Nearly 50 Years Post-Jim Crow

Abstract

This study addresses the segregation of English language learner (ELL) students in schools across Texas. We descriptively analyze levels of racial, economic, and linguistic isolation experienced by ELL students across the state of Texas. We also examine the association between segregation by race/ethnicity, economic disadvantage, and language proficiency with high-stakes accountability ratings. Despite nearly two decades of accountability policies that have promised equality, our statistical analyses show that a majority of ELL students in Texas still attend high-poverty and high-minority schools, and we find that segregation by socioeconomic status (SES) and race and ethnicity is highly significant for predicting whether schools will be low performing relative to high performing.

Keywords

English language learners minority students segregation accountability

Although U.S. schools are more racially diverse than ever before, they are growing increasingly segregated, with African American and Latina/o students attending more segregated schools than at any time in the past 20 years (Orfield, 2009). Although current levels of school segregation are reminiscent of the pre-Brown era, the demographics of students in U.S. schools have shifted dramatically since that time. One of the most notable shifts has been the growth of the English language learner (ELL) student population, a population that has more than doubled in the past 25 years, with high growth expected to continue (National Center for Education Statistics [NCES], 2011).

Although the ELL population is growing, the extent to which such students are affected by segregation by race, poverty, and language proficiency is not well understood. This study will first examine the segregation of ELL students in schools across Texas, the state with the second largest population of ELL students in the United States (NCES, 2010), by using school-level data from the Public Education Information Management System (PEIMS) data system. We will begin by considering how federal and state legal decisions and policies have directly and indirectly affected levels of school segregation for ELL students in the state. We then descriptively examine levels of racial, economic, and linguistic isolation experienced by students across the state of Texas. Finally, we conduct an inferential statistical analysis to understand the association between segregation by race/ethnicity, economic disadvantage, and language proficiency with high-stakes accountability ratings.

Background

Texas school districts have a long history of using native language as a rationale for segregating students. In the following discussion, we first detail the legal battles against segregation of ELL students into different school buildings, a type of isolation that has been termed “first-generation” segregation (Mickleson, 2001). We then examine challenges to the assignment of ELL students to separate classes within school, also known as “second generation” segregation (Mickleson, 2001). We then focus our discussion and analysis on the rise of a new type of segregation, which we call “third generation” segregation, which is the growing isolation of Latina/o students by race, language, and poverty.

First-Generation Segregation

In Texas, the issues of native language and segregation have been closely intertwined throughout the state’s history. Prior to Brown, many school districts segregated Mexican American students into separate schools, using their home language as a rationale for such practices (Valencia, 2008), even though the state constitution did not technically permit such segregation (as it did for African Americans). The practices of segregating students increased significantly as the state’s immigrant population from Mexico grew, especially in the 1920s (Valencia, 1991). By one estimate, by 1930, fully 90% of the schools in Texas were segregated (Gonzales, 1990; Valencia, 1991). The school segregation paralleled increasingly rigid residential segregation for Mexican migrants (Valencia, 1991).

The first suit to challenge this type of first-generation segregation was the case of Independent School District v. Salvatierra in 1930, which targeted Del Rio ISD’s practice of segregating Mexican American students into separate schools (Salinas, 1971; Valencia, 2008). Del Rio ISD justified its segregation of Mexican Americans on the grounds that segregation better served the needs of the children of migrant workers, even though Anglo migrant children were not subject to the same practices (Salinas, 1971). The district also argued that segregation was justifiable because the practice helped them to better meet the needs of ELL students. Although the parents lost in the lower court, the Texas State Court of Civil Appeals ruled that the district’s existing practice of segregating Mexican American students was not permitted because “the rules for the separation are arbitrary [and] applied indiscriminately to all Mexican pupils . . . without apparent regard to their individual aptitudes . . . while relieving children of other White races from the operation of the rule” (Independent School District v. Salvatierra, 1930). The court did, however, rule that the “children’s language deficiencies warranted their separate schooling” (Salinas, 1971, p. 15).

Although the judge in Salvatierra admonished districts against the “indiscriminate” segregation of Mexican American students, many school districts continued to segregate Latina/o students regardless of the “acceptable” rationale of language ability (Salinas, 1971). Such practices were again challenged in Texas in the Delgado v. Bastrop ISD case, filed in 1948. The case was filed after the 1947 decision in Westminster v. Mendez, in California, in which the Ninth Circuit Court of Appeals had ruled that segregation of Mexican American students in California was illegal, not because racial or ethnic segregation was illegal but because the California “Jim Crow” statutes did not include Mexican Americans (Westminster v. Mendez, 1947). Because the case was decided in the Ninth Circuit and Texas was in the Fifth Circuit, the case had no bearing on Texas law (Wilson, 2003). Lawyers in the Delgado case challenged the segregation in Bastrop ISD on the same grounds (Salinas, 1971; Valencia, 2008; Wilson, 2003). The court ruled in favor of the parents, deciding that while linguistic segregation was permissible, the “arbitrary” linguistic segregation as practiced by Bastrop ISD, in which all Mexican American students were assumed to be ELLs and segregated into separate schools on that presumption, was not. Segregation by language ability within schools, however, was deemed permissible as long as the district used objective standards (i.e., standardized tests or other assessments) to determine students’ levels of language proficiency before assigning students to separate classes (Salinas, 1971; Wilson, 2003; Valencia, 1991). However, many districts ignored the Delgado decision (Salinas, 1971; Wilson, 2003). As Wilson (2003) noted, “Most districts either ignored the mandate, or set standards that made it extremely easy for school administrators to prevent any Mexican Americans from sharing public classrooms” (p. 9).

After these rulings, many districts also enacted policies to resist integration orders, by instituting “freedom of choice” plans where Mexican American students were offered a “choice” between the formerly all Mexican school and the district’s other-all White school(s). As Gonzales (1990) writes, “Rarely did school districts exchange students as to integrate both schools. Usually these districts achieved integration only if Mexican parents sent their children to the old Anglo school” (p. 156). Little integration occurred through these choice plans, and these evasive practices were eventually struck down by the U.S. Supreme Court in Green v. New Kent County (1968).

