones in the top right cell of Table 1, who, despite the fact that they never did (or no longer) listen to rap songs or watch rap videos, are highly critical of the music. Their censures differ from those of the potential critics in the bottom right cell: these African Americans are highly critical heavy consumers who hold rap music to a loftier standard.Footnote 13 In contrast to the above critics are the Black youth in the bottom left cell of Table 1, who enjoy rap frequently and are not perturbed by some of the music’s content.
The different types of critics reflect the variability of Black youth’s skepticism about rap and their preference for the music. In the next section, we evaluate the role of rap critique and consumption as determinants of civic action. Politics is but one of many contexts in which rap can have an impact. That said, our focus on activism stems not only from the fact that there are fewer empirical studies of rap’s participatory consequences (for exceptions, see Bonnette Reference Bonnette2015; Hemphill Reference Hemphill2012; Spence Reference Spence2011), but also from the importance of studying activism in American society (Verba et al., Reference Verba, Lehman Schlozman and Brady1995). As you will see below, this focus on political participation inspires us to specify a hypothesis regarding the conditional relationship between rap consumption, rap critique, and Black youth activism.
CRITIQUE, CONSUMPTION, AND BLACK YOUTH’S POLITICAL PARTICIPATION
Conditional Hypothesis
Intuitions from the positive (rap) effects literature suggest that social, familial, and educational resources can shield Black youth from harm. As a proxy variable of rap consciousness, critique can represent the degree to which young people have amassed the necessary shielding. Since resources and interpersonal networks are known predictors of activism, we anticipate that rap critique, like time, money, or civic skills (Verba et al., Reference Verba, Lehman Schlozman and Brady1995), can enhance activism: the more rap critical a young person is, the easier it is for her to receive the uplifting, politically-empowering messages in the music.Footnote 14 Put differently, we expect a positive relationship between our variables of interest, with activism levels increasing as critique intensifies.
However, the mobilizing impact of rap critique on political participation should only apply to individuals who opt into higher levels of music consumption.Footnote 15 Just as traders get stock tips from the Wall Street Journal, Black youth rely on rap to inform them about relevant social issues (Bonnette Reference Bonnette2015; Harris-Lacewell Reference Harris-Lacewell2004; Spence Reference Spence2011). After all, and as Catherine Powell (Reference Powell1991) explains, rap music gives its audience an “education with a beat from the street.” But to reap the informational benefits of rap, African Americans must avail themselves of the music, and critique makes it possible for young Blacks to differentiate positive from negative messages. In this sense, we expect rap critique to exert a conditionally positive influence over participation levels. We expect activism should be highest among those young African Americans who: 1) are highly critical of rap music, and 2) consume large amounts of rap music.
To explain why this particular combination of critique and consumption fosters high activism levels, we turn to the political science research of the 1960s and 1970s on “political trust”—i.e., whether the government can be relied upon to produce outcomes that are consistent with voters’ expectations (see Hetherington Reference Hetherington2005)—and “political efficacy”—defined either as a person’s ability to grasp the complexities of politics (internal efficacy) or a person’s confidence that the government will respond to her political demands (external efficacy).Footnote 16 William Gamson (Reference Gamson1968), Patrick Gurin and colleagues (1969), Gurin and Edgar Epps (1975), and Arthur Miller and colleagues (1978) offer a “distrustful-efficacious” explanation of African American political participation (see also Milbrath and Goel, Reference Milbrath and Goel1977; Muller 1977). Specifically, the authors were responding to the discovery that the activism levels of African Americans often outpace those of Whites from similar socioeconomic backgrounds. In addition to demonstrating that Black people tended to participate at higher rates than their socioeconomic status would predict (a finding discussed in detail by Tate Reference Tate1991, Reference Tate1993), this discovery reveals that conventional models of political behavior do not apply as well to Black voters (Dawson Reference Dawson1994, Reference Dawson2001; Walton Reference Walton1985).
To address this puzzle of Black political participation, Gamson, Gurin, and their colleagues argued that, in the absence of economic resources, African Americans make use of psychological orientations that spur them into action. Scholars refer to these orientations as “Black consciousness” (see Olsen Reference Olsen1970; Verba and Nie, 1972; National Opinion Research Center, Almond and Verba, Reference Almond and Verba1960), and the impact of consciousness on political activism is summarized best by Richard Shingles (Reference Shingles1981):
… [T]he primary reason Black consciousness has such a dramatic effect on political participation is that it contributes to the combination of a sense of political efficacy and political mistrust, which in turn induces political involvement (p. 77).
