The views expressed by contributors are their own and not the view of The Hill

Hawaii illustrates that America’s alert system is fragile, fragmented and outdated

by John Lawson, opinion contributor - 01/19/18 9:30 AM ET

The false alert of a nuclear attack on Hawaii is an inexcusable, but hardly isolated example of the fragility and fragmentation of America’s emergency alerting system. False alarms are dangerous because they lead to complacency when real danger threatens. Even worse is when alerts fail to reach endangered people in real disasters — or when alerts are never sent in the first place. As natural and manmade threats persist and increase, it is clear that our alerting system is not up to the task of serving the mobile and connected America of the 21st Century.

Alerting failures occur with depressing frequency. During the fatal Northern California wildfires last fall, Napa and Sonoma Counties never issued an evacuation notice over the Wireless Emergency Alert system. Houston area officials complained to the Federal Communications Commission about their inability to geographically target flooding alerts after Hurricane Harvey. In the fatal Gatlinburg, Tennessee wildfires in 2016, state and county officials never issued a wireless alert, until late in the disaster when they asked residents to stay off their cellphones.

New York authorities commendably used a wireless alert to send a wanted poster for the New York-New Jersey bomber in 2016, but because the current system only allows 90 characters and no pictures, the alert was criticized as casting suspicion on tens of thousands of innocent young Muslim men.

{mosads}Denial of service — intentional or unintentional — is a recurring problem. Last year’s hurricanes produced widespread cellular outages in Florida, Puerto Rico, and the Virgin Islands, as have wildfires in other regions. Even intact wireless networks become overloaded in public emergencies and crash. In November, the Federal Emergency Management Agency’s national alerting backbone network that feeds local alerts reportedly was off-line for nearly 11 hours. Hackers have issued “zombie attack” alerts and other hoaxes on the legacy Emergency Alert System that feeds television and radio stations

One challenge is that no single entity “owns” the emergency alerting system and no dedicated funding source exists to sustain and improve it. Except for Cold War-era alerts from the president in case of nuclear attack, alerting is a local affair. And even nuclear alerts can be mishandled, as we saw in Hawaii. Alerting is controlled by a hodgepodge of national, state, tribal, and local authorities who depend upon private sector infrastructure and software for distribution. Training and technical capacity vary widely. Funding is highly inconsistent, and federal authority to cut across jurisdictional lines is basically non-existent.

Our decentralized and largely voluntary alerting structure, however, has some advantages and can be made to work. In a democracy, limiting government access to TV and radio stations and wireless networks, even in emergencies, is healthy. Single points-of-failure can be reduced. Costs can be lowered through the “dual use” of private communications infrastructure vs. dedicated government networks, while also leveraging the technological innovation of the free market.

Some progress is being made. In 2006, President Bush ordered development of the Integrated Public Alert and Warning System, although still only about a quarter of local authorities use the system. The FCC has been moving to improve wireless alerting over the past two years and is considering new rules late this month to improve geo-targeting. Congress established an alerting advisory subcommittee, on which I serve, to recommend new policies to FEMA. In addition, a coalition of television broadcasters and tech companies in the U.S. and South Korea, which I lead, is using Next Generation Television broadcasting to create a voluntary advanced alerting system.

However, the federal government, under its constitutional role, must step up its leadership and investment to bring all of the elements of our decentralized alerting structure together into an effective whole. First, a modest but dedicated federal funding source is necessary for training and equipment. When a state employee can “push the wrong button” and trigger a statewide panic, or when a county agency never sends an evacuation alert in a deadly wildfire, the problem is both human and technical.

Second, we need to invest much more heavily in social science research. A recent report from the National Academies of Science, Engineering, and Medicine identified major gaps in our knowledge for crafting effective alert messages. Even the best system will fail if people “mill” about for confirmation or more information and delay taking protective action, a major problem with current alerting approaches.

Finally, the federal government at both the White House and congressional levels should convene a nationwide conversation with key stakeholders in alerting at all levels, public and private. The goals would be: define requirements for a more resilient and advanced alerting system, identify resources for building out the system and staffing it properly, and setting a timeline for action.