It should be noted that the pre-Brown cases of Salvatierra and Delgado in Texas, and the Mendez case in California rested on the peculiar legal status of Latina/os, who were until that point considered legally “White”—or more specifically belonged to an “other White” category. Lawyers challenging the school segregation of Latina/os argued that because Latina/os were “other White,” that state Jim Crow laws (which permitted African American segregation, as well as often the segregation of students of American Indian and Asian descent) did not apply. Thus, the strategy of relying on the “other White” category as a basis for challenging segregation of Latina/os, as Valencia noted, came back to haunt the civil rights attorneys, as school boards began to argue that because Latina/os were “White” they had no grounds for a Fourteenth Amendment challenge:

Brown, the controlling authority in desegregation cases, could not be applied to Mexican Americans, the argument went, because in the absence of state laws allowing segregation, school boards could not be found culpable of intent, or de jure segregation. Furthermore, given that the law considered Mexican Americans as legally White, the CCISD (and other districts) desegregated their schools by pairing African American and Mexican American plaintiffs needed to develop an equal protection argument so as to be protected under Brown. (2008, p. 82)

The practice of integrating Mexican Americans (“Whites”) with African Americans for the purposes of desegregation, also found to occur in other districts, including Houston ISD (Valencia, 2008) was challenged in a 1970 case in Texas in Cisneros v. Corpus Christi ISD. Plaintiffs in this case challenged the segregation of Mexican Americans in CCISD, arguing that the Fourteenth Amendment of the U.S. Constitution did, in fact, apply to them. The federal district court ruled that Mexican Americans were entitled to the same Fourteenth Amendment protections extended by Brown (Cisneros v. Corpus Christi ISD, 1970). As Salinas (1971) noted of the case, “Cisneros is unique in that it is the first case in which a court officially recognized Mexican Americans as an identifiable ethnic minority group for the purposes of public school desegregation” (p. 3). The recognition of Latina/os as an identifiable minority group for desegregation lawsuits was affirmed by the U.S. Supreme Court in Keyes v. Denver School District No. 1, in 1973. In this case, the U.S. Supreme Court reasoned that Chicanos were “in fact recognized as an identifiable minority group and thus entitled to special services in desegregated settings” (Valencia, 1991, p. 39).

Second-Generation Segregation: Within-School Isolation for ELLS

Although the first-generation cases helped to dismantle the dual systems established by Texas for ELL students and Latina/o students, the issue of within-school segregation has been thornier for ELL students who have often been separated within schools into bilingual classrooms. As Valencia (1991) observes, after first-generation segregation was resolved, “a new form of school segregation began to emerge in American public schools—that is, language segregation within desegregated schools, a new form of resegregation” (p. 41).

One such form of segregation was the placement of students into “low ability” classes: As Gonzales (1990) notes, after court victories, particularly in the 1940s and 1950s, “schooling for Mexican children continued to come under the significant influence of pseudo-scientific intelligence testing, with its heavy tracking into slow learner, vocational, and EMR (educationally mentally retarded) classes” (p. 156).

Another tension arose around the issue of bilingual education—as many districts accused of segregating students on the basis of language were also accused of failing to provide appropriate language instruction for students to access the general curriculum (Vasquez Heilig, Dietz, & Volonnino, 2011). As the first-generation cases were resolved, the friction between bilingual education and desegregation became more apparent, as courts and districts sought to balance the need, on the one hand, to offer linguistically appropriate instruction for subgroups of students who do not yet speak English, and the danger, on the other hand, that such practices could result in racial and linguistic isolation of those students. As Gifford and Valdes (2006) write of the tension between those advocating for desegregation and those advocating for bilingual education,

For those concerned about segregation, bilingual education appeared to be a language policy that masked exclusion. For those concerned about the futility of educating students in a language that they did not understand, bilingual education was a compensatory education policy that focused on language, the condition that prevented students from accessing the curriculum. (p. 131)

Indeed, according to Valencia (1991), there were grounds for concern that bilingual education was used to discriminate against ELL students, as “many recalcitrant school systems circumvented the implementation of bilingual education programs by scattering Limited English Proficiency (LEP) children throughout their districts; others used bilingual education as an opportunity to segregate Chicano LEP children, thus separating them from White children” (p. 42).

Although courts began, in the 1970s, to recognize the rights of ELL students to integrated school settings, not all desegregation remedies addressed their linguistic needs (Arias, 2007; Valencia, 1991). This was because, as Arias (2007) notes, there was general wariness of any program that “might entail an aspect of ‘separate but equal’” (para 9). As a result, whereas some desegregation plans included exemptions for programs serving ELL students, others pushed for integrating students into schools regardless of whether or not there were programs available to meet their needs (Arias, 2007).

There has also been a concern around “tracking” of ELL students not only into bilingual classes, but also into lower or remedial academic tracks that offer little access to a college preparatory curriculum. Indeed, after many of the desegregation cases for Latina/os were won, other cases were brought challenging the use of standardized testing that often placed non-English speakers in classes for “educationally mentally retarded” students. Yet, as many scholars have observed, ELL students often remain in “dead-end” academic tracks (Suarez-Orozco, Suarez-Orozco, & Tordova, 2008). As Gandara and Orfield (2010) wrote, “Linguistic segregation at the classroom level for much of the day intensifies all the negative impacts of school segregation. For this reason, it is especially crucial to organize instruction in ways that can mitigate, not exacerbate, this segregation for students who are learning English” (p. 4).

Third-Generation Resegregation: Demographic Shifts and “Triple Segregation”

Struggles about how best to serve ELL students within integrated schools have become, over the past several decades, less relevant, as demographic shifts have left many ELLs concentrated into the same neighborhoods and the same schools. According to an analysis by Orfield (2009), levels of school segregation for Latina/os, who comprise 91% of the ELL student population in Texas, have increased substantially over the past 30 years: In 2006-2007, 40% of Latina/o students attended “intensely segregated” schools, up from approximately one third in 1988. Furthermore, the average Latina/o student that year, attended a school that was 57% low income (Orfield, 2009).

