Robert Kelley has done a series of interesting posts on his own blog (cross-posted to Duck of Minerva) exploring options for U.S. retrenchment and offering a template for thinking about U.S. alliance commitments. Consider what follows a set of variations on the theme he began.
Kelley asks: if U.S. leaders tried to pursue a policy of partial retrenchment, what alliances commitments might they choose to limit or terminate, and which allies would still be considered important? Framing the question this way acknowledges that there may be some reputational issues involved in downgrading a long-standing security partnership, even if its original strategic rationale has diminished or even disappeared. But what if we let our imaginations really run free and frame the puzzle a bit differently? What if we were starting from scratch, and doing a "zero-based" assessment of U.S. alliance options? If historical ties weren't an issue, what features would you look for in a strategic partner and how might America's future alliance portfolio differ from its current set of arrangements?
So here's my quick list of the qualities we ought to look for, notwithstanding some obvious tensions and tradeoffs between them. As you'd expect, I lean heavily on more-or-less realist considerations, and less on shared "values" or domestic political similarities.
1. Power: Up to a point, you want allies that are strong and capable so that they don't need a lot of protection from the United States and so they can make a real contribution to any necessary effort at collective defense. One of the reasons the US won the Cold War is that our alliance system contained a lot of wealthy and relatively powerful states, while the Soviet alliance system contained a lot of relatively weak and not-very-powerful clients. One can't take this logic too far (i.e., "concerts of great powers" usually don't work well because the strongest states are too worried about each other to be close allies), but on the whole, you'd prefer allies that can actually do something for you. (One might argue that this strengthens the case for NATO and the U.S.-Japan relationship, but not if these states continue to let their defense capabilities atrophy.)
2. Position: There are some allies who are valuable not because they have a lot of capabilities, but because they happen to sit in an especially valuable piece of real estate. Think Oman, or Singapore, for example, which sit right next to critical strategic waterways. If you define your interests in global terms, then you're going to need some allies in these places.
3. Political stability: On balance, you'd like to have allies whose governments are stable and legitimate, so that they can make effective decisions and so that you don't have to constantly worry that they might be overthrown. Unstable allies encourage adversaries to meddle in the hope of undermining them, or forcing you to spend a lot of time worrying about your allies' nternal political health.
This is not necessarily an argument for democracy, by the way, because a democracy that rests on unstable coalitions or where there are sharp divisions about foreign policy can be a pretty troublesome partner. For that matter, there's no guarantee that public opinion will always support the alliance (which is why some Americans worry about how Arab spring may ultimately affect U.S. relations with some traditional Middle East partners). But on the whole, a stable and legitimate ally is preferable to an unstable or fragile one.
4. Popularity: An ally that has few conflicts with other states, and that has a positive image in the world is less trouble than an ally that is unpopular or a pariah. The reason is obvious: if you join forces with a state that other countries resent or despite, you immediately pay a diplomatic price for your association and you may end up gaining more enemies than friends. Other things being equal, this is not smart. America's "special relationship" with Israel illustrates this problem perfectly, just as China pays a price for doing business with Sudan and that Russia is losing prestige by continuing to support the Assad regime in Syria. This concern can be ignored if the price is not too high or if other benefits are large, but on the whole, you want to be friends with countries that have lots of other friends too.
5. Pliability: The ideal ally is also easily influenced: you'd like to have partners who will do what you want at most of the time. In simple terms, you want allies whose interests are mostly compatible with your own (duh!). An ally that refuses to help when times are tough, that has to be constantly badgered into contributing its fair share, or that takes independent actions even when it knows that this will cause trouble for its partners, is more of a headache than an ally that usually does what you want. (This problem explains why U.S. relations with Pakistan are in such bad shape: both sides are deeply disappointed by what the other is doing). No two states have identical interests, of course, and any alliance will exhibit occasional strains. But on the whole, you want allies that are genuinely working with you, instead of at cross-purposes.
6. Potential impact: Finally, sometimes states form an alliance simply what happens to some other state could have an enormous effect on what happens to them. Neither Canada nor Mexico are major military powers, for example, and neither controls key strategic chokepoints. But their proximity to the United States also means that what happens there could have an enormous impact on U.S. security. To take this one step further: if either country were ever to align with a power that was hostile to the United States, the potential impact on American security would be enormous. So the United States has a powerful interest in keeping Canada and Mexico close, despite their relative military weakness.
As I suggested above, there are some interesting tradeoffs between these different criteria (which is one reason why diplomacy and grand strategy can't be reduced to a simple checklist or cookbook). You want strong allies, for example, but the more capable an ally is the less pliable it is likely to be. The United States generally likes allying with democracies, but democratic governments don't always act the way Washington wants (see under: Turkey). Ideally, you want you allies that are popular and do not have a lot of enemies (because that makes it harder to help protect them), but if they have a few enemies they will be more interested in your help and more willing to defer to your wishes.
So if we were really doing a "zero-based" alliance portfolio, what implications might one draw from this set of criteria? If one could really start from scratch, I doubt we'd give security guarantees to Taiwan, even in a period when we're worrying about the Asian balance of power. It's too small, and will be increasingly difficult to protect over time. But most of America's other Asian allies would still be valuable, and we'd probably be courting them today even if we didn't already have strong ties. As noted, one could make a case for NATO as a limited security partnership, but I doubt we'd try to build an elaborate multilateral institution in 2012 if it didn't already exist.
The United States would still need allies to maintain a balance of power in the Persian Gulf, but it's not entirely clear we'd pick the same allies we currently have. I can make a reasonable case for a normal relationship with Israel (though not the current special relationship): it's a strong country in a critical region and it shares some values with the US (though that rationale is eroding rapidly). The problem is that unconditional support for Israel damages US standing in lots of other places. India is an easy case on realist grounds, though it's also a country that has significant internal problems and lots of troublesome neighbors, which means there's the danger of getting sucked into its problems. And as Stephen Kinzer argues in his book Reset, if we were starting from scratch, one can actually make a fairly good case for a closer strategic realtionship with ... (drum roll) ... Iran.
Of course, nations don't get to wipe the slate clean and start from scratch (at least not very often), and so my speculations today are, well, somewhat unrealistic. Nonetheless, it is worth asking this sort of question from time to time, if only to force ourselves to think about possibilities that seem totally at odds with present circumstances. After all, just because things are one way today doesn't mean they have to stay that way forever. And they won't.
Stephen M. Walt is the Robert and Renée Belfer professor of international relations at Harvard University.