On Becoming a District of Choice: Implications for Equity Along the United States–Mexico Border

Choice and the Role of Regulation

There is significant variation in school-choice policies, particularly, in terms of how they are regulated, with implications for equity. Research analyzing the policy conditions under which choice actually leads to equity, particularly in terms of desegregation, makes a clear distinction between unregulated choice policies, those often advanced by market advocates, and more regulated choice policies designed carefully to ensure equity (Orfield & Frankenberg, 2012).

Advocates of a more unregulated, market-driven model of choice assume that by removing regulations such as residential school assignment and uniform curriculum across schools, open enrollment intensifies competition and choice, and subsequently creates incentives for schools to innovate and improve their academic outcomes (Chubb & Moe, 1990; Finn, Manno, & Wright, 2016; Friedman, 1955). Examples of these market mechanisms include school vouchers and charter schools. Some market advocates also claim choice promotes equity through increased access and as a means for breaking down segregated housing patterns as zip codes become less tied to school attendance. Advocates often assume that parents have equal ability and eligibility to act on choice policies. Therefore, parents of low-achieving students, or those with the greatest disadvantage, can choose to enroll in more advantaged schools or schools that better address their students’ needs, thereby decreasing segregation (Howell & Peterson, 2002). Theoretically, competitive forces will push the schools and principals to be more competitive through innovation, efficiency, and effectiveness (Holme et al., 2013; Lubienski, 2005), a theoretical outcome that lacks robust empirical support (Belfield & Levin, 2002; Lubienski, 2003; Ni & Arsen, 2010). Too many restrictions or regulations will, advocates argue, prevent school choice markets from operating effectively. Critics of these market-based models point out that rational choices are often constrained by access to information (Bell, 2009), geography and transportation (André-Bechely, 2005), school capacity (Holme et al., 2013), targeted marketing (Jabbar, 2016), and cream skimming (Fuller, 2014).

Long before the proliferation of charter schools and vouchers in the 1990s, school districts had and continue to have their own quasi markets including magnet schools, intradistrict transfer programs, and interdistrict transfer programs between two or more districts. Typically, district-directed choice programs come with greater regulation and more explicit equity goals than charter school and voucher policies, but they still vary immensely in terms of regulation. For instance, magnet schools—long used in both interdistrict and intradistrict plans—differ greatly in their level of regulation, strategy, and enrollment procedures. During the 1970s, magnets acted primarily as a means for racial integration by attracting White students from the suburbs and private schools (Grooms & Williams, 2015). However, with the end of most court ordered desegregation plans, only some magnet schools continue to promote desegregation while most emphasize choice for parents (Goldring & Smrekar, 2002). In some cases where magnets exist within a school, they make the overall school population seem diverse despite stark within-school segregation (Siegel-Hawley & Frankenberg, 2012). Today, some magnet schools use selective admissions, but most operate similar to charter schools, using lotteries or open applications (Bifulco, Cobb, & Bell, 2009).

Since the Supreme Court decision in Milliken v. Bradley (1974) limited the legal justification for mandatory interdistrict desegregation plans, the few existing regional enrollment policies tend to rely on districts’ voluntary participation (Grooms, 2016; Holme & Finnigan, 2013). Finnigan et al. (2015) highlight how eight metropolitan regions use interdistrict “collaboratives” to actively promote racial and economic diversity through transfers. Still, most interdistrict transfer policies in the United States, including the one in our study, are not collaborative or implemented with explicit goals of desegregation. According to Holme and Richards (2009), states promote interdistrict plans to expand choice while encouraging district competition through effectiveness and innovation. However, in their study of an interdistrict transfer program in the Denver Metropolitan Area, they found a stratified market where mostly parents with higher socioeconomic status (SES) chose to transfer to higher SES, higher achieving districts.

Over the past 25 years many districts have experimented with different types of open enrollment policies within the district, but they vary greatly in terms of regulation. For example, the Jefferson County District in Kentucky uses a highly regulated controlled choice desegregation program where student ranking of school choices is weighted by diversity considerations (Frankenberg, 2017; Siegel-Hawley, 2013). Other intradistrict programs such as New York City’s school choice program and EPISD’s policy are referred to as unregulated, or race/income-neutral, in that there are no provisions to ensure that school choice promotes racial and economic desegregation. Instead, similar to charter schools, these plans operate under the rationale that offering choices for various magnet and neighborhood schools inherently promotes equity by expanding access. In some cases, such open enrollment policies are effective in breaking up concentrations of minority and high-needs students, but without equity safeguards and transportation provisions, race/income-neutral polices often result in increased stratification (Cobb & Glass, 2009; Holme et al., 2013).

In summary, there are two competing theories on how choice policies—be they charter schools or district-run open enrollment plans—affect equity and student segregation. On one hand, because eligibility and ability to act on transfer policies is inequitably distributed across families, schools that have the highest concentrations of low-income and minority students may become more segregated because the most advantaged parents within those schools will transfer out, while few parents will transfer into such schools. On the other hand, low-income and other high-needs students living in low-income neighborhoods, often with poorly performing schools, may enroll in schools that are higher performing or provide some sort of value for their child. Despite research testing these theories, questions remain in terms of how school leaders influence the process as well as how the market operates in different geographic contexts, including a metropolitan border context such as El Paso.

Principals and Competition

A developing literature uses survey and qualitative methods to examine how school principals operate in a competitive market environment (Davis, 2013; Kasman & Loeb, 2013). However, the majority of these studies investigate how traditional public school principals respond to the presence of charter schools (Campbell, DeArmond, Guin, & Warnock, 2006; Davis, 2013), or how charter schools compete with other charter schools (Jabbar, 2015). In the case of competition between charter schools, researchers find that principals engage in an array of strategies including improved test scores, marketing, screening, and programmatic differentiation (Lubienski, 2005; Olson & Beal, 2016). Conversely, the response by traditional public-school principals from charter competition depends on the local context and the quality and market share of charter schools in the area (Arsen & Ni, 2012; Hess et al., 2001).