Restated, Shingles argues that Black activism grows best when there is the right mix of need and capacity. The need to act politically stems from the inadequacies in government that Blacks perceive (measured by their political [dis]trust), while capacity represents an African American’s wherewithal to get involved (operationalized as internal political efficacy). Applying Gamson and Gurin’s explanation to the study of rap, the need to participate is communicated through the music’s mobilizing messages, as rappers include lyrics in their songs and imagery in their videos that point out the failings of government and inspire listeners to seek solutions to sociopolitical problems. By consuming rap, Black youth give themselves the opportunity to receive these types of messages; however, the politically empowering content in rap music often reaches its audience as a “noisy signal,” one in which “bad” messages are frequently packaged along with “good” ones. The willingness to critique rap, therefore, tells us something about an individual’s capacity: highly critical rap fans can decipher their music’s content, gleaning the mobilizing signals from the noise. By this logic, the growth potential for activism should be greatest among those Black youth who are high-consuming heavy critics.
Returning to the critic types discussed earlier, we hypothesize that the potential for political involvement is highest among the African Americans in the bottom right quadrant of Table 1. These young people expose themselves to sufficient amounts of rap and are also “conscious” enough to be skeptical of their music. Highly critical minimal consumers (upper right quadrant) miss out on rap’s politicizing messages, heavy consumers who lack critique (lower left quadrant) are at risk, and Black youth who neither consume nor question the music (upper left quadrant) are least likely to take part in politics. Stated formally,
Consumption ➜ Critique ➜ Participation Hypothesis: Rap music consumption moderates the effect of rap critique on political participation. Rap critique has no impact on political participation among young Blacks who do not frequently consume rap. However, critique should enhance the participation levels of African Americans who are heavy consumers of rap music.
The rationale behind our “highly critical heavy consumer” hypothesis is straightforward. If a young person does not consume enough rap, then rap cannot mobilize her, regardless of how critical she is of the music. On a related note, if an African American is not able (or willing) to “call rap out” on its negative content, then it does not matter how much rap the person consumes; the mobilizing messages in the music are likely to go unheeded. Conversely, the fans who listen to lots of rap and offer critiques are the ones who can appreciate the music’s political content. Because they are open to receiving those messages, highly critical heavy consumers are uniquely qualified to translate their desires for social change into behaviors that create such change. To summarize, a strong skepticism of rap can, when combined with a commitment to consuming the music, contribute to heightened activism levels. Other combinations of rap consumption and rap critique levels should be less effective at mobilizing Black youth.
Survey Data
We test this claim about the role of rap consciousness on political activism using the 2005 Black Youth Project (BYP)’s Youth Culture Survey, a nationally representative poll of 1,590 Americans, ages fifteen through twenty-five (Cohen Reference Cohen2005).Footnote 17 Our analyses utilize the BYP’s oversample of self-identified African Americans (n = 635). Using regression analyses with endogenous treatment effects, we fit a model to estimate the differential values of our outcome variable (political participation) with respect to a presumably exogenous predictor (rap critique), conditional upon our endogenous variable (rap consumption).
This strategy allows us to address the fact that some youth listen to more rap than others for reasons that might plausibly be related to our dependent variable. In an experimental setting, we could have randomly assigned participants to receive varying levels of rap music exposure, but such an approach is not possible with the BYP survey. We therefore adopt a methodology that is suitable for extracting causal effects from observational data (Brown and Mergoupis, Reference Brown and Mergoupis2011; Guo and Fraser, Reference Guo and Fraser2015; Nichols Reference Nichols2007). Moreover, the polling data have two significant advantages over a laboratory setting: external validity or generalizability, and the ability to capture the impact of long-term exposure rather than a one-shot intervention.
Model Specification
Recall that we have a “selection problem.” Unlike sample selection bias, where certain observations are systematically excluded (i.e., “censored”) from a sample, we are concerned with treatment selection (Achen Reference Achen1986).Footnote 18 We doubt that there are data collection issues that prevent certain types of rap fans from being included in the BYP survey. However, because these respondents have different music exposure preferences, they sort themselves non-randomly into our “low” and “high” consumption categories—which, for us, function like treatment groups in a quasi-experiment. Endogenous treatment estimators adjust for treatment selection bias by using a statistical approach that includes the residuals from a “treatment model” into a “potential outcome model” (see Appendix 2 for a technical discussion).
