Breakdown: India GDP

From Farms to Finance: 75 Years of India’s GDP. In 1950–51, more than half of India’s income came from the land. Today a quarter comes from finance, real estate and IT. This is where we earned, what grew, what stalled, and what the numbers leave out. This is part of Breakdown, a series where I take one idea and pull it apart until the numbers make sense.

I have always read India’s GDP as one number. It is ₹318 lakh crore now, it grew 7% last year, and that is usually where I stop. But a GDP number is really a sum of many smaller numbers, one for each way a country earns its living. Those smaller numbers tell a much more interesting story than the total.

So I went back to the official tables and pulled nine snapshots, roughly one every ten years, from Independence to 2024–25. Here is what I found.

A small note before we start

There is no official GDP figure for 1947. India’s national accounts series starts in 1950–51, so that is my first snapshot. The others are 1955–56, 1965–66, 1975–76, 1985–86, 1995–96, 2005–06, 2015–16 and 2024–25, all financial years.

The statisticians have also changed their base year several times, most recently in February 2026, when MoSPI moved to a new series with 2022–23 as the base (PIB). Each change reshuffles some activities between lines. So I use three official sources, one for each era, and I flag the places where a jump comes from a change in method and not a change in the economy.

All figures are at current prices, in rupees.

How big is the pie?

In 1950–51, India’s total value added was about ₹10,000 crore (Economic Survey 2013–14, Table 1.3B). In 2024–25, it was ₹288.5 lakh crore, and GDP including product taxes was ₹318.07 lakh crore (MoSPI).

That is roughly 2,900 times bigger in rupee terms. Most of that is inflation and population growth. A fairer comparison is real income per person, which is about 9.8 times its 1950–51 level (Business Standard). Still a big jump, but a far more honest one.

One early surprise: between 1950–51 and 1955–56, the economy barely grew in rupee terms, from ₹10,036 crore to ₹10,861 crore. The value of farm output actually fell, from ₹5,274 crore to ₹4,847 crore. When more than half the economy is agriculture, crop prices can move the whole country.

Where did we earn? The five lines

To compare 75 years fairly, I grouped everything into the five broad lines that the Economic Survey uses:

• Farm & mining: crops, livestock, forestry, fishing, mining
• Factories, construction & power: manufacturing, building, electricity, gas, water
• Trade, transport & telecom: shops, hotels, railways, trucks, communication
• Finance, real estate & business: banks, insurance, property, professional services and, in the new series, IT
• Government & other services: public administration, defence, education, health, personal services

Each bar is 100% of what India earned that year.

What grew, and what did not:

1. Farm & mining: from 53% to 21%

This is the biggest shift. In 1950–51, 53 of every 100 rupees came from the land and from mining. By 2005–06, it was 22. Since then, it has held steady at around 20 to 21.

Agriculture did not shrink. In rupee terms, farm and mining output grew from about ₹5,300 crore to about ₹62 lakh crore. Everything else simply grew faster.

What stays with me is the gap between output and jobs. Agriculture now produces about a fifth of value added, but it still employs about 43% of India’s workers (Business Standard). That gap is the reason farm incomes remain such a hard problem.

2. Factories, construction & power: doubled, then stalled

This line went from 13.5% to about 25% by the mid-1990s. Then it stopped climbing. For three decades it has moved between 25% and 28%.

Within it, manufacturing on its own is 14.5% of value added in 2024–25. For a country that has spent years talking about becoming a manufacturing hub, this is the line to watch. The new series did give manufacturing a small boost: its average share of GDP between FY23 and FY26 is 13.3%, compared with 12.9% under the old series (Business Standard).

3. Finance, real estate, IT & business: the new number one

This line was 12.6% in 1950–51 and stayed in a narrow band of roughly 12% to 16% for four decades. After liberalisation it took off, and in 2024–25 it is 26.2%, the largest single line in the economy at ₹75.5 lakh crore.

