What Accommodations Do Neurodivergent Employees Actually Need at Work?

If you are a neurodivergent adult, there is a good chance you have never been offered a workplace accommodation, and an even better chance that you would not know what to ask for if you were. If you work in HR or lead a team, there is a good chance you believe your organisation already has this covered, and a good chance the data says otherwise.

Neurodivergent people make up an estimated 15 to 20 percent of the global population, and yet approximately 85 to 90 percent of neurodivergent individuals experience unemployment or underemployment (Kalmanovich-Cohen & Stanton, 2025). This article works through what accommodations actually look like in practice, what the research says about who gets them, why so many requests are rejected, and what a genuinely inclusive approach requires from organisations in Aotearoa New Zealand and beyond.

What Counts as a Workplace Accommodation for a Neurodivergent Employee?

A workplace accommodation is any adjustment to the work environment, schedule, communication style, or task structure that allows a neurodivergent employee to do their job without unnecessary barriers. This is broader than most people assume. It is not only about physical changes like lighting or noise. It includes how instructions are given, how meetings are run, how performance is assessed, and how much flexibility exists around when and where work happens.

Research consistently points to a small set of adjustments that come up again and again: flexible work hours, quiet spaces, job-sharing options, noise-cancelling headphones, written or recorded instructions that can be referenced later, and tools such as note-taking software or the ability to record meetings (Kalmanovich-Cohen & Stanton, 2025). None of these are exotic. Most are things that any thoughtful workplace could offer as a matter of course.

For a late-diagnosed adult, this list can be confronting in a different way. Many of these supports are things you have been engineering for yourself for years, on your own and without naming it, an extra monitor angled away from foot traffic, headphones you pretend are for music, asking for things in writing because you cannot trust yourself to remember a verbal instruction. The research is, in effect, naming the workarounds you already built and calling them what they are: legitimate accommodations that should have been available to you from the start.

For organisations:Audit your current accommodation list against this baseline set: flexible hours, access to a quiet space, noise-cancelling headphones, written instructions as standard, and meeting recording or note-taking tools. If these are not already offered to every employee by default, you are relying on individuals to ask for things that research identifies as foundational, not exceptional.

How Many Neurodivergent Employees Actually Receive Accommodations?

Fewer than half. That is the short answer, and for some groups it is considerably fewer than half.

A survey of 1,100 neurodivergent adults found that just 41 percent received any workplace accommodations at all, with a further 6.5 percent reporting they had been denied accommodations after requesting them (McGlauflin & Abrams, 2023). That means well over half of neurodivergent employees are working without any formal support, regardless of whether they need it.

The picture gets worse when you look at how support is distributed. Most accommodation research assumes that everyone has the same path to support: notice a need, disclose, request, receive. In practice, the path is far less even than that, and the unevenness falls along predictable lines, including gender, which the next section covers in detail.

Awareness of entitlement is also a major barrier in its own right. A 2025 survey found that nearly 7 in 10 employed neurodivergent adults had no idea what accommodations they were entitled to, and 51 percent did not know who to talk to in their workplace about requesting them (Understood.org, 2025, as cited in this cycle’s research brief). People are not failing to use a system that exists. For many, the system is effectively invisible.

For organisations:If fewer than half of neurodivergent employees in your organisation are likely to be receiving accommodations, the absence of requests in your HR system is not evidence that support is not needed. It is more likely evidence that people do not know support exists, do not know who to ask, or have already decided it is not worth asking.

Why Do So Many Accommodation Requests Get Rejected?

This is the part of the research that tends to surprise people who assume the main problem is that nobody asks. People do ask. The problem is what happens after they ask.

Deloitte Global’s 2024 survey of 10,000 respondents across 20 countries found that of those who had disclosed a disability, chronic health condition, or neurodivergence and gone on to request accommodations, 74 percent had at least one request rejected, and almost two in ten had every single request rejected (Deloitte Global, 2024). The same report found that six in ten respondents had been unable to attend a work event because it was not accessible to them.

Rejection rates at this scale point to a structural issue, not a series of unfortunate individual decisions. When three out of four formal requests meet at least one rejection, the system is not failing occasionally. It is failing as designed.

This is part of why fear of asking is so widespread and so rational. Deloitte’s respondents cited fears of negative perception, concerns about being seen as difficult, and past negative experiences as key reasons they had not asked their current employer for accommodations at all (Deloitte Global, 2024). For someone weighing up whether to disclose, a 74 percent rejection rate among people who already took that risk is not an abstract statistic. It is a fairly accurate prediction of what is likely to happen to them too.

