How the Market Can Save the Gulf of Mexico

Although the damages of the oil spillage in the Gulf of Mexico may exceed $50 billion, American law may limit liability to as little as $75 million. Seizing the opportunity of the BP disaster, Senator Menendez (D, NJ) proposed to raise the ceiling to $10 billion.

That may be a step in the right direction but I am left to wonder why there should be any liability limits at all. Liability sustains markets. According to Adam Smith, markets employ self-interest to the benefit of the public good because economic actors bear the costs of their activities.

Limiting liability creates an incentive for irresponsible behavior where corporations can make more money at the cost of third parties and the general public. As a consequence, liability limitations threaten property rights. We may destroy each others property not only with impunity but under the protection of the law.

That makes for a brutish, nasty and short life if you ask me. In the Gulf of Mexico, tens of thousands of fishermen are losing their jobs and their businesses as do hoteliers, gastronomers, and their employees. That is not their fault. They did not soil and pollute the Gulf. BP and its contractors did.

The logic of the market requires that the perpetrators pay for the consequences of their actions. Otherwise, there will not only be injustice but waste.

Murkowski and Inhofe claim that the insurance for $10 billion would be unaffordable for small drillers. Adam Smith might reply: “Tough, then you shouldn’t be in a business that can devastate an entire region of the United States of America.”

Even if you agree with Murkowski and Inhofe about curtailing markets, they are also wrong that insurance would be prohibitive for drillers. It turns out that liability insurance would be a relatively minor part of an oil rigs operating expenses.

NOAA maps over 3,500 rigs in the Gulf of Mexico. If they were all required to obtain $10 billion liability insurance and a disaster like the current one would occur every twenty years, then the annual premium per rig would be $200,000. That includes 40% overhead for the insurance’s costs and profits.

Stick it in a spreadsheet and see for yourself. $200,000 is barely the equivalent of four annual salaries on an oil rig. Of course, once we take smaller insurance cases into account, premiums will increase substantially. However, as a share of operating costs even a million dollars a year remains more affordable than most people’s automobile insurance.

Once we hold businesses accountable by requiring them to carry sufficient insurance, in effect, the insurance companies will protect the environment generally and the Gulf of Mexico particularly. It is not in the interest of the insurance companies to pay for damages. Therefore, the insurance companies will establish uniform minimum safety requirements, which the insurers will enforce with inspections.

Non-compliance will result in loss of insurance, which in turn, ought to suspend the operating license of the drillers.

I suppose that there is some merit in watching out for small businesses but no business should operate at the expense of the public. Being pro-business is not the same as pro-market. When Murkowski and Inhofe carry the water for the oil industry, we need to realize that their actions cost all of us and undermine the institution of the market.

By contrast, abolishing liability shelters and requiring liability insurance harnesses the power of markets on behalf of the environment and the public interest. It will probably protect the public interest and the environment more effectively than any other regulation.



  1. The idea is hardly unique to Welcher’s work. The notion that liability and property rules internalize costs, leading to efficient allocation of resources goes back at least to Ronald Coase.


  2. Right, Nate–this post is heavily doused in the scent of the Coase Theorem. A pity that we have no oceanic property rights to extend the analysis…since (I’m conjecturing here) that there aren’t an extremely high number of oil drilling firms in that region.


  3. I suppose you can subsume this argument under the Coase theorem. My point is actually much more narrow than the Coase theorem.

    The notion that markets require liability goes back directly to Adam Smith. During the 20th century, Walter Eucken has championed the opposition to liability exemptions.

    The implications of the Coase’s ideas, who never formulated the Coase theorem, by the way, probably go farther than Eucken’s work, which is why the former won a Nobel price and the latter did not.

    Eucken and Coase were roughly contemporaries. Since Coase was British and worked in the United States, is work is in English, which is a considerable competitive advantage in contemporary academia.


  4. The really difficult issues in imposing liability are not about ceilings. (Although, I think Hellmut is absolutely right that limiting liability to some dollar amount is a bad idea.) Rather, the tricky questions arise when one examines notions of fault and causation. If liability really is just about creating optimal incentives (and FWIW I don’t think it is and have argued for this in print), then you get into difficult questions about the cheapest cost avoider and transaction costs. It’s not clear without more argument that BP should bear the full costs of its mistake. We would also need to demonstrate that transaction costs are functionally zero or that BP is the cheapest cost avoider. In this case BP probably is the cca, and the transaction costs are high, so Hellmut’s analysis is sound but in other contexts this is probably not going to be true.

    While I agree with Hellmut on liability limits, one of the problems with these kinds of happy libertarian market stories is that they tend to implicitly assume a world of no or at any rate modest transaction costs. In such a world, however, the Coase’s Theorem suggests that the assignment of liability to one party rather than the other is irrelevant. If BP could dump oil with impunity into the gulf, then those forced to internalize the costs would simply pay BP not to dump.


  5. Thanks, Nate. It is true that I am not dealing with problems of evidence, which can be substantial. However, I am not really interested in guilt but in responsibility, which simplifies the matter. I need to be neither criminal nor negligent to pay for the cleaner’s bill if I spill coffee over your suit.

    You probably understand the economics of burden of proof standards better than me but intuitively I would lean towards minimal requirements, which would simplify matters.