This rise in school segregation for Latina/os, as Orfield (2009) observes, reflects increases in residential segregation—as opposed to the termination of court ordered school desegregation—because an increasing number of Latina/os reside in suburban districts that were never under court ordered integration plans. Overall, the percentage of minority students enrolled in suburban schools has increased from 28% to 41% between 1993-1994 and 2006-2007 (Fry, 2009). The enrollment growth of Latina/os in suburban districts outpaced the growth of all other minority groups: The growth rate in that 15-year period was 296% (Fry, 2009). In some metro areas, however, the increase in minority population in the suburbs is much higher: In 2006-2007, the percentage of Latina/o students enrolled in large suburban districts was nearly equal to the percentage of White students in those same districts in the 25 largest Metropolitan Statistical Areas (Tefera, Frankenberg, Siegel-Hawley, & Chirichingo, 2011). Analyses of the Census 2010 found that 59% of Latina/os now reside in suburban neighborhoods (Frey, 2011).

As the proportion of Latina/os has increased in suburban districts, they have become increasingly isolated into the same schools (Fry, 2009; Reardon & Yun, 2001). In 2006-2007, according to Fry, more than two thirds of non-White students in the suburbs attended a majority minority school, and such schools educated more than 73% of suburban Latina/o students, as compared with 68% of suburban Blacks, and just 13% of suburban Whites (Fry, 2009). In 2006-2007, Tefera et al. (2011) found that “nearly 30 percent of African American and Latina/o suburban students are in hyper-segregated suburban schools with 0-10 percent White students” (p. 4).

As a subgroup of Latina/os, it is ELLs who are the most likely to experience high levels of school and residential segregation because they often live in more segregated neighborhoods, and attend more segregated schools than their U.S.-born peers (Iceland & Scopolliti, 2008; Suarez-Orozco et al., 2008). In their study of immigrant students in Boston and San Francisco, for example, Suarez-Orozco et al. found that the Latina/o students in their sample (Dominicans, Central Americans, and Mexicans) were “much more likely than either Haitians or Chinese students to attend ‘intensely’ segregated schools where more than 90 percent of the students were of color” (p. 92).

Suarez-Orozco et al. (2008) term the uniquely high levels of school segregation experienced by ELLs as “triple segregation”—isolation by poverty, color, and language. In their study, they found that intensely segregated school settings provided a “separate and not equal educational environment” for ELL students (p. 92). They found that schooling environments that students experienced—the level of segregation and poverty—were correlated with students’ academic trajectories, with ELLs more likely to decline academically at the most segregated, high-poverty schools.

In the following section, we examine levels of school segregation by race and ethnicity and for ELLs in Texas, a state with majority “minority” student population and the second highest proportion of ELL students in the United States (second to California) in 2009-2010 (Vasquez Heilig, 2011b). In 2007-2008, Texas served 13.2% of all ELL students in the United States (Batalova & McHugh, 2010). In 2007-2008, the state served 775,000 ELL students in pre-K-12, an increase of 40% since 2000 (Vasquez Heilig, Rodriguez, & Somers, 2011). By 2011, ELLs comprised 17% of the state’s student population (Vasquez Heilig, 2011b). Little research, however, has examined the extent to which ELL students are isolated by race, poverty status, or language proficiency. As a result, the study seeks to ask the following questions: What is the current state of segregation by race/ethnicity, SES, and ELL status in the state of Texas? Are segregated majorities by race/ethnicity, SES, and ELL status associated with high-stakes accountability ratings?

Method

Descriptive Analyses

The study begins by conducting descriptive analyses of publicly available 2011 school-level PEIMS data. The PEIMS was created in 1983 for the Texas Education Agency (TEA) to provide a uniform accounting system to collect all information about public education, including student demographics, academic performance, personnel, and school finances. The PEIMS lies at the heart of the Texas student accountability system, and the wealth of information gathered from school districts offers the opportunity to examine segregation statewide.

We began by recoding PEIMS variables representing the school-level proportions of African American, Latina/o, economically disadvantaged, and ELL students in each school into dummy variables denoting school-level majority and nonmajority for each demographic variable. We also created a variable by combining the proportion of African Americans and Latina/os in each school to denote which Texas schools were majority African American and Latina/o. A locality variable (which we obtained directly from TEA) was matched to the data set and then recoded into three localities (urban/central city, suburban, and rural). A school-level variable (high school, middle, elementary, and multilevel) was also used. Descriptive cross-tab analyses were then undertaken using SPSS to consider segregation by each of these variables.

Multivariate Logistic Regression

The second set of analyses examined the association between segregation of students in Texas schools and TEA accountability ratings.¹ We used the school-level 2011 PEIMS data for all noncharter public schools that received low performing or exemplary accountability ratings (N = 1599) in the state of Texas to conduct a multivariate logistic regression. This method is based on the logit transformation of the probability of achieving the exemplary accountability rating relative to low performing based on a set independent segregation variables. The basic logit model is

log (P / 1 - P) = α + \sum β_{i} χ_{i}

The following variables were included as independent variables in the multivariate logistic regression: race/ethnicity, economic disadvantage, locality, ELLs, and school level.

Findings: Persisting Segregation in the State of Texas

In the following analysis, we describe the level of segregation of ELL students in Texas. We illustrate that, consistent with national trends, ELL students are highly isolated in both urban and suburban schools. We further illustrate that ELL students in Texas experiencing “triple segregation” by race/ethnicity, language, and poverty.

Race and Ethnicity

Our first set of analyses of 2011 Texas education data examined racial isolation by examining the proportion of schools that had a simple majority (50% or more) of any racial and ethnic group and schools that had a “vast majority” (90% or more) of any racial and ethnic group. We examined levels of segregation of almost 8,000 noncharter public schools in the state of Texas.

