Last week was my baptism as a blogger. The astuteness and vigor of many of the reactions I got seemed to me dwarf even the most successful academic workshop presentations I have had. To be sure, in bluntness too the reactions dwarfed what I am used from a workshop presentation, and did occasionally make me wince, but the tradeoff seems well worth it.
It was astonishing to receive not just ingenious hypothetical variations that I had never thought of (like someone’s suggestion of a bargain in which the prisoner’s brother agrees to serve his term for him) but also referrals to books, newspaper accounts and internet sites I did not know anything about.
Some readers obviously were annoyed by my habit of closing the first four posts of the week with what I thought was no more than a teasing invitation to look at the book on which they were based. They felt badgered. If you are one of those readers, I apologize. That’s a problem for us novice contributors to the blogosphere: we don’t yet have a sense of how something comes across. Next time I’ll know better.
Obviously I can’t respond in detail to the plethora of interesting points raised. I will just have to limit myself to those I judge to be of the widest possible interest.
Some Last Words on Loopholes:
Many readers rightly remained puzzled how exactly I conceive of the analogy between killer amendments and loopholes. Perhaps I erred in trying to condense the gist of the analogy within the confines of a post. Nailing it down actually takes me a fair number of pages in the book itself. But let me try to give more of a glimpse of how the analogy works.
Killer amendments illustrate a ubiquitous feature of voting rules: a kind of context dependence of choices. The choice between two alternatives can often be influenced by injecting a third alternative into the mix, or by changing voters’ thinking about some third alternative that is already in the mix.
That this should be so is quite bizarre. It means that collectivities routinely behave in the way one reader attributed to Sydney Morgenbesser: Facing the choice between apple and blueberry pie at his local diner, he opted for the apple. When subsequently informed that there was also some cherry pie available, he switched his choice to the blueberry pie.
Very crudely put, one might say that the essence of Arrow’s famous impossibility theorem is the demonstration that it is impossible to design a plausible voting system in which the collectivity does not behave somewhat like Sydney Morgenbesser.
Something similar is true even when we are not talking about collectivities, but single individuals who are ranking a set of alternatives (such as several cars they are considering) on the basis of several relevant criteria, like price, looks, safety and so on. This is something Arrow himself proved in a book on multi-criterial decision making in which he extended his insights about voting to multi-criterial decision making situations. His claim was that all individual decision-makers who rely on more than one criterion to rank the alternatives they are considering are doomed to behave exactly like Sydney Morgenbesser at times.
Legal doctrines can be thought about as being a form of multi-criterial decision-making and therefore should exhibit this kind of context dependence. And when they do, it’s what we often perceive as a loophole.
Let me illustrate this more concretely with the help of the immigration loophole I described in my first post, in which a visitor from a third world country makes a statement that turns him into persona non grata back home so that he can now claim political asylum in the US.
Before the would-be immigrant makes his provocative statement, we can think of him as having three options we can rank as to their permissibility: (1) staying in the US and speaking or not speaking out against his home country, (2)going home and risking persecution by speaking out against his home country, and (3)going home, keeping quiet and living unmolested. At the outset, (2) and (3) will be ranked in the “permissible” category, (1)will be ranked in the impermissible category. By making his provocative statement while in the US, the visitor has eliminated (3) from the choice set: he can no longer go home and expect to live in peace. And now the relative order of (1) and (2) has been changed. Both are now in the top category, permissible conduct. Making the provocative statement thus becomes the equivalent of injecting a killer amendment into the mix of legislative choices.
The context sensitivity here arises not from the presence of a plurality of voters but a plurality of criteria, or principles that are reflected in our immigration law. These include (a) the desire not to admit the entire world into our country, and (b)the desire not to deliver someone into the into the hands of a death squad or a concentration camp. Each of these desiderata would rank the three alternatives differently. When we combine them by giving weight to each, we end up with a context-sensitive procedure, in which the absence of presence of a certain alternative necessarily affects the relative ordering of the remaining two.
Put differently anyone applying a legal doctrine cannot avoid behaving at least occasionally like Sydney Morgenbesser. The opportunity to get a legal decision maker to switch from apply to blueberry pie by introducing cherry pie onto the menu is what we consider to be loophole exploitation.
Next, some last words on why the law is so either/or:
Many readers took issue with my characterization of law as either/or. They offered compelling counterexamples: sentencing, damages, the doctrine of promissory estoppel, and many others in which the law seems to operate on a graduated, or continuous, rather than a binary basis. They are absolutely right. If I had been more precise, I should have asked not why the law is so either/or but why it is predominantly either/or, in situations in which it seems it would be more natural not to be.
The answer I offered to that question seemed to convince just about no one. My sense is that that may have had a lot to do with the particular example I chose, the transition from life to death. I have the feeling, however, that many might find the argument quite a bit more compelling if I did nothing more than transpose it from the end-of-life setting to the beginning of life. Here is how it would then look.
Imagine we have a choice to make between aiding one neonate or two infants one day away from delivery.
Presumably the latter win.
Now compare two one-day-away-from-being-born infants with three two-days-away-from-being-born infants, and presumably the latter win out. In other words, it is better to save three two-days-away-from-being-born infants than one neonate.
Keep going like this and you end up with the conclusion that it is better to save 281 280-days-away-from-being-born “infants” than one neonate.
But that’s something not many people would endorse. So what went wrong? Where in the chain of inferences was there a mistake?
It would seem to have to be the case that there is some number at which we can no longer go back one day and make up for it by increasing the number of fetuses to be saved by one. (Nothing would change if we doubled or tripled or whatever the number of fetuses every time we went back another day.) So there is a decisive line somewhere such that however many creatures we have on one side of that line (say day 93-before-birth) they will always be worth less than even one creature on the other side of that line.
That line would seem to be the line at which human life begins.
The bizarre thing of course is that there is no particular reason to designate a particular day as being the one on which this happens. Although we know that there must be such a day, we do not have the slightest reason to pick a particular day as being the one. And it is hard to believe that there is some factual knowledge out there that could possibly lead us eventually “find” what that day is. It is this last fact that makes people naturally conclude that there cannot be such a line. And yet the logic of the foregoing argument strongly suggests that there must be.
Lastly, some last words on voluntary torture:
Many readers were disturbed by my focus on deterrence and retribution as the principal aims of punishment. What about incapacitation and rehabilitation? I felt free to simplify my life by putting them aside because I have a hard time believing that what really disturbs us about voluntary torture has to do with the absence of those two things.
Others were puzzled by my designating it as a win-win situation. I am being a little imprecise here. It seems like a win for society, in that deterrence and retribution are achieved at a fraction of the price. (Ignoring incapacitation and rehabilitation for now.) While it is also a marginal win for the prisoner, it is not clear that that should count, since we don’t really view a reduction of deserved punishment as an improvement. But even if we regard it as a loss rather than a win, it is so marginal a loss, that it hardly counts.
Others yet were annoyed by my insufficient appreciation of the inherently degrading character of torture that makes consent in their eyes irrelevant. It’s true I find it hard to accept the argument about degradation just like that, when we are dealing with something that benefits society at large and is actually desired by its “victim”. On the other hand, the answer I ultimately offered, which has to do with implicit valuations of life and limb in various legal rules, which make it difficult to give people complete discretion about the bargains they want to strike could be seen as giving analytical substance to the argument about degradation.