Measurement Is the Intervention

The 2025-2026 retrenchment ran a natural experiment on corporate disability and neurodiversity programs, and the seven programs that survived it share exactly one trait.

The 2025-2026 retrenchment ran a natural experiment on corporate disability and neurodiversity programs. Of 20 neurodivergent-specific hiring programs identified in the Council's review of 167 program entries at 144 North American public companies, seven show hard 2025-2026 evidence of activity, and all seven publish performance numbers: Wells Fargo, 82 percent retention on 364 hires; DXC, 92 percent. Programs framed as diversity commitments were scrubbed or genericized. Programs framed as talent performance with retention data attached kept operating. Build it as a measured performance program.

Nineteen of those 20 programs are FTE internship or direct-hire pipelines. One runs through a staffing arrangement. No employer, association, or standards body publishes a practice framework for the contingent channel. The whitespace is not underserved; it is empty.

Practitioners rated eight candidate solutions. Industry-wide recognition of DOBE-certified suppliers rated highest (9 of 12 agree), VMS-native accommodation features second, and standardized contract language, required recruiter training, and disability-specific scorecard metrics tied third at 9 of 12 each. Rate card allowances rated last. You do not need permission, budget, or new vendors. You need to build the machinery.

The Council Inclusion Maturity Model grades a program across five dimensions (policy, tracking, process, technology, metrics) at four stages (Unaware, Aspirational, Operational, Embedded). A program's stage is its lowest dimension score: machinery is graded by its weakest link.

An MSP leader's account shows the lever: one line on the quarterly supplier scorecard, accommodation-request response time against a 48-hour standard, no penalty, no bonus. Within two quarters every top-ten supplier had a named accommodation contact. Measurement is the intervention.

Five actions can start Monday: assign ownership, adopt the one-page policy statement, open the accommodation channel with a 48-hour SLA, fix the front end with requisition standards and the vendor question, and put one disability metric on the supplier scorecard. The Working Group moves from research to build in 2026-2027 and is recruiting programs to pilot the 90-day quick start.

Council members also receive the register behind this series: 167 documented inclusive hiring programs at 144 North American public companies, with source documents. Request access or talk with the Working Group at pathways@defiantsolutions.com.

This is Brief 4 of four drawn from The Missing Workforce: Disability and Neurodivergent Inclusion in the Contingent Workforce, a white paper of the Pathways Opportunity Council's Neurodivergence and Disabled Contingent Hiring Best Practices Working Group, published September 2026. This brief is the build: which programs survived the 2025-2026 retrenchment and why, which interventions practitioners believe would work, the Council Inclusion Maturity Model, recommendations for every party in the tri-party model, and the five actions a program can start Monday morning. Each brief stands alone; together the four carry the full argument. The complete report, including methodology, the survey instrument, and full response tables, is available from the Council at pathways@defiantsolutions.com.

The 2025-2026 period ran a natural experiment on corporate disability and neurodiversity programs, and the Council measured the results. Across its review of 167 publicly documented inclusive hiring program entries at 144 North American public companies (methodology in Section 4.4 of the full report), the Council identified 20 neurodivergent-specific hiring programs. As of July 2026, only seven show hard 2025-2026 evidence of activity: CIBC, DXC, Goldman Sachs, JPMorgan Chase, Microsoft, Travelers, and Wells Fargo. Several more are likely active but quiet. For the remainder, no public evidence of activity exists after 2023, and in ten cases program web pages remained live while the program content had been quietly removed. Absence of public evidence is not proof a program has closed, and this report names no program as discontinued. But the direction is unmistakable: the environment turned, and the population of visible programs contracted. The policy backdrop is a matter of record; alongside the OFCCP restructuring described in Section 3.3 of the full report, the Federal Housing Finance Agency terminated the housing enterprises' DEI initiatives in March 2025, cancelling roughly $6.4 million in programs.²

The programs that survived share one trait, and it is not size, sector, or budget. They publish performance numbers. Wells Fargo reports 82 percent retention across 364 hires through the fourth quarter of 2025, and a university research center now carries its name. DXC's Dandelion Program reports 92 percent retention (Section 3.1 of the full report). JPMorgan Chase marked its program's tenth anniversary with approximately 200 participants across eight countries. Programs framed as diversity commitments were scrubbed or genericized; programs framed as talent performance, with retention data attached, kept operating and kept publishing. For anyone deciding how to position disability inclusion in a contingent program, the retrenchment period supplies the answer: build it as a performance program with measured outcomes, because that is the framing that survives leadership changes, political cycles, and budget reviews. The recommendations later in this brief operationalize the lesson.⁶

