Skip to content
FreeFinance

Where the money comes from

Tax is the big one and the figure in the accounts is already net of what the states take. Non-tax revenue was at 72% of its annual estimate at the halfway point, which is a story about one cheque rather than about buoyancy.

Chapter 3 · Intermediate · Updated 3 November 2026

Three sources, in descending order of size and ascending order of surprise.

The receipts, in full

April to September 2026, in ₹ crore:

Budget Estimate Actual to Sept % of BE Same point last year
Tax revenue (net to Centre) 28,66,922 12,76,088 44.5% (43.3%)
Non-tax revenue 6,66,228 4,80,732 72.2% (79.9%)
Revenue receipts 35,33,150 17,56,820 49.7% (49.6%)
Recovery of loans 38,397 16,443 42.8% (39.1%)
Other receipts 80,000 62,159 77.7% (49.8%)
Non-debt capital receipts 1,18,397 78,602 66.4% (45.8%)
Total receipts 36,51,547 18,35,422 50.3% (49.5%)

Tax is 78% of the total at Budget Estimate. Everything else in this chapter is about the rest.

The word doing the most work: "net"

The tax line says Tax Revenue (Net), and that single word changes what the number means.

India's Union government collects income tax, corporation tax, GST, customs and excise. It does not keep all of it. A share of the divisible pool is transferred to the states, on a formula set by the Finance Commission for a five-year period, and that share leaves before this line is struck.

So ₹12,76,088 crore is not what the Centre collected. It is what the Centre collected and kept, after the states' share was devolved. Gross collections are considerably larger, and they are reported separately.

Three consequences.

Do not compare this to a gross collections headline. When a news report says GST collections hit a record, that is a gross figure. The number in these accounts is net, and the two differ by the devolution.

A good collections month is not fully a good month for the Centre. Higher collections mean a higher transfer to the states as well, so the Centre's net receipt rises by less than the gross.

The states' finances are the other half of Indian public finance, and these accounts do not cover them at all. A statement about "government borrowing in India" that uses only the Union figures is missing the states, whose combined deficits are of comparable scale and which this subject does not reach.

Tax at 44.5%, which is behind

Six months of twelve is 50% of a year. Net tax revenue at 44.5% is running 5.5 points behind a pro-rata pace.

That is not necessarily a shortfall, for a structural reason worth knowing: tax receipts in India are back-loaded. Advance tax instalments fall in a pattern that puts more in the second half, the final instalment lands in March, and compliance and assessment activity concentrates at the year end. A tax line at 44.5% in September is normal rather than weak.

The comparison that controls for seasonality is the column on the right: 44.5% this year against 43.3% at the same point last year. Slightly ahead of its own prior-year pace. Which is the honest reading, and it requires the year-ago column rather than the pro-rata intuition.

Non-tax at 72.2%, which is one cheque

Now the line that looks spectacular and is mostly arithmetic.

Non-tax revenue is dividends and profits from public sector undertakings and the Reserve Bank, interest on loans the government has made, spectrum and licence fees, and the receipts of government services. At 72.2% of its annual estimate by September, it is 22 points ahead of pro-rata.

The dominant reason is the surplus transferred by the Reserve Bank of India, which is declared once a year after the Bank closes its accounts and lands in a single tranche. A receipt that arrives in one lump does not accrue evenly, so its share of the annual estimate jumps in whichever month it is paid and then flatlines.

So "non-tax revenue is at 72% of target" is not evidence of a broad revenue surge. It is evidence that a large annual transfer has already been received, and the remaining months will add comparatively little.

Note too that it was 79.9% at the same point last year — higher. On the measure that controls for the lumpiness, non-tax revenue is running behind. The line that looks strongest in the table is the one weakening fastest, and only the year-ago column shows it.

Non-debt capital receipts, and the honest name

Two items. Recovery of loans — money coming back from states and others the Centre lent to. And other receipts, which is, in practice, mostly proceeds from selling stakes in public sector companies.

₹62,159 crore against a Budget Estimate of ₹80,000 crore, or 77.7%, against 49.8% at this point last year.

Two things worth saying about this line.

It is a sale of assets, not income. Selling a stake converts an asset into cash; it does not create anything. Chapter 1's test applies — it does not make the government better off, it changes what it owns. This is why these receipts are counted in the deficit calculation but sit on the capital side.

It is the least predictable line in the accounts, because it depends on market conditions and on decisions about what to sell and when. A government short of revenue can reach for it, and a disinvestment target met early says nothing about the following six months.

What the receipts side says, and does not

It says: total receipts are at 50.3% of the Budget Estimate at the halfway point, marginally ahead of pro-rata and of the 49.5% at this point last year. The aggregate position is close to plan.

It does not say that the components are. Tax is behind pro-rata and ahead of last year; non-tax is far ahead of pro-rata and behind last year; asset sales are well ahead of both and are the least repeatable of the three.

An aggregate that is on plan while its components diverge is the ordinary condition of a government's accounts, and reading only the aggregate is the commonest way to miss what is happening underneath it. That is the same lesson the inflation subject reached about a basket and the growth subject about an expenditure composition, which by this point in the course should be starting to feel inevitable.

The point

The Union government's receipts are net tax revenue, non-tax revenue and non-debt capital receipts, and the first is already net of the states' share, so it is not what the Centre collected. For April to September 2026, total receipts were at 50.3% of the Budget Estimate — on plan in aggregate. Underneath that, tax at 44.5% is behind pro-rata but ahead of last year because receipts are back-loaded; non-tax at 72.2% is far ahead of pro-rata because the Reserve Bank's surplus arrives in one lump, and is behind last year's 79.9%; and asset sales at 77.7% are the least repeatable line in the table.

Check yourself

4 questions. Every answer is explained afterwards, including the ones you get right — guessing correctly is not the same as knowing. Score 70% or more and the chapter is marked done.

Question 1 of 4

AccountingModerate
The accounts report "Tax Revenue (Net)". Net of what, and why does it matter?

0 of 4 answered. You can submit with questions unanswered — they simply score zero.

Now do it with your own numbers

From the latest monthly accounts, write down net tax revenue, non-tax revenue and non-debt capital receipts with each as a percentage of its Budget Estimate. One of the three is far ahead of a pro-rata pace and one is behind. For the one that is ahead, find out what single item is most likely responsible and why it arrives when it does. Then say what you would conclude about the government's revenue position, and what you would not.

Six months of twelve is 50%, so compare each percentage against that rather than against each other. The item that is ahead is a transfer from an institution that declares its result once a year, which is why its share of the annual estimate jumps in a single month rather than accruing evenly.

Who published this

Published by FreeFinance, an independent financial education project. How these chapters are written and checked. Nothing here is investment advice.

Sources