988 launched without 911’s plumbing: location, dispatch and a fee

The week

On Saturday, July 16, 2022, something simple and long overdue happened. Anyone in the United States in a mental health or suicidal crisis could dial, text or chat with three digits, 988, and reach help. Calls, texts and chats to 988 began reaching the National Suicide Prevention Lifeline, which was renamed the 988 Suicide & Crisis Lifeline.

The number had been a while in the making. The National Suicide Hotline Improvement Act of 2018 asked the Federal Communications Commission (FCC), the federal agency that regulates phone networks, to study a short dialing code. The FCC designated 988 on July 16, 2020 and gave carriers two years to make it work, which is why launch fell on a Saturday: the date followed the regulatory clock, not a publicity plan. The National Suicide Hotline Designation Act, signed that October, put the number into law. It also allowed states to fund 988 services through fees on phone lines, the same way most states pay for 911 through a small monthly charge on every phone bill.

The FCC’s 2020 order applied broadly. Wireless carriers, landline providers and providers of internet-based phone service all had to connect 988 by the July 2022 deadline. In the meantime, states prepared. With SAMHSA funding, the network administrator gave planning grants to states and territories in 2021. The grants paid for implementation plans covering who would answer calls, how centers would be paid and how they would connect to local services. The plans gave state agencies their first structured look at the gaps they would inherit, though few had the funding to close them by launch day.

Texting came later. In November 2021, the FCC required wireless carriers and text providers to route text messages sent to 988 to the Lifeline by the same July 2022 date. The order deliberately left out any requirement to pass location information with texts, a choice that shaped what states could do for the next several years.

The Lifeline itself dated from 2005. It is a network of local and state crisis centers rather than one big call center. A national administrator, Vibrant Emotional Health, coordinates it under the Substance Abuse and Mental Health Services Administration (SAMHSA), the federal agency responsible for behavioral health. More than 200 centers belonged to the network at launch. When someone calls, the system tries to connect them to a nearby center staffed by trained counselors who know the local services. Veterans could press 1 to reach the Veterans Crisis Line.

Federal funding rose ahead of the launch, and the old ten-digit number kept working, so nobody who had memorized it was left stranded. A separate federal measure took effect a few months before launch: from April 2022, the American Rescue Plan Act let state Medicaid programs cover community-based mobile crisis services with an enhanced federal match. A handful of states, starting with Virginia in 2021, had passed laws creating dedicated 988 fees. Most had not.

Two details of how calls were routed

Two details of how the system worked at launch mattered a great deal. First, wireless calls were routed by the caller’s area code rather than their location. The network looked at the first three digits of the caller’s phone number and sent the call to a center in that area, wherever the caller actually was. Second, calls that a local center could not answer rolled over to national backup centers. Those backup counselors could talk someone through a crisis, but they were far less likely to know the local mobile crisis team, the nearest walk-in clinic or the county’s emergency procedures.

The federal blueprint for what should sit behind the number already existed. SAMHSA’s 2020 national guidelines for behavioral health crisis care described three elements: a crisis line to call, a mobile team to respond and a crisis facility to go to. On launch day, the first element was national. The other two depended on where a caller lived.

The number was easy to remember; the local services behind it were still being built.

Take a nineteen-year-old college student in Ohio, late at night, in real distress. She has heard the new number on a poster in her dormitory and dials 988. Her phone number still carries the Texas area code she had in high school. Her call goes to a crisis center in Texas. The counselor who answers is skilled and kind, but has no idea which services exist near her campus, and could not send anyone to her even if it became necessary.

That scene captures the central point about 988’s launch. A short number is a promise of simplicity. Behind 911 that promise rests on infrastructure built over decades. Networks pass the caller’s location to the answering point. Dispatch links send police, fire or ambulance crews. A fee on phone bills in nearly every state pays for both. When 988 went live, it had the number and a network of crisis centers. It had none of those three supports in finished form.

