Key Takeaways

  • A rise in voluntary incident reports does not, by itself, demonstrate that incidents are becoming more frequent. It may reflect changing willingness or opportunities to report.

  • NASA explicitly warns that its Aviation Safety Reporting System (ASRS) database cannot establish how prevalent a problem is across aviation, because reports are voluntary, selective and generally unverified.

  • For organizational leaders, the distinction is between a reporting system that reveals possible hazards and a measurement system that can estimate incident rates. They serve different purposes.

Imagine a business launches a confidential channel for reporting near misses. The following quarter, reports double. One executive sees proof that the operation has become more dangerous; another celebrates the rise as evidence of greater trust. Neither conclusion follows from the count alone. The number of submitted reports is the product of what people experience, what they notice, what they consider worth reporting, and whether they believe reporting is safe and useful. Without knowing how those factors changed, the count cannot isolate the underlying level of risk.

Why does more reporting not necessarily mean more incidents?

Because reporting is a behavior as well as a record of an event. People can experience the same type of problem and make different choices about whether to describe it formally. A new channel, a reminder campaign, an assurance of confidentiality or a highly visible incident may change participation even if the underlying hazards have not changed. Conversely, a quiet reporting system might mean few events, or it might mean that employees see little value in using it. These are plausible explanations, not diagnoses that can be made from a trend line alone.

NASA makes the limitation explicit in its guidance on ASRS data: reports are voluntary, subject to self-reporting biases and often not corroborated by regulators or investigators. NASA states that the presence of records on a topic cannot be used to infer the prevalence of that problem in the National Airspace System. That is a particularly important warning because the ASRS database contains detailed, professionally coded accounts and can look authoritative enough to invite statistical conclusions it cannot support.

What can NASA's ASRS actually tell us?

ASRS is designed to capture safety concerns, near misses and operational experiences that might otherwise remain hidden. Its value is largely diagnostic: the narratives can reveal ways in which equipment, procedures, workload, expectations and communication interact. NASA's database description explains that analyst-coded records and de-identified narratives support human-factors research, policy development and education. It also says that NASA does not independently verify the details in submitted reports.

That distinction does not make the reports useless. A report can provide a credible reason to investigate a possible hazard without proving how often it occurs. If several people describe a similar confusing interface, for example, their accounts may illuminate a failure mode worth examining. But the reports alone cannot establish the proportion of all users who encountered it. A reporting system is particularly good at revealing possibilities; estimating frequencies requires a different design.

NASA also describes how its analysts remove identifying details, sometimes generalizing dates, locations and other fields to protect reporters (ASRS, About Data). That protection is important to the system's purpose, but it is another reason not to treat the published records as a conventional census of incidents.

What makes voluntary reporting useful if the counts are biased?

It can reveal mechanisms that routine outcome measures miss. A recorded accident tells an organization that harm occurred. A near-miss narrative may describe a confusing handoff, a misleading indication or an unexpected dependency that did not produce harm this time. These descriptions can help decision-makers ask better questions about how the operation behaves. NASA's ASRS program briefing describes uses that include identifying system deficiencies, alerting relevant authorities and sharing lessons through its publications.

Confidentiality is central to the program's design. NASA's confidentiality guidance describes de-identification and limited FAA enforcement-related incentives, subject to eligibility conditions. It would be inaccurate to summarize those arrangements as immunity from all consequences. The protections have boundaries, and the program does not replace other mandatory reporting obligations. For business readers, the transferable insight is not that every organization should copy aviation's legal arrangements. It is that the quality of a signal depends partly on the conditions under which people choose to share it.

What is the difference between a reporting count and an incident rate?

A reporting count is the number of submissions received under a particular reporting system and definition. An incident rate attempts to relate a defined number of events to a relevant exposure, such as operating hours or transactions, over a defined period. Even a rate is only as good as its event definition, coverage and data collection. Voluntary narratives generally lack a reliable denominator and may have unknown reporting probabilities. Dividing their count by a convenient exposure measure does not automatically repair those limitations.

For example, suppose a hypothetical company receives 40 reports in one quarter and 60 in the next. The arithmetic increase is 50%, but that figure describes submissions, not necessarily the incidence of the underlying hazard. If the workforce expanded, the reporting channel became easier to use or employees grew more confident about reporting, the interpretation changes. If the organization also has consistent operational event records, those might provide additional context, but the two sources should not be treated as interchangeable.

What can go wrong when managers reward a low report count?

AeroLeadership's interpretation is that a low-count target can confuse silence with reliability. If people believe fewer reports are evidence of better management, they may face an incentive to keep the record clean rather than make the operation understandable. That is a risk of the measurement design, not proof that every low-count organization suppresses reporting. Similarly, praising a rising count without investigating its contents can mistake activity for learning.

The more consequential question is whether the organization can turn uncertain, sometimes incomplete observations into appropriately investigated concerns. This connects to why employees stop sharing bad news with managers: access to a reporting channel is not identical to confidence that the information will be heard. It also connects to the broader problem of warning signs that exist but fail to reach a decision.

What should leaders conclude from a change in report volume?

They should conclude that the reporting system received a different number of submissions, then investigate what may explain the change. The reports may indicate a new hazard, greater awareness, altered trust, a change in operations or several of these at once. They should resist treating the trend as a direct safety score. The most useful contribution of voluntary reporting is often not the size of the number but the possibility of seeing a problem before a more consequential outcome makes it undeniable.

That is a bounded lesson from aviation, not a claim that every industry shares its reporting conditions. AeroLeadership's organizational reliability framework considers how leadership, communication and awareness interact with risk; reporting is one example of how those connections affect what an organization can know.

Sources and further reading