A few studies examine the principal role in intra- or interdistrict open enrollment plans (DiMartino & Jessen, 2016; Frankenberg, Ayscue, & Tyler, 2016; Holme et al., 2013; Kasman & Loeb, 2013). Holme et al. (2013) studied principal and teacher responses in two low-performing public high schools operating in a city with open enrollment and charters. They found that principals at these lower performing schools were aware of declining enrollments and the loss of higher achieving students, but, out of survival, the principals primarily focused their energy on raising test scores rather than curricular changes or marketing their schools. However, Holme et al. (2013) observed that district support for one of these schools helped the principal market the school and overcome some of the community’s negative perception of the school. After New York City Public Schools opened 250 small academy high schools, researchers observed that many principals were actively engaged in marketing and recruitment for their schools, but several of the principals preferred to focus on their role as an instructional leader of the school (DiMartino & Jessen, 2016). Although these studies further our understanding regarding the role of the principal in competitive environments in general, more research is needed specifically on how leaders navigate competition in open enrollment environments given how popular, yet understudied, these policies are.

Theoretical Framework

Contextual and organizational factors influence principals’ competitive behaviors. Pressures from declining enrollment in traditional public schools, potential school closures, and test-based accountability have expanded the role of school leaders beyond organizational and instructional leadership into marketing and recruitment. In an era of school accountability, student losses and school closures are thought to reflect poorly on the school leadership regardless of other underlying causes for low-achievement (Deeds & Pattillo, 2015). Therefore, to frame our study, we draw on research and frameworks that have been applied to competitive charter school marketplaces, or “portfolio districts” (Bulkley, Henig, & Levin, 2010), such as New York City, Detroit, or New Orleans.

The level of competition between principals in an open enrollment district also depends on the organizational history of the schools as well as the public perception of the neighborhood (Holme et al., 2013; Ladd & Fiske, 2001). Whereas a brand-new charter school, magnet, or private school initially benefits from a “new car smell” (Buckley & Schneider, 2009), the reputation of a traditional public school is very much embedded in the public perception of the school and neighborhood. Principals at schools with a long-standing reputation for academic achievement or extracurricular success have greater ability to screen and select students. Conversely, a school characterized by lower academic performance, behavior issues, or scandals puts that school and principal at a disadvantage. Furthermore, principals may have to overcome being in the “wrong” neighborhood, a term fraught with racial and economic connotation (Holme, 2002; Lareau & Goyette, 2014). In this competitive structure, the perceived “winners” can to some degree “rest on their laurels,” or not engage as actively in recruitment and competitive strategy, whereas the perceived “losers,” or even “average” schools, may be motivated to reinvent their narrative by establishing a niche program or through community outreach (Frankenberg et al., 2016).

Introducing an open enrollment policy in a district with a preexisting hierarchy of schools runs the risk of intensifying stratification or segregation. Parents with relatively higher levels of education are more likely to participate in choice programs (Ogawa & Dutton, 1997), and these parents are likely to seek opportunities at higher performing schools, which tend to enroll higher income students (Holme & Richards, 2009). Poorer schools are at risk of losing their most economically advantaged students. Ladd and Fiske (2001) describe this process as an “uneven playing field of school choice” where students are sorted by race and/or social class and achievement gaps widen. In turn, the less successful schools, and, in particular, their principals, can have an even greater competitive burden. Although Ladd and Fiske’s (2001) study took place in New Zealand, they find parallels with school choice plans in the United States and view their study as a cautionary tale. Existing market hierarchies influence school leaders’ perceptions of competition, which can reinforce rather than ameliorate long-standing inequities (Jabbar, 2016).

Given the dearth of literature on principal perceptions of competition in open enrollment districts, this study examines the following research questions:

Since understanding the local context is critical for situating these research questions, the following section provides some background on EPISD and the origin of their open enrollment plan.

Data Sources

We first used quantitative school-level data from Texas Education Agency (TEA) for the academic years 2009-2010 to 2015-2016, combined with school-level transfer data provided by EPISD for the year 2014-2015 (the first year of the open enrollment plan). TEA data include total enrollment and student demographics (e.g., race/ethnicity, percent FRL, and percent classified ESL). EPISD data include the number of students who transferred to a different school, and whether they live inside or outside the residential boundaries of the school. The EPISD data also include the number of out-of-district transfers, but not which district sent them. We define “net transfers” as the number of students enrolled in a school through the transfer program, minus the number of students assigned to that school that successfully transferred to another school. Schools with positive net transfers gained more students through the choice policy than they lost, whereas schools with negative net transfers were net losers of students. The “net transfer rate” is the net transfers during the first year of the program divided by the school’s enrollment that year.

As shown in Figure 1, we drew on our analysis of the transfer data to select a stratified purposeful sample of 12 school principals from across the school district. We wanted to capture principals in schools that held different positions in the market hierarchy, since their particular context likely shaped their competitive strategies. The categories included (1) net “winners,” schools that received large shares of in-transfers while losing few students resulting in positive net transfers; (2) net “losers,” schools that lost large shares of students via out-transfers and received few in-transfers resulting in negative net transfers; and (3) little movement schools, schools that experienced few transfers in either direction. ⁶ We initially selected one elementary, middle, and high school from each of these categories and we later added a second elementary, middle, and high school from the net winner category to further explore competitive practices at the most coveted schools (see Table 2 for school descriptions).

