How much of its budget does each ministry and division spend, which under-spend year after year, and how much spending bunches into the last quarter?

Not held

• Which budget figure a low rate reflects: The quarterly editions print the budget in force (headed 'Revised Budget' in every Fourth-quarter edition held); where the Yearly edition is held (FY2022-23, FY2024-25) its 'Total Budget' column is served beside it as yearly_edition. Some divisions' printed revised budget is far above their Yearly 'Total Budget' (Planning Division, code 114, is the extreme case), and the source does not say why, so a low rate there may reflect a block allocation spent elsewhere rather than slow spending.
• One annual actual per division: The sum of the four quarterly editions and the Yearly edition's printed actual agree within 3 thousand taka for every division in FY2022-23 but differ for some in FY2024-25; the year's actual here is always the quarterly sum, and yearly_edition.actual_crore shows the Yearly print.
• Budget against actual before FY2020-21 and after FY2025-26 Q3: The pinned iBAS++ table holds the quarterly editions FY2020-21 Q1 to FY2025-26 Q3 only.
• Development and operating budget separately: Statement-2 prints one budget figure per division; only actual expenditure is split into operating and development, so the development share is of spending, not of budget.
• ADP rows IMED printed but the readers did not agree on: The IMED table keeps only rows both readers agreed on, so a year-end report can hold fewer ministries than IMED printed (FY2024-25's July-June report holds 33). ADP totals here are sums over the rows held, not IMED's national total.
• Body links for divisions without a reviewed crosswalk decision: Code 163 (Supreme Court Secretariat, printed from FY2025-26) and IMED rows whose printed name matched no registry body carry body null.
• Causes of under-spending: The sources print amounts, not reasons; a low rate can be a late release, a cut, or a project delay.

Sources