Accessibility Statement Scanning

Stats as of 2026-04-09 02:54 UTC — last scan: 2026-04-09

107 scan batches run

72,735 of 82,714 available pages scanned (87.9% coverage) 67,302 of 72,735 scanned pages were reachable (92.5%) 32,622 of 67,302 reachable pages have an accessibility statement (48.5%) 28,212 pages have the statement link in the footer (86.5% of pages with a statement)

📥 Machine-readable results are available as the accessibility-data.json artifact in the latest Generate Scan Progress Report workflow run.

Each country entry in the JSON file includes page-level evidence for pages with and without accessibility statements, plus a per-domain summary you can share to validate the published counts.

Hover or focus any non-zero count in the country table to preview the matching pages. If there are 20 or fewer URLs, the preview shows all of them; otherwise it shows a short sample. Full machine-readable data is available as the accessibility-data.json artifact in the latest workflow run.

Accessibility Statement Scan by Country

Country Scanned Available Reachable Has Statement In Footer Statement % Scan Period
AUSTRIA 821 821 787 547 517 69.5% Mar 2026 – Apr 2026
BELGIUM 1,309 1,309 1,229 551 502 44.8% Mar 2026 – Apr 2026
BULGARIA 291 291 269 61 59 22.7% Mar 2026 – Apr 2026
CROATIA 233 233 231 85 61 36.8% Mar 2026 – Apr 2026
CZECHIA 843 843 804 427 365 53.1% Mar 2026 – Apr 2026
DENMARK 1,521 1,521 1,503 1,036 1,019 68.9% Mar 2026 – Apr 2026
ESTONIA 396 396 389 136 65 35.0% Mar 2026 – Apr 2026
FINLAND 180 180 172 111 104 64.5% Mar 2026 – Apr 2026
FRANCE 10,007 10,007 9,425 3,655 3,516 38.8% Mar 2026 – Apr 2026
GERMANY 6,555 6,555 6,479 4,612 3,872 71.2% Mar 2026 – Apr 2026
GREECE 1,748 1,748 1,627 409 238 25.1% Mar 2026 – Apr 2026
HUNGARY 390 390 366 64 48 17.5% Mar 2026 – Apr 2026
ICELAND 139 139 135 15 7 11.1% Mar 2026 – Apr 2026
IRELAND 522 522 498 226 205 45.4% Mar 2026 – Apr 2026
ITALY 5,338 5,338 4,851 2,622 2,556 54.1% Mar 2026 – Apr 2026
LATVIA 802 802 771 488 444 63.3% Mar 2026 – Apr 2026
LITHUANIA 120 120 108 0 0 0.0% Mar 2026 – Apr 2026
LUXEMBOURG 571 571 250 85 74 34.0% Mar 2026 – Apr 2026
MALTA 608 608 595 381 374 64.0% Mar 2026 – Apr 2026
NETHERLANDS 937 937 909 449 439 49.4% Mar 2026 – Apr 2026
NORWAY 239 239 233 108 102 46.4% Mar 2026 – Apr 2026
POLAND 13,875 14,938 12,866 5,309 3,369 41.3% Mar 2026 – Apr 2026
PORTUGAL 3,503 3,503 2,937 895 720 30.5% Mar 2026 – Apr 2026
REPUBLIC_OF_CYPRUS 24 24 24 0 0 0.0% Mar 2026 – Apr 2026
ROMANIA 799 799 351 28 9 8.0% Mar 2026 – Apr 2026
SLOVAKIA 434 434 414 192 177 46.4% Mar 2026 – Apr 2026
SLOVENIA 200 200 190 100 74 52.6% Mar 2026 – Apr 2026
SPAIN 6,069 6,069 5,374 2,379 2,083 44.3% Mar 2026 – Apr 2026
SWEDEN 1,558 1,558 1,486 860 785 57.9% Mar 2026 – Apr 2026
SWITZERLAND 2,117 2,117 2,082 980 978 47.1% Mar 2026 – Apr 2026
UNITED_KINGDOM_UK 10,586 19,502 9,947 5,811 5,450 58.4% Mar 2026 – Apr 2026
Total 72,735 82,714 67,302 32,622 28,212 48.5% —

Statement % is the percentage of reachable pages that contain at least one link to an accessibility statement.

Overview

The accessibility statement scanner checks whether each government page links to an accessibility statement as required by the EU Web Accessibility Directive (Directive 2016/2102).

Under the Directive, public-sector bodies must:

  1. Publish an accessibility statement describing the accessibility of their website or mobile app.
  2. Include a clearly labelled link to that statement on the page, ideally in the footer.