OPEN IN VIEWER

Figure 1.

Explanatory sequential mixed methods design.

OPEN IN VIEWER

Table 2.

Qualitative Sample.

School	Category	%FRL	%ELL	%URM
Mesa Verde MS	Winner/High-In	50	10	90
San Marcos HS	Winner/High-In	50	10	80
Crocket EL	Winner/High-In	50	30	90
Kennedy HS	Winner/High-In	40	10	80
Dewey MS	Winner/High-In	80	20	90
Chavez EL	Winner/High-In	50	20	80
Haskins HS	Loser/High-Out	70	20	100
O’Connor EL	Loser/High-Out a	80	30	90
Artis MS	Loser/High-Out	80	20	90
Rio Grande EL	Little movement	90	80	100
Worsley MS	Little movement	100	20	100
Alvarado HS	Little movement	70	10	90

Note. %FRL = Percent free/reduced-price lunch; %URM = percent underrepresented minority; %ELL = percent English language learners. All numbers rounded to nearest 10 to preserve anonymity of participants. HS = high school; MS = middle school; EL = elementary school.

a

O’Connor also had a high in-transfer rate, but further investigation revealed that this was a function of a specialized program for students with disabilities.

In the sampled schools, we conducted semistructured interviews on site with the principals that lasted between 30 minutes and an hour. Although principal experience varied, all the principals interviewed had been a principal at their present school for at least 2 years including the period of policy implementation. The principals also ranged in background; about half of the sample was female and half was Hispanic. However, we did not observe any relationships between principal background and competitive practices.

We conducted interviews during the 2016-2017 academic year, which allowed us to capture the principals’ reflection on their competitive practices and practices during the first 2 years of the program. In particular, our interview protocol focused on: (1) principals’ background and experience, (2) enrollment, (3) competition, and (4) marketing and recruitment. Interviews were audio-recorded and transcribed (we use pseudonyms for schools and principals). Principals were selected as the primary participants since they had final say in admissions. Although some principals involved other staff in marketing, recruitment, and screening; all principals oversaw and were heavily involved in all these open enrollment activities. For additional context, we also interviewed one district official, reviewed EPISD internal and external documents (e.g., school board minutes) pertaining to the choice policy as well as local media coverage of the plan.

Data Analysis

To analyze our data, and to explore alignment between the qualitative and quantitative sources, we used an iterative approach. First, to answer Research Question 1, on how principals behave within a competitive open enrollment environment, we analyzed qualitative interview data to understand principals’ perceptions and strategies. Two independent readers analyzed the transcripts, creating inductive coding categories (e.g., perceived competitors, control over student body, marketing strategy, and niche strategy). We then met to discuss and revise the codes and themes. Finally, we used the transfer categories, developed from the quantitative data, to explore differences between three groups of schools (“winners,” “losers,” and schools with little change), to understand how a school’s position in the market hierarchy influences its competitive practices.

To answer Research Question 2, which asks how principals’ behavior shapes students’ access to the school of their choice, we also drew on qualitative interview data to understand principals’ practices around the open enrollment policy (e.g., the extent to which they used student discipline in their decisions to accept or reject transfer applicants).

For Research Question 3, to understand how economic and racial segregation across schools changed over the period during which open enrollment was implemented, we analyzed the quantitative data and used the qualitative findings to explain some of the patterns we found. We used four approaches to address Research Question 3. The following explanations of our analysis use %FRL, but we also replicate each analysis for %URM. First, we examine whether high-poverty schools experienced increases in %FRL and whether that change is associated with their net loss of students (from 2013-2014 to 2014-2015, the years before and immediately after implementation of the choice policy). We define F R L i l o w and F R L i h i g h as schools in the lowest and highest quintile of %FRL, respectively (and use the middle three quintiles as the reference category). ⁷ We use net-transfer rate to capture the extent to which a school was affected by the choice policy—either by gaining or losing students. We regress the change in %FRL from 2013-2014 to 2014-2015 on a school’s net-transfer rate, high- and low-poverty indicators, and interactions between net-transfer rate and poverty indicators. In other words, we assess whether high-poverty schools experienced increases in %FRL, or whether low-poverty schools experienced decreases in %FRL, and whether those changes were related to the net-transfer rate. We estimate the following model:

Δ % F R L i = α 0 + α 1 F R L i l o w + α 2 F R L i h i g h + α 3 F R L l o w * n e t _ t r a n s f e r _ r a t e i + α 4 F R L h i g h * n e t _ t r a n s f e r _ r a t e i + X i γ + ε i ,

(1)

where X_i includes indicators for school grade level for school i (elementary, middle school, high school) and ε_i is an error term. A negative α₃ suggests that low-poverty schools experienced greater decreases in their concentration of FRL students as the net-transfer rate increased (i.e., the more students they gained). This change may happen if non-FRL students were more likely to transfer into low-poverty schools. Conversely, a negative α₄ suggests that high-poverty schools experienced greater increases in concentration of FRL students as the net-transfer rate decreased (i.e., the more students they lost). This decrease may happen if non-FRL students were more likely to leave high-FRL schools. Either a negative α₃ or α₄ would suggest that the choice policy contributed to greater segregation by students’ family income. To differentiate the types of transfers, we run the same models exchanging net-transfer rate with transfer-out and transfer-in rate (and replicate these models for URM students).

