Budget expenditure: revenue vs capital, committed spending and why spending grows

Detail

1. The basic split: revenue vs capital expenditure

• Art. 112 requires the government to place an Annual Financial Statement (the Budget) before Parliament. The Budget shows spending on the revenue account separately from spending on the capital account.
• The test: does the spending create an asset, or reduce a debt, for the Centre itself?
• If yes, it is capital expenditure.

• If no, it is revenue expenditure.

• PRS gives the same test. Spending that changes the government's assets or liabilities (for example, building roads) is capital expenditure. All other spending, such as salaries or interest payments, is revenue expenditure [2].

2. Revenue expenditure

• Revenue expenditure: spending that creates no physical or financial asset for the Centre. It is used up within the year.
• What it covers:
• running government departments and services;
• interest payments on past borrowing;
• subsidies (money the government pays so that a product becomes cheaper for users, e.g. food and fertiliser), salaries and pensions;

• grants to states and others, even when the grants build assets.

• The classic trap: a grant for building schools is revenue expenditure.

• The school is an asset of the state, not of the Centre.

• So in the Centre's books the money leaves and no asset comes back.

• Size (2026-27 BE): revenue expenditure is ₹41,25,494 crore. This is 6.6% more than 2025-26 RE (₹38,69,087 crore) [2].

• Subsidies (2026-27 BE): ₹4,54,773 crore, which is 3.1% below 2025-26 RE [2].
• Food subsidy is ₹2,27,629 crore and fertiliser subsidy is ₹1,70,799 crore. Together they make up 87% of all subsidies [2].

3. Capital expenditure

• Capital expenditure (capex): spending that creates an asset or reduces a liability (a debt the government owes).
• What it covers:
• land, buildings, machinery and equipment;
• investment in shares (e.g. buying equity in a PSU);
• loans and advances to states, UTs, PSUs and others. A loan is a financial asset because the money will come back;

• repayment of loan principal, because it reduces a liability.

• Its two parts (2026-27 BE) [2]:

• Capital outlay (spending on physical assets and equity): ₹9,43,042 crore (+6.3%).
• Loans and advances: ₹2,78,780 crore (+33.8%).

• Total capex: ₹12,21,821 crore, which is 11.5% more than 2025-26 RE.

• Loans to states are capital expenditure for the Centre. In 2026-27 the Centre has kept ₹1,85,000 crore as special loans to states for capital expenditure [2].

Item	Revenue or capital?	Why
Grant to a state for building schools	Revenue	The asset belongs to the state, not the Centre
Loan to a state	Capital	It is a financial asset; the money comes back
Interest paid	Revenue	It is only the cost of past borrowing
Repayment of loan principal	Capital	It reduces a liability
Buying equity in a PSU	Capital	It is a financial asset (shares)
Salaries, pensions, subsidies	Revenue	No asset is created

How big is each part? (2026-27 BE)

• Total expenditure is ₹53,47,315 crore [2].
• Revenue expenditure: 41,25,494 ÷ 53,47,315 ≈ 77% of total spending.
• Capital expenditure: 12,21,821 ÷ 53,47,315 ≈ 23% of total spending.
• So about three of every four rupees the Centre spends go on revenue items.

4. Effective capital expenditure

• Effective capital expenditure = capital expenditure + grants-in-aid for creation of capital assets.
• Why it exists: ordinary capex misses assets that states build with central grants, because those grants are booked as revenue expenditure. Effective capex adds them back, so it shows all spending on asset creation.
• Worked example (illustrative numbers):
• Say the Centre's capex is ₹12 lakh crore.
• Say its grants to states for roads, schools and hospitals are ₹5 lakh crore.
• Effective capex = 12 + 5 = ₹17 lakh crore.
• The ₹5 lakh crore still sits in revenue expenditure in the accounts. It is added back only in this wider measure.

5. Committed expenditure

• Committed expenditure: spending the government cannot easily cut in the short run. The main items are:
• interest (the government must pay lenders);
• salaries and pensions (these are legal promises to employees);

• defence (national security leaves little room for large cuts).

• Latest data: about 65.3% of revenue receipts in 2026-27 BE goes on committed expenditure (salaries, pensions and interest) [2].

• This means only about one-third of the Centre's revenue income is left for schemes, subsidies and capex before it has to borrow.

