What does India actually make, and why does it matter?

A wheat farmer in Punjab, a coder in Bengaluru, a welder in a Gurugram factory. Their work adds up to a single number: ₹323.5 lakh crore. That is India's Gross Value Added, GVA, for the year to March 2026. GVA is what an economy creates after subtracting the cost of inputs. The farmer, the coder, and the welder belong to three large buckets: agriculture, services, and industry.

Here is how ₹323.5 lakh crore broke down:

The snapshot is stark. Well over half of what India makes is a service, a trade, a transport, a software export, a rent. The factory floor accounts for just over a quarter. The farm, less than one in five rupees. But this snapshot hides a shift so unusual that if you looked only at the numbers, you would think you were reading the story of a very different country.

In 1950-51, agriculture was the economy. It made up 53.2% of GVA. Services were a distant second at 30.6%, and industry just 16.2%. Manufacturing alone was 12%. A farm economy with a thin layer of everything else.

Seventy-five years later, the transformation leaps off the table:

Agriculture's share collapsed, services more than filled the gap, and manufacturing barely budged. In 1951 it was 12%; in 2025-26 it is 14.1%. Over three generations, through planning and liberalisation and the IT revolution, the factory's share moved by just two percentage points.

The usual path of development, the one every rich country took, runs from farm to factory to services. Britain's textile mills pulled farmers off the land. Japan's heavy industries and its car plants did the same. South Korea's shipyards and steel mills built a middle class. In each case, the middle stage, the noisy, polluting, back-breaking factory stage, was the rung that lifted millions into steady wage work before anyone moved into offices. India seems to have skipped the rung altogether.

After the 1991 reforms, India opened up and found a different sweet spot: services. English-speaking engineers plugged into the global demand for IT, back-office processing, and business services. By the early 2000s, India was the world's back office. The jump brought wealth to a slice of urban India, but it was a jump only the educated could make. A worker with an engineering degree in Pune could code for a bank in London; a landless labourer in Bihar could not.

Economists call this services-led growth. When it happens before a country has built a proper manufacturing base, they call it premature deindustrialisation. India is not the only developing country to experience it, but its case is extreme because the services boom was so large and the factories so small. The government itself acknowledged the problem when it launched Make in India in 2014, a direct admission that the factory floor needed urgent reinforcement.

The answer lies not in output, but in jobs. A rupee of manufacturing output creates, on average, many more steady wage jobs for low- and medium-skilled workers than a rupee of software export or retail trade. The classic factory, a textile mill, an auto plant, a phone assembly line, needs hundreds of hands who do not need a college degree. It pays them a wage, offers some social security, and over time lifts their productivity. A large part of the services that absorb India's ex-farm workers are not high-end tech consultancies. They are petty retail, the kirana shop, the street vendor, low-productivity transport, security, and domestic work. Their output counts in GVA, but the wages and conditions often do not match a formal factory job.

Imagine a young man in Kanpur who finishes Class 12. If a phone assembly plant has come to his district, he might learn a skill on a line and draw a regular salary. If it has not, he is likely to end up as a Swiggy delivery rider, a security guard, or a helper at a construction site. These are real jobs. They pay, often better than the farm. But they rarely build the same kind of middle-class life over a decade.

We must be careful: this entry is about output shares, not employment shares. The data on who works where is not in front of us. But we can say something true and important: a sector can shrink as a share of what the economy produces while still holding a huge share of the people who work in it. Agriculture is the obvious case. Its output share is down to 16.8%, yet we know it still employs a very large portion of India's workforce. The factory's output share may be small, but every percentage point it gains could mean millions of decent jobs for people who have no other way into the modern economy.

India has one of the largest and youngest working-age populations on the planet. Economists speak of a demographic dividend: a once-in-a-generation window, roughly of a few decades, when the share of dependents, children and the elderly, falls and the share of people who can work rises. If those workers find productive jobs, they can drive a growth surge that lifts an entire generation out of poverty. If they do not, that same young population becomes a pressure cooker.

The window for India is now, and it will not stay open forever. A growth path led by capital-light, skill-heavy services rather than labour-absorbing manufacturing makes the job of providing those productive jobs much harder. The missing factory rung is not a curiosity from a chart. It is the reason that the country's single biggest asset, its young people, may end up under-used at the very moment they should be building the future.

Yes. The most visible policy response has been the production-linked incentive (PLI) scheme, launched in 2020 across multiple sectors from electronics to pharmaceuticals to automobiles. The government offers cash incentives to companies that produce in India and hit certain scale and investment targets. The idea is to lower the cost of manufacturing in India enough to attract global firms that might otherwise set up in Vietnam or Bangladesh, or stay in China.

Make in India, launched earlier, set the tone. PLI put serious fiscal backing behind the ambition, with incentives across 14 sectors. The policy is a bet that with enough push and a large domestic market, the factory floor can finally thicken.

The clearest and most cheered test case is mobile phones. Before 2015, India imported most of its handsets. By the mid-2020s, assembly had taken off, driven by PLI incentives, higher import duties on finished phones, and a global shift. After the pandemic and US-China trade tensions, companies began looking for supply chains outside China, the so-called China-plus-one strategy. Global brands and their contract manufacturers expanded assembly lines in states like Tamil Nadu, Karnataka, and Uttar Pradesh. India moved from assembling phones for the domestic market to exporting them.

That is a real manufacturing success. But there is a catch. Much of what happens in those campuses is assembly: screens, chips, and memory are imported, and Indian workers snap them together. That work creates jobs, but it adds less value to the Indian economy than making the components themselves would. A phone "Made in India" still has the bulk of its value created abroad. The kind of deep, value-adding manufacturing that builds a middle class, the way Japan's carmakers or South Korea's shipbuilders did, requires more than assembly. It needs an ecosystem of suppliers, a skilled workforce that goes beyond line workers, and decades of investment in research and process. The mobile phone story is a pointer to what is possible, not proof that the factory rung is finally in place.

At the end of all the graphs and percentages is a very human question. The shape of the economy is a floor plan for the kind of working life that is available to the majority of people. In a country where industry is small and most workers cannot get into high-end services, the fallback is often between low-productivity farm work and low-productivity urban informal work. That is not a development path that sustainably lifts tens of millions; it is a treadmill.

The missing factory stage is not a nostalgia for smokestacks. It is a calculation about whether a young workforce, the largest in the world, will spend its prime years in genuinely productive, fairly paid jobs or in insecure, piecemeal ones. That is what the numbers in this entry are really tracking. The economy has a size, which we measure with GDP, and a shape, which we measure with these sectoral shares. For most Indians, the shape matters more.