What Students Do Early College High Schools Serve? Unpacking Social Constructions of the Target Population

To support the nation’s college completion goals, early college high school (ECHS) reform creates opportunities for interested students to earn up to two years of free college credit during high school. ECHSs also have an equity objective: to target and enroll students who are historically underrepresented and/or might not otherwise go to college. Yet the extent to which ECHSs actually serve their target population in practice is unclear, especially in a marketized school environment. Using qualitative methods and the theory of social construction and policy design (Schneider & Ingram, 1993), this study explores the recruitment and selection practices at five ECHSs in the borderlands of Texas. Findings suggest that ECHS staff invited applications from the broad target groups. However, the admission process, shaped in part by patterns of self-selection, favored students who were academically inclined and relatively privileged compared to their district 1 The research reported in this article was made possible (in part) by a grant from the Spencer Foundation (#201700061). The views expressed are those of the author and do not necessarily reflect the views of the Spencer Foundation. Education Policy Analysis Archives Vol. 28 No. 173 2 peers. ECHS staff socially constructed narrower ideals of the target population than those articulated in the policy design based on their assumptions about who was likely to succeed in—and thus deserving of—an early college opportunity. Findings are discussed with particular attention to the equity implications of ECHS reform.


What Students Do Early College High Schools Serve? Unpacking Social Constructions of the Target Population
The Early College High School (ECHS) initiative, created by the Bill and Melinda Gates Foundation in 2002, is designed to facilitate postsecondary access and completion by combining high school with the first two years of college (Berger et al., 2014). Through partnerships between K-12 districts and institutions of higher education-most often community colleges-ECHSs allow high school students to earn up to an associate degree by 12th grade at little or no cost to students and their families (Muñoz et al., 2014). Students complete a comprehensive curriculum of dual credit courses, which confer both a college and high school credit without a post-course test requirement. ECHSs provide a host of academic and social-emotional support systems, such as tutoring and school-wide community building, to help students succeed in college coursework (Ari et al., 2017;Edmunds et al., 2013). A growing body of research has found positive effects of ECHS participation on credit accumulation, college enrollment, and degree attainment (Berger et al., 2014;Edmunds et al., 2017;Haxton et al., 2016;Song & Zeiser, 2019).
In addition to providing an accelerated pathway to an associate degree, the ECHS Initiative has an explicit equity agenda (Barnett et al., 2013): ECHSs are "committed to serving students underrepresented in higher education," which may include low-income students, students of color, English language learners, first generation students, and/or students struggling academically (Berger et al., 2009, p. 4;Webb & Gerwin, 2014). Although the specific subgroups vary across schools, the target population mandate distinguishes ECHSs from other college preparatory programs such as Advanced Placement (AP) and traditional dual credit coursework, which have historically enrolled college-bound higher achievers and disproportionately excluded low-income students of color (Barnett et al., 2015;Kolluri, 2018). In theory, ECHSs admit high school students who may struggle to access and/or persist in higher education and give them a jumpstart with college coursework.
Yet little is known about how ECHS educators understand and operationalize their target population (Cravey, 2013). This line of inquiry is warranted for a couple reasons. First, while most ECHSs are public schools that are accountable to the local school district, they are also schools of choice, meaning students must apply to attend. In school choice environments, patterns of selfselection can stratify students across schools (Monarrez et al., 2019), as more privileged or motivated parents may be more likely to utilize choice options (Altonji et al., 2015). As a result, the very students that ECHSs are tasked with targeting may be less likely to apply. Second, many ECHS staff feel pressure to ensure students succeed in college coursework (Thompson & Ongaga, 2011), which may incentivize the admission of more academically inclined students. These conditions potentially undermine the reform's equity promise; if ECHS students are relatively advantaged and/or likely to pursue college anyway, ECHSs may inadvertently increase attainment disparities. Thus, it is important to investigate qualitatively the local processes that shape the demographics of ECHSs.
To that end, this study examines the recruitment and selection practices at five ECHSs in a border region of Texas. These schools are uniquely suited to a study of ECHS admission due to the makeup of their populations; on the one hand, consistent with the target population mandate, their students are predominantly low-income, Hispanic, and first-generation. On the other hand, these ECHSs serve fewer disadvantaged students, on average, relative to the traditional high schools in their districts. The analysis is theoretically informed by Schneider and Ingram's (1993) theory of social construction and policy design, which analyzes how target populations are defined to better understand policy processes and outcomes. Specifically, I ask: (a) How do ECHS educators recruit and select their students? (b) What do these practices reveal about their social constructions of the ECHS target population? To answer these questions, I employ qualitative methods, using principal and teacher perspectives to ascertain the locally shared meanings and practices that influence how ECHSs operationalize the target population mandate.
In preview, findings suggest that ECHS staff invited applications from the broad target groups. However, the admission process, shaped in part by patterns of self-selection, favored students who were academically inclined and relatively privileged compared to their district peers. ECHS staff socially constructed narrower ideals of the target population than those articulated in the policy design based on their assumptions about who was likely to succeed in-and thus deserving of-an early college opportunity. Findings are discussed with particular attention to the equity implications of ECHS reform.

Literature Review
In this review, I first situate the ECHS initiative in its wider policy context, highlighting the emergence of the college completion agenda and the proliferation of school choice. I then discuss extant research on ECHS admission practices and student demographics.