The scanner detects these links using multilingual term matching across all 24 EU official languages plus Norwegian and Icelandic.

Scans run automatically every 4 hours via GitHub Actions so that the full set of ~80,000 URLs across 31 countries can be covered gradually without overloading government servers.

What Is Checked

For each scanned page the scanner:

  1. Fetches the page HTML.
  2. Searches first inside <footer> elements for links whose text or href matches known accessibility-statement terminology.
  3. If not found in the footer, searches the entire page.
  4. Records whether a matching link was found, where it was found (footer or page body), and what text triggered the match.

Multilingual Term Matching

The glossary covers the following languages:

Region Languages
EU official languages Bulgarian, Croatian, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Hungarian, Irish, Italian, Latvian, Lithuanian, Maltese, Polish, Portuguese, Romanian, Slovak, Slovenian, Spanish, Swedish
Allied nations Icelandic, Norwegian

Example recognised terms include “accessibility statement” (EN), “déclaration d’accessibilité” (FR), “Erklärung zur Barrierefreiheit” (DE), and equivalents in all supported languages.

Tier Classification

Each scanned page is assigned one of three outcomes:

Outcome Meaning
unreachable Page could not be fetched (network error, timeout, 4xx/5xx)
no_statement Page is reachable but no accessibility statement link was found
has_statement Page contains at least one link to an accessibility statement

Pages where the statement link was found inside a <footer> element are additionally flagged with found_in_footer = true, since placing the link in the footer is considered best practice.

Viewing Results

Scan Progress Report

The Scan Progress Report includes a per-country accessibility statement breakdown showing:

  • Total pages scanned and reachable count
  • Number of pages with a statement link
  • Number of pages where the link was found in the footer
  • Date range showing when each country was last scanned

GitHub Actions Artifacts

Each workflow run uploads a scan artifact containing:

  • data/metadata.db — the full SQLite results database
  • accessibility-scan-output.txt — the raw scan log
  • data/toon-seeds/countries/**_accessibility.toon — annotated TOON files

To download artifacts:

  1. Go to GitHub Actions → Scan Accessibility Statements
  2. Click on the relevant workflow run
  3. Scroll to the Artifacts section at the bottom of the run summary page
  4. Download accessibility-scan-<run_number> to inspect the database or TOON files

Running a Scan Manually

  1. Go to Actions → Scan Accessibility Statements
  2. Click Run workflow
  3. Optionally enter a country code (e.g. ICELAND) or leave blank to scan all
  4. Optionally adjust the rate limit (default: 1.0 req/sec)

Via the command line

# Scan a single country
python3 -m src.cli.scan_accessibility --country ICELAND --rate-limit 1.0

# Scan all countries (with a 110-minute runtime cap)
python3 -m src.cli.scan_accessibility --all --max-runtime 110 --rate-limit 1.0

Output Format

Annotated TOON file (*_accessibility.toon)

Each page entry gains an accessibility field:

{
  "url": "https://example.gov/",
  "is_root_page": true,
  "accessibility": {
    "is_reachable": true,
    "has_statement": true,
    "found_in_footer": true,
    "statement_links": ["https://example.gov/accessibility"],
    "matched_terms": ["accessibility statement"]
  }
}

Database table (url_accessibility_results)

Column Type Description
url TEXT Page URL
country_code TEXT Country identifier (e.g. ICELAND)
scan_id TEXT Unique scan run identifier
is_reachable INTEGER 1 = reachable, 0 = not reachable
has_statement INTEGER 1 = accessibility statement link found
found_in_footer INTEGER 1 = link was found inside a <footer> element
statement_links TEXT JSON list of resolved statement URLs
matched_terms TEXT JSON list of matched glossary terms
error_message TEXT Error message if fetch failed
scanned_at TEXT ISO-8601 timestamp of scan

Countries Covered

Scans cover all 27 EU member states plus 4 allied nations:

Region Countries
EU member states Austria, Belgium, Bulgaria, Croatia, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Poland, Portugal, Republic of Cyprus, Romania, Slovakia, Slovenia, Spain, Sweden
Allied nations Iceland, Norway, Switzerland, United Kingdom

Architecture

scan-accessibility.yml (GitHub Actions — every 4 hours)
    ↓
scan_accessibility.py (CLI)
    ↓
AccessibilityScannerJob.scan_country()
    ↓
AccessibilityScanner.scan_urls_batch()
    ↓
For each URL:
    httpx.get()  →  HTML content
    BeautifulSoup  →  find <footer> links, then full-page links
    Match against multilingual glossary terms
    ↓
Classify: has_statement / found_in_footer
    ↓
Save to url_accessibility_results table
    ↓
Write *_accessibility.toon output file