384 International Organization My third theoretical argument concerns the occurrence of change in and of regimes.The prevailing model postulates one source of regime change, the ascendancy or decline of economic hegemons,and two directions of re- gime change,greater openness or closure.If,however,we allow for the possibility that power and purpose do not necessarily covary,then we have two potential sources of change and no longer any simple one-to-one corre- spondence between source and direction of change.For example,we could have a situation in which there exists a predominant economic power whose economic program differs fundamentally from that of its leading rivals(e.g., Dutch supremacy in the 17th century).Or,we could have a situation in which power and purpose covary negatively,that is,in which neither a hegemon nor a congruence of social purpose exists among the leading economic pow- ers(the interwar period approximates this case).We could have a situation in which power and purpose covary positively (e.g.,Bretton Woods).There remains the situation of no hegemon but a congruence of social purpose among the leading economic powers (albeit imperfectly,the post-1971 in- ternational economic order illustrates this possibility). It is the last possibility that interests me most.It suggests the need for a more nuanced formulation of regime change than is currently available.If and as the concentration of economic power erodes,and the "strength''of international regimes is sapped thereby,we may be sure that the instruments of regimes also will have to change.15 However,as long as purpose is held constant,there is no reason to suppose that the normative framework of regimes must change as well.In other words,referring back to our analytical components of international regimes,rules and procedures (instruments) would change but principles and norms (normative frameworks)would not. Presumably,the new instruments that would emerge would be better adapted to the new power situation in the international economic order.But insofar as they continued to reflect the same sense of purpose,they would represent a case of norm-governed as opposed to norm-transforming change. Applying this argument to the post-1971 period leads me to suggest that many of the changes that have occurred in the regimes for money and trade have been norm-governed changes rather than,as is often maintained, reflecting the collapse of Bretton Woods and a headlong rush into mercan- tilism.Indeed,in certain cases earlier acts by the hegemon had violated the normative frameworks of these regimes,so that some post-1971 changes may be viewed as adaptive restorations of prior sets of norms in the context of a new and different international economic environment.Both occur- rences may to taken to demonstrate what we might call "the relative au- tonomy''of international regimes (with due apologies to the appropriate quarters). The various parts of my argument clearly stand or fall together.Ulti- 1sThe"hegemonic stability'school effectively demonstrates why this is so.See Keohane, Theory of Hegemonic Stability."384 International Organization My third theoretical argument concerns the occurrence of change in and of regimes. The prevailing model postulates one source of regime change, the ascendancy or decline of economic hegemons, and two directions of re￾gime change, greater openness or closure. If, however, we allow for the possibility that power and purpose do not necessarily covary, then we have two potential sources of change and no longer any simple one-to-one corre￾spondence between source and direction of change. For example, we could have a situation in which there exists a predominant economic power whose economic program differs fundamentally from that of its leading rivals (e.g., Dutch supremacy in the 17th century). Or, we could have a situation in which power and purpose covary negatively, that is, in which neither a hegemon nor a congruence of social purpose exists among the leading economic pow￾ers (the interwar period approximates this case). We could have a situation in which power and purpose covary positively (e.g., Bretton Woods). There remains the situation of no hegemon but a congruence of social purpose among the leading economic powers (albeit imperfectly, the post-1971 in￾ternational economic order illustrates this possibility). It is the last possibility that interests me most. It suggests the need for a more nuanced formulation of regime change than is currently available. If and as the concentration of economic power erodes, and the "strength" of international regimes is sapped thereby, we may be sure that the instruments of regimes also will have to change.15 However, as long as purpose is held constant, there is no reason to suppose that the normative framework of regimes must change as well. In other words, referring back to our analytical components of international regimes, rules and procedures (instruments) would change but principles and norms (normative frameworks) would not. Presumably, the new instruments that would emerge would be better adapted to the new power situation in the international economic order. But insofar as they continued to reflect the same sense of purpose, they would represent a case of norm-governed as opposed to norm-transforming change. Applying this argument to the post-1971 period leads me to suggest that many of the changes that have occurred in the regimes for money and trade have been norm-governed changes rather than, as is often maintained, reflecting the collapse of Bretton Woods and a headlong rush into mercan￾tilism. Indeed, in certain cases earlier acts by the hegemon had violated the normative frameworks of these regimes, so that some post-1971 changes may be viewed as adaptive restorations of prior sets of norms in the context of a new and different international economic environment. Both occur￾rences may to taken to demonstrate what we might call "the relative au￾tonomy" of international regimes (with due apologies to the appropriate quarters). The various parts of my argument clearly stand or fall together. Ulti￾l5 The "hegemonic stability" school effectively demonstrates why this is so. See Keohane, "Theory of Hegemonic Stability