You have already seen power shared sideways — split between the different arms of one government, and shared with the opposition a ruling party must still answer to. Now watch it shared a different way: not sideways, but up and down, between a government sitting in Delhi and dozens of governments sitting below it in the states. That is federalism — the vertical division of power among different levels of government, the form this chapter isolates from the horizontal and community forms you have already studied.
This chapter moves in one direction: from what federalism is, in general, to how India specifically qualifies as one, to how democratic politics — not just the words printed in the Constitution — has actually made India's federation stronger over the decades since, and finally down to the smallest tier of all, the panchayat or municipality nearest to where you live. Power divided sideways is one kind of sharing; power divided top to bottom is another — and this chapter follows the second kind all the way down.
Only 25 of the world's 193 countries run federal systems. A small number — except that those 25 countries hold 40 per cent of the entire world's population between them. Look down a list of the world's biggest countries by population and federations keep turning up: the USA, Brazil, Nigeria, and, as this chapter will show, India. Most of the world's large countries are federations — not because federalism happened to be fashionable, but because dividing power across levels is close to the only workable way to govern something that big. (Handbook of Federal Countries, 2002.)
So what actually is federalism? A system of government in which power is divided between a central authority and the country's constituent units. A federation usually runs on two levels: one for the entire country, holding a handful of subjects of common national interest, and others at the level of provinces or states, handling day-to-day administration. Both levels get to use their own power independently of the other — neither one borrows its authority from the other.
A federal system is doing two jobs at once, not one. It has to hold the country together as one unit, and at the same time leave room for real regional difference inside that unit — different languages, different economies, different histories, all inside one federation. Two things make both jobs possible together: governments at every level have to agree, up front, on the rules for sharing power, and each level has to genuinely trust that the other one will keep its side of that agreement. An ideal federal system runs on both at once — agreement on the rules, and mutual trust to abide by them; take either one away and the whole system strains.
Set federalism against its opposite and the definition sharpens. In a unitary system, there is either only one level of government, or the lower levels answer straight up to the one above them — the central government can simply order a state or provincial government to act, the way a head office orders a branch. Federalism refuses that chain of command. The centre cannot order a state government to act; the state holds powers of its own that it does not answer to the centre for, and both levels answer separately to the people who elected them. The test is not how many levels of government exist on paper — it is whether the centre can give the state an order it must obey.
Six tests, and a system that fails even one of them is not really federal, whatever it calls itself. Two or more tiers of government. Each tier answers to the same citizens, but over separate subjects — separate jurisdiction, not overlapping authority. Every tier's jurisdiction is written into the constitution, so no tier's existence depends on another tier's goodwill. The constitution's fundamentals cannot be changed by one level alone — that needs both levels' consent. Courts get the power to interpret the constitution and settle disputes between levels, with the highest court as final umpire. And each tier's sources of revenue are specified — financial autonomy, not financial dependence. Six tests, and India will need to clear every one of them before this chapter can honestly call it a federation.
Federations do not all arrive by the same road. In a "coming-together" federation, states that already existed as independent units choose, voluntarily, to pool their sovereignty into a bigger one — usually because banding together buys them a security none of them could manage alone. The USA, Switzerland, and Australia all federated this way. Because each state walked in as an equal, this route tends to leave the constituent states strong, and roughly equal in power, relative to the federal government they themselves created.
India took the other road. In a "holding-together" federation, a country that already governs as one unit decides to divide its existing power downward, between the states and the national government it keeps at the centre. India, Spain, and Belgium all federated this way. This route tends to leave the centre stronger than a coming-together federation's, the constituent units unequal in power, and some units carrying special powers the others do not. The road a federation took to get here decides, decades later, how the power inside it is actually shared out — and that is exactly the pattern India's own states will show.
Here is a claim worth testing before you accept it. In a classroom activity from Nepal, a student named Khag Raj argues that federalism inherently leads to reservation of seats for different caste groups — because that is supposedly what it produced in India. Test it. India did adopt caste-based reservation. But federalism only tells you how power gets divided across levels of government — nothing in that mechanism says anything about who gets seats reserved for them. Reservation is a separate policy choice a country may or may not layer on top of a federal structure; plenty of federations exist with no reservation policy at all. Khag Raj mistook one country's separate policy choice for federalism's own definition — a federal system does not, by itself, produce reservation of any kind.
A second claim, from a student named Sarita in the same activity: "our country is not very big, so we don't need federalism." Test that one against a country this chapter has already named. Belgium is small — smaller than several Indian states — and it still adopted federalism, not to manage size, but to manage the tension between its Dutch- and French-speaking regions. Size was never the deciding factor. What decides whether a country needs federalism is diversity and the demand for self-rule that comes with it; size is only one possible source of that demand, and a small country can carry just as much of it as a large one.
