Build Baby Build
Note: Independence day speech by PM was very practical. I want to write essay about same but due to lack of time, I put together this blog with help of LLM. Thoughts are mine but composition is by LLM.
I have a weakness for governments that announce factories.
There is something unusually reassuring about a politician standing before a country of 1.4 billion people and talking about semiconductors, nuclear power, artificial intelligence, manufacturing and infrastructure. These are not particularly good subjects for applause. They have terrible immediate political aesthetics. A factory takes years. A reactor takes longer. A semiconductor ecosystem may take a generation. Nobody gets a particularly satisfying photograph standing beside a transmission line.
Yet these are precisely the things from which prosperity tends to emerge.
The Prime Minister's Independence Day speech was interesting to me for this reason. The announcements - a push to train one crore young Indians in AI, seven to eight semiconductor plants, an eventual nuclear capacity of 100 GW, a national sporting talent programme and a broader emphasis on manufacturing, technology and strategic autonomy - look disparate when read as a list. Read as capital allocation, they are considerably more coherent.
They are attempts, in different forms, to increase the amount of productive capacity available to an Indian citizen.
That is the variable I care about.
Economics becomes strangely confused when it begins with consumption rather than production. We talk endlessly about how much money people have, how much the government transfers, how much subsidy a household receives, how many jobs exist. These are important measurements, but they are downstream measurements. Before there can be wages, there must be output. Before there can be tax revenue, there must be income. Before there can be redistribution, there must be something worth distributing.
The question preceding all of these is almost embarrassingly simple:
How much can the economy produce?
A country becomes rich when its people become extraordinarily productive.
The mechanism is not mysterious. A worker with a machine produces more than a worker without one. A worker with electricity produces more than one without reliable electricity. A farmer connected to a functioning road network can reach markets that another cannot. An engineer with modern computational tools can perform work that once required an entire team. A manufacturer with a port nearby can compete in markets thousands of kilometres away. A scientist with access to sophisticated instruments can investigate questions inaccessible to someone working with primitive equipment.
Capital is accumulated human leverage.
That is why I have always found the distinction between giving people resources and increasing their capacity to create resources so important. The former can improve a household's position today. The latter can alter the trajectory of the household, and potentially of its children, for decades.
The difference is compounding.
A ₹10,000 transfer is consumed once. A road can be used for thirty years. A power plant can produce electricity for decades. A trained engineer may design products for forty years. A semiconductor facility creates not merely chips but engineers, suppliers, processes, institutional knowledge and downstream companies. The first-order output matters; the accumulated competence around it may matter more.
This is why development is fundamentally a problem of capital formation.
And capital is not synonymous with money.
A country can possess enormous financial resources and remain poor if it cannot turn those resources into productive assets. The interesting question is always what the capital becomes. Does it become another consumption subsidy, another unproductive asset, another bureaucratic programme? Or does it become a railway, a factory, a power station, a laboratory, a university, a data centre, a logistics network?
The latter have an unusual property: other people can build on top of them.
This is the hidden power of infrastructure.
A road is not valuable because asphalt has intrinsic economic significance. It is valuable because thousands of economic decisions become cheaper once the road exists. A port is valuable because it changes the economics of trade. Electricity is valuable because it makes an enormous number of other machines economically viable. Digital infrastructure is valuable because it reduces the cost of transactions that previously required physical institutions.
Good infrastructure therefore behaves almost like a reduction in the friction coefficient of an economy.
This is why I am instinctively pro-infrastructure.
There is a certain sophistication in contemporary discourse that treats physical infrastructure as somewhat primitive. The fashionable country is supposed to be a "knowledge economy", a "services economy", an "innovation economy". Fine. But knowledge requires buildings, computers and electricity. Services require offices, telecommunications and transportation. Innovation requires laboratories, universities, capital and an industrial base capable of turning discoveries into products.
The physical world keeps sending the bill.
AI is perhaps the clearest example.
The public conversation around AI is dominated by models, agents, benchmarks and software. Underneath the entire thing sits a remarkably unglamorous stack of physical infrastructure: electricity generation, transmission, land, cooling, data centres, networking equipment and semiconductors.
The intelligence may be artificial.
The infrastructure is not.
This is why the semiconductor announcement matters. It would be easy to interpret chip manufacturing as a strategic exercise in import substitution. That would undersell it. The real prize is industrial competence.
Countries do not become technologically important because they possess a single factory. They become technologically important when a factory can exist inside an ecosystem of engineers, suppliers, universities, capital, specialised machinery, logistics, research institutions and customers.
Competence accumulates.
Once accumulated, it lowers the cost of acquiring the next competence.
That is how industrialisation becomes self-reinforcing.
There is a similar logic behind nuclear power.
India's aspirations are increasingly incompatible with energy scarcity. A country that intends to urbanise, industrialise, electrify transport, operate enormous computing infrastructure, manufacture sophisticated products and raise living standards for hundreds of millions of people is going to require extraordinary quantities of reliable electricity.
Energy is the primitive input into almost everything else.
We have a tendency to discuss energy policy as though it were one sector among many. It is closer to a constraint on the entire economic system. If energy becomes cheap, abundant and reliable, entire categories of economic activity become cheaper. If it becomes scarce or unreliable, the cost propagates through everything.
