NOTES AgriculTurAl & rurAl DEvElOpmENT 42695 Land Policy and Administration implementing low-cost rural land certification: The case of Ethiopia iSSuE 34 fEBruAry 2008 By klAuS DEiNiNgEr While many African countries have recently adopted highly innovative and pro-poor land laws, lack of implementa- tion thwarts their potentially far-reaching impact on productivity, poverty reduction, and governance. By contrast, and despite a policy environment that is lacking in many respects, Ethiopia has, over a short time period, distributed certificates to more than 20 million plots in a very decentralized process. We use a nationally representative survey of more than 2,000 households to decribe the process, explore participation by women and the poor, assess its impact, and quantify the costs. Results suggest that first-time registration has been very cost-effective and not biased against the poor while at the same time providing important lessons to enhance its sustainability and impact. BAckgrOuND and maintenance of documentary evidence that are There is now a growing consensus that, even in rural less demanding and costly than titles while at the African contexts where individual titling of land may same time offering transparent and non-discriminatory not be desirable or feasible and use of land as a col- options for upgrading as need arises. lateral for credit is at best a distant possibility, providing However, despite the attractiveness of such provisions poor land owners or users with options to have their in theory, progress in implementing them has lagged rights documented can yield significant benefits. These behind expectations. In fact, hardly any of the coun- benefits, which arise largely from the ability to invoke tries that introduced legal reforms with much fanfare formal mechanisms to protect property rights, include succeeded in developing, let alone rolling out, a low- incentives for land-related investment, enhanced cost system for land administration at a scale that is gender equality and bargaining power by women, sufficiently large to provide an option for the majority improved governance, reduced conflict potential, and of the poor. This impedes realization of many of the lower transaction costs for productivity-enhancing land expected benefits from such legislation and implies transfers through either rental or sale. A desire to reap that the poor often remain excluded from formal sys- these benefits provides the intellectual foundation for a tems and vulnerable to land loss. More generally, it has recent wave of land law reforms that aim to give legal raised doubts regarding the technical, institutional, and recognition to customary rights, strengthen women's political feasibility of such reform. To assess whether property rights, and establish processes for creation such doubts are justified and whether a decentralized and less demanding model of property rights reform can have impact on the ground, we draw on the expe- rience of Ethiopia which, over a period of 2-3 years, registered the majority of rural lands in a rapid process at rather low cost. EThiOpiA'S lAND pOlicy AND mODAliTiES Of prOgrAm implEmENTATiON Prior to 1975, Ethiopia's land tenure system was complex and semi-feudal. Tenure was highly inse- cure, arbitrary evictions were common, and many lands underutilized. High inequality of land owner- ship reduced productivity and investment, leading to political grievances and eventually the overthrow of the imperial regime in 1975. The Marxist government that took power transferred ownership of all rural land to the state, leading to declines in productivity ThE WOrlD BANk and soil degradation. Following the overthrow of the ment of common property resources and house plots; Marxist regime in the early 1990s, intentions to move and (iv) although many boundaries were mapped in the towards private land ownership did not materialize. field, measured by rope, and corner marks emplaced, The 1995 constitution vests land ownership in the state certificates identify owners of neighboring parcels but and upholds the right of every Ethiopian who wants to contain neither a map nor a sketch of the parcel. engage in agriculture to receive inheritable use rights to a piece of land for free, a principle that can be enforced Household- or parcel-level data also allow us to explore through administrative reallocation of land but that will gender dimensions of land certification and allow a pre- likely conflict with the goal of ensuring land users' ten- liminary assessment of the extent to which certification ure security. A 1997 federal proclamation (law) devolved has been pro-poor. We find that (i) there is little evidence responsibility for land policy to the regions. Although this of wealth bias in access to the program or the information has led to considerable inter-regional diversity, a number surrounding it; (ii) female participation in the early stages of common issues emerge, namely (i) administrative of registration was limited and there is considerable varia- land redistribution is not an empty threat; (ii) while land tion even within regions in the share of documents issued rental is now allowed, the fact that in most regions the jointly or in the name of a female; and (iii) there is con- amount that can be rented is limited and that land rights siderable interest in, and willingness to pay for, a map to are contingent on physical residence in the village could complement the current verbal description of the land. undercut efforts at non-agricultural development and migration from rural areas; and (iii) mortgaging and sale Economic impact: Regressions point towards positive and of land are universally prohibited. highly significant investment effects of land certification. In villages where plots had been certified, a plot was 5% Table 1 summarizes differences across regions in key more likely to receive new investment and the amount of program provisions as well as the level to which the land new investment was 4.4% higher than where this was certification program had been implemented in different not the case. regions. Under the program, each village elects a land Cost: Standard parameters obtained through field visits administration committee (LAC) that takes responsibility indicate a cost of first-time registration (including the for public registration of plots in a field-based process certificate) of about US$ 1per plot. Use of GPS to add with presence of neighbors to