The formula below outlines the treatment portion of our model, where “treatment” refers to the process by which Black youth self-select their preferred rap exposure levels.
$$consumption = f\matrix{ {church\,attendance,\,college\,degree,} \cr {linked\,fate,\,employment\,status,\,gender} \cr {race\,of\,interviewer,\,voting\,age} \cr } $$We measure our treatment (rap music consumption) using two BYP items asking respondents to rate how often they listen to rap music and watch rap videos. We dichotomize the variable using median splits: “heavy users” (n = 292) are respondents whose scores exceed the median consumption level on both survey items, and we classify Black youth with scores that fall below these medians as “light consumers” (n = 343). We code this binary variable so that heavy consumers get a score of “1” and all others receive a score of zero. Rap consumption is a function of church attendance (Alex-Assensoh and Assensoh, Reference Alex-Assensoh and Assensoh2001; Bonnette Reference Bonnette2015; Calhoun-Brown Reference Calhoun-Brown1996; Harris Reference Harris1999; McClerking and McDaniel, Reference McClerking and McDaniel2005), college education, linked fate (a measure of perceived racial group interdependence, see Dawson Reference Dawson1994; McClerking Reference McClerking, Olson and Green2008), employment status, gender, the interviewers’ race, and age group (see Appendix 1 for further details). Each of these covariates appears in previous studies as a determinant of rap consumption. For example, based on Table 2 in Michael Dawson (Reference Dawson and Lamont1999), we expect Blacks’ consumption to increase with linked fate, and decrease with age and education. Likewise, Melissa Harris-Lacewell (Reference Harris-Lacewell2004) finds that increases in wealth (employment status is our youth-appropriate proxy) correspond with a stronger preference for rap, and that Black men typically outpace their female counterparts in consumption levels—mainly because misogyny remains an unfortunate but prominent feature of the music (see also Adams and Fuller, Reference Adams and Fuller2006; Bonnette Reference Bonnette2015). Because the content of rap music can elicit “moral panics” (Cohen Reference Cohen2010), we expect consumption to decrease with rising church attendance. Sandra Barnes (Reference Barnes2008) makes a similar argument, and the connections (as well as tensions) between religion and rap were the subject of the spring 2009 issue of Culture and Religion: An Interdisciplinary Journal.
The Impact of Rap Music Critique on the Political Participation of Black Youth
Source: 2005 Black Youth Project (BYP) - Youth Culture Survey (Oversample of African Americans [N = 635]).
Notes: Table entries are coefficient estimates (with bootstrapped standard errors in parentheses) for a probit (for turnout) and Poisson (for non-voting participation) regression model with endogenous treatment effects. We calculated the treatment effects using an augmented inverse-probability weighting estimator. There are no coefficients for age group in the turnout models because only voting-age respondents can cast ballots.
+ p < 0.10, * p < 0.05, ** p < .01, *** p < .001, two-tailed test.
The next formula describes the “outcome” portions of our model. Specifically, it explains how we simulate the magnitude of political participation that a respondent would exhibit at different rap consumption levels.
or
$$participation|consumption = light$$
$$participation|consumption = heavy$$
$$ = f\left( {\matrix{ {rap\,critique,\,gov't.\,assistance,college\,educated\,parents,} \cr {took\,civics\,class,\,personal\,efficacy,parent^' s\,political\,interest,\,} \cr {experiences\,with\,discrimination,\,proselytize\,about\,politics,\,} \cr {neighborhood\,efficacy,gender,voting\,age} \cr } } \right)$$By distinguishing efforts to boost African American turnout (e.g., Diddy’s “Vote or Die” movement and Russell Simmons’s “Rock the Vote” initiative) from organizations like the Crunk Feminist Collective (CFC), the Hip Hop Summit Action Network (HSAN), and Conscious Hip Hop Activism Necessary for Global Empowerment [CHHANGE] (who take a more grassroots approach), we notice that political rap encourages both traditional and non-mainstream forms of activism (Bonnette Reference Bonnette2015). Accordingly, we include two dependent variables in our analyses: voter turnout and non-voter participation (referred to here as “civic activism”). Turnout is a dichotomous variable coded so that 1 = respondent voted in a recent national or local election, and 0 = otherwise (proportion = .489).Footnote 19 The second dependent variable is an additive index based on eleven of the thirteen remaining political activities queried about in the BYP survey (mean = 1.09, standard deviation = 1.42).Footnote 20 This index ranges from young Blacks taking part in zero non-voter activities (the minimum value) to seven acts (the observed maximum value) in the past year. Because of the latter is a “count” variable and the former is categorical, the functional forms we use for our model are probit (for voter turnout) and Poisson (for civic activism) regression.Footnote 21
Our theoretically central predictor is rap music critique, a five-item additive index that ranges from -5 (highly accepting of the messages in rap music) to +5 (highly critical), with zero as the middle category (mean = .93; standard deviation = 1.86). To rule out the possibility of rap critique having a spurious influence over political activism, we account for the impact of several predictors known to correlate with voter turnout and non-voting activity. Sidney Verba and colleagues (1995) organize these independent variables according to whether they pertain to a citizen’s capacity to take part in political life (measured in this article by dependency upon government assistance, parents’ college education, and whether Black youth took a civics class), their psychological motivation to do so (measured by personal efficacy, parent’s political interest [because there is no measure of respondent interest in the BYP survey], and perceived racial discrimination), or their access to networks of political mobilization/recruitment (measured by neighborhood efficacy and political proselytizing).Footnote 22 Generally, and with the exception of perceived discrimination, increases in the value of these predictors should yield higher rates of participation. The last two control variables are age group and gender, two variables that are common to both the treatment and outcome” portions of our model. Consistent with the recurring finding that women and men report different levels of political activism (Burns et al., Reference Burns, Lehman Schlozman and Verba2001) we create a binary gender variable and code it so that 1 = female and 0 = otherwise. Similarly, our binary age variable is based on Elan Hope and Robert Jager’s (2014) operationalization, which distinguishes respondents who are under the voting age from those who are at least eighteen years old.
