The accommodation tax is a myth: neurodivergent-first design helps everyone (and the honest cases where it doesn't)

There’s a frame that shows up in nearly every conversation about designing for neurodivergent people — in product roadmaps, in HR policies, in school IEP meetings, in the backlog grooming session where someone asks whether the team can “afford” to make the documentation accessible. The frame is simple, often unstated, and almost never examined: designing for the minority costs the majority something. Call it the accommodation tax. The belief that every accommodation is a line item on the majority’s ledger — slower processes, plainer interfaces, less polished experiences, fewer features shipped — and that the neurotypical default subsidizes the deviation. This frame treats the default as neutral and the accommodation as a deviation with a price. It is the frame that underwrites every version of “we’d love to design for everyone, but we can’t afford it.” It is also, in the cases that matter most, a myth. But — and this is the part most curb-cut writing skips — it is not always a myth, and the honest version of the argument says so.

The myth, stated plainly

The accommodation tax frame works like this. There is a default — a workplace, a product, a classroom, a code review process — designed around the modal user. The modal user is neurotypical, processes sensory input within typical tolerances, communicates in conventionally indirect ways, and accesses information through the channels the majority uses. This default is treated as neutral: frictionless, costless, the natural state of things. Then someone needs a deviation. An autistic employee needs the PR template to say exactly what’s expected instead of relying on tacit norms. A neurodivergent learner needs the lecture recorded and captioned. An ADHD employee needs the open-plan office to have a quiet zone. And the frame kicks in: this costs something. It costs dev time, it costs policy complexity, it costs the team’s default rhythm. The majority pays.

This is the frame that lives underneath “reasonable accommodations” as a legal concept — the language itself encodes the assumption that the accommodation is a concession extracted from a reluctant majority, measured against what’s “reasonable” for them to bear. It is the frame that makes accessibility a budget line, a compliance obligation, a favor granted. And it is the frame that makes builders defensive, because it positions every inclusive design choice as a tax on their velocity.

The problem is that the frame is built on a false premise: that the default is neutral.

The curb-cut effect, stated honestly

The curb-cut phenomenon is one of the most cited analogies in accessibility writing, and for good reason. Sidewalk curbs were cut at intersections to accommodate wheelchair users, and the result was that strollers, delivery carts, cyclists, luggage rollers, and people with temporary injuries all benefited. The accommodation designed for the edge case turned out to serve a broad population the designers never targeted.

There is a real cognitive analogue, and it’s worth being specific about it rather than gesturing at it.

Design the pull request for autistic literal communication — explicit norms about what comments mean, structured templates that say “this needs changes because X” instead of “maybe think about whether this is the right approach,” a convention of not reading into tone — and the neurotypical reviewer also benefits. The neurotypical team member who was guessing at what a vague comment meant now has a comment that says what it means. This isn’t charity for the autistic contributor; it’s a protocol that absorbs the empathy load in both directions, and the neurotypical contributor was paying a tax they didn’t notice: the cognitive cost of decoding indirect communication. We’ve written about this at length in our piece on the double-empathy problem in pull requests — the point is that protocol design that reduces the interpretive load helps both neurotypes, not just the one it was designed for.

Async-first communication over mandatory live standups helps the autistic employee who processes better in writing and finds verbal sync draining. It also helps the parent returning from leave who needs flexible hours, the person in a different timezone, the employee with social anxiety, and the deep-work contributor who loses an hour of flow to a 15-minute standup. The modal employee — the one the standup was designed for — was tolerating the interruption without noticing it as friction, because they’d acculturated to it.

Sensory-considerate environments help autistic people whose sensory processing makes fluorescent flicker or open-plan noise unbearable. They also help migraine sufferers, people with concussions, people who didn’t sleep well, and anyone whose cognitive load is already at capacity. The quiet, well-lit, low-stimulus environment isn’t a gift to the autistic minority; it’s a friction reduction the majority had learned to call normal.

The pattern is consistent: a design choice made for the neurodivergent edge case often removes friction for the modal case too, because the modal case was tolerating friction it had stopped noticing. The neurotypical default wasn’t friction-free. It was friction-acculturated.

Why the curb-cut works: friction-discovery, not subsidy

This is the mechanism that makes the curb-cut effect real, and it’s worth naming precisely because it’s the part that gets hand-waved. The majority doesn’t experience the default as frictionless; it experiences it as normal. Normal is not the same as optimal. The neurotypical team member navigating vague PR comments isn’t operating at their best — they’re operating at the level they’ve calibrated as acceptable. The employee sitting through a standup they could skip isn’t thriving on the sync; they’re enduring a format that’s slightly wrong but not wrong enough to complain about.

Designing for the person for whom the friction is unbearable surfaces friction the majority had learned to ignore. The autistic employee who can’t tolerate fluorescent flicker isn’t experiencing a different thing than the neurotypical colleague with a low-grade headache they’ve had for three years; they’re experiencing the same thing at a magnitude that makes it impossible to ignore. When you fix it for them, you fix it for everyone — and the everyone who benefited was paying a cost they’d stopped being able to name.

This is the real claim: neurodivergent-first design is a friction-discovery method, not a subsidy. It surfaces the friction in the default that the default was designed not to see. The accommodation isn’t a tax on the majority; it’s an audit of a majority environment that was charging everyone a fee it never disclosed.

The honest cases where it doesn’t generalize

Here is where most curb-cut writing stops, and here is where it shouldn’t. Not every neurodivergent-first design is a curb-cut. Some accommodations genuinely cost the majority and don’t return a universal benefit. Pretending otherwise — insisting that every inclusive design choice is a win-win — is the overcorrection that makes the argument brittle. When someone encounters an accommodation that does impose a real tradeoff and the curb-cut thesis said it wouldn’t, the whole frame collapses and they swing back to “accommodations are a tax” with renewed confidence.