The College Completion Agenda
ECHSs have gained popularity at a time when increasing postsecondary access and completion has become a national policy priority (Koropeckyj et al., 2017). One motivation is economic: The modern workforce demands a larger share of college-educated workers (Carnevale et al., 2018). The good news is a larger share of high school graduates are matriculating into college relative to prior generations (Twenge & Donnelly, 2016). Yet even with more students entering higher education, degree completion rates remain stagnant (Rubin & Hearn, 2018). Roughly 40% of students who enter 4-year colleges do not finish (NCES, 2018). Part of the problem is academic under-preparation; up to 68% of first-year community college students do not meet basic proficiency in math and English and require remedial coursework (Chen & Simone, 2016).
An adjacent concern is equity; data on higher education outcomes show ongoing disparities by race/ethnicity, class, and parent education. Students who are of color, from lower income households, and/or first in their families to attend college are significantly more likely to be placed into remediation and less likely to complete a degree than their more privileged counterparts (Cahalan et al., 2019;de Brey et al., 2019). These historically underrepresented populations are also more likely to be concentrated in under-resourced, low-performing schools (Anyon, 2014) and less likely to be tracked into advanced college preparatory courses (Giersch, 2018). To meet job demand at home, maintain competitiveness globally, and diminish persistent inequities, educational policymakers have called for better preparing K-12 students in general and underrepresented students in particular to complete postsecondary degrees and certificates (Duncheon, 2015).
ECHS reform is theoretically positioned to address both college incompletion and inequity. On the one hand, enabling high school students to earn college credit during high school may enhance the likelihood that those students continue on to university and complete a bachelor's (Edmunds et al., 2020). On the other hand, targeting students who have historically been excluded from advanced high school coursework and higher education may minimize gaps in postsecondary outcomes (Berger et al., 2010). The assumption of the original policy designers, for example, was that ECHSs could "motivate struggling students by raising expectations, and providing support for them to do more challenging work" (Webb & Gerwin, 2014, p. 8;Walk, 2020), thereby eliminating the need for remedial coursework upon college entrance. While not all ECHSs today target "struggling students" specifically, all are committed to serving underrepresented populations.

School Choice
The ECHS initiative has also emerged within an increasingly marketized educational policy environment that casts parents and students as consumers of schooling (Lubienski, 2005). Within this context, school choice is promoted on the assumption that having schools compete for customers (i.e., students) will incentivize them to improve (Chubb & Moe, 2011). Parents are able to select the school that is best for their child, be it a traditional district-run public school, charter school, magnet or vocational program, or private school paid for by vouchers (Bast & Walberg, 2004). ECHSs are an alternative to the traditional high school model in that students can begin college coursework early and graduate with an associate degree. Advocates of school choice argue that specialized curricular opportunities, such as that offered by ECHSs, can benefit particular student populations (Bast & Walberg, 2004;Chubb & Moe, 2011).
Yet some researchers have cautioned that school choice may enhance segregation by race, class, ability, and language proficiency (Bifulco & Ladd, 2007;Böhlmark et al., 2016;Brown & Makris, 2018;Frankenberg et al., 2017). Charter schools, for example, serve on average smaller proportions of students with disabilities and emergent bilinguals than traditional district-run schools (Epple et al., 2016). Parents tend to choose schools with students who are racially/ethnically similar to their children (Garcia, 2008;Stein, 2015). Higher income and more educated parents, meanwhile, may be more likely to know about and utilize school choice options (Diamond & Gomez 2004;Holme, 2002). Beyond race and class, parents who are more involved in their children's formal education may be more likely to exercise choice (Teske & Schneider, 2001). As a result, choice schools potentially siphon students who are higher-performing, higher-income, and/or easier to educate out of the traditional public system, a phenomenon referred to as cream-skimming (Altonji et al., 2015;Jabbar, 2015;Lacireno-Paquet et al, 2002).
Extant literature offers insight into factors that shape school demographics in a marketized environment. Public schools like charters and ECHSs cannot hand pick their students, but they can shape the pool of applicants through marketing and location decisions (Lubienski et al., 2009). For instance, Wilson and Carlsen (2016) found that charter schools signal "fit" to parents through their websites, sending implicit messages about race, culture, and academic achievement. Location is also important given that low-income parents may not have the time or resources to transport their student to a school far from home (Teske et al., 2009). Many charter schools open in high-minority, high-poverty areas and serve large proportions of low-income students of color (Epple et al., 2016), but some avoid areas that are highest-need, thus attracting applicants who are relatively better achieving and less expensive to serve (LaFleur, 2016). Because ECHSs require an application to attend, some of these challenges associated with school choice may influence how the target population mandate is operationalized.

