I haven’t wanted to write anything about Hurricane Katrina. Watching catastrophe unfold in Louisiana and Mississippi has been horrific, and what risk communication “lessons” I’ve been able to extract have seemed too secondary to bother with, compared with the raw, ongoing reality of destruction, death, and misery. Everyone, myself included, is still too emotionally wrought to want to think analytically.

Also, there are still so many unanswered questions. Was the official response really slower than it ought to have been, or did it just feel that way as we watched people call out from their rooftops and swelter in the Superdome? Was there really more violence and lawlessness than natural disasters usually provoke, or were a few isolated incidents and unconfirmed rumors magnified by the media? We’ll know more in a month or two about these and other issues. We’re guessing now, imagining that vivid television images constitute firm knowledge.

Communication lessons are going to be only a small part of the emergency preparedness and emergency response lessons of Hurricane Katrina. And the most important communication lessons are going to be about internal communications among emergency responders – everything from overlapping chains-of-command to nonfunctioning telephones. Risk communication lessons are not the heart of the matter.

But friends and colleagues keep asking what I think. So here are some early, tentative reactions.

1. We need to learn how to scare people … and we need to dare to scare people.

The most obvious risk communication lesson of Katrina concerns the tens of thousands of people in and around New Orleans who didn’t flee for high ground. I fervently agree that the authorities should have made it easier to leave. They should have used every bus they had; they should have required half-empty cars to pick up pedestrians; they should have assured people there would be food and shelter available when they got beyond the “bowl.” But many with cars, gas, and cash didn’t leave either – including resource-rich individuals like blues singer Fats Domino. And many able-bodied people who lived close enough to walk to high ground (or at least to a shelter) chose not to do so.

One big failure, much discussed, was not enabling people to leave – especially the sick and frail, and poor people without cars. Another big failure, rarely discussed, was not persuading people – sick and well, rich and poor alike – to want to leave. All too many heard about the “mandatory evacuation” order, rightly reckoned that it wasn’t all that mandatory, balanced the hardships of leaving against the risks of staying, and unwisely decided to stay.

If there had been a more successful risk communication effort to persuade people capable of evacuating to do so, the remaining task of helping those incapable of evacuating on their own would have been much more manageable.

I am waiting for the almost inevitable news stories about internal debates within various government agencies over how aggressively to insist on evacuation. Somebody has got to have written an email arguing that “we might panic people” if we put too much stress on the importance of getting out of town. Somebody has got to have warned that “the levees will probably hold, and then we will look like we overreacted.” Somebody has got to have pointed out that an “excessive” evacuation program could lead to charges of wasting desperately needed tax dollars, of assuming authoritarian control over people’s autonomy, of forcing citizens to leave their property unprotected from looters, of damaging community mental health by unduly frightening the fragile. Likeliest of all, some official must have said, “The more forceful and urgent we make that mandatory evacuation order, the more vulnerable we will be to accusations that we didn’t plan properly for ways to implement it – not to mention accusations of ignoring flood control priorities.”

Don’t get me wrong. Evacuation has real costs, and I respect the difficulty of deciding whether any particular hurricane merits such a disruptive precaution. New Orleans faced other hurricanes between Betsy in 1965 and Katrina in 2005. They weren’t catastrophic. New Orleans even announced other hurricane evacuations during that 40-year period, most notably for Hurricane Georges in 1998. Like this one, previous evacuations weren’t entirely successful; unlike this one, they turned out not to be essential. The calculation of risk-risk and cost-risk tradeoffs that goes into deciding whether to order an evacuation is necessarily a crapshoot. You can’t blame officials for guessing wrong sometimes – in either direction.

But I think you can blame them for guessing right and then implementing their guess half-heartedly. I can’t believe that they meant to leave more than a fifth of the city’s population behind. I am certain that people died because the authorities hesitated to scare them.

I routinely use hurricane forecasting as a risk communication good example. The Weather Channel and the National Hurricane Center, I point out, aren’t afraid to emphasize worst case scenarios or to “speculate ” responsibly on where the hurricane might be headed; they don’t fall into traditional risk communication traps like over-reassurance and over-confidence; they personalize the risk and rarely come off sounding bureaucratic. They’re pretty good at scaring people – at least at scaring the people who watch the Weather Channel. We need to work on ways to scare – and willingness to scare – the people who don’t watch the Weather Channel.

Two other failures-to-scare deserve mention.

First, consider the bloodless policy debates of past decades. Long before Katrina, experts were warning that New Orleans was vulnerable to devastating floods if a monster hurricane breached the levees. More broadly, experts have long argued that over-development of the Gulf Coast was exacerbating the threat. These weren’t secret debates. They were in the papers – but sporadically, and always calmly. What might have been accomplished if officials (or even a journalist or an activist) had made a concerted effort to arouse public fears about hurricane vulnerability? Would the levees have been strengthened? Would an evacuation plan for people without cars have been fashioned? Would individuals living at or below sea level have made their own preparations? Much will be done along these lines in the years ahead, as Americans vow never again. All of it could have been done beforehand, if we had been frightened enough. (I don’t mean permanently frightened – just frightened long enough to decide what to do and get it done.)

Scaring people about an abstract, hypothetical emergency isn’t easy. In a 2003 column on “Duct Tape Risk Communication,” Jody Lanard and I wrote: “When the Federal Emergency Management Agency urges precautions against natural disasters … it is fighting an uphill battle against apathy. No one accuses FEMA of ulterior motives; no one accuses FEMA of false reassurance or of panicking the public. No one accuses FEMA of anything in advance of a disaster; no one is even paying attention until after the disaster, when everyone accuses FEMA of not doing enough.”

Finally, as I write this there are still thousands of people in New Orleans declining to be rescued. They are weighing the risks if they leave their homes (looting; coping with shelter life; not being allowed back; losing track of their pets) against the risks if they stay (disease; fire; dehydration; starvation) – and they are choosing to stay. Day by day the risk of leaving declines and the risk of staying increases. But so far, law enforcement officials are trying to persuade people to leave rather than requiring them to do so. (The evacuation still isn’t really mandatory.) Judging from the TV reports, they are telling the holdouts that there’s nothing to stay for – no jobs, no food, no water, no power. They don’t seem to be acknowledging and addressing the fears that are keeping people in their homes. And once again, they don’t seem to be saying appropriately scary things about the danger of staying – most importantly about the litany of diseases, from cholera to West Nile Virus, that can accompany fetid, contaminated, standing water.