Interest payments: the largest single item

• Interest is paid on market loans, external loans and reserve funds. It is the single largest item of revenue expenditure (NCERT; this item was earlier classed as non-plan).
• Size: ₹14,03,972 crore in 2026-27 BE, up 10.2% over 2025-26 RE (₹12,74,338 crore) [2].
• As a share of the Budget (2026-27 BE):
• 26% of total expenditure [2];

• 40% of revenue receipts [2]. The NCERT-based range was 37-40%.

• Trend: interest as a share of revenue receipts rose from 37% (2013-14) to 42% (2020-21). It is estimated at 40% in 2026-27 [2].

• As a share of GDP: NCERT Table 5.1 gives 3.6% of GDP.
• A rough check with the 2026-27 BE: the fiscal deficit of ₹16,95,768 crore equals 4.3% of GDP [2]. So GDP is about ₹394 lakh crore.

• Then ₹14.04 lakh crore ÷ ₹394 lakh crore ≈ 3.6% of GDP, which matches NCERT's figure.

• Why interest crowds out other spending:

• The government borrows more, so its debt rises.
• A larger debt means a larger interest bill every year.
• Less money is left for health, education and capex.

Defence

• NCERT (Table 5.1): defence on the revenue side is 1.0% of GDP.
• Ministry of Defence (2026-27 BE): ₹7,84,678 crore. This is 15% of total central expenditure and the highest allocation of any ministry [2].
• This ministry figure includes revenue spending, capital spending and defence pensions. That is why it is much bigger than the 1.0%-of-GDP revenue figure.

6. Plan and non-plan: the old classification (abolished)

• The old split:
• Plan expenditure: central Five-Year Plan schemes and central help for state and UT plans.

• Non-plan expenditure: interest, defence, subsidies, salaries and pensions.

• Abolished from 2017-18:

• The Finance Minister announced the change, noting that several committees had questioned the value of the split [3].
• The Cabinet approved merging plan and non-plan in both the Budget and the accounts from 2017-18 [4].
• Earmarking of funds for the Scheduled Castes Sub-Plan and Tribal Sub-Plan was kept [4].

• The Rangarajan Committee (2011) had recommended removing the split and classifying spending as revenue and capital (NCERT scaffold).

• NCERT is outdated here: Table 5.1 and the chapter text still use the plan/non-plan split.

• Why it was dropped (NCERT footnote, confirmed by the Cabinet note):
• The split gave a fragmented view of how money was divided among schemes [4].
• It pushed ministries to start new schemes and to neglect maintenance of existing assets and establishment spending needed for essential social services [4].

• It made non-plan spending look wasteful. This hurt education and health, where salaries of teachers and doctors are the main cost.

• The current presentation (2026-27 BE, ₹ crore) [2]:

Head	2026-27 BE
Centre's own expenditure (central expenditure)	43,57,429
– Establishment expenditure of the Centre	8,24,114
– Central sector schemes (fully funded by the Centre)	17,71,928
– Other expenditure (includes interest ₹14,03,972)	17,61,387
Transfers to states (CSS and other transfers)	9,89,885
– Centrally sponsored schemes (costs shared by Centre and states)	5,48,798
– Finance Commission grants (16th FC)	1,29,397

• Finance Commission grants for 2026-27 (₹1,29,397 crore) are 15% below 2025-26 RE [2].
• Total transfers to states (tax devolution + grants + loans): ₹26,20,769 crore in 2026-27, up 12.2% [2].
• Tax devolution: ₹15,26,255 crore.
• Grants and loans: ₹10,94,514 crore.

7. The capex push

• Central capex as a share of GDP rose from about 1.7% (2019-20) to about 3.1% (2025-26).
• In money terms: public capex was raised from ₹11.2 lakh crore (2025-26 BE) to ₹12.2 lakh crore (2026-27 BE) [2].
• 2025-26 was later revised down to ₹10,95,755 crore (RE), against a BE of ₹11,21,090 crore [2].

• 2024-25 actual capex was ₹10,51,953 crore [2].

• Capex as a share of GDP in 2026-27: 12.22 ÷ ~394 lakh crore ≈ 3.1% of GDP (worked out from [2] figures). (NCERT Table 5.1: 3.2% of GDP.)

• Why governments favour capex:
• Roads, railways and ports cut transport costs for businesses.
• This encourages private firms to invest more ("crowding-in").

• Capex has a higher multiplier than revenue spending (each rupee raises total output by more than one rupee).

• Supporting measures in 2026-27 [2]:

• an Infrastructure Risk Guarantee Fund to give private developers confidence;
• ₹1,85,000 crore of capex loans to states.