ECHS Admission and Demographics
The ECHS model's commitment to serve historically underrepresented students has been adapted into the policies that govern ECHSs at the state and district levels (Jobs for the Future, n.d.). Though specific admission practices and target groups vary across localities, extant research offers some insight into ECHS recruitment and selection. To recruit applicants, ECHSs typically distribute materials to middle schools (Muñoz et al., 2014) and word spreads to parents and students by way of counselors (Fischetti et al., 2011). Many (though not all) ECHSs use a lottery to select students (Edmunds et al., 2012;North, 2011;Song & Zeiser, 2019;TEA, 2020). Often students only go into the lottery if they are deemed eligible by a screening process that can vary from school to school (Edmunds et al., 2020). Some ECHSs set academic requirements for enrollment (Berger et al., 2010). Others, such as ECHSs in Texas, are prohibited from basing enrollment decisions on academic records (TEA, 2016(TEA, , 2020. Leonard (2013aLeonard ( , 2013b) studied a suburban ECHS in Massachusetts that recruited students from the two middle academic quartiles. More than threefourths of ECHSs rely on some combination of essays and interviews to admit students (Berger et al., 2010). One case study found that ECHS staff used interviews and essays to identify "diamonds in the rough" and "good kids with potential" (Fischetti et al., 2011).
Research also reveals variability in the characteristics of the ECHS student population across ECHSs and compared to surrounding district high schools. According to one national evaluation study, ECHS student populations included about 60% low-income students and 66% underrepresented students of color (AIR & SRI International, 2009). Some studies have found that ECHS populations include higher proportions of students of color than comparison districts (Berger et al., 2010;Kaniuka & Vickers, 2010). Other research has reported that, on average, ECHSs "serve students who are similar to the student populations in their districts in terms of eligibility for freeand reduced-price lunch and race/ethnicity" (Edmunds et al., 2017, p. 304). Parent education levels of students also vary across ECHSs (Berger et al., 2014;Cravey, 2013;Edmunds et al., 2012;Howley et al., 2013;Muñoz et al., 2014). With respect to academics, some studies have found that entering ECHS students demonstrate significantly higher achievement, on average, compared to their non-ECHS peers (Berger et al., 2014;Edmunds et al., 2017). A case study of one ECHS identified three academic types: (a) students who would have excelled in a traditional school and earned AP credits; b) students who would have completed high school without college credit; c) students who likely would not have graduated high school (Fischetti et al., 2011).
Most extant studies have relied on quantitative metrics such as administrative data, evaluation reports, and surveys to snapshot the characteristics of students served and the methods used for admission. What remains underexplored are the nuances of who is targeted, how, and why-questions with implications for the reform's equity potential. The present study pursues these lines in inquiry, focusing on admission practices and social constructions of the target population.

Social Construction of Target Populations
Policies are designed to achieve an objective by changing behavior, either by incentivizing desired actions or punishing undesirable ones (Schneider & Ingram, 1993). To do so, policies define target populations, or "those groups actually chosen to receive benefits and burdens through the various elements of policy design" (Ingram et al., 2007, p. 95). ECHS reform, for example, aims to increase postsecondary completion by encouraging underrepresented students specifically to begin college coursework early. Yet the identification of target populations is not straightforward or neutral. Rather, target populations are socially constructed by policy designers and implementers, with consequences for policy outcomes. This section describes the theory of social construction and policy design (Schneider & Ingram, 1993) and its suitability to studying ECHS admission.
In a policy context, social constructions are "cultural characterizations or popular images of the persons or groups whose behavior and well-being are affected by public policy" (Schneider & Ingram, 1993, p. 334). Socially constructing a target population involves first identifying a group as socially significant based on common characteristics, and second, attaching value to those characteristics. Because social constructions are neither given nor objective, this process of attaching meaning to a specific group and/or phenomenon is political, value-laden, and often contested (Schneider & Ingram, 1997). How a policy constructs the target population-whether positively or negatively-casts the intended recipients as either deserving or undeserving of benefits. Schneider and Ingram (1997) categorized potential policy targets into four quadrants. Two are powerful, but one group, the advantaged, is depicted positively, while the other group, contenders, is seen as greedy. The remaining two groups lack political power. Of those without political power, dependents are constructed positively as needing and deserving of support, while deviants are constructed negatively (e.g., violent, lazy) and thus portrayed as undeserving. Some target populations are ill-defined and/or members of a social group may fit into multiple categories (Schneider & Ingram, 1993). For example, underrepresented students lack political power, but may be constructed positively or negatively. As dependents, they may be seen as aspiring, motivated, hardworking, and thus deserving of college-going support. As deviants, they may be portrayed as under-achieving, disengaged, disruptive, or another deficit-laden descriptor.
Social constructions matter for implementation because they influence who benefits or loses from a policy intervention (Bertrand et al., 2018). Schneider and Ingram (1997) posited that when policies target dependent populations, such as underrepresented students, implementers may feel pressure "to ensure 'deserving' have been separated from 'undeserving' and that ineligible persons (e.g., 'non-deserving ones') are not receiving benefits" (p. 138). Local actors therefore engage in their own process of social construction, interpreting who should benefit from a policy as they implement it (Spillane et al., 2002). The resulting social constructions of the target groups may diverge somewhat from the categories outlined by policy designers. In this case, ECHS staff may feel obligated to identify students whom they deem most "deserving" of the opportunity to earn free college credit, who may or may not reflect the target groups outlined in ECHS policy documents.
By exposing the targets of reform, social construction and policy design theory sheds light on the extent to which a policy "depart[s] from the typical reproduction of power and social constructions to introduce change" (Ingram et al., 2007, p. 93). In other words, the framework considers whether a reform disrupts existing hierarchies (Ingram et al., 2007;Schneider & Ingram, 1993). As designed, ECHSs level the college access playing field by offering a head start to students who have historically faced barriers on the path into and through higher education. Yet whether ECHSs are minimizing college attainment disparities depends on which students they serve. Thus, it is important to examine social constructions of the ECHS target populations by the ECHS staff who ultimately decide who is and is not admitted. This objective requires studying how admission practices unfold at the ground level.

Research Design
This study was set in Texas, which has been a leader in expanding the ECHS model, with roughly 200 ECHSs in operation during the 2019-2020 school year (TEA, n. d.). Data were derived from a larger project on the ECHS initiative in a Texas border region that included over 10 ECHSs across 8 school districts. In this region, where the median household income is roughly $45,000, Hispanics of any race account for about 80% of the population. Data collection for this study focused on five ECHSs (described below) and spanned three years from the fall of 2015 to the fall of 2018. Qualitative methods elicited the perspectives of principals and teachers to gain insight into admission practices (processes) and constructions of the target population (shared meanings).

