Efforts to plan economic development tend to apply uniform and supposedly optimal solutions, increasingly expressed in terms of targets, such as, in the area of interest here, the percentage of registered and properly mapped land.1 The present paper contends that the reason why these centralized and expert-led efforts tend to fail massively is because they still fail to adapt to local circumstances (Easterly, 2014). They therefore disregard the value of “knowledge of the particular circumstances of time and place” underscored by Hayek (1982:521) which, even after the fall of Socialist economies, has widely been ignored in development.

۱٫۱٫ Established policy

The survival of this planning misconception has affected land titling projects, which often aim at universal coverage in a given area, a policy often optimistically referred to as “systematic”, while the policy defended in this paper based on voluntary and, therefore, selective titling and demarcation is dismissed as “sporadic”. Thus, according to the World Bank’s good practice guidelines summarizing lessons learned since it started investing seriously in this area in the 1990s: “Systematic registration identifies, adjudicates, and registers rights to all adjacent land parcels in a selected locality and within a given period of time. Sporadic registration processes land rights on an ad hoc basis, usually when customers request registration of their parcels of land” (World Bank, 2005:397).2 The extreme version of this systematic-titling argument even led part of the “land administration” literature to propose “holistic” objectives, according to which surveying each land parcel was considered a requirement for good titling.3 Such land administration systems must not only provide complete information on land, including the different types of private rights and public restrictions, but ideally they must also be managed in a integrated manner to facilitate governments’ control of resources. It is claimed, for example, that “governments must proceed further and integrate the actual management processes of land use, tenure and development” (Bennett et al., 2008:127; summarizing Enemark et al., 2005). Similarly, the World Bank advised in 2005 that “cadastre and registration functions should be connected and managed by a single institutional entity wherever possible” (World Bank, 2005:398). This single-agency policy—despite some more nuanced judgments by the evaluation unit of the Bank (IEG, 2016:28)—is still advised by the Bank’s Doing Business 2018 report, when it considers as best practice in property transactions “to bring all agencies involved in property transfer under one roof”; ۴ and when the “coverage” component of its “Quality of Land Administration” index also scores countries higher if their mapping agency covers the largest business city in full and even higher if it covers the whole economy. Consistent with these recommendations, which have deserved little attention in the law and economics literature,5 for decades most land titling projects and land registration reforms have been trying to implement exhaustive, even mandatory, land titling and demarcation based on delineating boundaries and representing parcels graphically, often in conjunction with attempts to integrate land registries with cadastres to reap the benefits of alleged synergies.