The survey's final rating exercise asked practitioners (n = 12) to rate eight candidate solutions on a five-point agreement scale. The ranking is the report's roadmap, because it comes from the people who would have to execute it:

Three readings matter. First, the highest-rated single solution is collective, not programmatic: recognizing DOBE-certified suppliers in diversity spend is something no program can do alone and the industry can do together, and the industry recommendations later in this brief take it up. Second, the next four are all program-level levers: things a program owner or MSP can do unilaterally, inside existing governance, this quarter. Contract language, recruiter training, and scorecard metrics require no new technology procurement, new law, or new budget authority; VMS-native features require a vendor to build them, but the procurement question that forces the issue costs nothing. Third, the practitioners' ranking aligns with the economics: rate card allowances rated last, and where accommodations do cost something, the amounts are trivial against the cost of churn in an industry with triple-digit annual turnover. The survey's message to the industry is unusually coherent: you do not need permission, budget, or new vendors. You need to build the machinery.

Sections 5.1 through 5.6 of the full report prove the machinery gap from the demand side: inside programs, the policy, tracking, technology, and contracts are missing or unknown to the people running them. The Council's employer corpus proves it from the supply side. Of the 20 neurodivergent-specific hiring programs identified across the 167 entries (144 companies), 19 are FTE internship or direct-hire pipelines. The exception is KeyBank, which for five years has placed neurodivergent professionals as staffing contractors through a specialist neurodiversity staffing intermediary, with 19 contractors placed and 4 converted to full-time roles. One other staffing-mediated design appears in the record, an automaker's contract-to-hire program, but it shows no public activity in recent years. And no organization anywhere in the corpus publishes a practice framework for the contingent channel: skills-based assessment through a VMS, accommodation responsibility split between staffing supplier and client, job coaching for contractors, or conversion metrics through an MSP.⁸

The corpus also shows where the stable infrastructure actually sits. Employer programs churn; the layer around them persists. Specialist assessment and employment intermediaries, university centers such as the University of Connecticut's neurodiversity employment center, nonprofit alliances, and benchmarking bodies (Disability:IN, the Disability Equality Index, the Neurodiversity @ Work Employer Roundtable) operated continuously across the same period in which corporate programs opened, renamed, and disappeared. For a staffing industry deciding where to build partnerships, that durability ranking matters: the intermediary layer is the load-bearing structure. The Council describes commercial intermediaries by function rather than by name (Section 1.5 of the full report); the structural point does not depend on any vendor's identity.

Read together, the two evidence bases say the same thing from opposite directions. The survey says the machinery does not exist inside contingent programs, even in a network convened around the issue. The corpus says nobody outside has built it either: no employer, association, or standards body publishes the contingent practice framework. The whitespace this Council was formed to fill is not merely underserved. It is empty, and this report and its companion playbook are, as far as the Council can determine, the first occupants.

Three composite profiles illustrate how the machinery gap is experienced from inside. Each is assembled from conversations among Working Group members, who operate or oversee contingent programs themselves, and from informal exchanges with survey respondents and Council network members during the fielding period. Details are composited and anonymized so that no single organization is identifiable, and the quotations are composite renderings of what practitioners told us, not verbatim transcripts. The survey data in Chapter 5 of the full report confirm each pattern the profiles describe.

An MSP operations leader described the single change that altered supplier behavior in his program: one line on the quarterly supplier scorecard, accommodation-request response time with a 48-hour acknowledgment standard. No penalty structure, no bonus, just measurement and a ranking. Within two quarters, every top-ten supplier in the program had designated an accommodation contact, and two had rewritten their candidate onboarding scripts. His observation captures the central finding of Section 5.6 of the full report: in a supplier ecosystem built around scorecards, measurement is the intervention.

Suppliers optimize what the scorecard measures. That's not cynicism, it's the design of the system. So we measured the thing we cared about.

Four Stages: Unaware, Aspirational, Operational, Embedded

The Council Inclusion Maturity Model grades a contingent program across five dimensions (policy, tracking, process, technology, and metrics) at four stages. A program's stage is its lowest dimension score: machinery, like a chain, is graded by its weakest link.

