Graphic design by Michelle Jia; image by Alexander Monich.
I am a professor at the University of Texas who happens to be Latino with a son in Travis Heights Elementary (hereafter THES, Austin, Texas).
In 2017, I enrolled my 9-year old son in kindergarten at THES and without my knowledge, he (who was born in Austin and is a native English speaker), was enrolled as an English Learner. Children born in the USA whose households have more than just English as the language of everyday communication are by default considered inherently deficient all over the USA.
Without my knowing, I was condemning my son to take every year an exam to prove that he was making improvements in his “proficiency” in English. How was this possible? How did the son of two university professors wind up as demonstrably “non-proficient” in English in kindergarten? Who decided it? What kind of exam identified him as lacking in English skills? Did any other English native speaker, brought up in an openly acknowledged bilingual or plurilingual household, also get classified as deficient? Was he the only one? Does this classification have long-term consequences for the child?
Shouldn’t poverty be the main criterion to single out children for remedial help, not the parents’ language? West Virginia is the Whitest and the poorest of states in the USA and has the lowest number of ELs (0.8 %).
This essay is ultimately about my struggles to understand a system that seems so obvious that it has no need of justifying itself. No data is available or collected. The more I looked at this system of EL the more I understood that one needs to work as hard at creating ignorance about a topic as one needs to create knowledge. Silence demands work.
The other lesson I learned is that truth is the product of consensus. Something is good and truthful only if there is an upper middle class White liberal consensus that dignifies it with such status. Truth means a tangible class benefit that adds privilege to a group.
On evidence and ignorance
I have yet to find the criteria given to teachers to classify a native English speaker child as “a learner.” Teachers cannot rely on reading and writing to identify those falling behind (yet, given the age). I am therefore only left with my own imagination to recreate how the actual test of my son in kindergarten worked: a conversation between a cowed little kid and a teacher who desperately wanted to expand the pool of bilinguals to maintain dual immersion schools.
The exam is the main mechanism districts have to funnel federal funding programs into upper-middle-class Anglo immersion schools.
Federal funding, however, does nothing but reinforce a longstanding pattern in US causal accounts of poverty. Language and culture cause poverty. Within this logic, poverty originates in the very language of Hispanic “minorities.” Children are stunted by their encounter with the language of magic and incompetence, namely, Spanish. Poverty ensues. The two things reinforce one another.
Once the logic of the exam makes sense to an outside observer then one begins to wonder how it is done. Do teachers consider dialects, accents, and Spanglish as a manifestation of English deficiencies? How many kids in bilingual households pass the kindergarten test? We have no statistics available nationally, let alone in Texas, on this exam. As far I know there is no paper record of any meaningful evaluation, just a summary conclusion gets recorded: “not proficient.” Again, as far as I know, there is no record of whether any of those who have to take the kindergarten test pass it. My strong suspicion is that everyone targeted for evaluation is automatically declared deficient.
I did not know any of this until my son was in second grade when he was separated from his class to take a TELPAS exam. He had been identified two years prior as not proficient without his parents knowing.
My son asked me why he was special. He was upset. The purpose of the exam was to test kids to be removed from the list. No kid can be removed until grade 3 because the law says that students need to demonstrate improvement two years in a row.
The statistics of Texas show that removal happens rarely. It is therefore a tracking mechanism. The best schools in the state and in the city of Austin are the ones who have the lowest number of English Learners. EL proficiency is a marker of school status.
Upper middle-class consensus as truth
I have reached out with my findings in the report (see below) to many parents’ organizations, to the school, and to the school district. I have repeatedly been told: that all I say about tracking and statistics of removal are wrong. That there is no evidence that EL status determines access to the best schools. That there are plenty of studies by the school district demonstrating that by grade 6 most kids classified as EL are removed from this status. I have been twice asked to communicate with bilingual learning advocates who think that dual immersion programs ought to be encouraged to the benefit of everyone: Latinos and Anglo kids.
Every time I have requested these parties to guide me to their statistics, I have been greeted with silence. In the case of the Austin School District, there are all sorts of numbers on kids’ performance on proficiency exams. Yet the statistics on removal are unavailable even after a formal request. In Texas, removal is extraordinarily cumbersome and it takes five separate steps to remove an EL student from “the list” (see below).
