conclusions 155
This puzzle is analogous to the one motivating the study of conspicu-
ous consumption. Wars occur when actors choose a more expensive op-
tion (war) over a cheaper one (ex ante diplomatic agreement) even though
both options offer the same primary utility. Fearon offers a set of three ra-
tional mechanisms that can explain this counterintuitive choice: informa-
tion structures, credibility of commitments, and divisibility of goods. Yet
because he only relies on primary- utility calculations, he ignores the pos-
sibility of Veblen effects. Just like the Thai carrier, or the moon landing
program, war may be chosen precisely because it is more expensive than
negotiation, precisely because the additional cost may increase an actor’s
stature. Indeed, many contemporary studies of war are so caught up in the
language of material interest that they discount the possibility that pres-
tige and honor could still be motivations for war (Kagan 1995 , 8 ). Yet the
Veblenian framework, while focusing on prestige, does not depart from the
rational choice character of Fearon’s bargaining approach — it only points
out that one has to acknowledge a wider set of concerns when calculating
the utility functions of actors, especially when social and /or positional
goods are at stake. Indeed, analyzing war as a product of a Veblen effect
offers a fourth rationalist solution to Fearon’s puzzle.^1
The study of war is not the only topic that can benefit from the intro-
duction of conspicuous consumption. The logic underlying the puzzle of
war, as presented by Fearon, surfaces in similar terms in other rationalist
studies of international relations. Wagner’s ( 1983 ) study of arms races of-
fers another example of a similar puzzle:^2 Wagner argues that following
the logic of the security dilemma, a rational actor should be able to an-
ticipate that increasing his own defense expenditures is likely to trigger
similar reactions by his rivals. Once a rival matches this initial investment
in new armaments, our actor is likely to find himself just as insecure as he
was before making the initial investment, and considerably poorer.^3 Con-
sequently, arms races present a rationalist puzzle.
Why would he ever increase defense expenditures? Or in game theoretic terms,
the armaments game is a zero sum game with perfect information. It ought,
therefore, to have an equilibrium in pure strategies, and it appears that this
equilibrium should be at low levels of expenditure. How, then, are arms races
possible? (Wagner 1983 , 340 )