Capital Budget, Recurrent Trap

Totals. Capital demand is 180 + 140 + 60 + 45 = LKR 425 million, against LKR 320 million available. O&M demand is 12 + 10 + 4 + 6 = LKR 32 million, against LKR 22 million available. Both constraints bind, but the recurrent one is tighter relative to the money available: the O&M total is 145 per cent of the new recurrent space, while capital demand is 133 per cent of the capital envelope.

Feasible combinations (capital ≤ 320, O&M ≤ 22):

• Irrigation + roads: capital 320, O&M 22 - exactly on both limits.
• Irrigation + laboratories + clinic: capital 285, O&M 22 - fits.
• Roads + laboratories + clinic: capital 245, O&M 20 - fits.
• Irrigation + laboratories: capital 240, O&M 16 - fits, but leaves LKR 80 million capital uncommitted.
• Irrigation + clinic: capital 225, O&M 18 - fits.
• Roads + laboratories: capital 200, O&M 14 - fits.
• Roads + clinic: capital 185, O&M 16 - fits.
• Any three including irrigation and roads together: O&M 26 or more - not feasible.
• All four: not feasible.

What the arithmetic reveals. The obvious "fund everything technically sound" approach fails. The committee must choose a portfolio, not a project list, and must treat the O&M line as a hard constraint rather than an afterthought.

Recommendation: irrigation + laboratories + clinic (capital LKR 285 million, O&M LKR 22 million).

Criteria.

1. Poverty and equity: poverty is highest in the tank command area, and irrigation supports 2,100 farm families whose incomes depend on the tank. This is the strongest distributional case.
2. Human capital: the laboratories reach 6,000 students and the clinic 3,500 mothers and infants. These are recurring benefits with relatively low O&M per beneficiary: LKR 4 million ÷ 6,000 = about LKR 667 per student, and LKR 6 million ÷ 3,500 = about LKR 1,714 per mother or infant, compared with LKR 12 million ÷ 2,100 = about LKR 5,714 per farm family for irrigation.
3. Budget fit and risk: the portfolio uses LKR 285 million of LKR 320 million, leaves LKR 35 million as contingency, and exactly matches the O&M ceiling. It does not gamble on future recurrent allocations.

The omitted roads proposal. The 22 km road serves 14,000 users and has a strong access case, but adding it would breach O&M by LKR 10 million. Rather than drop it, refer it to the provincial road budget or a different financing window, since rural access roads often have another line ministry or provincial channel. If no alternative funding exists, phase 8–10 km in years four and five, with the O&M cost explicitly carried in the out-year estimates. Do not promise the full road and hope for a supplementary allocation.

Safeguards.

1. An O&M line per asset, not per project. For each completed scheme, record the annual maintenance requirement in the asset register and in the medium-term budget. The tank, laboratories and clinic should each have a named maintaining institution and a costed annual plan.
2. No commissioning without maintenance funding. Make practical completion conditional on the relevant institution confirming that staff and O&M funds are available. This prevents the common pattern of a completed building with no reagents, teachers or repairs.
3. Five-year O&M forecast. Show the recurrent cost of each year's new assets against the projected O&M envelope. The committee should see now that year three or four may be overcrowded, rather than discover it after commissioning.
4. Annual maintenance audit. Publish a simple table: asset, maintenance budget, amount spent, condition rating. This turns maintenance into a measurable performance item.

Reasoning. The district's problem is not a shortage of capital projects but a mismatch between capital and recurrent budgets. Safeguards must therefore operate at the point of commitment, not after construction. If the provincial council later provides additional O&M, the deferred road or a phased extension can be reconsidered through the same test.