Part of the jump comes from changes in method. For the same year, 2024–25, the new series shows this line at 26.2% while the old series showed 22.9%. Either way, it is now the biggest line (Economic Survey 2025–26, Table 1.4).

4. Trade, transport & telecom: up, then reshuffled

This line grew steadily from 9.7% to 25% by 2005–06, as roads, railways, shops, hotels and mobile phones spread across the country. The drop you see after that is largely a change in method. The 2011–12 series used different data and methods to estimate trade, and the new 2022–23 series regroups some services again. For the same year, 2024–25, the old series shows this line at 17.5% and the new series at 13.9%.

This was a useful lesson for me. When a line suddenly falls by several points, I should first check whether the ruler changed.

5. Government & other services: the quiet constant

Public administration, defence, education, health and other services were 11.2% in 1950–51 and are 12.9% today. Across 75 years, this line barely moved. Every other line reshaped itself around it.

2024–25 in detail

The new series gives a finer split for the latest year.

Three things stand out for me:

• Finance, real estate, IT and professional services earn more than manufacturing and construction combined.
• Agriculture, at ₹55.9 lakh crore, is still the second largest line. In the new series its share is 19.4%, higher than the 17.9% the old series showed.
• Electricity, gas and water are only 2.8% of value added, even though every other line depends on it. As someone who works in power, I found that number humbling.

What is missing, and what is new:

Missing lines:

• 1947 itself. There is no official GDP number for the year of Independence.
• The Plan and Non-Plan split. For decades, the Union Budget divided spending into Plan and Non-Plan. That split was dropped, and the latest tables simply show “n.a.” against it (Economic Survey 2025–26, Table 2.4).
• A long series on central government salaries. I could not find a consistent official series going back to the 1950s. More on this below.
• Jute and tea at the top of the export list. In the early years, India’s large exports were jute, tea, cotton and oilseeds (Swadeshi Shodh Sansthan). They are no longer the headline.

New lines:

• IT as a named line. The new series lists “Financial, Real Estate, IT and Professional Services” as one of its eight headline lines, something the 1950s accounts had no reason to track.
• Services exports. In 2024–25, India exported $387.5 billion of services, close to the $437.4 billion of goods exports (Business Standard).
• Smartphones. Exports reached $24.14 billion in FY25, up from $10.96 billion two years earlier, making them one of India’s top exported goods (Deccan Herald / PTI).

Exports and imports: open, but lopsided

In 1950–51, India’s goods exports and imports were both about 5.8% of GDP. The books were almost balanced: ₹606 crore of exports against ₹608 crore of imports (Economic Survey 2025–26, Table 6.1A).

For the next four decades, trade stayed small. Exports even fell to 2.8% of GDP in 1965–66. After 1991, both lines rose sharply. The difference is that imports rose faster. In 2024–25, goods exports were ₹37.0 lakh crore (11.6% of GDP) and goods imports were ₹61.0 lakh crore (19.2% of GDP). The goods trade deficit alone is about 7.5% of GDP.

Get Amar’s stories in your inbox

Join Medium for free to get updates from this writer.

Remember me for faster sign in

Services soften this. When goods and services are counted together, exports are 22.1% of GDP and imports are 23.9%, so the overall gap narrows to under 2% (MoSPI). India sells software, consulting and back-office work to pay for oil, electronics and gold.

The basics: debt, capex and the cost of running government

Debt to GDP

The central government’s total liabilities were 55.3% of GDP in 1990–91 and peaked at 63.1% in 2002–03 (12th Finance Commission, Chapter 3). They fell to 50.1% by 2015–16 (Economic Survey 2023–24, Table 2.5), and stood at 55.0% in 2024–25, or ₹181.7 lakh crore (Economic Survey 2025–26, Table 2.5).

One caution: that 55.0% uses the old GDP base. On the new, slightly smaller GDP of ₹318.07 lakh crore, the same debt works out to about 57%.

For the early decades, one research series estimates that public debt of the Centre and States together was around 18.5% of GDP in 1951 (Das and Ghate, Business Standard). It is not directly comparable with the figures above, but the direction is clear. We borrow far more today, relative to what we earn, than we did at the start.

