“Expressive Cool” and the Paradox of Black and White Males’ Neighborhood Socialization Toward Education

Abstract

This study explores how linkages between adolescents’ educational attitudes and achievement vary according to race, expressive culture, and neighborhood collective socialization qualities. Specifically, the study examines (a) racial differences in how males’ educational attitudes relate to their academic performance (i.e., “attitude–achievement paradox”); (b) how the attitude–achievement paradox varies according to Black and White males’ expressive culture; and (c) the relation of collective socialization to racial differences in expressive cool, educational attitudes, and behavior. Using Maryland Adolescent Development in Context Study (MADICS) data, I find that an attitude–achievement paradox among African Americans disappears when neighborhood collective socialization is considered; that expressive cool seems to have a stronger connection to adolescents’ achievement ideology rejection, and very little to their grade point average (GPA); and that neighborhood collective socialization decisively accounts for racial gaps in GPA. The concluding discussion considers why African Americans’ adherence to achievement ideologies fails to shield their GPAs from neighborhood socialization risks to the extent it does so for White males.

Keywords

neighborhood context education racial/ethnic identity culture

Introduction

Whether adolescents believe that hard work in school pays off with good grades and success in life (or “achievement ideologies”) has been a longstanding interest of researchers for good reason (Gurin & Epps, 1975; MacLeod, 1987; Willis, 1977). If adolescents’ rejection of dominant achievement ideologies is inspired by their social background, then unequal backgrounds may lead to educational inequality if achievement ideology rejection (AIR) causes lower achievement. Although the educational beliefs and achievement of disadvantaged children are frequent worries, these concerns also extend to gender, racial, and residential groups. The intersection of race, gender, and residency may embody multiple risks to the achievement of African American males in particular, because their gender performance and dispositions toward education are cultivated, in part, by observing their neighboring adult counterparts. By socializing the next generation, neighboring adults could contribute to the social reproduction of the dispositions toward education that enable social stratification. Pursuant to these possibilities, this study aims to (a) explore racial differences in how male adolescents’ educational attitudes relate to their achievement (i.e., “attitude–achievement paradox”); (b) examine how attitude–achievement correspondences vary according to Black and White males’ “expressive cool” gender performances; and (c) reveal whether neighborhood collective socialization accounts for racial differences in males’ expressive cool, attitudes and achievement.

Attitudes and Achievement According to Race, Gender, and Context

Racial stratification in society is an essential part of this discussion because it could unequally inspire among racial groups views that are critical of achievement ideologies. African Americans, for example, have been shown to endorse achievement ideologies until they are asked whether those ideologies apply to people like them, a distinction Mickelson (1990) notes as having “abstract” versus “concrete” attitudes. Others argue that African Americans question whether academic success is culturally appropriate for them because they may associate academic success with “acting White” (Fryer & Torelli, 2010; Ogbu, 1987). But skepticism has grown regarding the thought that African Americans are more likely than Whites to reject the importance of schooling (Ainsworth-Darnell & Downey, 1998; Harris, 2006), and face stiffer social penalties or fear stigmas for high academic performance (P. J. Cook & Ludwig, 1997; Tyson, Darity, & Castellino, 2005). Yet other studies conclude that African Americans can have both critical attitudes and high achievement, and that a covariation between African Americans’ educational attitudes and behavior should not be assumed (Akom, 2003; O’Connor, 1997).

Regarding gender, it has long been thought that males’ beliefs about achievement may be influenced by gendered social hierarchies and identities (A. A. Ferguson, 2001; Willis, 1977). Willis’ (1977) classic study, for instance, finds male adolescents reinforce masculine norms within their peer group that appear inconsistent with school success. However, studies of gender–race intersections leave questions unsettled about the ability of achievement ideology beliefs to shield different groups, equally, from the structural influences that undermine achievement. For example, White males in MacLeod’s (1987) study question linkages between schooling and employment opportunity and exhibit hyper-masculine behaviors. Their African American peers in contrast endorse achievement ideologies, are not hyper-masculine, but ultimately have greater difficulty in the labor market. One interpretation of these findings is that race limits the social returns of African American males’ positive educational attitudes even when compared with less optimistic, hyper-masculine, White adolescents. Another interpretation is that our gendered society rewards masculinity with social advancement, and that resistance to dominant achievement ideologies carries fewer sanctions when viewed as a part of adolescents’ masculinity.

That MacLeod (1987) finds these race–gender differences among boys who reside in a low-income neighborhood is of particular note. Other qualitative studies describe how achievement ideologies manifest among adolescents living in mixed-income and middle-class neighborhoods (Harding, 2010; Ogbu, 2003; Pattillo-McCoy, 1999). Although both experimental (O. Johnson, 2012b) and inferential studies (O. Johnson, 2008) have shown that the relationship between residency among the middle class and the achievement of Black males is not always a positive one, Harris (2006) shows that African Americans in a predominantly middle-class location often hold optimistic educational attitudes. However, none of the aforementioned studies systematically consider if neighborhood socialization opportunities qualify how Black and White adolescents’ expressive qualities relate to their educational attitudes and achievement.

Expressive Cool as Gender-Role Performances Within Contexts

I consider “expressive cool” as a type of masculine gender performance because research contends its elements resonate with male youth culture, are informed by neighborhoods and broader opportunity structures, and may be linked to adolescents’ academic orientations (Dance, 2002; R. R. Ferguson, 2001; Majors & Billson, 1993; Oliver, 2006; Pattillo-McCoy, 1999). Expressive cool extends from Majors and Billson’s (1993) recognition that Black males’ “cool pose” (i.e., a ritualized performance of masculinity to project pride, strength, and control in response to racial marginalization) has an expressive (i.e., externalizing) and non-expressive form (i.e., introversion). I next review key expressive cool constructs, their connections to neighborhood processes, and how they may relate to educational attitudes and behavior.

Code-Switching

I investigate code-switching as an aspect of adolescents’ expressive cool because it indicates knowledge of different expressive scripts and an ability to perform them. Neighborhood socialization may be a process in which individuals are taught to relate particular gender-role performances to the appropriate environmental circumstance and perform those behaviors when cued, just as urban residents have been found to switch between “street” and “decent” orientations (Anderson, 1999). In addition, neighborhoods may provoke a different gender-role performance than elicited at home or school; in different times of the day; and even across the different social groups within them (e.g., peers and strangers). Hence, gender performances are relational and fluid (Kimmel, 1986) and have an equally dynamic contextual basis. Although code-switching has a special relevance to African Americans due to their history of negotiating conflicting self and public images (DuBois, 1903), and desire to advance socially within dominant and non-dominant cultural contexts (Carter, 2003), this study will explore whether White teens within contexts heavily populated by African Americans also code-switch to socially navigate those environments.

Hype and “Swag”

I also consider male adolescents’ style of expression and appearance because qualitative research shows that “outfits, language, and walk” are important aspects of social organization in middle-class neighborhoods (Pattillo-McCoy, 1999). In addition to appearance, “acting Black” may be a vital component of adolescents’ hype, swagger, and ultimately their masculinity, because inasmuch as acting Black reflects being hard, tough, and having street credibility (Dance, 2002), it is as much a gender “burden” for males as it may be a racial one. The burden may be heavier for African Americans, and explain why Abreu, Goodyear, Campos, and Newcomb (2000) find a higher correlation between ethnic belonging and toughness among African Americans than for Whites and Latinos. Suggested racial differences find support in claims that “Blacks take their style ‘more seriously’ than Whites” and reside in contexts where “the chances that a stylistic faux pas will have negative consequences are higher” (Pattillo-McCoy, 1999, p. 119). A corollary concern is that racial variation in males’ expressive culture underlies racial differences in standardized test scores (R. R. Ferguson, 2001) and academic engagement among the middle class (Ogbu, 2003; Thernstrom & Thernstrom, 2003). However, White teens in near-urban contexts may be equally susceptible to these behaviors because other studies suggest that the burden of “acting White” is a concern for them too (Tyson et al., 2005), and that individuals become increasingly likely to perform a certain behavior as its prevalence in an area increases (Crane, 1991).

