Search Peter’s website:

Communicating about the BP Oil Spill:
What to Say; Who Should Talk

The Deepwater Horizon disaster is mostly a crisis – that is, a situation where hazard and outrage are both high – where people are rightly upset. It has elements of other kinds of risk communication. There is some unjustified outrage to manage (such as false or misleading accusations) and there are some “precaution advocacy” tasks to be done (such as convincing spill responders to protect themselves properly). But crisis communication is the main risk communication paradigm here.

In a crisis situation the principal communication task is to communicate honestly about the situation and help people bear their justified distress. Premature, dishonest, or disingenuous over-reassurance is a cardinal crisis communication sin. So is cavalier, unempathic dismissiveness.

Because crises are virtually always evolving, uncertain situations, sources should err on the alarming side. Thus BP’s Tony Hayward was very, very wrong to say on May 19 that “the environmental impact of this disaster is likely to have been very, very modest.” He did a much better job over the past several days in predicting that top kill had only a 60–70% chance of success, and in warning that early signs of what looked like success might well turn out misleading.

(Note: BP has been a client of mine at various times in the past, but not in recent years; nobody has hired me to advise on the current crisis in the Gulf of Mexico.)

BP’s failure to err on the alarming side, early on, has turned the ongoing uncertainty about how many gallons/barrels of oil are escaping every day into a controversy that has provoked accusations of cover-up. The ordinary citizen hasn’t a clue how many gallons or barrels constitute a really bad spill; we don’t even know how many gallons in a barrel. We take our cues from the accompanying language; “only” 5,000 barrels a day is a lot smaller spill than “as much as” 5,000 barrels a day. And we take our cues from how the numbers are changing. If BP keeps saying 5,000 barrels a day and then other experts start saying 12,000 to 25,000 barrels a day, we’re likely to conclude that BP has been intentionally downplaying the spill. That leaves us more distrustful and more alarmed than we would have been if BP had started with an estimate of 25,000 barrels a day.

In a crisis it is extremely damaging to come back later and say “it’s worse than we thought.” Far better to come back later and say “it’s not as bad as we feared.” So estimates and predictions about uncertain situations should be as alarmist as they need to be in order to virtually guarantee that “worse than we thought” won’t be tomorrow’s story. From the outset, BP should have been saying something like this: “We’re hearing estimates as high as 25,000 barrels a day, and some even higher. We’re also hearing lower estimates, some as low as 5,000 barrels a day. We hope the lower estimates turn out right, but we’re not counting on that. For now, we’re thinking of this as a 25,000-barrel-a-day disaster.”

Here are some rules of thumb in making crisis predictions:

• Be willing to make such predictions. The standard PR advice – “don’t speculate” – is usually misused to justify refusal to discuss internal hypotheses. Surgeons, plumbers, and crisis managers must all do better than “we haven’t a clue” or “I’m not going to speculate” when asked what’s likely to happen next. It’s usually better to offer a range of predictions rather than just one, and to call them “scenarios” rather than “predictions” – emphasizing that you’re not sure and you’re not putting all your eggs in one basket. But you can’t talk intelligibly about a crisis without talking about what might happen next.
• Emphasize endlessly that your predictions/scenarios are tentative – as tentative as you think they actually are. Try to replicate in your audience’s mind your own level of uncertainty. Overconfidence is as big a crisis communication sin as over-reassurance. The most common kind of speculation is overconfident, over-reassuring speculation. This is also the most damaging kind of speculation; it backfires disastrously. Many crisis communications experts advise their clients not to speculate at all. I advise my clients to speculate tentatively and alarmingly, emphasizing their hope that their alarming hypotheses will turn out wrong: “We can’t tell what’s going to happen, but here’s what we’re most worried about….”
• Put a lot of focus on the worst case scenario that isn’t vanishingly unlikely. This isn’t literally the worst case conceivable, which is a reductio ad absurdum. (“Not only that, but on the same day the Martians invade!”) The appropriate worst case scenario is the worst case that’s likely enough to be worth planning for. In the situation at hand, that’s presumably months of drilling to create a relief well before the exploration well can be plugged, with oil pouring out continuously until then and an occasional hurricane to exacerbate the disaster.
• Put about equal focus on the likeliest scenario. If the likeliest scenario is the worst case scenario, say so … with great regret. But usually they’re different, and those two scenarios should be your anchors: what you think is most likely to happen, and what you’re most afraid might happen. I don’t know what the experts’ likeliest scenario is for BP’s gushing well.
• Put less stress on scenarios that are better but less likely than your likeliest scenario – but do mention them. I’m not talking about miracles. Vanishingly unlikely deus-ex-machina scenarios will arouse false hopes and thus do more harm than good. But the realistic best case deserves a little attention, though less than the realistic worst case. “Top kill could still work, giving us a chance to cement this leak in the next few days. That’s what we’re hoping for, and working for. But we’re certainly not counting on it.”
• Don’t just make predictions, as if you were an observer. Talk about what you’re doing: both what you’re doing to improve the likelihood of the better scenarios and what you’re doing to prepare for the worse scenarios. It’s important to pair the two. You don’t want to leave the impression that you’re fatalistically awaiting your doom, nor that you’re overoptimistically counting on averting it.
• Pay special attention to communicating your preparations for the really bad scenarios. There’s seldom any risk that the authorities will neglect to explain what they’re doing to try to get the acute problem solved. The danger is that they’ll neglect to explain what they’re doing to help get us through the awful times that may be coming. “Suppose I can’t fish for years.” “Suppose this town’s whole economy collapses.” It’s a mistake to tell people they shouldn’t be asking such depressing questions, or to say something vague like “we’ll cross that bridge when we come to it.” President Obama’s promise on May 28 that “we will be on your side and we will see this through” isn’t bad – but a six-point plan would be better. “We hope it won’t come to that, but if it does, here’s what we’re planning to do….”
• Give people things to do. Military planners have long known that soldiers can bear their anxiety a lot better if they’re doing something useful than if they’re waiting around to see what’s going to happen. Wise crisis managers plan for volunteers – not only because the volunteers can do things to help, but also because it helps the volunteers to do things. Even better than offering people things to do is offering people choices of things to do, so they’re not just keeping busy but actually making choices.