Overall, we found that most Texas schools have a majority of one racial or ethnic group and that 43% of all schools in the state of Texas are majority Latina/o/a; 34% are majority White, 17% have no racial or ethnic majority, and 5% are majority African American (see Table 1). We found further that the Texas schools that served a “vast majority” of a single racial/ethnic group tended to be Latina/o/a: We found that 15.4% of schools in the state are “vast majority” (90% or greater) Latina/o, whereas just 2% are vast majority White, and 0.4% of schools are “vast majority” African American (results not shown).

Table 1.

Texas Public Schools by School Level and Majority Race/Ethnicity (R/E).

	No racial/ethnic majority	Majority Latina/o	Majority White	Majority African American	Majority Asian American	Majority Native American	Total
Elementary
Count	732	2,070	1,298	189	12	0	4,301
% within school level	17.00%	48.10%	30.20%	4.40%	0.30%	0.00%	100.00%
% within majority R/E	54.10%	61.00%	48.10%	51.90%	63.20%	0.00%	54.90%
% of total	9.30%	26.40%	16.60%	2.40%	0.20%	0.00%	54.90%
Middle
Count	316	627	585	69	3	2	1,602
% within school level	19.70%	39.10%	36.50%	4.30%	0.20%	0.10%	100.00%
% within majority R/E	23.40%	18.50%	21.70%	19.00%	15.80%	66.70%	20.50%
% of total	4.00%	8.00%	7.50%	0.90%	0.00%	0.00%	20.50%
High school
Count	249	586	666	87	4	1	1,593
% within school level	15.60%	36.80%	41.80%	5.50%	0.30%	0.10%	100.00%
% within majority R/E	18.40%	17.30%	24.70%	23.90%	21.10%	33.30%	20.30%
% of total	3.20%	7.50%	8.50%	1.10%	0.10%	0.00%	20.30%
Multilevel
Count	56	113	147	19	0	0	335
% within school level	16.70%	33.70%	43.90%	5.70%	0.00%	0.00%	100.00%
% within majority R/E	4.10%	3.30%	5.50%	5.20%	0.00%	0.00%	4.30%
% of total	0.70%	1.40%	1.90%	0.20%	0.00%	0.00%	4.30%
Total
Count	1,353	3,396	2,696	364	19	3	7,831
% within school level	17.30%	43.40%	34.40%	4.60%	0.20%	0.00%	100.00%
% within majority R/E	100.00%	100.00%	100.00%	100.00%	100.00%	100.00%	100.00%
% of total	17.30%	43.40%	34.40%	4.60%	0.20%	0.00%	100.00%

Segregation differs, we found, by the grade level of schools, with Latina/o/a students attending elementary schools in Texas experiencing the highest levels of segregation. Because elementary schools tend to more closely track patterns of residential segregation, this finding is consistent with the research on growing levels of Latina/o/a residential segregation. Our analysis found that 48% of elementary schools in Texas are majority Latina/o, whereas 17.5% of elementary schools in Texas are intensely segregated (90% or more) Latina/o schools. Latina/os are also segregated to a slightly lower degree at the middle school level: 39% of middle schools are majority Latina/o, with 13.8% being intensely segregated Latina/o/a schools (results not shown). Levels of segregation are slightly lower for African Americans: 4% of elementary schools in the state are majority African American, and just 0.5% are vast majority African American (results not shown).

Table 2 shows the demographics of public schools in Texas when African American and Latina/o proportions are combined. As much as 51% of all schools in Texas are majority African American and Latina/o. As a result, the majority of schools in Texas are now “majority-minority.” As with the data on majority one-race/ethnicity schools discussed previously, the highest levels of racial isolation in terms of “majority-minority” status are at the lower grades: Our data show that 57% of elementary schools are majority non-White, and 55% of middle schools are majority non-White, compared with 38% of high schools and 19% of multilevel schools.

Table 2.

Texas Public Schools by Level and Majority African American (AA) and Latina/o.

	Not majority AA and Latina/o	Majority AA and Latina/o	Total
Elementary
Count	1,835	2,466	4,301
% within school level	42.70%	57.30%	100.00%
% within majority AA & Latina/o	48.10%	61.40%	54.90%
% of total	23.40%	31.50%	54.90%
Middle
Count	723	879	1,602
% within school level	45.10%	54.90%	100.00%
% within majority AA & Latina/o	18.90%	21.90%	20.50%
% of total	9.20%	11.20%	20.50%
High school
Count	985	608	1,593
% within school level	61.80%	38.20%	100.00%
% within majority AA & Latina/o	25.80%	15.10%	20.30%
% of total	12.60%	7.80%	20.30%
Multilevel
Count	273	62	335
% within school level	81.50%	18.50%	100.00%
% within majority AA & Latina/o	7.20%	1.50%	4.30%
% of total	3.50%	0.80%	4.30%
Total
Count	3,816	4,015	7,831
% within school level	48.70%	51.30%	100.00%
% within majority AA & Latina/o	100.00%	100.00%	100.00%
% of total	48.70%	51.30%	100.00%

Analyzing the proportion of schools that are intensely segregated, serving 90% or greater populations of African Americans and Latina/os combined, reveals similar patterns. Our analysis shows that nearly a quarter (24.1%) of schools in Texas are intensely segregated (90% or greater African American/Latina/o combined); 68% of these intensely segregated schools are elementary schools (30.1% of elementary schools overall, 23% of middle schools, and 14% of high schools).

Our data also show that double segregation by race/ethnicity and economic disadvantage is still an issue in Texas (see Table 3). Of schools that are majority economically disadvantaged in the state of Texas, 93% (3,857) of them are also majority African American and Latina/o, whereas 45% are intensely segregated—90% or greater African American and Latina/o (results not shown). The stark difference in the socioeconomic status (SES) demographics of Texas schools that African Americans and Latina/os attend is readily apparent.

Table 3.

Doubly Segregated: Texas Public Schools by Majority African American (AA) and Latina/o and Majority Economically Disadvantaged (ED).

