Between 2019 and 2024, the reading scores of American fourth graders fell almost everywhere. The National Assessment of Educational Progress, the federal test known as NAEP and often called the Nation’s Report Card, is given to a sample of students in every state. Of the 52 jurisdictions it covers, 50 scored lower in 2024 than they did in 2019. Louisiana was the only state to post a meaningful gain. Its fourth-grade reading score rose 6.1 points while the national public average fell 5.2.
Louisiana is the reason the national average is a bad number to run an agency on. It tells you the country lost ground. It does not tell you that Louisiana went from about 9.5 points below the national average in 2019 to nearly 2 points above it in 2024, or that Mississippi has stayed above the national line since 2022, or that your own state may have slipped further than the headline number suggests. The average flattens all of that. A state agency has to work in the variance underneath it, and know where its own state sits.
State officials and several outside researchers credit Louisiana’s turnaround to a reading overhaul: a 2021 law that required its early-grade teachers to complete science-of-reading training, part of a broader “back to the basics” push under the state superintendent. Mississippi’s earlier climb, which took it from near the bottom of the NAEP rankings to the middle of the pack, is usually traced to a 2013 literacy law, though researchers disagree about how much of that gain reflects the law versus a third-grade retention rule that holds some of the lowest readers back before the fourth-grade test. The causes are still being argued over. The movement is not: two states went up while the rest of the country went down.
This piece walks down through the levels a state agency actually benchmarks at: the nation, the state, the district, and the individual campus. The campus-level examples are Texas and Illinois because, as of July 2026, those are the two state agencies whose files we’ve loaded so far. Every number below came from a query against public data, and any agency could run those queries itself once the files are cleaned and sitting in one place. That cleanup is the part OpenData has already done.
One altitude down: how states stack up
Put every state’s fourth-grade reading score on a map and scroll through what changed between 2019 and 2024. The map opens on score levels, which hold few surprises, then switches to five-year change. Louisiana gained 6.1 points and Alabama gained 1.5. Those are the only two gainers. Everyone else lost ground, and the losses run deep: Maine dropped 11.6 points, the worst in the country, with Oregon and Nebraska close behind at 10.7 each. The national public average fell 5.2.
The picture a commissioner already knows
The picture a commissioner already knows
This is fourth-grade reading in 2024, one shade per state. Massachusetts sits at the top at 224.8, New Mexico at the bottom at 200.6, and every other state falls somewhere in the 24-point band between them.
Now color by what changed
Recolor the same map by how each state moved from 2019 to 2024, and it turns almost entirely one color. The national public average fell 5.2 points, and of the 52 jurisdictions the test covers, 50 scored lower in 2024 than they did in 2019. Two states break the pattern, and they sit together in the South.
Where the map breaks
Louisiana gained 6.1 points, the only meaningful rise in the country. Alabama gained 1.5. Mississippi slipped 0.8 but still finished at 218.5, above the national line of 214.3. Three neighbors held or gained ground while the rest of the map fell.
The same shock, different recoveries
Every state took the pandemic hit. What separates them now is who came back. Louisiana did, and state officials credit a 2021 law that put its K-3 teachers through science-of-reading training.
NAEP grade-4 reading by state, 2024 score and 2019-to-2024 change. The national public average is excluded from the map (it is not a state) but cited in the text.
NAEP is a sampled federal assessment, so sub-point differences are effectively ties. Change is the 2024 average scale score minus the 2019 score for grade-4 reading. The 50 states and DC are mapped; the National public average is excluded from the map because it is not a jurisdiction on the map.
NAEP is a sampled federal assessment, so sub-point differences are effectively ties. Change is the 2024 average scale score minus the 2019 score for grade-4 reading. The 50 states and DC are mapped; the National public average is excluded from the map because it is not a jurisdiction on the map.
The obvious next question is whether money explains the gap, and the data says it mostly does not. Plot each state’s 2023-24 per-pupil spending against its 2024 reading score and the cloud of points has almost no slope: the correlation between the two is 0.07, which works out to spending explaining well under 1% of the variation in scores. Mississippi spent about $12,300 per student and scored 218.5; New York spent $31,900 and scored 214.6. That gap looks larger than it is, because these are raw dollars: salaries, rent, and everything else a district buys cost far more in New York than in Mississippi, so the same dollar buys less schooling there. The pattern survives the caveat, though, because it repeats among states with similar costs. Utah spent the least of any state, about $11,300, and scored 219.3, while Massachusetts scored highest at 224.8 on about $24,400. Group the states into bands of similar spending and every band contains a 13-to-23-point spread in scores, wider than the 5.2-point national decline this whole article is about. Knowing a state’s budget tells you very little about its reading scores on its own.
