Pete argued yesterday that raising or eliminating statutory caps on oil spill damages wouldn't increase safety - as he put it, "the incentive argument is overblown." There's some truth to this, but I don't agree with his conclusion. This is probably no surprise: in 2010 I recommended that spill liability caps be increased significantly, up to the estimated worst-case damages from each well (in a paper with Mark Cohen, Josh Linn, and Madeline Gottlieb). In terms of incentives this is effectively indistinguishable from no caps at all.
There's two ways to interpret Pete's argument - a weak version and a strong one. I think only the weak version holds up.
The weak version of Pete's argument says that companies won't fully respond to the increased incentive to avoid spills that full liability for damages would create. This is because, like all incentives that operate at the firm level, lack of information, agency problems, and other factors limit their ability to influence employees that are actually in a position to do something. This version of the argument is certainly true. Even small drilling firms are large organizations. Management and ownership have limited information about safety, and limited ability to get employees to take precautions even when management knows what should or can be done. There's only so much management can do, and it's true that there could be systemic problems with risk analysis that are hard to resolve with incentives alone.
The strong version says that raising liability caps won't have any effect because it does nothing to address the real causes of spills - as Pete puts them, "thinking about risk in the wrong way. . . a behavioral economics and risk perception problem."
I don't buy this. The whole point of giving firms additional incentives (whether through liability or regulation) is to push them to overcome the information, agency, and other problems that Pete correctly identifies. It's one thing to argue that these inefficiencies and behavioral factors are important, but a different matter entirely to say that, therefore, incentives don't matter. If this strong version of Pete's argument is right, then almost no policy would decrease the risk of spills.
It's hard to tell what policy prescription he has in mind, but it may be something like a safety case requirement, which would force firms to consider risk in specified ways, early in the development process. That might be helpful, but if it's such a good idea, firms will do it on their own given sufficient incentives. And it's hard to argue that full liability for multibillion dollar spill damages is inadequate.
That doesn't mean more liability exposure would eliminate spills, but caps mean that liability isn't generating nearly the incentives it could. It's also inherently calibrated to cost-benefit tradeoffs. Some safety measures are so expensive that they aren't worth taking, even if you consider the external damages from a spill. Liability only pushes firms to take precautions they think are justified - beyond that point, they'd rather pay damages to victims.
One argument you can make against raising caps is that companies end up paying anyway despite them. If so, raising them won't push firms much. For example, Clean Water Act fines for spills are not counted against the cap, and in any case BP voluntarily waived liability caps for the Macondo spill (possibly because it believed they would be waived in court due to regulatory violations). This argument is true to some extent, but caps do still matter. Victims that try to circumvent federal caps by suing in state court, for example, may find it harder to win their cases (for a variety of reasons, including the fact that federal law imposes strict liability for spill damages while many states do not). And we shouldn't expect future spillers to waive caps.
Moreover, liability isn't just about deterrence. It's also about compensation. Statutory fines and reputational damage don't help spill victims, and they don't do much to pay for environmental cleanup. To the extent victims are being compensated in the Gulf, it's due to payments BP has made to avoid litigation, or will make in the future as a result of it. Without liability there won't be compensation, and if liability is capped, it will likely be inadequate.