	Not majority AA and Latina/o	Majority American and Latina/o	Total
Not majority ED
Count	3,506	158	3,664
% within AA & Latina/o	95.70%	4.30%	100.00%
% within majority ED	91.90%	3.90%	46.80%
% of total	44.80%	2.00%	46.80%
Majority ED
Count	310	3,857	4,167
% within AA & Latina/o	7.40%	92.60%	100.00%
% within majority ED	8.10%	96.10%	53.20%
% of total	4.00%	49.30%	53.20%
Total
Count	3,816	4,015	7,831
% within AA & Latina/o	48.70%	51.30%	100.00%
% within majority ED	100.00%	100.00%	100.00%
% of total	48.70%	51.30%	100.00%

We next examined segregation by race/ethnicity and locality. In urban and central city areas, the vast majority of schools were majority Latina/o (63%) followed by no racial or ethnic majority (16%), majority White (14%) and majority African American (8%). Consistent with data on suburban diversification, it appears that there are high levels of diversity in suburban districts in Texas as well as high levels of school segregation. Our data show that suburban districts also had the largest proportion of majority Latina/o schools at 38%, with majority White schools close behind at 34%. In suburban areas, about 23% of schools had no racial or ethnic majority whereas about 4% were majority African American. In rural areas, 55% of schools were still majority White, 31% were majority Latina/o, 20% had no racial or ethnic majority and 3% were majority African American (see Table 4).

Table 4.

All Texas Public Schools by Locality and Majority Race/Ethnicity (R/E).

	No racial/ethnic majority	Majority Latina/o	Majority White	Majority African American	Majority Asian American	Majority Native American	Total
Urban/central city
Count	364	1,477	317	190	4	0	2,352
% within locality	15.50%	62.80%	13.50%	8.10%	0.20%	0.00%	100.00%
% within majority R/E	26.90%	43.50%	11.80%	52.20%	21.10%	0.00%	30.00%
% of total	4.60%	18.90%	4.00%	2.40%	0.10%	0.00%	30.00%
Suburban
Count	708	1,168	1,041	113	15	3	3,048
% within locality	23.20%	38.30%	34.20%	3.70%	0.50%	0.10%	100.00%
% within majority R/E	52.30%	34.40%	38.60%	31.00%	78.90%	100.00%	38.90%
% of total	9.00%	14.90%	13.30%	1.40%	0.20%	0.00%	38.90%
Rural/Ind. town
Count	281	751	1,338	61	0	0	2,431
% within locality	11.60%	30.90%	55.00%	2.50%	0.00%	0.00%	100.00%
% within majority R/E	20.80%	22.10%	49.60%	16.80%	0.00%	0.00%	31.00%
% of total	3.60%	9.60%	17.10%	0.80%	0.00%	0.00%	31.00%
Total
Count	1,353	3,396	2,696	364	19	3	7,831
% within locality	17.30%	43.40%	34.40%	4.60%	0.20%	0.00%	100.00%
% within majority R/E	100.00%	100.00%	100.00%	100.00%	100.00%	100.00%	100.00%
% of total	17.30%	43.40%	34.40%	4.60%	0.20%	0.00%	100.00%

Table 5 shows the demographics of public schools by locality in Texas when African American and Latina/o proportions are combined. As noted above, 51% of all schools in Texas are now majority African Americana and Latina/o. Our data show high levels of segregation in urban core districts: Driven by rapid growth in the Latina/o population, in urban and central city areas in Texas, 72% of schools are majority African American and Latina/o combined; with 46% of urban schools being “intensely segregated” (90% or more African American and Latina/o/a student population combined).

Table 5.

All Texas Public Schools by Locality and Majority African American (AA) and Latina/o.

	Not majority AA and Latina/o	Majority AA and Latina/o	Total
Urban/central city
Count	653	1,699	2,352
% within locality	27.80%	72.20%	100.00%
% within majority AA & Latina/o	17.10%	42.30%	30.00%
% of total	8.30%	21.70%	30.00%
Suburban
Count	1,602	1,446	3,048
% within locality	52.60%	47.40%	100.00%
% within majority AA & Latina/o	42.00%	36.00%	38.90%
% of total	20.50%	18.50%	38.90%
Rural/Ind. town
Count	1,561	870	2,431
% within locality	64.20%	35.80%	100.00%
% within majority AA & Latina/o	40.90%	21.70%	31.00%
% of total	19.90%	11.10%	31.00%
Total
Count	3,816	4,015	7,831
% within locality	48.70%	51.30%	100.00%
% within majority AA & Latina/o	100.00%	100.00%	100.00%
% of total	48.70%	51.30%	100.00%

We also find high concentrations of non-Whites in suburban districts. In suburban districts, 47% of schools are also now majority African American and Latina/o, with one fifth (20.2%) of suburban schools being “intensely segregated” (90% or greater African American and Latina/o/a combined). Historically, Whites in Texas have been the majority in rural areas overall across the state, but as of 2011, 36% of schools in rural areas are now majority African American and Latina/o.

ELLs

As stated previously, descriptive statistical analyses of the segregation of ELLs in schools in the state of Texas to date have been rare. Table 6 shows that about 9% of all schools in Texas are majority ELL. Examining the segregation of ELLs by grade levels, we find that the vast majority of schools (645) in the state of Texas that are majority ELLs are elementary—about 8% of all schools in Texas. There are very few schools in the state at the middle (16), high school levels (11), and multilevel (3) that are majority ELL. This is because many ELL students are reclassified by the end of elementary school (Gandara & Rumberger, 2009). In Texas in 2011-2012, the proportion of students classified as LEP fell from 28% to 16% between first and fifth grade, to 8% by eighth grade (Texas Education Agency, 2012). Nevertheless, although they are not the majority in middle and high school, considering that almost 17% of all students in the state of Texas are ELLs (Vasquez Heilig, 2011b), the next question is whether majority-ELL schools are dispersed broadly by locality if not by school level.

Table 6.

Texas Public Schools by School Level and Majority English Language Learner (ELL).

