Monday, November 17, 2008
Creeping consequentialism and insidious economics, part I
OK, the title is hyperbolic. But here is what troubles me.
You are teaching or analyzing legal rule A. Suppose rule A seems to express an underlying standard of undesirable or impermissible conduct C. For example, the legal rule is: "D must pay compensatory damages to P when D's unreasonably risky conduct harms P." The underlined phrase expresses a norm of impermissible conduct.
Unthinking consequentialism can creep in at both levels of this analysis--in your analysis of what counts as creating an unreasonable risk to others, and in your analysis of whether a legal rule expressing this standard of conduct is normatively desirable. This danger is especially pronounced in analyzing tort doctrine, but it arises elsewhere, too.
Consider first the standard of conduct. How do we decide what risks are unreasonable? (For example, whether it is unreasonable for D to speed to the hospital to obtain care for his sick child.) Here is an answer, and one that appears so self-evidently correct that any alternative analysis seems irrational.
1. Consider all the bad consequences of D's taking the risk.
2. Consider all the good consequences of D's taking the risk.
3. If the good outweighs the bad, it is permissible to take the risk. If not, not.
(The famous Learned Hand test is often viewed as expressing this approach: you are negligent if but only if the burden of taking a precaution against the risk is less than the benefits of taking the precaution, in the form of reduced risks of harm to those exposed to the risk.)
If you don't like this approach, what is wrong with you? Don't you care about consequences? Don't you prefer good to bad? (Or, under marginal versions of this approach, don't you prefer more good to less good, and less bad to more bad?) Are you fanatically opposed to balancing? Even though, in your own life, you routinely balance the advantages and disadvantages when deciding between option A and option B?
Consider next the desirability of the corresponding legal rule: if an actor fails to act as specified by this standard of reasonable conduct, he must pay compensatory damages to the victim. How do we decide the desirability of this rule? Why, in exactly the same manner. This is a matter of basic rationality, after all.
1. Consider all the bad consequences of enforcing this legal rule of negligence liability.
2. Consider all the good consequences of enforcing the rule.
3. If the good outweighs the bad, this is a desirable legal rule. If not, not.
Again, the logic appears to be inexorable.
This approach is enormously attractive for a number of reasons. It appears to be thorough, empirically grounded, pragmatic, rational, and focused on human welfare. It gives us a rigorous method, quite unlike the fuzzy soft-headed intuitions of those who defend rights-based and other nonconsequentialist approaches. (A colleague in the philosophy department recently remarked to me that many of her undergraduates, when first introduced to philosophy, are seduced by the simplicity and apparent inevitability of utilitarian thinking.)
I do believe that economic and utilitarian analysis have value, both descriptively and normatively. But the account I have just given greatly understates the difficulties with these forms of analysis, and overstates their ability to describe the world accurately and to prescribe norms that we should live by. (To be continued...)
TrackBack URL for this entry:
Listed below are links to weblogs that reference Creeping consequentialism and insidious economics, part I:
I'm confused. Isn't this exactly the discussion that takes place in every first semester torts class on the day the Learned Hand formula is introduced? The formula is one possible way of assessing what's "reasonable," explains the professor, and then invites students to debate whether it's the only way (no), the way used by the law (sometimes but not always), or the right way (long, long discussion follows).
Posted by: James Grimmelmann | Nov 17, 2008 3:09:01 PM
Its an interesting observation. I wonder whether part of the attraction of your consequentialist formulations has to do with eliding the distinction between "consequences" and "reasons," for of course one should consider the reasons for and against some course of action (or some rule), but the consequences are at best only one subset of the reasons one could marshal for or against some particular conduct or rule. I say "one subset" because there may be non-outcome-based reasons also. And I say "at best" because there are clearly instances in which some of the reasons on one side of the ledger preempt/exclude/silence certain kinds of reasons on the other side of the ledger (or indeed on the same side) and do not merely outweigh them -- a fact that Raz has made famous, but which has been seconded by many others, notably Scanlon. So again, evaluating conduct or rules in virtue of the reasons for and against is at least as intuitive as evaluating them in virtue of their consequences, but it also strikes me as a better way of framing our thought because it preserves the substantive and open question of what kinds of reasons there are, which is to say that it doesn't stack the deck in favor of consequentialism.
Posted by: John Oberdiek | Nov 17, 2008 10:02:16 PM
I don't get the implication in the thread that "rights-based" approaches are "non-outcome-based."
To put it simply, if you're concerned about someone's "rights," and you're against a rule because of your view of those rights, wouldn't your aversion for the rule exist because of its *consequences* on those rights? Perhaps I'm conflating "consequences" and "reasons" like Mr. Oberdiek alluded to above; but is it possible to take a 'reason' for disfavoring a rule that isn't based on any 'consequence' of that rule? If there's no consequence then nothing has changed...right?
Posted by: Aaron Williams | Nov 18, 2008 11:40:55 AM
I think John Oberdiek hits it on the nose. The key move is this one:
"Don't you care about consequences? Don't you prefer good to bad? (Or, under marginal versions of this approach, don't you prefer more good to less good, and less bad to more bad?)."
It's far from clear that preferring good to bad just is preferring good consequences to bad consequences. Curious to see where this goes, though...
Posted by: jsd | Nov 18, 2008 1:16:34 PM