For organisations:If your organisation tracks accommodation requests, start tracking rejections and partial rejections as a separate metric, reviewed at the same level as other risk indicators. A high approval rate on paper can still sit alongside a high rate of requests that were approved in part, delayed, or downgraded without explanation. Those outcomes function as rejections from the employee’s perspective even if they are not recorded as one.

Why Are Neurodivergent Women Less Likely to Receive Support?

The gender gap in accommodation access is one of the starkest findings in this research, and one of the least discussed.

Among the 1,100 neurodivergent adults surveyed by Hopper Health, 25.7 percent of openly neurodivergent men received accommodations without having to ask, and a further 30 percent received them after requesting. For women, only 6.41 percent received accommodations without asking, and 23.36 percent received them after asking. Women were also more likely to be denied accommodations when they did request them (McGlauflin & Abrams, 2023).

Put plainly: a neurodivergent man is roughly four times more likely than a neurodivergent woman to have support offered to him before he ever has to ask for it. For late-diagnosed women in particular, many of whom spent decades being told they were coping fine because they had built elaborate, exhausting systems to appear as though they were, this statistic will likely land as confirmation rather than surprise.

It is worth being direct about the limits of this evidence. This is industry survey data reported through a media outlet, not a peer-reviewed study, and there is currently no equivalent dataset for Aotearoa New Zealand. The gap it describes is real and widely felt, but the research base behind it is thinner than the size of the issue deserves. That gap in the literature is itself worth naming.

For organisations:If your organisation collects any data on who requests and receives accommodations, check whether that data can be broken down by gender. Most organisations cannot answer this question because the data was never structured to allow it. Not being able to answer is itself a finding, and a starting point for closing a gap that current research suggests is significant.

Do Accommodations Cost Employers Money?

Rarely, and when they do, not much.

A survey conducted for the United States Department of Labor’s Office of Disability Employment Policy found that 56 percent of workplace accommodations cost nothing at all to implement. Of the remainder, the typical cost was around 500 US dollars (Job Accommodation Network, 2020, as cited in this cycle’s research brief).

The accommodations that come up most often in the research, flexible scheduling, permission to wear headphones, written instructions, access to a quieter workspace, recording meetings, are policy and practice changes rather than purchases. They cost time and a willingness to do things differently, not a procurement budget.

This figure is US-specific and several years old, and there is no published New Zealand equivalent. But the broad pattern, that the cost barrier is more perceived than real, lines up with what the qualitative research describes: most of what neurodivergent employees ask for is closer to good management practice than to specialist equipment.

For organisations:Before assuming an accommodation request will be expensive or complex, check whether it is a policy or practice change rather than a purchase. The majority of commonly requested accommodations fall into the first category, which means the real cost is often the time it takes to have the conversation and adjust how a task or meeting is structured.

What Is Universal Design and Why Does It Matter More Than Individual Accommodations?

Most accommodation systems work like this: an employee notices they are struggling, decides to disclose, makes a request, and waits to find out whether it is approved. Every step in that chain is a point where the system can fail, and the research above shows that it frequently does.

An alternative approach, increasingly supported in the research, is universal design: building workplaces that are accessible by default rather than by request. This means offering things like flexible work hours, quiet spaces, job-sharing options, and noise-cancelling headphones, along with tools such as note-taking software or meeting recording, to everyone as standard, rather than as something that has to be individually justified (Kalmanovich-Cohen & Stanton, 2025). When offered this way, these supports do not single anyone out, and they tend to benefit people who would never have disclosed or asked in the first place.

A 2025 systematic review of 56 studies on neurodivergence and the workplace found that the proactive creation of inclusive team environments and adoption of neurodivergence-sensitive policies, both consistent with universal design, were associated with positive impacts on employee retention, commitment, and organisational image (Vargas-Salas et al., 2025). The same review noted that most existing studies focus on short-term interventions, so long-term outcome data is still limited.

For employees, the practical difference is significant. Universal design does not require you to identify yourself as different in order to access support. The quiet room is just a room. The flexible hours are just the hours. Nobody needs to know why you are using either.

For organisations:Pick three commonly requested accommodations from your current individual request data and ask whether each could become a standard offering available to everyone, rather than something that requires a disclosure and an approval process. Flexible hours, recorded meetings, and access to a quiet space are usually the easiest starting points and align directly with what the research identifies as highest impact.

What Does This Look Like in Aotearoa New Zealand Specifically?

Most of the large-scale research on neurodivergent workplace accommodations comes from the United States, the United Kingdom, and international survey panels. There is comparatively little published research specific to Aotearoa New Zealand, and what does exist tends to focus on physical workplace design rather than disclosure, accommodation requests, or policy outcomes.

