In May 2001, under the patronage on the Ministry of Agriculture of Tunisia, the CGIAR System-wide Program for Collective Action and Property Rights (CAPRi) co-organized an International Conference with the International Center of Agricultural Research in the Dry Areas (ICARDA) and the Tunisian National Institute of Agricultural Research (INRAT) on Policy and Institutional Options for the Management of Rangelands in Dry Areas. The conference's regional focus was on North African, Sub-Saharan and West Asian countries. Research findings on institutional issues regarding rangeland management and discussions between policymakers, community leaders and researchers identified critical issues and policy directions for the future. The research conducted under the CAPRi co-sponsored projects in Niger, Ethiopia, Burkina Faso, Tunisia, Morocco, Syria and Jordan focused on the role of cooperation and collective action at the local level and the effects of different rangeland management options on livestock production systems and livelihood strategies. The results indicate that communities can and do cooperate over the management of natural resources -- either through cooperative societies or more informal traditional mechanisms - but the extent to which they are successful differs across communities. The institutional options used to manage rangelands have had different effects on household production strategies. There is evidence that building on local knowledge and traditional structures to create more formal structures is likely to lead to better management, that increased market integration does not have a negative impact on cooperation (and in some cases increases cooperation), but that large communities with heterogeneous populations engaged in different livelihood strategies do find cooperation more costly. Participants highlighted the importance of identifying the appropriate roles and responsibilities of institutions at different administrative levels. This means that optimal devolution/ decentralization policies need to balance the flexibility and informational advantages of more informal local systems of resource management with institutional arrangements that offset high level of conflicts, greater capacity to free-ride, and encroachment of cropland that often accompany such flexible systems.

More specifically, participants largely agreed that the state's role in rangeland management should focus on the following: 1) the provision of legal frameworks to resolve uncertain property rights where these are seen as the root cause of degraded, unproductive rangelands and conflict, 2) ensuring that local institutions represent interests of all community members and developing policies that do not further hamper mobility of livestock herds, and 3) undertaking large-scale infrastructural and other investments in rangelands where these provide large public goods and where returns are uncertain and accrue over the long run.

Both research and policy discussions recognized the risk-reducing role of mobility and the need for access to a wide range of pastoral resources. From a policy perspective the study findings indicate that maintaining mobility has positive effects on livestock productivity and reduces vulnerability under poor rainfall conditions, but that overlapping claims and shared resources also make cooperation more costly.

Another important result from the research was the empirical observation that stock densities are generally lower in areas with high rainfall variability, because of the added risk to production. Policies and programs that successfully mitigate impacts of drought may in fact induce dramatic and unsustainable increases in stock levels. Therefore drought mitigation strategies should reduce vulnerability of herders to drought but at the same time prevent increases in stockholdings.

Finally, participants emphasized the need for integration of rangelands management within the larger context of national development plans. Rangeland restoration projects, drought contingency plans, and property rights policies must consider the overall functioning of the agro-pastoral system within which livestock production is but one, though important, activity.