Racial Inopportunity

Other research claims that racial inopportunity may be a precondition that activates male adolescents’ expressiveness, and is therefore important to consider. For example, Majors and Billson (1993) argue Black male expressiveness extends from the reality that opportunities to achieve a healthy masculine identity are more fully available to White than Black males. African American males may therefore enact externalizing behaviors to secure the respect that racial marginalization keeps beyond their reach (Majors & Billson, 1993). To this point, research shows that perceptions of racial discrimination in school and the labor market are likely to encourage teens to pursue other ways of gaining social acceptance, resulting in lower achievement (A. A. Ferguson, 2001; Smalls, White, Chavous, & Sellers, 2007), especially in under-resourced neighborhoods (Caughy, Nettles, O’Campo, & Lohrfink, 2006).

Masculinity

Perceived masculinity is also an important dimension of adolescents’ expressiveness, because how males see themselves and are perceived by others contribute to their gender identity and in turn their behavior. These behaviors tend to be explored within research as problematic responses to women and femininity (O’Neil, Helms, Gable, Davis, & Wrightsman, 1986; Reidy, Shirk, Sloan, & Zeichner, 2009), diverse sexualities (Kimmel & Mahler, 2003; Klein, 2006), society’s gender norms (Eisler & Skidmore, 1987; Lazur & Majors, 1995), and as manifestations of risk and vulnerability in contexts (Cassidy & Stevenson, 2005; Seaton, 2007; Spencer, 2001). Within schools, research argues that perceived masculinity informs adolescents’ peer group memberships (MacLeod, 1987), non-conformance with school expectations and higher rates of discipline (A. A. Ferguson, 2001), as well as modest aspirations within mixed income areas (Harding, 2010).

Yet, masculinities may play a constructive part in adolescents’ social and academic development. For example, Kirkland and Jackson (2009) find that Black adolescents’ use of language and styles that reflect their race and gender simultaneously facilitate a masculine cool self-concept, and an efficacy about the application of those expressive elements to successful literacy practices. Second, role strain, which hypothesizes males experience stress and problematic outcomes related to society’s gendered expectations of them (Eisler & Skidmore, 1987; Lazur & Majors, 1995), may prompt males to become critical of hegemonic masculinity, its linkage to achievement ideologies, and assume academically advantageous masculine counter-identities. In these ways, masculinity is possibly related to higher grades.

Neighboring Adults and the Socialization of Males

I examine adolescents’ dispositions toward education within neighborhoods because their understanding of opportunity structures, racial inequalities, and their chances of meeting male role expectations are possibly informed by observing male adults in their vicinity. The influence neighboring adults have on children other than their own is known as neighborhood collective socialization (Ainsworth, 2010; Jencks & Mayer, 1990), and neighbors’ informal social control of adolescent behavior (Sampson, 1997). Given that expressive cool is crafted in part by teens’ emulation of adult males, failing to account for their absence, presence, and characteristics within neighborhoods may lead expressive cool dimensions to appear more strongly tied to educational attitudes and achievement than they should. This study therefore considers the following theoretically relevant neighborhood attributes.

Male Joblessness

Local employment structures bring about the collective socialization of adolescents in various ways. First, the creation of norms that typically arise in response to work may be hindered in neighborhoods with high male unemployment (Wilson, 1996), including the modeling of professions, success, and the provider roles men serve in families. Second, joblessness within neighborhoods may lead teens to question the availability of opportunity and the economic returns to educational success (MacLeod, 1987), especially if increased competition for jobs has limited young people’s ability to find work (Newman, 1999). Alternatively, being shut out of the labor market might give teens more time to study and incentive to pursue more education to increase their employability (Ginther, Haveman, & Wolfe, 2000). Hence, dismal opportunity prospects could lead to higher grades.

College Graduates

A key indicator of neighborhood collective socialization is the presence of adults with an educational background capable of assisting teens’ educational advancement (O’Connor, 2000). Research hypothesizes that the neighborhood presence of highly educated adults increases the norms of college-going among peers within schools (Rosenbaum, 1995), and the pressure on local schools to provide an educational experience consistent with college-entry requirements (Jencks & Mayer, 1990). Having fewer college-educated neighbors limits the number of males teens can look up to as examples of educational success or rely on for sponsorship through educational systems.

Family Structure

Urban ethnographies have elucidated the important role “Old Heads” or adult males have as mentors and supervisors for young males in Black communities, and how their absence weakens a neighborhoods’ ability to collectively regulate adolescents’ maturation (Anderson, 1999; Duneire, 1994). It follows that female-headship at the neighborhood level may present consequences for adolescent males, because an absence of co-resident fathers could reflect a reduction in the number of male adults to engage adolescents in the area (W. E. Johnson, 2001; Wilson, 1996). Evidence suggests that neighborhood female-headship rates are negatively related to the educational attainment of Black males, but not White males (Duncan, 1994; Madyun & Lee, 2010). This study will expand this literature by estimating the influence of neighboring co-resident males on teens’ educational beliefs and grades.

Neighborhood Location

I also consider the location of adolescents’ neighborhoods within a broader geography of opportunity, because African American neighborhoods frequently lie between areas of greater risk and privilege (Pattillo-McCoy, 1999; Sampson, Morenoff, & Earls, 1999). In contrast to the solidly middle-class communities that Lacy (2007) explores within Prince George’s County, its neighborhoods that border Washington, DC, closely resemble the middle-class enclaves in Pattillo-McCoy’s study, which are situated near lower income African American areas. Pattillo-McCoy shows that adolescents’ proximity to disadvantaged areas contributed to their “ghetto trance,” or a greater affinity with expressive cultures that presumably differ from those associated with their higher income social status. Likewise, Harding (2010) suggests that economically heterogeneous neighborhoods provide adolescents models of successful and unsuccessful behavior, competing interpretations of opportunity, and differing standards of success. This study therefore takes into consideration youths’ location relative to Washington, DC, the major city that borders Prince George’s County. Following from the substance of the literature review, this study explores the following research questions:

Research Question 1: Does expressive cool account for racial differences in adolescents’ AIR and grade point average (GPA)?

Research Question 2: Is there evidence of an attitude–achievement paradox where adolescents’ expressive cool is more strongly related to their educational attitudes (AIR) than behavior (GPA)?

Research Question 3: Does neighborhood collective socialization account for racial differences in youths’ expressive cool, AIR, and GPA?

Research Methods

Data

Data for this study come from the Maryland Adolescent Development in Context Study (MADICS), a survey of 1,482 youths of Prince George’s County (http://www.rcgd.isr.umich.edu/pgc/home.htm). Although the perspectives of MADICS participants are of a time that may differ from today’s social climate, there are reasons why these data remain the best data for the proposed study despite their collection from 1997 to 2002. First, Prince George’s County is ideal for testing Pattillo-McCoy’s hypotheses concerning expressive cool and neighborhood effects in predominantly middle class, majority Black, near-urban areas. These sample features would not be achievable in a nationally representative survey, nor do such surveys contain as many measures of adolescents’ context. Second, this study features neighborhood and gender-role socialization processes that have not appeared in previous uses of these data (T. Cook, Herman, Phillips, & Settersten, 2002; Harris, 2006; Harris & Marsh, 2010).

This study utilizes data collected from 11th-grade adolescents, their parents, school personnel, and residential measures taken from the census. The risk of bias posed by school dropouts is minimal because 99% of those surveyed in Grade 8 (Wave 3) were also surveyed in Grade 11 (Wave 4). My analytic sample began with 524 males (approximately 61.4% African American, 29.1% White American, 9.5% Other racial/ethnic groups). Adolescents of other racial/ethnic groups, 50 in all, were eliminated. After eliminating those adolescents who had changed residences, I arrived at a final sample of 406 males (260 Black, 146 White) representing 143 census tracts of 183 census tracts within Prince George’s County.

One complication with these data is that the percentage of missing values for some variables reaches 11.8%, making listwise deletion procedures unacceptable. Rather than sacrifice the variation within each variable by inserting the variable mean where data are missing, I use multiple imputation methods to recover missing values. Multiple imputation uses values from other variables to assign a likely value to a missing case and repeats this process 5 times. The end product is five additional unique iterations of the original sample with no missing values. Statistical analyses generate a pooled estimate of all five samples. As seen in column 4 in Table 1, the pooled means of analysis variables appear similar to those of the original data set.