That framing matters for state agencies because each gap landed on a different owner and called for a different kind of work. Location was a telecom and regulatory problem. Dispatch was a service design problem shared by behavioral health, 911 and emergency services. Funding was a budget and legislative problem. Together they cover what stands between a person dialing and a person being helped, and none of them is solved by publicizing the number.

The three layers hidden behind 911

Most people think of 911 as a phone number. It is better understood as a system with three hidden layers, and seeing them clearly explains why 988 started with gaps.

The first layer is routing and location. When you dial 911 from a mobile phone, the network looks beyond your phone number. It works out roughly where you are, from the cell tower you are connected to and, increasingly, from your phone’s own location data, and sends the call to the local answering point, known as a public safety answering point. The dispatcher’s screen typically shows an approximate location before anyone speaks. Routing a call based on where the caller physically is, rather than on their number, is often called georouting.

The second layer is dispatch. A 911 dispatcher can see which ambulances, fire engines and police units are available, and send one. That rests on computer-aided dispatch systems, shared radio networks and agreements between agencies, built up over many years.

The third layer is funding. Most states charge a small recurring fee on every phone line and use it to pay for 911 call-takers, equipment and upgrades. Because the fee rises with the number of phone lines, the money keeps coming year after year without a fresh political fight.

988 arrived with a number and a network of skilled crisis centers. The three hidden layers were, at best, partly built. That is not a criticism of the people who launched it. Building a national number first and the plumbing afterward was a defensible choice. But it meant that the work of making 988 dependable fell largely on state agencies after launch day.

How 911 acquired its layers

The comparison is fair only if one remembers how long 911 took. The number dates from the late 1960s. Automatic location for landline calls spread over the following decades. Location for wireless calls came through FCC rules phased in from the late 1990s, first identifying the cell tower and later requiring a more precise estimate of the caller’s position. Fees on phone lines were adopted state by state. Even now, 911 is in the middle of another transition, to internet-based systems often called Next Generation 911.

Each of those layers needed carriers, regulators, state legislatures and local agencies to act, often in that order. 988 inherited none of that history. Its designers chose to launch the number on the existing Lifeline network and build the rest afterward, which was faster but left the sequencing to the states.

Who owns each piece

The ownership map explains why progress was uneven. Crisis centers are mostly nonprofit organizations working under state or county contracts. 911 answering points are usually run by counties, cities or regional authorities, governed by their own boards and funded by their own fees. Mobile crisis teams may be run by county behavioral health departments, community mental health centers or private providers billing Medicaid. Crisis receiving facilities may be run by hospitals, health systems or community providers.

No single official controls all of those. A state behavioral health authority can fund and contract, but it does not run 911. A 911 director can change call-handling protocols, but does not employ counselors. That is why the gaps below each have a different owner, and why closing them needs agreements across agencies more than new technology.

Location: where the call goes and where help can be sent

At launch, a wireless call to 988 went to a crisis center chosen by the area code of the caller’s phone. Americans keep their numbers when they move. A student in Ohio with a Texas number reached a Texas center, which knew nothing about Ohio’s services and could not send anyone. A 911 call from the same phone would have reached the local answering point with an approximate location attached.

Agencies could not fix routing on their own. It needed carriers, the FCC and the network administrator to agree a method that routed by approximate location. That method could not disclose a caller’s precise position, since many people call a crisis line precisely because they do not want a response at their door. This is a genuine tension, and worth spelling out for anyone new to the field. For 911, precise location is the point. For a crisis line, a caller’s trust depends partly on knowing that reaching out will not automatically bring police to the door. The routing solution had to be good enough to reach a nearby center, while respecting that privacy.

What agencies could do was measure the effect. Monthly reports showed how many calls from in-state numbers were answered in state, but not how many in-state residents were answered elsewhere, or how many out-of-state numbers reached local centers. A state program office that asked its centers and the administrator for both directions had an honest picture of who it was actually serving.