	Not majority ELL	Majority ELL	Total
Elementary
Count	3,656	645	4,301
% within school level	85.00%	15.00%	100.00%
% within majority ELL	51.10%	95.60%	54.90%
% of total	46.70%	8.20%	54.90%
Middle
Count	1,586	16	1,602
% within school level	99.00%	1.00%	100.00%
% within majority ELL	22.20%	2.40%	20.50%
% of total	20.30%	0.20%	20.50%
High school
Count	1,582	11	1,593
% within school level	99.30%	0.70%	100.00%
% within majority ELL	22.10%	1.60%	20.30%
% of total	20.20%	0.10%	20.30%
Multilevel
Count	332	3	335
% within school level	99.10%	0.90%	100.00%
% within majority ELL	4.60%	0.40%	4.30%
% of total	4.20%	0.00%	4.30%
Total
Count	7,156	675	7,831
% within school level	91.40%	8.60%	100.00%
% within majority ELL	100.00%	100.00%	100.00%
% of total	91.40%	8.60%	100.00%

Table 7 shows ELL segregation by locality, and the findings illustrate that most (55%) of the schools with a majority of ELL students are located in urban and central city districts (representing 16% of all urban schools), whereas a significant proportion (38%) are in suburban areas (representing 8% of suburban schools) and 6% are located in rural areas (representing 2% of rural schools overall).

Table 7.

Texas Public Schools by Locality and Majority English Language Learner (ELL).

	Not majority ELL	Majority ELL	Total
Urban/central city
Count	1,975	377	2,352
% within locality	84.00%	16.00%	100.00%
% within majority ELL	27.60%	55.90%	30.00%
% of total	25.20%	4.80%	30.00%
Suburban
Count	2,788	260	3,048
% within locality	91.50%	8.50%	100.00%
% within majority ELL	39.00%	38.50%	38.90%
% of total	35.60%	3.30%	38.90%
Rural/Ind. town
Count	2,393	38	2,431
% within locality	98.40%	1.60%	100.00%
% within majority ELL	33.40%	5.60%	31.00%
% of total	30.60%	0.50%	31.00%
Total
Count	7,156	675	7,831
% within locality	91.40%	8.60%	100.00%
% within majority ELL	100.00%	100.00%	100.00%
% of total	91.40%	8.60%	100.00%

We also find that ELL students tend to experience double segregation by language proficiency and poverty. We find specifically that 7% of all schools in the state of Texas are majority ELL and majority economically disadvantaged. We found that 15% of all schools that are majority economically disadvantaged are also majority ELL. Of majority-ELL schools, we found that 89% of them are also majority economically disadvantaged (see Table 8). Notably, we find that two thirds (66%) of the schools that are intensely poor—“vast majority” economically disadvantaged—are also majority ELL (results not shown).

Table 8.

Doubly Segregated: Texas Public Schools by Majority English Language Learner (ELL) and Majority Economically Disadvantaged (ED).

	Not majority ELL	Majority ELL	Total
Not majority ED
Count	3,592	72	3,664
% within majority ED	98.00%	2.00%	100.00%
% within majority ELL	50.20%	10.70%	46.80%
% of total	45.90%	0.90%	46.80%
Majority ED
Count	3,564	603	4,167
% within majority ED	85.50%	14.50%	100.00%
% within majority ELL	49.80%	89.30%	53.20%
% of total	45.50%	7.70%	53.20%
Total
Count	7,156	675	7,831
% within majority ED	91.40%	8.60%	100.00%
% within majority ELL	100.00%	100.00%	100.00%
% of total	91.40%	8.60%	100.00%

We also find that ELLs experience “double” segregation by language proficiency and race. Table 9 shows that 8% of all schools in Texas are majority African American and ELL. Of schools that are majority African American and Latina/o, 15% are majority ELL. Of those schools that are majority ELL, 89% of them are also majority African American and Latina/o, and a striking 81% of vast majority African American and Latina/o (90% or greater) schools are also majority-ELL schools (results not shown).

Table 9.

Doubly Segregated: Texas Public Schools by Majority English Language Learner (ELL) and Majority African American (AA) and Latina/o.

	Not majority ELL	Majority ELL	Total
Not majority AA & Latina/o
Count	3,744	72	3,816
% within majority AA & Latina/o	98.10%	1.90%	100.00%
% within majority ELL	52.30%	10.70%	48.70%
% of total	47.80%	0.90%	48.70%
Majority AA & Latina/o
Count	3,412	603	4,015
% within majority AA & Latina/o	85.00%	15.00%	100.00%
% within majority ELL	47.70%	89.30%	51.30%
% of total	43.60%	7.70%	51.30%
Total
Count	7,156	675	7,831
% within majority AA & Latina/o	91.40%	8.60%	100.00%
% within majority ELL	100.00%	100.00%	100.00%
% of total	91.40%	8.60%	100.00%

Our data also illustrates the problem of “triple segregation” for ELL students in Texas. Table 10 shows that about 15% of all schools in the state are triple segregated by majority ELL, majority economically disadvantaged and majority African American and Latina/o. Although triple segregation is negligible in middle schools (1.2%) and high schools (0%), about 24% of Texas elementary schools are triply segregated with majorities of African American and Latina/o, economically disadvantaged, and ELL students.

Table 10.

Triply Segregated Texas Public Schools: Majority Economically Disadvantaged, Majority African American and Latina/o, and Majority English Language Learner (ELL) by School Level.

	Not majority ELL	Majority ELL	Total
Elementary
Count	1,852	593	2,445
% within school level	75.70%	24.30%	100.00%
% within majority ELL	56.90%	98.30%	63.40%
% of total	48.00%	15.40%	63.40%
Middle
Count	836	10	846
% within school level	98.80%	1.20%	100.00%
% within majority ELL	25.70%	1.70%	21.90%
% of total	21.70%	0.30%	21.90%
High school
Count	504	0	504
% within school level	100.00%	0.00%	100.00%
% within majority ELL	15.50%	0.00%	13.10%
% of total	13.10%	0.00%	13.10%
Multilevel
Count	62	0	62
% within school level	100.00%	0.00%	100.00%
% within majority ELL	1.90%	0.00%	1.60%
% of total	1.60%	0.00%	1.60%
Total
Count	3,254	603	3,857
% within school level	84.40%	15.60%	100.00%
% within majority ELL	100.00%	100.00%	100.00%
% of total	84.40%	15.60%	100.00%