2. We’re NOT prepared for an emergency … and we need to keep saying so.

Explanations for the deficiencies in our responses to Katrina (and our preparedness to respond) fall into two categories. The left argues that it’s about malfeasance:

  • We ignored warnings about levee vulnerability in a monster hurricane.
  • We sent too much of the National Guard and its equipment to Iraq.
  • We eviscerated FEMA’s natural disaster response capability in order to build up terrorism response capability instead.
  • We let developers destroy the natural barriers that would have ameliorated the hurricane’s effect.
  • We let global warming spawn ever more violent storms.
  • We let clashing bureaucratic priorities and clashing egos take precedence over quick and efficient emergency response.
  • We planned an evacuation as if everybody had a car and cash for a hotel room.
  • We had the hubris to build below sea level in the first place.

The right leans more toward a tragic view, leavened with optimism: All the good intentions, big budgets, and master plans in the world can’t keep Mother Nature from her periodic bouts of devastation, but the American Spirit will rise to the occasion and rebuild New Orleans even better than it was before.

(The far left and the far right have somewhat different explanations, seeing New Orleans as a stunning example of what happens when racism and classism run rampant, or alternatively of what happens when the welfare state has sapped the entrepreneurial spirit.)

But we ought to be able to agree on one thing: If this is the best we can do when a hurricane hits the Gulf Coast, we’re obviously not in good shape to cope with a terrorist attack on Los Angeles or Chicago – or Los Angeles and Chicago at the same time. And think about our preparedness for an influenza pandemic that engulfs the whole world more or less at once, slowing travel to a trickle. Help has poured into Louisiana and Mississippi from everywhere. In a pandemic not much help would pour in; communities would be pretty much on their own.

It is conceivable that the failures of the past week are unique to Louisiana and Mississippi or to hurricanes, that a different sort of emergency in a different part of the country would have yielded a much better outcome. But no one really seems to be claiming that, and I don’t see any evidence of it.

That leaves only two alternatives. Either emergency preparedness needs a lot more work (and thought, and budget) or we all need to cultivate a spirit of resignation, recognizing that we are more or less naked before whatever the gods, the terrorists, and the viruses may throw at us. Or, most sensibly, both. We should be doing more to get ourselves ready for emergencies of all sorts … while acknowledging that we can never be really ready. Any emergency response professional who has been tempted to claim we are really ready should be rethinking in the wake of Katrina.

The main preparedness lesson of Hurricane Katrina is about deciding how much more we can do and gearing up to do it. Just as crucial is deciding who should do what. Already brewing are debates over local versus state versus national prerogatives, and civilian versus military roles. (The phrase “post-Katrina” is fated to become as common as “post-9/11” has become.) There has been less attention so far to individual preparedness – the need to upgrade our sense of what belongs in a family emergency kit (how many days’ supply of bottled water do you want now?) and how imperative it is to have such a kit. But that may be coming too.

The associated risk communication lesson is about leveling with the public about preparedness. If people expected miracles from the authorities in the wake of Katrina, it is at least in part because they have been led to expect miracles. The argument between those who say we must do more and those who say we can never do enough is a healthy argument; they’re both right. The public needs to hear the argument, and to participate in the argument. We can’t be allowed to keep imagining that we’ve done enough already.

The standard official line on any preparedness issue is the same: “We are ready. We will protect you.” This line encourages complacency beforehand and recriminations afterwards – when officials belatedly object that it wasn’t feasible to be that ready, that it wasn’t possible to protect you that completely. The things some experts are saying post-Katrina about people needing to understand what’s doable and what isn’t sound defensive and grossly lacking in compassion. They are exactly what should have been said pre-Katrina.

In risk communication, “safe” is not a dichotomy. The proper question is always a values question: “How safe is safe enough?” Disaster readiness is not dichotomous either. The proper question is also a values question: “How ready is ready enough?” In a democratic society, such questions should be answered by the people, not just the government. Coming up with an intelligible answer requires recognizing that you can always be better prepared for anything, if you’re willing to pay the price. It requires recognizing that you can never be completely prepared for everything, no matter how much you pay. And it requires recognizing that you will sometimes guess wrong and end up preparing for the disaster that didn’t happen instead of the one that did.

In this context, one of the most wrong-headed Katrina reactions I’ve heard so far came from Senator Joe Lieberman. Talking about plans for an investigation into what went wrong, he said such an investigation was needed “to rebuild the confidence of the American people … in our government’s ability to protect them from attack, whether it comes from nature or from terrorists.” I do not see why Lieberman believes that sort of blanket confidence is either justified or useful.

3. It is hard to scare people when you’re not ready. It is hard to get ready when you haven’t scared people.

How willing we are to scare people and how prepared we are to cope are interconnected questions.

A government that knows it isn’t ready to respond well to an emergency is very unlikely to issue dramatic warnings about that emergency. This has been apparent in U.S. policy on pandemic influenza. The U.S. missed its chance to stockpile a decent supply of Tamiflu early; it is still slow-walking its efforts to expand and modernize its vaccine manufacturing capability; it hasn’t yet published its updated pandemic plan. Having done too little to prepare, how urgently can federal agencies afford to push local governments, companies, civic groups, and citizens to do more?

In the run-up to Hurricane Katrina, local and national authorities must have realized that they were ill-prepared to evacuate the lowlands, rescue those who were left behind, or repair the levees quickly if they broke. This must have been on their minds as they contemplated how scary to make their hurricane warnings. Logically, their lack of preparedness should have motivated more strenuous efforts to persuade (or force) people to leave: “We didn’t prepare as much as we should have, so you really, really need to get out!” But psychologically, the authorities faced an internal conflict. If you have under-prepared, and thus need to “over-warn” to make up for it when disaster impends, you’re doomed whatever happens – guilty of under-preparing if things get bad, guilty of over-warning if they don’t. Having under-prepared, you tend to feel like you’re better off under-warning too, and praying that the levees hold. It’s hard to make yourself say scary things when you haven’t prepared properly.