Recall that we expect rap consumption to moderate the impact of rap critique on political participation, for this is the logic behind our consumption ➜ critique ➜ participation hypothesis. As the formula above shows, we make predictions regarding the participatory influence of rap critique given light consumption levels, and we will observe how the relationship between critique and participation changes given heavy consumption. There should be no relationship between these variables among young Blacks who do not frequently consume rap. However, the association should be positive and statistically significant for the African Americans who are heavy consumers. This prediction assumes the null hypothesis that the effect of rap critique on political participation does not vary depending on rap consumption levels.
FINDINGS AND DISCUSSION
To analyze the impact of rap music consumption on the relationship between rap critique and political participation, we display the results of probit (for voter turnout) and Poisson (for civic activism) regression models with endogenous treatment effects. In experimental research, investigators randomly assign subjects to treatment groups so that characteristics across groups will be similar, and any remaining differences after random assignment stem from the intervention (Campbell and Stanley, Reference Campbell and Stanley1963). To mimic the process of random assignment, models with treatment effects estimates reweight observational data to achieve experimental-like balance in results. Reweighting yields treatment groups of comparable sub-sample sizes, along with “balanced” covariates (i.e., the weighted distribution of each predictor in our outcome model should look comparable, if not identical, for the “untreated” and “treated” Black youths). In the bottom of Table 2, we see that the weighting scheme evens out the once asymmetrical number of participants across the levels of our rap consumption variable. We check for balance using a homogeneity of treatment effects test (see Angrist Reference Angrist2004). This method depends on balancing the two treatment groups (light and heavy rap consumers), and we know that we accomplished this goal because we cannot reject the null hypothesis that the covariates are balanced across groups (chi-squared = 2.56; p = .99).Footnote 23
We expect rap critique to exert a conditional influence on political involvement: critique enhances participation among young Blacks who listen to large amounts of the music; at lower levels of rap consumption, we should observe no effect of critique on participation. In conventional regression analysis, we would include an interaction term between rap consumption and rap critique to test this claim. The approach used here, however, fits separate regression lines for the light and heavy rap consumers, which also handles these differential effects of rap critique on participation. The critique coefficients in the “untreated group” portion of the model report the expected impact of this variable on participation levels if BYP respondents were light consumers. Likewise, the estimates in the “treated group” represent the average impact of rap critique on participation if those same respondents were heavy consumers.Footnote 24
The first and second rows of coefficients reveal that rap critique is not a predictor of turnout, for the influence of this variable falls short of statistical significance in the outcome models for the untreated (probit estimate = -.01; standard error = .06; p > .10) and treated (probit estimate = .03; standard error = .05; p > .10) groups. Critique does, however, have the anticipated association with our measure of non-voting participation. Rap critique has a positive and statistically significant impact on the activism of Black youth who are heavy consumers (Poisson estimate = .07; standard error = .03; p < .05). But critique loses its predictive power among light consumers (Poisson estimate = 0.03; standard error = .05; p > .10).Footnote 25
Table 2 confirms the presence of group differences in the change in civic activism associated with a unit-shift in rap critique.Footnote 26 Because critique has no effect on turnout, our focus henceforth will be on its impact on civic activism. To illustrate this particular result, we sort the expected number of participatory acts by levels of rap critique, holding all other variables at their relevant central tendencies (means for continuous variables, modes for dichotomous variables, and medians for ordinal variables). The dots in Figure 1 report the simulated values, and the capped vertical lines represent 95% confidence intervals. A shift in rap critique (from its minimum to its maximum value) contributes to a rise in non-voting participation. This pattern holds true for both “light” and “heavy” consumption levels, but the positive association between critique and participation is strongest for the Black youth in the “high exposure” quasi-treatment. Moving across the full range of the rap critique variable raises heavy consumers’ expected number of activities from roughly .53 to approximately 2.28, an increase of 1.75.