Consider a state that draws university students and seasonal workers from elsewhere. Its official report might show a high in-state answer rate, because calls from its own area codes mostly reached its own centers. But thousands of people actually living in the state carry out-of-state numbers, and their calls were going elsewhere. Without asking for data in both directions, the state could believe it was serving its residents well while a large group was invisible to it.

The same gap shaped data sharing. Without location, a counselor who judged that someone was in imminent danger had to ask where they were, and callers did not always say. Protocols for that moment, including when a counselor may request an emergency location from a carrier, needed to be written down and trained rather than improvised.

How routing by location works without revealing it

The approach eventually adopted for 988 calls separates routing from disclosure. The carrier uses the location of the cell site and sector handling the call, which is roughly the area the caller is in, to decide which crisis center should receive it. The center does not receive the caller’s position. The call simply arrives at a center that serves that area.

That design answers both concerns. A caller in Ohio with a Texas number reaches an Ohio center that knows Ohio services. The caller’s privacy is preserved because the center learns nothing about where they are unless they say. The approach has limits near state and county boundaries, where a cell site may serve people on both sides of a line. States sharing a metro area across a border have reason to agree how such calls are handled.

When a counselor needs to know where someone is

The hardest moments in crisis work come when a counselor believes a caller is in imminent danger and the caller will not or cannot say where they are. The Lifeline network has long had a policy for those situations. It favors working with the caller, and with people the caller trusts, and uses emergency services only when less intrusive steps have failed. When emergency services are needed, the counselor or center may need to request the caller’s location from the carrier under emergency procedures.

Those requests go through carriers’ emergency disclosure processes, and they take time. A center that knows which carrier to call, what information to provide and how to pass the result to the right 911 answering point will save minutes that matter. Those details belong in written procedures that are trained and tested, with clear rules on documentation, since each request involves overriding a caller’s privacy.

Text and chat: a separate routing problem

Texts and chats followed a different path. The 2021 texting order required carriers to deliver texts to the Lifeline but did not require location, so texts could not be routed by area at all. Chat came through a website, which knows even less about where a person is. At launch, text and chat were answered by a smaller set of centers than voice, many of them serving the whole country rather than one state.

That mattered because younger people favor text and chat, and because those conversations tend to run longer than calls. A state that wanted its own residents’ texts answered locally had to fund text-capable centers, train counselors for written conversations and wait for routing rules that could send texts its way. Until then, a state’s best lever was capacity: the more local centers that could answer text and chat, the more of its residents’ conversations would reach someone who knew local services once routing improved.

How the picture differs by state

Location problems weighed differently on different states. States with large universities, military bases or seasonal workforces had many residents with out-of-state numbers. Small states bordering large metro areas had heavy cross-border traffic in both directions. Rural states had fewer, larger crisis center catchment areas, so even a correctly routed call might reach a center far from the caller. Tribal nations, some of whose members live across state lines and use services from several jurisdictions, faced a further layer of complexity. A state program office that understood its own pattern could prioritize the right fixes.

Dispatch: what happens when a conversation is not enough

Most crisis contacts end on the phone, text or chat. A minority need someone to go to the person. For those, the vision behind 988 was a crisis continuum: mobile crisis teams of clinicians and peers, and crisis receiving and stabilization facilities as an alternative to emergency departments and jail.

The continuum has three parts, often summarized as someone to call, someone to come and somewhere to go. A mobile crisis team is a small group, typically a mental health clinician and a peer specialist with lived experience, who travel to the person in crisis. A crisis receiving and stabilization facility is a place where someone can be taken for short-term care that is calmer and more suitable than a hospital emergency room or a police cell.

In 2022 that continuum existed in some counties and not in others. Where there was no mobile team, a counselor’s main option in an emergency was 911, which often meant a police response the caller had hoped to avoid. Where teams existed, the link between a 988 center and the team was frequently a phone call between two systems that did not share records or status.