Table 1.

Descriptive Statistics for Full (N = 406), Black (n = 261), and White (n = 145) Samples.

Variables	Definitions	Eigenvalue	Alpha	Orig. M	Pooled mean	Black mean	White mean	Minimum	Maximum	Orig. SD
AIR factor (disagree–agree)		3.000	.78	1.662	1.803	1.755	1.889	0.00	5.040	1.000
Components	Schooling not so important for kids like me			2.093	2.194	2.165	2.246	1.00	5.00	0.978
	I don’t really care about school			2.213	2.336	2.258	2.477	1.00	5.00	1.083
	Studying in school rarely pays off with job			2.536	2.619	2.633	2.596	1.00	5.00	1.097
	Homework is a waste of time			2.392	2.499	2.397	2.684	1.00	5.00	1.064
	School is a waste of time			2.059	2.159	2.062	2.335	1.00	5.00	1.054
	We treated unfairly at work, no matter how educated			2.758	2.812	3.042	2.399	1.00	5.00	1.033
GPA	Final GPA			3.592	3.578	3.525	3.669	2.70	4.69	0.311
Income	Total family income (1-21)			13.490	13.960	12.71	14.83	2.00	21.00	5.392
Low SES	Family income quintile 1, US$0-US$34,999K			0.189	0.207	0.253	0.110	0.00	1.00	0.393
Middle low SES	Family income quintile 2, US$35-US$54,999K			0.185	0.215	0.226	0.172	0.00	1.00	0.389
Middle SES	Family income quintile 3, US$55-US$69,999K			0.185	0.209	0.195	0.297	0.00	1.00	0.389
Middle high SES	Family income quintile 4, US$70-US$89,999K			0.158	0.185	0.180	0.179	0.00	1.00	0.365
High SES	Family income quintile 5, US$90K and over			0.234	0.185	0.146	0.241	0.00	1.00	0.424
Friends’ AIR	Friends say you won’t get good job (1 = none, 5 = all)			1.650	1.720	1.67	1.62	1.00	5.00	0.971
Parent education	Highest grade of primary care giver in grade years			14.160	14.180	13.90	14.69	9.00	21.00	2.452
Dad’s interest	Dad takes interest in my activities (1 = never, 5 = always)			2.722	2.743	2.839	2.572	1.00	5.00	1.343
Race	What do most people call your race (0 = White, 1 = Black)			0.643	0.643	1.00	0.00	0.00	1.00	0.479
Age	Age at 11th grade start in years			17.46	17.46	17.48	17.43	16.26	19.68	0.487
Single parent	Single parent head of household (0 = no to 1 = yes)			0.338	0.349	0.414	0.200	0.00	1.00	0.474
Hype factor (not true–very true)		2.082	.70	0.820	0.916	0.959	0.981	0.00	5.28	1.000
Components	Parent says youth talks too much			0.220	0.280	0.31	0.24	0.00	2.00	0.486
	Parent says youth brags or boasts			0.390	0.460	0.42	0.53	0.00	2.00	0.583
	Parent says youth swears/uses obscene language			0.210	0.280	0.24	0.35	0.00	2.00	0.483
	Parent says youth shows off/clown			0.420	0.500	0.52	0.45	0.00	2.00	0.598
Masculine identity factor (not at all–very)		2.300	0.72	3.600	3.082	3.592	2.077	0.00	4.832	1.000
Components	I feel as though I am masculine			5.850	5.750	5.84	5.57	1.00	7.00	1.395
	I look as though I am masculine			5.840	5.740	5.77	5.68	1.00	7.00	1.401
	How important is it that you appear masculine			4.690	4.660	4.92	4.20	1.00	7.00	1.949
	How important is it that you avoid appearing feminine			4.640	4.640	5.01	3.97	1.00	7.00	2.286
Racial inopportunity factor (strongly disagree–strongly agree)		2.430	.81	1.679	1.796	2.079	1.269	0.00	4.374	1.000
Components	Little you can do to avoid job discrimination			2.220	2.270	2.38	1.97	1.00	4.00	0.836
	Because of race, you must do better than other kids			1.980	2.030	2.25	1.46	1.00	4.00	1.081
	Because of race, you will have to work harder			2.150	2.190	2.50	1.69	1.00	4.00	0.965
	Little you can do to avoid discrimination at school			2.25	2.300	2.35	2.21	1.00	4.00	0.857
Acting Black	Friends think it’s cool if I act Black (1 = uncool, 5 = cool)			3.190	3.170	3.49	2.60	1.00	5.00	1.102
Best dressed	Friends think it’s cool if I’m voted best dressed			3.840	3.800	3.92	3.62	1.00	5.00	1.000
Personality change	Can change personality (1 = strongly agree, 5 = disagree)			2.350	2.380	2.26	2.56	1.00	5.00	0.844
Looking good	How important is being good looking (1 = less, 7 = more)			4.150	4.180	4.36	3.80	1.00	7.00	1.646
Collective socialization factor (disagree–agree)		1.886	0.70	3.389	3.306	3.233	3.439	0.00	5.458	1.000
Components	Lots of adults our kids can look up to			3.603	3.569	3.449	3.783	1.00	5.00	0.922
	Can count on adults to tell me of child opportunities			3.213	3.206	3.130	3.342	1.00	5.00	0.991
	Neighbors share similar views about raising kids			3.561	3.548	3.571	3.506	1.00	5.00	0.875
Joblessness	Unemployment in area (1 = no problem, 3 = big problem)			1.809	1.807	1.900	1.635	1.00	3.00	0.802
% Degreed	% of college graduates			25.231	25.412	22.035	31.158	0.00	65.62	12.183
% Female heads	% of female-headed families			21.755	22.437	26.587	15.138	5.53	57.50	13.041
Inner ring	Residency in inner suburban ring (0 = no, 1 = yes)			0.204	0.232	0.279	0.152	0.00	1.00	0.404

Note. AIR = achievement ideology rejection; SES = socioeconomic status; GPA = grade point average.

Constructs and Measures

As it is impossible to include individually all relevant variables in the statistical models, I use Principal Components Analysis (PCA) to reduce the number of correlated covariates into a single factor that expresses the best summary of correlation between three or more variables. Factors were retained as analysis variables if the Kaiser–Meyer–Olkin measure of sampling adequacy was at .600 or higher and the eigenvalue (total variance explained by the factor) exceeded 1.00. Table 1 reports the factors’ principal components, eigenvalues, standardized means, and Cronbach’s alpha.

Study variables reflect adolescents’ expressive cool, social background, contexts, and behavioral outcomes. Beginning with outcomes first, I used six variables to create a factor that reflects youths’ AIR. As shown in Table 1, its principal components come from Mickelson’s (1990) attitude–achievement paradox scale and reflect teens’ concrete attitudes like “schooling is not so important for kids like me,” “I don’t really care about school,” and “studying in school rarely pays off with a good job,” and range from strongly disagree (1) to strongly agree (5). I juxtapose this attitudinal outcome with adolescents’ final GPA to assess possible attitude–achievement paradoxes.

I also create three factors to represent expressive cool measures (see Table 1). In the first factor, hype, parents say it is not true (0) to very true (2) that their child is now, or has been in the past 6 months, boastful, a show-off, verbose, and using obscene language. Racial inopportunity indicates teens strongly disagree (1) to strongly agree (4) that because of race, “there is little you can do to avoid discrimination at school, and at the job you will have in the future,” and that “you will have to work harder than others to prove yourself.” Last, masculine identity reflects adolescents’ opinion that they feel and look masculine ranging from not at all (1) to very (7), and feel it is not at all (1) to very (7) important to appear masculine and avoid appearing feminine. Other expressive cool variables include acting Black (“friends would think it was very uncool/cool if I acted Black”), best dressed (“friends would think it was very uncool/cool if I was voted best dressed”), both scaled 1 to 5, and being good looking (“how important is being good looking” scaled 1 for much less to 7 for much more important to me than it is for other kids). Finally, although adolescents were not asked whether they code-switch, they were asked to strongly agree (1) to strongly disagree (5) with the statement, “I can change my personality as I choose,” which reflects an ability to code-switch. I refer to this variable as personality change.