Our second approach to assessing changes in segregation before and after the choice policy tracks the percent of FRL students over time in high (positive) net-transfer schools and low (negative) net-transfer schools. ⁸ This second approach shows how the percent of low-income students changed over time for schools most affected by the choice policy (either negatively or positively). ⁹ A third analysis divides schools into quintiles of FRL based on 2013-2014 data, the year prior to implementation of the choice policy, and tracks the percent of FRL students in those schools over time. This analysis shows whether high-poverty schools experienced growth in poverty rates and whether low-poverty schools experienced declines in poverty rates, regardless of the extent to which these schools were affected by the choice policy. Finally, we analyzed the dissimilarity and isolation segregation indices for race and FRL for an additional vantage point.

Students on the Move

Families are taking advantage of open enrollment policies in El Paso, but clear winners and losers exist in this market. Secondary schools, and especially high schools, appear to be the most high-stakes in terms of competition, which makes sense given the greater specialization of secondary schools as well as the ability of many older students to get to school independently. Figure 2, Panel A (left) shows the number of students who transferred in (on the x-axis) and out (on the y-axis) across schools. Many schools had a greater number of students who live within the school’s residential boundaries transfer to another school than the number of students who transferred into a school, leading to a net loss of students. These schools are denoted with a triangle in Panel A of Figure 2. Conversely, schools represented with a circle accepted far more students than they lost through the choice program, a net gain. Some schools saw no substantial change in enrollment as a result of the choice policy either because they had few students transferring in or out; or because they had large shares of students both transferring in and out resulting in a low-net transfer rate. These schools are represented with an X in Panel A of Figure 2. Panel B demonstrates that the variance of net-transfer rates tends to be larger in middle and high schools. For example, most elementary schools experienced between −10% and 10% net transfer rate, with some outliers. In contrast, 7 of 10 high schools gained or lost about 20% of their students through transfers. Because middle and high schools have larger enrollment, these differences represent an even greater number of actual students transferring. Together, these figures demonstrate that students in EPISD widely use open enrollment while certain schools are far more active players in the market place.

OPEN IN VIEWER

Figure 2.

Transfer rates across schools in the El Paso Independent School District, 2014-2015. (A) Percentage of students who transfer in and out of schools. (B) Net transfers by school level.

As expected, the market hierarchy also maps onto the SES and racial composition of schools (see Table 3). Although the lowest poverty and lowest minority schools seem to be the most desirable transfer destinations, the pattern for other categories is less clear. We report the transfer data as (1) the total number of students transferring into schools and the percent of enrolled students who transferred in from another school (the “transfer-in rate”), (2) the number of students transferring out of a school and the number of transfer-out students divided by the number of currently enrolled students (the “transfer-out rate”), and (3) the total net transfers, which is the difference in the number of students who transfer in and transfer out. The “net transfer rate” is the net transfers divided by the number of enrolled students. The rate of transfers in to schools is decreasing with the percent of minority students (Panel A) and low-income students (Panel B). However, there is no linear relationship between the net transfer rate and URM or poverty quintile. The highest out-transfer rates come from the second highest and second lowest URM and FRL quintiles, suggesting that students in the most disadvantaged schools may lack the means—information or transportation—to transfer.

OPEN IN VIEWER

Table 3.

Transfers by Quintiles.

	Total	Low	Low-Middle	Middle	High-Middle	High
Panel A: Quintiles of percent underrepresented minority
Students who transferred into the school	132.3	190.6	165.9	133.7	89.1	79.0
Percent of enrolled students who transferred in	18.3	23.3	21.3	20.2	13.5	12.8
Students who transferred out of the school	145.2	140.9	175.8	110.2	187.8	106.7
Percent of enrolled students who transferred out	19.5	14.3	22.1	17.7	25.2	18.0
Net transfers	−12.9	49.6	−9.9	23.5	−98.7	−27.7
Net transfer rate, %	−1.2	9.0	−0.8	2.5	−11.7	−5.2
Panel B: Quintiles of percent low-income
Students who transferred into the school	132.3	177.2	157.6	132.4	105.8	85.7
Percent of enrolled students who transferred in	18.3	20.3	22.3	18.9	15.7	13.8
Students who transferred out of the school	145.2	153.8	165.0	113.7	184.0	105.1
Percent of enrolled students who transferred out	19.5	15.5	20.8	18.3	25.3	17.3
Net transfers	−12.9	23.4	−7.4	18.7	−78.3	−19.3
Net transfer rate, %	−1.2	4.8	1.5	0.6	−9.6	−3.5

Note. The total column is repeated in each panel. This table demonstrates a monotonic relationship between the percent of students transferring into a school and the quintile of poverty (and to some extent minority). In contrast, percent of students transferring out of a school is not systematically related to the percent of low-income or underrepresented minorities. Sources: Transfer data from El Paso ISD.

The extent to which schools experience net gains or net losses or even little change in enrollment is likely to influence the ways in which their leaders perceive and respond to the competition created by open enrollment. To unpack these differences, we sampled schools based on these transfer patterns. Specifically, we sampled six schools with high rates of “in-transfer,” and positive net transfer rates; three schools with high “out-transfer,” or net losses in students through the transfer program; and three schools with little change in terms of in or out transfers.

We first wanted to explore the extent to which these categories aligned with how school leaders perceived enrollment changes. For example, schools may lose students via transfer, but principals may be differentially attentive to or aware of these changes. We noted that there was strong alignment between the transfer category to which the school was assigned and their perceptions of how open enrollment had affected their enrollment. With one exception, school leaders’ perceptions of their enrollment patterns were in line with the categories we ascribed. For example, all high “in-transfer” schools perceived that they were “at capacity,” or had “steady enrollment” or growth in student enrollment. One high in-transfer school noted that they had seen growth but were not yet at capacity. Similarly, all the high “out-transfer” schools noted a decline in student enrollment that was in line with their net losses. For example, the school with the greatest net loss noted that they were “way below” capacity, and the school with the smallest change in this category noted, “some decline.” This suggests that school leaders were attentive to transfers and were aware of their relative positions in the market. Next, we describe how these transfer rates influenced school leaders’ behaviors and experiences.

