Two decision-makers choose hawkish or dovish actions in a conflict game with incomplete information. The decision-making can be manipulated by “extremists” who send publicly observed cheap-talk messages. The power of extremists depends on the nature of the underlying conflict game. If actions are strategic complements, a “hawkish extremist” can increase the likelihood of conflict by sending messages which trigger a “fear-spiral” of hawkish actions. This reduces the welfare of both decision-makers. If actions are strategic substitutes, a “dovish extremist” (pacifist) can send messages which cause one decision-maker to back down and become more dovish. This reduces his welfare but benefits the other decision-maker. The hawkish extremist is unable to manipulate the decision-makers if actions are strategic substitutes, and the pacifist is equally powerless if actions are strategic complements.

An electronic version of the paper may be downloaded • from the SSRN website: www.SSRN.com • from the RePEc website: www.RePEc.org • from the CESifo website: Twww.CESifo-group.org/wp T

We examine whether there is evidence that the US intelligence services pandered to their political masters when constructing intelligence estimates during the Cold War. We construct a model which shows how career concerns on the part of intelligence analysts could lead them to distort reports towards their president’s prior beliefs. We then take the model’s prediction, that errors in intelligence reports should be correlated with presidential ideology, to the data by constructing a unique measure of intelligence failures that compares CIA/ORE reports on the Soviet strategic, nuclear arsenal to credible, post-Cold War estimates of the Soviet Union’s actual nuclear capabilities. We find that report errors are systematically, positively correlated with both a conventional measure of presidential ideology and a unique, text-analytic measure of presidential “hawkishness”. This result is robust to controlling for a number of endogeneity issues and alternative mechanisms such as reverse causality, variation in intersuperpower relations, collusion between politicians and the intelligence agencies, and turnover in agency staff. Finally, there is evidence that longer-term forecasts are more sensitive to ideology in a manner that is consistent with our model. (JEL: H56, L82, N42) 1

Ambiguity can be used as a strategic device in some situations. To demonstrate this, we propose and study a framework for normal form games where players can use Knightian uncertainty strategically. In such Ellsberg games, players may use Ellsberg urns in addition to the standard objective mixed strategies. We assume that players are ambiguity–averse in the sense that they evaluate ambiguous events pessimistically. While classical Nash equilibria remain equilibria in the new game, there arise new Ellsberg equilibria that can be quite We thank Andreas Blume, Stephen Morris, Ariel Rubinstein, Burkhard Schipper, and Marco Scarsini for discussions about the decision–theoretic foundations of game theory. We thank Joseph Greenberg, Itzhak Gilboa, Jürgen Eichberger, and Bill Sandholm for

When does nuclear counterproliferation succeed? When does it lead to preventive war? We argue that the answers to these questions hinge on the effect of nuclearization on the balance of power relative to the cost of preventive war. When it is low, threats of preventive war are not credible; proliferation continues apace and peace prevails. Such was the case during the Cold War. When it is high, threats of preventive war are credible, slowing down the rate of proliferation. At the same time, since the decision to proliferate is not perfectly observable, there is a higher likelihood of mistaken preventive wars. This characterizes the post Cold War. We trace the logic of our argument by looking at the cases of Soviet nuclear acquisition in 1949 and the US-led invasion of Iraq in 2003.

University for helpful comments. We are grateful for the research assistance of Dr. Xi Chen of Yale, to the Glaser Foundation, and to the MacMillan Center at Yale for financial assistance. All our data and computations will be available on a website by the time of publication. We consider the influence of countries ’ external security environments on their military spending. We first estimate the ex ante probability that a country will become involved in a fatal militarized interstate dispute using a model of dyadic conflict that incorporates key elements of liberal and realist theories of international relations. We then estimate military spending as a function of the threat of armed interstate conflict and other influences such as arms races, the defense expenditures of friendly countries, actual military conflict, democracy, civil war, and national economic output. In a panel of 165 countries, 1950 to 2000, we find our prospectively generated estimate of the external threat to be a powerful variable in explaining military spending. A one-percentage point increase in the aggregate probability of a fatal militarized dispute, as predicted by our liberal-realist model of interstate conflict, leads to a three percent increase in a country’s military expenditures.

Agents with extreme agendas sometimes take provocative actions that inflame con-flicts. For example, Ariel Sharon’s symbolic visit to the Temple Mount in September 2000 helped spark the Second Intifada and derailed the Israeli-Palestinian peace process (Hefetz and Bloom 2006). How can extremists manipulate conflicts and when is it rational to respond aggressively to provocations? Provocations play a key role in the conflict between the two nuclear powers India and Pakistan.1 After September 11, 2001, Pakistani President Musharraf sent troops to the Afghanistan border, and tried to suppress militant groups within Pakistan. In December 2001, militants sponsored by the Pakistani intelligence agency ISI attacked the Indian Parliament. India mobilized for war, and Musharraf shifted his troops from the Afghanistan border to the Indian border. Similarly, in November 2008 a terrorist attack in Mumbai raised tensions at a time when Pakistani President Zardari wanted improved relations with India. ISI-sponsored militants seem to deliberately inflame the conflict between Pakistan and India, partly because India is seen as an implacable foe, but also because the conflict relieves the pressure on extremists supported by the ISI. For Pakistani and Indian leaders, a hawkish stance may be the best response, given the (correct) belief that their opponent will become more aggressive.2

Abstract It is often suggested that incentive schemes under moral hazard can be gamed by an agent with superior knowledge of the environment, and that deliberate lack of transparency about the incentive scheme can reduce gaming. We formally investigate these arguments in a two-task moral hazard model in which the agent is privately informed about which task is less costly for him to work on. We examine two simple classes of incentive scheme that are "opaque" in that they make the agent uncertain ex ante about the values of the incentive coefficients in the linear payment rule. We show that, relative to deterministic menus of linear contracts, these opaque schemes induce more balanced efforts, but they also impose more risk on the agent per unit of aggregate effort induced. We identify settings in which optimally designed opaque schemes not only strictly dominate the best deterministic menu but also completely eliminate the efficiency losses from the agent's better knowledge of the environment. Opaque schemes are more likely to be preferred to transparent ones when i) efforts on the tasks are highly complementary for the principal; ii) the agent's privately known preference between the tasks is weak; iii) the agent's risk aversion is significant; and iv) the errors in measuring performance on the tasks have large correlation or small variance. (JEL D86, D21, L22)

Would perfectly rational agents always negotiate peaceful outcomes at the bargaining table, or would they sometimes …ght costly wars? The Coase theorem suggests that when rational agents negotiate freely, they will reach a surplus-maximizing outcome. This seems to rule out war, since war will not in general be surplus-maximizing. However, the Coase theorem is valid only under certain assumptions, such as transferable utility (no restrictions on side-payments) and complete information. Brito and Intriligator (1985) showed how incomplete information may lead to war. An aggressor who demands concessions may simply be blu ¢ ng, so it may be rational to refuse his demands. If the aggressor is not blu ¢ ng, a war may ensue. We discuss how long such a war may last, and whether other kinds of “frictions” (such as limited commitment power and limits on side-payments) may also lead to war.