Inferential Statistical Analyses: Association Between Segregation and School Success

From Table 11 it can be seen that all of the school-level demographic variables considered in the descriptive cross-tab analysis are highly significant predictors of whether a school will be exemplary. Elementary schools were the referent group and are the most likely to be exemplary in Texas. Other schools levels are significantly less so as middle schools show a 91% (−2.454 log odds = e^.086; Table 11), high schools a 95% (−2.983 log odds = e^.051; Table 11) and multilevel schools also show a 76% (−1.413 log odds = e^.243; Table 11) decrease in the odds of being exemplary relative to elementary schools. Both majority economically disadvantaged (−1.645 log odds = e^.193; Table 11) and African American and Latina/o (−1.008 log odds = e^.365; Table 11) schools exhibit decreases in the odds (81% and 64%, respectively) for being exemplary relative to low performing. Locality is also a significant predictor of being exemplary as rural schools odds are 87% (2.048 log odds = e^.129; Table 11) less likely to be exemplary relative to urban schools with aforementioned demographics held constant.

Table 11.

Logistic Regression Analysis of Race/Ethnicity School Segregation and Exemplary Accountability Rating.

Predictor	β	SE β	Wald’s χ²	df	Significant	e β
Elementary			258.485	3	.000
Middle	−2.454	.200	149.816	1	.000	0.086
High	−2.983	.222	180.033	1	.000	0.051
Multilevel	−1.413	.443	10.149	1	.001	0.243
Majority economically disadvantaged	−1.645	.322	26.036	1	.000	0.193
Majority African American and Latina/o	−1.008	.331	9.258	1	.002	0.365
Urban/central city			130.905	2	.000
Suburban	0.227	.198	1.319	1	.251	1.255
Rural	−2.048	.216	89.687	1	.000	0.129
Constant	4.208	.249	285.425	1	.000	67.207

Model goodness of fit: Nagelkerke pseudo R² = .587.

In our analysis, we also examined the association between ELL segregation and school performance as measured by the likelihood of being rated “exemplary” by the state in 2011. Table 12 examines whether triple segregation of ELLs in majority African American and Latina/o and economically disadvantaged schools can predict whether schools will be exemplary relative to low performing. The odds ratios for economically disadvantaged schools are associated with an 80% (1.610 log odds = e^.200; Table 12) decrease in the odds of being exemplary. In triple segregated schools, majority African American and Latina/o schools are 48% (.662 log odds = e^.516; Table 12) less likely to be exemplary. Surprisingly, controlling for triple segregation, majority-ELL schools are 1.5 times (2.248 log odds = e^.4.125; Table 12) more likely to be exemplary. However, as noted in the descriptive analyses there are very few ELL majority middle and high schools. As a result, this positive coefficient is due to the fact that there were 72 exemplary and 15 low performing majority-ELL elementary schools in the state of Texas (results not shown). The state should be applauded for these numbers. However, there were no exemplary majority-ELL middle schools (0 of 2) and there were no majority-ELL high schools that were rated exemplary or low performing in 2011.

Table 12.

Logistic Regression Analysis of English Language Learner (ELL) School Segregation and Exemplary Accountability Rating.

Predictor	β	SE β	Wald’s χ²	df	Significant	e β
Majority economically disadvantaged	−1.610	.261	38.183	1	.000	0.200
Majority African American and Latina/o	−0.662	.255	6.746	1	.009	0.516
Majority ELL	1.417	.282	25.340	1	.000	4.125
Constant	2.248	.119	354.809	1	.000	9.473

Note: Model goodness of fit: Nagelkerke pseudo R² = .258.

Discussion and Conclusion

Our analyses of Texas data indicate that ELL students in Texas experience high levels of segregation in schools by race, poverty, and language status. The reasons for the growing segregation are less clear, and merit further research; but there are a number of potential causes indicated within the literature. One such cause indicated within the research literature is the growing residential segregation by race, language proficiency, and poverty—segregation that has increasingly been found, as our data show, in suburban school systems. It is also possible that educational policies contribute to high levels of triple segregation as well. Some policies that have been adopted with the intent of improving language acquisition, as in the past, have the effect of increasing linguistic isolation: for example, research has found that school systems that are racially diverse often adopt “clustered bilingual” programs in an effort to best serve the linguistic needs of ELL students. As in the past, a linguistic rationale used by school districts in terms of meeting the needs of ELL has the effect (intentionally or unintentionally) of increasing isolation of ELL students and reducing exposure to native speakers during the school day.

Another policy that can have the indirect effect of worsening ELL segregation is school choice: Many school districts in the state of Texas have adopted “open enrollment” policies that allow students to transfer between schools within the same district. These policies, as research has shown, tend to advantage more well-resourced students (particularly because transportation is not provided with most such policies). Due to differences in cultural and social capital, it is likely that students whose home language is not English are less likely to take advantage of choice (Vasquez Heilig, 2011a) due to lack of familiarity with the application process.

One of the most significant contributions to segregation in schools, however, is housing. ELL students, who are often Latina/o, are increasingly residentially isolated in urban and, increasingly, suburban neighborhoods. As Gandara and Contreras (2009) observed, “Housing segregation has particularly onerous effects on Latina/o students learning English. When students’ lack appropriate language models and individuals with whom to interact in English, their acquisition of academic English is delayed” (p. 74). This lack of opportunity is exacerbated when students residing in high-poverty and linguistically isolated neighborhoods attend schools isolated by race/ethnicity, poverty, and language.

In conclusion, nearly 50 years since Jim Crow, the intensity of segregation in Texas schools is still largely problematic. Our statistical analyses show that a majority of ELL students in Texas attend high-poverty and high-minority schools. One positive note is that elementary schools serving ELL students are more likely to be high performing than low performing schools. However, this finding is tempered by the fact that as ELL students progress in the education pipeline in Texas, they are more likely to attend low performing middle schools and high schools (results not shown). Furthermore, ELLs enrolled in secondary schools ultimately have the highest dropout rates and lowest tests scores and graduation rates in Texas (Vasquez Heilig, 2011a). Surprisingly, after almost two decades of Texas-style accountability, the overall finding that segregation by SES and race and ethnicity is still highly significant for predicting whether schools will be low performing relative to high performing suggests that high-stakes testing and accountability as systemic reforms have still not delivered as a cure-all in Texas.

