The College Preparatory Pipeline: Disparate Stages in Academic Opportunities

Abstract

The rise in college preparatory coursework across American high schools appears not to affect college enrollment and graduation rates. This study uses the Civil Rights Data Collection to evaluate three stages along the college preparatory pipeline: access to, enrollment in, and mastery of Advanced Placement® and International Baccalaureate® coursework to understand the cumulative academic opportunities shaping students’ college readiness. Leaks in the pipeline divert out historically marginalized students. An adaptation of the Herfindahl-Hirschman Index captures the magnitude of these racial and ethnic disparities. Social context explains where school and district resources alleviate disparities to provide more equitable (i.e., proportionally representative) academic opportunities. These findings offer substantive direction to improve equality in students’ college readiness opportunities.

Keywords

opportunities to learn college readiness tracking segregation Advanced Placement®International Baccalaureate®

Since the late 1980s, high school students have gained more and more access to Advanced Placement® (AP; Geiser & Santelices, 2004; Kolluri, 2018) and, more recently, International Baccalaureate® (IB) course offerings (Donaldson, 2017). Access has widened in the sheer number of schools offering these college preparatory (prep) courses, and the course choices offered are increasingly varied (Geiser & Santelices, 2004; Judson & Hobson, 2015; Kolluri, 2018). The increased access and normalization of college prep offerings has “increased the average” of high school course-taking expectations (Schneider & Stevenson, 1999).

The rise in access to prep coursework has, unexpectedly perhaps, not translated to an increased number of college admissions or baccalaureate degree completions (Snyder et al., 2016).¹ Considering the rhetoric surrounding AP and IB courses, one would expect the increased access to yield stronger results (College Board, 2018; Evans, 2019).

Since 2000, undergraduate college enrollments rose 20% overall (National Center for Education Statistics, 2018), but this enrollment gain was not evenly distributed. The rapid rise of students from Latinx backgrounds—who rose from the lowest enrollment rate to nearly the rate of White students—drove the overall increase in college enrollment. Otherwise, when we look deeper into college enrollment trends, prominent gaps remain, as shown in Figure 1. Persistently, Asian students enroll at rates 10 to 15 percentage points higher than White students; Black students enroll at a rate approximately 8 percentage points lower.²

Figure 1.

College enrollment after high school graduation by student race or ethnicity, 2- and 4-year college combined.

Despite the increase in college enrollments, graduation trends remain fairly flat by comparison. College graduation rates among Asian and White student subgroups persistently remain twice those of Black students and 1.5 times the rate of Latinx students (Snyder et al., 2016). College graduation is inching upward at 1 to 3 percentage points across most enrolled students, although Black and multiracial college students are experiencing a drop in graduation (Snyder et al., 2016).

The differences in these college enrollment and completion trends may be explained in part by the wide variation in high schools preparing students for college. To assess this explanation, this study captures the level of students’ exposure to the college prep coursework pipeline and the racial/ethnic disparities in exposure using a version of the Herfindahl-Hirschman Index (HHI). A framework of organizational resource allocations associated with social context then explains where schools and districts alleviate or exacerbate these disparities. This framework, the prior literature, and the conceptualization of these disparities are explained next.

Organizational Resources Framework

In the United States, academic opportunities are not supposed to be predetermined by the education level or wealth status of students’ parents. Rather, the ideal of equal opportunity shapes the delivery of U.S. K–12 education. Although this ideal is far from the reality, schools and districts use their available resources to expose students to an array of academic opportunities. With that said, the amount of resources varies vastly from district to district and even between schools within districts.

Coleman (1990) referred to the diseconomy of resources when describing the resource constraints that school administrators confront when allocating resources to schools in a district, which involves (1) the quantity of resources (how much and how many are available to allocate); and (2) the quality of resources (new or tattered buildings or textbooks, accredited or not, etc.). The diseconomy involves the allocation of the quantity and quality of resources to different schools and the experiences of the students receiving those resources. When both the allocation and the experience tilt in the direction of “loss of input,” there is a diseconomy of resources. For example, if teacher salaries within an urban district are equal between schools and some of the schools serve lower income neighborhoods where many students experience poverty (Title I schools) and some are located in more affluent neighborhoods in the city (not Title I schools), then the district is not competitive in salary and will lose teachers to the more affluent suburbs (Coleman, 1990).³ When thinking of academic opportunities, this diseconomy of resources offers a perspective to understand where and why leakage occurs along the college prep pipeline to ready students for college.

Drawing on Coleman’s diseconomy of resources perspective to examine the allocation of resources by administrators and the experience of those resources by students in schools, this study assesses how schools and districts use their organizational resources to shape AP and IB curricular opportunities for students by decomposing the pipeline into three interlocked stages:

Access: Do students attend schools with the curriculum offerings?

Enrollment: Who in the school with the curriculum participates in these particular courses?

Mastery: Which students in the classrooms acquire the credential to demonstrate that they learned the expected content knowledge?

To examine the allocation of resources, this study focuses on “where” aspect of the diseconomy of resources by focusing on this social context of schools and districts. Social context is partitioned into three elements: curricular policy, spatial location, and peers.⁴

This study analyzes how organizational resources induce educational equality or inequality as experienced by students in the same educational setting. Unlike many studies of AP or IB, this study does not take an individual, student-level perspective to understand the differences between students’ success in their coursework. Instead, this study assesses the group-level perspective in relation to the organizational resources. The study thus gives direction to future policy discussions on the loss of input between disbursement and reception of academic opportunities that supports student preparation for college success.

Persistent Inequality in Academic Opportunity

Research on the three stages in the curricular pipeline explain the persistence of inequality via students’ differential exposures to academics. At the most basic foundation, equality involves the notion of the absence (compared with the presence) of disparities in schooling opportunities: Whether there exist measurable sociodemographic differences between students who have and those who do not have the ability to engage in academic opportunities. The following literature section reviews research on persistent academic opportunity inequality in American high schools.

Access and Enrollment

Both between-school and within-school tracking research identifies structural differences in access and enrollment. Differences in levels and tracks of courses between schools contributes to stratified access to rigorous courses, such as whether or not a high school offers calculus. In addition, differentiation in levels and tracks of courses within schools contributes to stratified enrollment in the courses, such as who and how many students enroll in the calculus class. The most recent review of this literature as well as empirical testing of access and enrollment upholds the continued pervasiveness of stratified differences and shows how the disparities accumulate over students’ educational trajectories (Domina, McEachin, et al., 2016).

Between-school tracking research addresses the first stage in the pipeline: access. This literature answers the question: What are the differences in schooling opportunities for students? During the decades of Jim Crow and residential redlining, the answers were clear: students of different racial and ethnic backgrounds attended separate and segregated schools (Tyack, 1974). Ending de jure segregated schooling after Brown v. Board of Education altered between-school tracking on access to curriculum to de facto policy.

De facto tracking between schools is often addressed by demonstrating the spatial differences in district locale on students’ access to differentiated curriculum (Klopfenstein, 2004). Rural districts often do not have enough numbers of students to offer differentiated sections of a subject course, such as an honors history and a regular history course (Cisneros et al., 2014; Gagnon & Mattingly, 2016; Iatarola et al., 2017). This especially impacts students of American Indian heritage who predominately attend schools in overwhelmingly rural areas (Apthorp, 2016; Christenson, 1996; DeVoe & Darling-Churchill, 2008). Courses in these schools adopt a “teach to the middle” approach (Iatarola et al., 2017; Solórzano & Ornelas, 2004). In urban-core districts, the student numbers are larger, but academic priorities are not geared to prioritize accelerated courses, as these are considered optional offerings among other mandated academic services (Iatarola et al., 2017).

Although resources differ between the rural and urban-core districts, implications on educational inequality are similar between these spaces (Roscigno et al., 2006). Both types of districts struggle to attract and retain highly qualified teachers (Carver-Thomas & Darling-Hammond, 2017). Rural and urban-core districts inordinately staff their classrooms with more novice and fewer advanced credentialed teachers than urban-ring or suburban districts (Gagnon & Mattingly, 2016; U.S. Department of Education Office for Civil Rights, 2014).