The more I inquire, the more I realize that a consensus was formed a few years ago that bilingual education for liberal, White, upper middle-class urbanites is a good thing (it is). It is only when one begins to ask this obvious question (but at whose expense?) that things become opaque. Clarity and transparency disappear.
In the following, I rely on public documents available through the research sites of the Texas Department of Education and the Austin Independent School District (AISD). I have done my best to follow the data as presented online. A lot of the data is scattered and fragmented and percentages have to be reconstructed indirectly.
My summary is simple: Were this evidence made available, public support for dual immersion programs would plummet among the population that is most affected, namely, parents of US-born kids whose households are bilingual Hispanic. Most of these parents assume that participating with Anglo kids in the same classroom, following the same curricula, getting an education with the same teachers is greatly beneficial for their kids. No parent would deny that these are forms of equity, diversity, and inclusion. The logic is seemingly disarming: Hispanic kids are no longer considered ESL (English as a Second Language Students). They are no longer isolated in classrooms, away from their White peers.
The data, however, tells a very different narrative than the official spiel. There is nothing remarkable about this, of course. What is remarkable is that this gap between reality and rhetoric has rendered rhetoric into reality and reality into silence. The dismal statistics of EL list removal and tracking are invisible. Postmodern truth, just as modern truth, becomes nothing else than the consensus of one group, whose class, racial and ethnic privileges render the interests of any others invisible. Real reality surrenders daily to rhetorical consensus.
“English Learners” are USA-born kids whose families happen to be Hispanic and check the bilingual box in kindergarten application. They are not ESL students; ESLs are recent immigrants.
2. Who is EL and the racial and ethnic correlation
This system was created outside the control and knowledge of parents who have no saying in the classification of their kids. Classification had nothing to do whatsoever with command of the English language or lack thereof; it has everything to do with last names and the decision to check the box for a “bilingual” household. White kids as a default are not classified as ELs, regardless of whether they are proficient or not.
Nationally EL is an exam that tests ethnicity, not poverty. One of the poorest White states in the nation, West Virginia, has only 0.8 % EL whereas California, the richest, has 19%.
3. The problem with the K exam
To test English proficiency, teachers give an exam in kindergarten based entirely on listening and speaking, not reading and writing. Curiously TELPAS exams that include writing and reading show that EL students do much better. The lowest scores happen to be on “speaking.” How can one evaluate “speaking’ without openly discriminating against ethnic dialects and Spanglish?
4. Lack of transparency on data of EL removal
Data on removal in the EL list can only be found in the Biannual Reports by the Texas Education Agency (TEA) to the House of Representatives since 2011 when TELPAS got instituted and replaced TASK.
The Reports, I have found out, are the ONLY document available in the education research sites of both AISD (Austin) and TEA (Texas) that has data on “former ELs.” No other document in the state offers information on removal rates. The Report does it indirectly, so I had to work hard to disaggregate this data.
5. Ad hoc removal
For students to be removed it requires a committee of 6 individuals to quiz the kid. This is in addition to “exceptional” scores in TELPAS and STARR to be determined by the committee ad hoc. There are no objective criteria anywhere that regulate removal. Removal is at the pleasure of school administrations that get extra federal funding by having larger numbers of ELs. There is no transparency or accountability.
6. AISD misleading information
AISD offers statistics on TELPAS results per grade (not associated with race and ethnicity) and percentages for beginning, intermediate, advanced, and advanced high TELPAS results. Yet AISD has no statistics on percentages of removal of ELs. I have found that the AISD statistics of TELPAS are deeply misleading, for many reasons.
1) Because TELPAS is only one criterion for removal. There are three other criteria that get excluded from the AISD statistics: STARR results, the Subjective Letter of Evaluation by the student’s teacher, and the evaluation of the Learning Assessment Committee (LAC).
2) AISD misleads audiences of their research analysis (people like me) by avoiding one pesky piece of regulation: to pass, students need to prove advancement two years in a row in TELPAS scores. It is not enough to test well for one year to be removed, hence there is no possible removal from K, 1, and 2 grades. Statistics of removal begin only in 3 grade.
3) TELPAS statistics misleadingly suggest that the four different scores of TELPAS (readings, speaking, listening, and writing) are separate, so a grade with 46% advanced high in listening means nothing. What counts is the TELPAS composite. This leads me to surprising findings: ELs scores are much higher for TELPAS writing and reading than TELPAS speaking. TELPAS speaking is a veritable machine of policing proper speech and clamping down into dialects and Spanglish.