Closing that gap is concrete work. Agencies needed written agreements between 988 centers and 911 answering points on when calls transfer in each direction and what information goes with them. They needed a shared way to see which mobile teams were available, comparable to the unit status boards 911 dispatchers use. They needed consent and privacy rules that both sides understood, since behavioral health records carry stricter limits than dispatch notes, and counsel should review how state law and federal confidentiality rules apply. And they needed a metric that mattered to residents: how often a dispatched mobile team arrived, and how quickly.

Systems that show where help is available

SAMHSA’s 2020 crisis care guidelines described an approach often compared to air traffic control. A regional or statewide crisis hub tracks, in real time, which mobile teams are available, where they are, and which crisis beds are open. When a counselor decides a person needs a visit or a place to go, the hub can see the options immediately and coordinate the handoff. A few states had built versions of that hub before 988. Most had not.

The technology is not exotic. It needs a shared status board for teams and beds, a way for teams to update their status from the field, a secure method to pass referral details, and reporting on response times. The harder parts are organizational. Several providers must agree to update the same board consistently, the state must decide who runs it, and the data shared through it must respect each party’s confidentiality obligations. A status board that teams forget to update is worse than none, because counselors stop trusting it.

The confidentiality rules that apply

Data sharing between crisis services and emergency services runs into several overlapping rules, and which ones apply depends on who holds the data. HIPAA applies to covered entities, such as health care providers that bill electronically, and to their business associates. Some crisis centers are covered entities and some are not. Federal confidentiality rules for substance use disorder treatment records, known as Part 2, apply to certain programs and are stricter. Many states have their own mental health confidentiality laws, often stricter than HIPAA.

911 records follow different rules again. In some states, dispatch records can be subject to public records requests, with exceptions that vary by state. A behavioral health detail that is protected in a crisis center’s record may become far more exposed once it is written into a dispatch note. Agencies should work out with counsel what each party may share, what each may record and how long each keeps it, and build the data exchange to match.

Paying for the response as well as the call

Mobile crisis teams need their own funding stream. The American Rescue Plan Act gave states a new option, from April 2022, to cover qualifying community-based mobile crisis services in Medicaid. For a state’s first twelve quarters under the option, the federal government pays 85 percent of the cost of those services. To qualify, the teams have to meet federal conditions, including being multidisciplinary, including a behavioral health professional and being available around the clock.

The option changes the economics of mobile crisis for states that use it. Teams can bill for the people they serve who are on Medicaid, rather than relying entirely on grants. But billing requires systems: eligibility checks, documentation that meets Medicaid standards and claims processing. Small community providers often lack those systems, and states that want broad coverage may need to help build them. The enhanced match is also temporary, so the long-term cost returns to the state’s regular Medicaid budget.

Where health systems come in

Health systems are stakeholders whether or not they run crisis services. When there is no mobile team and no crisis receiving facility, people in crisis often end up in hospital emergency departments, where they may wait for hours or days for a psychiatric bed. A working crisis continuum diverts some of those visits to more suitable settings. Some health systems operate crisis stabilization units or partner with counties to run them. Others take referrals from mobile teams and need agreements on how patients are transferred and what records go with them. The same confidentiality questions apply, and the same need for written protocols.

Take a county that has done the work. A 988 counselor is on the phone with a man who is distressed and alone. She opens a simple dashboard showing that one mobile team is free, about twenty minutes away. With the caller’s agreement, she sends a request with the information the team needs, and the team confirms on the same screen. If the situation turns dangerous, a written protocol tells her exactly when and how to bring in 911, and what details to pass on. Afterward, the county can report how long the team took to arrive. Compare that with a county where the counselor must phone around to find out whether anyone is on duty: the difference is whether help arrives.

Funding: who pays for someone to answer at 3 a.m.