School Context Influences Experiences of Competition

As a central office staff member noted, the purpose of open enrollment was “to increase the friendly competition among our schools . . . that sense of urgency where we do need to be better than the guy down the street because . . . we want to attract students.” As she noted, the hope was that this policy would “get rid of complacency in our schools.” Principals in EPISD did view competition as an important part of their job, though focus on competition was varied based on school context. Despite the prevalence of competition in general, some schools were shielded from these pressures due to their market position and current enrollment. For example, many schools with high rates of in-transfers did not perceive as much competition. The principal at Dewey Middle School, a high in-transfer school, for example, noted that although competition was “out there,” it was not a concern for his school. Similarly, leaders at both Kennedy and San Marcos High School noted that they were at or above capacity, and generally felt little competition with other schools. As the principal at Kennedy High School noted, they had more than 100 students on their waitlist as a result of open enrollment, so there was little pressure to compete for students with nearby schools—many of which were struggling academically. Another school leader, of a school that was “winning” under the open enrollment system, noted that no other school “has the volume of transfers that we have here. Not even close.” The school “attracts kids from all over the city,” taking several hundred transfers.

On the other hand, school leaders at high out-transfer schools (“losers” under the open enrollment policy) and some high-in transfer schools perceived fiercer competition. As one leader said, “We are fighting for each student” (Haskins High School). Likewise, the principal at Mesa Verde, a relatively high-achieving middle school explained, “the level of competition is getting higher. We are all vying for well-rounded students.” The principal at another school similarly noted that they were losing students, and took to marketing the school, visiting feeder schools, and working to “do something that other schools weren’t doing” (Principal, Artis Middle School). The elementary school leader of the school with high out-transfers, however, while noting a loss of more than 150 students, sensed that it was a short-term issue. In this way, schools’ enrollment categories were related to and influenced by their perceptions of competition.

School Leaders Develop Strategies to Compete

Based on their position in the market, as winners or losers, schools varied in their behavioral responses and strategies when faced with competition. In particular, we identified three major strategic responses: developing a specialized academic program/market niche, marketing/ recruitment of students, and selecting or screening students. We discuss each of these below and describe how these varied by schools’ position in the market place (see Table 4).

OPEN IN VIEWER

Table 4.

Perceptions and Strategies of School Leaders.

School Category	Perceptions of Competition	Behaviors and Strategies
High out-transfer	Strong awareness of competitors and decline in student enrollment	• Marketing and recruitment
		• Beginning to adopt specialized programs
		• Few can select students
Little movement	Mixed. Some report a little competition in entry grades, but others note that parents have few other options because of language barriers and geographic constraints	• Lack of strategy at one school
		• Some marketing
		• Some adoption of specialized programs
		• Some selection, often in informal ways
High in-transfer	A little competition, but all in privileged position in market (e.g., high parent demand), benefitted from open enrollment policy	• Marketing, recruitment (proactive, use networks)
		• Specialized programs
		• Strong selection on discipline, attendance, capacity

Establishing or Protecting a Market Niche

The principals in our study either believed that their school occupied a preexisting market niche or were working to establish one. Some of these market niches included magnet programs, sports teams, and a unique school culture. The schools that were “winners,” or in a privileged position in the market due to high parent demand, had more established specialized programs and were tasked with maintaining that privileged position. In the case of one popular elementary school, Crocket Elementary, the long-serving principal of the school felt that the school already had a strong market position with their extremely affluent neighborhood as well as their well-established Spanish dual language and magnet programs. Crocket recently added Mandarin instruction. Although the principal at Crocket claimed this addition had nothing to do with competitive pressure, the Mandarin program certainly helped the school maintain market share among higher SES families in the area and differentiate themselves from the other dual language programs. Other “winners,” including Kennedy High, San Marcos High, and Mesa Verde Middle School, also promoted or relied on their specialized or magnet programs to maintain their student enrollment and reputation. Some of these “winners” were concerned about the future of competition in the city as a result of the rise in specialty programs. As one leader noted,

Competition is stiff. When we opened, there was no STEM program . . . It used to be that we were the only game in town. We’re not that way anymore. . . . Everyone wants those kids and, unfortunately, it is a business. And everyone has got a magnet [program] now. (San Marcos)

Therefore, even schools that were among the “winners” in the system, at the top of the market hierarchy, framed competition in terms of specialized programs, and feared they may lose their market position as other schools expanded or adopted such programs.

While some magnet programs existed prior to open enrollment, particularly, many of those at the “winner” schools, others had been recently added to coincide with the expansion of open enrollment. For example, while all high schools we studied had magnet programs, many of the middle schools were following suit in establishing specialty programs. Schools that were losing students were especially pursuing this strategy to stabilize or increase enrollment. For Haskins High, a lower achieving high school in a lower SES neighborhood, the principal realized that he would have a hard time competing with some of the more successful high schools, so he pushed a criminal justice magnet designed to attract “the tactile student” to help stave off a declining enrollment. The principal at Alvarado High described how he did a “market analysis” to understand why many of his neighborhood students were leaving as well as how to identify an in-demand niche capable of attracting in-transfers. He ultimately concluded that an early college high school program would help retain and attract students. In another case, a principal at a lower SES middle school, located near several more advantaged middle schools, had spent years setting up an International Baccalaureate program (Artis Middle School). However, the principal at Artis acknowledged that the school was still struggling to retain and attract students, so his goal was to get them in the door for a tour in order to break down preconceptions. While principals at high out-transfer schools adopted niche strategies, they also noted that in a few years, competition might become even stiffer. As the principal of one high school that was losing students said, “I can only imagine [that in] two, three, five years . . . it’s gonna get more difficult for public high schools to compete with all the specialty schools” (Principal, Haskins).

