The Gulf Oil Leak and Optimal Safety Precautions

Economists over the years have had a schizophrenic attitude toward low probability but highly costly events. One school of thought argues that most people give less attention to such events than is merited, and that this explains why many households do not to take out flood, earthquake, and other insurance on very small risks (although there may be other explanations of this fact). Some psychologists and economics claim the opposite, that people give too much attention to low probability events, as in much of modern behavioral economics.

Whether BP paid enough attention to the chances of and damages from a serious leak on its Deepwater Horizon rig in the Gulf of Mexico is at the heart of any evaluation of whether BP was negligent in its safety approaches to this well. The chief executive of BP, Tony Hayward, claimed that an accident of the magnitude of this one had "a one in a million" chance. It is ironic that when Hayward became chief executive, he vowed to focus "like a laser" on safety and reliable operations.

Suppose that his estimate is correct, in the sense that the probability of an accident of this magnitude in any year was one in a million. If the accident ends up causing a $100 billion worth of damages to fish, beaches, loss of lives, and in other ways, the discounted value (at a 5% interest rate) of this expected cost over time would then be approximately 20 times $100 billion divided by one million, or $2 million. A risk neutral BP would rationally not want to spend more than that amount to prevent such a leak from happening. Perhaps the damages from this oil spill will turn out to be as much as $200 billion, although this seems highly unlikely, especially since the value of BP's shares have declined by about $90 billion since the April 20th disaster. Even if the damages were more than twice as large as the decline in share value, the expected damages would only be about $4 million, a modest expected risk for any large business.

Of course, to justify their poor preparation to combat such a large leak of oil, Hayward may have greatly understated its probability. As Posner emphasizes, it is very difficult to estimate with any precision what is the actual probability of highly infrequent events, and such a leak would fall within this class. So if the annual probability were really more like 1/10,000 rather than 1/1,000,000, the expected discounted cost would be about $200 million, a more considerable sum. These examples show that only when the annual probability of such a disaster was reasonably large would it have paid for BP to spend a lot to prevent such a leak from happening.

Given the litigation that resulted from the 1989 Exxon Valdez oil spill off Alaska, and from other oil leaks, BP should have expected that it would be held liable in court for the damages caused by leaks from its deep water drilling. However, it could hardly have expected to be liable for new types of claims, such as the one proposed by the Obama administration that BP should be responsible for the wages lost by workers who are laid off as a result of the six month moratorium proclaimed by the president on all drilling off the American coast. It is ironic that Obama had on March 31st, just a few weeks before the Deepwater Horizon oil spill in April, proposed an extension of offshore drilling. Nevertheless, if the Deepwater leak alerted us to greater risks of offshore drilling than had been realized, BP should not be punished for (inadvertently) providing this useful information.

The president may be right in asking for a moratorium to spend a few months reviewing the safety precautions on all drilling off the US coastline. However, it would be unfortunate if this disaster led to permanently much greater restrictions by the US on offshore drilling for oil since that is at present the most promising source of additional US production of oil. The likelihood, and cost, of future spills has to be weighed against the gains to US national security and energy independence from having greater access to domestically produced oil. Since I believe these gains are substantial, I expect offshore oil production to continue to have an important place in US (and other countries) production of oil since offshore oil is the last frontier in the search for new deposits of oil.

** **

British Petroleum's drilling accident in the Gulf of Mexico this past April is the latest of several recent disastrous events for which the country, or the world, was unprepared. Setting aside terrorist attacks, where the element of surprise is part of the plan, that still leaves the Indian Ocean tsunami of 2004, Hurricane Katrina in 2005, the global economic crisis that began in 2008 (and has been aggravated by Greece's recent financial collapse), and the earthquake in Haiti last year.

The reaction to the latest accident has been surprising. Oil spills and underwater drilling accidents are common, and despite the media hype it is too soon to tell whether this one will prove to be the biggest yet. The amount of oil leaked so far is substantially less than the amount spilled or leaked in previous accidents, including at least one in the Gulf of Mexico.

It is also surprising that so much criticism has been directed at the Obama Administration, and indeed against Obama personally. Most of the criticism is absurd—his failure to react emotionally, and his inability to "just plug the hole," are not personal or professional failings. The Minerals Management Service in the Department of Interior does seem to have been asleep at the switch, but Obama unlike his immediate predecessor cannot be criticized as being hostile to regulation—if anything, he has too much faith in it. MMS is a small and obscure agency far below the horizon of a president's supervision. No president can eliminate all pockets of incompetence in the vast federal government.