The model's design reflects the survey: every Operational-and-above behavior named here is something a high-maturity respondent actually does, and every Unaware behavior is something a majority of respondents reported. It is descriptive before it is aspirational.

The employer corpus supplies market proof points for the model's upper stages, drawn from programs with published practices. Assessment redesign at the Process dimension: Microsoft's skills-based interviews and Charles Schwab's skills demonstrations replace the conventional interview rather than accommodating it. Applicant-stage accommodation, the differentiator between Operational and Embedded process: Boeing maintains more than 90 trained Reasonable Accommodation employees, AbbVie's policy spans application through hiring, and Micron proactively informs candidates of interview accommodation options. Readiness before arrival: Accenture and Jacobs train managers and recruiters before cohorts land, because programs fail at the hiring-manager level. Supported onboarding: SAP, Dell, and DXC budget job coaching through the first ninety days, where placements fail. ERGs as recruiting infrastructure rather than affinity groups: Aramark's resource group runs autism hiring events and ADP's business resource groups formally support talent acquisition. At the Metrics dimension, external benchmarking functions as accountability: Aramark has scored 100 on the Disability Equality Index for eight consecutive years and Travelers for seven, and annual benchmark participation has become the most reliable public signal a program still operates. Upstream, Travelers sources through state agencies and universities and helped found the University of Connecticut's neurodiversity employment center, and Freddie Mac runs an internship with the Autistic Self Advocacy Network. Every practice is an existence proof: each upper-stage behavior in the table is being done at scale by a named employer, in direct hire. The Council's work is porting them across the tri-party boundary.¹⁰

Self-Assessment Criteria

Scoring is deliberately blunt. For each dimension, assign 1-4 using the table above; the program score is the minimum across dimensions, not the average. A program with excellent policy and no tracking is, for a worker who needs an accommodation, an Unaware program. The full worksheet, with diagnostic questions per cell (ten per dimension) and a scoring sheet, is Playbook Appendix B. Two administration rules from the pilot: score from evidence, not intent (“show me the register, the clause, the field”), and score jointly, with client, MSP, and a top supplier in the same room, because the tri-party disagreements surfaced in the scoring conversation are themselves the most useful output.

Where the Market Sits Today

Mapped against the model, the survey sample clusters at the bottom: most responding programs sit at Stage 1 (Unaware) or Stage 2 (Aspirational), with the modal profile an Aspirational program: a written statement somewhere, goodwill everywhere, and no tracking, process, or metrics underneath. A minority of roughly three to five programs reach Operational on at least three dimensions (five have a defined process, four track status, three allocate duties by contract, two carry a scorecard metric, two verify training), and these are the programs whose practices anchor Chapter 6 of the full report.2. None scores Embedded across all five dimensions; technology blocks everyone, because the VMS gap is market-wide and cannot be solved unilaterally. The Working Group's read: the market's ceiling is currently set by its platforms, and its floor by the absence of ownership. Recommendations follow accordingly.

Program Owners and Clients

Clients own the program's terms and should act like it:

Own the policy. Adopt a written disability inclusion policy that names contingent labor explicitly (Playbook Section 2.3), and assign a named executive owner.

Frame it on performance. Position and fund the program as a talent-performance initiative with published retention and cycle-time outcomes, not as a diversity commitment. The employer corpus shows the framing decision is a survival decision: through the 2025-2026 retrenchment, programs framed on measured performance kept operating while programs framed as commitments were scrubbed (see the retrenchment test above).

Fund the accommodations. Establish a centralized program accommodation fund or an explicit allocation rule (Playbook Section 4.2) so cost never defaults to the supplier's margin or the worker's silence. JAN's data (61 percent of accommodations at $0, a $300 median one-time cost, and employee retention the most frequently cited employer benefit) makes this the cheapest retention investment in the program.⁷

Demand the metrics. Add the disability metric set (Playbook Section 9.2) to the supplier scorecard and the MSP's quarterly business review: self-ID invitation coverage, accommodation acknowledgment within 48 hours, accommodation cycle time, accommodation request volume, and 90-day assignment retention for accommodated workers.

Procure the technology. Put the Playbook Section 8.3 question set into every VMS and assessment-tool RFP and renewal: WCAG 2.1 AA conformance, a worker-facing accommodation workflow, and disability-inclusive AI validation.

Managed Service Providers

MSPs are the program's operating system, and every workflow in this report needs an orchestrator:

Build the scorecard. Own the disability metric definitions, data collection, and quarterly reporting (Playbook Chapter 9). Nine of twelve practitioners rate scorecard metrics a high-impact intervention, and operationally the scorecard is an MSP deliverable.