And it’s hard to prepare properly unless you build support by saying scary things. Emergency preparedness budgets are always controversial. Like insurance, the preparedness budget is money you hope you waste; that is, you hope the emergency you’re preparing for never comes. And there’s no limit to how much you could spend getting ever more ready for any number of possible catastrophes. Once again, logic and psychology conflict. Logic says you calculate (as best you can) the probability and magnitude of various bad outcomes, you calculate the cost of various precautions and how much good you think they’ll do if bad things happen, you pump the numbers, and you end up with a list of preparedness expenditures that look cost-effective. But one of the best-established findings in the psychology of risk is that people are risk-averse about gains but gamblers about losses. We’ll take a $100 prize in preference to a 1-in-100 shot at a $10,000 prize – but we’ll accept a one-in-100 chance of a $10,000 penalty rather than pay $100 for sure. So preparedness expenditures that are logically sound are nonetheless unpopular.

Making preparedness popular pretty much requires scaring people. This is what activists do best. When Greenpeace wants stronger precautions against dioxin coming out of smokestacks or genetically modified ingredients in foods, it paints scary pictures, builds a constituency, then mobilizes that constituency to put pressure on governments to act. The process isn’t any different when Christian fundamentalists want action against pornography or gay marriage. Activists for hurricane preparedness, terrorism preparedness, and pandemic preparedness – to the extent they exist – compete with environmentalists and fundamentalists for the public’s limited supply of fearfulness.

There may be alternative strategies worth considering. Many insurance salespeople, for example, sell peace of mind as much as preparedness. But the key to building a constituency for preparedness is almost always painting scary pictures of what might happen if we’re not prepared. Authorities who are unwilling to paint scary pictures are not likely to get increases in their preparedness budgets.

There is also a third relationship between fear and preparedness. When people first become aware of a new risk, or first start taking an old risk seriously, they usually go through an adjustment reaction phase – a brief period when they are seriously alarmed, perhaps more alarmed than the situation justifies. Their imagination is engaged; they may see far-off risks as here already or unlikely risks as likely; they often take precautions that are excessive or at least premature. They are galvanized into action. Then, unless disaster strikes almost immediately, their fear abates, turning to concern. They settle into the New Normal. Their precautions become more routine, more sustainable, less extreme.

The fearful adjustment reaction period is when you get a levee improvement bill passed, or an emergency evacuation launched.

4. The authorities need to focus more on their preparedness inadequacies.

I spent part of the Labor Day weekend on the telephone with one of the dozens of federal agencies that are struggling 24–7 to get their arms around this horrible event. Like most other agencies, this one lost its New Orleans office; many of its New Orleans employees lost their homes. The people I was talking with worked pretty much around-the-clock for days, right through the holiday.

Like everyone else, they must be wishing that they had been better prepared. They may even feel that they should have been better prepared. If they’re typical of emergency response professionals, in fact, they’ve probably spent years insisting that they ought to be better prepared – and arguing as persuasively as they could for the budget with which to become better prepared. But now they and their colleagues are under attack. And so they are making one of the most common mistakes in the book: aggressively “assuring” the public that they are doing a good job.

Elsewhere I have called this the “preparedness seesaw.” It’s much harder to be self-critical when you’re being criticized. But self-critical is the right stance right now. “There have been some awful glitches in getting the right supplies to the people who needed them most.” “If only we had acted more quickly on the warnings about levee vulnerability.” “Our confidence about national emergency preparedness has taken a body blow. After the worst is over, we will need to take a hard look at ourselves.”

You can even slip in a little self-congratulation in a subordinate clause: “Even though we are doing everything we can think of, we can’t help feeling that we’re just not doing enough.” Or: “Although I think the response effort on the ground today is really impressive, there were a few days when the need was urgent and the response was inadequate. I don’t know how we can do better next time, or even if we can do better next time, but it’s obvious we will have to look for ways to do better.”

Instead, the conflict of perspectives has been stark. Again and again, for days, citizens complained bitterly about being left to die. Then some official came on screen with a paean to the initiative, self-sacrifice, and wonderful spirit of cooperation displayed by all the responders. Then a reporter asked the official what about those complaints. And the official repeated the paean. The worst and most common official response to the complaints was to ignore them, to speak as if there were no complaints, to stay doggedly “on message” even when reporters asked about the complaints. Even those officials who managed to concede that the emergency response effort had been “imperfect” made the word sound more like a defense than an apology. Emergency response efforts are always imperfect, they invariably added, before “bridging” to yet another paean to emergency responder heroism.

The alternative, nearly as bad, was to tick off the reasons why instantaneous rescue wasn’t possible: Local responders were themselves victimized by the disaster; nonlocal responders had had to move their equipment safely away from the storm, then work their way back; help needed to be requested before it could be provided; it took time to transition from hurricane response to flood response; a helicopter can’t drop food and water on the Superdome while it is simultaneously rescuing a family from its attic; etc. These are all valid points. But people who endured the trauma of those first few days, even those who endured the trauma vicariously on television, needed officials to share their visceral feeling that the response just wasn’t good enough. They didn’t need to hear a rebuttal.

There were slightly better responses. Some officials said it was too soon to think about what could have been done better; there were still lives to be saved. Some officials got angry – a much more human and therefore much more credible response. (Lt. Gen. Russel Honoré did this especially well.) But I didn’t see any official say what I’ll bet nearly every official was feeling: “There’s got to have been a way to get people more help sooner. I don’t know what it is. There were some really serious screw-ups, but we all gave it everything we had. Obviously a lot of people feel that wasn’t good enough. Looking at what happened to people, how could anyone not feel that way? I feel that way. We did our damnedest, we broke records, and it wasn’t enough. It wasn’t nearly enough.”

