US exchange rate policy shifted in 1985 from unilateralist nonintervention to actively promoting dollar depreciation and multilateral cooperation. Pressures from producer interests, particularly multinational companies making manufactured goods, and from sympathetic members of Congress were the most important of multiple forces pushing the US Treasury toward dollar depreciation. Once the Treasury had chosen an activist course, a multilateral strategy had several benefits over a unilateral approach to depreciation. It could better counter the immediate threat to Administration trade policy from Congress, orchestrate depreciation, strengthen Treasury's influence within Washington and shift the burden of adjustment away from US fiscal policy, then frozen, onto other governments. When the trade account is in balance, individual policymakers have flexibility in determining exchange rate policy. But when large trade deficits arise, the domestic political pressures of trade-exposed sectors will dominate personalities and ideas.

Existing explanations of European monetary integration, emphasizing economic interdependence, issue linkage, institutions, and domestic politics, take a predominantly regional approach. In the international monetary thesis developed here, I argue that U.S. policy disturbances, transmitted through the international monetary system, created compelling incentives for European states to cooperate on exchange-rate and monetary policy. I develop a general theory of macroeconomic power, based on open economy macroeconomics, and show how the exercise of such influence can drive regional monetary integration. This article then tests the international thesis with reference to monetary integration within the European Union by examining four periods in which the United States acted to stabilize the international monetary system and seven episodes in which it disrupted the system. European governments and central banks reduced regional monetary cooperation when the United States supported system stability and strengthened it after each episode of disruption. The evidence thus strongly supports the inference that the link is causal.