7. Data available (1/2 to 1 kid out of 10 are removed in 12 years)
The only reliable data on removal, therefore, are statewide, and not specific to AISD, which itself is deeply troubling because they seem to suggest removal rates that do not really exist.
Given all these caveats, I have found the scores of removal Texas-wide, eye-opening. In the report of 2016 1,008,000. kids in the State of Texas got into ELS (separate from ESL that addressed recent immigrants), and only 68,000 got removed, for a shocking 6.8% or 6 kids of every 100 removed in 12 years.
In the biannual report of 2018, there is data on 120k removals out of a total population of 1,010,000. In 2018 twice as many were removed, but it is not entirely clear why (partly perhaps because this report has scattered information on former ELS for grades 11 and 12, missing in the previous one). But still, the statistics are shocking. An average of 1 in 10 gets out in 12 years. And again we have no idea of the race and ethnicity of those measured by TEA.
8. What is the destination of federal funding for EL?
We desperately need an honest and serious study of AISD for the past 10 years on rates of removal from EL status. We also need an evaluation of what is the destination of the federal money earmarked to benefit those ELS students. My impression is that that is not money spent necessarily on ELs students; it is used widely to cover expenses in many other areas.
9. Statistics of EL for AISD
There is evidence that EL testing works as a tracking system to keep Hispanics essentially away from the best White/Asian schools in Austin, which happen to be the best in the city and, in one case, the nation (Liberal Arts and Science Academy or LASA).
High schools/Early Colleges are predominantly Hispanic up to 60–70% and Black (Lyndon B. Johnson High School especially).
Everything that feeds the best high schools is predominantly White (and Asian in one case in particular) from 57% in the best HS of the city and 48% White and 28% Asian in the best magnet school of the state (LASA).
LASA, a white/Asian HS, has percentages of 1% EL. The other three (Bowie, Anderson, and McCalloun) oscillate between 3–7 % ELs.
10. How did statistics become tracking in Travis Heights?
THES has the same percentage of ELs as does the rest of high schools Early Colleges (the Hispanic Black high schools of the city of Austin): about 39 to 37%.
This is the very definition of tracking: Hispanic kids in THES move to a magnet school like Lively Middle School that is predominantly Hispanic (67%) with 20% White, low STARR scores, and EL 31%.
Lively, in turn, feeds into THHS that is an early college with even worse scores than Lively, more Hispanics (83%, and worse percentages of EL (40%).
In short, the population of THES is statistically divided clearly into two populations with two very different futures: White and Hispanic/Black.
Hispanic kids enter in great numbers into Lively (a magnet) only to end in one of the worst schools in the city TH early college.
The White population of THES, on the other hand, moves either to Kealing (with only 10% of ELs and 36 % white population) then LASA (with 1.5 % ELs and among the best scores in the nation) or to private schools.
11. The best two Magnet Schools of AISD
Per statistics provided in 2019–20, LASA reported 6 EL students (4 of which whom were foreigners/ESL) out of ca. 1,300 total. The Hispanic population of LASA is only 1/3 (20%) of the high school state average (56%). Nearly no Hispanic kid who is not removed from the EL state list can get to the best public school in the city.
This is called tracking, like in the old days. I am surprised there has not been a class action suit for structural racial discrimination at AISD.
"English Learner"( EL) is a perverse category, from which it is almost impossible to escape (1 % leaves the category per year in public schools in Texas). It organizes all Title III federal funding. “English learners” are found by teachers in kindergarten classrooms, without the parents’ consent. There is no formal testing as far as I know. All it takes is for parents to have “foreign” last names and to check “bilingual household” in applications to the school.
If the kid is a native speaker, she becomes EL, not ESL (English as a second language), automatically considered cognitively “deficient.”
The kid gets to be tested 11 years in a row and cannot leave unless she fulfills 5 criteria in Texas. Keeping kids in perpetual EL status allows schools to keep Title III federal funding.
EL has become in Austin the most reliable marker of ethnicity and vocational educational tracking. Four vocational high-school colleges for Hispanic kids in Austin are the embodiment of this policy today: Separate but equal. A return to Booker T Washington industrial education for Blacks at Black schools and colleges.
There is no transparency on any of the data offered by the Texas Department of Education and AISD