Marketing and Promoting the School

Although all schools engaged in some form of marketing and recruitment activities, the schools with high rates of out-transfers felt greater pressure to recruit students. At Artis Middle School, for example, a school that had lost a large population of students in the past year, the principal said that he was reaching out to feeder schools and pointing out the limitations of the open enrollment policy for low-income parents without transportation:

I take the time to go and introduce myself, to introduce the programs that we have here within my school so parents know. Why are you gonna drive your child over to [other school] if you can have him here at Artis? Why are you gonna go all the way to [other school]? Parents need to understand, if you transfer you don’t get any transportation. Those are the kind of things that you tell them. “You belong to Artis, just go to Artis. We will bring your child to our school and we will take your child back to your house.”

Similarly, other schools struggled with public relations because of their reputation. Another high out-transfer school leader noted that “negative publicity” hurt them, and he had to work “one-on-one” with parents to “get them to know who we are” (Principal, Haskins). Similarly, at O’Conner Elementary, the principal talked to parents who were considering leaving, one-on-one, to convince them to stay.

Schools that had high rates of in-transfers engaged in recruitment and marketing, but felt less immediate pressure to do so, and were also able to engage in more proactive strategies and leverage existing networks. For example, one successful high school leader (Dewey High School) said, “Competition is huge. I’m trying to figure out, how do I advertise publicly? How do I get our name out even further? Do it by TV, billboard, by just going by the new homes?” Another school noted that their strong reputation meant that they did not have to recruit as heavily: “With our magnet and I guess our reputation, we get kids from everywhere. Everywhere” (Principal, San Marcos High). Chavez Elementary, another high in-transfer school, had strong networks and partnerships with the military, ¹⁰ and used those relationships to help promote the school through word-of-mouth. Many schools thus engaged in some kind of marketing strategy, but schools that were losing most students through open enrollment were most pressured to do so.

Screening and Selecting Students

Most of the schools used both formal and informal ways to screen and select students, but the “winner” schools tended to use these strategies more since they could afford to lose students. “Winner” schools reported using discipline, attendance, and capacity constraints to screen out students frequently, while schools that were “losing” in open enrollment were often unable to do so. Schools at capacity were more able to select and control their student body, since they could turn away students. They could easily reject applicants due to capacity constraints, although “capacity” was a fluid term, without a clear definition. They also often sent students back or rejected students because of discipline or attendance. As the principal of one school said, they “absolutely” rejected applicants on those grounds, and admitted that it was somewhat subjective. One school believed that parents had too much power, and wanted to be able to consider more than just attendance and behavior but also grades:

But, yeah, we want to bring all those [high performing] kids in. Absolutely. Cause what’s the end result set by the state of Texas? STAAR scores. And we worry about that stuff. Yes. We look for kids to grow, but in the end, what are schools judged by? TEA. By the districts. The STAAR scores. . . . When you’re filling your basket with kids that you know are going to do well, it helps you out. That’s competition like a professional team. You’re not just gonna pick some schmoe, draft some schmoe off the bench. You’re going out for the best and that’s competition and we’re very competitive.

Therefore, while this high in-transfer, or winner, school felt competition, it was a qualitatively different form of competition—competing for the best, rather than simply competing for students to fill seats.

Schools with high out-transfers needed to accept students to maintain enrollments, but sometimes engaged in similar formal and informal strategies to shape their student bodies. For example, the principal at O’Connor Elementary said that their school never turned anyone down, but they might try to contact the parents first if they had concerns with the students’ record. Other informal ways of shaping enrollment even when the school was not at capacity were by encouraging parents to stick with their assigned school due to transportation challenges, as the principal of Rio Grande, a school with little movement noted. The Rio Grande principal could not reject the students since they met the transfer criteria and the school was way under capacity, but the principal encouraged them to withdraw after observing attendance issues related to the lengthy commute. Artis, a high out-transfer school, accepted students with discipline issues because the school needed to increase enrollment. As the principal stated, “I cannot not take anybody.” One high out-transfer school still rejected students based on discipline, suggesting that the leader was willing to remain underenrolled to avoid serving more challenging students. The leader at O’Connor noted that her school did not turn anyone away, but would put them on a waitlist if they were at capacity. Therefore, when it came to transfers, most principals used the screening system to recruit higher performing students or those with fewer perceived disciplinary issues, except for some schools that were more desperate to increase student enrollment.

Although the district no longer officially allows academic qualifications to be considered in a transfer request, some principals spoke openly about screening students. Principals suggested that although the academic criteria could not be formally used anymore, they did have access to it for within-district transfers and it is possible that such information influenced decisions on students with borderline attendance or discipline issues. One principal complained that they did not necessarily have access to information from out-of-district transfers, but there was pressure to accept them in order to get the district enrollment up. The principals at higher achieving, high-demand schools said they often rejected within-district applicants based on arbitrary attendance or discipline grounds. However, rejected students could appeal and request an in-person meeting at which time, the principal met with a parent or guardian and reviewed their academic and discipline record. One high in-transfer school rejected many “concern” students initially, forcing them to go through the appeals process. The principal explained that this allowed them to make expectations clear, and to ensure that it was a good fit, but this also created an additional hoop for students and parents. Although principals indicated that they were usually willing to give the students a chance if they went through this process, these were ways that principals could weed out more challenging students if the parents were unable or unmotivated to challenge the rejection. In interviewing a district official, the district was aware that the current attendance and discipline criteria were too vague and they were working on providing more formal guidance on what constituted grounds for a rejection. A district official acknowledged, “principals have very mixed feelings” about these guidelines because they feel “Why am I taking that problem child from somewhere else?” However, she felt that with some district guidance, schools would understand that most transfer students would not pose problems for the school. However, even with changing guidelines, principals would still be able to view academic information and use this information informally in their evaluation of transfers who did not meet the attendance or discipline thresholds.

