Bankslaughter, Tort Law, and Optimal Deterrence

Felix Salmon has been helping popularize Paul Collier’s idea of bankslaughter. (No, it’s not what you wish it were.) The idea is that there would be a crime called bankslaughter, or “managing a bank irresponsibly.” If a bank blows up, there could be a criminal investigation to determine if the bank managers behaved recklessly (more on that term later); if so, they would be convicted. The analogy is to manslaughter, which is actually a family of crimes; Collier probably means criminally negligent homicide, or causing death through negligent or reckless (more on those terms later) behavior.

Not surprisingly, the conservatives are not happy about this, even though it seems to conform to the conservative principle that people should bear responsibility for the consequences of their actions. (Or maybe that only applies if you are a pregnant teenager.) Salmon cites John Carney, who calls bankslaughter “the worst idea of the week.”

Here are some of his objections:

Collier doesn’t seem to have given much thought to the costs of over-deterrence. Bank executives faced with the prospect of a criminal investigation and possible conviction would likely be overly cautious. We’d lose a lot of socially beneficially risk taking by criminalizing bank failure.

There’s also a serious fairness issue. Only those executives whose risky bets blow up get investigated, prosecuted and punished. Those whose bets pay off are untouched. This means that being unlucky in the markets becomes a criminal matter.  Criminality becomes a kind of lottery. . . .

Because bankslaughter is backward looking but conducting business is forward looking, it would almost certainly result in wrongful convictions. Lots of activity that looks reckless after the fact can seem perfectly sensible ahead of time. Unless the crime required bankers to know they were being reckless—in which case it would deter almost no-one and result in approximately zero convictions—it would wind up punishing bankers for just being wrong.

Regular readers can guess what I think of the idea of “over-deterrence.” We need some more over-deterrence. People can talk all they want about “socially beneficial risk taking,” but the evidence that this happened during the last thirty years is pretty contestable. For a system to have the optimal amount of risk taking, it is necessary (but not necessarily sufficient) that the people who stand to benefit from the venture internalize the risk of failure in some way; as everyone on the Internet has written multiple times, that condition did not hold for the financial sector in all sorts of ways.

But the more interesting argument is whether there is a problem with retroactively holding people liable for harm that they cause unintentionally. Carney writes as if this is just a crazy idea: first, only the people who actually cause harm are prosecuted, as opposed to others who behave equally egregiously but are lucky enough not to cause harm;  second, it allows juries to evaluate behavior in the past with the benefit of hindsight.

What he doesn’t say is that this is exactly how a huge chunk of our legal system works today. It’s called torts, and even though I’ve only had one year of law school, and that was at Yale, where the joke is that you don’t actually learn any law, I know something about it. The general principle is that in most spheres of life, if you act negligently – defined to mean that you do not exercise the same degree of care a reasonable person would under the circumstances – and your action causes injury to someone else (to whom you owe a duty of care), you are liable for that injury. In some areas, the threshold is lower; for example, in product liability, the rule of strict liability applies, which means that you don’t even need to be negligent; if your product hurts someone, you’re liable. In other areas, the threshold is higher, and the requirement is not just negligence, but gross negligence, or willful wanton recklessness, or something like that. But that’s the basic principle; you don’t need intent, and the legal system certainly does assign liability after the fact.

So, for example, if you take a turn too fast on a wet road and you hit someone, you’re liable; but the twenty drivers ahead of you who all took the same turn too fast aren’t liable for anything. And juries decide whether you were being negligent with the benefit of hindsight; they know that the baby stroller was in the intersection, which you had no reason to know when you took the turn (and was highly unlikely, since it was late at night and raining). That’s just the way the law works.

Ironically enough, the modern tort system is a product of the second industrial revolution of the late nineteenth century, and was designed to cater to the needs of large businesses. By creating an objective, supposedly predictable and stable standard of negligence, it made it easier for businesses to manage the risks of their operations. And while there are certainly things they criticize, tort law is one of the main areas where the law-and-economics crowd has won out, and real judges actually talk about things like “cheapest cost avoiders.” For many scholars and jurists, the standard of reasonable care is defined as the degree of care such that the marginal benefit of accident avoidance equals the marginal cost of care; you can see how, in a tautological way, this creates a Panglossian “best of all possible worlds,” since it yields the perfect degree of deterrence. And the reason most economist types (and free marketers) like this way of thinking about tort law is that the whole point of the law is to create the right incentives without the need for all sorts of detailed regulation.

So Carney’s idea that this type of liability would produce over-deterrence is a bit curious. To get optimal deterrence in the bank-management context, you want managers to take precautions (e.g., maintain capital reserves) up to the point where the marginal reduction in the risk of a collapse is balanced by the marginal cost of those precautions. Today, we have no such thing, since we have no meaningful risk of collapse for bank managers, since their downside (relative to their upside) is limited by (a) leverage, (b) the implicit government guarantee, (c) the bonuses they got in the good years, and (d) the “resume put.”  Because there is no liability for blowing up a bank, our legal system provides no deterrence whatsoever, which is one reason why we need regulation. (In a perfect law-and-economics world, you wouldn’t need regulators, since the threat of liability would create optimal behavior all by itself.)

Now, I don’t think that Collier has all the details right. First, I think that criminal prosecution for mere negligence is overkill; civil liability would be plenty, since that way you could go after all of the negligent manager’s assets, including his bonuses from the good years. We could save criminal prosecutions for people who commit actual fraud. Second, Collier uses “recklessness” and “reasonableness” as perfect opposites, which, from a legal standpoint, they’re not; you can fail to behave as a reasonable person would (that’s negligence) without your conduct rising to the level of recklessness.

The thing standing in the way of a civil version of bankslaughter is the business judgment rule, which basically says that if you are a manager who makes an informed decision in a situation where you do not have a conflict of interest, you are not liable, period, no matter how risky or stupid that decision is. And as far as I know, no court has ever held that having a bonus tied to your company’s stock price, with no commensurate downside risk, counts as a conflict of interest (even though it is). The point of the business judgment rule was to enable managers to take risks without fear of liability (or to enable rich people to sit on boards of directors and become even richer doing very little work without fear of liability). But in an environment where it’s very clear that managers were taking too many risks, because there was no way they would suffer the potential consequences of those risks, one very logical and sensible solution is to turn the dial partially back the other way and increase liability, which I see as Collier’s basic point. Since the business judgment rule is basically common law, it could be pared back by statute, and it could be pared back selectively – for example, for businesses whose blowups can have large societal costs.

One weakness of the Obama Administrations financial regulation proposal is that it doesn’t increase real costs for the key actors – bank managers and directors. Instead, it relies on better regulation enforced by better regulators. A civil version of bankslaughter could help fix this problem.

• Become A Funded Futures Trader