Within-school tracking research addresses the second stage in the pipeline: enrolling in the course. The research on within-school tracking in the late twentieth century assessed the educational outcomes of assignment of students to a program of courses; typically, students were assigned to college prep, general, or vocational (Bryk & Thum, 1989;Gamoran & Berends, 1987; Hallinan, 1996; Rosenbaum, 1976; Sørensen, 1970). Due to the detrimental effects of tracking on inequality (Oakes et al., 1992), tracking as a program of courses evolved into subject-by-subject ability grouping (Lucas, 1999). The change from rigid programmatic tracking a more flexible approach still theoretically allowed students to enroll in courses per individual abilities, needs, and wants. In practice, the more flexible subject-by-subject ability grouping approach continued to reproduce parental, social, cultural, and educational advantages where parents’ assets of social networks and know-how advance their children into accelerated courses (Eaton, 2001; Kelly, 2004; Lareau, 2000; Lareau & Horvat, 1999; Lewis & Diamond, 2015; Oakes & Guiton, 1995). Tracking and ability grouping are persistently “sticky” in that students rarely advance beyond their initial track or ability group assignment (Gamoran & Mare, 1989; Kelly, 2009; Lucas, 1999; Lucas & Berends, 2002; Oakes & Guiton, 1995). This stickiness acts to de facto segregate high school course taking (Lucas, 1999; Lucas & Berends, 2002; Tyson, 2011). Together, this research on disparate access and enrollment highlights the differences in the allocation and use of schools’ and districts’ organizational resources.

Mastery

Although many advanced courses have similar titles, titles do not guarantee students’ exposure to the same level of rigor, content, or instructional quality (Domina, Hanselman, et al., 2016; Hallett & Venegas, 2011; Kelly & Price, 2011). Course quality is difficult to determine and certify in American schools due to the lack unified curricula and standards. Content considered advanced or accelerated in rural or urban-core districts may be considered the standard course of study in a neighboring urban-ring or suburban district (Domina, Hanselman, et al., 2016; Remillard et al., 2017). Examining the credentialing related to course mastery in the American primary and secondary school context is often relegated to course grades since no national course exams exist (Graham, 2005). State-based testing assesses general grade level proficiencies in subjects, but this does not link back to the nuanced differences of content between classrooms, such as between an honors and regular U.S. history course.⁵

Most of the research on differences in quality of curricula comes from qualitative comparative work. These qualitative studies often cite differences in teacher training, social-structural school resources, and family resources as reasons underlying differences in course rigor and mastery (Cisneros et al., 2014; Gagnon & Mattingly, 2016; Lareau, 2000; Oakes, 2005; Palmer, 2016).

Recent research on eighth-grade algebra exemplifies the idea of varied rigor in course content. Courses may share the same name, but content can and will vary.⁶ Domina et al. (2015) find that California’s algebra-for-all-eighth-graders policy induced schools to assign all students to eighth-grade algebra. In practice, teachers explain that they are only able to teach pre-algebra in many of their classrooms due to the preponderance of below grade-level math skills of students where teachers need to begin with more remedial teaching at the start of the course (Domina et al., 2015; Penner et al., 2015). On transcripts, all California students earn a grade in “eighth-grade algebra” yet mastery varies from general eighth-grade math to pre-algebraic to algebraic content (Domina et al., 2015; Penner et al., 2015). It is only the eighth-grade math exam that assesses proficiency in algebra across these different courses titled “algebra.”

A course exam offers a credential that attests to the mastery of the subject and provides a transferrable educational good (Bourdieu & Passeron, 1977). Without the exam credential, a course title on a transcript simply exchanges for lesser value because no third-party appraised the course quality (Bourdieu & Passeron, 1977; Evans, 2019). A transcript with an AP or IB course listed may provide a signal to college enrollment recruiters about students’ potential to succeed in college⁷, but the externally certified AP exam score or the IB diploma legitimates the course quality in exchange for college credit (Evans, 2019; Hallett & Venegas, 2011). In the case of AP, university policies of “3 or higher” exam scores set the exchange rate for mastery that bypasses three credits of college tuition (Evans, 2019; Kolluri, 2018).

Analytic Approach

AP and IB classes provide curricula with third-party recognized quality and extends beyond state variations in grade-level standards. In addition, colleges and universities legitimate the quality of these college prep courses by including them in admissions criteria (College Board, 2018; Kolluri, 2018).

This study thus uses data on AP and IB coursework to represent a substantial portion of college prep curriculum across the United States. With these data, the following research questions are asked,

Research Question 1: To what extent do racial/ethnic disparities exist in American high schools along the three stages in the college prep pipeline, from access to enrollment to mastery?

Research Question 2: Which social contexts related to resource allocations exacerbate or alleviate racial/ethnic disparities in the college prep pipeline?

Districts and schools are the units of analyses in the study since these organizations disperse the curricula among their students.

Data and Measures

The Civil Rights Data Collection (CRDC) gathers population data from all public schools on AP and IB course enrollments and test taking (U.S. Department of Education Office for Civil Rights, 2014). Unlike other school curricula that are state regulated, these courses are nationally monitored by agencies external to the local, state, or federal education agencies. The CRDC is the first of its kind to collect a census inventory that counts all AP and IB student enrollments and pass rates from every public traditional and charter school in the United States. The CRDC also provides data specific to each AP subject, but given the premises of this study, the data for IB and all AP subjects are combined to assess organizational-level disparities in educational opportunities.

In each year of the biannual data collection, the CRDC collects subgroup-level data on the gender and racial/ethnic enrollments for each course in each school. In the 2011–2012 and 2013–2014 school years, the racial and ethnic classifications were as follows: Latinx,⁸ American Indian, Asian, Native Hawaiian or Pacific Islander, (non-Latinx) White or Black, and multiracial.⁹ For the districts where more than 1% of their students were excluded from the racial/ethnic subgroup counts, this study adds a “no answer” subgroup to capture disparity among the students whose race or ethnicity were left unclassified. The data from schools managed by state or federal agencies, such as juvenile detention facilities and hospitalization schools, are excluded from these analyses.

This CRDC also gathers data on school enrollment totals, Title 1 school status, and school type (charter, magnet, traditional, state managed). District and school data from the Common Core of Data from the National Center for Education Statistics supplement the CRDC data to measure locale and teacher certifications.

Disparity in the College Preparatory Pipeline

Disparity measures provide metrics to understand the extent to which groups dominate a “market” (a form of opportunity hoarding, see Kelly & Price, 2011).¹⁰ This study uses a version of the HHI to measure the amount of disparity along the three stages in the college prep pipeline (Taagepera & Ray, 1977). The more equitable the market share across all groups, the closer the HHI gets to zero. The higher the HHI value, the greater the disparity in market share between groups. An extremely high value indicates a single group monopolizing a market. This HHI focus on concentration and clustering differs from the evenness assumptions of many segregation-based indices common to education research (Reardon & Firebaugh, 2002).

Social movement research and political social science research commonly use the HHI measure to represent degrees of segregation (Boydstun et al., 2014; Taagepera & Ray, 1977). The HHI allows for multiple group representations in a single measure rather than a set of dichotomous relations (White-Black, White-Latinx, etc.), which is the formulaic basis of the segregation indices of Gini, Theil, and others. If, for example, a measure with dichotomous relations was used, there could be up to 21 combinations analyzed for each of the three stages if all seven racial and ethnic affiliations were represented in the school. This study therefore adopts an adjusted HHI to estimate the amount of monopolization by subgroup affiliations over the college prep curriculum within school markets (Price, 2019).

The standard HHI assumes that each group has the same chance as any other group in the market (one group, one chance). Since district and school student subgroups are not the same sizes, this study normalizes the HHI estimate to adjust for the varied proportional representation of students in districts and schools at each stage in the pipeline (Price, 2019).

The HHI measure of market concentration also adjusts well to the “moving denominator” between stages in the college pipeline trajectory, whereas most segregation indices assume a single, static event (Price, 2019). As an example of this dynamic nature across the three pipeline stages, imagine a district where all students attend one high school. In this school, 90% of the students in AP Calculus are 12th graders and the other 10% are 11th graders. This school has a quartiled student body of ninth, 10th, 11th, and 12th graders and so all students in the school would have access at Stage 1 (HHI_access = 0), but the 12th graders hold a “corner on the market” of enrollment at Stage 2 (where HHI_enrollment = 2.3). This score above 1 indicates a near monopoly in the market of enrollment. At Stage 3, if all of the 11th graders demonstrate mastery (receive a 3 or higher score) but only half of the 12th graders do, then the disparity is greater than Stage 2 where HHI_mastery = 8.3. This extremely high score shows that one group overwhelmingly dominates the Stage 3 market compared to the other group. Put another way, this extremely high HHI shows that Stage 3 heavily favored one group (11th graders) while diverting out a substantial proportion of the other group (12th graders), all of whom were in the market for a mastery credential.

