Classroom Technology Statistics: What the Federal Data Actually Shows

A short list of sources supplies everything below. The National Center for Education Statistics (NCES), part of the Institute of Education Sciences at the U.S. Department of Education, runs two of them: the School Pulse Panel, a rolling nationally representative school survey that supplies the current 2024–25 provision and connectivity figures, and the older Fast Response Survey System study of instructional technology use, which supplies the 2019–20 detail. NCES Fast Facts summarizes the American Community Survey and the Current Population Survey, both federal household surveys, for the home-access figures. SETDA — the State Educational Technology Directors Association — supplies the state-leader survey. And USAC, the Universal Service Administrative Company, administers the E-Rate program and publishes its eligibility rules.

Three cautions apply throughout. First, the two NCES instruments are not interchangeable: the School Pulse Panel asks whether a school runs a 1-to-1 program, the 2019–20 FRSS study asked whether a school had a computer for each student and then asked a great deal more about how those computers were used and where they lived. Quote each against its own question. Second, school-level surveys and household surveys measure different things: a figure about schools providing devices and a figure about children having internet at home are not interchangeable, and combining them produces claims neither survey supports. Third, weighted national estimates carry standard errors — NCES publishes them alongside every table — so small differences between similar percentages are frequently not meaningful.

For scale, NCES reports that in fall 2022 about 49.6 million students were enrolled in public elementary and secondary schools, prekindergarten through grade 12. That is the denominator behind every percentage that follows.

Start with the current figure, because it is the one most searches are after. The NCES School Pulse Panel, collecting from 1,490 public schools between December 6 and December 20, 2024, found that 88 percent of public schools reported a 1-to-1 computing program — a program providing every student with a school-issued device. That is the defensible answer to "how many US schools are one-to-one," and it carries the 2024–25 school year with it.

What the School Pulse Panel does not do is describe how those devices are used, where they live, or how the network holds up under them. For that the source is a different and older report: Gray, L., and Lewis, L. (2021), Use of Educational Technology for Instruction in Public Schools: 2019–20, NCES 2021-017. It was collected in spring 2020 through the Fast Response Survey System, mailed to 1,300 public schools across the 50 states and the District of Columbia, and answered by principals or the staff member who knew most about instructional technology at the school. Critically, the report states that schools completing the survey after the pandemic began were asked to report on pre-pandemic experiences — so the whole dataset describes the 2019–20 school year as it was before March 2020.

The report defines computers as desktop, laptop and tablet computers, explicitly including Chromebooks and iPads, and explicitly excluding smartphones. That definition matters when comparing against a district's own asset count.

On provision, that 2019–20 survey found 45 percent of schools reported having a computer for each student, with a further 37 percent reporting a computer for each student in some grades or classrooms. Read together, roughly four in five schools had at least partial one-to-one provision before the pandemic — but fewer than half had it universally. Set the 45 percent against the School Pulse Panel's 88 percent for 2024–25 and the shape of the stimulus-funded device wave is visible in a single pair of numbers. The two are not the same question asked twice, so the difference is a trajectory rather than a precise growth rate — but the direction is not in doubt.

On take-home, 15 percent of schools let students in all grades take school-provided computers home and another 8 percent allowed it in some grades. A separate 15 percent permitted students to take computers home on a short-term basis. And 9 percent of schools provided mobile hotspots or web-enabled devices with paid data plans for students to take home — the pre-pandemic baseline for the home-connectivity programs that expanded sharply afterwards.

One of the most operationally useful tables in the NCES report is the one nobody quotes. Asked where student-use computers were located, schools reported that about one third — 34 percent — were assigned to individual students to carry during the school day. Thirty-nine percent stayed in a specific classroom. Sixteen percent moved between classrooms. Ten percent sat in resource rooms, computer labs, and library and media centers.

That distribution is a storage and power specification in disguise. Devices that stay in a classroom need a cabinet or a fixed charging solution. Devices that move between classrooms need a cart, casters, a lock and a power plan. Devices carried by students need neither, but need a check-in and asset-tracking process instead. A district whose ratio looks nothing like the national one is not doing anything wrong — but it should know which of the three models it is actually funding.

NCES also captured quality perception. A little over eight in ten schools rated the overall quality of computers used for teaching and learning as good (52 percent) or very good (30 percent), with software rated similarly at good (53 percent) or very good (31 percent). About nine in ten reported that their computers met the school's teaching and learning needs to a moderate (38 percent) or large extent (52 percent).

Against that, the same survey found 22 percent of schools describing outdated computers or software as a moderate challenge for teachers and a further 12 percent as a large challenge — a third of schools naming device age as a real instructional obstacle, in a survey where most respondents also rated their equipment as good.

NCES found that nearly two-thirds of schools — 64 percent — said internet connections in the teaching and learning areas of the school were very reliable. That is a figure about the classroom, not about the district's circuit to the internet, and the distinction matters: a school can have abundant wide-area bandwidth and still have rooms where a class set cannot all associate to the access point.