Read this carefully, because it is easy to overread. The chart does not say money doesn’t matter, and it says nothing at all about poverty, which is missing from the plot and is a stronger predictor of reading scores than any budget line. Mississippi is the poorest state in the country and it still finished above New York. That is a reason to ask what Mississippi changed, not proof that student poverty stopped mattering. One more limit: none of the dollars are adjusted for local costs, for the reason given above. The spending and the scores at least cover the same school year, since NAEP’s 2024 assessment was given during 2023-24. What the chart does say is that at any given spending level, some states get much more reading out of their dollars than others. For an agency, that reframes the question from “do we spend enough?” to “what are the states above the trend doing that we aren’t?” Louisiana, at mid-pack spending and an above-average score, is one of the states worth that call.
Teacher pay is usually the next question. Compare average teacher salaries in 1999-00 and 2021-22 in constant dollars and, after inflation, the average teacher earned less at the end of that stretch in 36 states, roughly two-thirds of the country. Indiana fell the hardest at 22.4%, followed by Michigan and North Carolina. Washington and Massachusetts went the other way, up 19.3% and 14.6%. The national average slipped 4.7%.
That last number needs a caveat: 2021-22 was a year of unusually high inflation, which drags the endpoint down. End the comparison one year earlier and the national figure is roughly flat. The direction holds either way, most states lost ground, but the size of the decline depends on which year you stop on. Anyone quoting the 4.7% should know that. Zoom out further and the two series don’t move together, though they also don’t share a start date: real per-pupil spending has grown about 2.7 times since 1969-70, while NAEP only goes back to 2003, and national fourth-grade reading in 2024 sits slightly below where it started. On the stretch the two do overlap in real dollars, 2010-11 to 2020-21, spending rose 12.6% and reading fell 3.9 points, though that endpoint lands inside the pandemic and carries its learning loss.
The problem with reading the source files
None of the numbers above are secret. They come from files that NCES, the states, and the Census publish for free. Free to download is not the same as ready to use, though. A raw CSV is a wall of comma-separated text until someone opens it in Excel, and it stops being useful the moment it outgrows what Excel can hold, which a 4.7-million-row attendance file does several times over. Past that point the file needs a database and someone who can write a query against it. Neither person is usually the one who needs the answer, so the question goes into a queue: a request filed with whoever owns the data, answered in days if the team is small and the question is easy, in weeks if it isn’t. By the time the number comes back, the board meeting it was for has often already happened.
Take Texas attendance data. The state’s Academic Performance Report publishes chronic absenteeism for every campus, but it does it through codes. Here is what four rows for a single Houston campus actually look like.
To decode CB0CA22R you need TEA’s reference document. The C means it’s a campus-level file, B0 means African American students, CA means chronic absenteeism, 22 means the school year ending in 2022 (so 2021-22, since these codes lag the report year), and R means it’s a rate. A value of -1 isn’t a real number, it’s a mask TEA applies when a student group is too small to report without identifying a child. Getting a single district’s chronic-absenteeism rate out of this means pairing every campus’s numerator with its denominator, dropping the masked cells, and knowing that a full-time virtual campus reports 0% by an accounting rule and will quietly skew any ranking that includes it.
That decoding is already done. Here are the same four rows, served from the platform:
This is not a Texas failing. Every agency publishes files that need a codebook. Illinois packs every school into one row with thousands of columns whose positions move year to year. Florida suffixes its column names with the year, so the same metric is called something different every time it’s published. The cleanup is real work, and it’s work that every researcher, journalist, and agency analyst currently redoes from scratch, each of them writing their own version of the same decoder for the same public file. That is the work OpenData does once, up front, so the next person doesn’t have to.
Drilling down: state and district
The payoff of putting campus-level files in one place is that a commissioner can move from a state number to a single school without switching tools. Start with the statewide picture in Texas, built by aggregating chronic-absenteeism counts up from about 8,500 campuses. Chronic absenteeism means a student missed 10% or more of the school year.
The 2019-20 dip is a counting artifact, not a good year: campuses closed in March 2020 and attendance accounting stopped, so that rate covers only the months schools were open. Treat it as a footnote, not a baseline. The 2023-24 figure at the other end is the most recent Texas has published, since the state’s attendance data lags a year behind its report label. Every point on that chart came from adding up 4.7 million raw campus rows on the fly. An agency usually buys that number once, from a contractor, as a spreadsheet that goes stale the day it arrives.