The adapted HHI calculations for measures of racial/ethnic disparity along the three stages in the pipeline are as follows:

Access disparity measures the disproportionate racial/ethnic representation of students who attend the schools that offer college prep curriculum compared to all the high school students in the district; a between-schools, within-district measure.

HH I_{access} = \sum_{j = 1}^{J} \frac{{(Π_{j} - π_{j})}^{2}}{Π_{j}},

(1)

where Π is the proportion of the district high school population, π is the proportion of high school students attending the district schools with AP or IB, and j is the subgroup designation.

Enrollment disparity measures the disproportionate racial/ethnic representation between students enrolled in the college prep curriculum as compared to their representation among the general student body within the same school; a between-classrooms, within-schools measure.

HH I_{enrollment} = \sum_{j = 1}^{J} \frac{{(π_{j} - (t_{j} + r_{j}))}^{2}}{π_{j}}

(2)

where π is the proportion enrolled in district schools with AP or IB, t is the proportion enrolled in AP classes, r the proportion enrolled in IB classes, and j is the subgroup designation.

Mastery disparity in college prep coursework measures the amount of disproportionate racial/ethnic representation among those students in AP or IB classes who pass the exam¹¹; a between-students, within-classroom measure.

HH I_{mastery} = \sum_{j = 1}^{J} \frac{{(t_{j} - q_{j})}^{2}}{t_{j}},

where t is the proportion enrolled in AP, q is the proportion passing at least one AP exam with a score of 3 or higher, and j is the subgroup designation.¹²

All calculations are restricted to high school age students only (Grades 9–12); population proportions exclude preK–Grade 8 students.

Proportionate representation (HHI = 0) could result when the student population comprises of only one racial or ethnic group. While 100% of a single racial or ethnic group composition is rare in U.S. districts, hypersegregation within schools and classrooms is less uncommon. Table 1 highlights the prevalence of districts (Stage 1) and schools (Stage 3) that serve diverse student populations that succeed at proportionate representation along the pipeline route (HHIs = 0).

Table 1

Distribution of Race/Ethnicity in Student Body for Proportionate Representation (HHI = 0)

	“Diverse” (Less Than Homogeneous) Student Body	90%–94.9% Single Race/Ethnicity	95%–100% Single Race/Ethnicity
HHI_access^a	51.3%	33.5%	15.3%
HHI_enrollment	N/A	N/A	N/A
HHI_mastery	61.2%	17.9%	20.9%

Note. HHI = Herfindahl-Hirshman Index, adaptation.

HHI_access tabulations at the district level.

Social Context

Social context shapes where resources are allocated and thus how students experience learning. Three elements of social context are measured in this study: curricular policy, spatial locale, and peers.

Curricular policies on college preparation can be understood using measures of course availability to students and teacher availability to teach the courses. Course availability is measured using the proportion of classroom AP and IB seats in relation to the total student population.^13,14 Teacher availability is loosely measured using the number of certified secondary education teachers in the district. A more ideal measure would link teacher qualifications to courses taught, but this information is not nationally collected.

Resources associated with spatial locale are measured using four categories of rural, town, suburban, and urban. In addition to the teacher qualification differences discussed earlier, Roscigno et al. (2006) discuss the family and school economic differences by locale impacting students’ schooling experiences.

Measures of school sociodemographics capture the peer context element. Some research stresses how the Whiteness of the American schooling system biases schooling experiences and expectations toward White students (Cherng, 2017; Fisher et al., 1996; Orfield & Eaton, 1996; Tilly, 1998). However, in cases of schools with only a few White families enrolled, bias dampens and the voices of people of color can be heard (Anyon, 2014; Yosso, 2005). Other research stresses socioeconomic class dominating the schooling expectations and student experiences (Bourdieu & Passeron, 1977; Dietrichson et al., 2017; Kelly, 2004; Lareau, 2000; Lareau & Calarco, 2012; Salloum et al., 2017). Moreover, contemporary American studies call on the inextricable intersectionality of class and race in American schools (Carter, 2005; Cipollone & Stich, 2017; Ellison & Aloe, 2018; Hallett & Venegas, 2011; Lareau & Horvat, 1999; Ndura et al., 2003; Posey-Maddox, 2012). Given these complementary ideas, this study assesses the prevalence of the schools’ Whiteness, majority minority composition, class (Title 1 status¹⁵), and the intersection of race and class.

Last, resource allocations in nontraditional schools differ. Charter and magnet schools operate under more independent funding and looser teacher certification regulations (Berends et al., 2009; Cannata & Penaloza, 2012). The self-selection to enroll in a charter or magnet school preemptively sorts social context (Berends et al., 2009; Riel et al., 2018; Sattin-Bajaj & Roda, 2018; Renzulli & Roscigno, 2005). For all these reasons, charter and magnet schools are flagged in the analysis to isolate the effect on the three focal social context elements.

Table 2 details the descriptive statistics of the data used in this study. The descriptions show 13% of schools with college prep offerings show no disparity between peer groups of passing test scores among their enrolled college prep students. Also, more than 40% of districts evidence no disparity among students’ access to these college prep offerings among their schools. However, five of six of these districts with equitable access only have one high school and so it is impossible for disparity to exist. Thus, this study adds a flag for districts that operate only one high school.

Table 2

Summary Statistics

	Count of Schools	Mean or Proportion	SD	Minimum	Maximum	Proportion HHI = 0
Outcomes
HHI_access	30,461	0.008	0.070	0	3.607	0.403
HHI_enrollment	25,797	0.404	3.514	0	318	0.006
HHI_mastery	22,034	0.221	0.325	0	6.055	0.131
School policy
College preparatory seats (proportion)	30,461	0.111	0.141	0	1
Spatial location
Rural	30,461	0.378		0	1
Town	30,461	0.144		0	1
Suburban	30,461	0.256		0	1
Urban	30,461	0.222		0	1
Peers
40% or more FRPL students (Title I school)	30,461	0.482		0	1
95%–100% White students	30,461	0.105		0	1
95%–100% students of color	30,461	0.115		0	1
95%–100% students of color and Title 1	30,461	0.080		0	1
95%–100% White students and Title I	30,461	0.052		0	1
Control variables
Magnet school	30,461	0.050		0	1
Charter school	30,461	0.098		0	1
Total school enrollment	30,461	771	746.254	2	23010
Count of secondary certified teachers	30,461	649	1397.094	0	8677
95%–100% single race-ethnic student body	30,461	0.240		0	1
Only one high school in district	30,461	0.425		0	1
ELL (proportion)	30,461	0.048	0.106	0	1
IDEA (proportion)	30,461	0.122	0.083	0	1.222

Source. Civil Rights Data Collection and Common Core of Data, pooled 2011–2012 and 2013–2014 school year data sets.

Note. HHI = Herfindahl-Hirshman Index, adaptation; FRPL = free or reduced-price lunch status; ELL = English language learner status; IDEA = Individuals with Disabilities Education Act status.

Additional controls in the models that mitigate resource allocations are size of the student population and special education needs.¹⁶ Schools and districts with 95% to 100% homogeneous racial or ethnic student populations are also controlled since the chance of racial/ethnic HHI disparity is arithmetically low when there is nearly no variation on racial or ethnic composition.¹⁷

Data Limitations

This study only speaks to organizational-level patterns. Using transcript data, Domina, Hanselman, et al. (2016) recently studied the longitudinal impact of curricular opportunities at the individual student level. They found clear evidence of divergent effects on individual student achievement and the reproduction and accumulation of inequality in educational opportunities within only 3 years based on students’ initial start course. With that said, the organization-level patterns tested here are important to understand loss of input impacts college readiness.

Unless they are titled with AP or IB the CRDC data do not ask about other advanced courses such as those addressing specialized science or literature topics. This limitation implies that these estimates narrow the definition of college prep from the broad scope of offerings that may expose students to college-level curricula. In addition, the CRDC does not track the credentialing of the IB program so this analysis imputes these few students across the exam measures to assume the acquisition of the credential. Extant analyses did test the sensitivity of the estimates by excluding these students; there were insignificant differences.