Two more voices from that same Nepalese classroom are not staged as traps at all — the source leaves them open, and so will this chapter. Babu Lal hopes the Terai areas will get their own autonomous state government: a plausible, ordinary benefit federalism can deliver, not a misconception to correct. Ram Ganesh welcomes federalism because it moves power that once sat with the king into the hands of elected representatives — a legitimate democratic argument, tying federalism to popular sovereignty rather than monarchy. Work out for yourself what you would say back to each of them; the chapter does not hand you an answer here.
India's own Constitution never once uses the word "federation." It calls the country a "Union of States." Read the label carefully, though, and it does not settle much, because every one of the six features this chapter just tested apply to India's constitutional provisions, whatever word the framers chose to print. India arrived at this arrangement the hard way: independence came bundled with a bloody partition, and in the years right after, hundreds of princely states had to be persuaded, one by one, to join the new Union rather than stand apart from it.
India did not start with three tiers. The Constitution originally set up two: the Union Government, representing the whole country, and the State governments below it. A third tier — Panchayats in villages, Municipalities in towns — was added only later, once the first two were already working. As in any federation, each of India's three tiers ended up with its own separate jurisdiction, not a shared one.
How does India actually decide who legislates on what? Through three lists, each with its own rule. The Union List holds subjects that need one uniform national policy — defence, foreign affairs, banking, communications, currency — and only the Union Government may legislate here. The State List holds subjects of state and local importance — police, trade, commerce, agriculture, irrigation — and only the State Governments may legislate here. The Concurrent List holds subjects both levels care about — education, forests, trade unions, marriage, adoption, succession — where both may legislate, and if their laws collide, the Union's law prevails. Three lists, three different answers to the same question: which government gets the final word on this subject.
Even three lists cannot cover everything — computer software did not exist when the Constitution was written, and neither did a hundred other subjects that would eventually need a legislator. Whatever falls outside all three lists is called a "residuary" subject, and the Constitution hands the power to legislate on residuary subjects to the Union Government alone.
Not every Indian state holds identical powers — consistent with the general rule this chapter already named: holding-together federations rarely divide power equally among their units. Assam, Nagaland, Arunachal Pradesh, and Mizoram hold special powers under Article 371, granted because of their own particular social and historical circumstances — chiefly to protect indigenous land rights and culture, and to reserve government jobs for local residents; someone from outside the state cannot simply buy land or a house there. Several other Indian states carry similar special provisions of their own.
At the opposite end sit Union Territories — units too small to stand as an independent state, but not a natural fit to merge into any existing one either: Chandigarh, Lakshadweep, Delhi. A Union Territory holds very little power of its own, none of a full state's powers, and the Central Government runs it directly, with special powers reserved for exactly that purpose. Special-status states hold more power than the norm; Union Territories hold less — two opposite ways of deviating from what an ordinary Indian state gets.
India deliberately made it hard to change how power is shared between Union and states — Parliament alone cannot touch it. Any amendment to that arrangement must first clear both Houses of Parliament with at least a two-thirds majority, and only then goes to the state legislatures, at least half of which must ratify it before it takes effect. That is not bureaucratic caution; it is feature four of federalism, built into working procedure — no single level gets to rewrite the fundamentals alone.
Someone has to referee when the Union and a state disagree about who has the power to do what — that job belongs to the judiciary. Any dispute over the division of powers goes to the High Courts, and ultimately the Supreme Court, to decide. Both levels also raise their own money, by levying their own taxes, to carry out whatever responsibilities the lists have assigned them. This is feature five and feature six, made concrete: a judicial umpire to settle disputes, and financial autonomy so each level can actually pay for the powers it holds.
Test yourself against three real questions this chapter's own worked scenarios pose. Could Rajasthan block the Centre from testing a nuclear device at Pokharan? Defence sits on the Union List, so no — Rajasthan has no standing to block it. Can the Union stop Sikkim from prescribing its own school textbooks without the state seeking the Union's permission first? Education is a Concurrent, largely state-administered matter, so the Union cannot simply override the state here. Can the Prime Minister force one uniform police policy against naxalite activity on Andhra Pradesh, Chhattisgarh, and Orissa alike? Police sits on the State List, so no — that decision belongs to each state separately. Work the three-lists rule against each case yourself before you check whether your answer matches the logic above.
A constitution on paper is necessary for federalism to work. It is not sufficient. India's federal experiment did not succeed merely because its provisions were well laid out — it succeeded because of the nature of India's democratic politics, which turned the spirit of federalism, respect for diversity, and the will to live together into shared ideals people actually acted on, not just words printed in a founding document. The next few sections test that claim against real cases, starting with the very first one Indian democracy faced.
