Is Accessibility a Supply or Demand Issue?

Jim Tobias has me thinking….Is the 508 Refresh delay a Supply (Industry) or a Demand (Government) issue? For years I’ve been advocating for economic stimulus and price subsidies – Federal/State sponsored rebates, grants, and refundable credits. Long term investment in the underlying technology, open source systems, applications and architecture to improve cost structure thereby generating organic market competition.

My position (okay, my rant!) has always been that industry and big business won’t invest in accessibility if there’s no business value proposition. Talk to any business – anywhere –  ask them where accessibility lies in their list of product/service priorities. Accessibility doesn’t make the top 10 list, ever. Lots of great intention; a few ‘feel good’ marketing messages; an upbeat Tweet now and then. But let’s face it, industry is in the business of making money. The accessibility business today is about compliance; it’s viewed as a cost of doing business, not one that generates revenue, never mind profit!

I strongly believe that economic stimulus is crucial to getting our industry — the accessibility industry — off the ground. I want to jump start it just like the Feds did for green technologies and the alternative energy industry.

But Jim really has me thinking about this Supply and Demand thing….If the Feds and States and all those agencies REALLY stuck to their guns and followed through on the fundamental principal of 508 procurement enforcement……if the Feds and States FUNDED Section 508 rather than starving it as an UNFUNDED mandate…..if xthe Feds and States INVESTED their its own infrastructure….well then….just maybe….just maybe we could get this vision of EMPOWERMENT truly off the ground….

…..I’m not sure….what do YOU think???

Categories: Accessibility Strategy, Development


Ricky Onsman says:

If every Government agency – everywhere – refuses to buy goods or services from providers who do not meet clear accessibility requirements (not guidelines, requirements), then accessibility becomes a business issue rather than a feel-good human-rights issue. Even providers who don’t pitch to Government will adopt accessibility as a matter of course, a principle of doing business. Right now, the “ifs” implicit in that are manifold. If all Government, if they refuse, if there are requirements rather than guidelines, if those requirements are clear and able to be implemented without significant cost … The model I wonder about is Responsive Web Design. As devices proliferated and their use extended to all forms of web contact, RWD went from unknown to totally accepted in four years. with almost no angst, no arguments about cost. Even if RWD is not implemented, everyone accepts that some form of accommodating the range of devices is necessary. How do we do that with accessibility?

Jim Tobias says:

Thanks for mentioning this, Mike. I think we all agree that accessibility, like any other business issue, is always about supply and demand to some extent. What I meant is that at this moment the supply is pretty good — more built-in accessibility in mainstream products, and pretty mature solutions available to developers and content creators on almost all platforms. It’s the demand side that’s weak right now: low adoption of even free AT solutions by people with disabilities, and low insistence on accessibility features by those who purchase and manage technology in large enterprises. There are plenty of ways to stimulate demand and clarify market signals about accessibility without additional subsidies or regulations. One way is to develop better tools for procurement, so CIO staffers don’t have to become such experts; GSA’s Section 508 Quick Links is an excellent example that should be adopted and extended. Another is to widen the targeted dissemination of information resources about accessibility so that consumers who are in the market can learn about products that have the features they need; the GARI database is a fine example there. My point is that there’s plenty we can do without passing new laws or funding new programs.

Mike Paciello says:

Jim –

Thanks much for clarifying the Demand/Supply situation and, more importantly, how critical tools like Quick Links and GARI can help stimulate the demand. Clearly we need to increase awareness of these tools and the need for more. However, I still wonder whether this is enough “pull” to “push” the accessibility economy into a business value proposition that increases its level of prioritization within industry.

Let’s face it, I’ll never be satisfied…

Mike Paciello says:

Ricky –

Thanks for your input. Certainly demand generated by government adherence (enforcement of) to their own procurement laws will ensure greater accessibility for government and citizen-focused services. I wonder how well this thinking works in the EU and in Australia?

Your thoughts regarding Responsive Web Design remind me of the long standing position involving universal design. Clearly design that is truly user-centric is valuable and could lead to a generation of inherently accessible interfaces. RWD and Apple’s Universal Accessibility architecture are noteworthy models.

H.E.A.T. says:

Firstly, this is a thought-provoking article that happens to hit the subject of accessibility from two distinct perspectives.

I actually see accessibility (as it exists today) as an issue…period. No need to point fingers at the extremes; we must first target the source of the problem.

It is said when things are messed up, it usually begins at the top.

HTML 5 and ARIA 1.0 are specifications coming out the W3C’s camp. So, I go to and guess what I see? I land on a page still coded with XHTML 1.0 with no ARIA nor even RDFa.

This is the issue. If the folks who publish these specifications do not feel the importance or need to apply them to their website, then why should anybody else?

Be honest with yourself. Gourmet cooks will taste their own food before serving it to others. Why should we expect less from the self-proclaimed stewards of the web?

First, HTML5 (even though I feel it is a far inferior tech than pure XHTML2 would have been) offers more semantic information than XHTML 1.0 by itself.

Secondly, ARIA 1.0 took so long to reach recommendation status that the entire website could’ve been retrofitted to use it by now.

Thirdly (and this is the problem), shouldn’t the W3C be accessible to all those looking for accessibility information, including those with disabilities?

Yes, I know there is a WAI site, but I am speaking about those who exist at the top of all this accessibility stuff.

