2026 State of Government Website Accessibility Report
A field assessment of WCAG 2.1 AA conformance across 2,400+ government websites in the United States. Sector-by-sector data. PDF risk findings. Platform-level patterns your procurement team needs to see.
Executive Summary
The DOJ's ADA Title II web accessibility deadline for larger government entities lands on April 26, 2027. That's the clock most compliance officers are working against. To see where agencies actually stand, WPPersona evaluated 2,412 public-facing government websites across the United States, spanning the full SLED landscape of local government, K-12 school districts, special districts, and public higher education.
The gaps are wide. They're also stubbornly consistent across sectors. Only 11% of the evaluated sites passed full WCAG 2.1 AA on automated criteria, and when manual testing criteria come into play, that number drops below 4%. Document accessibility is where the real damage sits. Small and mid-size entities are the furthest behind, and awareness of the 2027 deadline hasn't moved the numbers as much as you'd hope.
2,412
Government websites evaluated
11%
Fully conformant (automated)
68%
Had Critical or High-risk PDFs
34
Average WCAG violations per site
Key Findings
Finding 1: 89% of government websites fail at least one Level AA WCAG criterion
Automated scans of 2,412 government websites returned a pass rate of 11.3% (273 sites) on all detectable WCAG 2.1 AA criteria. The other 88.7% failed at least one. The most common trouble spots: insufficient color contrast (62% of sites), missing or weak form labels (54%), missing alt text on images (47%), and keyboard navigation failures (39%). Add the manual testing criteria (proper heading structure, meaningful link text, ARIA implementation quality, logical reading order in documents) and the fully conformant share drops to under 4%.
Finding 2: PDF accessibility is the most widespread and most severe compliance gap
78% of the evaluated sites published at least one PDF. Of those, 68% had at least one PDF classified as Critical or High risk. The average site with PDFs held 847 documents. 41% of those were Critical (image-only or untagged) and 32% were High (tagged but riddled with failures). K-12 school districts sat at the top of the pile, both in volume (average 2,340 PDFs per site) and in Critical-risk share (48%). Years of staff-scanned handbooks, course catalogs, and curriculum PDFs will do that to an archive.
Finding 3: The compliance gap between automated and full WCAG conformance is widening
Automated WCAG tools catch roughly 30-40% of violations. The 11% of sites that pass automated testing still carry an estimated 60-70% of their WCAG violations undetected. That's a false confidence problem, and it's the reason a passing axe score in one meeting turns into an OCR complaint six months later. The violations automated tools miss (ARIA that validates but confuses screen readers, keyboard traps in complex widgets, PDF reading order that scrambles content) are the same violations that render a site unusable for assistive tech.
Finding 4: CMS platform choice has a larger impact on compliance outcomes than agency size or budget
Platform type showed a stronger correlation with compliance than either agency size or budget. Sites on purpose-built compliant government CMS platforms hit 3.2x the compliance rate of sites on general-purpose platforms (WordPress, Drupal, Squarespace). Even inside the general-purpose category, skilled web teams still posted higher violation counts than agencies on purpose-built platforms. The reason is structural. General-purpose platforms don't enforce accessibility at the code level. They rely on human vigilance for every content decision, and human vigilance loses to a Tuesday afternoon deadline every time.
Finding 5: Agency size is not the primary predictor of compliance readiness
The assumption that bigger agencies with bigger budgets do better didn't hold up. Population size showed no significant correlation with WCAG conformance inside the local government sector. Small townships with the right platform and a named owner hit 100% automated conformance. Large counties on legacy platforms landed at 20-30%. The differentiator wasn't money or headcount. It was two things: platform-level accessibility enforcement, and one named person inside the organization who owned digital accessibility as their job.
Methodology
The 2026 Government Website Accessibility Report evaluated 2,412 public sector websites across four sectors: local government (cities, counties, townships), K-12 school districts, special districts (transit, water, library, housing), and public higher education. Sites were pulled using a stratified random sample from each sector, weighted by population size so large, mid-sized, and small entities all showed up in the data.
Automated WCAG scan
Each site homepage plus up to 5 subpages ran through a combination of the axe-core engine (version 4.8), the WAVE accessibility evaluation tool, and Lighthouse. Results were aggregated across all tested pages per site.
PDF inventory and classification
Each site was crawled for PDF documents. Up to 100 PDFs per site were sampled and classified using WPPersona's 5-criterion risk matrix. Sites with fewer than 100 PDFs had every document evaluated.
CMS platform identification
CMS platform was identified via HTML source analysis, HTTP headers, and public records where available. Sites were classified by platform type for the comparative analysis.
