Juliette Kayyem.

In excerpt from new book on our age of disasters, Kennedy School lecturer Juliette Kayyem ’91, J.D. ’95, examines how we take wrong lessons from history.

Nation & World

When disaster strikes, what you don’t know might kill you

long read

Juliette Kayyem says we often take wrong lessons from history in excerpt from new book

Excerpted from “The Devil Never Sleeps: Learning to Live in an Age of Disasters” by Juliette Kayyem, ’91, JD ’95, Belfer Senior Lecturer in International Security.

Many major disasters or events have some sort of commission or blue-ribbon group to determine what went wrong and what might be learned to prepare for the future. They can be thorough and help expose facts and lessons, such as the 9/11 Commission, whose historic words — “a failure of imagination”— captured the world’s attention. In so many cases, the reasons for the disaster are easy to come by: levees broke, intelligence dots weren’t connected, a network was vulnerable, a virus wasn’t contained early enough. Fixes are urged to ensure that an identical catastrophe doesn’t happen again. Who could be against that? But a thorough review has an additional purpose in an era of disasters: It not only confirms that people have died, but it can expose how people died. There is a difference.

In the 2020 hurricane season, there were 30 named storms, more than ever before. Storms were so plentiful that the National Hurricane Center (NHC) had to turn to the Greek alphabet — alpha, beta, and so on — once it had passed Z. Twelve of those storms made land in the U.S., another new record. Hurricane Laura in Louisiana would prove to be the biggest, creating 17-foot storm surges, the highest ever recorded. The NHC launched a massive messaging campaign throughout in an effort that minimized fatalities, using the dramatic word unsurvivable to impress upon people how serious Laura could be.

There was not one fatality because of the surge or hurricane. But still, 28 people died, most of these after the storm had passed. It wasn’t the waters. It was the gas. As the storm devastated the electrical grid, many communities had to rely on emergency generators; various areas in Southwest Louisiana had no access to power for weeks. Those generators proved to be unsafe for many. The majority of the deaths were in fact people dying from carbon monoxide poisoning rather than from the storm itself.

These are stupid deaths, often called indirect deaths. As hurricane forecasting has improved, information has helped make us safer and better prepared for surging waters. In turn, fewer people die from direct causes, such as flooding and high winds, yet people are still dying. These indirect causes include heart attacks, car accidents, electrocution, and carbon monoxide.

We’ve similarly learned about blizzards in the last few decades. It turns out most people do not die from the snow or cold. They mostly die from carbon monoxide poisoning as well, more often than not in their cars. In the 1978 blizzard in New England, nearly 100 people died during a surprise storm, one that came in so fast it was almost impossible to prepare for. Once the snow started falling, people got into their cars to rush home or check in on family members. Soon many got stuck. Without help on the horizon, while freezing, people kept their car engines on as the exhaust pipe froze as well. Carbon monoxide would ultimately kill 72 of them.

This is why today, governors throughout New England and in colder climates regularly institute travel bans well before the snow starts. If they wait too long, the disaster will kill in ways having nothing to do directly with snow. A goal is certainly to protect first responders and keep streets open to plow. But mostly it is to keep people from dying of carbon monoxide poisoning. It is the immediate aftermath period that can prove deadliest, when people turn to make-shift processes — generators barely used, fireplaces not cleaned out — and die. There is an irony here; as our systems of response become more and more sophisticated for recurring disasters, and people take heed of familiar threats, the disaster can still be challenging and lethal. How people die matters.

“The Devil Never Sleeps” proposes a guide to dealing with disasters.

The Devil Never Sleeps book cover.

Historical patterns, although helpful, cannot always serve as guides for what we might anticipate in the future. The threats are changing too quickly and occurring too rapidly. But that is not to say there is no role for history to promote better response and consequence management purposes. One obvious reason is because indirect deaths can always be avoided. Another reason, though, is that we often learn the wrong lessons from these disasters. We make the wrong assumptions, like water or wind being the cause of hurricane deaths. These initial assumptions about what occurred, and therefore how to fix it, will change over time. We must accurately memorialize how people die.

This happened with later studies of the mass shooting at Colorado’s Columbine High School on April 20, 1999. We believe a story about the two student murderers, Eric Harris and Dylan Klebold, that has not held up to examination: They were not misfits and goth advocates who lived in a dark world. All of that was a myth. They were well-adapted boys, beloved, who did something horrible. And as the events of that day were reviewed, it also became clear that the protocols for how to deal with active shooters had to change.

As the two student killers walked the hallways shooting, students were told to hide in the school’s library. The problem was that nothing protected them there. One after another, the killers targeted the captive students. After the massacre, 12 students and one teacher had lost their lives; the shooters committed suicide. In subsequent years, as more was learned about that day, it was clear that the students died because they were unable to escape from the library. And so those who began to help schools deal with the truly horrific American phenomenon of school shootings began to promote the concept of “run, hide, fight.” Run first if you can. Get out of harm’s way. Who died and didn’t die in that high school wasn’t merely a matter of luck; it was a question of location. An important lesson from the tragedy of Columbine is that we taught our children to run.