Method

To address the first research question, simple proportions of racial and ethnic subgroups of American students who remain in, compared with those who leak out, of the three stages of the college pipeline are presented. Since this study used population/census data, differences in subgroup populations are real and do not require sample-based statistical tests. Nonetheless, all t tests of the proportional differences between and within (“in” vs. “out”) subgroups are statistically significant, p < .001. Data are pooled across the two school years because it smooths averages for the smaller and more volatile subgroups of American Indian, Hawaiian or Pacific Islander, and multiracial student subgroups.¹⁸

To address the second question, negative binomial regression models are used since the dependent HHI variables follow a Poisson distribution with a truncated left end and a long right skew. For access and mastery, 40% of districts and 13% of schools, respectively, demonstrate proportionately equal representation (HHI = 0). In these two models, zero-inflated negative binomial regression models are estimated in order to parse the analysis to describe the social context of the organizational resources that (a) explain variation in disparity (an exponentiated outcome estimate) and (b) the likelihood to associate with no disparity (HHI score = 0, a binary logit outcome estimate). Social context could work differently since (a) predicts estimates of disparity and (b) predicts the likelihood of an absence of disparity.¹⁹ Logistic regression analyses (not shown here) precede these disparity models to support the necessary warrant that these social context elements relate to schools’ and districts’ likelihood to establish each pipeline stage (contact author for details).

Although these data are population data with no selection issues because there is no sampling, there may be differences related to resources associated with state-regulated funding formulae. The models therefore fix effects for the 50 states and the District of Columbia.

Findings

Research Question 1: Racial and Ethnic Disparity in the College Preparatory Pipeline

On average, seven per 100 high school students attend districts that do not offer any AP and IB curricula. Among the students with district access, only 30% of American students attend the schools that offer AP or IB courses while the others attend traditional or charter high schools that do not offer any AP or IB. Six per 100 American high school students enroll in AP and fewer than one per 100 are enrolled in at least one IB course. Between the start and the end of the pipeline, only three per 100 American students enrolled in regular high schools earned at least one IB or AP course credential of mastery.²⁰ Put differently, 97 per 100 high school students leak out of the pipeline at some stage.

Although it could be the case that all high school students have access to AP or IB in their school, a substantial number of high school students are likely too early in their high school education to enroll in AP or IB. The IB Diploma Programme requires students to be at least 16 years old to enroll (International Baccalaureate Organization, n.d.). For AP, most schools require substantial prerequisite courses to accumulate to be eligible to register for the AP class (Kelly & Price, 2011). Between the typical high school Grades of 9 to 12, we would expect nearly zero students in Grade 9 and 10 to enroll in any of the three dozen AP courses—this immediately pushes the ratio to 50 per 100. Although all seniors could theoretically enroll in AP or IB if their school offered courses, it would be unlikely under our current educational norms. Similarly, fewer juniors than seniors would be assumed to enroll in these courses. Among high schools that enroll all seniors in at least one AP course, their enrollment ratio would be 25 per 100 and their exam mastery would not likely be 100%.²¹

Figure 2 shows how the pipeline selection process systematically differs by racial and ethnic affiliation. Given the decades-long discussion on tracking, this analysis only focuses discussion three historically marginalized groups—American Indian, Black, and Latinx students—compared to White students. Although most traditional public high school students attend districts that offer college prep curriculum, only 87% of American Indian students attend districts with AP or IB courses, leaving American Indian students more than twice as likely as other students to not have access to AP or IB anywhere in their district. Moreover, fewer than 24% of American Indian attend the schools that offer AP or IB, whereas 27% of Latinx and 30% of Black and White students attend these schools.

Figure 2.

College preparatory curriculum pipeline by student race or ethnicity.

When using the moving denominator among those students attending schools with college prep courses, 18% of students in these schools are enrolled in at least one college prep course. The gap is wide, however, from 20% of White students to only 11% of Black, 13% of American Indian, and 15% of Latinx students enrolled. Included in these proportions are the small enrollments in IB courses, but there is little difference between the 1.5% of students from each of these four racial and ethnic groups enrolled in IB.

Demonstrating mastery on AP exams differs widely by racial and ethnic group affiliation. On average, 20% of students enrolled AP courses never take an AP exam, while another 54% students attain a passing score to demonstrate mastery of the course material. Approximately 55% of White students pass an AP exam, while 39% of Latinx and 30% of American Indian and Black students attain the mastery credential with a passing exam score of “3” or higher.

These results provide evidence in response to the first research question about the prevalence of racial and ethnic disparities along the three stages in the college prep pipeline. American Indian students leak out at every stage. After Stage 1, Black students leak out at Stages 2 and 3. Latinx students leak out disproportionately at the last Stage, 3, by failing to attain a passing exam score.

Research Question 2: Social Context Explaining Where the Pipeline Leaks

Table 3 shows the series of models that identify the social context elements associated with students’ experiences of racial/ethnic disparities along the pipeline route (Research Question 2). HHI columns estimate factors that exacerbate (positive coefficient) or alleviate (negative coefficient) the magnitude of disparities when they exist. The two HHI = 0 columns (Models 2 and 5) estimate factors that increase or decrease the likelihood of proportionate representation when no discernable racial/ethnic disparities exist at that stage in the school.

Table 3

Negative Binomial Regression Estimates of Disparity Along the College Preparatory Pipeline, With Zero-Inflated Estimates

	1	2	3	4	5
	Stage 1: Access		Stage 2: Enrollment	Stage 3: Mastery
	HHI_access	HHI_access = 0	HHI	HHI_mastery	HHI_mastery = 0
Prior disparity
HHI_access			0.014 (0.012)
HHI_enrollment				0.023** (0.009)	0.028 (0.016)
Curricular policy
College preparatory seats in district (proportion, ln)	−1.904*** (0.086)	0.309*** (0.034)
College preparatory seats in school (proportion, ln)			−1.006*** (0.034)	−0.095*** (0.010)	−0.126*** (0.013)
Count of secondary certified teachers (ln)	−0.070 (0.079)	−0.054 (0.038)	0.065* (0.031)	−0.003 (0.014)	−0.009 (0.017)
Spatial location (compared to suburban)
Rural	−0.213 (0.135)	−0.253*** (0.032)	−0.158*** (0.046)	0.097*** (0.021)	0.022 (0.026)
Town	−0.190 (0.126)	−0.282*** (0.033)	0.003 (0.053)	0.110*** (0.022)	0.003 (0.027)
Urban	−0.388*** (0.112)	−0.192*** (0.023)	0.088* (0.039)	0.054** (0.018)	0.053* (0.023)
Peers
Title I district	0.022*** (0.003)	−0.003*** (0.001)
40% or more FRPL students (Title I school)	0.201 (0.148)	−0.166*** (0.022)	−0.047 (0.033)	0.176*** (0.016)	0.015 (0.020)
95%–100% White students	−0.268 (0.196)	−0.108* (0.054)	−0.085 (0.121)	−0.029 (0.035)	0.114* (0.045)
95%–100% students of color	0.963*** (0.224)	−0.077 (0.065)	−0.150 (0.164)	0.614*** (0.063)	−0.033 (0.084)
95%–100% students of color and Title I	−0.448 (0.254)	−0.033 (0.071)	0.094 (0.189)	−0.457*** (0.067)	0.176* (0.090)
95%–100% White students and Title I	−0.091 (0.236)	0.023 (0.083)	−0.087 (0.109)	−0.026 (0.045)	−0.021 (0.056)
Control variables
Magnet school	0.328 (0.212)	−0.101*** (0.028)	0.220*** (0.062)	0.132*** (0.022)	−0.018 (0.032)
Charter school	^a	^a	−0.271*** (0.082)	−0.075* (0.035)	0.024 (0.046)
Only one high school in district	^a	^a	−0.136** (0.045)	−0.029 (0.019)	0.002 (0.025)
Total enrollment (ln)	−0.077 (0.048)	−0.041*** (0.009)	−0.260*** (0.035)	−0.170*** (0.016)	−0.205*** (0.016)
Total high school district enrollment (ln)	−0.360*** (0.099)	−0.023 (0.037)	−0.151*** (0.037)	0.014 (0.017)	0.036 (0.020)
ELL (proportion, ln)	−0.032 (0.042)	0.083*** (0.008)	−0.069 (0.057)	0.044*** (0.012)	0.029* (0.012)
IDEA (proportion, ln)	0.034 (0.030)	0.055*** (0.006)	0.034* (0.015)	0.012* (0.005)	0.001 (0.006)
95%–100% single race-ethnic district	0.083 (0.152)	−0.172*** (0.045)	−0.171 (0.121)	−0.034 (0.030)	−0.074 (0.042)
Constant	−2.669*** (0.617)	−28.963*** (0.185)	−1.020 (0.533)	0.149 (0.179)	−42.819*** (0.219)
Observations	17,562	17,562	22,613	22,034	22,034

Source. Civil Rights Data Collection and Common Core of Data, pooled 2011–2012 and 2013–2014 school year data sets.

