::::: : the wood : davidrobins.com

Compensating for fear: is forced Russian roulette just?

Political ·Tuesday May 24, 2011 @ 00:25 EDT (link)

Dr. Walter Block in a number of papers (e.g., Radical Libertarianism: Applying Libertarian Principles to Dealing with the Unjust Government, or Taking the Assets of Criminals to Compensate Victims of Violence: A Legal and Philosophical Approach) advocates, with Rothbard (The Ethics of Liberty), that criminals owe their victims:
  1. reparations, to make them whole (e.g., medical bills, compensation for lost work, past and future);
  2. retribution, that is, having their crime visited back on them, including death for murderers: this is fully negotiable by the victim and his agent, who may instead demand monetary compensation, incarceration, or rehabilitation (at the criminal's expense, of course).
Dr. Block also argues that the criminal must be made to suffer the same level of fear as their victim as part of the retributive justice the victim is owed. The first problem arises here: "fear" and other emotions are subjective: we cannot tell to what degree the victim was scared, and nor how to equally scare the criminal. While some may be satisfied with a reasonable approximation, that does not make it objectively valid in the same way that reparations and retribution can be (visiting the same physical invasion upon the criminal). Nor, too, does the fact that the victim and criminal may negotiate alternates hurt that objectivity: because it is the agreement that makes for equivalence, not arbitrary selection. In fact, a victim could entirely forgive a criminal and demand no retribution (or reparations, for that matter); just as Andy cannot forgive Bob for punching Chris, (absent any agency contract to the contrary) Dennis has no claim against Bob if Chris refuses to pursue it.

But if we assume that people are similar enough that what scares one person scares another, or at least that there is a retributive right to attempt to scare the criminal in the same manner that they assaulted the victim—equivalent surprise, equivalent threat of death—then we can proceed to examine what methods are just. First, one could attempt to combine the retributive harm and scaring into one and have the victim or their agent surprise the criminal with the same crime. There are problems here, though: third parties will not know that this is an act of retributive justice and not initiatory violence, and may intervene; the criminal is expecting the attack and may recognize it as such and not be scared as we hoped; or the criminal and victim may have agreed on monetary compensation for the injury (but not the fear, with the criminal claiming that it dealt no damage), so the "fear" attack will never be more than a paper tiger since the criminal knows no harm will be done to them. The criminal may be stronger and win the encounter, or may regularly travel with a bodyguard (which he scarce could be asked to forgo leaving him defenseless against other threats), or not be in a position to be attacked, robbed, etc. in the same situation (this does not prevent us from visiting the same harm upon him in a different locale, such as a security firm's office).

We might even consider that the "second tooth" ("two teeth for a tooth", being reparations and retribution) contains within it compensation for the manner of the attack and the fear inflicted upon the victim. The reparation makes the victim whole and (as best possible) makes it "as if the invasion never happened"; the retribution is the further discouragement of repetition (otherwise crime always has a positive expected value) and the just return to the criminal of his bad act, to take away what he took. But crimes differ in mode and it is different to rob someone at gunpoint in a dark alley versus anonymously slipping them a note demanding money with a vague threat for noncompliance. To ignore entirely the mode is to remove any risk premium from the first sort of attack: but how significant is that? There is a right to defend against imminent threat (gun pointed in one's direction) but what is the right to retribution after the fact in this case? The criminal will laugh if the victim points a gun back in their direction, knowing that he has no plan to pull the trigger in any case; even if a random person does it on behalf of the victim, it is such an uncommon event that the criminal would assume it was the retributive justice owed him and not be frightened by it.

Dr. Block handles the "scare" factor by requiring the criminal to play Russian roulette with himself, with the number of bullets and cylinders depending on the severity of the crime. (Presumably equivalents are fine, since one paper mentions a gun with one bullet in one of a thousand cylinders; I do not believe any such firearm exists, but it can easily be simulated with a random number generator.) The principle is that the criminal must, in retribution for the fear engendered in his victim, put himself in a situation where there is some probability p, depending on the severity of the threat, where he may be killed. Naturally this may be negotiated with the victim, either changing p or removing the requirement entirely in exchange for a consideration. As one paper mentions, a very wealthy person might laugh off the "two teeth" requirement, but be willing to part with significantly more to avoid any chance of their own death. It will be acknowledged that there is no right to inflict the death penalty on someone in return for a scare, it is presumably argued that the roulette game is not doing this because the odds of death are less than 100%.

However, I argue that this chance-of-death requirement is not a valid part of libertarian justice. I will stipulate that the correct level of fear, p, can be found, or a close enough approximation; but considering the Russian roulette claim in light of statistical mathematics, it is unjust. Statistically, we consider the expected value of the harm done, i.e., the sum of all probabilities multiplied by their values. In this case, we have a probability p (calculated by dividing the number of bullets by the number of chambers) of death and a probability (1-p) of no harm, yielding an expected value of "p death", i.e., perhaps for a robbery p is one-tenth, or 0.1 (one bullet, ten chambers), then the expected harm from the Russian roulette game the criminal will be forced to play is "one-tenth death". No amount of mere emotional trauma (i.e., excluding a scare that causes a heart attack) can be objectively translated into physical harm; the problem is much like Hume's guillotine in philosophy. While the victim may not have known if they were to live or die, the fact remains that they did not die, and so no (even partial) loss of life may be justly visited on the criminal in recompense. The best that could be done would be to ensure the criminal is on the hook for whatever psychiatric help the victim needs; but that is part of reparations, not retribution. Nor is it a defense if it is known that the criminal will always negotiate his way out of playing: the unjustified harm is still available to the victim as leverage to claim a similarly unjustified payout.

One might object that removing the Russian roulette aspect of retribution allows the very wealthy to in effect "buy" their way out of any crime. First, this is a consequentialist and not a deontological argument; and while it is in part true, allowing equivalent payment does require the consent of the victim, and as this wealthy criminal's reputation spreads, it may take more and more money for the victims to forgo equal physical retribution, depleting our wealthy criminal's fortune at a rapid rate. Enough depredations may bring our rich criminal face to face with someone who stands by the lex talionis and refuses to accept any compensation at all: and that risk (for violent crimes) is on the same order as the Russian roulette game. An extremely wealthy shoplifter (or similar nonviolent criminal), can indeed continually pay his way out, but this both enriches the merchants involved and causes very little if any fear in the first place.

A last problem is in the case where the Russian roulette game was played and the criminal died, let us say, for a holdup in a dark alley netting him $100. That is, the penalty for this robbery was death, clearly not a commensurate one; and so just as (according to Rothbard) a police officer that tortures a petty thief is guilty of the harm caused, and owes reparation and is owed retribution, those that forced the criminal to play Russian roulette and so caused his death are guilty of murder, and justice requires they die for it (as usual, depending on the expressed wishes of the criminal or his heirs)—something that should make any court very leery indeed of insisting that such a game be played.