Texas ECHS Policy Context
The Texas Education Agency (TEA) website defines ECHSs as "open-enrollment high schools that allow students least likely to attend college [emphasis added] an opportunity to receive both a high school diploma and either an associate degree or at least 60 credit hours toward a baccalaureate degree" (TEA, n. d). ECHS policy is laid out in the ECHS Blueprint 2 , which further defines the target population: ECHSs "shall target and enroll students who are at risk of dropping out of school as defined by [the state] and who might not otherwise go to college" (TEA, 2016, p. 2) 3 . The policy lists subgroups of students the state deems "at risk" academically who should not be "exclude [d] or discourage[d]" from enrollment, such as "students who are of limited English proficiency or who have failed a state administered assessment" (TEA, 2016, p. 2). The Blueprint further specifies that "in addition to those who are at risk," ECHSs should enroll subpopulations that are "historically underrepresented in college courses," including "first generation college goers, students of low socioeconomic status, African American, Hispanic" (TEA, 2016, p. 2). At the time of this study, the Blueprint prohibited admitting students based solely on academic record or discipline history.

Research Sites
The study drew data from five ECHSs that partner with the region's one community college system. These schools, each situated in a different K-12 district, represent a diversity of ECHS models, locations, and ages. Three ECHSs are located on community college campuses and two operate on stand-alone campuses. Two are urban and three are rural. These ECHSs opened between 2006 and 2012, and have student populations ranging from about 200 to 450.
While the majority of their students belong to one or more underrepresented groups, the five ECHSs serve, on average, smaller proportions of each target population than their traditional high school counterparts (See Table 1). For example, all the ECHSs serve predominantly Hispanic students (77% to 97%) and a majority of economically disadvantaged students (53% to 82%). However, the percentages of Hispanic and low-income students at three of the ECHSs are up to 10 and 16 points lower, respectively, than the average percentages served in regular high schools. Also compared to their traditional counterparts, all five ECHSs serve much smaller proportions of English language learners and "at risk" students, and larger proportions of gifted and talented students (up to 33%). These trends indicate there may have been cream-skimming happening at the local level (Jabbar, 2015) and motivate inquiry into admission practices.

Sample
Principals were initially contacted by email for permission to include their schools in the study. All teachers were invited to participate via presentations at faculty meetings and/or emails. Ultimately, 80 ECHS staff members participated: the principal from each school (5 total) and between 10 and 22 teachers from each school (75 total), which represented between 50% and 100% of the faculty. Consistent with the region's demographics, the majority of staff participants were 2 The Blueprint was recently updated in 2020 with small revisions to the target populations (e.g., students with disabilities have been added as a target subgroup). I quote from the 2016 version, which was in effect during data collection for this study. 3 Scholars have problematized the term "at risk" for portraying marginalized student populations as deficient and propagating stigma about youth who are underserved by our institutions (see, e.g., Rios, 2011;Swadener & Lubeck, 1995;Valencia & Solórzano, 1997). Despite these important critiques, Texas continues to use this term to define student subgroups in educational policy documents, and many Texas educators, in turn, use it as well. Thus, I use "at risk" when quoting the Blueprint or the participants, or when referring to student subgroups as defined by the state.
Hispanic. Three principals were Hispanic and two were white. Of the 75 teachers, 45 (~60%) were Hispanic, 29 (~39%) were white, and 1 (~1%) was Black. At the time of data collection, their years of experience spanned from 1 to over 20, and years at an ECHS specifically spanned from 1 to 10.

Data Collection
Data were collected via interviews, observations, and documents. I conducted semistructured one-on-one interviews with the principals, which lasted an average of 75 minutes. Teachers were interviewed either one-on-one or in a focus group setting, depending on principal preference and teachers' scheduling availability. At three schools, one-on-one interviews were scheduled with 37 teachers. These interviews lasted between 24 minutes and 3 hours. At the other two schools, the remaining 38 teachers were interviewed in a focus group setting scheduled during departmental meeting times. Teachers were given the opportunity to skip the meeting if they did not want to join the focus group; only one declined to participate. The eight focus groups-four at each school: math, English, social studies, and science-contained groups of 3-6 teachers, depending on the size of the department, and lasted from 60 to 90 minutes. The same interview protocol was used whether it was a one-on-one or a group interview. Questions focused on how teachers described their students, how their school recruited applicants, what the admission process entailed, how students were selected, and more generally, what their experiences were like teaching in an ECHS setting. Teacher interviews were conducted by me or one of two graduate student members of the research team, who shadowed two interviews with me before doing their own. All interviews were audio recorded for later transcription.
In addition to interviews, I conducted upwards of 50 hours of observations (10 per school) of events and meetings related to ECHS recruitment and selection. I attended recruitment nights at the district, presentations to middle schools about the program, and faculty meetings where students or admissions were being discussed. Finally, documents were collected pertaining to ECHS admission and the target population. Examples included the policy documents describing the ECHS mission and state-level demographic reports. Marketing, recruitment, and application materials relevant to the student selection process were also included in the dataset.

Data Analysis
Data analysis occurred concurrently with data collection. Early on in the data collection process, we noticed similar themes coming up across interviews and ECHSs. For instance, when describing their student population, teachers repeatedly rejected the idea that they served higher performing students than their traditional high school counterparts. They also emphasized the importance of motivation for ECHS student success, which became relevant to student selection. To unpack these notable patterns as they emerged in the data, and to distinguish any trends that were not immediately apparent but nonetheless important, our first round of coding was inductive, aided by the constant comparative method (Glaser & Strauss, 1967). To ensure interrater reliability, the research team met frequently to discuss and revise the codebook, clarify how codes should be applied, and compare coded transcripts. All coding was completed using N-Vivo software.
Subsequently, I completed a round of deductive coding, guided by prior literature and theory. For example, the theory of social construction and policy design helped me to identify instances where participants portrayed ECHS applicants as "deserving" or "deviant," and to track the characteristics participants ascribed to those constructions. This process was helpful to develop theoretical interpretations in line with the themes identified during the inductive analysis (Miles & Huberman, 1994).
Trustworthiness was supported by member checking and prolonged engagement. Member checking was conducted with a handful of key informants across schools, including three principals and five teachers, each of whom had longevity at their campuses. These participants served as sounding boards throughout the research process as I reflected on what we were seeing in the field and what it meant. That this project was ongoing over multiple school years also enabled us to confirm that themes we observed early in data collection remained salient.