Check the durability signal. When evaluating employer programs, partners, or practices to emulate, weight current-cycle evidence over program-page existence: a cohort announcement, a live requisition, or annual benchmark participation (Disability Equality Index, Neurodiversity @ Work Employer Roundtable) within the past 18 months. A live web page is not an active program (Section 4.4 of the full report).¹¹

Orchestrate training verification. Require supplier recruiter training and verify it, through attestation plus audit, rather than accepting announcements (Playbook Section 7.2).

Manage the accommodation workflow. Operate the intake register, enforce the 48-hour acknowledgment SLA, run the escalation path (Playbook Sections 5.1 and 5.4), and chair the quarterly review of cycle times and outcomes.

Broker the technology gap. Until VMS platforms ship native features, run the workaround playbook (Playbook Section 5.3) on the program's behalf, and carry the program's consolidated feature demands to the VMS provider. An MSP representing 40 programs has leverage a single client does not.

Staffing Suppliers

Suppliers touch the candidate first and the worker longest:

Train recruiters, and prove it. Deploy the curriculum in Playbook Section 7.1 (disclosure, accommodation basics, neurodivergent candidate engagement, screening-tool limits) and provide completion evidence to programs that ask.

Fix job descriptions and screening. Apply the requisition standards (Playbook Section 6.1): strip non-essential physical requirements, separate genuine demands from boilerplate, and offer assessment alternatives as standard practice, not special favor (Playbook Section 6.2).

Build pipeline partnerships. Register with DOBE directories and specialist neurodivergent talent networks, connect with university channels such as the USC Lancaster pipeline (Section 9.2 of the full report), and track source-of-hire for disabled and neurodivergent candidates (Playbook Section 6.3).

Pitch by sector. The employer corpus shows where neurodiversity and disability hiring infrastructure already concentrates: finance and technology run the mature cohort programs and skills-based assessment redesigns, industrials build accommodation capacity through centralized teams and technician pipelines, and healthcare has the least named-program infrastructure of any major sector, making it the largest greenfield for a supplier arriving with a ready framework.

Make asking safe. Adopt the application-stage accommodation notice and intake script (Playbook Section 6.4). The practitioner evidence is unambiguous: when asking is safe, request volume rises. That is the system working, not failing.

VMS and Technology Providers

The technology tier sets the market's ceiling, and the survey says the ceiling is low:

Build native accommodation support. A worker-facing request field at assignment, routing logic between supplier and client contacts, status tracking against SLAs, cost-code allocation, and reporting. One of sixteen programs has anything like this today; the first platform to ship it owns a differentiator every MSP in this report has said it would pay attention to.

Meet and publish accessibility standards. WCAG 2.1 AA conformance with a current VPAT for every candidate- and worker-facing surface, tested with assistive-technology users, not only automated scanners.

Open the AI black box. Publish validation documentation for screening and matching models that includes disability: subgroup adverse-impact analysis, accommodation pathways inside assessments, and audit rights for customers. Local Law 144 does not require it; the market should. The AskEARN/PEAT eRecruiting checklist offers a ready framework.¹⁴

The Industry Collectively

Some machinery only works if it is shared:

Recognize DOBE suppliers in diversity spend. This was the single highest-rated solution in the Council's survey (9 of 12 agree, 5 strongly). Route contingent-labor spend, one of the largest indirect procurement categories, through Disability:IN-certified Disability-Owned Business Enterprises where they exist, and count them in Tier-1/Tier-2 diverse-spend goals alongside other certified categories. Staffing firms themselves can be DOBE-certified.¹⁵

Adopt shared standards. The Council offers the maturity model, the metric definitions, and the model contract language in this report and the playbook as open industry standards. Use them, improve them, and converge on them so that suppliers face one set of expectations rather than seventeen.

Benchmark across programs. The Working Group's next phase (Section 9.1 of the full report) includes a voluntary annual benchmark: participating programs submit the Playbook Chapter 9 metrics, anonymized, and receive the cross-program distribution. Seventeen programs' data is a signal; a hundred programs' data is a market.

The full program takes a quarter; the playbook's 90-day quick start sequences it. Five actions can begin Monday morning, each cross-referenced to the companion Implementation Toolkit:

Assign ownership. Name who owns disability inclusion in the contingent program (client, MSP, or supplier, task by task) using the RACI matrix in Playbook Chapter 2. Today the honest answer in most programs is “nobody, severally.”

