An R package for fetching and processing Texas school enrollment, graduation, and assessment data from the Texas Education Agency (TEA). Texas public schools enroll 5.5 million students across 1,200+ districts – the second-largest school system in the United States. One function call pulls it all into R.
This package is part of the njschooldata family of packages providing consistent access to state education data.
Documentation: https://almartin82.github.io/txschooldata/
Installation
R
# install.packages("remotes")
remotes::install_github("almartin82/txschooldata")What can you find with txschooldata?
library(txschooldata)
library(dplyr)
library(tidyr)
library(ggplot2)
library(scales)
# Fetch 5 years of enrollment data
enr <- fetch_enr_multi(2020:2024, use_cache = TRUE)Here are fifteen narratives hiding in the numbers.
1. COVID erased a decade of growth in one year
state_trend <- enr %>%
filter(is_state, subgroup == "total_enrollment", grade_level == "TOTAL") %>%
select(end_year, n_students) %>%
arrange(end_year) %>%
mutate(
change = n_students - lag(n_students),
pct_change = round(change / lag(n_students) * 100, 2)
)
stopifnot(nrow(state_trend) > 0)
print(state_trend)
#> end_year n_students change pct_change
#> 1 2020 5479173 NA NA
#> 2 2021 5359040 -120133 -2.19
#> 3 2022 5402928 43888 0.82
#> 4 2023 5504150 101222 1.87
#> 5 2024 5517464 13314 0.24-120,133 students vanished between 2020 and 2021 – equivalent to the entire enrollment of El Paso ISD.
2. One in four students is now an English learner
lep_trend <- enr %>%
filter(is_state, subgroup == "lep", grade_level == "TOTAL") %>%
select(end_year, n_students, pct) %>%
mutate(pct_display = round(pct * 100, 1))
stopifnot(nrow(lep_trend) > 0)
print(lep_trend %>% select(end_year, n_students, pct_display))
#> end_year n_students pct_display
#> 1 2020 1112674 20.3
#> 2 2021 1108207 20.7
#> 3 2022 1171661 21.7
#> 4 2023 1269408 23.1
#> 5 2024 1344804 24.4
From 20.3% to 24.4% in five years. Schools need more ESL teachers than ever.
3. Coppell ISD is Texas’s first Asian-majority district
# Districts with highest Asian percentage
asian_top <- enr %>%
filter(is_district, subgroup == "asian", grade_level == "TOTAL", end_year == 2024) %>%
inner_join(
enr %>% filter(is_district, subgroup == "total_enrollment",
grade_level == "TOTAL", end_year == 2024) %>%
select(district_id, total = n_students),
by = "district_id"
) %>%
filter(total >= 10000) %>%
arrange(desc(pct)) %>%
select(district_name, total, n_students, pct) %>%
mutate(pct = round(pct * 100, 1)) %>%
head(10)
stopifnot(nrow(asian_top) > 0)
print(asian_top)
#> district_name total n_students pct
#> 1 COPPELL ISD 13394 7591 56.7
#> 2 FRISCO ISD 66551 28349 42.6
#> 3 PROSPER ISD 28394 8312 29.3
#> 4 ALLEN ISD 21319 6201 29.1
#> 5 FORT BEND ISD 80034 22080 27.6
#> 6 PLANO ISD 47753 11207 23.5
#> 7 ROUND ROCK ISD 46042 10126 22.0
#> 8 KATY ISD 94589 16311 17.2
#> 9 WYLIE ISD 19166 3227 16.8
#> 10 LEWISVILLE ISD 48356 8123 16.856.7% Asian. Frisco ISD jumped from 33.6% to 42.6% in just three years.
4. Fort Worth ISD lost 14% of its students
# Calculate 2020-2024 changes using district_id (2020 data has NA names)
d2020 <- enr %>%
filter(is_district, subgroup == "total_enrollment",
grade_level == "TOTAL", end_year == 2020) %>%
select(district_id, n_2020 = n_students)
d2024 <- enr %>%
filter(is_district, subgroup == "total_enrollment",
grade_level == "TOTAL", end_year == 2024) %>%
select(district_id, district_name, n_2024 = n_students)
losses <- d2020 %>%
inner_join(d2024, by = "district_id") %>%
mutate(
change = n_2024 - n_2020,
pct_change = round((change / n_2020) * 100, 1)
)
# Largest percentage losses among big districts
top_losses <- losses %>%
filter(n_2020 >= 10000) %>%
arrange(pct_change) %>%
select(district_name, n_2020, n_2024, change, pct_change) %>%
head(10)
stopifnot(nrow(top_losses) > 0)
print(top_losses)
#> district_name n_2020 n_2024 change pct_change
#> 1 FORT WORTH ISD 82704 70903 -11801 -14.3
#> 2 ALDINE ISD 67130 57737 -9393 -14.0
#> 3 BROWNSVILLE ISD 42989 37032 -5957 -13.9
#> 4 HARLANDALE ISD 13654 11781 -1873 -13.7
#> 5 YSLETA ISD 40404 34875 -5529 -13.7
#> 6 LAREDO ISD 23665 20557 -3108 -13.1
#> 7 ALIEF ISD 45281 39451 -5830 -12.9
#> 8 HOUSTON ISD 209309 183603 -25706 -12.3
#> 9 LA JOYA ISD 27276 23995 -3281 -12.0
#> 10 ABILENE ISD 16456 14482 -1974 -12.0Urban districts are bleeding students to suburbs and charters.