Does Open Enrollment Lead to Segregation of Poverty and Underrepresented Minorities?

Although our qualitative findings and policy analysis raised concerns that open enrollment would lead to increased socioeconomic and racial segregation, our quantitative analysis shows mixed results for this scenario (see Table 5 and Figure 3). The regression results in Table 5 suggest that greater transferring may have been associated with declines in percent FRL for schools that already had the lowest poverty rates. Specifically, we report in Table 5, regression coefficients based on Equation (1), which show the extent to which transfer rates contributed to changes in the concentration of low-income students, for schools already at the highest and lowest quintiles of percent FRL. The first row in Model 1 shows that, holding constant the net-transfer rate, schools in the highest-poverty quintile experienced decreases in their percent of FRL students in the first year of the choice policy implementation (relative to the last year before implementation). Net transfer rates were not associated with changes in percent of FRL students from before to after the choice policy was implemented. However, as shown in Model 2, low-poverty schools with average net-transfer rates did not see changes in the percent of FRL students; these schools saw a decline in the percent of FRL students as the net-transfer rate increased. A 10-percentage point increase in the net-transfer rate is associated with a decrease in the percent of FRL students of 2.01 percentage points for schools already in the lowest quintile of percent FRL. In other words, the low-poverty schools most affected by the choice policy (those with highest/most positive net transfer rates), saw decreases in their percent of FRL students, suggesting that students transferring into low-poverty schools were more likely to be non-FRL.

OPEN IN VIEWER

Table 5.

Regression Coefficients Predicting the Change in the Proportion of Students Eligible for FRL, from 2013-2014 to 2014-2015.

	Net Transfer Rate		Transfer-Out Rate		Transfer-In Rate
	(1)	(2)	(3)	(4)	(5)	(6)
High-poverty schools (highest quintile %FRL)	−0.033*	−0.038**	−0.034**	−0.065*	−0.032*	−0.012
	(0.013)	(0.013)	(0.013)	(0.031)	(0.013)	(0.029)
Low-poverty schools (lowest quintile %FRL)	0.008	0.020	0.007	−0.018	0.011	0.059 +
	(0.014)	(0.016)	(0.014)	(0.031)	(0.014)	(0.030)
Net transfer rate	0.014	0.073
	(0.039)	(0.047)
Percent of student who transferred out			−0.048	−0.103
			(0.062)	(0.076)
Percent of student who transferred in					−0.012	0.087
					(0.059)	(0.081)
Interactions
High-poverty school × net transfer rate		−0.13
		(0.116)
Low-poverty school × net transfer rate		−0.201*
		(0.097)
High-poverty school × transfer-out rate				0.166
				(0.150)
Low-poverty school × transfer-out rate				0.138
				(0.173)
High-poverty school × transfer-in rate						−0.116
						(0.176)
Low-poverty school × transfer-in rate						−0.224 +
						(0.123)
School level
Middle school	−0.008	−0.005	−0.011	−0.009	−0.01	−0.006
	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)	(0.014)
High school	−0.048**	−0.044**	−0.047**	−0.044**	−0.049**	−0.047**
	(0.016)	(0.016)	(0.016)	(0.017)	(0.016)	(0.016)
Constant	0.008	0.009	0.018	0.029	0.010	−0.009
	(0.009)	(0.009)	(0.016)	(0.019)	(0.014)	(0.017)
N	78	78	80	80	79	79
R 2	0.181	0.233	0.185	0.202	0.177	0.214

Note. %FRL = percent free/reduced-price lunch.

+

p < .1; *p < .05. **p < .01. ***p < .001.

OPEN IN VIEWER

Figure 3.

Changes in school-level student demographics before and after implementation. (A) %FRL for high– and low–net transfer schools. (B) %FRL for schools in the highest and lowest quintile of %FRL in 2013-2014 (year prior to implementation). (C) %URM for high– and low–net transfer schools. (D) %URM for schools in the highest and lowest quintile of % URM in 2013-2014 (year prior to implementation).

Models 3 to 6, which replace net-transfer rate with the percent of students transferring out (Models 3 and 4) and transferring in (Models 5 and 6), show that the relationship between net-transfer rates and changes in percent FRL for low-poverty schools is driven by more students transferring into low-poverty schools (based on the negative and marginally significant coefficient for the interaction between low-poverty schools and transfer-in rate of −0.22 in Model 6). Thus, low-poverty schools experienced decreases in their percent of FRL students most likely because of non-FRL students transferring in (rather than FRL students transferring out). The relationship between transfer rates and changes in the percent of FRL students was not significantly different for high-poverty schools (compared with those in the middle three quintiles).

We also examined changes in the concentration of underrepresented minorities in schools. Results show little changes in the percent of URM students associated with transfer rates. As shown in Row 5 of Table A1 in the appendix, low-minority schools generally saw small increases (about one percentage point) in the percent of students of color from 2013-2014 to 2014-2015, regardless of the amount of transfers in or out. For example, Model 1 shows that the increase from 2013-2014 to 2014-2015 in the percent of students who identify as an underrepresented minority was 0.7 percentage points larger in low-minority schools than for schools near the median of percent minority. Model 4 shows that low-minority schools experienced larger increases in the percent of URM students as the percent of students transferring out of those schools increased (as evidenced by the positive and significant interaction term, which suggests that among low-minority schools, a 10–percentage point increase in the percent of students transferring out is associated with a 1.1–percentage point greater increase in the percent of URM students, relative to schools near the median percent minority).

