DIGITARISE
esc

6 of 55 indexed · ↑↓ to move · ↵ to open

Government portals are accessible. They are not private.

We measured 44 of 51 US state and District of Columbia portals. They expose a main landmark more often than any private sector we have measured — genuine accessibility engineering, consistent with the obligations that bind them. Yet 82% ship a recognised tracker while just 2% surface any consent mechanism at all.

The numbers

n = 44
386 msmedian server response

Half of the 44 measured homepages return their first byte within this time; the slowest takes 1552 ms.

76 KBmedian HTML delivered

Size of the delivered document before any subresource loads. 36 of 44 serve it compressed.

82%ship a known tracker

36 of 44 homepages reference a recognised analytics or advertising tracker in the delivered HTML.

12median third-party domains

The median homepage references this many distinct external domains before any interaction.

2%surface a consent mechanism

Only 1 of 44 expose a detectable consent platform — yet 21 set a cookie before any choice is possible.

89%expose a main landmark

39 of 44 provide a main landmark for assistive technology; 8 ship no first-level heading at all.

The observatory

44 portals · Rev A · 2026-07-26
Have main landmark39/44
Single h126/44
Declared lang44/44
  1. 01California100/100checks5/5alt100%h11
  2. 02Florida100/100checks5/5alt100%h11
  3. 03Georgia100/100checks5/5alt100%h11
  4. 04Hawaii100/100checks5/5alt100%h11
  5. 05Idaho100/100checks5/5alt100%h11
  6. 06Missouri100/100checks5/5alt100%h11
  7. 07Montana100/100checks5/5alt100%h11
  8. 08Nebraska100/100checks5/5alt100%h11
  9. 09Oregon100/100checks5/5alt100%h11
  10. 10Nevada99/100checks5/5alt96%h11
  11. 11Wisconsin99/100checks4/5alt93%h11
  12. 12Mississippi98/100checks4/5alt89%h11
  13. 13North Carolina97/100checks4/5alt86%h11
  14. 14North Dakota97/100checks4/5alt83%h11
  15. 15Pennsylvania97/100checks4/5alt83%h11
  16. 16Maryland96/100checks4/5alt80%h11
  17. 17Oklahoma96/100checks4/5alt78%h11
  18. 18Vermont96/100checks4/5alt79%h11
  19. 19Connecticut93/100checks4/5alt67%h11
  20. 20Tennessee93/100checks4/5alt67%h11
  21. 21Illinois89/100checks4/5alt47%h11
  22. 22Alaska87/100checks4/5alt37%h11
  23. 23Iowa86/100checks4/5alt31%h11
  24. 24Colorado80/100checks4/5alt100%h12
  25. 25Delaware80/100checks4/5alt100%h13
  26. 26Indiana80/100checks4/5alt100%h18
  27. 27Michigan80/100checks4/5alt100%h12
  28. 28New Jersey80/100checks4/5alt100%h12
  29. 29West Virginia80/100checks4/5alt100%h10
  30. 30District of Columbia79/100checks3/5alt94%h11
  31. 31Kentucky79/100checks4/5alt95%h11
  32. 32Washington79/100checks3/5alt93%h10
  33. 33Louisiana75/100checks3/5alt75%h10
  34. 34Kansas74/100checks3/5alt71%h11
  35. 35South Carolina74/100checks3/5alt68%h13
  36. 36Rhode Island73/100checks3/5alt67%h10
  37. 37Alabama67/100checks3/5alt33%h10
  38. 38Texas67/100checks3/5alt33%h12
  39. 39South Dakota65/100checks3/5alt27%h12
  40. 40Ohio64/100checks3/5alt20%h12
  41. 41Utah63/100checks3/5alt14%h10
  42. 42Virginia60/100checks3/5alt100%h12
  43. 43Wyoming60/100checks3/5alt0%h10
  44. 44Maine58/100checks2/5alt89%h10

44 of 51 US state (and DC) government homepages · measured 2026-07-26 · score = transparent composite of the raw signals in each row

Switch the dimension to re-rank the sector. Download the dataset (JSON) →

Against the private sector

identical method

The same measurement script produced every sector here, which is what makes the comparison legitimate. Switch to the consent metric to see the gap that matters most for public services.

Share exposing a detectable consent platform. Higher is the better posture — the visitor is at least asked.

Healthcaren = 2924%
Legaln = 3263%
DTC ecommercen = 2167%
Governmentn = 442%

DTC ecommerce holds the best posture here — 67% against 24% and 63% and 2%.

Higher is better · all sectors measured with the same script and method · 2026-07-20

Instrument — every measured sector, one metric at a time.

Analyses

sector sheet

Public services measured on public, observable signals — one request each, no JavaScript executed.

Read the methodology