By 1996, it was widely recognized that the CGIAR's description of its work was no longer consistent with frameworks used by others, including those who financed the System. For example, lines of work were described in terms of inputs rather than outputs and the classification system was deemed less useful than alternative approaches. In effect, better practice was now available. Interest mounted in reassessing and, perhaps, remodelling the format then in use.
In the autumn of 1996, TAC initiated a series of discussions and workshops to review the existing classification scheme for CGIAR research activities and to develop a logical framework ("logframe") in support of research planning and impact assessment. In order to ensure a broad spectrum of expertise and perspectives, a cross section of actors and experts both from within and outside the CGIAR System was invited to participate in this development. At each stage, feedback from the centres was sought and incorporated. Progress reports were circulated by TAC, following workshops in Feldafing, Germany, in February 1997 and 1998. The process was explicitly designed as a consultative one.1
1
While TAC served as organizer and convener, inputs were made by many members of the CGIAR and of its Finance and Oversight Committees as well as IAEG, Centre Directors, Board Chairs, CGIAR Secretariat, NARS and a number of outside experts. TAC gratefully acknowledges support received from the Government of Germany and from DSE.
The present document sums up results at a decisive stage - the concept is ready for step-by-step implementation. At the same time, TAC wants to reaffirm the notion that the framework is open to further improvement and adaptation. While not all questions can be answered simultaneously on the basis of a clear logical framework, solutions can be incorporated in a collaborative and constructive manner.
The logframe discussed here clearly reflects the change away from an activity/input to a project/output focus. It offers a higher degree of transparency in linking outputs to goals than in the past. The positive effects of these changes will include better co-ordination and management within the System and within the centres, more effective participation of NARS and others, and greater accountability to members and other stakeholders.
Among its characteristics, the logframe requires the development of a clear, tightly drawn logical relationship between outputs, purposes, and the ultimate goals of the System. In the case of the System's more basic research, the framework must recognize that its outputs will be directed as inputs to other projects which, in turn, move toward the attainment of the same goals. One benefit of these characteristics is that impact assessment and performance evaluation can be more effectively integrated and utilized. Taken together, the logframe should further strengthen priority setting by the System, by centres, and by members. As well, it will assist members' efforts to be accountable to their constituencies.
Applying the logframe described here for future research planning should not be an additional burden for researchers and managers. In due course, it will replace the present approach(es). While it would be unrealistic to believe that a logframe reduces the overall time to be devoted to research planning, it is safe to assume that the final product can be delivered more efficiently. As well, reporting about implementation and achievements will be greatly simplified. Finally, once basic principles and conventions are established, it will support creativity and encourage new ventures by providing a clear framework for their entry into the research agenda.
To sum up, the rationale for adopting the logframe approach at System, centre and project level is:
· to increase transparency at all levels for members by establishing clear logical relationships between inputs, expected outputs, and impact to be achieved;· to aid scientists at the centres as well as their collaborators in partner organizations to better plan, implement and evaluate research efforts;
· to provide a meaningful and concise way of aggregating projects both at the centre as well as System level;
· to streamline research management instruments.