Figure 3 shows two other approaches to assessing the choice policy’s influence on student segregation by race and poverty. In Panel A, we plot the percent of students eligible for FRL each year for schools that fell in the highest, middle three, and lowest quintiles of the net transfer rate. Panel B shows how percent FRL changed over time for high- and low-poverty schools. In each case, schools that would ultimately be high transfer schools had stable percent FRL leading up to the choice policy implementation. Following implementation of the choice policy, schools with the highest transfer rates did not experience substantial changes, on average, in their percent of FRL students. These graphs show that the schools with the highest concentrations of low-income students did not experience increases in the percent of low-income students, whereas schools with the lowest concentrations of low-income students did not experience significant decreases in the percent of low-income students. Similarly, Panels C and D indicate that racial segregation has been relatively stable throughout the first 2 years of implementation. ¹¹ In fact, the lower minority schools have experienced slight gains in their proportions of URM students (which aligns with the regression results shown in Table A1 in the appendix). Finally, we also analyzed the racial and economic segregation indices: dissimilarity and isolation (see Figure A1 in the appendix). The segregation indices reinforce our general finding that socioeconomic and racial segregation were generally stable during the first 2 years of open enrollment in EPISD. In other words, implementation of the choice policy did not appear to substantially alter the already high degree of segregation in the district.

Explanations for Segregation or Lack Thereof

The quantitative analysis reveals that the lowest poverty schools were indeed receiving slightly more non-FRL students, but increases in poverty concentration were lower than expected given the principals’ descriptions of intense competition, screening practices (many of which were allowed under the district policy), and the lack of equity provisions regulating the policy. Although we caution that this plan is still in its infancy and greater hurdles may lay ahead, there seem to be several current conditions preventing more widespread segregation. Despite our concerns, rejections for students seem to be extremely rare at this point in time. Given the district’s recent cheating scandal, the Board “did not want principals to be gatekeepers of, you know, we only take high [performing] students here.” This perhaps led them to modify the original plan to eliminate grades and test scores as criteria for accepting transfers. While we found evidence that principals at successful schools were screening and rejecting students, there was ultimately pressure from the district to accept students who went through the appeal process.

Although the district packaged the open enrollment plan as something completely new, several EPISD schools already had long-standing magnet programs that attracted out-of-neighborhood students. For instance, the principal at Crockett Elementary, a school with very high rates of in-transfers, expressed indifference to open enrollment, pointing to her school’s long-standing magnet program and strong reputation. This reputation had made Crocket a prime destination for children of district employees and other professionals for over a decade. Conversely, the principal at Alvarado High School, a school with little movement, described open enrollment as an equalizer:

[Open enrollment] leveled the playing field because I can go to your backyard and recruit your kids. And you can do that to my kids. That is fine, as long as there is a leveled playing field and we all have something to offer. Then, I am okay with competition, but four years ago, when we were not in the game, it was really difficult to say “OK, you have an advantage and I don’t, but I am held to same standards and I am pushing for the same kids. But yet, you can legally take my kids . . .”

As these newer magnet and specialized programs take root, parents may become more aware of options and enrollment/transfer patterns will likely stabilize. As this occurs, concerns persist regarding whether poor students, ELLs, and students with disabilities will be concentrated in certain schools.

Transportation and geography created additional challenges and most principals felt they were primarily competing for students in nearby neighborhoods, often with similar SES. EPISD has a unique geographic layout straddling the Mexican border with Ciudad Juárez and the Franklin Mountains dividing the city into a Y-like configuration with various other districts surrounding El Paso. Notably, these surrounding districts allowed EPISD students to transfer into their schools. With a mountain range dividing the city (there is one winding road connecting the West Side and Northeast) and limited busing options, principals typically perceived their main competition to consist of their immediate neighbor schools. The principal at Rio Grande Elementary, near the entry point to Mexico, did not view her school as being involved in the competitive environment so she was not concerned with losing the few non-FRL students at her school. The proximity to the bridge meant she could only draw in students from one direction. It also meant that her students were recent immigrants or possibly international commuters who, she noted, were “without the means” to navigate open enrollment and transport themselves across the district. Conversely, she hinted that the lack of in-transfers made her rely heavily on border crossers to maintain enrollment and keep the school open.

Although our quantitative analysis indicates that the distribution of FRL and URM students has remained relatively stable across the board, the plan is not breaking up the existing concentrations of poverty and minority students either. Moreover, increased economic segregation could potentially occur more as the plan matures with successful schools hitting capacity and principals feeling more freedom to screen students. Notably, principals at some of the more successful schools noted with surprise that more private school students and out-of-district students had enrolled than any previous year during their tenure. For instance, the principal at Dewey Middle School observed, “Eight kids want to transfer from the parochial schools [this year]. . . . We’ve never had that. I’ve had a couple here and there, but not like that.” Such trends demonstrate success for the district in attracting private school and out-of-district students and it carries potential for creating more economically diverse schools in a district with high-poverty rates, overall. However, since the current plan privileges out-of-district transfers, this trend also raises concerns that more disadvantaged students will not be able to secure seats at the more popular schools. As one school with a high in-transfer rate, noted “the demographics have changed.” As the principal said, “We don’t have as many low-income [students].” Therefore, some of these patterns of stratification may exist, but be confined to particular schools.