noted that ‘those of us involved for a long time have an intuition that CSCAP has fallen behind the policy curve’ (interviewee 7, 19 March 2008). A senior Chinese participantin CSCAP complained that many meetings had become ‘too formulaic’ (interviewee 8, Beijing, September 2009). The organization’s 2004 reconstitution of its open-ended five working groups into new ‘study groups’ with fixed sunset clauses was premised on the need to inject new life into a process that was widely regarded as moribund. Some of CSCAP’s outputs, such as a 2005 memorandum on ‘Enhancing Efforts To Address the Factors Driving International Terrorism’ were so watered down by the need to achieve consensus that some members did not consider them sufficiently useful to be worth forwarding to the ARF (interviewees 1 and 6).A related explanation focuses on the role of the key individuals and elites who press particular ideas. Peter Hall’s study of Keynesianism stresses the importance of administrative and political processes in shaping the diffusion of economic ideas (Hall, 1989). Analyses of policy networks similarly stress the role played by key individuals in spreading ideas. Richard Higgott talks about the importance of ‘understanding the dramatis personae of the policy process’ (1994: 367). Stuart Harris points to the critical role of a ‘small group of individuals or intellectual elites’ in building support for regional economiccooperation (1994: 384). This approach is picked up by many accounts of ASEAN-ISIS and CSCAP. One author notes the crucial role played by the ‘Asia Pacific track two elite’, a set of individuals ‘whose commitment transcends national barriers and who, by virtue of their advantaged positions as directors of think tanks, holders of prominent academic posts and the like, have been remarkably effective in coordinating their energies to advance institutional cooperation’ (Job, 2003: 253). Reliance on key individuals, however, is a double-edged sword. Well connected elites may ‘whisper in the ear of power’ and help spread ideas and create domestic constituencies for policy change, but dependence on individualscan also weaken networksif those people move on(Kao, 2003). Job notes that ‘the ebb and flow of regional track one and track two innovation over the last several decades can be correlated to a considerable extent with the appearance(and the inevitable receding) of figures from the regional political scene’(2003: 253). Tay notes that within ASEAN-ISIS, some member institutes have been weakened by leadership transitions and have ‘lost impetus and strength’ (2006: 139). Carolina Hernandez (personal communication, 20 March 2009) has said a ‘generational shift’ within some ASEAN-ISIS member institutes has gravely weakened the network.