Pat Robertson's recent suggestion — now recanted — that the U.S. government assassinate Venezuelan president Hugo Chavez led me to think a bit more about this subject. Let me pass along some tentative thoughts (especially tentative as to point 2 below), and see what the rest of you think.
1. Morality: Government-sponsored assassination is essentially an act of war; it's an attempt to affect another nation's government policy by military force. Nonetheless, if an invasion is morally justified, it seems to me that an assassination is if anything more so. It would be an odd morality that allowed the killing of enemy soldiers, many of whom are personally morally innocent, but forbade the killing of their commander-in-chief — or even ostensibly civilian leaders of the enemy government — who may be morally culpable indeed.
Idi Amin was ultimately driven from power by a Tanzanian invasion (which was prompted by a Ugandan invasion of Tanzania, but would have been eminently justified even without that). If the Tanzanians or others could have stopped Amin's murders by assassinating Amin, and without killing any Ugandan soldiers, that would have been even better.
The same goes for many other tyrants, though naturally not for every leader you dislike: Just as invasions are unjustified in most certain circumstances, so are assassinations (especially of democratic leaders, where the people's self-government as well as the leader's right to live are implicated). My point is simply that assassinations are no morally worse than other acts of war, and likely morally better than many such acts.
(I don't know enough about Hugo Chavez to express a view on which category he falls into; I mentioned Chavez only because Robertson's remarks prompted me to post this, not because I have any views about him. I also set aside for purposes of this post questions about whether and when such assassinations violate either domestic law or international law; the "morality" inquiry is about whether they're inherently wrong, not just about whether they violate the legal rules, since presumably we can change any executive orders of statutes, or withdraw from any treaties, that we think are too constraining. For a couple of past posts about one corner of the legal question, see here and here.)
[UPDATE: For whatever it's worth, George Stephanopoulos seems to take this view, too, in a Dec. 1, 1997 Newsweek article urging the assassination of Saddam Hussein.]
2. Practicality: The chief problems with assassination, it seems to me, are practical ones. First, assassination will only do so much — it will remove one person, but it may see him replaced with someone who is equally bad, or it may lead to a bloody fight for succession, which may yield more deaths of innocents than the tyrant was responsible for. Especially if the assassination is done for humanitarian reasons (which may sound odd, but as the Amin hypothetical shows, would be eminently plausible), a humanitarian would want to make sure that the act will really do more good than harm. In many (though not all) situations, an invasion is a much surer way of accomplishing your goals than an assassination would be.
Second, democracies have much more to lose from an increase in the number of foreign policy assassinations than do tyrannies. As best I can tell, foreign policy assassinations are quite rare, even among countries that are quite hostile to each other, to the point of war or near war.
This condition — perhaps a tacit understanding — is very good for democracies. Civilian leaders in democratic governments, including ones who have a great deal of power, are generally soft targets except at the very highest levels (e.g., President or Vice President): They are often seen in public, and they generally live as the people do. That's good; we want our political leaders to meet with ordinary citizens, and to live like ordinary citizens. In autocratic governments, power is generally much more centralized, and the few who have power can much more easily live in bunkers, and always be under heavy guard.
If a spate of foreign policy assassinations leads more countries (and nongovernmental groups) to adopt this tactic, tyrants could protect themselves to a considerable extent, with little effect on their ability to govern. What's more, they probably won't be much deterred by the risk of assassination, unless we can make the risk very high: They knew the job was very dangerous when they took it, and they are likely the sorts of people who can live quite well with this sort of risk.
On the other hand, politicians in democratic countries could protect themselves only through steps that would substantially change, and change for the worse, the way our democracies function. Also, quite a few democratic politicians might conclude that the job isn't worth the risk. Naturally, there will always be some who are willing to take the risk; but I'm not sure that we want that sort of self-selection effect, in which political positions are increasingly taken only by people who want them so much that they're willing to ignore mortal danger to get them.
Of course, this all presupposes that our attempts at assassinating others will lead others to assasinate us; and this may be a mistaken presupposition, because I suspect that most of our enemies and potential enemies aren't exactly animated by a sense of fair play here. Yet the fact is that, despite the softness of many of our targets (again, not the President or Vice President but many other important leaders), our enemies — even in past shooting wars — have not generally gone after them. Likely this was often because they feared retaliation on our part. But I suspect that there was a bit of a tacit deal involved there, and if I'm right, then some violations of the deal could lead the whole deal to break down. (And, yes, I know that there have been violations of the deal already; but sometimes such deals survive a few breaches, but fall apart once a tipping point is reached.)
This having been said, there will of course be exceptions. Hitler was an unusual enough figure even within Germany, and the worth of killing him was so great, that assassinating him would have been very much worthwhile (though also very hard, as many would-be German assassins learned). When we're fighting a broader war, the killing of high military and quasi-military officers may be necessary, and might not materially increase the risk of retaliation beyond what it would be in any case. Still, it seems to me that if you look at the big picture, the seemingly cheap step of assassination ("just take him out") may generally be much more expensive than it appears.
In any case, these are just some tentative thoughts — I'm most surely no expert on the question — and I'd love to hear what others think.
UPDATE: I use the term "assassinate" because I don't want to sugarcoat what would be happening — the deliberate killing of a particular person. But if you think that "assassinate" inherently carries a connotation of improper deliberate killing, just mentally replace the use of "assassinate" in my post with "targeted killing."
Dictionary definitions of the term in fact reflect this ambiguity in the term. Some define it using the term "murder," which generally means kill immorally or illegally; others define it using the term "kill," which doesn't carry that meaning. I generally lean towards the latter category: A deliberate attempt to kill Hitler would in my view be commonly called "assassination," though it would of course be morally praiseworthy. And I think that this is a common enough use of the term that saying "targeted killing" instead would make it sound like I was pussyfooting around the subject. But, again, if you think that assassination definitionally can't be permissible, just use the term targeted killing instead.
FURTHER UPDATE: Professor Bainbridge has a Catholic perspective on the issue.