5. IDEA Public Schools grew 55% in five years
# Use district_id join to include 2020 (which has NA district_name)
idea_trend <- losses %>%
filter(grepl("IDEA", district_name))
stopifnot(nrow(idea_trend) > 0)
print(idea_trend %>% select(district_name, n_2020, n_2024, change, pct_change))
#> district_name n_2020 n_2024 change pct_change
#> 1 IDEA PUBLIC SCHOOLS 49480 76819 27339 55.3
# Year-by-year for 2021-2024 (name filter works for these years)
idea_yearly <- enr %>%
filter(district_name == "IDEA PUBLIC SCHOOLS", is_district,
subgroup == "total_enrollment", grade_level == "TOTAL") %>%
select(end_year, n_students) %>%
arrange(end_year)
stopifnot(nrow(idea_yearly) > 0)
print(idea_yearly)
#> end_year n_students
#> 1 2021 62158
#> 2 2022 67988
#> 3 2023 74217
#> 4 2024 76819From 49,480 to 76,819 students (+55.3%). Charter networks are reshaping Texas education.
6. Fort Bend ISD has no racial majority
fb_demo <- enr %>%
filter(district_name == "FORT BEND ISD", is_district, grade_level == "TOTAL",
subgroup %in% c("white", "black", "hispanic", "asian")) %>%
select(end_year, subgroup, pct) %>%
mutate(pct = round(pct * 100, 1)) %>%
pivot_wider(names_from = subgroup, values_from = pct)
stopifnot(nrow(fb_demo) > 0)
print(fb_demo)
#> # A tibble: 4 × 5
#> end_year white black hispanic asian
#> <int> <dbl> <dbl> <dbl> <dbl>
#> 1 2021 14.8 27.5 26.4 27.3
#> 2 2022 14.7 27.8 26.6 26.7
#> 3 2023 13.8 27.8 26.7 27.3
#> 4 2024 13.2 27.8 26.7 27.6No group exceeds 28%. One of the most diverse large districts in America.
7. Kindergarten enrollment dropped 5.8%
grade_trend <- enr %>%
filter(is_state, subgroup == "total_enrollment",
grade_level %in% c("PK", "K", "01", "05", "09", "12")) %>%
select(end_year, grade_level, n_students) %>%
pivot_wider(names_from = end_year, values_from = n_students) %>%
mutate(
change = `2024` - `2020`,
pct_change = round(change / `2020` * 100, 1)
)
stopifnot(nrow(grade_trend) > 0)
print(grade_trend)
#> # A tibble: 6 × 8
#> grade_level `2020` `2021` `2022` `2023` `2024` change pct_change
#> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 PK 248413 196560 222767 243493 247979 -434 -0.2
#> 2 K 383585 360865 370054 367180 361329 -22256 -5.8
#> 3 01 391175 380973 384494 399048 385096 -6079 -1.6
#> 4 05 417272 395436 387945 395111 399200 -18072 -4.3
#> 5 09 448929 436396 475437 477875 472595 23666 5.3
#> 6 12 352258 362888 360056 364317 365788 13530 3.8-22,256 kindergartners since 2020. The pipeline is narrowing.
8. 62% of students are economically disadvantaged
econ_trend <- enr %>%
filter(is_state, subgroup == "econ_disadv", grade_level == "TOTAL") %>%
select(end_year, n_students, pct) %>%
mutate(pct = round(pct * 100, 1))
stopifnot(nrow(econ_trend) > 0)
print(econ_trend)
#> end_year n_students pct
#> 1 2020 3303974 60.3
#> 2 2021 3229178 60.3
#> 3 2022 3278452 60.7
#> 4 2023 3415987 62.1
#> 5 2024 3434955 62.3
Up from 60.3%. Nearly two-thirds of Texas students qualify.
