Torture Versus Death and the Greater/Lesser Problem

By Mike Dorf

My new column on Verdict asks whether President Obama's targeted-killing policy is worse than former President Bush's torture policy.  I use as my point of departure a recent Wall Street Journal op-ed by John Yoo, who makes that claim.  (I don't link to the op-ed because the WSJ remains behind a pay wall.)  I then explore the intuition behind the claim.  As Jane Mayer acknowledges in a New Yorker piece, "it’s better to be alive with no fingernails than dead."  Nonetheless, I end up agreeing with Mayer that Yoo is wrong--which is not to say there aren't legitimate grounds on which to criticize the targeted killing policy.

Here I want to explore the broader logic underlying Professor Yoo's claim. That broader logic is sometimes captured by the proposition that the greater power to do X includes the lesser power to do Y.

Let me make that more concrete by giving a famous example.  In the 1892 case of McAuliffe v. New Bedford, Oliver Wendell Holmes, Jr., writing for the Massachusetts Supreme Judicial Court, rejected the free speech claim of a petitioner who had been dismissed from his job as a police officer for violating a rule forbidding various external political activities.  Holmes wrote that "[t]he petitioner may have a constitutional right to talk politics, but he has no constitutional right to be a policeman."  The core idea is that because the city didn't have to hire McAuliffe at all, it could condition his employment on his complying with rules restricting his political speech.  The greater power not to employ McAuliffe includes the lesser power on the condition that he limit his political speech, Holmes reasoned.

Modern First Amendment case law rejects the greater-includes-the-lesser logic of McAuliffe.  Under the modern employee speech doctrine, government employees retain their free speech rights, although the necessities of running a workplace allow the government greater regulatory leeway with respect to its employees than it enjoys in its capacity as regulator for the society as a whole.  Still, the core point stands.  Rejecting the Holmesian logic, the Court has made clear that while the government indeed does have the power not to hire particular indiiduals, once it enters the hiring and firing business, people retain free speech rights.

The application of that principle may not be self-evident in a case like McAuliffe itself.  After all, that case turned on a New Bedford rule that forbade political canvassing by police officers.  Even under the modern doctrine, McAuliffe might have lost his case because such a prohibition appears justified on an anti-corruption/anti-extortion rationale.  The government could have legitimately worried that citizens would perceive a tacit threat behind political canvassing by a police officer, even if he were not in uniform. So Holmes's principle was overly broad but perhaps he was right about the particular outcome.

We can see the overbreadth of the Holmesian principle more clearly in a hypothetical case.  Suppose that a city owns a vacant lot.  If the city sells the lot to a private developer, the developer may build private homes and sell them to purchasers who may then exclude the public, including members of the public who want to gather for expressive purposes.  However, if the city turns the lot into a public park, then it may not exclude people from the park because they want to give political speeches or hold rallies, except insofar as the case law developing the time, place and manner rules allows.  The city's greater power to sell the lot to a private developer and thus extinguish any expressive rights of the public does not include the lesser power to retain the land as a park and censor speech.

What about torture and killing?  Once we have developed a healthy skepticism towards greater/lesser arguments, it's easy to see what's wrong with the claim that torture is allowed whenever killing would be allowed on the ground that killing is worse than torture.  For one thing, killing isn't always worse than torture.  It's not uncommon for torture victims to wish for death during torture.  But let's put that aside and imagine that for some set of persons and circumstances, death is worse than torture.  It still doesn't follow that the power to kill implies the power to torture.

Suppose that the U.S. army in wartime comes upon an encampment of enemy soldiers.  Under the law of war, the U.S. army can engage in a surprise attack, killing all of the enemy soldiers.  But let's suppose that instead of doing so, the commander of U.S. forces orders that the U.S. troops encircle the enemy soldiers.  They do so, whereupon the enemy soldiers surrender and are taken prisoner.

Can the U.S. now torture the enemy captives?  Of course not.  And that would be true even if, ex ante, each of the enemy soldiers would prefer to be captured and tortured than to be killed in a surprise attack.  Put differently, torture is not a lesser act than killing, or at least it's not only a lesser act than killing; torture is also a different act.

Does this mean that greater-includes-the-lesser arguments never work?  No.  I think they are useful starting points and can serve as a check on the consistency of rules, especially for policy makers with finite enforcement resources.  For example, if the government forbids some substance X in a product but permits a more harmful substance Y, then that should trigger an inquiry into whether that juxtaposition makes sense.  Perhaps it can be justified on the ground that there are ready substitutes for X but not for Y, or on the ground that while Y is more harmful than X, Y also has compensating benefits that X lacks.  Etc.

So comparisons of "greater" and "lesser" harms are relevant to normative reasoning.  But they should not be conversation stoppers.