Capex to GDP:

I looked at capex in two ways.

The whole economy. Gross fixed capital formation, meaning everything spent on buildings, machines, roads and equipment by the government, companies and households, was 9.3% of GDP in 1950–51. It is 31.6% in 2024–25 (Economic Survey 2013–14, Table 1.6; MoSPI).

Who invests has changed completely. In 1985–86, the public sector invested more than the private sector, 11.3% against 9.3% of GDP. By 2005–06, the private sector was at 23.0% and the public sector at 7.3%. The state built the base, and private business took over the heavy lifting (Economic Survey 2021–22, Table 1.9).

The Centre alone. The central government’s capital expenditure was 5.6% of GDP in 1990–91, but that older figure also counted large loans the Centre gave to the States. It fell to 2.3% by 2000–01 and to 1.8% by 2015–16. It has since risen to 3.2% in 2024–25 (Economic Survey 2021–22, Table 2.4; Economic Survey 2025–26, Table 2.4).

Here is the number that made me stop: in 2024–25, the Centre spent 3.4% of GDP on interest and 3.2% on capex. We pay slightly more to service old borrowing than we spend on building new assets.

Salaries and admin to GDP:

This was the hardest number to find. The Centre does not publish a consistent salary series going back to the 1950s, so I used two measures.

The Centre in 2024–25. Pay, allowances and travel for about 33 lakh central employees cost ₹3.35 lakh crore, which is about 1.1% of GDP. The broader establishment expenditure of the Centre, which covers the running of ministries and departments, was ₹8.29 lakh crore, or about 2.6% of GDP (Expenditure Profile 2026–27, Statements 3 and 22).

All government, over time. Government final consumption expenditure measures what the Centre and States spend on running the government, mostly salaries plus the goods and services they buy. It rose from 6.6% of GDP in 1960 to about 10.9% in 1985, and has stayed around 10% to 11% ever since (World Bank data via IndexMundi). In 2024–25 it was 10.7% (MoSPI).

So the cost of running government, relative to the economy, grew until the mid-1980s and has been flat for about 40 years.

The whole picture in one table:

What I take away from this:

India did not follow the usual path. Most countries move from farms to factories and only later to services. India moved from farms almost straight to services, while factories doubled their share early and then stalled at about a quarter of the economy.

The trade numbers tell the same story from the outside. We pay for our imported goods largely with exported services.

And the fiscal basics show a government that invests more than it did a decade ago, but that now spends as much on interest as on new assets.

When I hear the next GDP headline, I will be asking a different question. Not how big the number is, but which line moved it.

This is part of Breakdown, a series where I take one business idea and pull it apart until the numbers make sense. All figures are from official Indian government sources, linked above. Figures from different base years are not perfectly comparable. I have flagged the major breaks, so please read the trends rather than the decimals.

Sources

• Economic Survey 2013–14, Statistical Appendix (Tables 1.3B, 1.5, 1.6)
• Economic Survey 2021–22, Table 1.9 and Table 2.4
• Economic Survey 2023–24, Table 2.5
• Economic Survey 2025–26, Statistical Appendix, Table 2.4, Table 2.5, Table 6.1A
• MoSPI, Press Note on GDP Estimates for Q4 2025–26 and PE 2025–26
• PIB, New Series of GDP Estimates with Base Year 2022–23
• Union Budget, Expenditure Profile 2026–27 and Statement 22
• 12th Finance Commission Report, Chapter 3
• Business Standard: How India’s economy transformed over 80 years
• Business Standard: New GDP series and manufacturing
• Business Standard: Public debt in India (Das and Ghate)
• World Bank data via IndexMundi: Government final consumption expenditure, India
• Business Standard: CSO nominal GDP 2015–16
• Deccan Herald / PTI: Smartphone exports FY25
• Swadeshi Shodh Sansthan: Major exports of India 1947 to 1990