Adopt the model policy statement. One page, adapted from Playbook Section 2.3, extending the organization's disability inclusion commitments explicitly to contingent labor. Seven of sixteen programs have a written policy in any form today, and five more do not know whether one exists; this closes the gap in a week.

Open the accommodation channel. Stand up the intake workflow in Playbook Chapter 5, meaning a named mailbox, a 48-hour response SLA, and a funding rule decided in advance, even before the VMS can support it natively. Playbook Section 5.3 is the workaround playbook.

Fix the front end. Issue the job-description and requisition standards from Playbook Section 6.1, and ask every screening-tool vendor the disability question from the AI audit checklist in Playbook Section 8.2: does your validation measure adverse impact on candidates with disabilities? Almost none do; New York's bias-audit law, the first in the nation, does not require it.¹⁶

Put one metric on the supplier scorecard. Nine of twelve practitioners rate it a high-impact intervention, and it is the one an MSP can execute inside existing governance this quarter. Playbook Section 9.2 provides definitions and weightings; start with accommodation-request response time and self-ID invitation coverage.

The Working Group Next Phase

The Working Group moves from research to build in the 2026-2027 program year, and the Council invites program owners, MSPs, suppliers, and technology providers to join. The operating cadence: monthly 60-minute virtual working sessions organized around the playbook chapters (contracts one month, technology the next), quarterly steering reviews against the network benchmark, and an annual in-person summit hosted on rotation by the founding partners. Contribution paths are deliberately concrete: pilot the 90-day quick start and report results; contribute contract language or workflow designs to the playbook's next edition; join the annual anonymized benchmark; or host a working session. Membership carries one obligation: implement something and report what happened, because the next edition of this report should be written from results, not intentions.

The Council also commits to maintaining its employer program dataset as standing infrastructure: quarterly re-verification against the program-active standard (Section 4.4 of the full report), publication of aggregate durability findings, and direct outreach to programs whose public status is uncertain. That outreach serves double duty, as primary data nobody else collects and as an invitation into this Working Group.

The Internship Connection Pipeline

The Council's university partnership converts research into pipeline. The University of South Carolina Lancaster, institutional partner for this report's survey phase, works with the founding partners to place students, including disabled and neurodivergent students, into structured internship and project roles inside Council member programs: staffing operations, MSP program offices, and client workforce teams. The mechanics are deliberately simple: partners commit internship slots per semester; USC Lancaster's career office and disability services coordinate candidate preparation, disclosure counseling, and accommodation planning with the host; and the Working Group collects conversion and retention outcomes for the benchmark. The pipeline serves two purposes at once. It is a talent channel for member programs and a live training ground for the inclusive-program practices this report describes, operated by the people who wrote it.

Adopting the Companion Playbook

Everything this report recommends is operationalized in the companion volume, The Inclusion Playbook: Policies and Practices for Disability and Neurodivergent Inclusion in Contingent Programs. The playbook contains the maturity self-assessment and scoring sheet (its Chapter 1 and Appendix B), the RACI matrix and model policy statement (Chapter 2), model contract clauses and rate-card structures (Chapter 4), the end-to-end accommodation workflow, funding decision tree, and VMS workaround playbook (Chapter 5), sourcing and assessment standards (Chapter 6), training curriculum and verification mechanics (Chapter 7), technology and AI checklists (Chapter 8), the scorecard metric definitions (Chapter 9), the co-employment plain-answer guide (Chapter 10), caregiver flexibility provisions (Chapter 11), and the 90-day quick start (Chapter 12) that sequences all of it. Read this report for the case; pick up the playbook for the build. Both are Council publications, and both will be revised annually from benchmark results and member experience.

The Pathways Opportunity Council is an invite-only council of workforce professionals focused on expanding access and economic mobility for underrepresented talent, including Black, Hispanic, disabled, and veteran communities, as well as other historically disadvantaged groups. The Council’s mission is to elevate industry practice and the professional standing of its members by convening leaders committed to measurable improvements in access, hiring, and advancement outcomes. The Council prioritizes organizations that demonstrate sustained, authentic commitment to expanding opportunity. Research and thought leadership are sponsored by Primary Talent Partners and AgileOne and executed by Defiant Solutions as the Council’s named consulting partner. For more information email us at pathways@defiantsolutions.com.