The load question was measured separately. About half of schools — 52 percent — reported having problems to a small extent with internet connections or speed when large numbers of students were online simultaneously, while 20 percent reported no such problems at all. In other words, the majority of schools in 2019–20 experienced at least some degradation under concurrent load.

That is the single most transferable finding on this page, because concurrency is a design problem rather than a bandwidth problem, and it has not gone away. A room that supported fifteen devices comfortably in 2019 is being asked to support thirty today, on access points that may be the same ones. If your building is showing symptoms under load, the honest first diagnosis is access point density and placement, not the internet circuit.

NCES Fast Facts reports, from the 2021 American Community Survey, that 97 percent of 3- to 18-year-olds had home internet access, that 93 percent had access through a computer, that 4 percent relied on a smartphone for home internet access, and that 3 percent had no internet access at home.

The headline is genuinely encouraging and the distribution underneath it is where the planning happens. By parental education, computer-based home access ranged from 98 percent where a parent held a bachelor's degree or higher down to 78 percent where the highest attainment was less than a high school credential. By family income, access ran from 99 percent in the highest quarter to 94 percent in the lowest. By race and ethnicity, NCES reports 99 percent for Asian children and 89 percent for American Indian and Alaska Native children.

The reasons for non-access also matter for program design. From the 2021 Current Population Survey, NCES reports that among households without home internet access, 48 percent cited not needing it or not being interested, 24 percent cited cost, and 6 percent reported no service available. A district planning a hotspot program is therefore addressing a smaller and differently-shaped population than a headline access gap implies — and a district in a rural area where the no-service-available share is higher than the national figure is solving a different problem from one where cost dominates.

One caution repeated: these are household survey figures about children, collected in 2021. They do not describe school-provided connectivity, and they cannot be combined with the school-level device figures above to produce a claim about students' ability to complete assigned work at home.

Line up the collection dates and the shape of the evidence is clear. The School Pulse Panel is current: 2024–25, headline provision and connectivity. The detailed instructional-technology survey describes the 2019–20 school year, deliberately capturing pre-pandemic conditions. The household access figures come from 2021. The enrollment denominator is fall 2022. Meanwhile the devices in most district fleets today were purchased in the federal stimulus wave that followed the FRSS study, and are now moving through their second or third refresh decision.

So there is a current federal headline — 88 percent of schools running a 1-to-1 program in 2024–25 — and almost no current federal detail underneath it. Nothing published today tells you the national split between carried, classroom-fixed and cart-mounted devices, the national picture of concurrency under a class set, or the national device-age distribution. Where you see a 2026 figure for any of those, check whether it is a vendor survey, a market estimate or an extrapolation, and label it accordingly.

For a district technology office, the practical consequence is that the headline is not the tool for a refresh case. National benchmarking answers whether you are unusual; it does not build a budget. Your own asset system — device counts by model, purchase year, end-of-support or Auto Update Expiration date, and failure rate by model — is both more current and more persuasive than any national figure. A finance committee responds to a per-model count with vendor-published dates against it. It does not respond to a national average, and it should not.

Use the federal data for what it is genuinely good at: describing structure and distribution. The location split of student devices, the concurrency finding, the shape of the home-access gap by income and parental education — these describe patterns that remain useful even as the absolute levels move.

The most current sector-level datum available comes from SETDA's 2025 State EdTech Trends Report, produced with Whiteboard Advisors and drawing on responses from 47 states and the Department of Defense Education Activity. SETDA reports that only six percent of respondents indicated they have plans in place to continue funding edtech initiatives previously supported with federal stimulus dollars — a sharp decline from 27 percent in 2024.

The same report records artificial intelligence ranking as the top state edtech priority for the first time, displacing cybersecurity, which had held that position for two years.

This is a survey of state education agency leaders, not a count of district budgets, and it should be cited as such. What it establishes is directional and important: the sustainability question that districts have been raising internally is now documented at state level, by the officials who administer the programs.

Because E-Rate figures are frequently misquoted into device conversations, it is worth stating the eligibility structure precisely, in USAC's own terms. USAC describes Category One as "Data Transmission Services and/or Internet Access," covering "broadband connectivity and basic conduit access to the internet." Category Two covers the internal connections — USAC lists "eligible equipment (such as routers, switches, wireless access points, and cabling)" needed to distribute broadband within school or library buildings.

On devices, USAC is explicit: Category One services "cannot include charges for content, end-user devices or equipment purchases (other than those necessary for transmission of Category One services)." Laptops, tablets, desktops, charging carts, interactive displays, projectors and document cameras are end-user equipment and fall outside the program.

USAC also publishes the funding-year calendar, and April 1, 2026 was the last date of the FY2026 application filing window. Program-level demand and commitment totals for the year are published by USAC and the FCC; we have deliberately not reproduced dollar figures here that we could not read directly from a USAC or FCC source, which is the same standard applied to every other number on this page.

The practical consequence for a district buying hardware is simple and often missed: build two lists. The connectivity and internal-connections list can be pursued through E-Rate on its own timetable. The device, storage, power and classroom AV list has to be funded from local, state, bond or general-fund money. Mixing them in one requisition delays both.