The last gap was money. 911 is funded largely by recurring fees that rise with the number of phone lines. 988 started on a mix of federal grants, state appropriations and local contributions. Grants pay for launches. They are poorly suited to paying a counselor’s salary for twenty years.

That mattered because the binding constraint was people. Crisis work requires recruitment, training, clinical supervision and retention in a field with high burnout. A center funded by a two-year grant hires on short contracts, loses staff when the grant ends and rebuilds. Calls it cannot answer roll over to backup centers, which further weakens the local connection to services.

The 2020 law let states choose a 911-style fee, and a handful did so early. Others relied on appropriations, meaning money voted by the legislature in each budget cycle. Either way, the useful budget question was the cost of a fully staffed day: voice, text and chat counselors on every shift, supervisors, mobile teams on call and the systems that connect them. Text and chat needed separate staffing models, because conversations run longer and peak at different hours from voice, and younger people prefer them.

It is worth making the staffing arithmetic concrete, because it is where budgets most often go wrong. A crisis line has to be answered every hour of every day, including weekends and holidays. Covering a single seat around the clock takes several full-time people once shifts, leave, training and turnover are counted. Each counselor also needs a supervisor available for hard calls, and supervisors need their own cover. Add text and chat, whose conversations are longer and cluster in the evening and late at night, and the number of people grows again. A budget that counts “seats” rather than people, or that prices only voice calls, will look affordable on paper and fall short in practice.

Technology costs are smaller but real. Centers need telephony and text platforms that connect to the national network, a record system for contacts and follow-up, and, if dispatch is to work, a link to mobile teams and 911. Those systems need maintenance, security and staff who know how to run them. It is not exotic, but all of it has to be paid for every year, which is precisely what one-off grants do badly.

Who else might pay

Public budgets are not the only possible source. People in crisis are often covered by Medicaid, Medicare or commercial insurance, and their insurers pay when the same person later reaches an emergency department or an inpatient unit. Advocates have argued that insurers should also pay for crisis services that prevent those costlier admissions. Some states have looked at requiring commercial insurers to cover crisis services. The difficulty is that a crisis line is designed to answer anyone, anonymously, without checking coverage. Billing works better for mobile visits and crisis facility stays, which involve identified people, than for calls.

Fees, appropriations and the 911 precedent

The states that moved first on fees chose the 911 model. Virginia passed a 988 fee in 2021, and Washington, Nevada and Colorado also legislated that year. California’s fee took effect at the start of 2023. Most states relied instead on appropriations, federal grants or a mix.

Each route has trade-offs. A per-line fee is stable and grows with the number of lines. But it falls equally on every phone line regardless of income, and collecting it requires rules for prepaid phones, internet calling and other services. Appropriations can be sized to need, but they must be won again in each budget cycle and can be cut in a downturn. Federal grants help with start-up costs but carry their own timelines.

911 offers a warning as well as a model. The FCC reports each year on whether states divert 911 fees to unrelated purposes, and some states have done so. The 2020 law required similar reporting for 988 fees and expected states to spend them on 988-related services. A fee is only a recurring source if the money stays in the program.

The case for launching first

There is a serious argument that launching the number before the plumbing was right. Building georouting, dispatch links and funding first could have delayed the number for years, and each of those layers would have been designed without knowing how people actually used 988. A national number created demand and visibility that pushed carriers, regulators and legislators to act. The ten-digit Lifeline had the same gaps and far less attention.

That argument holds as far as it goes. The risk lies in what came after: treating launch as success. A state that saw the number working and moved on left its residents with the gaps the launch had made more visible. The launch was a reasonable first step. The work that followed it was the test.

Agencies also had to decide which numbers they would publish. Answer rates are easy to report and look good quickly. In-state answer rates, wait times by channel, mobile team arrival times and follow-up contacts completed are harder to produce. They are also the measures that show whether the three gaps are closing.

