Bangladesh has been given three more years for LDC graduation. The UN Committee for Development Policy has recommended deferring our graduation from least developed country status to November 24, 2029. The General Assembly is due to take the decision before November this year.
The Committee attached a condition in its own words, that the extension should not be viewed as an opportunity to delay reforms but should accelerate them. The question that follows is a plain one: What are the three years for?
My answer is regional economic development, and I do not mean it as a matter of fairness between divisions. I mean it as the fastest route we have to a diversified export base, to jobs at the scale the country needs them, to women's employment that does not require a woman to move house, and to growth that adds less to the environmental load we already carry.
We usually pursue these as four programs with four budgets. But they meet in one place, which is the region.
By regional development, I do not mean spreading projects more evenly across divisions. I mean treating a division as an economic unit with a declared specialization rather than as an administrative area that receives its share of everything. That has four parts: (i) each division carries a declared economic identity based on what its enterprises already do best; (ii) public investment there is sequenced to that identity instead of spread thinly across every sector; (iii) the support functions the identity needs, being finance, standards, testing, skills and logistics, are built at the cluster rather than in the capital; and (iv) one institution is answerable for the division's economic outcome and is measured on it.
The first part should not be difficult, as our productive geography is already specialized. The specialization is also sharper than our policy acknowledges. In the agro-logistics system mapping we undertook for the Embassy of the Kingdom of the Netherlands this year, we found that the top three mango producing districts account for roughly 50% of national output, and the top three papaya districts for 43%. Watermelon has been shifting into Patuakhali, Bhola, and Barguna, where the top three districts now hold around 30%.
The pattern holds well outside agriculture. Bogura carries around 70 foundries and 1800 manufacturing and workshop units. These meet some 70% of national demand for agricultural machinery and spare parts, are worth Tk500 crore a year, and employ about 15,000 people. Bhairab in Kishoreganj holds over 10,000 shoe making units across some 20 villages. The trade there supports close to two lakh people and turns over Tk15-20 crore a day.
None of this was planned. These are comparative advantages that enterprises built themselves, and no policy has yet been written to them.
Writing one would look concrete, and it would look different in each place. Rajshahi would be developed as a horticulture economy. The packhouses, hot water treatment plants and phytosanitary certification that mango and papaya exports require would be sited there rather than at the airport in Dhaka.
Bogura would be developed as a light engineering economy. It would need metallurgical testing, tool and die capability, industrial training, and working capital lines, as these are what let a foundry district move from spare parts into finished machinery it can sell abroad.
Bhairab would be developed as a footwear economy, with component supply, design services, and the standards testing that an international buyer asks for before placing an order.
Each of these is a different economy and each needs a different set of institutions. That is not what we build.
What we build is one national system, and we build most of it near the capital. Our logistics costs stand at 15-20% of the GDP, which is amongst the highest in Asia. This cost falls hardest on the production areas that are farthest from the capital. Nearly 95% of our cold storage is dedicated to potatoes.
Producers of fresh vegetables are thus forced to send their produce fast to urban markets. The highways cut travel times to those markets by 30-40%, but did not address the issue of preservation. We lose 6.0-8.1 million metric tons of cereal a year, worth $2.1-2.8 billion. A 30% reduction would save $630-850m and feed 15-20m people without bringing another hectare into cultivation.
The environmental case is not separate from this. We have little new land or water to bring into cultivation. Additional food will therefore have to come from what we stop losing.
Activities that we will require to overcome this loss would also push the frontiers for jobs. Grain spoils as nobody dried, graded, or stored it, and that work would be employment. Agriculture contributes about 11% of GDP but employs 43-44% of the labour force. Most of our workforce is thus in the part of the economy that produces the least value per worker.
Only around 12% of our agricultural output is processed. The work that would close the gap is drying, chilling, grading, packing, machining, assembly, and stitching. It belongs in the cluster rather than in Dhaka.
This matters most for women. Our garment industry brought women into waged work, but it required them to move to Dhaka, Gazipur, Narayanganj, and Chittagong. The move cost them in the form of rent, travel, and separation from their families. Work sites at the cluster do not ask that price, as they are available where women already live and where their labour is often already offered without wages. A quarter of the Bogura light engineering workforce is already made up of women, despite the fact that it was not designed to employ women.
The question therefore is, why does economic development remain centralized? It is because we are yet to have the institutional system to support regional development. There is no institution that owns a region's economic outcome. Our Annual Development Program is allocated by ministry and by project, never by region. Nobody is therefore answerable for what Bogura's or Barishal's productive base looks like in five years.
The Local Government Division takes the largest share of the ADP and carries no economic development mandate. We have already tried the alternative of announcing sites. Of the 100 economic zones announced, roughly a tenth reached the ground in a decade before the program was cut back to five.
A zone is a site. A site with neither an owner nor a declared purpose fails to become an economy.
Other countries in the Global South have given their regions an owner, and I have watched this at close quarters in the work I do outside Bangladesh. In Nigeria, every state carries a declared economic identity. Lagos is the Centre of excellence and Kano the Centre of commerce. Those declarations came first, and the instruments were built to them afterwards, most recently Lagos's industrial policy and the state-level agro-industrial processing zones.
The declaration is what made the investment legible to everyone who had to act on it. Vietnam supplied the piece that Nigeria does not, which is measurement. Its Provincial Competitiveness Index gives every province a published score on economic governance, and provinces compete on it openly. We have neither the declaration nor the score.
Both are available to us within the current fiscal year. The FY2026-27 Annual Development Program is Tk3 lakh crore, 30% above last year. It already names region-based balanced development amongst its five pillars, with the northern region, the coastal belt, Chittagong, and Mongla named specifically.
The intent is on paper and the money is appropriated. What is absent is the machinery.
The government should declare an economic identity for each division on the comparative advantage its enterprises have already built. It should name a single owner for that identity within the Planning Commission. It should then publish a divisional measure the owner is held to.
Tk74,000 crore of this year's ADP sits in block allocation, undirected; that is the instrument. Routing it through the divisional plans would cost nothing that has not already been budgeted.
How this should be done -- meaning the sequencing, the institutional design, and the financing -- needs more space than a column allows, and I will set it out separately. What I am arguing here is the prior point: The decision to organize the economy regionally has to be taken in this fiscal year for any of that design to matter. Three years is not long for this kind of work. If we begin in the next budget rather than this one, we will reach November 2029 with the structure half built and the preferences gone.
The Committee for Development Policy was right that the extension is not an opportunity to delay. It is the time available to build in, and what we should build is regional.
Md Rubaiyath Sarwar is Managing Director, Innovision Consulting.
