skip to Main Content
Text to 911

Imagine that you are home alone and suddenly hear glass shattering as someone breaks in. You can hear at least two persons whispering to each other.

As you seek an exit or hiding place, your first instinct is to call 911, but you fear that if you can hear the burglars whispering, they will hear you as you make the call. Since you use “texting” or Short Message Service (SMS) to communicate with your family and friends – you try to send a text to 911. Surprisingly, a message pops up stating: “911 Error, Invalid Number,” meaning that the 911 system in your area cannot receive a text message.

Many people in the U.S. today cannot reach 911 — or more accurately the local “Public Safety Answering Point” (PSAP) — by sending a text message from their mobile phone. That is changing — but slowly and sporadically. The fact that Text-to-911 is not widely available impacts everyone, but is especially critical for three specific groups: 1) people in emergency situations who do not want to speak for fear of being discovered, (as with break-ins, kidnappings, and students desperately trying to reach help during school shootings); 2) the speech and hearing impaired who have historically used non-portable TTY/TDD devices to reach and communicate with 911; and 3) persons unable to “get through” to emergency numbers during natural disasters and other critical situations when the number of calls, or “call load” overwhelms the answering capacity of the 911 system.

Text to 911 Background

The current 911 system still uses legacy telephone technology, built in the 1950s-70s to handle calls from landline phones. However, mobile phones became popular in the 1990s, and by the 2000s, the number of mobile phones in use exceeded the number of landline phones. For 911 operators, mobile phones added three new problems: 1) increased call loads, 2) difficulty in locating the mobile phone callers to send help to them, and 3) callers unable to text to 911.

Concerning the first problem, rather than just one landline phone per family, as existed when the 911 system was developed, nearly every individual today has a mobile phone and carries it with them all the time. Thus, the 911 system receives thousands more calls than ever before. They receive numerous calls about accidents on highways, altercations in public, and calls for help when people are trapped in floods, fires, avalanches, and earthquakes. This increased call load from mobile phones has overwhelmed the legacy 911 network during every major event since Hurricane Katrina in 2005. In such cases, the extra calls typically ring with no answer or are sent to a voice message system and checked later.

Concerning the second problem, the legacy 911 system was built to locate callers at the specific address where the landline phone being used by the caller was installed. However, since most callers are now “mobile,” the system can locate callers only within an approximate area. As more phones have advanced Global Positioning Satellite (GPS) chips in them, locating callers has become easier, but is still not as advanced as ONStar and other similar systems.

Concerning the third problem, the analog 911 system has been cited as the primary technical reason that U.S. residents cannot text to 911. For years, telecommunications companies such as AT&T, T-Mobile, 4G Americas, CTIA, and Motorola, argued that — while text messaging is common in private calls – it was technically unfeasible to provide Text-to-911 service. Other telecommunications companies disagreed, but no one had completed lab tests or other measurements to explore these concerns.
Instead, efforts have been underway for decades to develop “Next Generation 911 (NG911), ” that will replace the current 911 analog technology with digital, IP-based packet-switched technology capable of supporting a full range of multimedia-to-911 communications including the transmission of text, photos, video, and data. However, full deployment of NG911 on a national scale will require complex, expensive and time-consuming changes to the 911 system. Thus, it is expected to take years and will not occur throughout the U.S. at the same time. Until then, some companies have discounted text messages to 911 as ‘fundamentally unsuitable” for emergency communications and questioned if text messages were as “reliable” as voice calls.

To investigate these concerns, in September 2011, the Federal Communications Commission (FCC) opened PS Docket No. 11-153 “In the Matter of Facilitating the Deployment of Text-to-911 and Other Next Generation 911 Applications.” A few days later, it issued a Notice of Proposed Rulemaking (2011 NPRM), requesting comments on facilitating the deployment of Text-to-911. Responses were submitted by hundreds of entities including the National Emergency Number Association (NENA), the Association of Public-Safety Communications Officials (APCO) International, various state, county and city 911 groups, some universities and many private companies.