In short, Cassandra is a better guide to crisis communication than Pollyanna. Yes, you’re hoping for the best, and working to avert the worst – and you should say so. But you’re also preparing for the worst, and urging the rest of us to prepare for the worst: emotionally as well as practically. Say that too.

There is one significant downside to this recommended strategy. If the situation turns out a lot less serious than your worst case prediction, perhaps even less serious than your likeliest prediction, you’re due for some criticism for having “hyped the oil spill.” (Or the pandemic. Or the hurricane.) People will blame you for all the worry you put them through, all the tourists you scared away unnecessarily, all the expense you wasted on preparedness that wasn’t needed. The risk perception literature calls this “hindsight bias.” After the results are known, people tend to imagine you should have been able to guess right. So there’s a fair amount of anger in the world today at public health officials for having taken the H1N1 pandemic more seriously than it ended up deserving. In the U.K. a few months ago, some of the same newspapers were simultaneously criticizing the U.K. government for buying too much pandemic vaccine and too little road salt and grit. Editorialists felt that officials should have guessed correctly that the pandemic would turn out mild and the winter would turn out severe.

But it’s not “damned if you do and damned if you don’t.” The criticism for under-responding to what turns out a full-bore emergency is long-lived and devastating. The criticism for over-responding to what turns out a minor problem is briefer and milder. Not to mention that under-responding costs lives, while over-responding costs mostly money and anxiety. If you are trying to figure out how much to sound the alarm about an evolving uncertain situation, remember this rule: It’s “DARNED if you do, damned if you don’t.”

This emphasis on how to talk about the uncertain future shouldn’t overshadow an even more important crisis communication criterion: total transparency about the present. BP was unwise (and naïve) to think it could deny the world real-time access to a video feed of its efforts to plug the hole. Both BP and the feds were unwise (and naïve) to think they could get away with implying that the top kill strategy was proceeding as planned without mentioning a 12-hour hiatus. Reports of efforts to thwart overflights of the plume and videos of stricken wildlife, if true, are further examples of unwise (and naïve) departures from transparency.

As for who should be doing the communicating, my answer is: everybody with knowledge, a viewpoint, or a potential role to play. It is now well established that the public wants to see top leadership personally involved in crisis management. In a disaster this big, we want to see President Obama flying to the Gulf, not once but often; we want to see him in shirtsleeves talking not just with senior advisors but also with volunteers and victims; we want to see him actually placing a boom or washing a bird. This is even more the case for BP CEO Tony Hayward, who would be wise to give virtually hourly sweat-stained news conferences from various hotspots.