9. White students dropped below 25%
demo_shift <- enr %>%
filter(is_state, grade_level == "TOTAL",
subgroup %in% c("white", "black", "hispanic", "asian", "multiracial")) %>%
select(end_year, subgroup, pct) %>%
mutate(pct = round(pct * 100, 1)) %>%
pivot_wider(names_from = subgroup, values_from = pct)
stopifnot(nrow(demo_shift) > 0)
print(demo_shift)
#> # A tibble: 5 × 6
#> end_year white black hispanic asian multiracial
#> <int> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 2020 27 12.6 52.8 4.6 2.5
#> 2 2021 26.5 12.7 52.9 4.7 2.7
#> 3 2022 26.3 12.8 52.8 4.8 2.9
#> 4 2023 25.6 12.8 53 5.1 3
#> 5 2024 25 12.8 53.2 5.4 3.1
Hispanic students are now 53.2% of enrollment.
10. 439 districts now have Hispanic majorities
hisp_majority <- enr %>%
filter(is_district, subgroup == "hispanic", grade_level == "TOTAL") %>%
mutate(majority = pct > 0.5) %>%
group_by(end_year) %>%
summarize(
total_districts = n(),
hispanic_majority = sum(majority),
pct_majority = round(hispanic_majority / total_districts * 100, 1),
.groups = "drop"
)
stopifnot(nrow(hisp_majority) > 0)
print(hisp_majority)
#> # A tibble: 5 × 4
#> end_year total_districts hispanic_majority pct_majority
#> <int> <int> <int> <dbl>
#> 1 2020 1202 419 34.9
#> 2 2021 1204 428 35.5
#> 3 2022 1207 433 35.9
#> 4 2023 1209 438 36.2
#> 5 2024 1207 439 36.4Up from 419 in 2020. Now 36.4% of all Texas school districts.
11. Houston ISD vs Dallas ISD: two giants, two trajectories
# Use district_id to include 2020 (which has NA district_name)
big_two_ids <- c("101912", "057905") # Houston ISD, Dallas ISD
big_two <- enr %>%
filter(is_district, subgroup == "total_enrollment", grade_level == "TOTAL",
district_id %in% big_two_ids) %>%
mutate(district_label = case_when(
district_id == "101912" ~ "HOUSTON ISD",
district_id == "057905" ~ "DALLAS ISD"
)) %>%
select(end_year, district_label, n_students) %>%
pivot_wider(names_from = district_label, values_from = n_students) %>%
arrange(end_year) %>%
mutate(
houston_change = `HOUSTON ISD` - lag(`HOUSTON ISD`),
dallas_change = `DALLAS ISD` - lag(`DALLAS ISD`)
)
stopifnot(nrow(big_two) > 0)
print(big_two)
#> # A tibble: 5 × 5
#> end_year `DALLAS ISD` `HOUSTON ISD` houston_change dallas_change
#> <int> <dbl> <dbl> <dbl> <dbl>
#> 1 2020 153784 209309 NA NA
#> 2 2021 145105 196550 -12759 -8679
#> 3 2022 143430 193727 -2823 -1675
#> 4 2023 141042 189290 -4437 -2388
#> 5 2024 139096 183603 -5687 -1946
Houston lost 25,706 students (-12.3%) vs Dallas’s 14,688 (-9.5%) between 2020 and 2024.
12. Nearly 1 in 7 students receives special education
sped_2024 <- enr %>%
filter(is_state, subgroup == "special_ed", grade_level == "TOTAL", end_year == 2024) %>%
select(end_year, n_students, pct) %>%
mutate(pct_display = round(pct * 100, 1))
stopifnot(nrow(sped_2024) > 0)
print(sped_2024 %>% select(end_year, n_students, pct_display))
#> end_year n_students pct_display
#> 1 2024 764858 13.9764,858 students (13.9%) receive special education services in 2024.
13. Suburban boomtowns are reshaping Texas education
# Top growing large districts (using district_id join for accurate 2020-2024 comparison)
suburban_growth <- losses %>%
filter(n_2020 >= 5000) %>%
arrange(desc(pct_change)) %>%
select(district_name, n_2020, n_2024, change, pct_change) %>%
head(10)
stopifnot(nrow(suburban_growth) > 0)
print(suburban_growth)
#> district_name n_2020 n_2024 change pct_change
#> 1 HALLSVILLE ISD 11452 21266 9814 85.7
#> 2 PROSPER ISD 16789 28394 11605 69.1
#> 3 PRINCETON ISD 5414 8671 3257 60.2
#> 4 CLEVELAND ISD 7559 11945 4386 58.0
#> 5 IDEA PUBLIC SCHOOLS 49480 76819 27339 55.3
#> 6 MEDINA VALLEY ISD 5847 8656 2809 48.0
#> 7 PREMIER HIGH SCHOOLS 5345 7819 2474 46.3
#> 8 YES PREP PUBLIC SCHOOLS INC 12074 17622 5548 45.9
#> 9 FORNEY ISD 11944 16962 5018 42.0
#> 10 ROYSE CITY ISD 6585 9338 2753 41.8Hallsville ISD nearly doubled (+85.7%), Prosper ISD grew 69.1%. While urban cores lose students, suburban boomtowns absorb the growth.
