Arkansas Lost More Students This Year Than COVID Took

Mon, 22 Dec 2025 12:00:00 GMT

The COVID-19 pandemic was supposed to be the shock. In 2020-21, Arkansas public schools lost 6,428 students in a single year, the kind of enrollment hit that prompts emergency budget meetings and anxious headlines. It took three years to claw back roughly half of what was lost.

Then 2025-26 erased all of it, and then some.

Arkansas public schools enrolled 465,421 students this year, down 8,916 from the prior year. That is a 1.9% decline in a single year, the largest on record in 21 years of state data, and 39% larger than the COVID drop. The state now sits 14,011 students below its pre-pandemic peak of 479,432 in 2019-20 and at its lowest enrollment since 2005-06.

A record nobody wanted

The scale of this year's loss is visible in the year-over-year record. Before 2020-21, the largest single-year decline in the dataset was just 940 students in 2018-19. The COVID year shattered that pattern with a loss of 6,428. But even COVID left the state above 473,000. The 2025-26 figure of 465,421 is a level Arkansas has not seen since the 2005-06 school year.

The post-COVID recovery, such as it was, peaked in 2022-23 at 476,579 students, recovering 3,575 of the 6,428 lost, or about 55.6%. Then the trajectory reversed. The state shed 1,372 students in 2023-24, another 870 in 2024-25, and then 8,916 this year. The three-year combined loss of 11,158 amounts to 2.3% of the 2023 enrollment.

Three out of four districts shrank

The decline was not concentrated in a handful of struggling urban cores. Of 258 districts with data in both years, 192 lost students, 74.4% of the total. Only 65 grew, and one was flat.

Little Rock^↗ led all districts with a loss of 601 students (-3.1%), followed by Springdale^↗ at 559 (-2.6%) and Fort Smith^↗ at 478 (-3.5%). Conway^↗ lost 369 (-3.7%) and Pulaski County Special^↗ lost 362 (-3.0%). These five districts alone account for 2,369 students, about 27% of the net statewide loss.

But the breadth matters more than the concentration. Mid-size districts like Russellville^↗ (-183, or -3.4%), Siloam Springs^↗ (-265, or -5.8%), and Lake Hamilton^↗ (-207, or -5.4%) posted losses well above the statewide average. eStem Public Charter School^↗, a well-regarded B-rated charter, lost 340 students, a staggering 14.4% of its enrollment.

The single notable exception: Bentonville^↗ gained 369 students (+1.9%), driven by the ongoing population boom in Northwest Arkansas anchored by the Walmart, Tyson Foods, and J.B. Hunt corporate campuses. Bentonville is now the state's second-largest district at 19,944 students, closing the gap with Springdale's 21,097.

The voucher question

The most obvious variable that changed between 2024-25 and 2025-26 is the full expansion of the Education Freedom Accounts program created by the Arkansas LEARNS Act. For its first two years, participation was capped and restricted to specific student categories. This year, every K-12 student in Arkansas became eligible, and participation surged.

According to state budget documents, roughly 28,100 students received EFA accounts for private school attendance and another 18,500 for homeschooling or microschool enrollment in 2025-26, at a projected cost of $326 million. That is a dramatic increase from the 14,297 participants in the program's second year and 5,548 in its first.

But the relationship between those numbers and the enrollment loss is not straightforward. A fiscal analysis by Reason Foundation estimated that only 27.5% of second-year EFA participants were "switchers" who would have otherwise attended public school. In the first year, the rate was 34.8%. The rest were students already enrolled in private schools, homeschooled, or entering kindergarten for the first time. If the switcher rate held at roughly 25-35% for the expanded third year, that would account for somewhere between 7,000 and 16,000 actual departures from public schools, a range wide enough to explain most, all, or more than all of the 8,916-student loss.

The honest answer: nobody knows the precise switcher rate for 2025-26 yet. The data does not exist in the enrollment files.

