ADP Project Deliverability

The Pipeline, Not the Budget

Bottom line: the constraint on Bangladesh's public investment is delivery, not approval. Across the 1,356 approved projects in the 2025-26 Revised ADP, Tk 1,798,749 crore of cost has been approved and Tk 783,912 crore ever spent, an execution rate of 43.6%. This year's allocation of Tk 166,900 crore is 9.3% of the approved stock, so on current budgeting the portfolio as approved would take decades to complete.

674 projects, 49.7% of the portfolio, are past their own scheduled completion date and still short of approved cost. They carry Tk 648,772 crore of approved commitments. A schedule that roughly half the pipeline misses is not a schedule; it is an accounting convention. The implication for planning is direct: any completion date quoted from this table should be read as a lower bound, and any projection that assumes on-time delivery is assuming something the portfolio has not achieved for half its projects.

290 projects record no expenditure at all. These are approved, costed entries carrying no delivery. Whatever the individual reasons, a pipeline that approves faster than it can start is generating commitments rather than assets, and each one occupies both fiscal space and administrative attention.

Where the Money and the Weakness Sit

The portfolio is concentrated: the ten largest implementing agencies hold 56.7% of approved cost. That concentration is useful, because it means execution reform is a problem of a manageable number of institutions rather than a system-wide one. Among those ten, the weakest execution sits with সড়ক ও জনপথ অধিদপ্তর (36%), ঢাকা ম্যাস ট্রানজিট কোম্পানি লিঃ (27%), বাংলাদেশ পানি উন্নয়ন বোর্ড (বিডব্লিউডিবি) (38%).

Foreign-aided projects account for 57.3% of approved cost. Schedule slippage is more expensive here than in domestically financed work: commitment charges accrue on undisbursed balances, and disbursement-linked financing converts delay directly into a smaller envelope.

What Would Change the Picture

1. Stop approving into a full pipeline. Make new approvals conditional on retiring or restructuring overdue projects in the same agency, so the stock of commitments cannot keep growing faster than delivery capacity.

2. Publish a project-level delivery record. Every figure in this section is computed from a table the government already maintains. Publishing execution and schedule performance per project, per agency, quarterly, costs nothing new and puts the number that matters in front of the people who control it.

3. Triage the zero-spend set explicitly. Each of the 290 projects with no recorded spend should be either started, restructured, or cancelled within one budget cycle, with the decision recorded.

4. Treat schedule as a financing variable. For the foreign-aided 57.3% of the portfolio, model commitment charges and disbursement loss from slippage and report them alongside project cost, so delay has a visible price at approval time rather than only at completion.