8. Social spending

• Social sector expenditure: spending on education, health, social security and welfare.
• It builds human capital (the skills and health of people).

• Its returns come slowly, over the long run.

• Welfare expenditure: tax-funded spending on programmes that improve people's quality of life.

• Class 11 critique (LPG chapter): the post-1991 reforms limited public spending, especially in the social sectors.
• Accounting note: most social spending is revenue expenditure.
• Salaries of teachers and nurses are revenue items.
• Grants to states for schools and hospitals are also revenue items.

• So a rule that only "cut revenue expenditure" can end up hurting health and education.

• Example (2026-27 BE): the rural employment guarantee scheme (VB-G RAM G, which replaced MGNREGA) got 42.8% more than 2025-26 RE [2].

9. Why public spending keeps growing

• Wagner's law (law of increasing state activity):
• Public spending grows faster than national income as a country industrialises.
• Reasons: more regulation is needed, cities need services (water, transport, policing), and people demand more education and health care as incomes rise.

• Indian sign: total central spending rose 7.7% in 2026-27 BE [2], close to the assumed 10% nominal GDP growth [2]. The long-run rise in the state's role fits Wagner's idea.

• Peacock-Wiseman hypothesis (displacement effect):

• Spending rises in steps, not smoothly.
• It jumps during wars and crises, because people then accept higher taxes and borrowing.
• After the crisis, spending does not fall back to the old level. New spending habits and the debt taken on keep it high.
• Indian example: COVID-19 in 2020-21.
  • Spending and borrowing jumped.
  • Interest as a share of revenue receipts peaked at 42% (2020-21) [2].
  • The fiscal deficit was still 4.8% of GDP in 2024-25 [2].
  • It is targeted at 4.4% (2025-26 RE) and 4.3% (2026-27 BE) [2].
• The central debt target is about 50% ±1% of GDP by March 2031, against 55.6% of GDP in 2026-27 [2].

Prelims Hooks

• Grant to a state for building assets is revenue expenditure for the Centre. A loan to a state is capital expenditure.
• Repayment of loan principal is capital expenditure. Interest paid on the same loan is revenue expenditure.
• Effective capital expenditure = capital expenditure + grants-in-aid for creation of capital assets.
• Interest payments are the largest single item of revenue expenditure. They are about 40% of revenue receipts and 26% of total expenditure (2026-27 BE) [2].
• Committed expenditure (salaries, pensions, interest) takes about 65.3% of revenue receipts (2026-27 BE) [2].
• The plan/non-plan classification was abolished from 2017-18 [3][4]. The Rangarajan Committee (2011) had recommended it. SCSP/TSP earmarking continues [4].
• The Ministry of Defence has the largest single allocation: ₹7,84,678 crore, 15% of central spending (2026-27 BE) [2].
• Capital expenditure has two parts: capital outlay (₹9.43 lakh crore) and loans and advances (₹2.79 lakh crore), 2026-27 BE [2].
• Wagner's law: state spending grows faster than national income. Peacock-Wiseman: spending rises in steps through the displacement effect of wars and crises.

Mains Points

• Quality of spending matters as much as its size.
• The capex share has risen: ₹12.2 lakh crore, about 3.1% of GDP (2026-27 BE) [2].
• But committed spending takes about 65% of revenue receipts [2], so there is little room left for new priorities.

• Cutting the interest burden needs steady fiscal consolidation (lowering deficits and debt step by step) toward the 50%-of-GDP debt goal by 2031 [2].

• The revenue/capital label can mislead.

• Grants to states for schools and health centres, and salaries of teachers and doctors, are "revenue" spending.
• Yet they build human capital.

• So cutting the revenue deficit alone can squeeze the social sectors. Effective capex is a fairer measure of asset creation.

• The fiscal-federalism angle (GS-II):

• The Centre now spends through ₹1.85 lakh crore of capex loans and centrally sponsored schemes, which rose 30.6% [2].
• At the same time, Finance Commission grants fell 15% [2].

• This raises the debate between tied transfers (money that must be spent on schemes the Centre chooses) and untied transfers (money states may spend as they decide).

• Why spending does not come back down:

• Displacement effects (Peacock-Wiseman) and Wagner's law mean crisis-era spending tends to stay.
• This is why rule-based limits such as the FRBM Act 2003, and clear medium-term targets, are needed to bring deficits back down.