14. Pre-K enrollment still hasn’t recovered from COVID
pk_trend <- enr %>%
filter(is_state, subgroup == "total_enrollment", grade_level == "PK") %>%
select(end_year, n_students) %>%
arrange(end_year) %>%
mutate(
change = n_students - lag(n_students),
pct_change = round(change / lag(n_students) * 100, 1),
vs_2020 = n_students - first(n_students)
)
stopifnot(nrow(pk_trend) > 0)
print(pk_trend)
#> end_year n_students change pct_change vs_2020
#> 1 2020 248413 NA NA 0
#> 2 2021 196560 -51853 -20.9 -51853
#> 3 2022 222767 26207 13.3 -25646
#> 4 2023 243493 20726 9.3 -4920
#> 5 2024 247979 4486 1.8 -434
Pre-K cratered 21% during COVID. By 2024 it has clawed back to 247,979 – still 434 students short of 2020 levels.
15. Multiracial students are the fastest-growing demographic
multi_trend <- enr %>%
filter(is_state, subgroup == "multiracial", grade_level == "TOTAL") %>%
select(end_year, n_students, pct) %>%
mutate(pct_display = round(pct * 100, 1))
stopifnot(nrow(multi_trend) > 0)
print(multi_trend %>% select(end_year, n_students, pct_display))
#> end_year n_students pct_display
#> 1 2020 138434 2.5
#> 2 2021 143368 2.7
#> 3 2022 155887 2.9
#> 4 2023 166128 3.0
#> 5 2024 173425 3.1
From 138,434 to 173,425 – a 25% increase in raw numbers. The fastest-growing demographic group in Texas.
Graduation rates
Track four-year longitudinal graduation rates for the Class of 2003-2024:
library(txschooldata)
# Get Class of 2024 graduation rates
grad <- fetch_grad(2024, use_cache = TRUE)
# State graduation rate by subgroup
grad |>
filter(type == "State") |>
select(subgroup, grad_rate) |>
arrange(desc(grad_rate))
#> subgroup grad_rate
#> 1 asian 95.59655
#> 2 female 95.40719
#> 3 white 94.74369
#> 4 all_students 94.39027
#> 5 hispanic 93.69867
#> 6 male 93.38859
#> 7 econ_disadv 93.24518
#> 8 multiracial 93.09000
#> 9 at_risk 92.12205
#> 10 lep 91.25872
#> 11 special_ed 87.80512
#> 12 black NA
#> 13 native_american NAData available for all students plus race/ethnicity, gender, and special populations (economically disadvantaged, special education, LEP, at-risk).
Data availability
This package pulls from three TEA reporting systems:
| System | Years | Notes |
|---|---|---|
| AEIS CGI | 1997-2002 | Older AEIS format via CGI endpoint |
| AEIS SAS | 2003-2012 | Academic Excellence Indicator System |
| TAPR | 2013-2025 | Texas Academic Performance Reports |
All data is fetched directly from TEA servers – no manual downloads required.
What’s included
Enrollment data: - Levels: State, district (~1,200), and campus (~9,000) - Demographics: White, Black, Hispanic, Asian, American Indian, Pacific Islander, Two or More Races - Special populations: Economically disadvantaged, LEP/English learners, Special education, At-risk, Gifted - Grade levels: Early Education (EE) through Grade 12, plus totals
Graduation rate data: - Years: Class of 2003-2024 (four-year longitudinal rates) - Levels: State, district, campus - Subgroups: All students, race/ethnicity, gender, special populations - Note: Uses federal calculation for consistency across years
Data notes
-
IDs: District IDs are 6 digits (e.g.,
101912for Houston ISD). Campus IDs are 9 digits (district + 3-digit campus number). -
Tidy format: By default,
fetch_enr()returns long/tidy data withsubgroup,grade_level, andn_studentscolumns. Usetidy = FALSEfor wide format. -
Percentages: The
pctcolumn is a proportion (0-1), not a percentage. Multiply by 100 for display. -
Caching: Data is cached locally after first download. Use
use_cache = FALSEto force refresh. - Suppression: TEA applies data masking to protect student privacy for small subgroups.
Caveats
- Asian/Pacific Islander: Pre-2011 data combines Asian and Pacific Islander into a single “asian” category. Separate Pacific Islander data only available from 2011 onward (federal reporting change).
- Two or More Races: Only available from 2011 onward (federal reporting change)
- Column names: Standardized across years, but underlying TEA variable names differ between systems
- Historical comparisons: Definition of “economically disadvantaged” and other categories may shift over time
Part of the State Schooldata Project
This package is part of the njschooldata family of packages providing consistent access to state education data.
A simple, consistent interface for accessing state-published school data in Python and R.
All 50 state packages: github.com/almartin82