The target of 100 GW of nuclear capacity by 2047 therefore deserves to be read as industrial policy rather than merely energy policy.
The same applies to AI training.
The most valuable interpretation of the proposal to train one crore young Indians is not that India needs to produce ten million people who can use a fashionable piece of software. It is that India is attempting to increase the productivity of an enormous population at the exact moment when a general-purpose technology capable of changing the economics of cognitive labour is becoming available.
That is potentially enormous.
India's demographic advantage has always been slightly misunderstood. Population is not an economic advantage by itself. It becomes an advantage when the population is combined with capital, skills, institutions and technology.
A billion people with low productivity are a burden on the capital stock.
A billion increasingly productive people can become the world's largest economic engine.
The difference is not demographic.
It is capital per person and output per person.
This is also where the argument for enablement becomes more interesting than the simplistic welfare-versus-market debate.
I have no ideological objection to a safety net. A civilised society should protect people against circumstances in which individual agency is insufficient. But a country that makes permanent redistribution the centre of its economic philosophy eventually encounters an unpleasant mathematical problem: the state can redistribute output indefinitely only if someone continues increasing output.
There is a limit to how much prosperity can be redistributed before the productive base begins to matter.
The more interesting function of government is therefore to enlarge that base.
Make transportation cheaper.
Make electricity more abundant.
Make education better.
Make capital easier to deploy.
Make cities more productive.
Make it easier to start and close businesses.
Make research commercially useful.
Make manufacturing competitive.
Make technology available to ordinary people.
Make the legal and physical infrastructure around enterprise reliable enough that an ambitious twenty-five-year-old can attempt something without needing to become an expert in navigating the state.
This is not an argument against government.
It is an argument for a government with a larger conception of its job.
The state has access to something private individuals do not: an ability to coordinate resources across decades and across enormous populations. That power can be squandered on trying to administer every economic outcome, or it can be used to construct the foundations upon which millions of private decisions become productive.
The second use is vastly more interesting.
Because once the foundation exists, the state no longer needs to create every subsequent unit of prosperity.
A highway does not need a government employee driving every truck.
A power plant does not need a government employee operating every factory.
A semiconductor ecosystem does not need a government employee designing every electronic product.
A university does not need the state to decide which student becomes an entrepreneur.
The infrastructure creates optionality.
And optionality is one of the most underrated forms of wealth.
A poor person has fewer choices because the surrounding system makes more choices economically impossible. A richer society is not merely one in which people possess more things. It is one in which far more courses of action are viable.
You can start a company.
You can move cities.
You can export.
You can study something obscure.
You can build a product for a market on another continent.
You can fail and try again.
You can specialise.
You can become absurdly productive at something that did not exist twenty years earlier.
Infrastructure, capital and institutions expand this possibility space.
That, ultimately, is why I find the language of "empowerment" somewhat unsatisfying when it is used as political decoration. Real empowerment is not a slogan. It is an increase in the set of things a person can actually do.
A road empowers more effectively than a speech.
Reliable electricity empowers more effectively than a slogan about opportunity.
A good school empowers more effectively than an employment guarantee.
Cheap computing empowers more effectively than a lecture about digital India.
A functioning capital market empowers more effectively than telling young people to dream big.
The difference is that the first set of things remains useful after the politician has left the stage.
This is the standard I would apply to the Prime Minister's announcements.
The speech itself is irrelevant in the long run.
The reactors matter.
The fabs matter.
The transmission lines matter.
The factories matter.
The trained engineers matter.
The exports matter.
The productivity numbers matter.
And, eventually, the wages matter.
Announcements are cheap. Capacity is expensive.
India has no shortage of announcements. What it needs is an almost obsessive national preference for completion.
Build the thing. Operate the thing. Measure the thing. Improve the thing. Build the next thing.
There is nothing glamorous about this process.
That may be precisely why it works.
Great economic transformations rarely look particularly impressive while they are happening. Someone is pouring concrete. Someone is laying cable. Someone is calibrating a machine. Someone is training an engineer. Someone is debugging a manufacturing process at two in the morning. Someone is trying to make a component for ₹80 that currently costs ₹120.
Years later, the aggregate result is called development.
At the time, it mostly looked like work.
This is why the phrase that keeps coming to my mind after the speech is not "Viksit Bharat", "self-reliance", or even "Amrit Kaal".
It is something considerably less dignified.
Build, baby, build.
There is a particular kind of optimism in construction that I find more convincing than political optimism. It does not require believing that the future will be good.
It requires believing that the future can be made better through deliberate accumulation of capability.
That is a much harder belief.
And a much more useful one.
If India can become extraordinarily good at building energy, infrastructure, manufacturing capacity, technological capability and human capital, the consequences will eventually escape the control of whoever happens to be sitting in the Prime Minister's chair.
That is the point.
The greatest development policy is one that eventually makes itself less important.
The state builds the platform.
Capital builds the businesses.
Businesses create employment.
Workers acquire skills.
Skills create higher productivity.
Higher productivity creates higher incomes.
Higher incomes create demand.
Demand attracts more capital.
Capital builds more capacity.
And the cycle continues.
At some point, you stop talking about development and realise that development has become the default behaviour of the economy.
That is the country I want to see.
Not a country whose citizens are permanently waiting to receive a larger share of what exists.
A country in which what exists keeps becoming larger.
Build Baby Build.