Findings
Recruitment and selection practices revealed that the ECHS target population was constructed in two main ways. The first was a broad construction of the target population, based on the general socio-demographic and academic categories outlined in TEA's Blueprint. As one principal explained, "Part of the TEA mandate is to get your low-income, at-risk kids to attend so that they will go to college." Another characterized the program as intended for "kids who traditionally might not have an opportunity to go to college." Teachers echoed this mission, suggesting that "the very idea of an early college high school" was to "target first generation" and "assist economically disadvantaged students." Others focused on academic characteristics of the target population. Per one principal, "Part of the Blueprint is that they're not looking for highachievers. They're looking for kids who are in that middle ground." Participants relayed that ECHS was not "meant for gifted kids" or "kids who are going to go to college anyway." These statements capture participants' perceptions of the target population as constructed in the policy design: students traditionally excluded from higher education socio-demographically and academically.
The second social construction of the target population was narrower, based on participants' beliefs about which students were likely to perform well in an ECHS setting. ECHS staff felt that students needed motivation and maturity to complete up to two years of college coursework on top of high school requirements. As one principal explained, "The student has to be ready, they have to be willing, they have to be committed to this whole idea of an early college." Teachers suggested that the school's unique expectations, such as attending class independently at the college, required a certain type of student: "There has to be a certain level of initiative and drive to succeed over at the community college because I'm not going to stand over their shoulders when they go to class." Thus, while participants acknowledged the broad target populations constructed in the TEA Blueprint, they created a narrower construction defined by pro-academic behaviors and mindsets.
While both of these social constructions surfaced during the ECHS admission process, each received emphasis at different points. The broader construction of the target population was referenced most often in relation to recruitment and marketing. The narrower construction of the target population came to the fore when applicants were actually admitted. Embedded in a school choice policy context, ECHS admission practices intersected with self-selection patterns to privilege higher achieving students for enrollment. Below, I present these emergent themes across two sections: recruitment and selection. Throughout, I draw connections to the two social constructions.

Recruitment
Recruitment was designed to introduce a broad swath of 8th graders from the Blueprint's target populations to the ECHS opportunity. At the same time, ECHS staff signaled to potential applicants that the school would be demanding academically, encouraging self-selection among students in the narrower construction of the target population.

Casting a Wide Net: "I'm Talking to Everybody in this Room"
ECHS recruitment practices were designed to cast a wide net-that is, encourage a broad pool of applicants-with the goal of reaching the target populations outlined in the Blueprint. However, with the exception of two ECHSs whose marketing materials specified the intent to serve "low-income, minority, and/or first-generation students," rarely were target groups mentioned overtly. Rather, ECHS staff made strategic choices about where and how to advertise the program to alert students in the broad target populations that the ECHS opportunity was available to them.
To recruit applicants, ECHS ambassadors-typically the principal and sometimes also the counselor, teachers, and select students-held presentations or information sessions for prospective 9th graders and/or their parents. Often these events took place at high school fairs hosted by the district. Other times, ECHS ambassadors put on assemblies at local middle schools, prioritizing high-needs schools. One teacher described this strategy: "There is recruiting going on over at [the high poverty middle school] because a lot of their kids fit the demographic that we want. But we are not going to go up to [the low poverty middle school]." Decisions about where to market the ECHS were therefore geared toward students in the Blueprint's target populations.
At these recruitment events, ECHS recruiters used coded language to appeal to low-income and first-generation students. For example, all the principals discussed the financial benefits of attending an ECHS. In the words of one principal, "I tell [eighth graders], 'Between books and tuition, this is how much money you can save.'" They also emphasized time to degree. At one high school fair, a principal spoke to a large room full of eighth graders and their parents: ""If you will be the first person in your family going to college, this program can provide you the support you need to get ahead." Though recruiters did not explicitly suggest low-income or first-generation students apply, highlighting cost and time benefits was one way to pique the interest of those target groups.
In a similar vein, ECHS recruiters did not overtly ask "middle ground" students to apply, but instead stressed that past academic under-performance was not a barrier to admission. One principal described her sales pitch thusly: "I say, 'There are lots of reasons why you may not have been a good student in middle school, but that doesn't mean that you can't come to our school." Other principals spoke in more general terms, reiterating that all students, regardless of their academic levels, were invited to apply: "I'm talking to everybody in this room. If you want to come here, you can come here." Part of this outreach effort was dispelling the myth-which participants perceived was widespread in the community-that ECHSs were elite and selective. One principal used a truefalse quiz to educate teachers at feeder middle schools. "They said, 'Well you have to be a perfect student to be accepted,' and 'You should be gifted and talented.' And all of those things are false." In this way, ECHS ambassadors appealed to students in the target population, broadly constructed, by trying to minimize perceived barriers to admission.