Of the social background characteristics, my interest in differences among Black and White adolescents required a dichotomous race variable (0 = White, 1 = Black). Single parenting (0 = no, 1 = yes) and dads’ interest indicating “my father almost never (1) to almost always (5) takes an interest in my activities” is considered because an engaged father in the home may reinforce adolescents’ traditional masculinity (Mandara, Murray, & Joyner, 2005) or lessen the impression other males in the neighborhood may have on them. I include two additional family measures; parent’s education (primary care giver’s highest grade completed) to account for parental influences on adolescents’ educational aspirations, and total family income, which I segment into equal quintiles in case its effects are non-linear. As peers may confound relationships between neighborhoods or expressive orientations and schooling outcomes, I also include friends’ AIR beliefs (“How many friends that you spend most of your time with think you won’t get a good job even if you do well in school”). This variable ranges from none of them (1) to all of them (5).

Table 1 includes neighborhood characteristics of two types; the first are more objective measures of neighborhood composition taken from the census. I include the percentage of college degree holders because the neighborhood presence of those who have completed degrees may communicate the importance of education to area youths and indicate a greater availability of effective role models. I also include the percentage of female-headed households to explore variation across neighborhoods in non-resident fathers. Finally, I use youths’ distance from DC because Pattillo-McCoy (1999) posits that males in nearby cities offer suburban teens alternative examples of masculine expressive behavior. I dichotomize this variable to indicate adolescents’ inner suburban ring residency (0 = no, 1 = yes). The final two measures reflect parents’ perceptions about their neighborhoods. Neighborhood joblessness has been dichotomized to indicate parents’ view that unemployment in the area is a big problem (1) or not a big problem (0). Collective socialization is a factor consisting of parent reports that they strongly disagree (1) to strongly agree (5) that “there are lots of adults in the area our kids can look up to,” “I can count on neighbors to tell me about opportunities for children,” and “neighbors share similar views about raising kids.”

Hierarchical Linear Modeling

As the adolescents in this study are nested within neighborhoods, I use hierarchical linear models (Raudenbush & Bryk, 2002) to model between adolescent-measures of AIR according to their social background and expressive cool qualities at Level 1, and ecological measures at Level 2. In the unconditional model specification below, youths’ AIR Y_cn is viewed as a function of an intercept for adolescent c in neighborhood n yielding the Level 1 equation:

Y_{c n} = β_{0 n} + e_{c n},

where Y_cn is the outcome of adolescent c within neighborhood n, β_0n is the mean outcome level in neighborhood n, and e_cn is an error term assumed normally distributed.

Level 1 of the conditional model includes the social background and expressive masculinity variables. As this research investigates adolescents’ expressive cool net of other social background characteristics, there are two Level 1 conditional specifications for each education outcome. In the first Level 1 specification, AIR, Y_cn is a function of the quintiles of family socioeconomic status (SES; with the middle quintile excluded); the variables of parents’ educational attainment, dads’ interest in activities, marriage status, friends’ AIR, and age. The second specification adds the expressive cool measures: hype, masculine identity, racial inopportunity, acting Black, best dressed, personality change, and being good looking. The full Level 1 equation is

\begin{array}{l} Y_{c n} = β_{0 n} + β_{1 n} {(Age)}_{c n} + β_{2 n} {(Race)}_{c n} + β_{3 n} {(Single Parenting)}_{c n} + β_{4 n} {(Parents’ Education)}_{c n} + \\ β_{5 n} {(Dad’s Interest)}_{c n} + β_{6 - 9 n} {(SES Quintiles)}_{c n} + β_{10 n} {(Friends’ AIR)}_{c n} + β_{11 n} {(Masculine Identity)}_{c n} + \\ β_{12 n} {(Hype)}_{c n} + β_{13 n} {(Racial Inopportunity)}_{c n} + β_{14 n} {(Good Looking)}_{c n} + β_{15 n} {(Acting Black)}_{c n} + β_{16 n} (Best \\ Dressed)_{c n} + β_{17 n} {(Personality Change)}_{c n} + e_{c n} . \end{array}

The Level 2 equation models neighborhood-to-neighborhood variation for all of the neighborhood variables. Hence, AIR, β_0n is a function of the percentage of female-headed households, degree holders, parents’ perception of high unemployment, collective socialization, and adolescents’ residency in the inner suburban ring surrounding DC. I express the Level 2 equation as follows:

\begin{array}{l} β_{0 n} = γ_{00} + γ_{01 n} {(% Degreed)}_{n} + γ_{02 n} {(% Female Heads)}_{n} + γ_{03 n} {(Joblessness)}_{n} + γ_{04 n} (Collective \\ Socialization)_{n} + γ_{05 n} {(Inner Suburban Ring)}_{n} + r_{0 n} . \end{array}

The neighborhood variables indicate the estimated deviation from the mean AIR γ_00n associated with a unit increase among those factors.

Analysis

Descriptive Analysis

The descriptive analysis reveals that this sample has low levels of AIR and high GPAs, with modest variation in the latter as evident from its small standard deviation. The racial differences in mean outcomes are also minor, but more pronounced among the social background and neighborhood characteristics. Although Prince George’s consistently holds the designation as the nation’s wealthiest majority Black county, Table 1 shows that the average total annual income of White families is higher, falling at the upper end of the US$65,000 to US$69,999 range, whereas for Black families, it is in the mid-US$55,000 to US$59,999 range. The SES quintile means provide a glimpse at how these incomes are distributed according to race. Only 28% of Whites are in the two lowest income quintiles compared with 48% of Black youth. Another notable racial difference appears within the family context, where the single parenting average for African Americans is double that of Whites. Despite this disparity, Black adolescents report their dads having greater interest in their activities than do White teens.

Among the expressive cool constructs, Black and White adolescents differed most in regards to their masculine identity, with the Black average being substantially higher (3.59 vs. 2.08). Perceptions of racial inopportunity and the social acceptance of acting Black were also much higher for Black boys. For the former measure however, the Black average was still on the low end of the scale indicating more optimism than pessimism about discrimination. In the neighborhood context, White youth live in neighborhoods that have a higher proportion of college graduates and a lower percentage of female-headed families than do African Americans. A smaller percentage of White youth reside in the inner suburban ring that borders Washington, DC, than do their Black peers (15.2 vs. 27.9%). Minor racial differences were evident among the remaining indicators.

Full Sample Analysis

Table 2 shows the fixed and random effects of the full sample analysis. Model 1 reports the AIR intercept and social background estimates; Models 2 and 3 add the expressive cool and neighborhood dimensions, respectively. Considering teens’ AIR first, Model 1 shows African Americans have a significantly lower level of AIR ( $\overset{⌢}{β}$ ₀ = −.24) than their White counterparts of nearly a quarter standard deviation. A smaller but significantly higher likelihood to reject dominant achievement ideologies appears for adolescents who have friends who also reject achievement ideologies ( $\overset{⌢}{β}$ ₁₀ = .21) and dads who show interest in their activities ( $\overset{⌢}{β}$ ₀₅ = .10). Once expressive cool dimensions are added in Model 2, those significant relationships in Model 1 weaken and the racial difference in AIR loses significance. Racial inopportunity ( $\overset{⌢}{β}$ ₁₃ = .13) appears to heighten AIR, whereas adolescents’ masculine perception ( $\overset{⌢}{β}$ ₁₁ = −.13) and being considered well-dressed ( $\overset{⌢}{β}$ ₁₆ = −.25) lowers AIR, the latter effect amounting to a quarter standard deviation unit reduction. My introduction of the neighborhood variables in Model 3 appears to have little influence on the magnitude of expressive cool estimates, and none of them appear significantly related to teens’ AIR.

Table 2.

AIR and GPA, Full Sample (N = 406).