It is possible that the number of recent disasters has created a public sense that something is wrong with government: that it ought to be able to prevent all disasters. But this is an unrealistic expectation. Everything conspires against a government's being able to protect its people against disasters, whether natural or man-made. A factor that retards prevention of man-made disasters is the rapid and relentless advance of technology. Regulation lags innovation. The Federal Reserve, Treasury Department, and SEC were no more able to keep abreast of advances in financial engineering than MMS was to keep abreast of advances in drilling for oil at very great depths under water. Slack regulation encourages private companies to adopt a high-risk business model. Risk and return tend to be positively correlated, in finance because risky loans command higher interest rates and in underwater drilling because risk abatement is costly. Business is particularly reluctant to take preventive measures against unlikely disasters because they do not pose a serious near-term threat. If there is a 1 percent annual probability of a disastrous drilling accident or financial collapse, the probability that the disaster will occur any time in the next 10 years is only 10 percent. Business managers have finite planning horizons just like politicians.

Of course, if the consequences of a disaster would be very grave, the fact that the risk is low is not a good reason to ignore it. But there is a natural tendency to postpone taking costly preventive action against dangers that are only likely to occur at some uncertain point in the future ("sufficient unto the day is the evil thereof," as the Bible says), especially if prevention is expensive, if the probability and often the consequences of disaster cannot be estimated with any confidence, if remediation after the fact seems like a feasible alternative to preventing disaster in the first place—and because there is so much else to do in the here and now than worry about remote eventualities. The overcautious business will lose profits, investors, and staff to its bolder competitors; the overcautious regulator will be harassed by politicians pressured by business, labor, and other interest groups.

All the factors that I've identified came together to enable the economic crisis, despite abundant warnings from reputable sources, including economists and financial journalists. Risky financial practices were highly profitable, and giving them up would have been costly to financial firms and their executives and shareholders. The Federal Reserve and most academic economists believed incorrectly that in the event of a crash, remedial measures—such as cutting interest rates—would suffice to jump-start the economy. Meanwhile, depending on how they were compensated, many financial executives had a limited horizon; they were not worried about a collapse years down the road because they expected to be securely wealthy by then. Similarly, elected officials have short time horizons for policy; with the risk of a financial collapse believed to be low, and therefore a meltdown unlikely in the immediate future, they had little incentive to push for costly preventive measures, and this in turn discouraged the appointed officials of the Federal Reserve and other regulatory agencies from taking such measures. Finally, with no reliable probability estimate of a financial collapse available, it seemed prudent to wait and see, hoping that with the passage of time at least some of the uncertainty about risks to the economy would dissipate.

The BP oil leak reveals a similar pattern, though not an identical one. One difference is that the companies involved must have known that in the event of an accident on a deepwater rig prompt and effective remedies for an oil leak would be unlikely—meaning that there was no reliable alternative to preventing an accident. But the risk of such an accident could not be quantified, and it was believed to be low because there hadn't been many serious accidents involved in deepwater drilling. (No one knew how low; the claim by BP chief executive Tony Hayward that the chance of such an accident was "one in a million" was simply a shorthand way of saying that the company assumed the risk was very small.)

But other causal factors were similar in the leak and the financial crisis. If deepwater oil drilling had been forbidden or greatly curtailed, the sacrifice of corporate profits and of consumer welfare (which is dependent on low gasoline prices) would have been great. Elected representatives did not want to shut down deepwater drilling over an uncertain risk of a disastrous spill, and this reluctance doubtless influenced the response (or lack of it) of the civil servants who do the regulating.

The horizon of the private actors was foreshortened as well. Stockholders often don't worry about the risks taken by the firms in which they invest, because by holding a diversified portfolio of stocks and other financial assets an investor can largely insulate himself from risks taken by any particular firm. Managers worry more about the fate of their company because they often are heavily invested in it terms both of human capital and of reputation. But they rarely are held personally liable for the debts of the firms they oversee and, more important, the danger to their own livelihood posed by seemingly small risks is not enough to discourage risk-taking.

Two final problems illuminate the nation's vulnerability to disasters. First, it is very hard for anyone to get credit for preventing a low-probability disaster. Because such a disaster was unlikely to occur, the benefits of taking action beforehand could not be assessed unless the preventive action took the form of a dramatic last-minute save. Had the Federal Reserve raised interest rates in the early 2000s rather than lowering them, it might have averted the financial collapse in 2008 and the ensuing global economic crisis. But we wouldn't have known that. All that people would have seen was a recession brought on by high interest rates. Officials bear the political costs of preventive measures but do not usually reap the rewards.

The second problem is that there are so many risks of disaster that they can't all be addressed without bankrupting the world many times over. In fact, they can't even be anticipated. In my 2004 book Catastrophe: Risk and Response, I discussed a number of disaster possibilities. Yet I did not consider volcanic eruptions, earthquakes, or financial bubbles, simply because none of those seemed likely to precipitate catastrophes.

It would be nice to be able to draw up a complete list of disaster possibilities, rank them by expected cost, decide how much we want to spend on preventing each one, and proceed down the list until the total cost of prevention equals the total expected cost averted. But that isn't feasible. Many of the probabilities are unknown. The consequences are unknown. The costs of prevention and remediation are unknown. And anyway, governments won't focus on remote possibilities, however ominous in expected-cost terms. A politician who proposed a campaign of preventing asteroid collisions with Earth, for example, would be ridiculed and probably voted out of office.