Because the underlying rows are per campus, the same query drops straight down to the district level. In 2023-24, Houston ISD’s chronic-absenteeism rate was 21.0%, close to the statewide 19.5%. Austin ISD was 27.0%, about 7.5 points above the state. Both numbers came out of one query against the same campus rows.
Illinois shows the other half of what this enables. Its report cards run back 29 years, though not every metric goes back that far: the state only began reporting chronic absence in 2018, having tracked chronic truancy under a different rule until 2017. The two series never overlap, so the chart starts where the current measure does.
One thing this chart cannot do is sit next to the Texas one. Illinois measures chronic absence under its own state rule, which is not the same as the Texas definition, so the two states’ rates should never be dropped onto a single chart and compared directly, which is why they aren’t here. Within Illinois the comparison is fair, and it gets more interesting the closer you look. Take one county. Lake County sits north of Chicago and contains 35 districts with at least 1,000 students, all under the same state rules, the same regional labor market, and the same county health services.
Poverty and absence move together almost in lockstep: the correlation across those 35 districts is 0.80. Lake Forest District 67, where 1.3% of students are low income, posted 7.9% chronic absence. North Chicago District 187, at 91.3% low income, posted 34.4%. This is the variable missing from the spending chart earlier, and inside one county it explains far more than any budget line did. We told that story in full in our Illinois chronic-absence analysis.
The last level: one campus at a time
Everything so far has been an average of something. The last step down is the individual school, and it is where the article’s opening question comes back: not who is absent, but who is reading. Texas publishes both per campus, so the same map can show each in turn.
Start with who the students are
Start with who the students are
Every Austin ISD campus, one dot each, sized by how many students it enrolls. Red dots are schools where most students come from low-income families. They sit east of the center of town. The blue dots run west. Hold that geography in mind.
Now color by who shows up
Same dots, new measure: the share of students who missed 10% or more of the school year. The map barely changes shape. Absence tracks the same east-west line poverty does, which is most of why the two look so closely related.
The red cluster is mostly one kind of school
Eight of the reddest dots are alternative and early-college campuses, and they are a category of their own. Across all twelve such campuses in the district, 5,617 of 9,466 students were chronically absent, a rate of 59.3% against 24.6% everywhere else. These schools enroll students who already stopped attending somewhere else, so a high rate here describes who walked in the door.
The measure that matters
Color by reading now: the share of students meeting grade level on the state test. Blue is high, orange and red are low. The east-west pattern is still visible, but it is looser than the first two maps. Poverty and absence explain a lot. They do not explain everything.
Some schools beat the line
The five enlarged campuses are all at least 70% low income and post the best reading scores in that group. Pleasant Hill reaches 51%, Campbell and Zavala 48%, Guerrero Thompson and Palm 44%. Guerrero Thompson is 95% low income, the poorest campus in the district, and it reads better than schools with a quarter of its poverty rate.
Two schools, one question
Pleasant Hill is 87.3% low income and gets 51% of its students to grade level in reading. Dobie Middle is 88.8% low income and gets 8%. A point and a half apart on poverty, 43 points apart on reading, in the same district under the same board. The district-wide numbers a commissioner gets briefed on show neither school.
110 Austin ISD campuses, 2023-24, recolored in turn by economic disadvantage, chronic absence, and STAAR reading at grade level. Dot size is enrollment, except on the steps that single out specific campuses, where the named campuses are enlarged instead.
Campus coordinates joined to TEA TAPR attendance and STAAR reading on the 9-digit TEA campus number, 2023-24. Campus poverty is derived from the economically-disadvantaged and all-students chronic-absenteeism denominators, which TAPR publishes per campus. Of Austin ISD's 114 campuses carrying all three measures, three are excluded for enrolling fewer than 100 students and one (Rosedale, a special-needs campus) has no coordinates, leaving 110 mapped. Basemap: Census TIGERweb 2020 ZIP Code Tabulation Areas, Austin metro.
Campus coordinates joined to TEA TAPR attendance and STAAR reading on the 9-digit TEA campus number, 2023-24. Campus poverty is derived from the economically-disadvantaged and all-students chronic-absenteeism denominators, which TAPR publishes per campus. Of Austin ISD's 114 campuses carrying all three measures, three are excluded for enrolling fewer than 100 students and one (Rosedale, a special-needs campus) has no coordinates, leaving 110 mapped. Basemap: Census TIGERweb 2020 ZIP Code Tabulation Areas, Austin metro.