	AIR						GPA
	Model 1		Model 2		Model 3		Model 4		Model 5		Model 6
Fixed effects	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE
Intercept	1.763***	0.057	1.763***	0.054	1.753***	0.055	3.583***	0.027	3.583***	0.026	3.577***	0.022
Age in years	0.117	0.112	0.014	0.107	0.022	0.106	−0.028	0.029	−0.016	0.028	−0.010	0.026
Race	−0.228^†	0.122	0.029	0.149	0.134	0.157	−0.106**	0.039	−0.145**	0.047	−0.007	0.036
Single parent	0.049	0.115	0.010	0.112	0.004	0.111	−0.003	0.023	−0.001	0.022	−0.007	0.024
Parent education	−0.059*	0.024	−0.057*	0.023	−0.064**	0.023	0.013*	0.005	0.012*	0.005	0.004	0.004
Dad’s interest	0.101*	0.039	0.084*	0.038	0.084*	0.039	−0.001	0.010	−0.003	0.010	−0.004	0.009
Low SES	0.061	0.135	0.072	0.128	0.095	0.127	−0.080*	0.037	−0.075^†	0.039	−0.059	0.039
Mid-low SES	0.214	0.137	0.199	0.136	0.204	0.133	−0.060	0.039	−0.055	0.042	−0.050	0.039
Mid-high SES	0.160	0.165	0.159	0.160	0.147	0.165	−0.011	0.036	−0.015	0.037	−0.033	0.038
High SES	−0.101	0.185	−0.061	0.183	−0.105	0.183	0.002	0.046	0.006	0.047	−0.046	0.040
Friends’ AIR	0.232***	0.065	0.187**	0.060	0.193**	0.062	−0.003	0.014	−0.002	0.015	0.013	0.012
Masculine appearance			−0.144*	0.060	−0.153*	0.060			0.025	0.015	0.016	0.013
Hype			0.067	0.053	0.065	0.050			−0.008	0.014	−0.008	0.014
Racial inopportunity			0.121*	0.090	0.120*	0.058			0.014	0.013	0.015	0.012
Being good looking			0.000	0.034	0.006	0.034			−0.013^†	0.008	−0.008	0.008
Cool if acted Black			−0.020	0.052	−0.023	0.048			−0.003	0.013	−0.002	0.012
Cool if best dressed			−0.249***	0.070	−0.145***	0.070			−0.003	0.011	0.000	0.010
Personality change			0.125	0.076	0.117	0.077			0.014	0.013	0.002	0.012
% Degreed					0.007	0.005					0.012***	0.003
% Female heads					−0.004	0.005					−0.005^†	0.003
Joblessness					0.002	0.071					−0.059^†	0.030
Inner suburban ring					0.029	0.131					0.098	0.072
Collective socialization					−0.020	0.049					0.015	0.020
Levels 1 and 2 variance/SD	0.548***	0.740	0.431***	0.657	0.425***	0.652	0.00002	0.004	0.00002	0.004	0.00002	0.004
Level 3 variance/SD	0.052^†	0.228	0.047*	0.216	0.044*	0.210	0.0431***	0.208	0.040***	0.200	0.0162***	0.127

Note. AIR = achievement ideology rejection; GPA = grade point average; SES = socioeconomic status.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Columns 4 through 6 of Table 2 specify the same models in relation to GPA. Unlike the AIR analysis, Models 4 and 5 show that social background and expressive cool characteristics explain little, if any, variation in GPA. The significant estimates for the neighborhood percentage of degree holders ( $\hat{γ}$ ₀₁ = .01) and the tenuous negative effects of neighborhood joblessness ( $\hat{γ}$ ₀₃ = −.06) shown in Model 6 imply that neighborhood features are better able than individual characteristics to account for variation in adolescents’ GPA. The amount of between-neighborhood variation in teens’ GPA explained from Model 5 where no neighborhood dimensions are specified, to Model 6 where they are 37.3% is calculated by differencing the variance component, τ, of two models and dividing by the value of the former model variance component (τ₁ − τ₂/τ₁).

In sum, and in relation to the research questions, Tables 2 shows (a) evidence of an attitude–achievement paradox where African Americans have attitudes that are more supportive of achievement ideologies, but not higher GPAs than their White counterparts; (b) that expressive cool dimensions seem to have a stronger connection to AIR, and very little if any, to adolescents’ GPA; and (c) that neighborhood qualities appear unrelated to AIR but do account for some variation in teens’ GPA.

Disaggregated Sample Analysis

The full sample analysis provided important insights about racial differences in educational attitudes and the mediating role of expressive cool. The fact that racial differences in AIR became insignificant once expressive cool dimensions were added intimates that racial differences in expressive cool might be revealed once these data are disaggregated according to race. To explore this possibility, Tables 3 and 4 summarize the analysis disaggregated according to race. The model specifications are identical to those presented in Table 2 except that the race variable is withheld.

Table 3.

AIR Disaggregated According to Race.

	Black sample (n = 261)						White sample (n = 145)
	Model 1		Model 2		Model 3		Model 4		Model 5		Model 6
Fixed effects	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE
Intercept	1.756***	0.059	1.758***	0.053	1.747***	0.054	1.873***	0.100	1.882***	0.090	1.898***	0.085
Age in years	0.067	0.115	−0.019	0.123	−0.003	0.123	0.199	0.212	0.103	0.194	0.136	0.185
Single parent	−0.033	0.166	−0.045	0.146	−0.042	0.148	0.208	0.194	0.148	0.154	0.157	0.164
Parent education	−0.106***	0.027	−0.091***	0.025	−0.096***	0.026	−0.001	0.038	0.011	0.038	0.014	0.036
Dad’s interest	0.089^†	0.049	0.078^†	0.042	0.079^†	0.042	0.133	0.094	0.056	0.082	0.047	0.082
Low SES	0.141	0.204	0.124	0.170	0.139	0.173	−0.145	0.290	−0.017	0.272	0.057	0.252
Mid-low SES	0.375*	0.165	0.317*	0.154	0.318*	0.152	−0.176	0.209	−0.035	0.220	−0.107	0.224
Mid-high SES	0.177	0.199	0.189	0.196	0.174	0.211	0.021	0.237	−0.109	0.258	−0.156	0.265
High SES	0.196	0.218	0.137	0.213	0.108	0.228	−0.500*	0.216	−0.309^†	0.205	−0.297	0.199
Friends’ AIR	0.201**	0.073	0.165**	0.065	0.163**	0.067	0.254*	0.123	0.181	0.105	0.135	0.101
Masculine appearance			−0.122*	0.057	−0.125*	0.055			−0.076	0.097	−0.083	0.096
Hype			0.069	0.061	0.064	0.060			0.070	0.083	0.051	0.082
Racial inopportunity			0.035	0.062	0.031	0.057			0.301**	0.093	0.307***	0.087
Being good looking			−0.003	0.038	0.002	0.037			0.018	0.070	0.025	0.067
Cool if acted Black			−0.004	0.063	−0.007	0.063			−0.080	0.079	0.093	0.080
Cool if best dressed			−0.243***	0.063	−0.241***	0.064			−0.306***	0.180	−0.286***	0.079
Personality change			0.209**	0.079	0.204*	0.080			−0.030	0.099	−0.065	0.095
% Degreed					0.006	0.006					0.008	0.010
% Female heads					−0.001	0.005					−0.008	0.015
Joblessness					0.000	0.078					0.293*	0.123
Inner suburban ring					−0.064	0.127					0.174	0.218
Collective socialization					−0.026	0.061					−0.110	0.092
Variance/SD	0.022	0.148	0.006	0.080	0.011	0.107	0.213**	0.461	0.155*	0.394	0.122^†	0.349
Level 1, r/SD	0.806	0.898	0.702	0.838	0.705	0.840	0.757	0.870	0.675	0.821	0.671	0.819

Note. AIR = achievement ideology rejection; SES = socioeconomic status.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Table 4.

Grade Point Average Disaggregated According to Race.