As a mother I found these school shootings devastating.  With decades of mass shootings, we’ve now learned that there is no benefit for first responders to delay entry into a facility. Previously, they had assumed that a shooter had some agenda and that by not entering, police could convince them to stop their violence. After Columbine, police were trained in a new tactic: immediate action rapid deployment. Speed, in other words, could have saved those children. It is worth noting that years later, conventional wisdom has begun to change again. The new understanding is that students could know what to do if there was an active shooter if it was explained to them but that formal active shooter drills are less beneficial than once thought. The trauma to students, especially younger ones, outweighs any benefit they may gain.

In design and planning, the same is true. Bridges falling are headlines. It is a tragedy. But we must return to the site to determine how, in fact, it fell. On Nov. 7, 1940, the Tacoma Narrows Bridge, the third-largest suspension bridge in the world, collapsed. The bridge connected Tacoma to the Kitsap Peninsula in Puget Sound and had opened just a few months earlier. It was a spectacular bridge failure, a technological wonder that didn’t last a year.

What brought the bridge down was wind. It was not just any wind, though, or the wind that examiners originally believed brought the bridge down. For some time, engineers believed the collapse was due to something called resonant frequency. Resonant frequency describes how much an object can absorb vibrational energy. Too much resonant frequency, too much pressure on the system unable to absorb it, and catastrophe follows. It was assumed that the wind moved the bridge naturally at first, but then pushed the frequency too hard, too strong, for too long, and it couldn’t sustain the pressure.

“There is an irony here; as our systems of response become more and more sophisticated for recurring disasters, and people take heed of familiar threats, the disaster can still be challenging and lethal. How people die matters.”

That simple assumption proved incorrect. Decades later, science would later change the narrative. When an object is suspended between two points, it is built to move to absorb impacts such as wind. The capacity to vibrate is built in, and we know how to build bridges to do so. That November 1940 day, the wind was so strong and continual that it caused something new, a flutter. The flutter served as an extra push at the ends of the suspended object, causing them to move perpendicular to the wind (rather than with the wind). Airplane manufacturers have learned to account for flutter in the design of a plane’s wings. But no engineer thought it could happen on a bridge. With the unique intense wind, the flutter was uncontrolled, twisting back and forth, breaking a steel suspension cable. The bridge could just not hold.

Fixing resonant frequency is a very different effort than addressing flutter in a suspension bridge. The latter requires buttressing end posts. Without such knowledge, bridges would continue to be built without a focus on flutter. Modern science led to a new engineering subfield called bridge aerodynamics and aeroelasticity. It pushed engineers to monitor new bridges as well that might be prone to flutter-like damage, including London’s Millennium Bridge and Russia’s Volgograd Bridge. Both of these major bridges had delayed openings and abrupt closings due to concerns over flutter.

Any review of what went wrong or how we can do better has to begin with the fundamentals, not the results. Take Facebook, for example, if we must. Mark Zuckerberg created a product, not just a platform. It connected people, he told us. He let us share memories and pictures, reacquaint ourselves, and meet strangers. Life would be better because we would be together. Facebook was sold as a benign company with a leader who seemed young enough to avoid judgment. But then reality hit the company: It had to monetize all the fun. So it turned to an advertising-based model, where we — Facebook’s users — actually became the product. Our information and our desires were targeted by the company for sale; advertisers would use that data to focus their efforts.

Zuckerberg was the perfect salesperson for the pitch. And he told regulators and legislators, privacy advocates, and those who would want to protect democracy not to worry about his growing power to control what we knew. By 2016, most Americans were absorbing their news through Facebook; it was no longer a platform but a publisher. The “news” became a sold commodity, targeted to those who would want to read it. Whether it was true or not was not Facebook’s worry.

As complaints grew about all the disinformation, Zuckerberg defended himself by what seemed a completely rational explanation. We wouldn’t want him to have the power to decide what is true, he would argue to congressional investigators and reporters. For Facebook to be the adjudicator of truth, the CEO claimed, was worse than letting information flow, even if some of it was false. The explanation sounded pretty solid.

Over time, it was clear that the argument was a total self-serving manipulation. He was playing with our assumptions about information. By claiming he didn’t want such authority to decide the truth, he was hiding the fact that he had already asserted considerable authority. His decision not to decide was a value-laden decision itself. It was favoring treachery; it let the misinformation and disinformation flourish. Zuckerberg was claiming he was agnostic. He was instead running the devil’s errands.

Facebook would spend a large chunk of its efforts post-2016 defending its business model, one that was flawed by design. It would promise to get better; piecemeal efforts, including the creation of a “Supreme Court” to independently assess questions of truth and usage, were created. It has not changed, though it apologizes a lot. It will not learn because it refuses to look at the primary, fundamental, even existential decision it had made — deciding not to decide. It will repeat history because it has no interest in learning from it.

Copyright © 2022. Available from PublicAffairs, an imprint of Hachette Book Group, Inc.