Manual review (sample)
10% of sites (241 sites) received a manual WCAG review by trained accessibility evaluators covering keyboard operability, screen reader compatibility, and logical content structure.
Section 1: Compliance Status Across Government Sectors
| Sector | Sites Evaluated | Avg. Violations | Fully Conformant | High/Critical PDF Risk |
|---|---|---|---|---|
| Cities & Counties | 847 | 31 | 13% | 63% |
| K-12 School Districts | 612 | 44 | 8% | 81% |
| Special Districts | 494 | 38 | 9% | 72% |
| Public Higher Education | 459 | 28 | 14% | 58% |
| Overall Average | 2,412 | 34 | 11% | 68% |
K-12 districts topped every sector on both violation count and PDF risk. If you've worked in a large district, you already know why. Individual school offices upload documents without central review, teachers post their own PDFs, principals send home scanned handbooks, and nobody owns the archive. Decentralized publishing accumulates inaccessible documents faster than any policy memo can slow it down.
Avg. WCAG violations per site by sector
Overall avg 34
Automated WCAG 2.1 AA scan results across 2,412 government websites.
Section 2: PDF Accessibility, The Scale of the Problem
The PDF numbers are the ones that keep compliance officers up at night. Of the 2,412 sites evaluated, 1,881 (78%) published at least one PDF. The aggregate inventory across those sites hit 1.94 million documents. Classification of the sample produced this risk distribution:
29%
Critical
Image-only, no text layer
38%
High
Tagged with multiple failures
21%
Medium
Minor failures only
12%
Low
Passing or near-passing
Extrapolate from the sample and you're looking at roughly 565,000 government PDFs in this dataset that are image-only scans with no text layer. Completely invisible to screen readers. That's a compliance exposure staff training alone won't close. It needs automated tooling at scale, vendor support, and in a lot of cases, external remediation help.
Section 4: CMS Platform Analysis
The link between CMS platform and compliance outcome was one of the strongest signals in the dataset:
| Platform Category | Avg. Violations | % Fully Conformant | PDF Enforcement |
|---|---|---|---|
| Purpose-built compliant gov CMS | 4 | 34% | Platform-enforced |
| Enterprise municipal CMS (CivicPlus, Granicus) | 18 | 17% | Advisory only |
| Managed WordPress (hosting only) | 31 | 9% | None |
| DIY WordPress | 52 | 4% | None |
| Website builders (Wix, Squarespace) | 39 | 6% | None |
| Abandoned/unsupported legacy CMS | 78 | 0% | None |
Avg. violations by CMS platform category
Purpose-built: 4
Lower bars mean fewer average WCAG 2.1 AA violations per site.
Section 5: What Compliant Agencies Do Differently
Looking at the 273 fully conformant sites, a few patterns showed up again and again. If you're building your 2027 plan, this is your checklist:
94%
Platform-level enforcement
94% of conformant sites ran on platforms that enforced at least some accessibility requirements structurally, as platform constraints rather than editorial guidelines.
89%
Named accessibility owner
89% had one named person with explicit responsibility for digital accessibility. Not an IT director juggling it alongside dozens of other duties. One owner, one job.
100%
Published accessibility statement
100% of conformant sites published a findable, up-to-date accessibility statement. Zero of the most severely non-conformant sites did.
87%
Regular automated monitoring
87% had evidence of regular automated accessibility scanning, based on scan metadata, compliance dashboard indicators, or public compliance reports.
78%
PDF accessibility verification
78% had a documented process or platform mechanism for verifying PDF accessibility before publication.
Recommendations
Run a PDF inventory first. For most agencies, PDFs are the biggest and most severe accessibility risk. Do a bulk scan before you spend a dollar on HTML template remediation.
Ask your CMS the enforcement question. If your platform can't block a non-compliant PDF upload, can't require alt text, or can't hold heading hierarchy, you have a structural problem no amount of staff training will fix.
Name your accessibility owner. Give them the authority to say no. Compliance isn't a project with an end date. It's an operational discipline, and it needs someone whose job depends on it.
Publish an accessibility statement now, even if you're not fully conformant. It shows good faith, gives your users a feedback channel, and starts the documentation record OCR will ask for.
Stop treating a passing automated scan as proof of accessibility. A green axe score misses 30-40% of violations, not confirms zero. Invest in manual testing for the pages your residents actually use, and for the services they can't opt out of.
Download the Full Report
The full 2026 report includes sector-specific breakdowns, state-level data, complete methodology, and recommendations for each agency type. Available via demo.
Schedule Demo to Get Full ReportReport authored by Logan Nathan, WPPersona Research.
Back to all resources