	Black sample (n = 261)						White sample (n = 145)
	Model 1		Model 2		Model 3		Model 4		Model 5		Model 6
Fixed effects	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE	Estimate	SE
Intercept	3.522***	0.036	3.523***	0.036	3.523***	0.030	3.721***	0.041	3.654***	0.043	3.643***	0.033
Age in years	−0.009	0.021	−0.004	0.025	−0.003	0.024	−0.012	0.046	0.003	0.051	0.014	0.053
Single parent	−0.006	0.018	−0.007	0.017	−0.009	0.018	0.011	0.050	0.003	0.046	0.001	0.045
Parent education	0.003	0.007	0.002	0.007	0.005	0.006	0.010	0.011	0.009	0.012	0.001	0.011
Dad’s interest	−0.001	0.009	−0.000	0.009	0.001	0.007	−0.010	0.021	−0.010	0.023	−0.022	0.017
Low SES	−0.011	0.023	−0.010	0.023	−0.006	0.021	−0.053	0.116	−0.012	0.080	0.040	0.085
Mid-low SES	−0.018	0.026	−0.017	0.027	−0.021	0.028	−0.042	0.080	−0.038	0.089	−0.028	0.088
Mid-high SES	0.001	0.041	0.002	0.040	−0.001	0.040	−0.035	0.099	−0.054	0.058	−0.032	0.080
High SES	0.003	0.051	0.006	0.048	−0.006	0.050	−0.009	0.063	0.081	0.081	−0.015	0.061
Friends’ AIR	0.001	0.017	0.002	0.018	0.004	0.018	0.018	0.036	0.021	0.039	0.026	0.037
Masculine appearance			0.006	0.012	0.001	0.009			0.021	0.042	0.012	0.036
Hype			−0.005	0.009	−0.008	0.008			−0.027	0.029	−0.009	0.023
Racial inopportunity			0.007	0.012	0.007	0.012			0.001	0.017	0.004	0.016
Being good looking			−0.004	0.012	−0.004	0.011			−0.014	0.023	−0.017	0.022
Cool if acted Black			−0.006	0.011	−0.005	0.011			0.003	0.036	0.002	0.031
Cool if best dressed			−0.000	0.009	0.001	0.009			0.012	0.019	0.010	0.017
Personality change			0.000	0.012	−0.002	0.011			0.019	0.038	0.020	0.031
% Degreed					0.010**	0.003					0.016***	0.004
% Female heads					−0.005^†	0.003					−0.005	0.006
Joblessness					−0.113**	0.037					0.002	0.049
Inner suburban ring					0.112	0.069					0.063	0.112
Collective socialization					0.004	0.030					0.023	0.039
Variance/SD	0.085***	0.292	0.084***	0.290	0.053***	0.230	0.074***	0.271	0.075***	0.273	0.038***	0.194
Level 1, r/SD	0.013	0.115	0.014	0.116	0.013	0.116	0.034	0.185	0.033	0.182	0.034	0.185

Note. SES = socioeconomic status; AIR = achievement ideology rejection.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Regarding teens’ AIR, Table 3 shows remarkable racial differences among social background, expressive cool, and neighborhood dimensions. Considering social background first, the significantly higher levels of AIR related to dads’ interest in their son’s activities that emerges in the full analysis seems, in Table 3, most likely to apply to African American teens. Being within a mid-lower SES household ( $\overset{⌢}{β}$ ₀₇ = .38) is related to significantly higher AIR levels for Black adolescents while being in a high income family is related to a half standard deviation reduction for Whites ( $\overset{⌢}{β}$ ₀₉ = −.50). Higher AIR levels remain for low SES African Americans with the consideration of expressive cool and neighborhood features, whereas lowered AIR levels disappear for high SES Whites. This is also the case for the association between adolescents’ AIR and that of their friends’; AIR levels remain elevated for Blacks in all models as the AIR of their friends are considered ( $\overset{⌢}{β}$ ₁₀ = .19), while the larger AIR effect for Whites ( $\overset{⌢}{β}$ ₁₀ = .25) loses significance as their expressive cool and neighborhood qualities are considered.

Considering expressive cool next, the two racial groups are similar in that being regarded as best dressed is associated with less AIR, especially for White teens ( $\overset{⌢}{β}$ ₁₆ = −.30), indicating the high SES connection to lowered AIR in Model 4 was accounted for by their dress. However, there were interesting differences too. First, African American adolescents’ masculinity ( $\overset{⌢}{β}$ ₁₁ = −.12) is related to less skepticism about achievement ideologies in the full model, but not for White males. Second, African Americans’ inability to change their personality ( $\overset{⌢}{β}$ ₁₇ = .21) was associated with higher AIR levels, suggesting the absence of this social skill accompanies more critical assessments of dominant achievement ideologies. Third, White AIR levels were more sensitive to beliefs in racial inopportunity ( $\overset{⌢}{β}$ ₁₃ = .30). This curious finding prompts the question of whether this association is due to their own underrepresented status in this context or if it reflects their worldview about race relations.

Last, Models 3 and 6 imply that neighborhood features, while unrelated to AIR in the full analysis and for Black adolescents, are associated with the AIR of White teens. Parent concerns about joblessness are related to nearly a third of a standard deviation increase ( $\hat{γ}$ ₀₃ = .29) in White’s AIR. In addition, the random components, τ, show that there is significant neighborhood variation in White but not Black adolescents’ AIR levels. This variation becomes more likely due to chance as expressive cool and neighborhood dimensions are added to the models.

Table 4 presents the analysis of adolescents’ GPA according to race and affirms that the findings of the full analysis apply to both racial groups with just a few exceptions. First, none of the social background or expressive cool characteristics are related to GPA. Second, Models 3 and 6 show that it is at the neighborhood level where we find significant associations with GPA of greatest magnitude. Among the associations found for African Americans is a modest positive effect for the neighborhood presence of degree holders ( $\hat{γ}$ ₀₁ = .01) and a larger association for joblessness ( $\hat{γ}$ ₀₃ = −.11). Only the neighborhood percentage of degree recipients ( $\hat{γ}$ ₀₁ = .02) is positively related to White adolescents’ GPA. Significant between-neighborhood variation in GPA is shown in the random components but dwindles considerably for both racial groups as neighborhood characteristics are included, but more among White teens (49%) than Black teens (37%).

We can conclude from the disaggregated analysis that (a) expressive cool has varied associations to both racial groups’ AIR but are largest for Whites; (b) expressive cool is insignificantly related to GPA; and (c) neighborhood linkages to GPAs are stronger than they are to AIR.

Discussion

The scarceness of quantitative research on adolescent expression and style stands in stark contrast to their relevance in today’s public concern about Black males. In suburban areas too, Black males are often assumed by their observers to “think and behave as they look.” It is an assumption that has left apparel, such as the “hoodie” worn by the late Trayvon Martin, to symbolize the tragedy of stigma. Setting aside assumptions, this study sought to understand if Black and White adolescent males’ expressive cool relates to their attitudes and behaviors after accounting for neighborhoods’ collective socialization, and whether the pattern of those relationships constitutes an “attitude–achievement paradox.” Also motivating this study are the claims of research that racial differences in expressive gender performances contribute to educational inequality. Given that few studies address these issues within a multilevel analysis, this study’s findings are especially timely and offer implications for research and how society regards Black males.

The first implication extends from the fact that while African Americans’ expressive measures were slightly higher than their White peers (Table 1), these performances (i.e., masculinity, code-switching, and dress) were related to stronger achievement ideology acceptance, not lower. For White males, not only were expressive cool associations with AIR among the largest in this analysis, being best dressed in particular mediated and rendered insignificant this study’s largest association between White adolescents’ high income status and lowered AIR. This result emphasizes the extent to which social class is expressed through White teens’ dress, and to a greater degree than it is for high income African Americans. Although Majors and Billson’s (1993) and Pattillo-McCoy’s (1999) claim that middle-class Blacks take their style more seriously than Whites may be supported in the descriptive analysis, the inferential analysis shows the educational consequences of style’s importance are greatest for White teens.

Some may reason that this finding among White teens is evidence of a cultural spillover effect due to their location in a majority–minority context. Although this is plausible, more research is needed to determine whether the salience of White adolescents’ expressivity is exogenous and would appear similar in contexts with a different demographic composition before we could conclude racial heterogeneity has led to this outcome. Rather, this finding might extend from between-school differences in how educational policy manages the expressive culture of racial groups. Although the district has a system-wide dress-code that does not mandate uniforms (Prince George’s County Student Conduct Handbook, 2014), the individual schools attended by African Americans often have a mandatory uniform policy. Uniforms in largely African American schools would limit the importance of dress while their absence in largely White schools might facilitate youth’s expression of social class. Although this study found only positive effects related to being considered well dressed, the relative benefit of style and uniformity related to adolescents’ schooling is a topic for future research.