How Music Consumption Shapes the Influence of Rap Critique on Non-Voting Participation (Civic Activism)
Source: 2005 Black Youth Project (BYP)’s Youth Culture Survey (Oversample of African Americans [N = 635]).
Notes: Data points are the expected number of participatory acts (hollow dots) with 95% confidence intervals (capped vertical lines), simulated using Stata’s margins command, and measured across values of the rap critique variable while holding other covariates at their means (for binary and continuous variables), medians (for ordinal variables).
While it may seem like an underwhelming result, this increase is both statistically and substantively meaningful. Considering the metric of our dependent variable, the finding in Figure 1 translates into a young African American taking part in nearly two additional political acts. Again, the “theoretical” range of our civic activism scale is from zero to eleven, but the “observed” range goes from zero to seven. Therefore, the effects we uncover in Table 2 should be interpreted based on seven, rather than eleven, political acts. To put this effect size in more relatable terms, a shift of two activities on a seven-point scale represents nearly a 30% increase. Since civic activism is naturally low—the mean score on the non-voting participation index is 1.09 (approximately one civic act), and the modal score on this index is zero, which means most respondents abstain from politics—our results show that rap critique can double activism levels among a certain segment of the Black youth population. In several respects, ours is a conservative test of the consumption ➜ critique ➜ participation hypothesis. For example, the conditional impact of rap critique on civic activism would be even larger had we not weakened it by controlling for alternative types of participatory resources—in this case, the predictors measuring whether a respondent took a civics class, was (or currently is) relying on government assistance, or has college-educated parent(s).Footnote 27 Given that young Blacks tend to lag behind their older and White peers when it comes to conventional modes of political participation (Cohen Reference Cohen2010), our choices regarding sample selection and model specification make the job of observing the effect of rap music on participation especially difficult.
CONCLUSION
Overall, our findings lend qualified support to the consumption ➜ critique ➜ participation hypothesis. Heavy rap consumers are more inclined than their light-consuming counterparts to translate their music critiques into civic activism. Although we demonstrate that skepticism can stimulate Black youths’ non-voting participation, readers should guard against the temptation to (mis)interpret this as an anti-rap conclusion. To use William Gamson’s (Reference Gamson1968; Reference Gamson, Abcarian and Soule1971) language about “dissident” (low trust, low efficacy) versus “assured” (high trust and efficacy) citizens, we are not implying that being a rap curmudgeon (a low consuming heavy critic) makes people civically active. Rather, the participatory impact of critique exists only among connoisseurs—those Blacks who can enjoy the music (as evidenced by strong consumption rates) while demanding a higher caliber of rap (indicated by their willingness to voice skepticism). Critique has no impact on turnout, however, and this non-finding holds true no matter how frequently an individual consumes rap. This lack of a “turnout effect” might be a function of the cynical manner in which many rappers discuss voting, or maybe it is diagnostic of a waning sense of political efficacy among Black youth. Either way, the absence of a relationship between critique and turnout makes sense when one considers the music’s non-mainstream nature, and, more importantly, its emphasis on grassroots empowerment (Kelley Reference Kelley1996; Kitwana Reference Kitwana2002, Reference Kitwana2004; Tyson Reference Tyson2002, Reference Tyson2006). Despite the mobilization efforts of prominent rap artists, producers, and organizations, young Blacks who embrace Hip Hop tend to eschew electoral politics in favor of less traditional modes of participation—perhaps because conventional modes of activism appeal less to people who tend to feel “outside” of politics due to a history of disenfranchisement and marginalization (Bonnette Reference Bonnette2015; Spence Reference Spence2011).
