Emergency and crisis lines make queue performance unusually visible. They also make comparison unusually hazardous. A percentage reported by the 988 Suicide & Crisis Lifeline, a 911 public-safety answering point’s local target, and a company’s telephone service level may all describe “answering,” but they do not necessarily share a denominator, clock, call type, or consequence.
This review uses federal and federally sponsored sources available through August 12, 2026. Its principal periods are the post-launch 988 era beginning July 16, 2022, the latest periods displayed in SAMHSA’s changing metrics series, and the reporting years represented by National 911 Program and GAO materials. The units include contacts offered, contacts answered, answer rates, speed measures, 911 requests for service, and organizational records. The cohorts are people contacting 988 and people or devices reaching emergency communications systems—not a sample of ordinary business callers. The primary geography is the United States, although operational reporting can be national, state, territory, tribal, or local.
988 supplies a national series, not a generic telephone benchmark
SAMHSA’s performance-metrics page is the most direct official source for demand and response trends across the 988 system. It reports the measures in defined time intervals and separates modalities such as voice, text, and chat. That separation matters: a “contact” is not always a telephone call, and modality-specific staffing and handling patterns can differ. The sound comparisons are like-for-like—for example, voice contacts answered as a share of voice contacts offered during the same reporting period.
The page’s figures should be read with their labels intact. Offered contacts measure attempted demand presented to the network; answered contacts measure successful connections under the system’s definition; an answer rate is a ratio, not a count; and a speed measure concerns elapsed time rather than whether a crisis was resolved. A rise in answered volume can coexist with a lower answer rate if offered demand rises faster. Conversely, a high answer rate says little by itself about clinical quality, repeat contacts, transfers, abandonment timing, or outcomes after connection.
SAMHSA’s 988 FAQ establishes the service boundary. 988 provides access to suicide, mental-health, and substance-use crisis support by call, text, or chat. It is distinct from 911, even though escalation to emergency services can occur when necessary. The 988 network description further explains that the national number is supported by a distributed network of crisis centers. National aggregation therefore combines locally handled work, network routing, backup capacity, and multiple contact modes; it is not the performance record of one homogeneous queue.
Routing rules can also affect both experience and measurement. The FCC’s 988 overview describes the national dialing framework. In 2024, the Commission adopted georouting rules intended to route wireless 988 calls using approximate location while protecting precise location information. The relevant cohort changes as providers implement georouting: routing by approximate physical location can send some calls to a different crisis center than routing based on a telephone number’s area code. Before-and-after local comparisons may consequently reflect routing changes as well as demand or staffing changes.
911 data are operationally important but structurally fragmented
The National 911 Program’s statistics and data resource documents the effort to improve knowledge about emergency communications nationwide. The central analytical fact is decentralization. 911 is delivered through state, territorial, regional, tribal, and local arrangements, with differences in governance, technology, terminology, and reporting coverage. A national total can describe scale, but it does not automatically produce a nationally comparable answer-time distribution.
That limitation is the reason the 911 DataPath initiative matters. DataPath addresses consistent exchange and use of 911 data rather than assuming that every jurisdiction already records the same event in the same way. Definitions must distinguish calls received, calls answered, incidents created, transfers, abandoned or misdialed calls, texts, and machine-originated communications. Even the starting and stopping events for an answer interval can differ among systems.
The National 911 Program’s Next Generation 911 overview describes the transition toward standards-based, Internet Protocol emergency-services networks capable of carrying voice and additional data. NG911 is not simply a faster version of legacy 911. It changes what can enter the system, how traffic can be routed, and how centers can interoperate. Historical voice-call series therefore may not be fully comparable with future multimedia request-for-service series.
Federal communications rules provide another layer, but they should not be mistaken for a universal queue target. The FCC’s 911 and E911 guide explains carrier and location obligations. 47 CFR Part 9 contains federal requirements addressing 911 capabilities, including provisions for covered services and systems. These rules concern access, routing, notification, and location-related responsibilities. They do not convert a private company’s incoming sales or support queue into a public-safety answering point, nor do they establish one federal business answer-time service level.
GAO report GAO-24-106070 adds an oversight perspective on the 911 environment. GAO evidence is valuable for institutional and implementation questions, but its scope, agencies, selected sites, interviews, and review period must govern interpretation. Findings about participating emergency communications entities should not be projected mechanically onto every U.S. center, much less onto commercial reception.
“Non-emergency” is a routing class, not one comparable dataset
Non-emergency demand is often discussed as though it were a third national queue beside 911 and 988. It is not. A local police or municipal non-emergency line may accept delayed crime reports, noise complaints, administrative questions, welfare concerns, or service requests. Some calls are transferred into or out of 911; some jurisdictions use 311 or separate agency numbers. Consequently, unit, cohort, and geography vary by agency. The cited national sources do not supply a comprehensive U.S. series for all non-emergency telephone demand or a single answer-performance standard.
This gap has practical implications for analysis. A jurisdiction can examine its own hourly arrivals, transfer rate, abandonment distribution, and call-purpose mix, but it should not label those records “national non-emergency demand.” When an administrative line receives work displaced from 911, lower 911 volume may coincide with higher non-emergency volume without any decline in total public requests. Cross-line analysis requires stable categorization and a documented period.
Public-safety measures do not define a business SLA
Public-safety answering objectives are designed around emergency risk, statutory or regulatory duties, local policy, and specialized operational standards. A business service-level agreement is a negotiated commitment for a defined customer population and service scope. The two may use similarly worded measures while remaining non-equivalent.
At minimum, any comparison should retain five definitions: which inbound events enter the denominator; when the timer begins; what event counts as answered; whether very short abandons are excluded; and whether the statistic is a mean, median, percentile, or threshold share. It should also identify the reporting period, operating hours, channel, overflow and transfer treatment, outage handling, and geography. Without these elements, “answered in seconds” can conceal substantial differences.
Emergency queues also face demand and consequences unlike ordinary commercial contact. 911 can involve immediate threats to life and location-dependent dispatch. 988 combines crisis support, risk assessment, and network routing. A non-emergency public line can still surface an emergency requiring reclassification. Those facts justify public-safety-specific controls; they do not provide evidence that a particular emergency target is optimal, feasible, or contractually appropriate for every business queue.
Limits and responsible conclusions
These sources are authoritative for their respective programs, but not interchangeable. SAMHSA’s dashboard is an administrative network series, not a randomized study of all people in crisis; contacts never attempted are absent, repeat contacts may appear more than once, and displayed periods or definitions can be revised. 988 network materials describe the participating system and may emphasize program operation. FCC materials establish regulatory context, not caller satisfaction. National 911 Program data depend on participating jurisdictions and harmonization. GAO’s conclusions inherit the scope of its audit design. None measures the universe of private business calls.
The defensible conclusion is narrow but useful: 988 demonstrates how a national crisis network can publish modality-specific demand and answer measures; 911 evidence shows why decentralized emergency data require shared definitions; and non-emergency demand remains locally heterogeneous. These records can inform measurement design. They cannot, by themselves, set a business answering SLA. Any business target requires its own demand distribution, caller risks, operating period, channel mix, abandonment definition, and contractual context.