Inaccessibility in the NZ government's COVID-19 social media
As the new year 2021 began, the NZ COVID-19 group posted a new video on its social media channels, particularly Twitter. The video included flashing images and strobe effects and motion, which could induce seizures, migraines and other adverse reactions. The video was posted without a warning. For safety, the original tweet and the @covid19nz Twitter profile are not directly linked here, but a still-image screenshot is reproduced below:
![Screenshot of a tweet by covid19nz, dated 11.55 PM Dec 31 2020, with the text, Thank you for everything [hashtag] team of 5 million and Happy New Year](https://i.snap.as/sXriTlOZ.png)
It is not uncommon for such inaccessible media to appear on the web, but it is interesting that this particular video is published by a government body — an arm of the Ministry of Health, which one might expect would support Disabled people better.
So how bad is it? Why did this happen?
And what did they do after being informed? Why didn't that make things better?
Timeline
Let's establish the sequence of events, at least as I saw it:
Video posted on Twitter by @covid19nz.
I saw the video for the first time.
I posted a thread in response, on my timeline, tagging @covid19nz, highlighting accessibility problems.- Note: I avoided engaging directly with the original tweet so that I would not spread it into more people's view through reply activity.
I caved in and replied directly to the tweet (risking adding to its spread on Twitter).
@covid19nz posted a reply to their own tweet appending a warning.
Then @covid19nz replied to me saying that a warning was added and captions would follow later.
I looked for the warning and could not find it (I expected it to accompany the video, appearing before it, but there was nothing).
I replied asking where I could find the warning. I got no response.
I looked up @covid19nz's profile and found their separate reply including the warning.- The warning looked defective: it appeared after the hazard, it used language against recommendations, and it contradicted itself (saying the video was, “not for social media”).
I replied in multiple posts objecting to the warning and its defects, as well as on my own timeline with additional detail as to why.- As of this blog posting, I received no response and the video remains posted with the defective warning.
What's the big deal?
The timeline above asserts a few things that should be elaborated. What was the problem in the first place and why is it serious? Why was the warning reply defective? What should @covid19nz have done, and what should they do now?
Hazard
The major problem, to recap, is that the video could be hazardous. It presents images that flash, strobe or move, in ways that could induce a seizure or migraine or similar reaction in some people. Films and video games often have these problems, and in various jurisdictions, they may be required to present a warning before the hazardous content to allow affected people to opt out of viewing it. Here, no warning was given.
Surprise
The hazard is made worse by the fact that the video is posted by a trusted public source, @covid19nz, an informational government account that usually posts key public alerts and warnings about COVID-19. The risk posed by the COVID-19 pandemic is of special concern to Disabled people, so they are likely to follow the account for meaningful information. There is no expectation that @covid19nz will publish hazardous material on such a channel, so the video is likely to take people by surprise. A prior warning, therefore, is of utmost importance.
The video itself is designed to be misleading: it starts off like a regular video with Ashley Bloomfield addressing the audience with his usual greeting. It then transitions into an unusual music video sequence, in an intentionally surprising way. This bait-and-switch trick is almost malicious, let alone negligent, given how hazardous the content can be.
Without any prior warning available, a viewer has to watch the video to discover whether it is unsafe, or important, or not. It inherently exposes them to risk before they can determine it is risky. A prior warning would allow a chance to opt out of watching the video. Alternative content (such as a text post) would give those who have opted-out a way to receive the same content and judge whether it is important to them.
Generally inaccessible
There are even more problems with accessibility because the video does not include captions, a transcript, or alternative content, for Blind or Deaf users. The COVID-19 press briefings were well-known for the sign language interpreters typically present. However, that minimal standard of accessibility does not seem to extend to social media or even the COVID-19 tracer app. The hazardous video posted now is part of this odious trend.
The law
There are legal and policy requirements for government digital content to be accessible. As a State Party, @covid19nz is obliged to publish material with reasonable accommodations for equal access to information. The hazardous video, also lacking captions, transcripts or alternative content, does not meet the legal or policy standard for a variety of disabilities.
Defective response
The response by @covid19nz was to post a reply to their own original post, to include a warning after the fact. This is useless. Why? Because — and this may seem obvious when I say it — a warning should come before a hazard.
This kind of after-the-fact warning is particularly defective on Twitter, because the video is generally visible without replies being shown. That is, if any user retweets, likes or replies to the original post, only that original post and the user's interaction are shared to their followers. Other replies in a thread are not shown. So the warning becomes detached from the hazardous content.
