Here's a number that should keep every Program Director up at night: among workers with disabilities, only 72.8% are still at their job one year later. Compare that to 81.9% for workers without disabilities. That 9-point gap isn't about ability. It's about what happens after the initial placement ends.

We spend months—sometimes years—building skills, documenting progress, celebrating milestones. Then the case closes. The file moves to "completed." And for too many participants, the progress we worked so hard to create begins to unravel.

This isn't a failure of the people we serve. It's a failure of how we define success.

The 90-Day Illusion

Most programs measure success at a single point: discharge. Did they complete the curriculum? Did they get placed in a job? Did they hit their ISP goals? Check, check, check. Case closed.

But here's what the data tells us: for adults with autism in supported employment, initial job retention starts at 100% at three months. By eighteen months? It drops to 74.3%. A quarter of the progress—gone.

The real question, as one education researcher put it, "is not where did they go after the program, but how well are they doing once they've left."

This is the mindset shift every program needs to make. Success isn't about the discharge report. It's about what life looks like six months later. A year later. Five years later.

Why Outcomes Fade

When we dig into why progress doesn't stick, a few patterns emerge consistently across research and practice.

Skills don't transfer automatically. Someone might master a task in your day program but struggle to perform it at home or on a job site. Different people, different environments, different expectations. If we don't intentionally teach skills to work across contexts, they stay trapped in the setting where they were learned.

Support disappears abruptly. During services, participants have structure, check-ins, coaches, and accountability. Then the case closes and all of that vanishes overnight. It's like removing the scaffolding before the concrete has fully cured.

Nobody's watching for warning signs. Without ongoing visibility, early signs of struggle go unnoticed. A missed shift here, a skipped routine there—small slips that compound into major setbacks before anyone realizes there's a problem.

Consistency breaks down. When a participant moves from your program to a job site to a residential setting, information doesn't always follow. The new supervisor doesn't know about the morning routine that prevents meltdowns. The family doesn't know what strategies worked during the day program. Each environment operates in isolation.

These aren't character flaws or lack of motivation. They're predictable, preventable system failures.

Reframing What We Measure

The field is starting to recognize this gap. A recent national consensus effort identified 43 priority outcomes that actually matter to people with intellectual and developmental disabilities—things like autonomy, relationships, health, and quality of life. Not just "services received" or "goals met."

As one industry expert frames it: "True success isn't just completing a task; it's completing it with progressively less support until the person can do it on their own."

That's the key insight. We should be measuring not just whether someone can perform a skill, but how much support they still need to perform it. And we should be tracking that over time—not just at discharge, but months and years into the future.

This requires a fundamental shift in how we think about our work. We're not in the business of completing programs. We're in the business of building lasting independence.

The Four Pillars of Durable Outcomes

Based on the research and hard-won lessons from programs that get this right, outcomes that stick rest on four foundations.

Pillar 1: Maintenance Systems

Skills fade without practice. This isn't speculation—it's one of the most consistent findings across behavioral research. When skills aren't maintained, programs spend significant time and resources re-teaching what was already learned.

The solution isn't complicated: create lightweight systems for ongoing practice and reinforcement after the formal program ends. Some programs schedule follow-up sessions at regular intervals. Others build in "booster" check-ins when early warning signs appear. Any structured follow-up is better than none.

We go deeper on what this looks like practically in The 30-60-90 Day Follow-Up Playbook.

Pillar 2: Generalization by Design

A skill that only works in one setting isn't really mastered—it's memorized. If someone learns to follow a schedule in your program but can't apply that skill at a job site, you haven't taught scheduling. You've taught compliance in a specific context.

Generalization doesn't happen automatically. It requires intentional practice across different settings, with different people, using varied materials. It means teaching the principle behind the skill, not just the rote behavior.

The good news: skills that generalize are also more durable. They're less likely to fade because they're not dependent on specific environmental cues. We explore this distinction more in Maintenance vs. Generalization.

Pillar 3: Early Warning Detection

Research on workers with autism found that at the six-month mark, many needed increased support to prevent decline—not less. The ones who got that additional support maintained their jobs. The ones who didn't often lost them.

This means building workflows that make early warning signs visible to the team—e.g., consistent check-in notes, routine adherence observations, incident logs, and qualitative changes staff notice over time. By the time someone is in full regression, you're doing damage control instead of prevention.

Leading Indicators of Regression breaks down what to watch for and when.

Pillar 4: Environmental Consistency

Outcomes collapse when support doesn't follow the person across settings. What works in the classroom fails at the job site. What the day program knows, the residential team doesn't.

The most successful programs build bridges between environments—shared documentation, consistent strategies, warm handoffs. Not just information transfer, but genuine coordination so that support stays consistent regardless of who's providing it or where.

We dig into this challenge in Why Outcomes Collapse Across Environments.

The Interdependence Reframe

Here's where we need to be honest about something. "Lasting independence" doesn't mean abandoning people to fend entirely for themselves. As one disability advocate pointedly observed: "Life for all of us is an existence of interdependence. Independence is not my goal."

She's not arguing against skill-building. She's arguing against a framework that treats any ongoing support as failure.

The goal isn't zero support. The goal is sustainable support—the right mix of personal capacity and community connection that allows someone to thrive long-term. For some people, that means a weekly check-in call. For others, it means daily coaching. Neither is "better" or "worse." Both can be lasting success.

This reframe matters because it changes what we're building toward. Not an exit point where someone is finally "fixed" and sent off alone. But an ongoing life where they have the skills they need and the support network to fill in the gaps—just like everyone else has.

What This Means for Your Program

If you're running a CRP, a day program, or an employment service, here's the practical takeaway: your work doesn't end at discharge.

Start building these elements into your program design:

Extend your outcome window. Stop measuring success only at case closure. Track what happens at 30, 60, 90 days—and ideally 6 and 12 months. You can't improve what you don't measure.

Design for maintenance from day one. Every skill you teach should come with a plan for how the person will continue practicing it after services end. Build in the follow-up cadence before you need it.

Teach for transfer. If you're only teaching skills in one setting with one staff member, you're setting people up to struggle. Vary the context, vary the people, vary the approach.

Create handoff systems. When someone transitions out of your program, make sure the people in their next environment know what works. Document strategies, not just goals. Share what you've learned about this specific person.

Watch for warning signs. Build systems to catch early indicators of regression. A brief check-in call at the right moment can prevent months of lost progress.

None of this requires massive new infrastructure. It requires a shift in how you define your job—from "delivering services" to "building outcomes that last."

The Bottom Line

The programs that achieve lasting outcomes share one thing: they don't celebrate too early. They understand that the discharge report is the beginning of the real test, not the end.

Every person who walks through your program has decades of life ahead of them. The question isn't whether you helped them during their time with you. The question is whether that help made a lasting difference in the life they live after.

That's the outcome that matters. That's what we should be building toward.

The rest of this cluster explores how to make it happen—from identifying who's at risk of regression to building lightweight follow-up systems that catch problems early. The research is clear: lasting outcomes aren't luck. They're the result of intentional design.

‍