Note. HHI = Herfindahl-Hirshman Index, adaptation; FRPL = free or reduced-price lunch status; ELL = English language learner status; IDEA = Individuals with Disabilities Education Act status.

Access HHI model excludes charter and single-school districts since the school high school representation is the same as the comparison population.

p < .05. **p < .01. ***p < .001. Fixed effects for state and year.

Stage 1: Disparity in Access

Disparity in access at Stage 1 regards whether, in the same district, students who attend schools with AP or IB curricula affiliate with racial or ethnic identities differently than their peers in other high schools that do not offer these college prep curricula. Models 1 and 2 therefore restrict the estimates only to the multischool districts because single-school districts and charter schools will immediately score HHI = 0 since the school population is the arithmetically identical to the whole high school population in the district or charter school.

Model 1 shows that curricular policy on the volume of AP or IB availability does reduce disparity; districts with more AP and IB seats per pupil average lower rates of racial disparity in access to college prep curricula. The availability of secondary-certified teachers does not matter one way or the other regarding disparity in access. Spatial location advantages students attending urban schools: Students attending urban schools experience less racial disparity in access than other students. Urbanicity stands independent of the peer context in the school where the predominance of poverty in schools does not relate to differentiated access. Students attending a hypersegregated school serving students of color experience higher levels of disparity, however, compared to their peers in other district high schools.

Stage 1: Equality in access

Model 2 shows that policy on the volume of AP or IB availability improves the chances of racially proportionate representation in access to college prep curricula (Model 2, an HHI_access = 0). Suburban schools are the most likely to equalize access, while rural and township schools are the least likely. The peer context element works quite differently than disparity in Model 1: The predominance of student poverty reduces the chance of proportionate representation in access and hypersegregated schools serving White students are less likely to allocate access equally to their students even though less than 5% of their students are not White. Students of color attending hypersegregated schools do not experience this worsened chance of proportional representation.

Stage 2: Disparity in Enrollment

Disparity in enrollment at Stage 2 identifies the differences in racial or ethnic representation between students who sit in the AP or IB classrooms compared to the students who sit in the lunchroom. Model 3 shows that policy on the sheer availability of AP or IB seats in a school reduces enrollment disparity; schools with greater AP or IB availability average lower rates of racial disparity in enrollment to college prep curricula. Greater availability of secondary-certified teachers, however, associates with higher levels of disparity. By spatial location, rural schools average lower enrollment disparities while urban schools average slightly higher disparities compared to other locales. Peer context elements of poverty or hypersegregation levels do not play a significant role in racial/ethnic disparity in AP or IB enrollment.

Stage 2: Equality in enrollment

Racial/ethnic disparity at Stage 2 occurs ubiquitously (in 99.4% of schools). Thus, no model estimates where HHI_enrollment = 0.

Stage 3: Disparity in Mastery

At the final stage in the pipeline, disparity in mastery at Stage 3 assesses the extent of racial/ethnic disproportionality between students earning a passing score of 3 or higher on their exam. Students can fail to earn mastery two ways: they either not take any exam one or they take it and earn a low score of 1 or 2. Again, the results show that the volume of AP or IB offerings in the school reduce disparity, albeit a much smaller influence at this stage. The availability of secondary-certified teachers does not sway mastery disparity levels. The magnitude shrinks for the spatial location element of social context; students attending urban schools experience slightly more disparity than their suburban counterparts, but less disparity than if they attended a rural or township school.

What matters the most at Stage 3 is peer context. Schools receiving Title I funding average greater racial/ethnic disparity in mastery among their students than other schools. Hypersegregated schools serving 95% or more students of color also average greater disparity in mastery between their students who enrolled in AP or IB. The two characteristics do not compound, however, Title I schools that predominantly serve students of color overcome much of the disparity that otherwise occurs. Title I schools serving 95% to 100% White students do not act the same way.

Stage 3: Equality in mastery

Only a few social context factors relate to proportional representation in mastery. Policies related to offering more volume of college prep offerings lessens the chance of proportionate racial/ethnic representation on mastery. Students enrolled in college prep in urban schools also experience slightly greater likelihood of proportionate representation. Students enrolled in AP or IB who attend hypersegregated White schools or students who attend Title I schools that predominately serve students of color are both more likely to experience proportionate representation on mastery.

Control Variables

These social context elements stand independent of school type, school or district size, the proportion of special education or limited English proficiency needs, single high school districts, state education funding, or whether or not the district Grades 9 to 12 student body is composed of 95% or more of all the same race or ethnicity. Thus, findings such as the impact of students in rural schools experiencing less enrollment disparity is not because those students all attend a single high school where nearly all of the students are White. Instead, the spatial location of rural schools stands independent, no matter the district size, design, or student composition.

Research Question 2: Summary

Together, these models show that some social context elements at an earlier stage continue to impact later stages in pipeline while other elements are more stage-specific. The prevalence of curricular policies to make more available college prep curriculum to students reduces disparity throughout the pipeline stages of access, enrollment, and mastery. Additionally, the policy of hiring teachers with a basic secondary certification does not help reduce disparities.

Spatial location is stage-specific. It particularly distinguishes proportionate representation in access and disproportionate representation in mastery. In both cases, students attending suburban schools fare better than students attending other schools. However, suburban locale is not always advantageous: Students attending urban schools experience less discrimination in access and students attending rural schools experiences less discrimination in course enrollment.

The social context question of whether the racial composition or economic status of peers affects the level of monopoly in the college prep marketplace proves quite nuanced. First, attending a hypersegregated White school does not reduce disparity at any stage. This means that the few (less than 5%) students of color attending these hyper-White schools do not experience more or less discrimination than if they attend a more diverse school. However, attending a hypersegregated school that overwhelmingly serves students of color with few to no White students does widen discrimination in access and mastery. In addition, high schools serving many students experiencing poverty substantially widens discrimination at Stage 3 of mastery. However, the schools’ resource uses associated with these peer contexts work independently of each other as evidenced by the finding that hypersegregated Title I schools serving nearly all students of color do not compound disparity on mastery but instead overcome much of it.

Last, these models account for prior disparity in the pipeline. Results show that disparity in-and-of-itself does not substantially accumulate from one stage to the next stage in the pipeline.

Discussion

This study explores a new 21st-century high school tracking issue: The rise and newly established prevalence of college prep curricula in high schools and the disparate opportunities that result from it. Before this study, the extent to which this new tracking system differentiated academic opportunities among American students throughout the pipeline stages was not understood. Prior studies on AP access and enrollment for one group of students or another proved invaluable to hypothesize on this larger structural issue of the complex, interlocked stages in this pipeline. This study traced the disparities along each of the three stages using two guiding questions: Which groups of students experience disparities at which stage in the pipeline and where do social contexts exacerbate or alleviate disparities for those students remaining in the pipeline? Focusing the results on the Latinx, American Indian, Black, and White student subgroups spoke to the intersection of the urban/rural divides and different experiences with racial discrimination in schooling.

Answers to both questions clearly evidence unequal opportunities. The findings related to the first research question show stark racial/ethnic differences in college prep opportunities among students. American Indian students are disproportionately diverted out of the pipeline early with relatively low access to college prep curricula anywhere in their district or their schools. These students are also disproportionately underenrolled and do not pass the exams at higher than average rates. Black students are disproportionately diverted from enrolling in courses and, among those one-in-ten who enroll, two thirds leak out without a passing score. Latinx students tow the average along the pipeline route with White students until Stage 3 where they experience disproportionately low pass rates on exams. These findings coincide with case studies (Ndura et al., 2003; Solórzano & Ornelas, 2004; Tyson, 2011) and state (Cisneros et al., 2014; Klugman, 2013; Moore & Slate, 2008) and nationally representative (Gagnon & Mattingly, 2016; Judson & Hobson, 2015; Moore & Slate, 2010) studies that look at single stages along the AP course-taking pipeline.

