dations are followed through on. In fact, in Peru, panel members suggested that putting in place a structure to provide such follow-up would be possible with modest resources. Discussions along that line are ongoing. This function for the LGAF is very similar to that of the land observatories that have already been established in various contexts, and the framework could build on that structure to work toward establishing broad-based land working groups at the national level. Such a group could provide regular input into national forums, such as the Comprehensive Africa Agriculture Development Programme roundtables to provide specific operational guidance to policy. This group could be linked to agreement on specific steps and report on progress made toward improving land governance in response to multilateral initiatives such as that of the African Union or of other institutions. Second, beyond the discrete indicators, the LGAF points to a number of areas that change relatively quickly and in which the design of quantitative indicators to monitor land governance on a more frequent basis will thus be useful. Although more work will be needed to agree on the specific definition of variables, the LGAF experience suggests that key areas of concern include (a) the coverage of the land administration system (that is, the extent to which primary or secondary rights of groups or individuals are recorded) and the extent to which different types of transfers are registered, with particular attention to women; (b) the amount of land tax revenue that is raised; (c) the total area of public or private land that is mapped with publicly available information; (d) the number of expropriations and the modalities for compensation (including amounts and delays in receipt of payment); and (e) the number of conflicts of different types entering the formal system. The fact that each of these indicators is related to one or more core areas of the land administration system suggests that the collection and publication of these indicators on a regular basis, and in a way that can be easily disaggregated by administrative units,5 should be routine in any land administration system and should be integrated in future donor support in the land area. In fact, addressing the two elements, that is, monitoring discrete measures of policy and designing specific quantitative indicators, would provide the basis for a more results-based way of providing support to the land sector that could help increase accountability at the national level. It would also help with the sharing of experience and collaboration across countries to effectively address some of the challenges in trying to improve land governance. NOTES 1. In a number of instances during the pilot, involvement of midlevel government officials unwilling to admit to shortcomings in the way the system operated made it impossible for the panel to come to a consensus view. If there is a danger of this happening, a useful approach would probably be to have the panel present a user view of the system and to gather consolidated government comments thereafter but before a public workshop is held.
CONCLUSION AND NEXT STEPS
137