Residents, police and the budget cycle

For a governor’s office, a health secretary or a county executive, 988 is easy to treat as a success the moment the number works. The three gaps are where the real risk lies.

Residents. A caller routed to a distant center, or a counselor with no team to send, is a person who may not get help in time. Mental health crises are among the most consequential moments in a person’s life.

Public safety and police. Where the continuum is missing, crisis calls default back to police, which many communities and police departments themselves want to avoid. A working dispatch link is how a state reduces that burden.

Budgets. Short-term grants create a cycle of hiring, losing and rehiring staff, which is expensive and erodes quality. A recurring revenue source, whether a fee or a stable appropriation, is the difference between a program and a pilot.

Legal and privacy exposure. Behavioral health information carries strict confidentiality protections. Sharing it between 988 centers, 911 and mobile teams without clear rules invites both harm to callers and legal risk to agencies.

Public trust. Publishing only the flattering numbers can buy short-term credit and long-term embarrassment. Honest measures build durable support from legislators and the public.

What the workforce and its systems need

Behind every answered contact is a counselor who has been recruited, trained, supervised and supported. Lifeline network centers follow network standards for training and for handling callers at risk. Many centers also employ peer specialists, people with lived experience of mental health crises, who bring credibility that clinical training cannot supply. Peer certification requirements vary by state, and states that want to expand the workforce can review whether their rules make it easier or harder for qualified peers to work.

Remote work changed crisis centers as it changed contact centers. Counselors answering from home need secure telephony, a records system they can reach safely and supervisors who can join a difficult call. Crisis centers hold some of the most sensitive data any organization can hold: notes on people at risk of suicide. They are often small nonprofits with limited security staff, and they are as exposed to ransomware and data theft as any hospital. State contracts can require baseline security controls and fund them, rather than leaving each center to manage alone.

The consequences also differ by type of government. A state behavioral health authority bears the program risk and the budget risk. A county running 911 bears the operational risk when crisis calls default to police. A city police department bears the risk of officers responding to situations they are not trained for. A health system bears the cost of emergency department boarding. Each of them benefits from a working continuum, which is why the most effective crisis systems tend to have a governance group in which all of them sit together.

Signs of a working crisis system

A state with a working crisis system tends to share a few traits. Its program office knows, in both directions, who its centers are serving and who is being answered elsewhere. Its 988 centers and 911 answering points have signed, tested agreements on transfers. Counselors can see mobile team availability in real time. Its budget reflects the full cost of round-the-clock staffing across voice, text and chat, and that budget comes from a recurring source. And its public reports include the harder measures as well as answer rates.

The data a program office should hold

A program office that wants to manage the three gaps needs data that comes from several sources. From the network administrator, it needs answer rates by channel, the share of calls answered in state, and the share of residents answered elsewhere. From its centers, it needs wait times, transfers to 911 and referrals to mobile teams. From mobile teams, it needs requests, dispatches, arrival times and outcomes. From 911, it needs the number of behavioral health calls transferred to 988 and the reverse.

Bringing that together is a data-sharing project with its own agreements and privacy reviews. The state does not need individual records for most of these measures; aggregate counts by month and region answer most management questions. Starting with aggregate data keeps the privacy burden low and lets the program show progress quickly.

Publication matters as much as collection. A dashboard released on a fixed schedule, with the same measures each time and short notes explaining changes, lets legislators and the public see trends rather than snapshots. It also disciplines the program office, because a measure that stalls in public draws questions that a measure kept internal never does. Measures that are not yet reliable can be published with a caveat, which is better than leaving them out and letting the easy numbers stand alone.

