“Presidential Alert” test: what emergency communicators need to know about effective messaging

The "presidential alert" system is the USA's highest emergency alert system.
Example of how a presidential alert would appear on mobile phones nationwide. A test is set for October 3, 2018. (FEMA photo.)

The upcoming nationwide test (now rescheduled for Oct. 3) of the nation’s highest level of emergency alert, the federal “Presidential alert” system, raises important questions about how government agencies communicate urgent safety information to the public when lives are on the line.

Being the source of credible emergency information is one of the most important responsibilities of any elected official, and especially for those who head executive positions such as mayor, town manager, governor or president. Knowing what makes an emergency message effective should be mandatory learning for all elected officials.

We have many examples from recent emergencies how these factors play out. Based on real life examples, there are three key factors that determine whether emergency messages are effective:

1. The messages must actually reach the intended audience.

Facing a fast-moving firestorm last year, Sonoma County, Calif., officials relied in part on automated calls to landline telephones, unaware that most phones were inoperative because winds or fire knocked down utility lines. A more common error is sending only English-language messages to neighborhoods with many non-English speaking residents. Emergency plans must account for translation and multi-lingual communication (not Google Translate!,) as well as the complexities of communicating with concentrations of elderly or those with mobility limitations, blindness, hearing impairments or other special circumstances. There are few emergencies in which a single mode of communication is sufficient to reach everyone in danger. This is also why the only method that consistently has better than 50 percent success is door-to-door, person-to-person contact by emergency personnel.

2. People who receive the message must believe that it is real.

The credibility of a message’s sender is paramount. Early efforts at creating text messaging alert systems failed when organizations conducted too many tests or used the alert system to convey non-urgent information. These systems got messages out, but people tuned them out. If every message makes your cell phone light up, people tend to either shut off or ignore messages. An emergency alert system should be used only for true emergencies. At least one major sheriff’s department has used their Nixle text messaging system to announce non-emergency public events — hardly an emergency. That kind of message makes people more likely to ignore future alerts. (Besides, there are ways to send these “community messages” via Nixle without using the emergency alert function.)

3. Messages must include clear instructions.

Informing people that there is an emergency is only the first step. People need to know what to do. Instead of saying “take shelter,” say “stay in your home with doors and windows closed and locked.” Instead of saying “avoid the area,” say “stay away.” Instead of “water may contain toxins” say “do not drink the water.” Use clear, direct language so that people can act swiftly on the directions provided. If there is nothing for people to do, at least let them know how to get updates or when the next update will be sent.

Hawaii false missile alert
Hawaii’s emergency management agency sent this false alert on January 13, 2018. A correction went out 38 minutes later.

In January, when the Hawaii emergency management office mistakenly transmitted a missile attack alert, people who paid attention to prior communications about emergency alerts disregarded the message because they knew that a real message would have been accompanied by other messages with additional details. However, thousands of others were terrified for the 38 minutes between the errant alert and the follow-up message retracting it. Some of the panic was because the alert provided no actionable instructions. People had no idea where they were supposed to take shelter.

As we approach the test of the presidential alert system, we must consider whether such an alert could ever be effective. In a real emergency, the president is the only one with authority to activate this system. However, activation means directing designated officials to issue one or more emergency messages via both traditional radio and television channels and the nationwide cellular networks. It does not necessarily mean that the president personally crafts or transmits the message.

The emergency broadcast system, which forms an important part of the current alert structure, was created during the Cold War fear of a Russian nuclear attack. It was highly effective in large part because there was a concentration of media. Instead of hundreds of different television channels, most communities had five or six channels, and a majority of people watched one of the three major network evening newscasts. While the laws now require cable TV systems to join broadcasters and transmit a national level emergency alert, the messages would only reach people watching live TV programming. Streaming via the Internet or watching shows on your your DVR won’t be interrupted, and fewer people watch live TV every year.

This is why the current system adds Wireless Emergency Alerts — the kind that trigger cell phone alarms for an Amber Alert or impending major weather emergency. Based on geography instead of individual cell phone numbers, the system is highly effective at delivering emergency messages to individual cell phones.

Even though the systems have been used with great restraint, there is major risk that people will shut off emergency alerts if one or more ill-advised messages go out. If a local government official starts off a news conference in an emergency touting how well he is doing as the emergency manager, there’s little chance the public will listen to or care about anything that comes later — even if it has life-safety implications. If a public information officer from one agency puts out a message that turns out to be completely wrong, that PIO will have trouble getting other emergency messages out. Credibility and trust are keys to effectiveness.

That brings us back to the current situation. While elected officials are political by definition, disaster or other emergency response must always be separated from politics. Natural disasters do not take paths based on the political persuasions of people along the way. Neither does this impact who gets rescued (at least, it certainly should not.)

One reason that many elected officials defer to their law enforcement, fire or public health leaders to make emergency announcements is to avoid any perception of politics or other bias. However, those people are still acting on authority granted by the elected leader.

Should there be a real national emergency, the president must personally authorize use of the presidential alert system, but others must be involved, both for practical reasons and because the law includes a chain of authorities for the system’s use. Effectively communicating what to do when a dire threat looms may best be delegated to a military, national security, public health or public safety official with the right expertise and credibility. If the goal is saving lives, knowing who people trust has to be considered before any emergency message goes out.

For more about crafting and disseminating emergency messages, check out Doug Levy’s book, The Communications Golden Hour: The Essential Guide to Public Information When Every Minute Counts (Public Safety Press, 2018.)