In the back of our minds, we all know that Obama and even Hayward have better things to do, that they have people working for them who are much more qualified than they are to make on-the-spot tactical decisions, that their job is to stand back and see the big picture. Nonetheless, we want to know that they’re willing to immerse themselves in the details as well. For Obama, this is essential evidence that he cares, that he’s not just our wonk-in-chief but also our cheerleader-in-chief and our consoler-in-chief. For Hayward, it’s evidence that he knows he has sinned and is prepared to begin the process of atonement.

But we also want to get to know lesser officials as real personalities, people who become the daily “face” of the crisis response. During the Three Mile Island (TMI) crisis of 1979, Harold Denton became the federal government’s point man at the site. President Jimmy Carter made a quick visit, but it was Denton to whom the public looked for support. Rumpled, sleepless, unfailingly calm but never over-reassuring, he was quickly dubbed “Dr. Denton” although he had no doctorate; we gave him all the credentials we could, and we trusted him implicitly.

Metropolitan Edison (TMI’s BP equivalent) had no such figure; the CEO of the utility was virtually invisible, leaving his top engineer, Jack Herbein, to do the best he could. Herbein always wore a freshly laundered suit and always looked well-rested; like everyone else at TMI, he was working incredibly long hours, but he thought it would be unprofessional to let it show. Worse, he always sounded confident, almost uncaring. I asked him why he so consistently ignored the advice of his PR specialist, Blaine Fabian. (Risk communication hadn’t been invented yet.) He told me, “PR isn’t a real field. It’s not like engineering. Anyone can do it.” That attitude, I think, cost MetEd and the nuclear power industry dearly.

National Incident Commander (and just retired Coast Guard Admiral) Thad Allen is the closest Deepwater Horizon has to a Harold Denton. Time will tell how close that is.

As far as I can tell, BP has no Harold Denton equivalent. Some of the BP spokespeople are okay; some come off sounding defensive or hyper-technical. Except for Hayward himself, none has yet to acquire a public personality. A Reuters story recently described Hayward as having “perspiration dripping from under a white plastic BP safety hat” during an interview from a drilling ship in the Gulf. President Obama was photographed in shirtsleeves kneeling on a contaminated Louisiana beach.

One of the conventional recommendations of crisis communication experts is to “speak with one voice” – to decide who the spokesperson will be and try to get everybody else to shut up and funnel all inquiries to that person … or require everyone else to read from the same script. I disagree mightily with this dictum. “Speak with one voice” has never worked very well. Sources leak. Worse, journalists who don’t want to settle for a one-source or two-source story search out a bunch of additional sources with something different to say. If all the knowledgeable people are saying exactly the same thing or nothing at all, reporters interview less knowledgeable people – kooks, pseudo-experts, bystanders, even each other. Long before the era of tweets, blogs, and 24/7 news cycles, “speak with one voice” was a lost cause.

In crisis situations, moreover, people triangulate. We gather information from all the sources we can find – official sources, web sources, friends and neighbors – and we look for commonalities. If BP says it and the Coast Guard says it and your favorite blog says it and your mom says it, it might actually be so. If BP, the Coast Guard, your favorite blog, and your mom all say similar but not quite identical things, that’s actually even more persuasive; it sounds like they got to roughly the same place by different routes. If you’re hearing something very different from one source than from all the others, then you have to decide whom to believe – a decision that may be grounded in wishful thinking or animus or ideology … or, conceivably, careful assessment of the evidence. Whoever you decide to believe, you’re stuck with the anxiety engendered by competing claims.

In many cases, this anxiety may be justified. The range of opinion may accurately reflect the lack of definitive knowledge about the situation, and the lack of consensus about what to do about it. When this is the case, the crisis communication task is to validate that opinions differ, and help people bear the anxiety that provokes – not to try to promulgate a phony consensus.

But when there is actually considerable agreement among the various sources, “speak with one voice” is the wrong way to showcase that agreement. The most credible claim is the claim that seems to come in a wide range of similar but not identical variants from a wide range of different sources. So the best answer to the question of who should be talking to the public is everybody. The various government agencies and companies involved in the crisis should focus on keeping their people (and their adversaries and critics) well informed, not on keeping them quiet or making them ventriloquist’s dummies.

A single overriding official source with everyone else forced to shut up or parrot the script is bound to backfire. It helps turn the kooks into significant sources. A cacophony of competing official sources with radically different stories is also bound to backfire. It conveys chaos. Both trigger skepticism and confusion, if not paranoia. Neither is conducive to a united front against the common enemy: in this case, the oil spill. The goal is a symphony of intertwining, compatible but not identical messages from many well-informed sources who are free to speak their minds.

