The single most useful fact about disability and work is also the most uncomfortable one: the disability employment gap did not close during the tightest U.S. labor market in fifty years. According to the Bureau of Labor Statistics' annual Persons with a Disability: Labor Force Characteristics release, the employment-population ratio for people with disabilities was 22.7% in 2024, compared with 65.5% for people without disabilities — a gap of roughly 43 points that has barely moved in a decade. When you cannot move a number that much during full employment, the problem is not the cycle. It is the architecture.
That architecture has three load-bearing walls: accommodations treated as charity rather than infrastructure, hiring funnels that screen out variance before it can be assessed, and disability metrics that almost no public company actually reports. Rebuilding any one of them would close more of the gap than another decade of awareness campaigns. Rebuilding all three is what the next phase of inclusion has to look like.
Accommodations are cheap; the myth that they aren't is expensive
The U.S. Department of Labor's Job Accommodation Network has, since the early 2000s, surveyed employers about what workplace accommodations actually cost. Their finding has been remarkably stable: roughly half of accommodations cost nothing, and the median one-time cost for the rest sits near $300. JAN's most recent multi-year summary, Workplace Accommodations: Low Cost, High Impact, found 56% of accommodations were no-cost. The number that matters more, though, is the one employers themselves report afterward: most say the change improved retention, productivity, and team morale — not just for the accommodated employee.
Put against that, the standard manager objection ("we can't afford it") is empirically false. What it usually means is "we don't have a process." Companies with central accommodation funds — Microsoft, IBM, EY and a handful of federal agencies — have shown that pulling accommodations off individual managers' budgets removes the perverse incentive to deny them. Microsoft's published guidance on its enterprise accommodation program is the cleanest public template; the cost has been described internally as a rounding error on benefits spend.
The funnel screens out disability before talent gets to compete
If accommodations are not the binding constraint, hiring is. A 2023 audit study published in the ILR Review by economists Mason Ameri and Lisa Schur (Rutgers) sent matched résumés to thousands of accounting positions, identical except for a disclosed disability. Applicants disclosing a disability received 26% fewer expressions of employer interest, with the gap concentrated among smaller firms that are not covered by federal contractor rules. That is roughly the same penalty earlier audit work had found in the 2010s — a discouraging non-improvement.
The mechanism is not usually animus. It is variance aversion in a hiring system that has been optimized to filter, not to evaluate. Algorithmic résumé screens punish gaps. One-way video interviews penalize non-neurotypical eye contact. Timed cognitive assessments handicap candidates with motor or processing differences. Each tool, taken alone, is defensible. Stacked, they form a screen that filters out exactly the candidates the ADA was written to protect, without anyone ever making a discriminatory decision in the legal sense.
The fix is not abolishing assessment; it is unbundling it. Companies that have moved to skills-based, work-sample hiring — Deloitte's apprenticeship program, JPMorgan's Office of Disability Inclusion pilots, and a growing cluster of European firms following the EU's 2024 Pay Transparency Directive — consistently find that disabled hires perform at parity once given a structured task instead of a structured interview. That is also the finding of the Disability Inclusion at Work → pillar research at NWLB.
Neurodiversity programs are the proof of concept
The clearest natural experiment is the corporate neurodiversity hiring programs SAP, JPMorgan, Microsoft, EY and Dell have run since the mid-2010s. Their published retention data — summarized in Robert Austin and Gary Pisano's Harvard Business Review piece "Neurodiversity as a Competitive Advantage" — shows 90%+ retention and productivity that matches or exceeds general hires in software testing, data quality, and cybersecurity roles. The programs work because they replace the interview with a multi-week assessment, give candidates the tools they will use on the job, and train managers explicitly. None of those interventions are disability-specific. All of them are good hiring.
What measurement would actually change
The third wall is data. Disclosure rates for disability inside U.S. corporations sit somewhere between 4% and 6%, against an underlying workforce prevalence the CDC puts above 25%. Most disability is invisible — mental health, chronic illness, neurodivergence — and most employees correctly judge that disclosing it carries career risk. As long as that risk-reward calculation holds, every internal "disability dashboard" is reporting on the wrong denominator.
Two structural moves change the math. The first is the SEC's human-capital disclosure rule, which since 2020 has required public companies to report material workforce information; the next refresh, expected to push toward standardized metrics, is the single highest-leverage policy lever available. The second is the EU's Corporate Sustainability Reporting Directive (CSRD), which from 2024 onward requires reporting under the European Sustainability Reporting Standard S1, including disability representation. Multinationals will not maintain two reporting regimes for long, which means EU rules will drag U.S. practice with them — the same way GDPR did for privacy.
The Valuable 500, a global coalition of 500 CEOs that has pushed disability reporting since 2019, found in its 2023 ESG and Disability Data review that only 12% of its members were publicly reporting disability workforce data, despite all having committed to do so. That gap between commitment and disclosure is the real story of corporate disability inclusion right now — and it is also the most fixable.
What the next five years should look like
The agenda is unglamorous and almost entirely operational. Centralize accommodation budgets out of line managers' P&Ls. Replace résumé screens and one-way video interviews with structured work samples for any role where the work product can be observed. Audit assessment vendors for disparate impact under the EEOC's 2023 guidance on AI in hiring. Set a public disability representation target tied to executive compensation, the way many firms now do for gender. Train managers on accommodation conversations the way they are trained on performance reviews — because they happen at roughly the same frequency, and one is currently treated as routine while the other is treated as a crisis.
None of this requires a moral conversion. It requires treating disability inclusion as a process problem with a known solution set, which is what the evidence has been saying for a decade.
Half of workplace accommodations cost nothing; the other half cost less than a single bad hire. The expensive thing is pretending the gap is about willingness rather than process.
Updated May 21, 2026. This piece was substantively rewritten as part of NWLB's 2026 editorial refresh.