I have heard speculation that the excessive, defensive self-congratulation coming out of the Federal Emergency Management Agency, the Department of Health and Human Services, and the other major federal responders was orchestrated by the White House. This wouldn’t surprise me, though I have no evidence. Certainly there’s nothing wrong with the White House exerting influence in a major national crisis. What’s wrong is the excessive, defensive self-congratulation itself, wherever it’s coming from.

5. The rest of us need to dial back the recriminations.

One of the truisms of crisis communication is that the recriminations are mostly suppressed until the worst of the crisis is over. But the recriminations this time around were fierce long before the crisis was anywhere near over. This bodes very badly indeed for the likely level of anger in the coming months.

Part of the reason is the excessive, defensive self-congratulation of the federal authorities. That’s how the risk communication seesaw works. Public criticism provokes government defensiveness. Government defensiveness provokes more public criticism.

And of course the survivors of Katrina have much to be angry about. Many of them had much to be angry about even before Katrina. A disproportionate percentage of those left behind in the evacuation were poor, elderly, or sick. (I don’t know if a disproportionate percentage were African-American. Most of New Orleans was African-American.) Katrina gave them additional, powerful reasons for their anger.

Beyond that, I think we are all projecting our feelings of helplessness. We live now with the searing image of people waiting day after day in their attics for rescue, in intense heat, without food or water – and of all too many who perished before the rescuers get to them. That’s hard to bear. On television, person after person segued almost seamlessly from misery to anger; some cried and yelled at the same time. Even for those who didn’t cry, you could often hear the cry in their voices.

We need to blame someone, and it is more soul-satisfying to blame the feds than the gods or ourselves. You can sense this in the angry public comments of Louisiana Governor Kathleen Blanco and many other local and state officials. I can sense it in my own desire to mouth off at the television.

Recriminations may well turn out to be justified. Nearly everyone (except some emergency response professionals) seems convinced that the response was badly flawed; even President Bush has implied as much. I think there was a decent case to be made, beforehand, against the flood control expenditures that might have saved the day. I think there’s a decent case to be made that evacuations always leave people behind. But it’s awfully hard to make a case for not ensuring that the Superdome and the convention center were adequately stockpiled, adequately replenished, and adequately staffed. In fact, the more convinced you are that not all disasters can be prevented, the more convinced you should be that pre-disaster preparedness and post-disaster response are worth spending money on.

Katrina should provoke a thorough reconsideration of our ability to cope with natural disasters and other emergencies. But some of the heated rants and snide carping I’ve seen on TV and in the papers are part of the problem, not part of the solution.

6. Allegations of lawlessness need to be dialed back too.

I don’t have a good sense of how frequent and how dangerous the looting and gunfire actually were in the days after Hurricane Katrina struck.

But every news story has its own storylines, the “myths” that enable reporters to convert atomistic facts into a compelling narrative. Heroism became a big piece of the 9/11 storyline. By contrast, lawlessness was a big piece of the Katrina storyline. I’d bet my mortgage it was a lot smaller piece of the Katrina reality. I talked with one national television journalist who was working on a story about Katrina violence. Was he really sure there had been a lot of violence, I asked. “Well, no,” he conceded. “Somebody hears someone screaming in the distance and decides it was a woman being raped. It could have been a woman with a bad toothache. How would we know?”

At least temporarily, the focus on lawlessness may have led to a downturn in national public concern. Viewers recoiled at the horrible things people were doing, were distracted from the horrible things people were enduring, and for a news cycle or two the call for law and order competed with the call for a more urgent rescue effort.

Much of what was initially called looting, moreover, might better be described as scavenging. A hungry resident who takes a bagful of food from a flooded-out grocery store may technically be stealing, but surely he or she is in a different moral category than a neighbor who liberates a dozen TVs from an electronics outlet. In emergency conditions, scavenging is rational and pro-social. If diabetics were dying for lack of insulin while abandoned hospitals and drugstores had plenty, you could make a case that there was too little “looting” rather than too much. A young man commandeered a school bus and drove a few dozen people to Houston, where authorities debated whether or not to let the unauthorized driver and passengers into the Astrodome (though at least they didn’t arrest him for stealing the bus). Note, too, that the looting that wasn’t emergency scavenging was still more about loss-of-control or irrational desperation or unbridled rage than about greed. What could you do with a dozen TVs in doomed, powerless New Orleans? You couldn’t watch them or pawn them.

A September 3 Google News search for “Katrina lawlessness” got 5,110 hits. “Katrina heroism” yielded only 67. (Not that lawlessness is the biggest single piece of the Katrina storyline. “Katrina victims” got 61,800 hits, compared to just 8,770 for “Katrina survivors.” Where are those resilient, law-abiding New Yorkers when we need them?)

Like the criticisms of emergency responders, the focus on lawlessness is at least in part a response to our feelings of helplessness and inadequacy. At the Twin Towers we had the terrorists to be enraged at. In New Orleans, lacking a perpetrator, we got enraged at bureaucratic ineptitude and sporadic lawlessness instead.

It’s not just the media and the public that may have exaggerated the extent of the lawlessness. Apparently, so did the law-enforcers. When the authorities stopped evacuating the Superdome because shots had been fired, some of those charged with implementing the evacuation protested that it seemed safe enough to them to continue – certainly safer than leaving people to suffer inside. I wasn’t there, and I don’t want to second-guess the decision to postpone the evacuation. Could it have been more a response to the Katrina storyline than to the actual situation on the ground? Similarly, many survivors inside the Superdome and the convention center made impressive efforts to hang onto their humanity under enormous stress. They picked up litter; they cared as best they could for children and the elderly. At one point, apparently, law enforcement officials actually used stun grenades to maintain order. Again you have to wonder whether this was a response to the actual situation or to the Katrina storyline.

There are now far too many anecdotal accounts of authorities who seemed terrified of violence that never materialized. Virtually any civilian attempt to band together and organize risked being interpreted as a threat. One of many Katrina blogs tells the story of a group of tourists – some of them EMS professionals who had been attending a conference in New Orleans – whose collective efforts to get out of town were consistently impeded by the authorities. When they camped outside a police command post on Canal Street, the police commander talked them into leaving by promising that there would be buses waiting on the Pontchartrain Expressway bridge to Gretna. When they made it to the expressway there were no buses. When they started walking across the bridge to safety, Gretna sheriffs fired over their heads.

