Preemptive Compensation

[Cross-posted on the parent blog]

One question I've been thinking about recently is the idea of preemptive compensation; that is, compensation paid in advance of a particular damaging event actually happening. I wanted to broach the subject here, as a starting point for future discussion. I'll start by sketching why the idea of preemptive compensation is important at all.

Let's say that Betsy launches a rocket into space, and calculates that it will eventually return to the Earth and land on Murray's house, but that this will take 10 years. Betsy acknowledges that when the rocket returns to Earth, it will damage Murray's house, and she should be held accountable for that damage. But sadly, Betsy is quite old, and doesn't expect to live anywhere near long enough to see the impact.

Murray considers the situation in which he finds himself, and sees a problem looming. He anticipates that Betsy will spend all of her savings by the time she dies (Betsy is willing to acknowledge that this is likely to be the case), and if he waits until the rocket lands on his house, there will be no way to do anything about it. Betsy won't be around anymore, and there will be no estate from which he can demand compensation.

It seems to me that Murray would be right to demand that he be guaranteed compensation for the future damage which will be dealt to him. To require that Murray wait until the damage eventuates would be to deny him compensation entirely. And given that we already know what will happen to Murray, and we agree that Betsy ought to be held accountable for what happens to him, it seems unfair that we should deny Murray restitution for the damage that he will surely be dealt.

One way to approach this would be to say that Murray should only be entitled to compensation when the harm eventuates, and so Betsy is only responsible for ensuring that when the rocket lands on his house, proper arrangements have been made to compensate Murray. So, for example, if there will be $50,000 of damages that Murray can legitimately claim (including the damage to his house, as well as any compensation for the inconvenience), Betsy only needs to ensure that when the rocket lands, there is $50,000 available to compensate him. Betsy wouldn't have to pay Murray the $50,000 immediately, but would only have to guarantee that it would be given to him when the rocket landed. Because it might be difficult to administer such a rule, it might be fair for the courts to require that Betsy pay Murray the present value of the damage in the future, or to require that Betsy secure a no-risk, interest-bearing investment on Murray's behalf which would pay the $50,000 at the appropriate time. This would ensure that Murray was compensated, but also that Betsy was not deprived of more than we might think she ought to be.

But in reaching this conclusion, we have not mentioned what is undoubtedly the most important consideration in this problem, which is uncertainty. We have spoken as if we knew exactly what would happen to Murray 10 years after Betsy launched the rocket, when it landed on his house and damaged it. It's quite probable that in any case like this one, we won't actually know what will happen to Murray's house. Perhaps there is a range of levels of damage we can conceive of as being possible scenarios for when Betsy's rocket hits Murray's house. What then?

One possible way to approach this problem would be to gather more information. But it might not be possible to collect the necessary information before it's too late. As Murray is unable to wait to see what the damage actually is, it could be that he has no real choice but to go to court with considerable uncertainty over the consequences of Betsy's action. Perhaps he can demonstrate that the actual damage will, with a high level of confidence, be within the range of $40,000-60,000, with a "best guess" of $50,000. What should Betsy be held accountable for?

One answer would be to say that since Murray can't prove the precise nature of Betsy's offense, that he fails to meet the burden of proof which would be required to hold Betsy responsible at all. But this seems clearly wrong. We agree that Betsy's actions will cause damage to Murray, and we agree that we won't know the precise nature of the damage until it's too late. It seems patently unfair to completely deny Murray's case on the basis of him lacking knowledge of the exact amount of the damage.

Another route would be to come down on Murray's side, and to say that he should be entitled to the maximum amount of damage that could reasonably be expected. The reasoning behind this would be that Betsy is the one who's at fault here; if there's any uncertainty, the burden should fall on her. After all, she's the one who launched the rocket, not Murray; it would be a real shame if the rocket landed and ended up doing more damage than Murray had been compensated for, especially if we knew that there was a reasonable chance that things would turn out the way they did.