Two things in that sequence are worth an agency’s time. The first is that absence looked like the answer and mostly wasn’t. Campuses with more chronic absence do read worse, at a correlation of -0.55, but absence and poverty are tangled together at 0.78, and among campuses at the same poverty level the effect nearly disappears: a 14-point gap in absence is worth about two and a half points of reading. Pleasant Hill has worse-than-average absence and the best reading score of any high-poverty campus in Austin. Getting students in the building is necessary and not sufficient, which is worth knowing before an agency spends a year on an attendance campaign and expects reading to follow.
The second is that the campuses at both ends are the ones a commissioner can act on. The five schools beating their poverty band are where to send someone to find out what they are doing. Dobie and the other low readers are where to send help. Neither group appears in the district-level numbers, and both serve different grade bands on a single year of one subject, so the map is a place to start asking, not a verdict on anyone’s staff.
What an agency does with this
Put the levels together and the use cases are the ones a commissioner’s office already handles, minus the wait. When a legislator asks how the state’s reading recovery compares to its neighbors, the answer is a chart pulled that morning rather than a request routed to a contractor. A board member who wants to know which districts sit furthest above the state absenteeism rate gets a ranking in minutes, without filing a request and waiting for someone else to fill it.
The person doing the pulling does not have to be the analyst, either. Every chart in this article came from a query, but a staffer can get the same answer by asking for it in plain English. An office with one data person becomes an office where anyone can check a number before a meeting.
The part that makes all of this defensible is provenance. Every row on the platform carries a hidden column pointing back to the exact authoritative file it was parsed from. Ask for it and the source URL appears next to the data, so a number in a board packet traces to the TEA, NCES, or ISBE file it came from, not to a screenshot of a spreadsheet nobody can find again. When someone challenges a figure in a hearing, that trail is the answer.
Your staff’s AI assistant, pointed at real data
Most commissioner’s offices now have someone using an AI assistant like Claude or ChatGPT. The catch is that those assistants, on their own, will happily invent an education statistic that sounds right, because they’re guessing from training data rather than reading a file. OpenData closes that gap by acting as a custom tool the assistant can call. In plain terms: your staff keeps using the AI tool they already have, and it gains the ability to search real government datasets, run the query, and draw the chart, with every number traced to its source.
There’s no new app to learn and no protocol jargon to wade through. Whoever administers your tools connects OpenData once, the way you’d add any add-on, and after that anyone on the team can ask a question in ordinary language and get a data-backed answer. No SQL, no spreadsheet gymnastics, no ticket to the analytics team. Here’s what that looks like when the board meeting is on the calendar and a staffer needs the chronic-absenteeism one-pager from earlier in this article.
The assistant does the parts your staff shouldn’t have to: it finds the right dataset, understands that TEA’s coded rows need to be summed into a district rate, runs the numbers, and renders the chart. That is the same CB0CA22R decoding problem from earlier in this article, handled by the person asking the question rather than routed to someone who knows the codebook. What it does not do is make anything up. The 27.0% for Austin and the 19.5% statewide are the figures in the state’s file, and the assistant can hand back the source link alongside them.
What’s in the catalog
The datasets behind this article are a slice of the education coverage on the platform. At the national level there’s NAEP scores by state and grade, per-pupil spending and teacher salaries back to 1969-70, plus enrollment, graduation, and civil-rights collections from NCES and the Census. At the state level, Illinois has 29 years of report cards and Texas has roughly 45 agency datasets covering attendance, test scores, finance, staffing, and the K-12-to-college pipeline. As of July 2026 the full catalog spans hundreds of datasets and is still growing, and they connect to each other. School data links to child-poverty estimates through shared geographic codes; poverty links to income; income links back to outcomes.
Where you sit
Every level in this article told the same kind of story. Nationally, fourth-grade reading fell 5.2 points, and that average hid both Louisiana’s 6-point gain and Maine’s 11-point drop. Texas reported 19.5% chronic absenteeism statewide while Austin ISD ran 7.5 points higher. Inside one Illinois county, chronic absence ran from 7.9% in Lake Forest to 34.4% in North Chicago. Inside one Austin district, two schools with the same poverty rate were 43 points apart in reading. The number you get from the level above is almost never the number you need.