Encouraging Self-Selection: "It is a Lot of Work"
While inviting applications from a wide range of students in the Blueprint's target groups, ECHS recruitment and marketing practices also encouraged particular kinds of students to select into the program. For instance, some participants noted that simply becoming aware of the ECHS opportunity was easier for certain families than others. As one teacher surmised, "economically disadvantaged parents might not have the time or transportation resources to [attend] our general information meetings, so […] the opportunity is presented to a limited group of students." In her view, families who could attend middle school presentations or high school fairs-often held in the evening-might be relatively advantaged compared to those who could not.
Recruiters also wanted to raise awareness about what an ECHS experience would entail, subtly dissuading applicants who might fall behind. For example, while they emphasized that prior academic performance was not a barrier to admission, ECHS staff were transparent about the school's demands: "It is a lot of work." In addition to stressing the course load, ECHS ambassadors made clear that ECHSs did not have athletics programs. During one recruitment presentation, for example, a principal said, "We have lots of clubs, but if you really want to play football-if that's your dream for your high school years-we might not be the school for you." These marketing strategies were designed to signal to potential applicants whether the ECHS environment would be a good fit. In so doing, ECHSs tailored their target population, even while they presented the opportunity broadly.

Selection
All five ECHSs had a variety of admission requirements that were used to select students (See Table 2). All utilized an application essay and an interview as part of the admission process. They also required students to submit academic, discipline, and attendance records for reporting purposes only, though one required applicants to have a "good discipline record." Two used a lottery to make the final selections. The implementation of these practices varied somewhat from school to school. For example, one ECHS had students rotate through all the teachers to answer interview questions, similar to speed dating, while other ECHSs had students interviewed by pairs of teachers or just by the principal. Common across the schools, however, was a focus on identifying students who fit the narrow construction of the target population.

Essay
One required about challenges student has faced and why he/she wants to attend.
Two of three required about motivation for attending, challenges faced in middle school, and/or ingredients for academic success.
One required about motivation for attending and preparation to succeed.
One required about students' desire to attend and preparation for success.
One required about students' preparation to handle college courses and the perceived benefits of attending early college Teacher recommendat ions One required Two required, one from a core subject teacher

Additional requirements
Scheduling availability over the summer for remediation testing and summer bridge (all early colleges).

Finding the Right Fit: "The 'I'm going to work hard' kind of mentality"
The admission process was geared toward assessing students' interest in and motivation to complete the ECHS program. Participants across schools explained that "we want to make sure that students are [not] coming […] because their parent wants them here […] or they're going to fail." To that end, the written application and in person interview were designed to assess students' desire to attend. As one principal reminded his teachers during a faculty meeting, "If the student is sitting there looking uncomfortable-and sometimes they'll just tell you their parent made them applywrite that down." Part of determining intent was assessing whether applicants were interested in high school athletics. One teacher shared, "When a student sits there and tells me that all they ever want to do is play soccer, it's a red flag that they should probably go to a school where they can play soccer." If students did not seem to understand and show enthusiasm about an ECHS experience, they were rejected and/or counseled to attend their regular district high school.
School staff also tried to decipher whether applicants had the motivation to tackle the program's academic demands. One proxy for assessing students' willingness to work hard was attendance record, which was a "large decider" in the selection process across schools: "If they're missing school," said one teacher, "that's where the difficulty comes in in terms of [students] passing [college] classes." The interview process was another opportunity to assess students' commitment. As one principal shared, "We ask about what I like to call grit, perseverance, and resiliency. Do they have the 'I'm going to work hard' kind of mentality or do they quit when the going gets tough?" Another teacher recalled interviewing a less-than-enthusiastic applicant: "One of them last year told me, 'I'm lazy. I don't like to do homework.' So right away you know which students won't make it." ECHS staff prioritized applicants who demonstrated work ethic and academic commitment, qualities they deemed necessary for completing the program.
Participants' preference for these qualities was further evident in the way they discussed applicants' academic records, which were reviewed but, as mandated by the Blueprint, did not preclude acceptance. One principal stated, "I don't take a certain grade point average or a certain [test] score. I see where the student is at and it's part of the discussion I have with the student." If the student understood the program's demands and still wanted to attend, the principal overlooked past under-achievement. Teachers across schools confirmed this strategy. One referenced "a few" current ECHS students who performed poorly in middle school: "There are a few students that the principal lets in that have potential, despite poor grades, and we've seen those kids really do well." In these examples, ECHS staff placed more value on a student's potential, or demonstrated interest in completing the program, than achievement in middle school (Fischetti et al., 2011). In other words, the selection process privileged the narrow construction of the target population; 8th graders in the academic middle (broad) were admitted only if they showed academic motivation (narrow).
The broad construction of the target population as underrepresented demographically also took a back seat at this juncture in admission. All ECHSs required applicants to specify whether their parents had a college degree, and two formally assigned points to those who were first generation. For the most part, however, participants simply took for granted that, in the words of one teacher, "we're always going to have the first-generation and the low income, just because of the area we live in." A principal similarly noted that "for us in a predominantly Hispanic community," underrepresented applicants "could be eight out of ten people." Due to their location on the border, ECHS staff shared the implicit assumption that most students applying to their program represented at least one of the Blueprint's target demographics, which absolved them from focusing on student background during selection decisions.