University of Colorado ITP: Technical Analysis and Response to NRPM

Throughout 2011 and 2012, I was part of a four-person team from the University of Colorado’s (CU) Interdisciplinary Telecommunications Program (ITP) in Boulder, Colorado that studied the concerns around Text-to-911 and responded to the FCC’s 2011 NPRM. Our team included Professors Sharon K. Black and Timothy X. Brown, and two graduate students, Suzana Brown and Martin Saint. The purpose of our efforts was to provide the FCC with the solid, unbiased engineering work that had not been done before.
After months of extensive laboratory research, our CU Team’s work found that texting, or “Short Message Service” (SMS), not only works with the legacy 911 system, but is relatively affordable, and would provide an important new lifesaving option for people trying to reach 911. That was surprising to everyone and important news for the emergency communications industry. Our engineering tests were later duplicated by several other researchers who found the same results.
Our more detailed findings, as submitted to the FCC, included:

  1. Reliability. Concerning the issue of reliability, our CU Team tracked hundreds of text messages and found that all were received by the cellular network, with no data loss (0%). This resulted in a delivery “reliability level” comparable to voice, and in some situations, even more reliable than voice. Such situations include:
    a. In Weak Signal Areas. Our Team’s tests found that text messaging is actually more reliable than voice communications when a weak signal exists, “such as when the caller is in the mountains, in the midst of high-rise buildings, inside a building, under a collapsed building following an earthquake or explosion, or in a trunk of a car or closet.” This is because when a caller is at the edge of wireless coverage communications can become sporadic, allowing only momentary windows of communications signal coverage. In such cases, the parties often hear only every third word spoken, or the call is “dropped” by the communications system. Thus, these windows are not long enough to support an intelligible voice call — but are long enough to allow the short burst of data required for a text message to get through. In addition, many implementations of text messaging (SMS) automatically keep trying to send the text message, every second or so, until a transmission window opens. Voice systems do not keep retrying voice calls, but instead, the caller must redial and wait for a new call “set up.” This typically requires several seconds, if the weak signal can connect at all.
    b. During Heavy Call Loads at PSAPs. Second, Text to 911 is more reliable than voice calls when the 911 system is heavily congested or overloaded, such as when a major accident, explosion, or natural disaster (earthquake, fire, flood, tsunami, etc.) occurs. In such cases, the 911 call-takers face crushing call loads and cannot answer all of the emergency calls. In contrast, our Team’s tests found that in the time a 911 call taker could communicate with one person via a voice call, a second 911 call taker could communicate with 17- 20 persons simultaneously via text messages. The ratio of 17 text messages to each voice call was with 2012 technology, and the number of text messages to voice calls increases with every new generation of technology. Therefore, our Team found that implementing Text-to-911 systems can greatly enhance the 911 system’s ability to respond to more calls, more efficiently, during major events.
    c. Ability to Quickly Sort Calls. Third, since information can get lost in the unanswered calls and voicemails left during heavy call loads, automatic filters in the PSAPs could help sort through the information provided in text messages and pull out new details about a specific event that might be useful to First Responders. The filters could also detect the one or two calls in the string of messages that are not related to the major event, but instead are seeking help for someone having a heart attack, reporting a fire, or other emergency.
    d. When Batteries are Dying or Caller is Injured. Fourth, when the caller’s phone battery is dying, or the caller is injured and concerned about losing consciousness, the caller may not have time to make a voice call, especially if the caller receives a busy signal, dropped call, or other delay. In such cases, text messages can be a more reliable method to send a request for help. Since Text-to-911 has not been an option in the past, most callers in such situations have sent a text to family or friends and asked them to contact 911. This has complicated finding them if their text message did not include location information.
  2. Time to Compose and Send a Text Message versus Time to Call 911. During our work, our Team also asked whether it took longer to compose and send a text message than to call 911. From the hundreds of texts and calls tested, we found that a text message of 60 characters could be created and sent in approximately the same time as it takes to press the numbers 9-1-1 on a phone and wait for the receiving phone to start ringing — usually 4 or more seconds. Our CU Team also found that the size of a text message did not significantly impact that time. As the message size increased from 60 characters to 160 characters, the time to transmit the message increased by only 20% or 1 second. Concerning the time to compose a longer message, we noted that there were an increasing number of smartphone applications and other texting short cuts that enabled quickly composed messages and pre-stored information, such as contact information for an epileptic having a seizure.
  3. Connection Time. We further asked if the time to connect a text message or voice call changed if the caller were stationary, such as when hiding in a closet or under a desk, or if the caller were moving, such as when trapped in a moving vehicle. Historically, it has been well-known in the communications community that the telephone system requires slightly more time to establish a connection with a moving caller than a stationary one. Our tests found that this remained true, but we found no significant difference in the time to connect text messages versus phone calls.
  4. Continuity and Consistent Contact with 911 Call Takers. Finally, our Team acknowledged that most people appreciate the benefit of staying connected with the same 911 call taker through a crisis and being reconnected with that person if the call is interrupted. This is because valuable information for the call taker and feedback to the caller may be lost if a gap in the conversation occurs. Since text messages can be a string of separate messages, we anticipated that some callers would worry that their texts might be routed to different call takers. However, our research showed that the SMS technology, even in 2012, could route a string of text messages to the same call taker. Callers have since stated that, if texting were their only reasonable option to reach 911, continuity with a call taker would not be as important to them as being able to use all available avenues to reach help.
    We filed these results in our response to the FCC’s 2011 NPRM and they were cited by the FCC in their various subsequent orders in the docket.