Footnotes

Authors’ Note

Both authors contributed equally to the content of this article.

Declaration of Conflicting Interests

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

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Julian Vasquez Heilig is an award-winning researcher and teacher. He is currently an associate professor of Educational Policy and Planning and African and African Diaspora Studies (by courtesy) at the University of Texas at Austin. His primary work considers equity and social justice in relation to various educational policies. He blogs at Cloaking Inequity.

Jennifer Jellison Holme is an assistant professor of educational policy and planning in the Department of Educational Administration at the University of Texas at Austin. Her research examines the politics and implementation of educational policy, with a particular interest in the relationship among school reform, equity, and diversity in schools.

References

Arias

(2007). School desegregation, linguistic segregation and access to English for Latino students. Journal of Educational Controversy, 2(1). Retrieved from http://www.wce.wwu.edu/Resources/CEP/eJournal/v002n001/a008.shtml

Batalova

McHugh

(2010). Number and growth of students in US schools in need of English instruction. Washington, DC: Migration Policy Institute.

Cisneros v. Corpus Christi ISD, 324 F. Supp. 599 (U.S. Dist. 1970).

Delgado et al. v. Bastrop Independent School District, Civil No. 388 (W.D. Tex, 1948).

Fry

(2009). The rapid growth and changing complexion of suburban public schools. Washington, DC: The Pew Hispanic Center.

Frey

(2011). Melting pot cities and suburbs: Racial and ethnic change in metro America in the 2000s. Washington, DC: The Brookings Institution.

Gandara

Contreras

(2009). The Latino education crisis: The consequences of failed social policies. Cambridge, MA: Harvard University Press.

Gandara

Rumberger

(2009). Immigration, language, and education: How does language policy structure opportunity? Teachers College Record, 111, 750-782.

Gifford

Valdes

(2006). The linguistic isolation of Hispanic students in California’s public schools: The challenge of reintegration. Yearbook of the National Society for the Study of Education, 105(2), 125-154.

10.

Gonzales

(1990). Chicano education in the era of segregation. Philadelphia, PA: The Balch Institute Press.

11.

Iceland

Scopolliti

(2008). Immigrant residential segregation in U.S. Metropolitan areas, 1990-2000. Demography, 45, 79-94.

12.

Independent School District et al. v. Salvatierra et al., 33 S.W. 2d 790 (Tex. App. 1930).

13.

Keyes v. School Dist. No. 1, 413 U.S. 189 (U.S. S.Ct. 1973).

14.

Gandara

Orfield

(2010). A return to the “Mexican room”: The segregation of Arizona’s English learners. Los Angeles: UCLA Civil Rights Project/Proyecto Derechos Civiles.

15.

Mickleson

R. A.

(2001). Subverting Swann: First- and second-generation segregation in the Charlotte-Mecklenberg schools. American Educational Research Journal, 38, 215-252.

16.

National Center for Education Statistics. (2010, July). Status and trends in the education of racial and ethnic groups. Washington, DC: Author.

17.

National Center for Education Statistics. (2011). The condition of education 2011. Washington, DC: Author.

18.

Orfield

(2009). Reviving the goal of an integrated society: A 21st century challenge. Los Angeles: UCLA Civil Rights Project/Proyecto Derechos Civiles.

19.

Reardon

Yun

J. T.

(2001). Suburban racial change and suburban school segregation, 1987-1995. Sociology of Education, 74(2), 79-101.

20.

Salinas

(1971). Mexican-Americans and the desegregation of schools in the Southwest. Riverside: University of California at Riverside, Western Regional School Desegregation Projects.

21.

Suarez-Orozco

Todorva

(2008). Learning in a New Land. Cambridge, MA: Harvard University Press.

22.

Tefera

Frankenberg

Siegel-Hawley

Chirichingo

(2011). Integrating suburban schools: How to benefit from growing diversity and avoid segregation. Los Angeles: UCLA Civil Rights Project/Proyecto Derechos Civiles.

23.

Texas Education Agency. (2012). State AEIS multi-year history report. Retrieved from http://ritter.tea.state.tx.us/perfreport/aeis/hist/state.html

24.

Valencia

(2008). Chicano students and the courts: The Mexican American legal struggle for educational equity. New York: New York University Press.

25.

Valencia

(1991). Chicano school failure and success. New York, NY: Falmer.

26.

Westminster School Dist. of Orange County et al. v. Mendez et al., 161 F.2d 744 (U.S. App. 1947).

27.

Wilson

S. H.

(2003). Brown over “other White”: Mexican Americans’ legal arguments and litigation strategy in school desegregation lawsuits. Law and History Review, 21(1), Retrieved from http://www.historycooperative.org/journals/lhr/21.1/forum_wilson.html

28.

Vasquez Heilig

(2011a, December 9). What and why: Secondary education and ELLs in an environment of Texas-style accountability (Keynote Speaker, Lecture conducted from Washington, DC). Washington, DC: U.S. Department of Education, Office of English Language Acquisition.

29.

Vasquez Heilig

(2011b). Understanding the interaction between high-stakes graduation tests and English language learners. Teachers College Record, 113. Retrieved from http://ows.edb.utexas.edu/sites/default/files/users/jvh/Vasquez%20Heilig%202011%20ELs%20and%20Exit%20testing%20proof.pdf

30.

Vasquez Heilig

Dietz

Volonnino

(2011). From Jim Crow to the top 10% plan: A historical analysis of Latina/o access to a selective flagship university. Enrollment Management Journal: Student Access, Finance, and Success in Higher Education, 5, 83-109.

31.

Vasquez Heilig

Rodriguez

Somers

(2011). Immigrant DREAMs: The Texas 10% admissions plan, ELL student college choice and academic success. Journal of Latinos and Education, 10(2), 106-126.