increase transparency and a spatial reference would cost about US $ 1.5 per plot, reduces scope for manipulation or error. Once results 50% above the current process based on traditional tech- have been discussed publicly, households receive a pre- nology. Although similarly low costs have been achieved liminary registration certificate identifying their holdings, with systematic processes using modern technology in to be followed by an official certificate with holders' other contexts, this is at least an order of magnitude pictures (including a separate picture of the spouse in below what is reported in other literature. This provides Amhara and Oromia) and space for maps. The program an important benchmark for other African countries and has registered about 20 million plots by some 5.5 million suggests that technical factors can not explain the lack of households in a very short time. Household data provide progress with implementing innovative new land laws. evidence on (i) the extent to which the prescribed process have been adhered to in the field; (ii) the impact of first time registration; and (iii) the cost of the process. fuTurE chAllENgES Comprehensiveness: Although the program expends con- EviDENcE Of siderable effort to register individual holdings, it includes prOgrAm EffEcTS neither house plots nor common property resources (CPRs) on a consistent basis. This makes it difficult to use the data Implementation: Community-level data from our survey generated as a basis for an integrated land administration suggest that the process was largely implemented as system and contradicts users' desire as well as interna- planned, in particular that (i) in most locations public tional best practice. Demand for inclusion of house plots meetings were held before and during the certification is already high and consistent mapping of CPRs, possibly process; (ii) LACs were popularly elected and represented in combination with land use planning and assignment of most of the sub-kebeles; and (iii) adjudication made group rights, could help confront serious resource degra- use of traditional village authorities (elders) to resolve dation and soil erosion at low additional cost. disputes and demarcation was carried out in the field in the presence of neighbors. Shortcomings include that Keeping registries up to date: While evidence on first- (i) with only 20% of LACs including a female member, time registration is encouraging, the data generated will women's representation was limited; (ii) access to written quickly become obsolete unless self-contained processes information on the law and the process and purpose of to keep records up to date are in place. Evidence sug- certification was far from universal; (iii) the process was gests that failure to register changes in land rights due focused on mapping of agricultural holdings to the detri- to inheritance, divorce, long-term lease, and land taking 2 for public use already lead to gaps between records and would, with cost recovery levels in line with users' stated reality. Priority steps to address this include (i) redesigning willingness to pay, yield a self-sustaining system. To avoid registry books to achieve consistency across regions and obsolescence of the investment in first-time registration, ensuring that no new certificates be given out without steps to advance with this are urgent. cancelling old ones; (ii) clarifying institutional respon- sibilities to update information quickly in response to Legal awareness: Households will fully exploit the oppor- changes and eliminate gaps and time lags which are a tunities afforded by certificates only if they are aware potent source of possible inconsistency; and (iii) spelling of their legal rights and confident to be able to enforce out protocols to ensure public access to information at them. We find lack of knowledge or gaps between per- local and higher levels. Rough calculations suggest that current demand would not warrant computerization of ceptions and the law regarding tenure security (migrants' records for villages or even at county (woreda) level even ability to retain land rights), transferability of land, if electricity were available. However, computerizing the and women's rights in case of inheritance or divorce. 30% of woredas with the highest transaction frequency Remedying these while at the same time enhancing land and integration of all records into a national database rights will be important. Table 1: key features of the land registration process Ethiopia's four main regions Tigray Amhara Oromia SNNPR Basic information Year started 1998 2003 (pilots 2002) 2003 2005 (pilots 2004) No of kebeles 630 2895 6000 4431 Households registered a 632,000 (88%) 2.4 million (79%) 2.4 million 700,126 (40 %) No. of certificates issued n.a. (delays) 1.3 million n.a. (delays) Only in pilot areas Fee for certificate 3 birr c Free of charge 5 birr 2 birr Women's rights Certificate in name of head; Certificate joint: One certificate per family; Joint cert. W. 2 no photo. 2 photos 1 photo names & photos b Divorce Wife gets half the land but Wife gets half the land, but Wife gets share of land Wife gets half the land no certificate no certificate CPRs demarcated No (only pilots) Yes No Kebele discretion Procedures Initial training by Kebele; woreda staff Woreda officials Woreda Kebele; training centers Written materials used Few copies of Proclamation Posters (delayed) Yes; proclamation & book No due to language problems Supervision of data Students backed up Woreda survey team (3) & Expert advice from Rural development agent collection by technicians Land Adm. team (1 + 2) woreda on demand Data collection in field Not always Yes; not always done Yes Yes Presentation of results Discussed in kebele Meeting with woreda Yes, details at Yes, details at council official present discretion of LAC discretion of LAC Registers & certificates Plot map/sketch No No, except in pilot kebeles No, pilot proposal No Certificates written by Kebele, signed by Woreda LA team Kebele; signed by woreda Woreda LA expert kebele chairman. chairman Registry book written by Kebele Woreda LA team Woreda LA expert Woreda LA expert Book kept at kebele? Yes Yes No, only copies of data Yes collection forms Book kept at woreda Yes Yes Yes, very condensed Yes, data at household level (only few plots/hh) (clumsy size) Updating procedures General format Discretion of kebele Interplay between Woreda book with Woreda book still kebele & woreda` columns for transfers being developed Procedure for inheritance Transfer of certificate or New certificate; old cert. not New certificate; no clear Not clear return & new issue closed; no cross-reference procedure yet. Recording of leases Unclear Not in book Not in book Not in book Source: Field visits to regional Environmental Protection, Land Administration and Use Authority (EPLAUA) offices and 24 kebeles in all four regions; Sept. 2005. a Numbers as of August 2005, the latest date for which information was available. b Additional wives receive a joint certificate with their name first and their photo included. c US$1 = 9.09 Ethiopian birr as of 20 November 2007. 