“In the 21st century,” as I write with my co-authors in the companion piece to this article, “land administration systems will nearly always be developed using digital systems. Assuming they are well designed, then they are more secure, more efficient, more transparent, and more accessible.”

But the test of any tool is with users in the field. This axiom of development is borne out by DAI’s experience implementing a land tenure registration activity in Tanzania, where a pilot of the open source Mobile Application for Secure Tenure (MAST) has yielded important lessons that we are now incorporating into the U.S. Government’s Feed the Future Land Tenure Assistance (LTA) activity.

In Tanzania, the United States is supporting investment in agriculture to fight hunger and improve food security. Achieving this objective requires formalizing and strengthening rights to land and other productive assets. While Tanzania’s legal framework provides protections for land users, village-level land tenure can be insecure and vulnerable to outside interests. Land in many villages is often neither mapped according to use nor individually demarcated and legally registered. There is also disparity in how domestic and international investors access land, which may lead to illegal transactions or land grabs. These factors constrain economic growth and investment.

LTA, a four-year project launched in December 2015, is designed to reduce land tenure-related risks and lay the groundwork for sustainable agricultural investment for both small holders and commercial investors. It focuses on the Southern Agricultural Growth Corridor of Tanzania and specifically the districts of Iringa and Mbeya, centered in Iringa in years 1 and 2 and extending to Mbeya in years 3 and 4.

LTA has three main objectives:

• Assist villages in completing the land use planning process and delivering individual Certificates of Customary Rights of Occupancy (CCROs) through the use of MAST;
• Build the capacity of village and district land governance institutions to complete land use plans, issue CCROs to village land owners, manage land resources, respect women’s land rights, and build agriculture-related business skills; and
• Raise awareness and uptake of the MAST technology within the Government of Tanzania, civil society, academia, and the private sector.

LTA’s initial design was informed by a 2015–2016 pilot that tested MAST as a field tool for capturing and mapping parcel boundaries and attribute information—the claimant’s name and details—on an image background. Using mobile devices (in this case Android-based), the pilot tested MAST’s ability to formalize land for subsequent registration and delivery of titles in three villages in Iringa.

Following completion of the pilot in June 2016, LTA was given a broader mandate to develop rapid, low-cost, transparent, and replicable procedures for land use planning, first registration, and subsequent land transactions. Implicit in these ambitions is the need to streamline procedures and service delivery at village and district levels so as to achieve full public buy-in at the lowest possible cost. Opportunities to integrate cost recovery and potential revenues are to be explored throughout the course of the work.

MAST will play a key role in clarifying and registering land rights at scale and completing village land use plans. We will extend the outputs of MAST (individual CCROs) to provide a sustainable solution that can capture, manage, and maintain land rights and support transactions at district and village levels. For the latter, DAI is developing the Technical Register under Social Tenure (TRUST), which is intended to link seamlessly with the (improved) MAST functionality to provide for ongoing maintenance, transactions, and long-term sustainability.

Guiding Principles

Are MAST and by extension TRUST fit for purpose? LTA has established the following core principles that our technology must support and to which it must adhere:

• LTA will clarify land rights through community/local attestation of boundaries and claims, and subsequent registration;
• The process will be public, participatory, and transparent;
• It will allow all land users to record a claim, then determine whether title can be granted;
• The decision to accept or deny title will be based on evidence;
• The adjudication and award of title will be justly administered, with equal access to all;
• The systems used will comprise a model based on clear procedures that can be adapted, replicated, and scaled; and
• These systems will be manageable and sustainable at district and village levels.

The MAST pilot—reviewed in the LTA Inception Report and a separate needs assessment—made it evident that if these guiding principles are to be followed, LTA would have to make significant improvements and customisations to MAST. To take just one example, the use of small hand-held devices in the field is problematic because the screen is effectively restricted to a single operator—with the added difficulty of reading in the glare of strong sunlight—which hinders people from gathering around and agreeing on what they’re seeing, in turn reducing participation and understanding on the part of villagers and adjudicators.

Typically, we can address this challenge by printing out and/or posting the material in question, but the initial design of MAST did not allow for the printing and display of maps and attribute lists for public consultation. Other obstacles included the stability of internet connections and power supply in rural Tanzania, a material hindrance to the use of cloud-based platforms which in turn affects the management of data and devices in the field.

Certain issues such as the low quality of input data—and hence low-quality outputs—which in turn necessitate time-intensive “data cleaning,” are not necessarily technology problems. They have to do with the training of field operators and adjudicators. But they are exacerbated if the technology and associated processes are not conducive to accurate, participatory inputs; do not readily enable screening of inputs or prevent duplication of entries; and do not readily accommodate sorting and editing. Based on the needs identified through the pilot, LTA is making the following changes to MAST:

• Enhancing the accuracy of the mapping software and improving the training of field operators;
• Upgrading the parcel numbering system and display to support mapping and work in the field;
• Refining attribute editing capabilities to prevent erroneous entries in the field and back office;
• Revising the attribute process to make individual claimants unique, thereby allowing for better management and analysis of patterns in land holdings (such as gender ratios, the number of multiple parcel holders, family links, and so on);
• Tightening claimant attribute data entry criteria to enable claims to be checked more effectively prior to titling;
• Modifying claimant attributes so that attestation of claims, disputed claims, boundary issues, unclaimed land, and planning breaches can be properly recorded and addressed;
• Adding editing and sorting functions to enable more effective data quality checks and claimant searches; and
• Enabling more effective batch processing and printing—for public display, consultation, adjudication, monitoring and evaluation, and publication of basic statistics and reporting. Eventually, LTA will require production of maps, titles, and field issuance procedures on an industrial scale.

In addition to these changes in MAST, LTA is working to address factors such as capacity in the district and village registries, and the availability of secure and appropriate floor space. In the short to medium term, as the modifications to MAST are rolled out and tested anew, LTA will move forward the current, paper-based systems and procedures in a phased transition to a more automated approach. This gradual introduction of TRUST into village and district land administration, appropriate for the current levels of support infrastructure, will be critical to longer-term sustainability.

Trials of the new version of MAST will commence in May and June 2017; detailed work on the design and implementation of TRUST begins in July. Overall, we must concede that the review and redesign of the MAST software has slowed LTA’s progress in the field and indeed raised questions about the effectiveness and cost of using mobile devices at scale in rural Tanzania. But that iterative process has turned out to be essential, and still guides our efforts to ensure the project and related processes and procedures can function at scale.

The overall procedural requirements for large-scale, participatory land use planning and registration at the village level are well known—proven models exist. It is incumbent on LTA to demonstrate that MAST can add significant value to field data capture and TRUST to sustainable registry maintenance systems. We are pushing hard for those outcomes and look forward to reporting on our progress.