Benefitting from Self-Selection: "Not Good Students…Don't Even Apply"
ECHS staff also acknowledged that certain types of students were more likely to apply than others. As one teacher said, "students do self-select." Echoing this point, another teacher compared his ECHS students to "the same kinds of students I got at [the traditional high school] in [Advanced Placement] and dual credit, in the sense that it's a certain kind of student who takes those classes." In other words, ECHS applicants tended to be students who were already inclined to seek out academically challenging curricular experiences.
In part, the model simply did not appeal to everyone; not all rising 9th graders want to spend their high school years getting a head start on college coursework. The application process itself was also cumbersome. In the words of one teacher, "we make kids jump through hoops." He elaborated, You can't do a strict screening process. The laws governing early college prevent that. But you can make kids jump through hoops. We can't say, "You didn't pass [your 8th grade English test] so you can't come here," but we can say, "You've got to write an essay, and you've got to come in and do the couple of interviews, and you better show us that you really want to do this." By making applicants "jump through hoops," the ECHS admission process favored students who were predisposed to be academically engaged and motivated. As one principal confessed, "the [eighth graders] that are just not good students usually don't even apply because I mean that's... it takes work to apply." This perspective is consistent with the disparate enrollment patterns between the five ECHSs and their traditional district counterparts (Table 1). Still, participants routinely rejected the widespread assumption of non-ECHS educators in the region that ECHSs served the "cream of the crop." Though one teacher admitted, "I mean we are still taking top students," she added the caveat, "but it is not as much as everybody thinks."

Discussion
The data offer insight into how the ECHS admission process unfolded to influence which students were served, revealing two distinct social constructions of the target population. To recap, recruiting practices cast a wide net in service to the broad construction of the target population, students who were demographically and academically underrepresented. To admit students, however, ECHS staff looked for pro-academic behaviors and mindsets, aligned with a narrower social construction of the target population. Bolstered by self-selection effects, ECHS admission practices ultimately favored students who were, in certain ways, more advantaged than their district peers and perhaps more likely to pursue college-even without an early college intervention. In what follows, I use the theory of social construction and policy design (Schneider & Ingram, 1993) and prior literature to analyze the findings, and close with implications for theory, policy, and future research.

Distinguishing Deserving from Deviant
The ECHS Blueprint broadly socially constructs the target population as students from nondominant socio-demographic subpopulations and students who may struggle academically (TEA, 2016). These policy targets are constructed positively, as dependents who are deserving of an early college intervention (Schneider & Ingram, 1997). Echoing this set of broad social constructions, ECHS staff articulated their commitment to serve underrepresented students, especially Hispanic students, economically disadvantaged students, and first-generation students. They also invoked the ECHS mission to target students in the academic "middle ground," as opposed to high-achieving students "who are going to go to college anyway." These broad constructions were evident in recruitment efforts to target lower income middle schools and dispel the myth that ECHSs was an elite program for high-achieving students. Importantly, the broad target populations were rarely recruited explicitly, but instead reassured all students were welcome to apply.
Yet admission practices revealed that, in the implementation of the ECHS model, not all students in the broadly constructed target populations were deemed worthy of the ECHS opportunity. A narrower social construction emerged as ECHS staff weighed the target population mandate with the other objective of the ECHS model, to confer associate degrees. The goal, in their view, was to identify applicants most likely to succeed in college coursework during high school. To that end, ECHS staff had to differentiate students who deserved access-those with pro-academic behaviors and mindsets-from those who were undeserving (Schneider & Ingram, 1997). Thus, while marketing the program, ECHS ambassadors signaled that the school might be a better fit for academically inclined students by describing the rigorous curricular standards and lack of sports. Though prior achievement was not disqualifying, the admission process required "jump[ing] through hoops," such as writing application essays or sitting for interviews, so that staff could assess applicants' motivation to attend. In turn, particular subgroups of students were socially constructed to be deviant, including students whose parents made them apply, students who wanted to play sports, "lazy" students, students who "don't do homework," and students with poor attendance or discipline records. Students perceived to belong to one or more of these categories were presumed unlikely to take advantage of the program and/or perform well in college classes, and thus less deserving of ECHS admission than their more motivated peers.
On the one hand, the pro-academic mindsets and behaviors that ECHS staff used to narrowly define their target population-eagerness to attend, motivation, maturity, good attendance-were well suited to the model's mission to confer associate degrees. Students who did not want to attend an ECHS were unlikely to do well, especially if they wanted to play sports during high school. Academic perseverance and maturity made success in college coursework more likely (Fischetti et al., 2011). Meanwhile, as participants were quick to point out, their location in a predominantly Hispanic, lower-income border community meant that the majority of their students were already coming from historically underrepresented populations.
On the other hand, the narrow construction of the target population favored students who were likely to be higher-achieving and college-bound, the very students who already benefit from traditional college preparatory programs such as Advanced Placement (Kolluri, 2018) and who were not the original policy targets. Recall the rationale that drove the ECHS initiative: "Even reluctant or discouraged high school students […] can be motivated at a relatively early age to view themselves as successful participants in the college experience" when provided with adequate academic and social supports (Berger et al., 2010, p. 334). This theory of action disrupts long-standing structural inequities in which students who are white, middle-class, and presumed to be high-achieving are disproportionately tracked into college preparatory courses (Oakes, 2005). However, the narrow social construction of the ECHS target population that emerged during implementation embodied the same elitist assumptions about which students are deserving of rigorous college preparation that produce and replicate educational hierarchies in the first place (Oakes et al., 1997;Hatt, 2012). Despite the policy intent to disrupt existing power dynamics (Ingram et al., 2007), these ECHSs were largely not enrolling students "least likely to attend college" (TEA, n. d.).