To reiterate, our research is not a repudiation of rap. Although we discuss some of the music’s perceived negativity, of course, only a portion of rap’s messages are negative. Instead, our research is a celebration of the possibilities for Black youth. In this regard, we agree with the work of Cohen (Reference Cohen2010). It is clear when reading Democracy Remixed that Cohen celebrates Black youth, and she certainly believes that Black youth have the potential to create a strong future for communities of color. In creating that strong future, it is very important that positive messages of various types and from various sources take center stage in the learning process for young Blacks. Therefore, to fight against potential miseducation, sufficient consciousness-raising factors (i.e., caring teachers, supportive loved ones, and quality peer networks, etc.) need to be in place for Black youth. These factors protect young African Americans from negative messages, as these youth are more easily able to “call out” any source of information on its problematic content. In other words, proper inculcation into Black culture, supervised by one’s teachers, mentors, family, etc., can train young Blacks to see more clearly the discrepancies between the better worlds they desire and the harsher ones they may experience that are sometimes depicted in rap. Recognizing this contrast need not prejudice young people against any form of popular culture, but it can motivate those who consume music to be critical of certain aspects of it.
Our findings in regard to civic participation demonstrate that we actually need rap in our political world. More importantly, we need thoughtful young folks to listen to rap. Here is a rationale for why. One of the critiques launched at young people who abstain from politics is that “you can’t complain about the outcomes in a process in which you don’t take part.” This old saying has constantly had relevance, and it is relevant here when thinking of getting young people engaged. It seems that some young Black people might not be motivated enough to participate in higher levels of community action unless they listen to and observe their social world with a critical eye. And a critical eye is a great thing to have. Our social world has recently become much more increasingly untrustworthy. We are now seeing a new heyday of something that had moved from the forefront of American life with the decline of yellow journalism in the late 19th century and early 20th century—fake news. Fake news is a recent term to describe stories that are just not true (Ember Reference Ember2017). The point of fake news is to deceive a populace. Thus, a critical eye is something that we want Black youth to possess in bountiful supply. This is obviously not a new concern, as Black critics of American society have long argued that mainstream American institutions can be quite faulty in their treatment of Blacks themselves and the subjects that Blacks hold dear (Walton et al., Reference Walton, Smith and Wallace2017). That is why indigenous Black media sources have been available to the Black population since 1827 and the founding of the first Black-owned newspaper, Freedom’s Journal.
Many observers may think of rap as nothing more than entertainment, but entertainment is only one aspect of the world of rap and Hip Hop. We noted earlier in this paper that Chuck D saw rap as a “Black CNN.” In a world with fake news all around them, Black youth need “rap as CNN” more than ever. Black youth need clarity and truth-telling concerning our world, and rap can be a great resource in this search for truth. Black youth can critically utilize and learn from that resource. In addition to helping them “stay woke,” this critical engagement enables our Black youth to both cultivate their imaginations while making society better for all of us. The need for (and importance of) such critical engagement connects strongly with our major research finding that “imagining a better world”—in this case, by envisioning improvement in rap’s messages—may inspire some young African Americans to “build a better world” through enhanced community involvement.
APPENDIX 1: DESCRIPTION OF VARIABLES
Dependent Variables
Voter Turnout (D3): “Please tell me if you voted in a national or local election during the last 12 months.” {1 = yes, 0 = otherwise; proportion = .49 (N = 415)}
Civic Activism: We create an additive index from the 11 non-voter participation items below {Kuder–Richardson coefficient of reliability (KS20) = .69; mean = 1.09; standard deviation = 1.42}. Each item begins with the following prompt: “Tell me if you have done any of the following in the last 12 months. Have you…” {1 = yes, 0 = otherwise}
• D4: Contacted a public official or agency?
• D5: Signed a paper or e-mail petition?
• D6: Attended a protest meeting, demonstration, or sit-in?
• D7: Participated in a boycott?
• D8: Engaged in “buycotting” (buying a certain product or service because you like the social or political values of the company that produces or sells the product)?
• D9: Been active in or joined a political group?
• D10: Written and sent an email or written a blog about a political issue, candidate or political party?
• D11: Given money to a candidate, party, or political issue?
• D12: Worked or volunteered on a political campaign for a candidate or a party?
• D14: Worked with the people in your neighborhood on a political issue or problem?
• D15: Written an article or letter to the editor about a political issue or problem?
Theoretically-Central Independent Variables
Rap Music Critique: “I’m going to read you a number of statements about rap music videos. Please tell me whether you {1 = strongly agree, .5 = agree, 0 = neither agree nor disagree, -.5 = disagree, or -1 = strongly disagree}.” We create an additive index from these items (Cronbach’s alpha = .73) that ranges from -5 (strong rap approval) to +5 (strong rap critique).
• B8: Rap music videos contain too many references to sex.
• B9: Rap music videos should be more political.
• B10: Rap music videos have too many references to violence.
• B11: Rap music videos portray Black women in bad and offensive ways.
• B12: Rap music videos portray Black men in bad and offensive ways.
