【正文】 se they have owned the expertise needed for ?best practice? in their members of staff, contractors or consultants. This has had the effect of colocating power and knowledge in such a way as to make consideration of alternative policy and implementation approaches difficult. We have previously argued that in the late 1990s in Australia policy and economic advantages converged in the (re) emergence of the concept of munity as a powerful public idea (Adams and Hess, 2022). This was not a new idea internationally or in Australian public policy and echoed the 197275 Labor Government?s Australian Assistance Plan which had proposed munity involvement as a means of positioning government activity in local environments. The difference in the late 1990s was that with political conservatism dominant the use of munity in policy reflected a new set of priorities. The previous focus on redistributive activity had been shifted to (in Prime Minister Howard?s words) a mix ? which bines liberalism in economic policy and ? . “Modern conservatism” in social policy? as ?mutually reinforcing? elements (Howard, 1999). In practice, however, governments have found the balance hard to strike with critics pointing to a tendency to emphasize budget relief and the fa231。ade of consultation rather than genuine munity engagement (Public Policy Forum, 1997。 ACOSS, 1998). As the century ended the struggle to make munity a practical part of policy processes continued. We had argued that an uncritical rush to munitybased practices ran the danger of ?being another policy fad? with little actual benefit (Adams and Hess, 2022, p. 21). The solution, we felt, was in a more thoughtful approach, which assumed a sharp learning curve in which quickfix solutions were rejected in favor of careful consideration of ideas and how they might be made to work. Consideration of the changing knowledge base of good policy would be a central aspect of this effort (Hess and Adams, 2022). Economic knowledge had bee so dominant under NPM as to be selfreferential and therefore functionally unchallengeable. The difficulties this created were of two kinds. The first were those, which arose from the exclusion of consideration of other knowledge frames. The second had to do with the positivist basis of economic knowledge. Our epistemological argument has been that historically public administration applied ideas and used instruments arising from frameworks of knowledge and meaning which were relatively stable (Hess and Adams, 2022). They changed quite slowly over time and were closely linked to socially normative concepts underpinning and legitimizing administrative action. In the 1990s NPM privileged functional knowledge drawn primarily from economics and management, pushing other knowledge frames into the background. This was consistent with earlier changes in so far as it continued the reliance on knowledge provided by experts. In this period, however, the expertise was drawn increasingly from outside the administration itself with private sector models and panies providing many of the new ideas and ways of doing things. Noheless, knowledge was still assumed to be something to be sought and, once found, applied by the experts who were in, or were contracted by, central departments.