Accessibility debt and usability erosion
Accessibility problems rarely arrive all at once. They accumulate, they compound, and the cost falls on the people least able to absorb it.
What accessibility debt is
Accessibility debt is the build-up of unresolved usability and accessibility problems that raise the cost of future remediation and, in the meantime, transfer that cost to users. Like technical debt it is invisible in short-term metrics, and like technical debt the interest compounds.
Cognitive distance explains the mechanism. Each decision that privileges internal logic over user understanding widens the gap slightly. No single decision looks like a defect. The accumulation is the defect.
Why compliant systems still exclude
Accessibility work has focused, productively, on perceptual and physical barriers: contrast, keyboard access, screen-reader compatibility. Those remain essential and they are not the whole picture.
A system can satisfy every one of those checks and still be unintelligible. When instructions are vague, workflows inconsistent and system states hard to interpret, users with cognitive impairments face a barrier that no audit tool reports. The system is usable in a narrow sense and not understandable in any sense that matters.
This is why cognitive accessibility resists checklist treatment. Understanding is not a property you can assert about an interface in isolation; it only exists in relation to a particular person's model of it.
How erosion happens
Usability erosion is the slow version of failure. A system ships coherent, then accumulates features, exceptions and special cases, each locally reasonable. The conceptual model that made the original version learnable stops describing the current one.
Users adapt before they complain. They memorise sequences, keep private notes, ask the one colleague who understands the billing screen, and avoid the parts they cannot predict. Because task completion rates hold up, dashboards report health. The erosion is only visible if you ask whether people can explain what the system did.
Who pays
The cost is borne disproportionately by users least able to compensate: people with cognitive disabilities or memory limitations, neurodivergent users, older adults, and anyone under stress or time pressure. For these groups distance converts directly into exclusion, dependence on others, or error.
That distribution is what gives accessibility debt an ethical dimension rather than merely an operational one. Framing it as debt makes the transfer explicit: an organisation that defers this work has not avoided the cost, it has moved the cost onto the people with the least room to absorb it.
Paying it down
Evaluation has to change first, because you cannot repay a debt you do not measure. Traditional usability testing captures task success and can miss long-term misunderstanding entirely. Assessing cognitive distance means observing learning, recall and transfer across sessions, not just completion within one.
Useful proxy indicators already exist in most systems: error-recovery patterns, reliance on support channels, repeated visits to the same help article, and the gap between what users predicted would happen and what did. None require new instrumentation. They require deciding that understanding is a metric.