States that struggle tend to repeat the same errors:

• Mistaking awareness for capacity. Marketing the number without the staff and teams behind it raises demand faster than the system can meet it.
• Reading one-directional data. Reporting only in-state numbers answered in state hides residents who are being served elsewhere.
• Improvising the hardest moments. Leaving emergency location requests and 911 transfers to individual judgment, rather than written protocol and training.
• Funding staff with launch money. Paying for permanent roles out of temporary grants all but guarantees turnover.
• Treating text and chat as an add-on. These channels have different patterns and need their own staffing plans.
• Ignoring privacy design. Building data sharing between behavioral health and emergency services without legal review of what may be shared and when.

Testing the location, dispatch and funding gaps

The questions below are designed for a governor’s office, a health secretary or a county executive reviewing the crisis system once a year. Each maps to one of the three gaps or to the evidence that shows whether it is closing.

• How many of our residents are being answered by centers outside the state, and how many out-of-state callers are we answering here?
• Do our 988 centers and 911 answering points have written, tested agreements on when calls transfer each way, and what information goes with them?
• When a counselor needs to send a mobile team, can they see which teams are available right now, or do they have to phone around?
• What does a fully staffed day cost across voice, text, chat, supervision and mobile teams, and how much of that is paid from recurring revenue?
• Which measures do we publish, and do they show whether the gaps in location, dispatch and funding are closing?

The number was the easy part. These questions test the location, dispatch and funding behind it, which a caller never sees.

With hindsight

Contact volumes rose after launch, and answer rates improved as funding grew. SAMHSA reported nearly five million contacts in the first year, well above the old ten-digit line’s volume. The number did its first job: people used it.

Location was fixed in stages. In September 2024, Verizon and T-Mobile began routing 988 calls by approximate location voluntarily, covering about half of Lifeline calls according to HHS. In October 2024 the FCC required all wireless carriers to georoute 988 calls, giving the nationwide carriers a short window after the rule took effect and smaller carriers until late 2026. In July 2025 it adopted rules for texts, with compliance deadlines running into 2027 and 2028. The problem of the Ohio student with a Texas phone number took roughly two years from launch to address for calls, and longer for texts. It required carriers and the federal regulator to act, exactly as the original analysis suggested.

More states adopted dedicated 988 fees, and mobile crisis capacity expanded, though unevenly. The Medicaid mobile crisis option gave states that adopted it an enhanced federal match for their first twelve quarters, and the option itself is scheduled to run through March 2027. By 2026, states that built mobile crisis on the enhanced match are deciding how to carry the full cost in their regular Medicaid budgets.

Federal priorities also shifted. In March 2025, HHS announced a reorganization that would fold SAMHSA into a proposed new Administration for a Healthy America. Staff reductions at the agency drew concern from members of Congress about support for 988. In July 2025 the Lifeline ended its specialized option for LGBTQ+ young people, a reminder that program features can change with federal policy. That is a practical point as well as a political one. Services that depend on federal decisions can change on federal timelines, and states that want certainty for particular groups may need to plan for it themselves.

The funding analysis held up as well. States that had built 988 on short-term federal grants found themselves exposed whenever federal priorities or budgets shifted. States with a dedicated fee or a stable appropriation had more room to plan. The 2020 law requires the FCC to report each year on how states collect and spend 988 fees. Those reports gave legislators and advocates a way to check that fee revenue was going where it was meant to go.

Dispatch remained the least visible gap. Mobile crisis coverage grew where states combined the Medicaid option with sustained funding and local partnerships, and stayed thin where they did not. Coordination between 988 centers and 911 developed locality by locality rather than through a national standard. The places that made the most progress treated the three elements of crisis care as one system with one governance group: a line to call, a team to respond and a place to go.

By 2026 the number is widely known. Of the three gaps, location is closest to closed, at least for calls. Dispatch and stable funding still separate states with a working crisis system from those where 911 remains the default response.

A well-known front door is not the same as a working house. The remaining gaps are local work rather than technical puzzles waiting on a national fix: agreements between agencies, live visibility of mobile teams, and budgets built on recurring revenue. Those are choices within reach of a state or county leadership team, and they are where the difference between a number and a system now lies.