The Role of School Choice and Self Selection
The narrowing of the target population was further exacerbated by self-selection patterns. Because the ECHS is a choice model in a marketized educational policy context, students by necessity self-select when they apply. One issue was which families learned of the opportunity in the first place. Though presentations at high school fairs were offered to a broad audience (Jabbar, 2016), as one teacher pointed out, not all parents were able to attend. The data also show how ECHS ambassadors shaped the applicant pool through marketing (Lubienski et al., 2009;Wilson & Carlsen, 2016). While they encouraged all interested students to apply, ECHS presenters stressed the program's academic emphasis and lack of sports, signaling that the school was best fit for hard workers and/or high achievers. Finally, the application process itself functioned to weed out students who were unable or unwilling to "jump through hoops." In these ways, the dynamics of self-selection-coupled with purposeful recruitment choices and application requirementsincreased the likelihood that the students applying to the program were more academically inclined and/or motivated than the average rising ninth grader.
Collectively, these findings help explain the patterns evident in the quantitative district data (Table 1). Even though a majority of ECHS students were Hispanic and low-income, the ECHSs served, on average, smaller proportions of these underrepresented subgroups and "at risk" students and larger proportions of gifted students than their traditional district counterparts. This trend is consistent with the finding that some charter schools serve high proportions of disadvantaged students, but not necessarily the most disadvantaged in a given district (LaFleur, 2016). To the extent that the ECHSs attracted and admitted more academically engaged students, they may have contributed to cream-skimming (Altonji et al., 2015).

Implications for Theory, Policy, and Future Research
Whether the ECHSs in this study were exacerbating inequality or fulfilling the equity portion of the mission depends, in part, on one's perspective. Admission was largely driven by ECHS staff members' desire to move students successfully through college coursework, which-coupled with self-selection-favored the enrollment of academically capable, college-bound students. On a microlevel, therefore, the ECHSs may have increased stratification between students in the region who were likely to enter higher education and those who were not. Yet from a macro lens, due to their unique borderland context, the ECHSs were still serving populations that were racially, ethnically, and/or socioeconomically underrepresented in higher education writ large. The goal here is not to evaluate whether these ECHS population outcomes were right or wrong; rather, the takeaway is that the target population of an ECHS should not be taken for granted, especially given its importance to advancing postsecondary equity.
A few theoretical implications are worth noting. First is the value of focusing on how policy implementers socially construct target populations. Actors in institutions must define the boundaries of the target population as they implement new policy (Schneider & Ingram, 1997). Because they ultimately differentiate the deserving from the undeserving, how street-level bureaucrats construct the target population matters as much if not more for policy outcomes than the goals of policy designers (Spillane et al., 2002). Second, the findings illustrate how other aspects of a policy design influence the target population that implementers socially construct. In this case, ECHS staff used the college completion objective as justification for narrowing the target population to favor academically inclined students. Third, and relatedly, just because a policy targets disadvantaged groups as dependents does not mean that, in practice, it is benefiting those who are most in need in a particular local context. In the realm of education, the principals and teachers who implement policy in schools may be predisposed-implicitly or explicitly-to construct narrow ideas about the kinds of students who deserve access to an elite academic experience, based on systemic biases related to race, ethnicity, class, ability, gender, etc. (Steele et al., 2004;Sternberg, 2007). These historically and institutionally ingrained narratives about the kinds of students who are academically successful can undermine equity goals (Lewis & Diamond, 2015).
This study also has implications for policy. First, the findings highlight tensions inherent in ECHS policy design. The model is intended to facilitate early completion of college coursework for high school students who are unlikely to access and persist in higher education, thereby enhancing both postsecondary completion and equity (Barnett et al., 2013;TEA, n.d.). ECHSs have often been lauded for increasing degree attainment rates Hoffman, 2003), with the implicit assumption that the target population is being served. This study cautions that such an assumption may be faulty. In practice, ECHS staff may perceive a tradeoff between increasing attainment outcomes and enrolling the target population. ECHS admission practices may dissuade more vulnerable eighth graders from applying and enrolling. Eighth graders who have been marginalized by the education system may be less able and/or willing to apply to a high school that requires extra work. An ECHS may-or may not-be located in a majority-minority or predominantly low-income district. In short, when a special program offers an elite academic experience in a marketized educational policy context, the student population will be shaped by a variety of factors beyond the social constructions outlined in the policy design. These complexities are important to consider to ensure ECHS reform advances equity as well as completion.
If serving eighth graders "least likely to attend college" (TEA, n. d.) is a priority, policy designers and implementers need to think creatively about how to reach marginalized students who may be less willing or able to apply. One strategy might be to have middle school teachers and counselors identify students most likely to benefit. Another option is to eliminate admissions requirements that can impede access. As was evident in this dataset, standardized metrics (e.g., achievement, discipline, attendance) are not the only barriers to access; "hoop jumping" is itself a form of gatekeeping (West et al., 2006). Of note, recent updates to the Texas ECHS Blueprint now require ECHSs to use an open access lottery or a weighted lottery-one that favors students in the target groups-for admission (TEA, 2020). Although this change suggests the state's commitment to ensure access for the target populations, a lottery-only approach might also limit the discretion of ECHS staff to identify students who genuinely do not want to attend (i.e., whose parents made them apply). A compromise might be preferable, whereby most admission barriers are eliminated but ECHSs still retain some autonomy to survey applicant interest.
Several avenues exist for future research. Given the recent revisions to the Texas ECHS Blueprint, new studies would be helpful to determine whether and how the target population has changed, and what those changes mean for equity. Research is also needed to unpack how ECHS admission works and who is served in different state and regional contexts that have different target subgroups and/or pools of applicants. Another topic for future inquiry is what happens at the middle school level. What do middle school staff know about ECHSs, and what messages do they convey to eighth graders? Interviews with middle school stakeholders would offer further insight into what types of students apply and why. Finally, studies that look at how ECHSs shape enrollment district-wide, and how, if at all, they influence the composition of student bodies in regular high schools would be useful to complement studies of ECHS students. More broadly, research that is attentive to which students are served in ECHSs and why can support ECHS reform in advancing its dual objectives of completion and equity. education policy analysis archives editorial board