Text to 911 Outcomes

Accelerate Text-to-911. A year after the 2011 NPRM, on December 12, 2012, the FCC ordered several actions to accelerate the availability of Text to 911 nationwide. Among these, the FCC required, effective January 29, 2013, that all mobile phone companies and other text message service providers implement Text to 911 service within an area six months after a PSAP area requests it.
In its Order, the FCC specifically rejected the claims that texting (SMS) is “fundamentally unsuitable” for emergency communications, or that the concerns about reliability should delay its implementation. Instead, the FCC stated that the information provided in the responses to its 2011 NPRM showed that, while Text-to-911 is not a perfect solution, it provides an important, reliable, additional way for the public to reach 911.

Bounce-back Messages. The FCC was also concerned that because the public expected that they could text to 911, they would be at risk if they were not informed when their emergency text messages could not go through to the PSAP. To mitigate this risk, on May 8, 2013, the FCC issued its “Text to 911 Bounce-back Message Order” requiring that by June 28, 2013, and later extended to September 30, 2013, all mobile phone companies and other text message service providers must send an automatic reply or “bounce-back” text message to its customers whenever that customer tried to text to 911 in an area where Text-to-911 service is not available. The purpose of the message is to inform the customer that: 1) his/her message could not be delivered to 911, 2) no error occurred in sending the text, and 3) the mobile phone company or other text messaging provider did not experience a network failure causing the non-delivery of the text to the PSAP. Not all messages include all three items.

In acknowledging these orders, it is important to note that, while the FCC can order communications companies to implement certain services by a specific date, it cannot order the PSAPs to do so. This is because, in most states, the governance, decision-making, and operation of PSAPs is decentralized and locally-ruled. Each PSAP service area in the U.S. selects its own infrastructure technology for 911 and makes its own decisions concerning implementation and operations. To pay for its equipment, communications lines, and personnel, each PSAP service area receives a share of the 911 surcharge that all phone customers pay each month in their phone bills. Each PSAP’s share is based on the population it serves. Thus, larger, urban PSAP service areas usually have more money for upgrades and the implementation of new features in their systems, such as Text-to-911, while smaller, often more rural PSAP areas do not.

This decentralization of decision-making and funding is a significant reason why implementation of Text-to-911 and NG911 takes decades and is not uniform across the U.S. Additionally, since the FCC’s orders in 2012 and 2013, many companies have asked the FCC for extensions of time and exemptions from the FCC’s rules – further extending implementation.


Text to 911 is a critical service that most of the U.S. population believes is currently available, even though it is not in many locations. To be truly effective, the 911 system must be: 1) as accessible as possible from any communications device that persons would expect to use to reach help, and 2) usable in both spoken and written format to ensure that all requests for help get through.

After discerning from the research studies submitted to it, including our University of Colorado Team’s input, that Text to 911 is technically feasible and would save lives, the FCC acted to encourage the implementation of Text to 911 nationwide.

While this implementation has been slow, impacted by the fragmented structure of the 911 system and the various requests for exemptions and extensions — it is occurring – providing an important new way for people to reach 911.

[1] For current locations with “Text to 911” see

[2] “Teletype Writer/Telecom Device for the Deaf” (TTY/TDD).  Texting from a mobile phone provides a portable means of communication for speech and hearing-impaired persons.

[3] This effort has been trackedby the National Emergency Number Association (NENA) (  and the FCC’s2010 PSDocket No. 10-255 titled: “In the Matter of Framework for Next Generation 911Deployment.

[4] In the Matter of Facilitating the Deployment of Text to 911 and Other Next Generation 911 Applications, Framework for Next Generation 911 Deployment, PS Docket No. 11-153, PS Docket No. 10-255, Notice of Proposed Rulemaking, 26 FCCRcd 13615 (2011).

[5] We could not test their delivery to the 911 center since that capability did not exist at the PSAPs.

[6] Bounce-BackOrder,

Back To Top