Or consider the tone and content of this September 3 press briefing by Lt. Gen. H. Steven Blum, chief of the National Guard Bureau:

[There was] a potentially very dangerous volatile situation in the convention center where tens of thousands of people literally occupied that on their own. We had people that were evacuated from hotels, and tourists that were lumped together with some street thugs and some gang members that – it was a potentially very dangerous situation.
We waited until we had enough force in place to do an overwhelming force. Went in with police powers, 1,000 National Guard military policemen … yesterday shortly after noon stormed the convention center, for lack of a better term, and there was absolutely no opposition, complete cooperation, and we attribute that to an excellent plan, superbly executed with great military precision…. There were no soldiers injured and we did not have to fire a shot.
Some people asked why didn’t we go in sooner. Had we gone in with less force it may have been challenged…. As soon as we could mass the appropriate force, …they were immediately moved off the tail gates of C-130 aircraft flown by the Air National Guard, moved right to the scene, briefed, rehearsed, and then they went in and took this convention center down.

If it turns out that there really was a lot of lawlessness in the wake of Katrina, it will be important to try to understand why. Decades of disaster research say there usually isn’t. Some people inevitably lose control or take advantage of the chaos, but in general crisis brings out the best in people. Researchers distinguish “consensus situations” like disasters, when people pull together and looting is minor, individual, covert, and socially disapproved, from “dissensus situations” like riots, when people are in conflict and looting is major, done in groups, overt, and socially approved.

There are disaster researchers in New Orleans now, trying to understand whether Katrina provoked “dissensus behavior” in what should have been a consensus situation – and, if so, why. Was New Orleans different (if it was different) because of its long history of poverty and crime? Did historic mistrust of government lead people to ignore efforts to maintain order – or, conversely, have Americans come to expect a higher standard of rescue and protection than government could deliver this time? Was the violence a response to survivors’ sense that help wasn’t coming, that they had been abandoned and needed to fend for themselves, and that their abandonment was intentional and racially motivated? Was it an underclass exploding in pent-up rage that had nothing really to do with the hurricane at all? Did the stories about violence themselves lead some people to take their guns with them when they hit the streets? Above all, what should we do differently next time? Figuring out how to improve security during catastrophes is a legitimate concern – but figuring out how to help people stay civilized is the more fundamental task.

My own hypothesis is that people are likeliest to lose control when they feel powerless and trapped – and when they feel (rightly or wrongly) that those with real power, the people in charge, are oblivious to their need. Arguably that describes the situation in much of New Orleans for several days last week.

7. Inadequate communication has been one of the major problems in Katrina emergency response.

I strongly suspect that one of the least noticed yet most important emergency response failures in the early days of Hurricane Katrina was the failure of officials to talk to the survivors … even when there wasn’t much to say.

Most of the people reading this column know what it feels like to wait in an airport for news of a delayed flight. This is a very minor, everyday “crisis.” Even though we’re completely safe and reasonably comfortable, as the minutes pass (minutes, not days!) we tend to regress. We chafe not just at the absence of information but at the absence of any communication at all, at what feels like the total indifference of the staff to our need to know what they know.

People in crisis require a steady stream of communication. Real information is best, of course, but telling people you don’t know when the buses will arrive either is enormously better than telling them nothing at all. Those of us who were following Katrina from afar learned all we needed to know from television and the Internet. Those who were living through it, or trying to live through it, needed to know far more – and in most cases learned far less. One vivid difference between New York City in the days after 9/11 and New Orleans in the days after Katrina was that New Yorkers knew what was going on. Many journalists have noted that their interviews in New Orleans were often two-way; as much as they wanted to tell their stories, people were desperate to find out from reporters what was happening. (Officials, too, often asked reporters as many questions as they answered.)

This isn’t just about the power being out. I’d love to get transcripts of all the communications aimed at Katrina survivors – not just what was put out over the radio, but even what went out over the bullhorns. My hypotheses: Too little was said. Too much of what was said was instructions. There was a dearth of explanations; of apologies for not having more information; of promises to try to get more information; of expressions of solidarity, compassion, and hope.

What’s needed here is a willingness to keep communicating with people, even when you have no help to offer. Too many responders seemed to be avoiding crowds unless they had something specific and useful to say. Maybe they were embarrassed by their inability to help, and so like defensive airline gate agents they treated the survivors as if they were the problem. The focus on lawlessness played a role here also. At least sometimes, responders must have been afraid to talk to the crowds, and especially afraid to admit they couldn’t help – which may explain the widespread reports of false official promises of food, water, or buses just a few blocks that way.

Communicating adequately with Katrina survivors will continue to be a challenge for many months. We have all seen stories of families separated in the evacuation and still struggling to find each other. The Gulf Coast diaspora is nationwide – several hundred thousand people desperate for word of their homes, their neighbors, their communities, their prospects and options. Rumors will fly, provoking needless anxiety and impeding psychological recovery. People will need accurate information, compassionately expressed. Once again, instructions are not enough.

8. Emergency communication equipment and internal communication logistics still need work.

The other side to the inadequate communication, of course, is equipment and logistics. The first major crisis I ever worked on was the 1979 nuclear power accident at Three Mile Island. For days, TMI control room operators couldn’t get a call out because the phone lines were jammed with people (many of them reporters) trying to call in. Now, 26 years later, I’m reading about widespread cell phone failures in New Orleans. I don’t know if there are portable emergency cell towers that can be brought to the scene of a disaster to keep communications going, but I’ll bet that’s feasible. I do know there are wind-up cell phones and radios that don’t become useless when the battery runs down – but apparently most emergency responders don’t carry them. And it is a little shocking four years after 9/11 that emergency responders are still coping with incompatible radio frequencies.