Pinning the tweet
It is especially egregious in the case of the @covid19nz Twitter profile page, where they have chosen to pin the hazardous tweet. That means visitors to the @covid19nz page will be exposed to the video even if they do not follow the account. And the video will be visible without the warning in the replies, because the reply does not get pinned.
A lack of understanding and standards
Incidentally, another hint is their wording, “… suffer from migraines and headaches”. The phrase, suffer from, is one that digital.govt.nz advises against in their very clear guidance on disability language. It is evident that the social media operators behind @covid19nz have not been through any training on digital accessibility or disability inclusion.
The final clue is the self-contradictory excuse that @covid19nz posted in their warning: that the video was “not for social media”. If that is the case, then why does it appear on social media? Regardless of the intent of the video, the impact is that it will appear via social media on the screens in front of affected users, because — and this must also seem obvious — @covid19nz posted it on social media.
If it is not accessible enough for social media, why are there no standards in the COVID-19 group or the Ministry of Health to stop it being posted?
Let me repeat that these failures are occurring in a body within the Ministry of Health — a ministry that, as a primary concern, nominally caters for people with disabilities and chronic illnesses.
The legal and policy obligations demand that the government, including the Ministry and COVID-19 group, should promote accessibility also to urge private entities and mass media to adopt better accessibility practices. @covid19nz is required to be an exemplary leader in publishing accessible content. It is not doing so.
So where to from here?
Solutions
The immediate problem is that a potentially hazardous video is visible without a prior warning. The original post must be deleted.
The bare minimum: re-post with a text warning
At a minimum, a re-post may be acceptable if it includes at least a text warning in the content of the tweet. That may be suitable because most seasoned Twitter users who need to avoid hazardous flashing or strobe or motion in videos may have switched off autoplay.
The responsible choice: re-post with a warning in the video
But a text warning is not enough for the high standard required of @covid19nz as a State Party which must be exemplary in its promotion of digital accessibility. So @covid19nz should also edit the video to include a leading title screen with a warning, allowing enough time for users to pause the video before the flashing content starts.
Given that the video is designed to be surprising, with Ashley Bloomfield starting off speaking normally, then adding a warning screen at the start could be a spoiler. So be it.
The video was clearly produced with a substantial budget, so it is reasonable to expect some of that budget to ensure accessibility, such as producing a warning title screen.
One benefit of adding a title screen with a warning to the video itself is that it will work even in cases of video autoplay. Not everyone who will have an adverse reaction to this kind of moving image will already know their condition or have a diagnosis or a history of reactions. So it is useful to inform users before they consent to the risk.
Another benefit of a title screen with a warning is that it would be portable across social media platforms and channels, such as Instagram (where @covid19nz has also posted the video). A text warning in a tweet only covers the tweet, but a title screen at the start of the video is likely to be usable everywhere.
It's not too late
The @covid19nz social media response to this ongoing accessibility problem has been reluctant and ineffective. It took about twenty hours before I, a member of the general public, noticed the post and reported the problem. It looks like it will take longer for @covid19nz to abandon their post's social media high-score to re-post it. In the meanwhile, the hazard is visible to users without a proper warning. This is unacceptable.
But does it mean it is too late to do anything? The second-best time to do the right thing is now. Remedying the problem in the proper way would set an example for future publications by @covid19nz, but also the digital sector as a whole, which State Parties are obliged to lead on.
The future
For next time, it would seem there is a serious gap in the way the COVID-19 group approaches accessibility, especially in social media. The digital.govt.nz guidance and standards does not seem to have penetrated the minds of the Ministry of Health and the COVID-19 group. Perhaps training in this area is in order.
Whatever internal standards and policies the COVID-19 group has for its social media publications seems to be deficient. Perhaps an overhaul of these rules is in order.
Meanwhile, it appears that there will a next time, because @covid19nz will be producing more videos:
Let's hope they can pivot their accessibility strategy in time.
Conclusion
The NZ COVID-19 group related to the Ministry of Health has shown time and again that it views accessibility and disability as a second-class concern. This occasion of the hazardous video being posted without a suitable warning is yet another example. It is important to reverse the trend by rectifying the COVID-19 group's disableist approach wherever possible. Fixing the warnings on this video is as good a place as any to start, in anticipation of the year to come.