Put together, this study’s interlocked pipeline analysis points to the accumulation of structural inequality. Understanding where these inequalities occur can inform policy discussions to address these academic juggernauts that particularly disenfranchise students from historically marginalized groups.

The diseconomy of resources theory provides a means to test how organizational resource allocations associated with social context structurally exacerbate (Coleman’s “loss of input”) or alleviate these persistent racial/ethnic inequalities. The administrative decision to allocate resources to add college prep curriculum holds promise to reduce racially and ethnically disparate opportunities. Adding in-name-only titled AP courses, however, do not deliver on mastery of the course content (Cipollone & Stich, 2017; Domina, Hanselman, et al., 2016; Hallett & Venegas, 2011) although these findings suggest that even if courses were added in-name-only, the declines would not be unequally shouldered by some students but rather all students would equally experience poorer quality. Focusing policies on the most basic threshold of staffing secondary-certified teachers is not adequate enough to benefit students’ experiences in college prep.

The social context of spatial locale especially differentiates mastery at Stage 3. The limitations of resources in non-suburban areas create disparate outcomes. This finding may point to the differences in teacher quality across locales that impact the rate of learning (see Carver-Thomas & Darling-Hammond, 2017), or it may point to the differences in parental educational advantages in the rate of learning (see Roscigno et al., 2006). These data cannot speak to the mechanism related to spatial locale, but future research would want to gather national data to further inform policy.

The peer context findings speak less to the premise that class (income) of peers associates with differentiated resource allocation and instead speaks much more to the research strand regarding Whiteness hindering learning opportunities for students of color (see also Davies & Rizk, 2018; Lewis & Diamond, 2015; Tyson, 2011). If disparity was truly an underlying class-based mechanism, the measure of Title 1 schools serving 95% to 100% White students would stand out, but it does not. Moreover, the impact of concentrated poverty (Title I) on disparity levels pales in comparison to racial or ethnic hyper-segregation measures. This finding coincides with a recent test of the “school socioeconomic status effect” on learning by Armor et al. (2018) who find the impact of free or reduced-price lunch rates are mostly an artifact of other school contexts, including racial composition.

To push these social context ideas further, this study also controlled for specialty schools with different regulatory standards since some scholars propose that these schools offer a condition where the norms and expectations around schooling are shared from the onset with selection to attend the school (Ellison & Aloe, 2018; Riel et al., 2018; Sattin-Bajaj & Roda, 2018). Independent of the homogeneity of the specialty schools’ peer composition, attendance selection appears to pan-out in charter schools to reduce the disparity between students within schools. Magnet schools, however, follow an inverse effect where racial/ethnic disparity magnifies in these schools. Further research will want to understand why resource allocations differ among these nontraditional school types.

In conclusion, the diseconomy of resources framework argues that it is not just about the availability of resources but also their allocation. These findings lend support to the notion that administrators can differentially allocate their resources to improve the chance of equal opportunities for students, reduce discriminatory educational opportunities, and stop leakage from the pipeline. Even if a district is fraught with disparity, the finding that disparity at one stage along the pipeline has no substantial impact on the next stage holds promise because it means that school administrators’ resource allocations can work internally to alleviate in-school disparity.

Implications

The prevalence of disparate opportunities along this new college prep pipeline indicates that the current system fails to deliver on the promised exchange value of college prep coursework. Students can be short-changed on the exchange rate on the educational good they think they are acquiring in high school by taking college prep courses but not actually gaining mastery of the content.

The evidence here shows that this educational pipeline disparages historically marginalized students at rates higher than their advantaged peers and this disparity structurally exists between schools and districts. For some students, like American Indian students, the pipeline is leaky at all stages. For other students, the sediment in the pipeline slowly builds up over time and eventually creates a significant and detrimental effect on their learning. These CRDC data allow the mapping of the AP and IB from access to enrollment to mastery in efforts to assess the extent to which social context elements compound disparities in academic opportunities.

Although college prep curriculum is in-place for the overwhelming majority of American high school students to access, there are still a number of districts with no access (the first sequence in the diseconomy of resources regarding resource quantity). If students are lucky enough to be in a district with access, they then need to attend the schools with these courses, get enrolled in them, and master the content. Currently, all resources allocated for these stages vary by the school or district and depend on the elements of social context (the second sequence in the diseconomy of resources). Historically marginalized groups of American Indian, Black, and Latinx students are disenfranchised and experience a loss of input under the current pipeline structure.

These findings build on years of equality research to evidence the cumulative divergence of curricular opportunities in the U.S. educational system. The findings for research question one on racial and ethnic disparities in access, enrollment, and mastery dovetail with the educational studies on resegregation of educational opportunities (Clotfelter, 2004; Frankenberg & Orfield, 2012; Mickelson, 2003; Orfield & Eaton, 1996; Orfield & Lee, 2005, 2006; Quillian, 2014) as well as the current research regarding the unequal learning experiences of students in American high schools (Frankenberg & Orfield 2012; Judson & Hobson, 2015; Lewis & Diamond, 2015; Mickelson, 2003; Tyson, 2011). The findings complement studies on within-school segregation by moving the discussion forward to understand the diseconomy of resources that stratify academic opportunities and learning. This study brings to the forefront the prevalence of the allocation differences that have been theorized (Coleman, 1990; Lucas, 1999), but seldom tested with national population data.

Future Research

Where students attend high school shape their opportunities. Differences in how high schools ready students for college provides some explanation to the persistent gaps in college attainment. It is imperative to prepare students in high school to be “college ready” in this era where college enrollment is common, but completion still evades many college students (Snyder et al., 2016). Understanding why “where” matters for loss of input in the leaky pipeline is the next step in this research.

A strength of this study is the comprehensiveness of the population data so that no racial or ethnic subgroup might go unaccounted due to small counts. The gaps measured for the dependent variable are therefore not subject to sampling error. Using these data, however, limit the resource measures to that which could be matched in other federal databases of the population of American schools, like the Common Core of Data. Further studies would benefit from richer resource measures to better understand why leakage occurs. For example, variables on teacher training, social-structural resources, family resources, and links of teachers to the students in their courses could deepen the conversation about resource quality in the leaky college prep pipeline.

To be sure, this study cannot test the intentions of school administrators or teachers. These data cannot reveal if this stratified pipeline of college prep curricula opportunities is due to intended actions of a few in power to maintain segregation and inequality, if this is a product of unintended consequence of school and district policies, and/or if this stratified pipeline is a function of historical and systemic racism in America. Although intention is important to study in order to induce social change, this study begins by evidencing the social problem in order to drive research to understand it more so that students do not endure continued discrimination in their readying for college.

Footnotes

ORCID iD

Heather E. Price

Notes

HEATHER E. PRICE, PhD, is assistant professor and interim Dean of Undergraduate Studies at Marian University, 45 South National Avenue, Fond du Lac, WI, 54935, USA; email: hprice@alumni.nd.edu . Her research focuses on sociology, school climate, teacher commitment, school leadership, and educational policy. Internationally, she continues project work on the OECD’s Teaching and Learning International Survey (TALIS). Prior to Marian, Price worked in the private educational policy sector and taught for years in the Milwaukee Public Schools district.

References

Anyon

(2014). Radical possibilities: Public policy, urban education, and a new social movement. Routledge. https://doi.org/10.4324/9780203092965

Apthorp

H. S.

(2016). Where American Indian students go to school: Enrollment in Seven Central Region States (REL 2016-113). Regional Educational Laboratory Central & Institute for Educational Sciences.

Armor

D. J.

Marks

G. N.

Malatinszky

(2018). The impact of school SES on student achievement: Evidence from US statewide achievement data. Educational Evaluation and Policy Analysis, 40(4), 613–630. https://doi.org/10.3102/0162373718787917

Baldi

Warner-Griffin

Tadler

(2015). Education and certification qualifications of public middle grades teachers of selected subjects: Evidence from the 2011-12 Schools and Staffing Survey (NCES 2015-815). National Center for Education Statistics. https://nces.ed.gov/pubs2015/2015815.pdf

Berends

Springer

M. G.

Ballou

Walberg

H. J.

(2009). Handbook of research on school choice. Routledge. https://doi.org/10.4324/9780203881781

Bourdieu

Passeron

J. C.

(1977). Reproduction in education, culture and society. Sage.