Or is the W3C simply employing a “do as I say and not as I do” attitude toward the rest in the web community?

I am not trying to come off as a jerk (and I really could care less), but I am trying to make a point about how when things are messed up at the top, it can trickle down to others.

Take a look at this site, When I look at the source code for this comment box, I see that you are using the @aria-required=true on the TEXTAREA element. Good, right?

But wait. Steve Faulkner and Hans Hillen, two of your own members, stated in the first rule of using ARIA in HTML5 to use native attributes when possible. So, where is @required=required?

Again, another example of not applying a specification to your own stuff.

So, when folks in the trenches (the wild, I guess) hear all this parental mumble jumble about the importance of accessibility and see it not being used by the W3C on their own precious website and not being used by so-called professionals in the area of accessibility, this raises cognitive dissonance.

Cognitive dissonance is actually a cognitive disability being produced by the W3C and imposed on others.

Looking to the extremes when there is a clean mirror right in front of you is never a good thing to do.

If accessibility is such a good idea, if it is so important that we may need governments to step in and doing something to force…I mean…encourage others to apply to their websites, then the folks at the top need to invest in themselves and get their acts together as well.

Design like we give a damn, right? Lead by example.

Patrick H. Lauke says:

So, I go to and guess what I see? I land on a page still coded with XHTML 1.0

HTML5 does not “replace” XHTML 1.0. And, particularly in the case of a large existing corpus of content, if your documents don’t make use/have a need for any of HTML5’s new features, there is no practical advantage (or indeed need) to change those documents (and it would simply be a case of swapping out the DOCTYPE in most cases).

Take a look at this site, When I look at the source code for this comment box, I see that you are using the @aria-required=true on the TEXTAREA element. Good, right?

But wait. Steve Faulkner and Hans Hillen, two of your own members, stated in the first rule of using ARIA in HTML5 to use native attributes when possible. So, where is @required=required?

The required attribute and aria-required are not completely equivalent. The former also triggers automatic validation and built-in form submission error messages in browsers – which are currently not 100% accessible – while the latter only signals that a form field is required to assistive technologies (without any further validation/error bubbles being triggered in browsers). So, the decision is a conscious one. This is a nice example of how accessibility is more than simple adherence to technical standards. Accessible development in the real world requires nuanced, pragmatic decisions on when to apply certain technologies, and when to look for alternatives.

Steve Faulkner says:

also the required attribute as implemented in browsers exposes both the required and invalid states to AT by default, for example in Firefox – Bug 666544.

H.E.A.T. says:

HTML5 must be a replacement for XHTML 1.0 because the W3C decided to shelve XHTML 2 because of HTML 5 (and its XML/XHTML syntax feature). This is documented already and does not require any further discussion.

Besides, HTML5 is being touted as the next step in marking up the web: a core product. has elements marked up with IDs as header, main, and so forth. HTML5 provides these element in full semantic form.

The argument around using HTML or XHTML seems to always come to “if you don’t need it, don’t use it.” Guess what? I agree. I definitely agree.

But the is not simply another website. represents the W3C as a public-facing demonstration of web stewardship. If for any other reason, they should lead the way by setting the example.

Also, whatever maladies that exists between the HTML5 specification and implementation by browsers could’ve been caught and addressed firsthand within the same consortium that has full access to editors and the browser vendors.

Look, if I a car maker, I am going to drive my top-or-the-line car to push the product and to support my own brand. Not just that. There’s also another more important reason the W3C should use HTML5 on its website.

When HTML5 became a Recommendation, it was automatically placed 10 years behind the power curve. With all the politics, organizational restructuring, public arguments amongst consortium members, and the like, HTML5 had already aged so much by the time it was fully approved.

So, XTHML 1.0, XHTML 1.1, working on XHTML 2…and now, HTML 5 looking at HTML 5.1. is due for a major face-lift here in 2015.

The arguments about the @required attribute does not seem to hold water. If the attribute exposes both the required and invalid states in AT by default (@Dr. Faulkner) and triggers automatic validation and error messages in browsers (@Mr. Lauke), then the attribute is definitely ready for prime time usage.

ARIA can also be used to supplement native semantics. One can add @role=alert to the SECTION element to make the semantics even stronger.

The point of all this is that the W3C sits at the top of the web world as developers and stewards. Any new tech coming out of that camp should be tried and tested in their own web pages to identify implementation problems before pushing the product out to the world. If problems are found and cannot or will not (in the WHATWG’s case) be resolved, then simply do not specify it.

Design like we give a damn is not the same as “Design like we have a choice.” Lead by example.

Steve Faulkner says:

Your comments have nothing to do with the subject of the article. Please refrain from continued trolling, thanks in advance 🙂

Patrick H. Lauke says:

The arguments about the @required attribute does not seem to hold water.

Only if you don’t actually understand the issue, as outlined. If it triggers in-browser validation which, in itself, is not accessible, and if by default it flags an empty field immediately as “invalid” to assistive technologies, before anything has actually been entered, then that’s not currently an acceptable solution to use.

Design like you understand when a technology is fit for purpose, and when it currently isn’t. Design with an understanding of what impact your technology choice has on your actual users.

As for the rest of your arguments about what the W3C should or shouldn’t do (combined with your slightly outmoded view of what the W3C actually is, other than a consortium of big companies like Google and co)…I suggest you take it up with them.