Getting to the level below it used to mean codebooks and a two-week data request. That part is done.
Datasets used:
nces/naep— National Assessment of Educational Progress scale scores by jurisdiction, subject, and grade, 2003-2024 (11 assessment years, 52 jurisdictions). Reading and grade-4 scores used here. Source: NCES / NAGB.nces/per-pupil-expenditure— Per-pupil spending by state, 1969-70 to 2023-24. Historical years come from NCES Digest table 236.65 (current and constant 2022-23 dollars); 2021-22 through 2023-24 come from the Census Bureau’s Annual Survey of School System Finances (F-33), the survey NCES itself draws from, in nominal dollars only. Nominal FY2023-24 values used for the scatter; constant dollars used for the long-run 2.7x comparison. Sources: NCES Digest table 236.65, US Census Bureau F-33 FY2024 summary Table 20.nces/teacher-salary-by-state— Average teacher salary by state, 1969-70 to 2021-22, current and constant (2021-22) dollars. Source: NCES Digest table 211.60.tea/tapr-attendance-campus— Texas Academic Performance Report attendance metrics by campus, including chronic absenteeism from school year 2019-20 (4.7M rows). Source: Texas Education Agency.tea/tapr-staar-campus— Texas Academic Performance Report STAAR results by campus, grade, subject, and proficiency level, 2013-2025 excluding 2019-20 (35M rows). Grade-3-8 reading at “Meets grade level” used here. Source: Texas Education Agency.isbe/report-card— Illinois State Board of Education Report Card, school/district/statewide rows, 1997-2025. Source: ISBE.census/saipe— Census Small Area Income and Poverty Estimates, county and state child poverty and median income, referenced for the district poverty context. Source: US Census Bureau.
Calculations & transformations:
- NAEP change is each jurisdiction’s 2024 grade-4 reading score minus its 2019 score. Louisiana +6.1, national public -5.2.
- The spending-vs-score scatter joins nominal FY2023-24 per-pupil spending to spring-2024 grade-4 reading scores for the 50 states and DC. NAEP 2024 was administered during the 2023-24 school year, so the two series cover the same year. Pearson r = 0.073 (r-squared = 0.005) across the 51 rows. The similar-cost band comparison groups states into four spending bands ($11-14k, $14-17k, $17-21k, $21-32k); within-band score ranges are 12.9, 15.0, 21.3, and 23.2 points. It is a snapshot correlation, not a controlled estimate, and is uncontrolled for cost of living and student poverty.
- Teacher-pay change is the percent change in constant-2021-22-dollar average salary from 1999-00 to 2021-22, excluding a non-state footnote row. A decline in 36 states. The 2021-22 inflation spike depresses the endpoint: ending in 2020-21 instead leaves the national change roughly flat, so the magnitude is endpoint-sensitive while the direction is not.
- Texas chronic-absenteeism rates aggregate campus numerator and denominator pairs from the TAPR file, excluding pairwise any campus without a usable numerator and denominator. Masked cells (denominator 1-4) are about 1-2% of campuses; counting the ‘.’ placeholder and other unreported rows, total pairwise exclusion reaches roughly 5% in the 2019-20 through 2021-22 source years. The excluded campuses have tiny denominators, so the statewide rate is insensitive to their removal. TAPR attendance metrics lag one year, so the code year yy corresponds to the school year ending in 20yy; the latest available year is 2023-24.
