4. Equivalent Experience

What functional and intrinsic do not say

Functional accessibility — page 2 — gives a floor: at least one path through must succeed. Intrinsic accessibility — page 3 — gives a breadth: the underlying interface must admit many such paths for many user-and-context profiles. Both definitions count successful task completion. Neither says anything about what the successful completion cost the user who completed it.

That silence is deliberate: cost-of-experience is a different question from existence-of-path, and conflating them produces an analytical mess. But the question of cost cannot be left out of a serious account of accessibility, because the entire field has a long-running pattern of declaring the question answered the moment a path exists, with no further attention to what using that path is actually like. That declaration is wrong, and naming why it is wrong is the work of this page.

The timing dimension

The most measurable cost is time. A sighted user completes a checkout flow in forty seconds; a screen-reader user completes the same flow in eight minutes; both interfaces are functionally accessible by the floor definition. The difference is twelve times the effort spent on the same business transaction. The screen-reader user pays that difference every time they buy something online, and the sighted user does not.

The temporal asymmetry is rarely accidental. The ordinary path through an interface is the one the designer optimised for; the assistive path inherits whatever the designer left over. Tab-order issues, redundant ARIA-region announcements, dynamic content that pulls focus away, modals that trap keyboard users, focus-loss after AJAX updates — each adds a few seconds. The cumulative effect, measured honestly, is a hidden tax on the user whose only accessible path is the assistive one. The interface can be functionally accessible and tax its minority users at a rate that no commercial product would tolerate if it were a minority of revenue.

Equivalent-experience analysis names that asymmetry and asks the explicit question: are users on different paths paying comparable amounts of effort for comparable outcomes? The asymmetry is measurable. The ratio is not subtle. A definition of accessibility that cannot ask this question is a definition that cannot tell you what is wrong with an interface where the screen-reader path takes ten times longer than the visual one.

The equality argument

The lede of this page is the political claim underneath the temporal one. The claim is not that every user must have the same experience — experiences differ, and the differences are not themselves the problem. The claim is that every user has the same right to the goods and services on offer, and that the cost of the difference in experience should not be paid systematically by the users whose bodies fall outside the designer’s default assumptions.

Two readings of that claim are common, and one of them is wrong.

The wrong reading treats the claim as a demand for identical experience. Identical is not what the claim asks for. A blind user cannot have the same perceptual experience as a sighted user; the difference is not the accessibility problem. Equivalent-experience analysis is about whether the functional outcome— the goods, the services, the participation in society — is available to both at comparable cost in effort, attention, and time. The shape of the journey can differ; the accessibility of the destination is the question.

The right reading treats the claim as a demand for equality of access. The user who is blind, the user with a tremor, the user navigating in a noisy environment, the user with low literacy in the interface’s language — all of them are members of the society the interface serves, and all of them are entitled to participate at comparable cost. The equality is in the entitlement, not in the perceptual or motor sequence by which the entitlement is exercised.

Why utilitarianism is the wrong frame

The argument deserves to be made explicitly, because utilitarian framings are widespread in accessibility practice and largely unexamined. The shorthand version: we are doing the greatest good for the greatest number, so we are right to optimise for the majority case and treat the minority as a residual accommodation.

The framing is appealing because it reads as hard-headed and as resource-aware. It is also wrong, for two reasons.

The first reason is structural. Utilitarian logic evaluates an action by summing utility across a population. To compute the sum, you have to be able to add the utility you give one user to the utility you give another. Inter-personal utility comparisons are notoriously contested even in the philosophical literature that takes utilitarianism seriously; in accessibility practice, they are typically not attempted at all. What replaces them is a simple numerical headcount — the majority outvotes the minority — which has nothing to do with utility maximisation and everything to do with the convenience of the designer.

The second reason is categorial. Accessibility is not a question about how to allocate utility; it is a question about who is recognised as a member of the society the interface serves. A user who is excluded from a banking app because the screen-reader path never worked is not someone whose utility is too low to bother optimising for; they are someone the designer treated as not-a-member. The harm is not that they got less utility; the harm is that they were not in the population the designer counted in the first place. Utilitarian framings cannot represent that harm because their operative vocabulary is summing across members, and the harm is precisely the failure to recognise membership.

The right vocabulary is rights, not utility. Rights attach to membership; membership is not earned by being in the majority; the failure to provide accessible access to a recognised member is a violation, not an unfortunate trade-off. That framing is what the lede captures. It is also why the legal vocabulary (which is rights-shaped) and the utilitarian vocabulary (which is sum-shaped) cannot be reconciled into a single hybrid: they are answering different questions about different objects.

What “equivalent” permits

Equivalent-experience analysis is not a demand for perceptual sameness, and recognising this matters in practice. Some examples:

• Different sequencing. A screen-reader user often navigates a page non-linearly via heading or landmark navigation, while a sighted user scans visually. The two routes are different; neither is wrong. Equivalent-experience analysis accepts both as long as both arrive at the same outcome at comparable cost.
• Different modality. A deaf user reads transcripts while a hearing user listens. The modalities are different; the content can be equivalent. The analysis does not insist that everyone listen, only that the content be conveyed.
• Different affordances. A motor-impaired user may invoke an action via a single switch press rather than a click-drag-release; the action sequence is different; the action is the same. Whether the affordances are equivalent depends on whether the cost of using the alternative path is comparable to the cost of using the default.

What the analysis does not permit is asymmetric cost. A screen-reader path that works but takes ten times as long is not equivalent; it is a tax. A switch-access path that requires twenty stops to invoke an action a mouse user completes in one click is not equivalent; it is a tax. The analysis is comfortable with shape-difference; it is not comfortable with cost-difference at scales that fall systematically on the minority user.

What this leaves to the next two pages

Pages 1 through 4 establish the position: the question is political; functional and intrinsic accessibility are the formal measures of two distinct properties; equivalent experience is the cost-aware analysis that holds them honest.

Pages 5 and 6 turn from definition to construction. Page 5 treats the methodological substrate — the recursive-design framing that makes the formal definitions buildable rather than just stated. Page 6 reframes inaccessibility itself as a kind of community dysfunction, and opens onto the multi-agent and game-theoretic territory the 2029 framework resumes.

Continue

← Previous: 3. Intrinsic Accessibility · Next: 5. The Shlaer-Mellor lens →

See also Accessibility as a property of the dialogue, not the device — a complementary 2006 reframing, with the coverage-set formulation and the commercial-and-legal version of the equality argument.