The web is getting less accessible to people with disabilities

And that’s a miserable thing during the current COVID-19 crisis if you are one of them or close to someone who is

One thing coming out of the current COVID-19 pandemic is that the lack of web accessibility is really hurting people with disabilities, especially those who use assistive technology. In the past two weeks alone, I’ve seen:

• Government agencies and NGOs releasing completely inaccessible COVID-19 materials.
• Children being discriminated agains in home schooling in the US.
• Stories about people with severe disabilities ending up in hospitals when their personal care assistants were blocked from coming to their homes.
• Stories about people who are blind not being able to order groceries online.

Online shopping used to be a convenience.

Now, it is a necessity.

And if your online shopping is not accessible, you are excluding and DISCRIMINATING AGAINST your customers with disabilities.

Little of this is surprising if you are in the accessibility field, and the statistics back up the anecdotal evidence. WebAIM came out with it’s second annual report on the accessibility of the top million webpages, and the news was not good for people with disabilities, while simultaneously being GREAT news for the 10 lawfirms who file 82 % of the ADA lawsuits in the United States. You can read my article on the inaugural report here.

According the 2020 report:

• The number of accessibility errors increased 2.1% with an average of 60.9 errors per page across the one million home pages analyzed. Let’s break that down. WebAIM says their WAVE tool detects between 25 and 35 % of errors. Let’s call it 30 % for argument’s sake (that is the number I typically use when talking about WAVE, which my employer uses). That means there are likely over 200 accessibility errors per home page if the rate of errors from manual testing is the same rate as the rate of errors found in automated testing. In my experience, the rate of errors found in manual testing is usually higher, not lower than, the rate of errors found through automated testing.
• Home page complexity increased 10.4% to 864 elements. That means a user is likely to run into an accessibility error in one out of every 4.3 elements.
• 98.1% of home pages had automation detectable accessibility failures, up from 97.8% in last year.
• Interior pages were almost as bad with a slightly lower error rate but a 20 % higher number of components.

The word appears to be getting (somewhat) out about the need for alt-text. 66 % of home pages in 2020 had missing alt text, an improvement from 68 % in 2019. However, all other basic components whose errors can be detected in an automated manner got worse. Low contrast text, empty links, missing form input labels, empty buttons and missing document language statistics all got worse in 2020 over 2019. One in every 20 headers was badly structured. Even though interior pages had more elements, they had fewer headers. Interior pages likely should have had more headers, since more components == more things to navigate == more headers.

A surprising statistic that continued from last year is that home pages containing ARIA averaged 60% more errors than those without ARIA. Using ARIA implies you know something about accessibility? Yet it correlates to worse results. To paraphrase Yoda — in accessibility, do or do not, there is no try. There are no points granted to defendants in accessibility lawsuits for trying. If anything, using ARIA and having more errors than average makes an organization look worse — it confirms they KNEW enough about accessibility to put ARIA code in, but didn’t care enough to do a comprehensive job at it. That is not a message organizations want to facing in a legal setting.

When looked at by category, the “best” (where “best” is defined as “least worst”) site categories were:

• Law, Government, and Politics
• Technology and Computing
• Business
• Society
• Careers
• Health & Fitness
• Science
• Personal Finance
• Education

And the worst categories were:

• Home and Garden
• Family and Parenting
• Pets
• Food and Drink
• Religion and Spirituality
• Travel
• Automotive
• Hobbies and Interests
• Arts and Entertainment
• Sports
• Style and Fashion
• Real Estate
• Shopping
• Adult Content
• News/Weather/Information

What can we learn about the information reflected by the categories? Regulations work. Most of the categories that had fewer errors than average are in topics that are either highly regulated (government, education, health, banking) or have the peaks of their lawsuits in the past (personal finance). And not suprisingly, most of the categories that have more errors than average are current hotbeds for litigation.

Another area of interest is the default language for the page. English, Dutch, and Japenese sites had fewer defects than average, while sites in Russian, Spanish, German, French, Portuguese, Italian, Chinese, Polish or Farsi had between 2 % and 61 % more errors than average.

And, finally, *ANY* page that included ads had higher than average defect counts, didn’t matter who the ad server was. With the inexplicable statistic that pages using Google Ad Sense (where Google is know for being a company that claims they are committed to accessibility) had 60 % more defects than average.

If an organization is voluntarily choosing to enter into a contract with any third party, such as providers of ads, cookies, mapping, search, zip codes, or surveys — The organization doing the choosing becomes liable for the third-party vendor’s inaccessibility. While an organization may be able to recover damages against the partner, both parties will likely get sued. Practice safe accessibility — choose your partners carefully !

---

The winners here are the 10 lawfirms that file 82 % of the ADA lawsuits in the United States. These firms largely use WAVE to find their next victim. WAVE results have frequently been seen as part of the filings in the initial complaints. There is not just low hanging fruit here, there is fruit rotting on the ground for the plaintiff’s trolls to make money from.

Easy pickings, unfortunately for the organizations who continue to ignore the warnings.