Sometimes, of course, the “symphony” goal is genuinely and appropriately not achievable. Sometimes authoritative sources really do disagree on important questions, and the public deserves to know that that is the case. Then the communications task – to underscore this important point again – is to help people bear the anxiety caused by the reality of expert disagreement (or expert ignorance), not to try to squelch the disagreement in order to avoid the anxiety.

But often the appearance of disagreement is an artifact. Journalists, of course, are famous for trying to get sources to sound like they disagree with each other, whether the disagreement is real or not. They do that a lot less in a genuine crisis than on a routine story – but the habit runs deep. And in a slow-motion crisis, there is lots of time for stories about disagreements among officials and experts.

The appearance of disagreement about the oil spill – and especially about issues of responsibility and blame – has been manufactured chiefly by the government – mostly by Congress, but sometimes by the President as well. When BP says that it is responsible for cleaning up the oil but Transocean is responsible for the safety of crewmembers on the rig, it is simply explaining the way the law works – albeit in oversimplified form. When Transocean says that it is basically an agent of BP and BP has ultimate responsibility for all decisions made aboard the rig, it is simply explaining the way the contract reads – again, in oversimplified form. Several times in recent weeks various government officials have demanded these explanations and then attacked the companies for providing the answers. As President Obama put it on May 14: “I did not appreciate what I considered to be a ridiculous spectacle during the congressional hearings into this matter. You had executives of BP and Transocean and Halliburton falling over each other to point the finger of blame at somebody else.”

There have been some real examples of BP, Transocean, and Halliburton unwisely scapegoating each other – presumably at the behest of their legal departments. This is unwise because the dynamics of blame reliably follow the principle of the risk communication seesaw – when there is ample blame to go around, as there clearly is here, the organization that gets blamed most is always the one that tries hardest to blame everybody else instead. Seesaw aside, BP should and inevitably will bear the lion’s share of the blame. It’s the iconic company, the one whose logo and name are universally recognized, the one that claimed to have progressed Beyond Petroleum, the one that hired everybody else. It can only worsen its situation by trying to point fingers elsewhere.

One glaring example of finger-pointing is testimony from the chief Deepwater Horizon mechanic, Douglas Brown, who told federal officials about an argument between a BP man on the rig and Transocean employees, shortly before the explosion on April 20. Brown said there was “a [verbal] skirmish,” and then: “The company man was basically saying, ‘Well this is how it’s going to be.’” The rig’s employees “reluctantly agreed,” Brown said. What neither Brown nor most journalists mentioned was the fact that everyone aboard Deepwater Horizon – BP and Transocean employees alike – had Stop Work Authority , a right and an obligation long enshrined in offshore drilling operations. Anyone who believed a practice to be dangerous could force a halt, without risking retribution, even if a visiting honcho wanted the work to proceed.

On the whole, though, I don’t believe BP (or Transocean, or Halliburton) is trying to point fingers elsewhere. These companies are getting set up by politicians who want it to look like they’re trying to point fingers elsewhere … which is largely a tactic on the politicians’ part to wriggle out from under their own responsibility for insufficiently skeptical and aggressive regulation of oil development in the Gulf of Mexico, including insufficient requirements that companies (and the government itself) be prepared to cope with a catastrophic blowout if one should occur.

Or maybe there was nothing major wrong with the regs. Maybe an occasional oil spill, even an occasional disastrous oil spill, is the price we have tacitly agreed to pay for the benefits of domestic oil development. Not every disaster is a scandal. Some are just tragedies. And some, of course, are scandals. It’s hard to tell the difference. What we know is that humans in general and Americans in particular greatly prefer scandals. We have no stomach for the idea that horrible things can happen with nobody to blame. As we assess the blame game, then, we need to consider not only the possibility that the wrong entities are being blamed, but also the possibility that “blame” is the wrong frame by which to assess what is happening.

Though it looks that way to me, I am not qualified to judge whether BP and the federal government jointly underestimated the risk of a deepwater blowout. Nor am I qualified to judge whether BP, Transocean, Halliburton, and the federal government – or some subset of those four – cut corners aboard the Deepwater Horizon. I do have a sense that the government is doing everything it can to blame the companies, and the companies are maladroitly allowing it to look like they’re trying to blame each other (and sometimes actually trying to blame each other).

Even “speak with one voice” would be an improvement over that pattern.

Copyright © 2010 by Peter M. Sandman

Contact information page:    Peter M. Sandman