38
weapons program or would it generate security anxieties in China over the US ability to
defend Taiwan if it were to declare independence? It might do both, and in this case,
debates on TMD might revolve more around the value, or harm, of increased allies’
anxieties versus increased rival’s anxiety.
Valorizing competing outcomes depends upon normative theories and principled
beliefs. Many expansion opponents doubted that NATO expansion was in US interests, a
distinct question from what the likely outcomes of expansion versus non-expansion
would be. However, these discussions neglected the issue of how the US or other NATO
allies would or should respond to violent conflicts in Central Europe. NATO was
becoming embroiled in Bosnia in 1994 and 1995, and backers argued that deterring or
preventing conflicts was preferable to intervening after they had intensified. Atlanticists,
however, believed that the US should avoid intervention in internal ethnic conflicts where
it lacked strong strategic interests; imperial multilateralists, by contrast, endorsed such
actions as a necessary part of effective collective security. In short, NATO expansion
predictions reflected disagreements over larger questions of grand strategy.
Future debates on security policy are likely to display similar problems.
Normative and empirical theorizing are often separated even in academic journals that
publish policy-relevant research. For example, arguments over whether partition is a
good solution to ethnic conflicts rest in part on beliefs about the normative value of
ethnic separatism and its implications for multi-ethnic democracy more broadly as well as
the dangers of forced migration. These questions, however, are rarely debated in journals
Convention