adopted sensitive approaches. The agreement of NGOs concerned with agroforestry
 in South Nyanza District of Kenya to coordinate their action with local government
 offices is particularly noteworthy (Musyoka et al., 1991).


 What Can NGOs, Research Services and
 Funding Agencies Do?

In many countries the potential exists for mutually-beneficial links between NGOs and public
sector research and extension services, though numerous obstacles also exist, and paths
towards closer links have to be charted carefully.

Two forewarnings are essential: first, the prospects will vary widely among countries,
according to the wider context of NGO-State relations and according to how far NGOs and
Government share a common view of the future of the rural poor, and of strategies for
achieving that future. Second, within countries there is a wide diversity of NGO types, and
their relations with government will lie along a spectrum from outright hostility to willing
collaboration. Recognition of this diversity is implicit in our continued use of the loosely
positive term 'link' to describe potential modes of interaction between NGOs and GOs.
'Collaboration', by contrast implies a high degree of mutual interdependence.

Linkages can be promoted by the following actions:

i) Efforts by GOs to keep a small percentage of their budgets (whether from central
 government or from externally-assisted projects) 'unallocated' to allow for rapid
 response to NGO requests as they arise. Needs and opportunities for potentially
 productive interaction often occur unexpectedly: they cannot always be held over to
 the next financial year. Alternatively, a percentage of staff time might be kept
 unallocated, and NGOs provided with the resources by their funding agencies to
 contract GO staff.

ii) GO and NGO staff can jointly participate in training courses (ideally led by a joint
 team) in the 'learning-by-doing', 'action-oriented' methods favoured by NGOs such
 as participatory rapid appraisal (for examples see the RRA Notes series published by
 the International Institute for Environment and Development). The relevance of these
 methods to individual GO staff will vary, but their capacity to consolidate farmer-
 oriented perspectives is important.

iii) Efforts need to be made by NGOs to interact more fully with each other than hitherto.
 Interaction may begin with exchanges of information and joint meetings, perhaps
 extending in some cases to fully collaborative projects. Most countries are character-
 ised by large number of NGOs of varying size, and GOsmay find it easier to work
 through effective NGO networks. Continuing attention is therefore needed to the


GATEKEEPER SERIES NO. SA43