Another implication of this research begs the question of whether past research on the attitude–achievement paradox would have produced similar results had expressive cool features been considered. Although I find greater achievement ideology beliefs among African Americans do not translate into relatively higher GPAs, this paradox disappears as the consideration of expressive cool factors accounts for the higher AIR levels of White teens. Nonetheless, that African Americans do not appear to convert their stronger achievement ideology beliefs into relatively higher GPAs reinforces the need for social programs and educators that can leverage these beliefs toward optimal academic performances.

But there are other notable “paradoxes” that this research presents for discussion. First, significant associations that I report between expressive cool and AIR do not exist in relation to GPA, for the full sample and both racial groups. Although this pattern might apply only to the expressive characteristics I examine, it nonetheless implies that the dispositions adolescents hold toward achievement do not necessarily impinge on their subsequent behaviors. This outcome contrasts with the claims of Thernstrom and Thernstrom (2003) and Ronald Ferguson (2001) that expressive culture is an academic distraction, and linked to lower achievement for African Americans in particular. Assumptions that correspondences exist between ideological beliefs and behavior paint adolescents’ culture with broad strokes and overlook the apparent complexity of adolescents’ behavioral responses to their gendered achievement dispositions. These findings echo the qualitative work of Dance (2002) who laments that these assumptions are overly deterministic and disregard the “contradictions and hybridity” of Black males’ culture in ways that deny their agency (p. 17). Indeed, the best way for schools and society to support African American male achievement may be to discontinue the stereotypical association of their culture with underperformance.

Second, this work also reveals what might be called a “neighborhood paradox” in which Black teens’ achievement ideology seems robust to neighborhood socialization risks while their grades seem vulnerable. One interpretation of this phenomena is that achievement ideology beliefs are unable to shield African Americans from the neighborhood influences that undermine their grades to the extent they appear to do so for Whites. Put differently, believing in equal opportunity neither ensures equal educational outcomes nor retires the material reality Black adolescents must contend with to secure academic success. Yet another interpretation might view the non-existent linkages between neighborhoods and AIR as evidence that their function as sites of African American males’ cultural production is secondary to their existence as environments that African Americans must negotiate to achieve in school. Hence, neighborhood joblessness, for example, which many have argued undermines Black males’ opportunity beliefs, did not do so in this study, but was nonetheless linked to lowered academic performance. Such obstacles to excellence require structural intervention, not merely cultural intervention, on behalf of youth.

There are reasons to exercise caution when using research of this kind to make causal claims. One reason is that simultaneity effects are especially difficult to address in observational research. For example, although this research finds that dads’ interest in their sons’ activities is related to higher AIR levels in the full sample, it is unknown if dads’ interest causes higher AIR or if adolescents’ AIR causes greater interest from their dads. Additional research is needed to disentangle transactions of this sort before causal conclusions can be reached. A second reason is that adolescents’ school contexts are also important to consider, but whether they provide an independent effect or serve to convey neighborhood effects remains a question within research (O. Johnson, 2012a). School effects were not presented here because the models in which I included school gender and racial bias factors provided no significant or mediating effects while other school factors were too highly correlated with neighborhood dimensions for analysis. Future studies with larger samples might avoid some of these analytical challenges.

A third reason is that it is not clear whether children’s opposition to achievement leads to lowered achievement, or develops in response to lowered achievement. This is an important point that I believe existing studies on youth culture have woefully neglected. We have forgotten Liebow’s (1967) classic study, which observed that men hold mainstream values about work and family, but adjust downward their importance upon losing their jobs and partners. AIR may similarly be a “coping disposition” that emerges post hoc among youth whose mainstream achievement ideologies were not rewarded with good grades.

A final reason for cautious consideration of these study results is that neighborhood selection effects remain a concern in residential studies because many of the reasons that parents decide to live in certain neighborhoods are unknown and possibly related to the welfare of their children. This caution, and the others I mentioned, can be addressed with methodologies such as counterfactual modeling that are not suitable for samples of this study’s size. It is encouraging to note however that other examinations of simultaneity and selection effects related to education provide little evidence that methodological approaches similar to the ones taken in this study are prone to overstate neighborhood influences (Foster & McLanahan, 1996; Harding, 2003).

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biography

Odis Johnson Jr. PhD, is associate chair and director of Graduate Studies in the Department of Education and an associate professor in the Departments of Sociology and Education at Washington University in St. Louis. Prior to his appointments at Washington University, Dr. Johnson chaired the African American Studies Department at the University of Maryland. Dr. Johnson’s research and teaching explore how neighborhoods, schools and public policies relate to social inequality, youth development and the status of African American populations.

References

Abreu

J. M.

Goodyear

R. K.

Campos

Newcomb

M. D.

(2000). Ethnic belonging and traditional masculinity ideology among African Americans, European Americans, and Latinos. Psychology of Men & Masculinity, 1, 75-86. doi:10.1037//1524-9220.1.2.75

Ainsworth

J. W.

(2010). Does the race of neighborhood role models matter? Collective socialization effects on educational achievement. Urban Education, 45, 401-423. doi:10.1177/0042085910372346

Ainsworth-Darnell

J. W.

Downey

(1998). Assessing the oppositional culture explanation for racial/ethnic differences in school performance. American Sociological Review, 63, 536-553.

Akom

A. A.

(2003). Reexamining resistance as oppositional behavior: The Nation of Islam and the creation of a Black achievement ideology. Sociology of Education, 76, 305-325.

Anderson

(1999). Code of the street: Decency, violence, and the moral life of the inner city. New York, NY: W.W. Norton.

Carter

(2003). “Black” cultural capital, status positioning, and schooling conflicts for low-income African American youth. Social Problems, 50, 136-155.

Cassidy

E. F.

Stevenson

H. C.

(2005). They wear the mask: Hypervulnerability and hypermasculine aggression among African American males in an urban remedial disciplinary school. Journal of Aggression, Maltreatment & Trauma, 11, 53-74. doi:10.1300/J146v11n04_03

Caughy

Nettles

O’Campo

Lohrfink

(2006). Neighborhood matters: Racial socialization of African American children. Child Development, 77, 1220-1236. doi:10.1111/j.1467-8624.2006.00930.x

Cook

P. J.

Ludwig

(1997). Weighing the “burden of ‘acting White’”: Are there race differences in attitudes toward education? Journal of Policy Analysis and Management, 16, 256-278.

10.

Cook

Herman

Phillips

Settersten

(2002). Some ways in which neighborhoods, nuclear families, friendship groups and schools jointly effect changes in early adolescent development. Child Development, 73, 1283-1309. doi:10.1111/1467-8624.00472

11.

Crane

(1991). The epidemic theory of ghettos and neighborhood effects on dropping out and teenage childbearing. American Journal of Sociology, 96, 1226-1259.

12.

Dance

L. J.

(2002). Tough fronts. New York, NY: RoutledgeFalmer.

13.

DuBois

W. E. B.

(1903). The souls of Black folk. New York, NY: Random House.

14.

Duncan

G. J.

(1994). Families and neighbors as sources of disadvantage in the schooling decisions of black and white adolescents. American Journal of Education, 103, 20-53.

15.

Duneire

(1994). Slim’s Table: Race, Respectability, and Masculinity. Chicago: University of Chicago Press.

16.

Eisler

R. M.

Skidmore

J. R.

(1987). Masculine gender role stress. Scale development and component factors in the appraisal of stressful situations. Behavior Modification, 11, 123-136. doi:10.1177/01454455870112001

17.

Ferguson

A. A.

(2001). Bad boys: Public schools in the making of black masculinity. Ann Arbor: University of Michigan Press.

18.

Ferguson

R. F.

(2001). Test-score trends along racial lines, 1971 to 1996: Popular culture and community academic standards. In Smelser

Wilson

Mitchell

(Eds.), America becoming: Racial trends and their Consequences (pp. 348-390). Washington, DC: National Academy Press.

19.

Foster

McLanahan

(1996). An illustration of the use of instrumental variables: Do neighborhood conditions affect a young person’s chance of finishing high school? Psychological Methods, 1, 249-260. doi:10.1037/1082-989X.1.3.249

20.