Boydstun

A. E.

Bevan

Thomas

H. F.

(2014). The importance of attention diversity and how to measure it. Policy Studies Journal, 42(2), 173–196. https://doi.org/10.1111/psj.12055

Bryk

A. S.

Thum

Y. M.

(1989). The effects of high school organization on dropping out: An exploratory investigation. American Educational Research Journal, 26(3), 353–383. https://doi.org/10.3102/00028312026003353

Cannata

M. A.

Penaloza

(2012). Who are charter school teachers? Comparing teacher characteristics, job choices, and job preferences. Education Policy Analysis Archives, 20(29), 1–25. https://doi.org/10.14507/epaa.v20n29.2012

10.

Carter

P. L.

(2005). Keepin’ it real: School success beyond black and white. Oxford University Press.

11.

Carver-Thomas

Darling-Hammond

(2017). Teacher turnover: Why it matters and what we can do about it. Learning Policy Institute. https://learningpolicyinstitute.org/sites/default/files/product-files/Teacher_Turnover_REPORT.pdf

12.

Cherng

H. Y. S.

(2017). The ties that bind: Teacher relationships, academic expectations, and racial/ethnic and generational inequality. American Journal of Education, 124(1), 67–100. https://doi.org/10.1086/693955

13.

Christenson

(1996). What academic programs are offered most frequently in schools serving American Indian and Alaska Native students? (Issue Brief No. IB–3–96). Office of Educational Research and Improvement, U.S. Department of Education. https://nces.ed.gov/pubs/96841.pdf

14.

Cipollone

Stich

A. E.

(2017). Shadow capital: The democratization of college preparatory education. Sociology of Education, 90(4), 333–354. https://doi.org/10.1177/0038040717739071

15.

Cisneros

Gomez

L. M.

Corley

K. M.

Powers

J. M.

(2014). The Advanced Placement opportunity gap in Arizona: Access, participation, and success. AASA Journal of Scholarship and Practice, 11(2), 20–28.

16.

Clotfelter

(2004). After Brown: The rise and retreat of school desegregation. Princeton University Press.

17.

Coleman

J. S.

(1990). Equality and achievement in education. Westview Press.

18.

College Board. (n.d.). The AP experience: Get the most out of your classes. https://apstudent.collegeboard.org/exploreap/the-ap-experience

19.

Davies

Rizk

(2018). The three generations of cultural capital research: A narrative review. Review of Educational Research, 88(3), 331–365. https://doi.org/10.3102/0034654317748423

20.

Dietrichson

Bøg

Filges

Klint Jørgensen

A. M.

(2017). Academic interventions for elementary and middle school students with low socioeconomic status: A systematic review and meta-analysis. Review of Educational Research, 87(2), 243–282. https://doi.org/10.3102/0034654316687036

21.

Domina

Hanselman

Hwang

McEachin

(2016). Detracking and tracking up: Mathematics course placements in California middle schools, 2003–2013. American Educational Research Journal, 53(4), 1229–1266. https://doi.org/10.3102/0002831216650405

22.

Domina

McEachin

Hanselman

Agarwal

Hwang

Lewis

(2016). Beyond tracking and detracking: The dimensions of organizational differentiation in schools [Working Paper]. RAND. https://doi.org/10.7249/WR1155

23.

Domina

McEachin

Penner

(2015). Aiming high and falling short: California’s eighth-grade algebra-for-all effort. Educational Evaluation and Policy Analysis, 37(3), 275–295. https://doi.org/10.3102/0162373714543685

24.

Donaldson

(2017). The implementation of the International Baccalaureate Diploma Program: Equity, access, and effectively maintained inequality [Unpublished doctoral dissertation]. University of Notre Dame.

25.

DeVoe

J. F.

Darling-Churchill

K. E.

(2008). Status and trends in the education of American Indians and Alaska Natives: 2008 (NCES 2008-084). National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education.

26.

Eaton

S. E.

(2001). The other Boston busing story: What’s won and lost across the boundary line. Yale University Press.

27.

Ellison

Aloe

A. M.

(2018). Strategic thinkers and positioned choices: Parental decision making in urban school choice. Educational Policy, 33(7), 1135–1170. https://doi.org/10.1177/0895904818755470

28.

Evans

B. J.

(2019). How college students use Advanced Placement credit. American Educational Research Journal, 56(3), 925–954. https://doi.org/10.3102/0002831218807428

29.

Fisher

C. S.

Hout

Sanchez Jankowski

Lucas

S. R.

Swidler

Voss

(1996). Inequality by design: Cracking the bell curve myth. Princeton University Press.

30.

Frankenberg

Orfield

(2012). The resegregation of suburban schools. Harvard Educational Press.

31.

Gagnon

D. J.

Mattingly

M. J.

(2016). Advanced Placement and rural schools. Journal of Advanced Academics, 27(4), 266–284. https://doi.org/10.1177/1932202X16656390

32.

Gamoran

Berends

(1987). The effects of stratification in secondary schools: Synthesis of survey and ethnographic research. Review of Educational Research, 57(4), 415–435. https://doi.org/10.3102/00346543057004415

33.

Gamoran

Mare

(1989). Secondary school tracking and educational inequality: Compensation, reinforcement, or neutrality? American Journal of Sociology, 94(5), 1146–1183. https://doi.org/10.1086/229114

34.

Geiser

Santelices

(2004). The role of Advanced Placement and honors courses in college admissions. Center for Studies in Higher Education. https://cshe.berkeley.edu/sites/default/files/publications/rop.geiser.4.04.pdf

35.

Graham

P. A.

(2005). Schooling America: How the public schools meet the nation’s changing needs. Oxford University Press.

36.

Hallett

R. E.

Venegas

K. M.

(2011). Is increased access enough? Advanced placement courses, quality, and success in low-income urban schools. Journal for the Education of the Gifted, 34(3), 468–487. https://doi.org/10.1177/016235321103400305

37.

Hallinan

M. T.

(1996). Track mobility in secondary school. Social Forces, 74(3), 983–1002. https://doi.org/10.2307/2580389

38.

Iatarola

Conger

Long

M. L.

(2017). Determinants of high schools’ advanced course offerings. Educational Evaluation and Policy Analysis, 33(3), 340–359. https://doi.org/10.3102/0162373711398124

39.

International Baccalaureate Organization. (n.d.). Information for students and parents. https://www.ibo.org/information-for-parents/

40.

Judson

Hobson

(2015). Growth and achievement trends of Advanced Placement (AP) exams in American high schools. American Secondary Education, 43(2), 59–76.

41.

Kelly

(2004). Do increased levels of parental involvement account for social class differences in track placement? Social Science Research, 33(4), 626–659. https://doi.org/10.1016/j.ssresearch.2003.11.002

42.

Kelly

(2009). The black-white gap in mathematics course taking. Sociology of Education, 82(1), 47–69. https://doi.org/10.1177/003804070908200103

43.

Kelly

Price

(2011). The correlates of tracking policy: Opportunity hoarding, status competition, or a technical-functional explanation? American Educational Research Journal, 48(3), 560–585. https://doi.org/10.3102/0002831210395927

44.

Klopfenstein

(2004). The Advanced Placement expansion of the 1990s: How did traditionally underserved students fare? Education Policy Analysis Archives, 12(68), 1–15. https://doi.org/10.14507/epaa.v12n68.2004

45.

Kolluri

(2018). Advanced Placement: The dual challenge of equal access and effectiveness. Review of Educational Research, 88(5), 677–711. https://doi.org/10.3102/0034654318787268

46.

Klugman

(2013). The Advanced Placement arms race and the reproduction of educational inequality. Teachers College Record, 115(5), 1–34.

47.

Lareau

(2000). Home advantage: Social class and parental intervention in elementary education. Rowman & Littlefield.

48.

Lareau

Calarco

J. M.

(Eds.). (2012). Class, cultural capital, and institutions: The case of families and schools. Russell Sage.

49.

Lareau

Horvat

E. M.

(1999). Moments of social inclusion and exclusion race, class, and cultural capital in family-school relationships. Sociology of Education, 72(1), 37–53. https://doi.org/10.2307/2673185

50.

Lewis

A. E.

Diamond

J. B.

(2015). Despite the best intentions: How racial inequality thrives in good schools. Oxford University Press.

51.