Moderator Variable
Rap Music Consumption (Continuous Variable Used in Interactive Models): We create an additive index of the items below, re-scaled to fit a 0 to 1 interval (Cronbach’s alpha = .75).
• B1: “How often do you listen to rap?” {1 = every day, .8 = several days/week, .6 = once/week, .4 = once or twice/month, .2 = rarely, 0 = never}
• B5: “On average how often do you watch rap music programming on television?” {1 = every day, .8 = several days/week, .6 = once/week, .4 = once or twice/month, .2 = rarely, 0 = never}
Rap Music Consumption (Dichotomous Variable Used in “Regression with Treatment Effects” and “Heckman Selection” Models): We generate a new variable using a median split of the re-scaled “listen to rap” and “watch rap videos” items mentioned above, with values lower than the medians for both variables (.8 and .6, respectively) signifying “light rap consumption” and scores exceeding the median indicating “heavy rap consumption.”
Control Variables
Parents’ Education Levels: Respondents get a score of “1” if one or both of their parents have at least a college education; respondents whose parent(s) fall short of this educational attainment get scores of “0” {proportion = .53}. The items upon which we build this variable appear below.
• L15: What is the highest level of schooling [your father/the person who acted as a father to you] completed?
• L20: What is the highest level of schooling [your mother/the person who acted as a mother to you] completed?
Dependency upon Government Assistance (L26): “[AGE 18-25: While growing up was] [AGE 15-17: Has there ever been] a time when the people you lived with received any form of public assistance from the government like welfare, TANF, WIC, public housing, section eight, Medicaid, or food stamps?” {1 = yes, 0 = otherwise; proportion = .56}
Civics Curriculum (A10): “Have you ever taken [a high school / an] American government or civics course?” {1 = yes, 0 = otherwise; proportion = .60}
Political Efficacy: Degree to which respondents agree with the following four statements {Cronbach’s alpha = .55; mean = .37; standard deviation = 1.43}. C12, C13 and C15 are coded so that -1 = “strongly disagree,” -.5 = “disagree,” 0 = “neither agree nor disagree,” .5 = “agree,” and 1 = “strongly agree.” C14 is reverse coded so that higher values signify greater compassion among politicians.
• C12: I believe that by participating in politics I can make a difference.
• C13: I have the skills and knowledge necessary to participate in politics.
• C14: The leaders in government care very little about people like me.
• C15: The government is pretty much run by a few big interests looking out for themselves and their friends.
Parent’s Interest in Politics (D1): “Okay, now think about your parent or the person who raised you who was most interested in politics. How often did that person follow what was going on in government and public affairs?” {1 = most of the time, .66 = some of the time, .33 = only now and then, or 0 = hardly at all; proportion = .67}
Experiences with Discrimination: Seven-item additive index {responses for each item range from 1 = strongly agree to 0 = neither agree nor disagree to -1 = strongly disagree; Cronbach’s alpha = .70; mean = .48; standard deviation = 2.43}
• D24: “Generally, I feel like a full and equal citizen in this country, with all the rights and protections that other people have.”
• D25: “The government treats most immigrants better than it treats most Black people born in this country.”
• D26: “In the United States, everyone has an equal chance to succeed”
• J2: “It is hard for young Black people to get ahead because they face so much discrimination.”
• J3: “Sometimes young Black people have to act White to get ahead.”
• J4: “On average, Black youth receive a poorer education than White youth.”
• J5: “On average, the police discriminate much more against Black youth than they do against White youth.”
Mobilization/Recruitment:
• Proselytize about politics (D13): Does respondent discuss politics with others? {1 = yes, 0 = otherwise; proportion = .68}
• Neighborhood efficacy (K4): This item asks a respondent to agree/disagree with a statement regarding the ability of people in her neighborhood to get the government to respond to their needs {“strongly agree” to “neither agree nor disagree” to “strongly disagree”; we code this item to fit a -1 to +1 interval (higher scores represent greater perceived government responsiveness, with 0 as the middle category); mean = -.09; standard deviation = .55}
Age (AGE_GROUP): {1 = over 18 years old, 0 = otherwise; proportion = .65}
Gender (MAIN_SEX): {1 = female, 0 = male; proportion = .55}
Church Attendance (E5): How often does respondent attend a church/place of worship? {responses range from “never” to “daily”; we code this item to fit a 0 to 1 interval (higher values represent frequent church attendance); mean = .32, standard deviation = .33).