One relevant example is a 2025 conference paper examining universal design for neurodiverse individuals in New Zealand workspaces, which found that when workplace design does not adequately consider the cognitive styles, sensory sensitivities, and communication preferences of neurodivergent people, it creates barriers, reduces productivity, and contributes to feelings of exclusion (Muneer & Burfoot, 2025). The paper’s focus is architectural and based on case studies of ten projects, so it cannot speak to disclosure rates, accommodation approval rates, or employer practice in the way the international survey data can.

This is a genuine gap, not a minor footnote. Aotearoa has its own employment law context, its own cultural norms around disclosure and disability, and its own workplace cultures. International statistics on rejection rates, gender gaps, and accommodation costs are useful as a guide to what is plausible here, but they are not a substitute for local data that, as far as current research shows, does not yet exist at scale.

For organisations:If your organisation operates in Aotearoa New Zealand, treat international statistics in this article as indicative rather than locally verified. Consider what internal data you could start collecting now, accommodation requests, approval and rejection rates, and any available breakdown by gender or neurotype, so that in a few years your organisation has its own evidence base rather than relying entirely on overseas figures.

What Should a Late-Diagnosed Neurodivergent Adult Do if They Have Never Been Offered Support?

Start by knowing that the research backs up what you are likely feeling. If you have never been offered an accommodation, you are in the majority, not the exception. If you have never asked, the data on rejection rates suggests that hesitation has not been irrational.

That does not mean asking is pointless. It means going in with realistic expectations and a clear sense of what you are asking for. Vague requests are easier to deprioritise than specific ones. Rather than asking for general support, naming a specific adjustment, written meeting summaries, a quieter desk location, the option to turn cameras off in video calls, flexible start times, gives whoever receives the request something concrete to act on.

It can also help to separate the question of disclosure from the question of accommodation. Many of the adjustments that make the biggest difference, flexible hours, recorded meetings, written instructions, are increasingly being offered as standard in workplaces that have adopted universal design principles, which means they may already be available to you without needing to explain why you want them.

If a request is knocked back, that is data about the organisation, not a verdict on whether your need was reasonable. The research is clear that rejection is common and that it happens regardless of how reasonable or low-cost the request was.

For organisations:When an employee makes an accommodation request, the most useful first response is often simply to ask what would help, rather than starting from what your organisation is willing to offer. Many requests, once specific, turn out to be low-cost adjustments to existing practice rather than new resources, and starting from the employee’s actual need rather than a predetermined list reduces the likelihood of a request being only partially met.

What Should Organisations Do Differently, Starting Now?

The research points to a small number of changes that consistently make the most difference, and none of them require a large budget or a long implementation timeline.

Move from individual requests to default offerings

Identify the accommodations that come up most often in requests, and make them available to everyone by default. This removes the disclosure step entirely for the most common needs and reduces the administrative load of processing individual requests.

Track rejections, not just requests

Given that international data shows rejection rates as high as 74 percent among those who disclose and request support (Deloitte Global, 2024), any organisation that is not separately tracking how many requests are rejected, partially approved, or delayed is missing the metric that matters most.

Train managers, not just HR

Accommodation decisions are frequently made or shaped at the manager level, often informally. Confidence in recognising and responding to neurodivergent employees’ needs varies widely across management, and training that reaches managers directly, rather than sitting solely within HR policy documents, closes a gap that policy alone cannot.

Build local evidence

Given how little Aotearoa-specific data currently exists, organisations that start collecting their own data on accommodation requests, approvals, rejections, and outcomes will be in a stronger position in years to come, both to demonstrate progress and to identify where their own practice diverges from international patterns.

For organisations:None of the four changes above require waiting for a budget cycle or a policy rewrite to begin. Identifying your three most common accommodation requests, starting to track rejection and partial approval rates, and having one direct conversation with your management team about how confident they feel responding to these requests can all start this week.

The Bottom Line

The research on neurodivergent workplace accommodations tells a fairly consistent story. The supports that make the biggest difference are simple and usually low-cost. Fewer than half of neurodivergent employees receive them. Of those who ask, most face at least one rejection. Women are far less likely than men to be offered support without having to ask, and far more likely to be turned down when they do.

None of this is a story about individual employees not trying hard enough, or individual managers being unreasonable. It is a story about systems built around disclosure and individual request, in a context where disclosure is risky and requests are often refused. The research increasingly points toward a different model: building workplaces that are accessible by default, so that support does not depend on someone first identifying themselves as needing it.

For late-diagnosed neurodivergent adults, this research offers something close to validation. The exhaustion of building your own workarounds, the hesitation to ask, the sense that support seemed to come more easily to other people, all of this tracks with what the data shows. For organisations, it offers a clear and achievable starting point: the changes that matter most are not the expensive ones.