The Evidentiary Gap Runs Both Directions

Keisha's analysis in Complaint Volume Isn't the Right Metric for Community Voice lands on something practitioners often dance around: the definition of harm in accessibility enforcement isn't neutral. It was constructed through legal and regulatory processes that largely excluded the people most likely to experience that harm. That structural critique is correct, and it should inform how agencies and covered entities think about their accountability frameworks.

But there's a second evidentiary problem that the original framing doesn't fully surface — one that complicates the picture for practitioners trying to act in good faith. Covered entities are also operating with systematically degraded information about accessibility failures. The problem isn't just that communities can't get complaints heard. It's that the feedback loops connecting accessibility failures to institutional decision-making are broken in both directions.

What Covered Entities Don't Know About Their Own Failures

The Great Lakes ADA Center (opens in new window) and similar regional technical assistance providers have documented for years that most accessibility barriers are never reported through formal channels. People encounter inaccessible websites, untranslated documents, or broken physical accommodations and they leave — they find workarounds, they abandon the interaction, or they simply don't access the service at all. That exit behavior is invisible to the covered entity unless they're actively measuring it.

This isn't a defense of institutional inaction. It's a description of an information environment that produces bad outcomes even when institutional intent is relatively good. A transit agency that genuinely wants to improve language access has no reliable mechanism to know which routes are generating the most navigation failures for limited-English riders if those riders aren't filing formal complaints — and the DOJ's Title VI guidance (opens in new window) doesn't require them to build one.

The Pacific ADA Center (opens in new window) has made a similar observation about physical accessibility: the gap between formal ADA complaints and actual barrier prevalence is enormous, and complaint data systematically underrepresents barriers in communities with lower civic infrastructure and less familiarity with formal grievance processes. Relying on complaint volume as a proxy for barrier prevalence produces a distorted map of where problems actually exist.

The Accountability Gap That Compliance Testing Can't Close

At this publication, we approach accessibility accountability through what we call the CORS framework — examining Community impact, Operational realities, Risk distribution, and Strategic alignment together rather than in isolation. Keisha's piece is strongest on the Community dimension: who gets to define harm, and whose testimony counts as evidence. That analysis is essential.

But the Operational and Risk dimensions reveal a different kind of accountability gap. Covered entities that want to do better than minimum compliance have very few reliable tools for identifying where their systems are actually failing. Section 508 conformance testing (opens in new window) catches technical failures in digital products, but it doesn't tell you whether the people who need those products can actually use them to accomplish their goals. WCAG 2.1 success criteria (opens in new window) define technical thresholds, not functional outcomes. An agency can pass every automated accessibility check and still produce a website that's effectively unusable for someone with low digital literacy, cognitive differences, or limited English proficiency.

This is where the evidentiary gap compounds. Communities lack power to define harm. Covered entities lack reliable mechanisms to detect harm. The result is a system where the DOJ's complaint resolution process (opens in new window) becomes the primary feedback mechanism for both — which is precisely why that process is so overloaded and why resolution timelines stretch into years.

Community-Defined Metrics as Evidentiary Infrastructure

As explored in the original analysis, the solution isn't simply lowering barriers to formal complaint submission. That's necessary but insufficient. The deeper intervention is building community-defined metrics into the evidentiary infrastructure that covered entities actually use.

The Northeast ADA Center (opens in new window) has done work on participatory accessibility auditing that points toward what this could look like in practice: structured processes where community members define the functional outcomes that matter to them, test systems against those outcomes, and generate documentation that carries evidentiary weight in both internal accountability processes and formal enforcement contexts. This isn't a replacement for legal enforcement — it's a way to generate the kind of granular, community-grounded evidence that formal complaint processes were never designed to produce.

The Southeast ADA Center (opens in new window) has taken a similar approach in its technical assistance work, emphasizing that covered entities need feedback mechanisms that don't depend on community members knowing their rights, finding the right intake portal, and sustaining engagement through a multi-year resolution process. Those are high barriers even for people with significant civic resources.

What this requires from covered entities is a different conception of their accountability obligations — not just responding to formal complaints, but actively building the feedback infrastructure that makes community-defined harm visible. That's a harder ask than improving intake systems, but it's the intervention that actually addresses the evidentiary problem from both directions.

What Practitioners Should Do Differently

Covered entities that rely exclusively on formal complaint data to understand their accessibility failures are making a bet that the absence of complaints reflects the absence of problems. That bet is almost certainly wrong. It creates compounding risk over time — both legal risk as enforcement capacity increases, and reputational risk as communities develop better documentation tools and advocacy infrastructure. More fundamentally, it means real people are being turned away from services they're entitled to, invisibly, with no institutional record of the failure.

Building proactive feedback mechanisms isn't only a risk management decision. It's a more accurate way to understand whether an institution is actually serving the people it exists to serve. An agency that knows where its language access failures are concentrated can allocate resources more effectively than one waiting for the DOJ to tell it.

The practical implication for practitioners, building on Keisha's framework, is this: treat community-defined harm as a data source, not just a political consideration. The evidentiary gap she identifies as a power problem is also a measurement problem — and measurement problems have solutions that don't require waiting for enforcement systems to reform themselves. The starting point is asking what feedback infrastructure you would need to know, right now, whether the people you serve can actually use what you've built.