"asal bapak senang" that may be translated as "as long as father is happy" (with the sense of "to avoid upsetting your boss with negative information"). At each level in the hierarchy, the bad news about the devastation in the rice fields was watered down. It was only when the people from his own village came to the President directly to ask for help that he learned that something was seriously amiss. In our assessment, policy initiatives that aim at empowerment and endogenous development would be most accepted where democratic forms of government and a strong civil society exist; most poor people live in countries where these conditions are not present.

7.5.4 Innovation systems (IS)

Innovation is the emergent property of the interaction among organizations and people who make the complementary contributions required for innovation to take place (Röling and Engel, 1991; Bawden and Packam, 1993). The configuration of actors is not fixed (Engel and Salomon, 1997).

      The empirical research of successful and innovative economies that stimulated the recent interest in innovation Systems has found that "the essential determinant of innovation appeared to be that the suppliers of new knowledge were intimately engaged with the users of that knowledge" (Barnett, 2006).

      Older traditions of systems thinking and practice (e.g., Checkland, 1981; Checkland with Scholes, 1990) drew attention to linkages, relationships, interfaces, conflicts, convergence, and reciprocity in innovation processes. The application of such thinking and practice to pro-poor development in agriculture has been stimulated also by the evidence that it appears to be suited to dealing with the kind of institutional development that The New Institutional Economics (North, 2005) sees as a precursor to growth.

      The "innovation systems" approach in recent years has become an ex-ante policy model (World Bank, 2007) that draws on the aforementioned traditions as well as on empirical research on the emergence of Asian economies. Such models are an increasingly important tool for stimulating innovation at the interface of agriculture, sustainable natural resource management and economic growth, for instance in the context of the EU's Water Framework Directive (e.g., Blackmore et al., 2007) and Land Care and more recently Catchment Management Authorities in Australia (Campbell, 1994). These experiences also show up the weakness of the IS approach: absent appropriate enabling policy frameworks and economic drivers at higher system scales, successful lower scale innovations can peter out or become

frustrated. The lessons may be linked to the widespread confidence that rational choice theory offers an appropriate foundation for policy designed to support innovation; the empirical evidence suggests to the contrary that, given the public good character of development and sustainability goals, policies based on an understanding of the role of collective management in innovation processes may be more appropriate (Ostrom et al., 1993; Gunderson et al., 1995).

Conditions under which the policy options may be conducive to meeting development and sustainability goals. The following concrete steps have been proposed to make an innovation systems approach work in resource-poor environments (see Tripp, 2006; McCann et al., 2006; Van Huis and Houkonnou, 2007):

• Public, private and civil society agencies identify a number of priority themes based on national plans, or poverty reduction strategies;
• For each theme, rapid appraisal of agricultural knowledge systems (RAAKS) (Engel and Salomon, 1997) or other methods are used to identify configurations of stakeholders (including researchers, farmer organizations, etc.) that constitute promising innovation systems. Such configurations include actors at the both the national and the decentralized local government level;
• Key representatives of these stakeholders are facilitated to form a "Community of Practice" (COPs) (Wenger, 1998) at decentralized (e.g., district) and national levels, where the national level has the power and ability to create conducive institutional framework conditions for the concrete activities at the decentralized level. An IS approach thus requires trained facilitators who operate within a national mandate that recognizes the importance of IS;
• For each COP, diagnostic studies identify concrete opportunities that can be realized through concerted action by the stakeholders;
• Each COP submits proposals to a national fund set up for this purpose;
• Each COP is monitored to allow national learning about the IS approach as a basis for staff training and increasing management effectiveness.

The IS approach assumes considerable political will and an understanding of processes that cannot be captured by hierarchy and market since creating windows of opportunity for small-scale producers will require new kinds of institutional innovation (Egelyng, 2000).