Birth rates: the slow-motion factor

The other force at work predates the LEARNS Act by more than a decade. Arkansas births peaked in 2007 and have declined nearly every year since, with roughly 4,000 fewer children born per year by 2017 compared to the peak. Those smaller cohorts have been working their way through the K-12 pipeline. By fall 2021, every grade from kindergarten through eighth consisted of students born during the declining-birth-rate era.

"The state will likely lose more than 15,000 students" over the following five years as smaller birth cohorts replace larger graduating classes. -- Talk Business & Politics, August 2019

That projection, made before anyone had heard of COVID-19 or Education Freedom Accounts, anticipated sustained demographic losses on roughly the scale the state is now experiencing. Hot Springs^↗ Superintendent Mike Hernandez told the Democrat-Gazette that he attributes his district's 3.4% enrollment drop since 2023-24 to shrinking birth rates, a trend visible in districts across the state regardless of school grades.

Virtual schools grew while everything else fell

The sector-level data offers one more clue about where students went. Arkansas Connections Academy^↗ grew by 1,205 students (+26.3%) and Arkansas Virtual Academy^↗ by 510 (+9.7%), bringing the combined virtual enrollment to 11,559, nearly triple its pre-COVID level of 4,071 in 2019-20. Virtual schools have grown every year since the pandemic, a pattern not reversed by the return to in-person schooling.

Traditional districts absorbed the full force of the decline and then some. The charter sector (brick-and-mortar charters, identified by name) was essentially flat, with losses at eStem (-340) and others roughly offset by gains at Exalt Academy^↗ (+355) and Graduate Arkansas Charter^↗ (+250). The virtual sector was the only one to post clear growth.

Whether virtual enrollment growth represents families choosing a different public school model or an intermediate step before leaving public education entirely, the data cannot say.

The racial composition of the loss

White students accounted for 7,863 of the 8,916-student decline, or 88.2% of the total loss. Black enrollment fell by 1,593, and Hispanic enrollment fell by 1,157. Only Asian students (+327) and multiracial students (+1,449) posted gains.

The disproportionate white loss is consistent with both the EFA program's initial demographic profile and longer-running demographic trends. White students have declined from 69.4% of Arkansas enrollment in 2004-05 to 56.5% in 2025-26, a 12.9 percentage-point drop over two decades. Hispanic enrollment has grown from 6.0% to 15.4% over the same period, and multiracial students from near zero to 5.4%. This year's loss accelerated those trajectories but did not create them.

Hispanic enrollment declined for the first time in 21 years of data, after growing every single year since 2004-05. The 2025-26 drop cut across demographic lines, not just along them.

96 districts at their lowest point ever

Among the 259 districts in the 2026 data, 96 now sit at their lowest enrollment in the full 21-year dataset, 37.1% of all districts. That figure includes small rural districts that have been declining for decades and mid-size suburban districts that were growing as recently as 2022.

The funding implications are immediate. Arkansas allocates foundation funding on a per-pupil basis, and state officials have argued that historic funding increases mean districts can absorb a 2.5% enrollment loss before budgets are affected. But 122 of 258 districts, nearly half, lost more than 2.5% this year.

"The enrollment shift means already underfunded public schools face challenges that are 'only going to get more dire.'" -- April Reisma, Arkansas Education Association president, Arkansas Democrat-Gazette, January 2026

What 2026-27 might reveal

The 2025-26 drop cannot be pinned on any single cause. Birth-rate-driven pipeline shrinkage was already forecast to cost Arkansas 15,000 or more students by the mid-2020s. The LEARNS Act's universal EFA expansion added a powerful pull factor in the same window. Virtual enrollment tripled over six years for reasons distinct from either.

The 2026-27 count will show whether this year was a one-time adjustment as pent-up demand for the EFA program was released, or the beginning of a steeper decline. If the switcher rate stabilizes and no new cohort of public school families applies for vouchers, the losses could moderate. If the program continues to grow and smaller birth cohorts continue to enter kindergarten, the state could fall below 460,000 within a year.

Detailed code that reproduces the analysis and figures in this article is available exclusively to EdTribune subscribers.

Lake Hamilton - EdTribune AR - Arkansas Education Data