As for logistics, chaos goes with the territory in emergencies. There are always detailed flow charts detailing who gets to make which decisions – and the flow charts never seem to work very well. Unexpected people and institutions seize power they weren’t supposed to have, while others shrink from exercising theirs. Questions arise that weren’t considered; decision-makers show up who aren’t in the flow charts. From TMI in 1979 to SARS in 2003, I’ve never seen a crisis where the post-event analysis didn’t say internal communications were terrible, especially in the first few days.

Were internal communications more terrible than usual this time? Or does the problem just seem less forgivable because the disaster itself was so terrible? I don’t think we know yet. Either way, internal communication logistics need work.

One of the most predictable results of poor internal communications, by the way, is conflicting messages to the public – which is the main way the public discovers how poor internal communications really are! I routinely advise my clients that “speak with one voice” is an unachievable crisis communication goal. If dissent exists, as it almost always does, it’s going to come out. A sounder goal is to try to express conflicting opinions in ways that show mutual awareness and mutual respect. “We know the Corps of Engineers is desperate for helicopters to start fixing the levees, but right now we need every helicopter we have to rescue people who are trapped.” As Katrina showed, even that is a difficult goal to achieve.

9. Bureaucracy needs to be relaxed in emergencies.

Often accompanying the complaints about logistical foul-ups were complaints about bureaucratic red tape. Everyone has a favorite example. Mine is the story that FEMA, fearing that terrorists might take advantage of Katrina chaos, wouldn’t allow evacuation planes to fly out of New Orleans unless an air marshal was available for each plane. Runner-up: A story that firefighters were getting sexual harassment training before being allowed to conduct rescues.

There is a long history of officialdom hesitating too long to relax standard procedures under emergency conditions. At Three Mile Island, hundreds of automatic alarms in the reactor control room made it impossible to think. But Nuclear Regulatory Commission officials objected at first when the utility wanted to disconnect the alarms.

But it is worth noting that one function of bureaucratic red tape is to prevent logistical foul-ups. Standard operating procedures aim to make sure that everything and everyone is properly coordinated. If we want governments to cut bureaucratic corners in emergencies and Just Get It Done, we need to be a little forgiving when responders end up working or talking at cross-purposes sometimes. If we want responders to Get Started Now without waiting for confirmation that everything is in order, we need to recognize that sometimes they’ll end up at the wrong place with the wrong equipment.

I do believe that bureaucracy needs to be relaxed in emergencies. I believe that officials running the response effort need to understand that doing the wrong thing is less harmful than doing nothing out of a paralytic fear of doing the wrong thing. But I see few signs that the public is prepared to cut them slack when they do the wrong thing. So I end up as sympathetic as I am critical.

The focus on lawlessness, moreover, may have had a lot to do with the inflexibility of the bureaucracy. Officials who believed New Orleans had sunk into anarchy had that much more reason to enforce rules rigidly – not letting rescue workers cross the state border until the paperwork was done; not letting people off evacuation buses anywhere short of the destination; not letting the Red Cross into New Orleans because everyone was supposed to be leaving; not letting pedestrians out of New Orleans because everyone was supposed to muster at an official shelter and be processed onto a bus; not exceeding their authority no matter how urgent the need. There was a feeling that only the right people could be trusted to take action, and then only if they had the right authorization. Initiative looked dangerous; initiative on the part of citizens looked extremely dangerous.

Ego probably also played a role in bureaucratic rigidity. Certainly early on, it was harder than it should have been for officials to acknowledge how bad things were getting, how urgent it was to call for reinforcements. This was visible in early local reluctance to push hard for federal help, in early federal reluctance to accept foreign help, in early reluctance at every level to break the bank or break the rules in ways that might turn out unnecessary or even harmful.

Part of relaxing the bureaucracy is relaxing the communication bureaucracy. Government organizations have elaborate, hierarchical procedures for getting public statements approved. In an emergency, there’s no time to let a dozen senior officials argue over wordsmithing. Streamlined procedures are needed. Messaging decisions need to be made quicker and at lower levels. The cost of streamlining: Some of what gets said is wrong, or in conflict with what others are saying. The cost of not streamlining: Not enough gets said. Better to streamline, get the word out, and live with the errors and conflicts. (But note: One of the errors may be exaggerated reports of lawlessness. You can’t win.)

Part of relaxing the bureaucracy, too, is letting the amateurs in. There are endless stories about volunteers having trouble getting the authorities to accept their help – equipment waiting for the right permits, individuals told they’d have to go through a training process, etc. Once again, I recognize how unforgiving we are all likely to be if the contributed equipment turns out substandard or the untrained volunteers end up hurting themselves and others. But officials have severely underestimated the importance of letting people do something to help. Spurning help, whether it comes from a rock star or a Cuban dictator, isn’t just a lost opportunity; it’s a thumb in the eye of both those who want to help and those who need the help. And not making a priority of accepting and using people’s (and countries’) help is spurning their help.

Spurning the help of ordinary citizens who are desperate to help is especially cruel. When I do crisis communication training for government officials, one of my themes is the need to prepare to use volunteers wisely. I usually get a fair amount of pushback. Volunteers are more trouble than they’re worth, I’m told; there’s no supervisory structure for them, no liability insurance, no training program. We’re professionals here. We run a professional operation. Let them give us money … and stay out of the way.

It’s not going to happen. In moments of severe crisis, those who are safe suffer vicariously with those who are not. Meeting their emotional need to be part of the solution is as important as harnessing their genuine ability to help. Logistical and legal concerns are real, and need to be addressed. (The guild mentality is also real, and much harder to address.) But in a crisis like Katrina, the dangers of a closed-off bureaucracy are worse than the dangers of volunteer chaos.

10. It’s not about Katrina, or hurricanes, or New Orleans.

So far only three journalists have called me to talk about Hurricane Katrina. Two of them wanted to know what the lessons were for New Orleans, and for being ready to cope with a Category Four or Category Five hurricane. I said I thought those were the wrong questions. What are the lessons for San Francisco and being ready to cope with a monster earthquake, I asked the reporter from San Francisco. What are the lessons for Toronto and being ready to cope with a really big SARS outbreak, I asked the reporter from Toronto.

