Will Better District-Level Data Help Address Regional Inequalities?

At the 11^th Governing Council meeting of NITI Aayog, Prime Minister Narendra Modi urged state chief ministers to undertake estimation of gross domestic product at the district level and track their progress. Such district domestic product (DDP) is expected to help uncover lagging districts across India’s approximately 800 such administrative units, and encourage better policy formulation to overcome developmental gaps.

Sub-regional calculation of GDP – the calculation of the value of goods and services produced in a small defined area over a specific period – is conducted by various countries and groupings, including the US at the country level, the EU under its NUTS 3 (Nomenclature of Territorial Units for Statistics, level 3) geocode standard, and the UK for local authority districts. Among large developing countries, China, Brazil and Mexico calculate GDP for certain municipalities but the data quality is uneven. India’s venture to develop the methodology for DDP is thus an ambitious and inclusive project, rightly prioritised by the Ministry of Statistics and Programme Implementation (MOSPI).

MOSPI brought out draft uniform guidelines for compiling gross state value addition and DDP on April 7, 2026. This was part of the exercise to update the base year of national accounts to 2022-23, and was followed by the issue of final guidelines on June 3, just ahead of the Governing Council meeting. Earlier, 26 states and UTs practiced the top-down approach for estimating DDP, wherein state GDP estimations were distributed across districts based on selected indicators such as population. However, this was not comparable across states and sectors, providing only a rough estimation that often masked intra-state deficiencies.

The new bottom-up framework aims at harmonisation and consistency in data tracking across the country. In its guidelines, MOSPI recommends states to use improved data sources to estimate the primary, secondary and tertiary sector production value through administrative and survey processes or, when not possible, through top-down measures. Updated sources such as the Annual Survey of Industries (ASI), Annual Survey of Unincorporated Sector Enterprises (ASUSE), Periodic Labour Force Survey (PLFS), Goods and Services Tax and other existing and upgraded databases are to be deployed by states to compile the data at the district level.

The concept of backward districts for targeted developmental activities is not new. It has been considered an apt measure for fostering equitable growth right from the beginning of the five-year plans process. Various area-based programmes have been launched over the years, including for drought-prone, hilly, tribal, border and other regions. The Rashtriya Sam Vikas Yojana of 2003-04 identified almost 150 districts as backward, and requiring special attention. Subsuming this programme, the Backward Regions Grant Fund came about in 2007 to support delivery of schemes and promote planning at the panchayat level in over 250 districts.

With the end of the Planning Commission and the launch of NITI Aayog, backward districts assumed a new nomenclature under the Aspirational Districts Programme (ADP) in 2018, covering 112 districts for special measures. NITI Aayog tracks these districts monthly for their progress on 49 key performance indicators relating to healthcare, education, agriculture, financial inclusion and infrastructure. Significant progress has been noted in these indicators in underdeveloped districts, especially in the health and agriculture sectors.

The experience of backward districts or aspirational districts over the years implies that much more needs to be done towards equitable growth, rather than merely identifying districts and datapoints.

States track their district data through directorates of economics and statistics. As per the Economic Survey 2024-25, Rangareddy district in Telangana enjoys the highest GDP per capita, followed by Gurugram (Haryana), Bengaluru Urban (Karnataka), Gautam Budh Nagar (Uttar Pradesh), Solan (Himachal Pradesh), and districts in Goa, Sikkim, Maharashtra and Gujarat. The differential between the richest and poorest districts within a state often skews its overall position in state rankings. For example, Gautam Budh Nagar hosts major global companies, and has emerged among the top districts for exports and FDI, thus pulling up the rank for Uttar Pradesh as a whole.

The MOSPI guideline, when implemented across states, would greatly improve the estimation of DDP and thereby contribute to evidence-based policymaking. It would provide granular information on key macro indicators that can be expected to yield standardised and reliable estimates of economic output to feed into decentralised planning.

It would also contribute to improve coverage of India’s vast informal sector and identify district-level economic hurdles and shocks, such as closure of factories or area-based excessive rains and droughts. Financial allocations can also be made in an informed manner. Another advantage would be to evaluate and monitor districts’ progress in development initiatives.

Several challenges would need to be addressed for the all-round success of the DDP initiative. Implementation would rest on resolving data gaps persistent at the district level, capacity of the state government statistics departments to maintain a high-level of granular data, and correctly assessing self-employment and the informal sector where employment is often flexible across sectors. For example, a farmer on a small plot may shift between agricultural work and retail activities in a single year.

The uniform guidelines for DDP mark a significant step towards better data on the economy, and, if executed well, could support evidence-based, decentralised policymaking and targeted interventions. The central and state governments must continue to collaborate strongly and invest in data collection systems and facilities, but with the understanding that building a geographically equitable India will require much more than just data.