Employment Status: {1 = currently employed (part time or full time), 0 = otherwise, proportion = .52}
APPENDIX 2: REGRESSION WITH ENDOGENOUS TREATMENT EFFECTS
Consistent with the reasoning behind experimental analyses, a “treatment-effect” estimate refers to the impact of a binary grouping variable (1 = subject receives treatment, 0 = otherwise) on an outcome of interest (for details about these and related models, see Cattaneo Reference Cattaneo2010; Green 2009). Applying this terminology to our observational data, the “treatment” (Black youth’s exposure to rap music) interacts with our theoretically-central predictor (rap music critique) to affect our set of dependent variables (measures of political participation). Ideally, we would explore the impact of heavy rap consumption on the critique ➜ participation relationship in a laboratory setting, using a within-subjects research design in which we observe a Black youth’s participation level when s/he is “untreated” (i.e., not heavily exposed to rap music) and again once s/he is “treated” (i.e., given heavy rap exposure). We would record these observations under identical circumstances so that the only difference is the presence or absence of treatment, and then average the difference between treatment groups across all subjects to infer the impact of rap consumption on participation. Unfortunately, such analyses are not possible with BYP data because we cannot randomly assign individuals to receive varying levels of rap music exposure.
Respondents decided their music preferences long before completing the survey, so we have no control over their consumption patterns. The design of the BYP allows us to compare Blacks’ varying rap exposure levels while controlling for observed differences between these respondents. However, since heavy consumers might differ in important and unmeasured ways from light consumers (see Table 1), we worry about the potential for endogeneity: unobserved factors—and not the treatment itself—might be producing the differences we detect between these groups of young Blacks. This is especially problematic if these unobserved factors are associated with the characteristics affecting treatment assignment. Additionally, the process by which BYP data were generated produces a unique “missing data” problem: we only observe a respondent receiving one level of the treatment (light vs. heavy rap consumption); it impossible to examine the rap critique and participation levels of someone who both has and has not been assigned to the “heavy exposure” condition.
As Rubin (Reference Rubin1974) notes, treatment-effect estimates account for the potential outcome(s) that each individual would obtain under different treatment levels, the process by which subjects are assigned to levels of the treatment, and the dependence of the potential outcomes on the treatment-assignment process (see also Holland 1986; Wooldridge Reference Wooldridge2010, chapter 21). Specifically, treatment-effects estimates are derived from a system of equations, with a binary variable representing the treatment condition entered directly into the “outcome equation,” and the outcome variable being observed for both levels of the treatment. Stated formally, these models generate data in which y i is the observed outcome for respondent i, t j is the treatment variable that can take on j levels, x i is a vector of covariates presumed to influence the outcome, and w i is a vector of covariates that affect the treatment assignment.Footnote 28 Such a model would specify that the observed outcome variable (y) is y 0 when t = 0 and y 1 when t = 1:
$y = \left( {1 - t} \right)y_0 + ty_i $This suggests the following functional form for y 0 and y1:
$y_0 = x^' \beta _0 + \varepsilon _0 $
$y_1 = x^' \beta _1 + \varepsilon _1 $where β 0 and β 1 are coefficients to be estimated, and ϵ 0 and ϵ 1 are residual terms that are exogenous (not correlated with x or w). This modelling strategy separates each potential outcome of the dependent variable into two components: the model’s prediction (x t) and the unobservable error term (ϵ t).
We use the notation below to describe the treatment model, which characterizes the process by which respondents in the BYP are assigned to levels of rap music exposure:
$t = \left\{ {\matrix{ {1,} &amp; {if} &amp; {w^' \gamma + u_i &gt; 0} \cr {0,} &amp; {if} &amp; {otherwise} \cr } } \right.$where γ is a coefficient vector, and u i is a residual term that is unrelated to either x or w. Like in the outcome model, the treatment assignment process divides into its predictable component (w’γ), and its unobservable component (u i). Because t determines how the data on y 0 and y 1 are “missing,” treatment effects models address the missing data problem (that respondents are present in either treatment condition, but not both) by providing estimates of respondent’s “potential outcomes.” Consider an African American that did not receive treatment so that we observe y 0. What would y 1 be for this subject if s/he were heavily exposed to rap music? For this young person, y 1 is the potential outcome or “counterfactual.” Likewise, if we observe y 1 for a Black youth that received treatment, then y 0 would be the counterfactual outcome for that respondent. Because they involve “switching regressions” (Quandt Reference Quandt1958, Reference Quandt1972), treatment-effects estimates overcome the challenges of calculating such counterfactuals using survey data.
APPENDIX 3: ROBUSTNESS CHECKS
Heckman Selection Models of the Impact of Rap Music Critique on Political Participation
Interactive Models of the Impact of Rap Music on Political Participation