The lesson of Katrina cannot possibly be that we should be better prepared for any unlikely catastrophe that is actually going to happen … but not waste our money on unlikely catastrophes that won’t happen. Let me be explicit here about what should be obvious: Until it happens, we don’t know which unlikely catastrophe we need to prepare for.

After every catastrophe, evidence emerges that “the authorities were warned this might happen,” and they took insufficient precautions. It was true for the Challenger disaster. It was true for 9/11. It is certainly true for Katrina. The U.S. Army Corps of Engineers has long worried aloud that New Orleans’s levees might not hold in a major storm. State and local agencies have agitated (politely) for more federal money to reinforce them. The Federal Emergency Management Agency has had the levees high on its list of natural disaster vulnerabilities to be addressed pretty soon.

Citizens were warned too, of course. New Orleans’s major newspaper, The Times-Picayune, ran a five-part series in June 2002 entitled “Washing Away.” It began as follows: “It’s only a matter of time before South Louisiana takes a direct hit from a major hurricane. Billions have been spent to protect us, but we grow more vulnerable every day.” Most of what went wrong in Katrina’s wake was covered three years earlier by the Times-Picayune. (Part two includes this prescient passage: “Evacuation is the most certain route to safety, but it may be a nightmare. And 100,000 without transportation will be left behind.”) The series focused attention on the threat and the ways it might be better addressed – but as it turned out, too little attention to provoke much improvement.

The problem is that the authorities weren’t just warned about the catastrophe that happened; they were also warned about dozens of catastrophes that didn’t happen. They took insufficient precautions against those too – that is, precautions that would have turned out insufficient if those other possible catastrophes had happened. The argument should always be about how much we want to spend for how much preparedness for how many possible catastrophes in the future. It doesn’t help to argue instead about how much we now wish we had spent for maximum preparedness for the catastrophe we just endured.

Tsunamis are far more common in the Pacific Ocean than in the Atlantic and Indian Oceans. But after last year’s horrific Indian Ocean tsunami, plans were made to develop a tsunami warning system for the Indian Ocean, modeled on the Pacific Ocean system that has worked well for decades. Plans have even been proposed for an Atlantic Ocean tsunami warning system. So Banda Aceh will probably be better prepared if it ever faces another tsunami, and New York might be better prepared for its first in modern times. But as far as I know, neither region responded to the Indian Ocean tsunami by beefing up its hurricane or typhoon preparedness plan. Nor, of course, did New Orleans.

It is natural for people to think that way. That’s why we have so many metaphors that address this particular error: locking the barn door after the horse has escaped; fighting the last war. And it is natural for the media to cover issues that way, too. But recriminations about how yesterday’s emergency was poorly prepared for and poorly handled are meaningless until they are recast as recommendations about how tomorrow’s possible emergency should be prepared for and handled.

In the wake of Hurricane Katrina, absolutely nobody is saying publicly what most experts are thinking privately: You can bankrupt a country preparing for possible emergencies, most of which will never happen. The media are now reporting that Hurricane Katrina was the worst natural disaster in U.S. history. Leave aside the question of whether the flooding of a coastal city constructed below sea level is a “natural” disaster. If it’s really the worst in U.S. history, then it’s almost a foregone conclusion that we weren’t adequately prepared. In fact, it’s almost a foregone conclusion that we shouldn’t have been adequately prepared. What kind of society devotes scarce resources to preparing “adequately” for a worse disaster than has ever before happened? We are a rich country, but no country is that rich!

Although it’s probably too soon to say those things on television (I’m hesitant even about writing them here), eventually people need to understand. There are lessons to be learned from Hurricane Katrina, ways to improve our ability to prevent and respond to various sorts of disasters. But if we now over-prepare for Category 5 hurricanes in coastal cities below sea level while changing absolutely nothing about our preparedness for anything else, we will have learned one wrong lesson. And if we imagine that we can prepare adequately for every possible disaster, we will have learned a different, equally wrong lesson.

Risk communicators need to teach these unpleasant truths to the public. Otherwise, the public will demand too little – and our officials will focus excessively and exclusively on hurricane preparedness. Or the public will demand too much – and our officials will pretend, yet again, to be ready for every possible disaster.

11. Low-probability high-magnitude risks are hard to think about.

Until it happened, the devastation wrought by Hurricane Katrina in Louisiana and Mississippi was a low-probability high-magnitude risk. Hurricanes are common; really bad hurricanes are occasional; really bad hurricanes that strike the U.S. are infrequent; really bad hurricanes that wipe out a major U.S. city are unprecedented. Some people have been saying for years, and many people are saying now, that it was bound to happen eventually – which is true as a matter of principle. Everything that isn’t impossible is bound to happen eventually; monkeys playing on typewriters will write Hamlet eventually. I don’t mean to be flippant. A disaster on the scale of Katrina was always more likely than monkeys writing Hamlet, and a disaster on that scale was always likelier in hurricane-prone, below-sea-level New Orleans than most other places. Still, the probability was low. And as with any unprecedented event, the probability was in a real sense incalculable.

Risk perception experts know that low-probability high-magnitude risks are hard to think about. The implication of “low probability” is: Don’t worry about it. The implication of “high magnitude” is: Take precautions. The combination induces ambivalence.

We resolve the ambivalence in one of two ways. Either we decide that low probability is tantamount to zero probability, so the risk’s high magnitude becomes irrelevant and action becomes unnecessary. Or we decide that the probability isn’t really all that low after all, so the risk is serious and the case for action is clear. What we don’t seem to be very good at is holding in our minds simultaneously the thought that this is really, really bad and the thought that it’s really, really unlikely – then deciding what precautions make sense. We over-react or we under-react.

Before Katrina, most people thinking hypothetically about a disastrous hurricane shrugged off the risk as too unlikely to deserve attention. As Katrina approached, many in its path thought a disaster was likely, evacuated, and turned out right. Others thought a disaster was unlikely, stayed, and turned out wrong. Comparatively few decided that a disaster was unlikely but evacuated anyhow. Taking precautions that will probably turn out not to be needed is often the wisest response to a low-probability high-magnitude risk. It’s always the rarest response. Those who are left in New Orleans are making similar judgments now about the risk of getting looted if they leave versus the risk of getting sick if they stay. Both risks are high-magnitude. Many are grossly inflating the probability of the former, while imagining the probability of the latter to be close to zero.

