Bangladesh’s Land Records Must Be Digitally Integrated

Bangladesh does not suffer from a shortage of land records. In fact, the same piece of land may appear in a cadastral map, a Record of Rights or khatian, a registered deed, a mutation record, a Land Development Tax account, an acquisition file, a local-government holding record, a housing-authority plot register, a bank's mortgage file and, increasingly, several digital systems.

The problem is that these records were not originally designed to work as parts of one land-information system. That distinction is important. Different generations of land administration created different records to meet the needs of their own time. Survey, tenancy and revenue administration, registration, taxation, housing, acquisition and local government consequently developed their own views of land.

Why were the records compiled separately? The answer begins with the legal history of land administration. The Survey Act, 1875 provides for the survey and demarcation of land; section 41 directs the Collector, in determining a disputed boundary, to determine it according to ‘actual possession’. Survey was therefore developed within a historical administrative system concerned with identification of land, estates, tenures, possession and boundaries. It was not supposed to be the geospatial foundation of a twenty-first-century multipurpose land-information infrastructure.

The State Acquisition and Tenancy Act, 1950 emerged in a different institutional context. Among other things, it provides for the preparation and revision of Records of Rights. Section 144A gives entries in a Record of Rights prepared or revised under section 144 evidentiary and presumptive value: an entry is presumed correct until evidence proves otherwise.

This distinction is significant. A khatian is an important statutory land record, but it should not automatically be equated with a state-guaranteed title. Government land-administration materials treat the khatian as the Record of Rights and link its maintenance to changes in ownership, possession and land-revenue administration. Registration of land evolved under another legal framework. But these laws did not originate from a single conceptual model of a parcel that would require every institution to maintain a common identity, geometry and lifecycle for the same land. That historical fact helps explain today's fragmentation.

A national parcel identifier, geospatial reference framework, land-data dictionary, interoperability standards, APIs, event notifications, data provenance and audit trails can allow one institution to consume verified information from another without duplicating institutional mandates.

One land, many institutional views
Over time, the number of institutions requiring land information expanded dramatically. The Ministry of Land and its field administration need land records for administration, mutation, settlement, revenue and related services. DLRS produces survey records and cadastral maps. Registration authorities record instruments affecting immovable property. Deputy Commissioners maintain acquisition cases. Local-government institutions maintain holdings and taxation-related information. Housing authorities maintain their own plot, allotment and ownership-related records. Banks maintain mortgage and collateral information. Planning agencies need land-use and zoning information.

Each institution has a legitimate statutory or operational purpose. The weakness is horizontal. Bangladesh never fully developed a common mechanism through which all these institutions could recognise that they are describing different attributes, rights, transactions or uses relating to the same physical parcel. A housing-authority plot number, municipal holding number, cadastral dag, mutation reference and registered-deed description may consequently coexist without a permanent digital relationship among them. This is no longer merely a land-record problem. It is a national information-architecture problem.

A common language for land information
International standards now provide useful conceptual tools.The ISO 19152 series, known as the Land Administration Domain Model or LADM, provides an international reference framework for describing land administration. LADM does not dictate a country's land law, nor should Bangladesh simply import it. Bangladesh needs a common understanding of the parcel as a legal, spatial, administrative and economic object.

Asian countries have already begun this transition
Bangladesh does not need to design this transformation in isolation. Several Asian countries offer useful lessons, particularly those that began with fragmented or historically inherited land-record systems.

India offers a particularly relevant regional comparison. Its Digital India Land Records Modernization Programme (DILRMP) goes beyond computerised Records of Rights to include cadastral-map digitisation, RoR�"map integration, registration-office computerisation, registration�"revenue connectivity, survey/resurvey and geo-referencing. India later introduced ULPIN/Bhu-Aadhaar for geo-referenced land parcels. The key progression is from digitised records towards parcel identification, spatial integration and information sharing.

Punjab’s World Bank-supported PULSE programme is especially relevant because Pakistan and Bangladesh share similar inherited revenue and cadastral traditions. PULSE combines digital land records, cadastral maps, deed digitisation, urban records, systematic property registration and a spatial framework for land information. It also seeks to incorporate records held by municipalities, development authorities and other entities into a more integrated, parcel-oriented system.

Sri Lanka moved from traditional deed registration towards parcel-based title registration under the Registration of Title Act, 1998 and the Bimsaviya programme. Its system links title registration with unique parcel identification and cadastral information on location, boundaries and extent, while the Survey Department maintains the cadastral information supporting the title-registration process. Its experience also illustrates the institutional challenge involved in moving from deed registration towards a parcel-based title system.

Thailand’s long-running land-titling programme similarly combined cadastral surveying and mapping, title issuance, land-registration improvement and institutional strengthening. Although these countries differ legally and institutionally, the common direction is clear: land administration is moving from isolated records towards parcel-based, geo-referenced and increasingly integrated land-information systems.

From land records to a land-information ecosystem
Bangladesh's future land architecture therefore needs to connect several dimensions of the same parcel: Parcel Identity, Spatial Identity, Rights and Interests, Transaction History, Record History, Fiscal Relationship, Planning and Development, and Economic Information. A parcel is not only a legal and spatial object. It is also an economic asset. Government needs valuation information for acquisition, taxation, public-land management and infrastructure planning. Banks need reliable property information for collateral assessment. Investors need it for investment decisions. Local governments need it for urban management. Yet valuation and economic information can themselves become additional silos if they cannot be associated with the same parcel identity and spatial framework.

The future land-information architecture should therefore allow rights, location, use and value to meet around the parcel, while preserving the statutory responsibilities, access controls and confidentiality requirements of the institutions producing each dataset.

What changes is that these institutions agree on the common objects they share. A national parcel identifier, geospatial reference framework, land-data dictionary, interoperability standards, APIs, event notifications, data provenance and audit trails can allow one institution to consume verified information from another without duplicating institutional mandates. The objective is not necessarily to create one database. It is to make different authoritative records capable of recognising and communicating about the same land.

Each was built for a purpose. The problem arose because the purposes multiplied while the underlying land remained the same. Bangladesh's land records were built separately because different laws and institutions needed different information at different points in history. That history should be respected, not erased. But the next generation should be built differently. The records can remain institutionally distributed. The unique land they describe must finally become connected.