    Clearly, the oil disaster is a consequence of BP’s economic activities. While other considerations might apply, such as the logic bankruptcy rules, BP should pay for the damages.

    The matter of the cheapest cost avoider is actually a secondary question. If no one is liable for damages then there will be more waste than with liability. Therefore, there needs to be liability. Then we need to determine how to do that more efficiently.

    I suppose there might be cases where poorly organized liability can lead to even greater waste, but to be honest, I have a hard time imagining such circumstances. Since you seem to have worked on the issue, I would be grateful for your suggestions.

    With respect to transaction costs, the back of the envelope calculation takes that partially into account by assuming 40% overhead in insurance premiums. That’s a substantial proportion, isn’t it?

    Of course, the costs of adjudication remain unaddressed. That seems to be a core state function. Having the tax payer absorb those expenses might only lead to minor allocation problems. On the other hand, perhaps there is a fee structure or a functional equivalent that can efficiently absorb those costs.


  6. In such a world, however, the Coase’s Theorem suggests that the assignment of liability to one party rather than the other is irrelevant. If BP could dump oil with impunity into the gulf, then those forced to internalize the costs would simply pay BP not to dump.

    Exactly, that’s one of the reasons why there ought to be an insurance requirement. While the insurance would not literally pay BP, it would create incentives for BP to internalize costs.

    Such incentives might range from payments such as premium rebates to loss of insurance, which ought to cancel the rig’s operating license unless BP can obtain another insurer.


  7. “However, I am not really interested in guilt but in responsibility, which simplifies the matter. I need to be neither criminal nor negligent to pay for the cleaner’s bill if I spill coffee over your suit.”

    In the language of tort, you are in favor of strict liability. I actually tend to agree with you. On the other hand, opting for strict liability rather than a fault system doesn’t eliminate as many problems as it initially appears to. This is because one still has arguments about causation. In order to impose liability, what sort of causation must be shown. Sufficient? Necessary? Necessary plus something else? Most legal systems fudge this issue by invoking something like “proximate causation.” Proximate causation, however, ends up being an amalgem of normative judgments — e.g. was the harm foreseeable? — that start looking much like a fault-based analysis. It is not exactly the same thing but it is similar.

    BTW, if you are really interested in debates about the economics of competing conceptions of liability, the locus classicus for this in tort law theory is the debate between Richard Posner, who is a proponent of fault-based negligence liability, and Richard Epstein, who is a proponent of strict liability. I think that Epstein has the better of the argument, but I think the divide is less stark that folks are willing to admit.


  8. “If no one is liable for damages then there will be more waste than with liability. Therefore, there needs to be liability. Then we need to determine how to do that more efficiently.”

    This doesn’t follow. In the absence of a legal rule shifting costs, then someone, somewhere will bear the costs. The question then becomes whether that person can avoid the costs more cheaply than someone else. If transaction costs are low, even if the person who bears the costs is not the cheapest cost avoider, there will be no economic loss because they will simply buy off the cheapest cost avoider. This will have distributive effects, but will not result in any efficiency loss. On the other hand, if bargaining is expensive then it matters a great deal where we assign liability as it is likely to remain on that party. If our goal is economic efficiency, we want to assign liability to the cheapest cost avoider. If we assign liability in a high transaction cost world to a person who can only avoid damages at a higher cost than someone else our liability rules will create inefficiencies.

    Here is a simple example. If I am driving along the interstate at 75 mph and a person walks in front of me, I will maim them with my car. There is an undeniable sense in which I caused the maiming. Certainly, had I been driving more slowly or not at all I would have avoided causing the maiming. In this sort of a case, where do we assign liability? One answer would be to stick to one’s strict liability guns and say I am responsible for all damage that I cause. Of course, one might argue within the strict liability paradigm that I didn’t “really” cause the maiming because the victim’s agency was involved in stepping in front of the car. Of course, victim agency is virtually always involved. Shrimp fishermen in the gulf of mexico could have paid BP not to drill oil. Their failure to do so contributed to the spill. We don’t, however, regard this are relevant. Why? I suspect it is because we (correctly) regard the shrimp fisher’s failure to buy off BP as entirely innocent. (Notice, this brings in a notion of fault.)

    Return to the freeway example. One reason we don’t assign liability to me as the driver in this case is because it would be quite expensive for me to avoid the accident. I would have to forego the benefits of swift travel, etc. On the other hand, it would be relatively costless for the victim to avoid the accident. Don’t go running on the freeway. So what does the example show? It shows that we can’t rely on the notion of causation to assign liability, not without significantly qualifying what we mean by “causation.”


  9. Thanks for the Posner and Epstein suggestion, Nate. I will have to look into that.

    With respect to liability, I am less concerned about the fair attribution of costs than about properly defined incentives that privilege responsible behavior. That way, we avoid costs in the first place.

    The absence of liability leads to waste because it undermines prevention.

    My apologies, I will take a more careful look at your response when things calm down around here.


  10. I get the point about incentives vs fairness. FWIW, debates in the l&e literature about fault and causation are debates about incentives not fairness. In this context, I am using fault in the technical sense of a duty of care rather than in it’s ordinary moral sense.


Comments are closed.