I can definitely see why someone would want to argue this way, and I don't think that they'd be wrong to do so. But I could also see why someone would object to this treatment. One could point to that fact that libertarians have generally taken the position that it's somehow worse to let an existing damage go unrepaired than to punish an innocent person for something for which they are not responsible. As Murray Rothbard wrote in his essay, "Law, Property Rights, and Air Pollution," when "...it is unclear whether an individual is committing aggression...the only procedure consonant with libertarian principle is to do nothing; to lean over backwards to ensure that the judicial agency is not coercing an innocent man." But of course, in our example, Betsy is not innocent. In fact, that's the entire intuition behind making her bear the burden of the potential harm.

But the fact remains that if the rocket struck and did less damage than the maximum amount that could have been reasonable expected, then it would mean that Betsy would have been punished for a degree of damage which she did not cause. And because we're talking about the largest reasonably expectable damage, it's extremely likely that this sort of overpunishment will happen under this policy. So for people who believe that we should be protected from being punished more than we deserve, the policy in question here would be extremely troublesome.

Accordingly, we might place the burden of uncertainty on Murray, and say that he is entitled to compensation for the lowest amount of damage that could reasonably be expected. The argument here would be that the burden of proof is on the victim to show that someone harmed them, and the mere possibility of greater damage doesn't satisfy this burden of proof. The lowest amount of damage that could reasonably be expected would constitute the sort of thing that could be looked at as "proven" in court, but moving beyond that would seem to indict Betsy on "the possibility" that she is guilty.

It should be noted that the same sort of problem that confronted the last policy applies to this one, except in reverse: In all likelihood, Murray will be compensated for less damage than he will actually incur. This might make people uncomfortable, for the simple fact that Murray's the victim here. It could seem like we're going out of our way to protect the person who did something wrong, and chose to do it, very likely at the expense of the person who did nothing wrong, and in fact did nothing at all. So this policy is somewhat questionable as well.

A final approach would be to hold Betsy liable for the amount of damage representing the scenario where the chances of more damage eventuating are the same as the chances of less damage eventuating. The core virtue of this policy is obvious: it represents a compromise, placing an equal portion of the burden of the uncertainty on both parties. But of course, in the eyes of many, this virtue will also be its main weakness: by placing half of the burden on Murray, it ensures a good chance that he will be undercompensated, and by placing half on Betsy, it ensures a good chance that she will be overcharged.

It seems to me that this discussion highlights a very simple point: what will appeal to you as the correct way to deal with uncertainty will depend on your views regarding the purpose of the burden of proof. If the spirit of the burden of proof is to ensure that innocent individuals are protected, then you will likely feel that the victim -- who is the only innocent party -- should bear less of the burden of uncertainty. On the other hand, if you view the burden of proof as protecting individuals from having to pay for damage that they have not been proven to have caused, then you will likely come down on the side of placing the burden on the victim. I will not take a stance on this issue here, since I honestly don't know what to think.

Before calling it quits on this subject, I want to touch on one more potential problem for Murray's case against Betsy. Namely, Murray might not be able to prove that Betsy's rocket will hit his house at all. It seems to me that uncertainty of this sort is very different from the kind of uncertainty discussed above. It is uncertainty regarding whether or not the harmful event will occur at all, rather than uncertainty regarding how severe the damage caused by the harmful event will be. On this issue, I think our current system has the tools to tell us what should be done. That is, our normal ideas regarding the burden of proof seem more or less adequate for dealing with cases like Murray's. Perhaps we might want to allow for a slightly more relaxed standard, given that our current standard might create a systemic bias towards placing burdens on victims, but that seems like something that can be discussed without breaking new ground.

So I guess I haven't really reached any firm conclusions about preemptive compensation in this post, but I think that as far as providing a starting point for discussion, I'm happy with this. It seems like the next step needs to be a discussion of what the proper role of the burden of proof ought to be. But I think I'll stop for now and deal with that later.