Fryer

Torelli

(2010). An empirical analysis of “acting white.” Journal of Public Economics, 94, 380-396. doi:10.1016/j.jpubeco.2009.10.011

21.

Ginther

Haveman

Wolfe

(2000). Neighborhood attributes as determinants of children’s outcomes. Journal of Human Resources, 35, 603-642.

22.

Gurin

Epps

E. G.

(1975). Black consciousness, identity and achievement. New York, NY: Wiley.

23.

Harding

(2003). Counterfactual models of neighborhood effects: The effect of neighborhood poverty on dropping out and teenage pregnancy. American Journal of Sociology, 109, 676-719. doi:0002-9602/2003/10903-0004

24.

Harding

(2010). Living the drama: Community, conflict, and culture among inner-city boys. Chicago, IL: The University of Chicago Press.

25.

Harris

(2006). I (don’t) hate school: Revisiting oppositional culture theory of Blacks’ resistance to schooling. Social Forces, 85, 797-832.

26.

Harris

Marsh

(2010). Is a raceless identity an effective strategy for academic success among Blacks? Social Science Quarterly, 91, 1242-1263. doi:10.1111/j.1540-6237.2010.00730.x

27.

Jencks

Mayer

(1990). The social consequences of growing up in a poor neighborhood. In Lynn

L. E.

Jr. McGeary

M. G. H.

(Eds.), Inner-city poverty in the United States (pp. 111-186). Washington, DC: National Academy Press.

28.

Johnson

Jr. (2008). Who benefits from concentrated affluence? A synthesis of neighborhood effects considering race, gender and education outcomes. Journal of Public Management and Social Policy, 14, 85-112.

29.

Johnson

Jr. (2012a). A systematic review of neighborhood and institutional relationships related to education. Education and Urban Society, 44, 477-511. doi:10.1177/0013124510392779

30.

Johnson

Jr. (2012b). Relocation programs, opportunities to learn and the complications of conversion. Review of Educational Research, 82, 131-178. doi:10.3102/0034654312449489

31.

Johnson

W. E.

(2001). Paternal involvement among unwed fathers. Children and Youth Services Review, 23, 513-536.

32.

Kimmel

M. S.

(1986). Toward men’s studies. American Behavioral Scientist, 29, 517-529. doi:10.1177/000276486029005002

33.

Kimmel

M. S.

Mahler

(2003). Adolescent masculinity, homophobia, and violence: Random school shootings, 1982-2001. American Behavioral Scientist, 46, 1439-1458. doi:10.1177/0002764203046010010

34.

Kirkland

Jackson

(2009). “We real cool”: Toward a theory of Black masculine literacies. Reading Research Quarterly, 44, 278-297. doi:10.15.98/RRQ.43.3.3

35.

Klein

, (2006). Sexuality and school shootings: What role does teasing play in school massacres? Journal of Homosexuality, 51(4), 39-62. doi:10.1300/J082v51n04_03

36.

Lacy

(2007). Blue-chip black: Race, class, and status in the new Black middle class. Oakland: University of California Press.

37.

Lazur

R. F.

Majors

(1995). Men of color: Ethnocultural variations of male gender role strain. In Levant

R. F.

Pollack

W. S.

(Eds.), A new psychology of men (pp. 337-358). New York, NY: Basic Books.

38.

Liebow

(1967). Tally’s Corner: A study of Negro streetcorner men. New York, NY: Little, Brown.

39.

MacLeod

(1987). Ain’t no makin’ it: Aspirations and attainment in a low-income neighborhood. Boulder, CO: Westview Press.

40.

Madyun

Lee

M. S.

(2010). The influence of female headed households on Black achievement. Urban Education, 45, 424-447. doi:10.1177/0042085910372348

41.

Majors

Billson

(1993). Cool pose: The dilemmas of Black manhood in America. Clearwater, FL: Touchstone Books.

42.

Mandara

Murray

Joyner

(2005). The impact of fathers’ absence on African American adolescents’ gender role development. Sex Roles, 53, 207-220. doi:10.1007/s11199-005-5679-1

43.

Mickelson

(1990). The attitude-achievement paradox among black adolescents. Sociology of Education, 63, 44-61.

44.

Newman

(1999). No shame in my game. New York, NY: Russell Sage.

45.

O’Connor

(1997). Dispositions toward (Collective) struggle and educational resilience in the inner city: A case analysis of six African-American high school students. American Educational Research Journal, 34, 593-629. doi:10.3102/00028312034004593

46.

O’Connor

(2000). Dreamkeeping in the inner-city: Diminishing the divide between aspirations and expectations. In Danziger

Lin

, (Eds.), Copping with poverty (pp. 105-140). Ann Arbor: University of Michigan Press.

47.

Ogbu

J. U.

(1987). Variability in minority school performance: A problem in search of an explanation. Anthropology of Education Quarterly, 18, 312-336. doi:10.1525/aeq.1987.18.4.04x0022v

48.

Ogbu

J. U.

(2003). Black American students in an affluent suburb: A study of academic disengagement. NJ: Lawrence Erlbaum Associates Press.

49.

Oliver

(2006). “The streets”: An alternative black male socialization institution. Journal of Black Studies, 36, 918-937. doi:10.1177/0021934704276445

50.

O’Neil

J. M.

Helms

B. J.

Gable

R. K.

Davis

Wrightsman

(1986). Gender role conflict scale: College men’s fear of femininity. Sex Roles, 14, 335-350.

51.

Pattillo-McCoy

(1999). Black picket fences: Privilege and peril among the black middle class. Chicago, IL: The University of Chicago Press.

52.

Prince George’s County Student Conduct Handbook. (2014). Encouraging positive student conduct and safety handbook. Prince George’s County, MD: Prince George’s County Public Schools. Retrieved from pgs.k12.va.us/dmdocuments/StudentConductHndbk.pdf

53.

Raudenbush

Bryk

(2002). Hierarchical linear models. Newbury Park, CA: Sage.

54.

Reidy

D. E.

Shirk

S. D.

Sloan

C. A.

Zeichner

(2009). Men who aggress against women: Effects of feminine gender role violation on physical aggression in hypermasculine men. Psychology of Men & Masculinity, 10, 1-12. doi:10.1037/a0014794

55.

Rosenbaum

J. E.

(1995). Changing the geography of opportunity by expanding residential choice: Lessons from the Gautreaux program. Housing Policy Debate, 6, 231-269. doi:10.1080/10511482.1995.9521186

56.

Sampson

R. J.

(1997). Collective regulation of adolescent misbehavior: Validation results from Chicago neighborhoods. Journal of Adolescent Research, 12, 227-244. doi:10.1177/0743554897122005

57.

Sampson

Morenoff

Earls

(1999). “Beyond Social Capital: Spatial Dynamics of Collective Efficacy for Children.” American Sociological Review, 64, 633-660.

58.

Seaton

(2007). Toward a theoretical understanding of hypermasculine coping among urban black adolescent males. Journal of Human Behavior in the Social Environment, 15, 367-390. doi:10.1300/J137v15n02_21

59.

Smalls

White

Chavous

Sellers

(2007). Racial ideological beliefs and racial discrimination experiences as predictors of academic engagement among African American adolescents. Journal of Black Psychology, 33, 299-330. doi:10.1177/0095798407302541

60.

Spencer

M. B.

(2001). Resiliency and fragility factors associated with the contextual experiences of low-resource urban African American male youth and families. In Booth

Crouter

(Eds.), Does it take a village? Community effects on children, adolescents, and families (pp. 51-77). Mahwah, NJ: Lawrence Erlbaum.

61.

Thernstrom

(2003). No excuses: Closing the racial gap in learning. New York, NY: Simon & Schuster.

62.

Tyson

Darity

W. A.

Jr. Castellino

D. R.

(2005). It’s not “a Black thing”: Understanding the burden of acting white and other dilemmas of high achievement. American Sociological Review, 70, 582-605.

63.

Willis

(1977). Learning to labour: How working class kids get working class jobs. New York, NY: Columbia University.

64.

Wilson

W. J.

(1996). When work disappears: The world of the new urban poor. New York, NY: Vintage.