3 ConClusion and Beyond Ethiopia, and the need for follow-up in this poliCy impliCations country notwithstanding, the massive scale and posi- tive impact of first-time land certification highlights The evidence presented here leads to a number of con- that technical obstacles or resource constraints can not clusions. First, the rapid speed, participatory nature, and low cost of Ethiopia's land certification, together with explain the near-universal failure by African countries the positive results from this process and the absence of to put the innovative aspects of recent legal reforms bias in favor of the wealthy demonstrate that, contrary into practice. Instead, the main obstacle to doing so to what one might to conclude from experience in other and documenting land rights of current occupants on countries, large-scale and rapid delivery of land cer- a massive scale seems to be political. While some fac- tificates in a participatory way is possible. Users' positive tors, such as a top-down administrative structure and assessment of the process, readiness to pay to replace the absence of chiefs may have been propitious to the lost certificates, high demand -and willingness to pay implementation of such a program, none of them are modest sums- for a spatial reference, and their positive unique to Ethiopia and other factors (land scarcity in par- assessment of likely impacts suggest that the modalities ticular), make implementation of such a program much of implementation responded to local needs. The fact harder. To the contrary, our results suggest that, if there that most disputes could be resolved in the field and out- is political commitment to doing so, documenting land comes were not biased in favor of the wealthy or against rights at a massive scale--and at a cost that is affordable women, reinforced by initial evidence of positive invest- under African circumstances--is feasible and can have ment- and transfer-effects, all support this view. significant benefits even in the short term. In view of the critical role of land-based activities for poverty reduction At the same time, realizing the full potential of the in virtually all of sub-Saharan Africa, greater attention to country's first time registration in a sustainable man- securing land rights in relevant policy documents may ner is likely to require action in four areas. First, lack thus be warranted. Documenting the size, evolution, and of updating could quickly undermine the reliability of incidence of such benefits, and ways in which they can, the system, especially in areas with higher transaction frequencies, thus undermining its reliability and trust- in a given situation, be enhanced by specific policies and worthiness. Second, failure to register house plots and institutional arrangements will be an important topic for CPRs is inconsistent with users' demands and will greatly future research. reduce the ability of the system to help address key land use problems in the Ethiopian context. Third, although REFEREnCEs registration demarcates boundaries in the field, it does not create a graphical record and may thus fall behind Deininger, K., D. Ayalew, S. Holden and J. Zevenbergen, "Rural expectations in terms of reducing boundary disputes. As Land Certification in Ethiopia: Process, Initial Impact, and technology to meet demands for a low-cost map in a way Implications for Other African Countries." World Development consistent with the decentralized process is available, forthcoming. ways to include it should be explored. Finally, experience Deininger, K. & Jin, S. 2006. Tenure Security and Land-Related illustrates that certificates will be kept up to date only if Investment: Evidence from Ethiopia. European Economic land users perceive that they add value by increasing their Review, 50, 1245-1277. tenure security. Key aspects of this include (i) well defined compensation in case of expropriation; (ii) protection of Fitzpatrick, D. 2005. `Best practice' options for the recognition contracts and security against arbitrary redistribution; of customary tenure. Development and Change, 36, 449-475. and (iii) transferability of land use rights for longer time Rahmato, D. 2007. A Review of Land Tenure Legislation in periods. Policy action on these areas will be critical to Ethiopia. Addis Ababa: Forum for Social Studies. make certificates sustainable. Ethiopian policy makers recognize the importance of these issues and have start- Holden, S. and T. Tefera 2007, Early Impacts of Land Registration ed to address them on a pilot basis. While our analysis and Certification on Women in Southern Ethiopia: Findings points to a large potential and a very successful start, the from a Household Survey, Report prepared for the Global Land long-term success and sustainability of certification will Tools Network. depend on the extent to which the lessons from such Holden, S., K. Deininger, and H. Ghebru 2007, Land Certification, pilots are learned and implemented on a larger scale. Land Related Investment, and Productivity Impacts, mimeo. The ARD Notes series on Land Policy and Administration aims to disseminate results from research and Bank ESW, describe innova- tive operational practices, or point towards areas meriting further analytical attention. Significant contributions to their publication and content come from the DFID-World Bank land policy partnership, the World Bank-FAO collaborative program, the Knowledge for Change Trust Fund, the Global Land Tools Network, the multi-donor trust fund supporting implementation of the Gender Action Plan, and the Norwegian ESSD Trust Fund. The views expressed are those of the author(s) and not necessarily those of the World Bank Group or supporting institutions. tHE WoRld BanK 1818 H Street. NW Washington, DC 20433 www.worldbank.org/rural