It’s worth pondering some other high-magnitude low-probability risks in recent history:

  • A bioterrorist attack on the U.S. using smallpox as the weapon. The intelligence community imagined that the probability was high. The public health community imagined that it was zero. There was a credible case to be made for a vaccination program, grounded in the huge magnitude of the risk notwithstanding its unknown but small probability. Nobody managed to make that case.
  • Weapons of mass destruction in Iraq. Once again, the proper argument was that the risk of WMDs in the hands of Saddam Hussein was intolerable even though the intelligence information was vague and the probability of an actual threat might be low. The Bush administration argued instead that it was confident the WMDs were there. So far, at least, it turned out wrong. (If you blame the President for taking the WMD risk too seriously and the hurricane risk too lightly, try to come up with a rationale that explains why – other than the fact that both decisions happened to turn out wrong.)
  • Overproduction of greenhouse gases leading to global warming and worldwide devastation. Environmentalists started out claiming that we couldn’t afford to wait for really strong evidence before taking action. They were mostly ignored. So they claimed instead that the evidence was really strong already, and got a little more action. The evidence is somewhat stronger today, but the controversy still hinges foolishly on the strength of the evidence, rather than on the need to act before we’re sure. (Climatologists expressed similar fears about global cooling in the mid-1970s. They were ignored and turned out wrong.)
  • Meltdown at Three Mile Island, flooding central Pennsylvania with deadly radiation. I was there when Pennsylvania Governor Dick Thornburgh made the decision not to evacuate central Pennsylvania. (Instead he asked pregnant women and preschool children to leave town, less because that made sense than because he felt he ought to order something.) The Unit Two reactor was in bad shape; there was a hydrogen bubble in the containment that threatened to explode; experts differed on how bad things might get. But evacuations can kill people too. They cost a lot of money, and they threaten the career of a politician who orders one that turns out unnecessary. Thornburgh decided to gamble, and the radiation release from TMI was minimal.
  • A bird flu pandemic killing hundreds of millions of people and plunging the entire world into a deep economic depression. We know flu pandemics happen roughly three times per century; we know we haven’t had a pandemic since we acquired the surveillance and medical capability to make preparedness feasible; we know bird flu has been nominating itself as the likeliest candidate since it first hit Hong Kong in 1997. Some governments and some individuals are taking some precautions. I think we should all take more. But even I don’t think we should do as much as a post-pandemic Congressional investigation will think we should have done.
  • A tsunami or a major hurricane striking Manhattan and taking out Battery Park City. It’s a disaster waiting to happen. Why did the Mayor let them build there? Why isn’t he making plans to evacuate the city? If the forecasters see a credible threat and he dares to attempt an evacuation, why will he leave thousands behind? Why won’t the feds get there in time?

It’s not clear what the “right answer” is in any of these situations. What is clear is that deciding what answer you think is best starts with recognizing that the horrific risk you’re worrying about is neither likely nor impossible.

What determines whether people magnify a risk’s low probability or “round off” to zero? As the preceding examples show, advocates on both sides make this decision tactically. Their assessment of the probability depends on their precaution policy preference more than the other way around. But ordinary people respond most to what I call outrage. If the “outrage factors” – voluntariness, control, dread, familiarity, and the rest – are on the high-outrage side, people are inclined to think the unlikely risk is serious, and thus decide it must be pretty likely as well. If the outrage isn’t so high, their low concern persuades them that the risk is too unlikely to worry about.

Before Katrina, hurricane outrage was modest. Hurricanes did have some of the outrage factors going for them – memorability, in particular. But living in hurricane country seemed more or less voluntary and leaving seemed more or less under your own control; the hurricane itself was natural and familiar; the hurricane’s choice of victims wasn’t unfair or morally freighted; hurricane forecasters and emergency responders were highly trusted. And hurricanes weren’t an especially dreaded risk. Most Americans’ response to hurricanes has traditionally been excitement rather than actual fear; and anger seemed completely beside the point.

Katrina changed all that, at least for now. There’s more than enough Hurricane Katrina outrage to fuel all sorts of hurricane precautions, both wise ones and foolish ones. And in hindsight people are imagining that a Katrina-level disaster was far likelier than it actually was. (It happened and it was horrible, so it must have been likely.) Other low-probability high-magnitude risks aren’t getting the benefit of the hurricane makeover.

Is it possible to teach people to focus simultaneously on a risk’s low probability and its high magnitude, so they can reconcile the two into a coherent understanding of the actual situation? Yes, it’s possible. You get it. I get it. So normal people can get it too. But teaching them to get it is hard work. And too many communicators don’t try. They’d rather pretend the probability is high (if they support precautions) or zero (if they oppose precautions). And they’d rather build or suppress the outrage.

If we want a society that deals sensibly with low-probability high-magnitude risks, we will have to teach people to think straight about those risks.

12. We need to beware of second waves … and third waves.

During the SARS outbreak in Toronto, there came a time when the number of new cases fell to zero. Officials and citizens alike breathed a sigh of relief, and joined in the demand that the World Health Organization lift its travel advisory. They were caught unawares by the second wave of cases.

Many catastrophes come in waves. Pandemics roll around the world for a year or two, hitting some areas three or four times. Earthquakes have aftershocks. Terrorists plant secondary bombs to kill those who responded to the initial attack. In the case of Hurricane Katrina, the hurricane itself was the first wave. It struck New Orleans a glancing blow, devastating the Mississippi coast instead. For a day the news was that New Orleans had dodged the bullet. Then the floods breached the levee system and destroyed the city.

It is human nature to relax too soon. One of the tasks of risk communication is to alert people that the crisis isn’t over yet.

As I write this, the rescue of people in acute need is coming to an end, as is the evacuation of those willing to evacuate. The threat now, the third wave, is the risk of infectious diseases. Once again the big news this morning is New Orleans’s mandatory evacuation that still isn’t quite mandatory.

Copyright © 2005 by Peter M. Sandman