Lucas

S. R.

(1999). Tracking inequality: Stratification and mobility in American high schools. Teachers College Press.

52.

Lucas

S. R.

Berends

(2002). Sociodemographic diversity, correlated achievement, and de facto tracking. Sociology of Education, 75(4), 328–348. https://doi.org/10.2307/3090282

53.

Moore

G. W.

Slate

J. R.

(2008). Who’s in the Advanced Placement courses and how are they doing? A two-year statewide study. High School Journal, 92(1), 55–67. https://doi.org/10.1353/hsj.0.0013

54.

Moore

G. W.

Slate

J. R.

(2010). Advanced Placement exams and American Indian performance. American Secondary Education, 38(2), 73–94.

55.

Mickelson

R. A.

(2003). When are racial disparities in education the result of racial discrimination? A social science perspective. Teachers College Record, 105(6), 1052–1086.

56.

National Center for Education Statistics. (2018). Digest of education statistics, 2016 (NCES 2017-094). U.S. Department of Education. https://nces.ed.gov/pubs2017/2017094.pdf

57.

Ndura

Robinson

Ochs

(2003). Minority students in high school advanced placement courses: Opportunity and equity denied. American Secondary Education, 32(1), 21–38.

58.

Oakes

(2005). Keeping track: How schools structure inequality. Yale University Press.

59.

Oakes

Gamoran

Page

R. N.

(1992). Curriculum differentiation: Opportunities, outcomes, and meanings. In Jackson

(Ed.), Handbook of research on curriculum (pp. 570–608). Macmillan.

60.

Oakes

Guiton

(1995). Matchmaking: The dynamics of high school tracking decisions. American Educational Research Journal, 32(1), 3–33. https://doi.org/10.3102/00028312032001003

61.

Orfield

Eaton

S. E.

(1996). Dismantling desegregation. The quiet reversal of Brown v. Board of Education. New Press.

62.

Orfield

Lee

(2005). Why segregation matters: Poverty and educational inequality. Civil Rights Project at Harvard University. https://civilrightsproject.ucla.edu/research/k-12-education/integration-and-diversity/why-segregation-matters-poverty-and-educational-inequality/orfield-why-segregation-matters-2005.pdf

63.

Orfield

Lee

(2006). Racial transformation and the changing nature of segregation. Civil Rights Project at Harvard University. https://civilrightsproject.ucla.edu/research/k-12-education/integration-and-diversity/racial-transformation-and-the-changing-nature-of-segregation/orfield-racial-transformation-2006.pdf

64.

Palmer

(2016, December 22). Many parts of state are high-level math and science “deserts.” Oklahoma Watch. https://oklahomawatch.org/2016/12/22/parts-of-state-are-high-level-math-and-science-deserts/

65.

Penner

A. M.

Domina

Penner

E. K.

Conley

A. M.

(2015). Curricular policy as a collective effects problem: A distributional approach. Social Science Research, 52, 627–641. https://doi.org/10.1016/j.ssresearch.2015.03.008

66.

Posey-Maddox

(2012). Professionalizing the PTO: Race, class, and shifting norms of parental engagement in a city public school. American Journal of Education, 119(2), 235–260. https://doi.org/10.1086/668754

67.

Price

(2019). Large scale datasets and social justice: Measuring disparity in opportunities to learn. In Strunk

K. K.

Locke

L. A.

(Eds.), Research methods for social justice and equity in education (pp. 203–216). Palgrave-McMillian. https://doi.org/10.1007/978-3-030-05900-2_17

68.

Quillian

(2014). Does segregation create winners and losers? Residential segregation and inequality in educational attainment. Social Problems, 61(3), 402–426. https://doi.org/10.1525/sp.2014.12193

69.

Reardon

S. F.

Firebaugh

(2002). Measures of multigroup segregation. Sociological Methodology, 32(1), 33–67. https://doi.org/10.1111/1467-9531.00110

70.

Remillard

J. T.

Baker

J. Y.

Steele

M. D.

Hoe

N. D.

Traynor

(2017). Universal algebra I policy, access, and inequality: Findings from a national survey. Education Policy Analysis Archives, 25(101), 1–29. https://doi.org/10.14507/epaa.25.2970

71.

Renzulli

L. A.

Roscigno

V. J.

(2005). Charter school policy, implementation, and diffusion across the United States. Sociology of Education, 78(4), 344–366. https://doi.org/10.1177/003804070507800404

72.

Riel

Parcel

T. L.

Mickelson

R. A.

Smith

S. S.

(2018). Do magnet and charter schools exacerbate or ameliorate inequality? Sociology Compass, 12(9), Article e12617. https://doi.org/10.1111/soc4.12617

73.

Roscigno

V. J.

Tomaskovic-Devey

Crowley

(2006). Education and the inequalities of place. Social Forces, 84(4), 2121–2145. https://doi.org/10.1353/sof.2006.0108

74.

Rosenbaum

J. E.

(1976). Making inequality: The hidden curriculum of high school tracking. Wiley.

75.

Salinas

Jr. Lozano

(2019). Mapping and recontextualizing the evolution of the term Latinx: An environmental scanning in higher education. Journal of Latinos and Education, 18(4), 302–315. https://doi.org/10.1080/15348431.2017.1390464

76.

Salloum

S. J.

Goddard

R. D.

Larsen

(2017). Social capital in schools: A conceptual and empirical analysis of the equity of its distribution and relation to academic achievement. Teachers College Record, 119(7), 1–29.

77.

Sattin-Bajaj

Roda

(2018). Opportunity hoarding in school choice contexts: The role of policy design in promoting middle-class parents’ exclusionary behaviors. Educational Policy. Advance online publication. https://doi.org/10.1177/0895904818802106

78.

Schneider

Stevenson

(1999). The ambitious generation: America’s teenagers motivated but directionless. Yale University Press.

79.

Snyder

T. D.

de Brey

Dillow

S. A.

(2016). Digest of education statistics 2015 (NCES 2016-014). National Center for Education Statistics, Institute of Education Sciences, U.S. Department of Education. https://nces.ed.gov/pubs2016/2016014.pdf

80.

Solórzano

Ornelas

(2004). A critical race analysis of Latina/o and African American advanced placement enrollment in public high schools. High School Journal, 87, 15–26. https://doi.org/10.1353/hsj.2004.0003

81.

Sørensen

A. B.

(1970). Organizational differentiation of students and educational opportunity. Sociology of Education, 43(4), 355–376. https://doi.org/10.2307/2111838

82.

Taagepera

Ray

J. L.

(1977). A generalized index of concentration. Sociological Methods & Research, 5(3), 367–384. https://doi.org/10.1177/004912417700500306

83.

Tilly

(1998). Durable inequality. University of California Press. https://doi.org/10.1525/9780520924222

84.

Tomasello

Brand

(2016, April). State policies to support competency-based education for overage, under-credited students. Ask the CCRS Center. College and Career Readiness and Success Center at American Institutes for Research. https://ccrscenter.org/sites/default/files/AskCCRS_CompetencyBasedEducation_OAUC.PDF

85.

Tugend

(2017, September 7). Who benefits from the expansion of A.P. classes? The New York Times. https://www.nytimes.com/2017/09/07/magazine/who-benefits-from-the-expansion-of-ap-classes.html

86.

Tyack

D. B.

(1974). The one best system: A history of American urban education. Harvard University Press. https://doi.org/10.2307/j.ctv136c61j

87.

Tyson

(2011). Integration interrupted: Tracking, Black students, and acting White after Brown. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199736447.001.0001

88.

U.S. Department of Education Office for Civil Rights. (2014). Civil Rights Data Collection Data snapshot: Teacher equity (Issue Brief No. 4). https://www2.ed.gov/about/offices/list/ocr/docs/crdc-teacher-equity-snapshot.pdf

89.

Yosso

T. J.

(2005). Whose culture has capital? A critical race theory discussion of community cultural wealth. Race Ethnicity and Education, 8(1), 69–91. https://doi.org/10.1080/1361332052000341006

90.

Zerquera

D. D.

Haywood

DeMucha

(2020). More than nuance: Recognizing and serving the diversity of the Latinx community. In Teranishi

R. T.

Nguyen

B. M. D.

Alcantar

C. M.

Curammeng

E. R.

(Eds.), Measuring race: Why disaggregating data matters for addressing educational inequality (pp. 154–169). Teachers College Press.