Be specific. A strict no-talking-allowed office policy that helps one hypersensitive autistic employee harms the collaborator who thinks out loud and needs verbal processing to do their best work. That’s not a curb-cut; that’s a genuine tradeoff. A meeting-elimination that helps the autistic employee who finds verbal sync draining starves the neurotypical (or other neurodivergent) employee who thrives on real-time verbal exchange and generates their best ideas in conversation. That’s not a curb-cut; that’s a genuine tradeoff. A documentation-only onboarding that helps the autistic employee who processes text best harms the social learner — neurotypical or otherwise — who absorbs through conversation and apprenticeship. Some people learn by reading the manual. Others learn by shadowing someone for a week, asking questions in real time, building a mental model through dialogue and osmosis. A documentation-only onboarding optimized for the text-first autistic learner doesn’t surface hidden friction the majority was tolerating. It replaces one default with another and shifts the cost onto the people who were served by the old one. That’s a genuine tradeoff, not a curb-cut.

These three examples — the no-talking office, the meeting elimination, the documentation-only onboarding — aren’t fringe edge cases. They’re the predictable shape of what happens when you take a single neurodivergent need and universalize it as the default for everyone. The pattern is the same: one person’s accommodation becomes another person’s barrier. The friction doesn’t disappear. It moves. And when it moves onto someone who was already served by the old default, you haven’t designed anything. You’ve redistributed the tax and called it inclusion.

The honest design move: options, not a replaced default

The answer to a genuine tradeoff is not “impose the autistic person’s need on everyone,” and it’s not “impose the neurotypical default on the autistic person.” It’s to build the environment with options that absorb the tradeoff.

A quiet zone and a collaboration space. Async-first communication and optional sync for the people who genuinely do their best thinking in real-time dialogue. Documentation and a conversational onboarding path. The point isn’t to pick the neurodivergent-optimized default and declare it universal. The point is to stop pretending there’s a single default that serves everyone and start building environments where the tradeoff is structurally absorbed rather than personally paid.

This is universal design — not in the diluted sense where it becomes a synonym for “accessibility,” but in the architectural sense the term was coined for: design that works for the widest range of people not by averaging their needs but by offering multiple paths through the same environment. A ramp isn’t universal design because it replaces stairs. A ramp plus stairs is universal design, because it gives everyone the path that fits their body. The same logic applies cognitively.

The principle has a precise analogue in learning design. Universal Design for Learning is the same idea applied to education: multiple means of engagement (different ways into the material), multiple means of representation (different formats for the same content), and multiple means of action and expression (different ways to demonstrate understanding). Not one optimized path. Many paths, same destination. The learner who needs structure gets structure. The learner who needs open exploration gets that. The learner who needs text gets text; the one who needs conversation gets conversation.

The honest curb-cut thesis, then, isn’t “design for the edge and the center benefits for free.” It’s UDL: options, not a replaced single default. The curb-cut is the case where the option you added for the edge turns out to serve the center too, because the center was paying a friction cost it had normalized. The genuine tradeoff is the case where it doesn’t, and the answer is still options — just without the claim that everyone benefits from the same one.

Honest about the evidence

The curb-cut evidence base is real. Captions help people in noisy environments, not just Deaf users. Ramps help people with strollers and deliveries, not just wheelchair users. Structured PR protocols reduce ambiguity for everyone, not just autistic engineers. Async-first communication reduces meeting load for everyone, not just the people who can’t tolerate standups. These are documented, replicated, and they follow a discoverable pattern: the majority was acculturated to friction it had stopped seeing, and designing for the person who couldn’t tolerate it surfaced the cost for everyone.

But the universalizing claim — “design for the edge and the center benefits for free” — is overstated in parts of the accessibility literature. Not every edge-optimized design generalizes. Some are genuine tradeoffs with no universal return. The no-talking office helps the autistic employee who needs quiet and harms the social learner who needs verbal processing. The documentation-only onboarding helps the text-first thinker and costs the apprenticeship learner. These aren’t failures of curb-cut thinking. They’re cases where the curb-cut frame doesn’t apply because the friction isn’t shared — it’s genuinely distributed differently across people.

The honest version is this: neurodivergent-first design is a friction-discovery method that often helps broadly, because the friction was real for everyone, just bearable for the majority. But where it doesn’t help broadly, the answer is universal design — options that absorb the tradeoff — not pretending the tradeoff is free. The evidence supports the friction-discovery claim. It does not support the universal-free-benefit claim as a blanket statement. The universalizing claim needs the honest middle, not the over-rosy version that every accommodation is a hidden gift to the majority.

Coda

The accommodation tax is a myth where the design is a friction-discovery curb-cut: the majority was paying a friction cost it had normalized, and designing for the person who couldn’t bear it surfaced that cost for everyone. In those cases, the “tax” was always a subscription to friction everyone was already paying.

The accommodation tax is real where the design is a genuine tradeoff with no universal return: one person’s accommodation imposes a cost on someone else, and no amount of reframing makes that cost disappear.

The honest design move is universal design — options that absorb the tradeoff — not a false claim that every accommodation is free, and not the defensive frame that every accommodation is a tax. Build the quiet zone and the collaboration space. Build the documentation and the apprenticeship. Build the ramp and the stairs. The myth isn’t “there’s no cost.” The myth is “the cost always falls on the majority and is always a tax.” Mostly it’s a curb-cut. Honestly, sometimes it’s a tradeoff. And the answer to a tradeoff is options.