- The Austin ISD campus map joins two TEA campus files on the 9-digit campus
id: STAAR reading at “Meets grade level” (all students, all tested grades)
from
tapr-staar-campus, and chronic absenteeism fromtapr-attendance-campus, both for 2023-24. The three STAAR proficiency tiers reproduce TEA’s published statewide rates (Approaches 62.3%, Meets 30.6%, Masters 8.7%), which is how the metric codes were confirmed. Campus poverty is derived as the economically-disadvantaged chronic-absenteeism denominator over the all-students denominator, since TAPR publishes no direct campus-level economically-disadvantaged percentage. Campus coordinates come fromtea/school-locations, joined on the same campus id. 114 of Austin ISD’s campuses carry all three measures. Three are excluded for having fewer than 100 students in the attendance denominator (Graduation Preparatory Academy and its Navarro campus at 66 each, and Austin State Hospital at 34, which is a hospital rather than a school), and Rosedale, a special-needs campus with no TEA accountability rating, has no coordinates, leaving 110 mapped. The map separates alternative and early-college campuses from neighborhood campuses so the two groups can be read on their own terms rather than against each other. Among the 46 neighborhood campuses at 70% or more low income, reading runs 8% to 51%; the five highest are Pleasant Hill (51%), Campbell (48%), Zavala (48%), Guerrero Thompson (44%), and Palm (44%). Pleasant Hill Elementary (87.3% low income, 51% reading) and Dobie Middle (88.8%, 8%) differ by 1.5 points of poverty and 43 points of reading; they serve different grade bands, so the pair is a pointer for further questions, not a performance judgment. - The absence-and-reading correlations cover the 103 neighborhood campuses on the map. Pearson r is -0.554 for reading against chronic absence, -0.625 for reading against low-income share, and +0.778 for absence against low-income share. Controlling for poverty, the partial correlation between reading and absence falls to -0.137 (r-squared 0.019), so absence explains under 2% of the remaining variation in reading. The within-band test splits the 46 campuses at 70%+ low income at the median absence rate: the lower half averages 27.0% absence and 24.4% reading, the upper half 41.2% and 21.9%, with average poverty of 86.2% and 87.0% respectively. These are descriptive correlations on one district, one year, and one subject, not a causal estimate of what attendance does to reading.
- The alternative and early-college comparison sums chronic-absenteeism numerators and denominators across all 118 Austin ISD campuses with a usable pair, including the three too small to appear on the map. Campuses are classed as alternative or early college by name (early college / ECHS, Garza Independence, Graduation Prep, the county JJAEP, International High School, and the Alternative Learning Center): 12 campuses, 9,466 students, 5,617 chronically absent, a 59.3% rate. The other 106 campuses total 60,284 students and 14,803 chronically absent, a 24.6% rate. So 13.6% of the district’s counted students sit on campuses carrying 27.5% of its chronic absence, rounded in prose to 14% and 27%. Name-based classification is a judgment call; a campus TEA codes as alternative under a different name would land in the “other” group.
- Illinois statewide chronic-absence rates come from the ISBE statewide aggregate row. 2018 is the first year ISBE reported chronic absence; the earlier chronic-truancy measure (2002-2017) uses a different state rule and shares no overlap years, so the two cannot be joined into one series.
- The Lake County scatter covers the 35 districts in that county with 1,000 or
more students, 2025. Both axes are ISBE-reported district percentages
(
student_enrollment_low_incomeis published as a percentage, not a count). District rows are identified by a null school name and an RCDTS ending in 0000; filtering on the row type alone returns duplicates. Pearson r = 0.800. Holding one county fixed controls roughly for cost of living and regional labor market, but this is a descriptive correlation, not a causal estimate.
Limitations:
- NAEP is a sample assessment. This dataset carries no standard errors, so differences smaller than about a point should be treated as ties.
- Louisiana’s and Mississippi’s gains are credited by state officials and some researchers to reading-reform laws, but the evidence is observational and, in Mississippi’s case, partly disputed because of a third-grade retention rule. No causal claim is made here.
- Teacher salaries are base averages that mask within-state and experience-mix variation. Constant dollars in this dataset are indexed to 2021-22 and are not comparable to the spending dataset’s 2022-23 base.
- The per-pupil spending series changes source in 2021-22, from NCES Digest table 236.65 to the Census F-33 survey NCES draws from. On the FY2021 overlap year the two agree to within 1.6% on average across 52 geographies (Massachusetts is the largest gap at 5.4%), reflecting NCES revisions and a slightly different enrollment denominator. Census years are nominal only, so cross-year real-dollar comparisons should stay inside the NCES range (1969-70 to 2020-21). Census Table 20 also omits the territories NCES covers.
- The 2019-20 Texas rate and the 2020 Illinois rate are pandemic counting artifacts, flagged on both charts.
- Illinois “chronic absence” and Texas “chronic absenteeism” are defined under different state rules and are not directly comparable; no cross-state absenteeism comparison is made.
- Chronic absence is a threshold measure, not an average: it counts students who missed 10% or more of the school year. A campus rate near 90% means nearly every enrolled student crossed that threshold, not that the school sat empty. At credit-recovery and alternative campuses, where students often enroll after falling behind on attendance elsewhere, a rate that high reflects the population served.
- Campus reading and campus poverty are single-year, single-subject snapshots. Elementary and middle campuses test different grades, so campus-to-campus comparisons across grade bands are indicative rather than like-for-like.
Data accessed on 2026-07-24 via the OpenData API.
OpenData is in active development. The datasets in this article live at tryopendata.ai.