The creation of linguistic states was that first real test. Compare a political map of India from 1947 with one from today, and the change is extensive — old states vanished, new ones appeared, boundaries and names shifted across the country. In the years after 1947, many state boundaries were redrawn specifically so that speakers of one language ended up living inside one state. A few states — Nagaland, Uttarakhand, Jharkhand — were carved out for a different reason again: to recognise a difference of culture, ethnicity, or geography, not language.
National leaders feared, at the time, that carving states along language lines would tear the country apart — and the Central Government resisted linguistic states for a while, on exactly that fear. Experience proved the opposite. Forming linguistic states made India more united, not less, and made the country genuinely easier to administer. The fear was disintegration; what actually happened was the opposite — worth remembering the next time someone predicts that recognising a difference will fracture a country.
India's Constitution made a deliberate choice most new countries do not make: it never granted "national language" status to any single language. Hindi became the official language — but Hindi is the mother tongue of only about 40 per cent of Indians (a figure this chapter will refine later, using 2011 Census data, to 44 per cent). So the Constitution built in safeguards for everyone else: 21 other languages, besides Hindi, are recognised as Scheduled Languages, a candidate sitting a Central Government exam may take it in any of them, and every state runs most of its own government work in its own official language, not in Hindi.
The Constitution originally planned for official use of English to stop in 1965. It did not stop. Many non-Hindi-speaking states demanded English continue — in Tamil Nadu the demand turned into a genuinely violent movement — and the Central Government backed down, agreeing to keep English alongside Hindi for official purposes. Critics point out this compromise mostly favoured an English-speaking elite already comfortable in that language. Promoting Hindi remains official Government of India policy to this day — but promoting a language is not the same as imposing it, and the Centre cannot force Hindi onto a state where people speak something else.
Set this against Sri Lanka, from the contrast this book has already drawn in an earlier chapter. Indian leaders chose a cautious, flexible approach to spreading Hindi rather than an imposed one — and that flexibility is credited with keeping India off the same collision course that language policy sent Sri Lanka down. The lesson repeats across this whole chapter in a new costume: refusing to impose a solution, and choosing instead to accommodate difference, is what keeps a diverse federation from breaking apart.
A rule written into the Constitution is only as real as the people running the government make it. For a long stretch after Independence, the same party held power at the Centre and in most states, so state governments rarely exercised the autonomous rights the Constitution actually gave them — there was no real political need to. When a different party did run a state, the party ruling the Centre often worked to undermine it, including misusing the Constitution to dismiss state governments run by political rivals. That was not federalism failing on paper — it was federalism working exactly as written, and being undermined anyway, in practice.
That changed after 1990, for a structural reason, not a change of heart. Regional political parties rose across many states, and no single national party could any longer win a clear majority in the Lok Sabha on its own. To form a government at all, national parties now had to build coalitions — alliances that had to include regional parties, on regional parties' own terms.
Coalition dependency did something the Constitution's own text never could on its own: it built a genuine culture of power-sharing and respect for state autonomy, because the Centre could no longer afford to ignore the states it needed as coalition partners. A major Supreme Court judgement reinforced the same shift, making it far harder for the Central Government to dismiss a state government arbitrarily. Federal power-sharing runs more effectively in India today than it did in the years right after the Constitution came into force. Before 1990: dominance and arbitrary dismissal. After 1990: coalition dependency and a court ruling that closed the door on it. The Constitution did not change between those two pictures — the politics around it did.
The ruling behind that shift deserves naming, even though this chapter's own source does not name the case itself: a major Supreme Court judgement made it difficult for the Central Government to dismiss state governments in an arbitrary manner. The source treats this ruling's binding force as settled fact — a decision that structurally shut off a power the Centre had, until then, been willing to misuse.
Just how linguistically diverse is India? The 2011 Census recorded more than 1300 distinct languages as mother tongues. Group closely related languages together — Bhojpuri, Magadhi, Bundelkhandi, Chhattisgarhi, and Rajasthani, all counted under "Hindi," for instance — and the number still does not collapse to a handful: 121 major languages remain. Of those, 22 sit in the Eighth Schedule of the Constitution and are called "Scheduled Languages"; everything else is "non-Scheduled." By number of languages spoken, India is arguably the most linguistically diverse country in the world.
Look at how those 22 Scheduled Languages actually split by number of speakers, according to the 2011 Census. One fact stands out immediately: Hindi leads at 43.63 per cent. Bengali follows at 8.03 per cent, Marathi at 6.86, and Telugu at 6.70. Tamil comes next at 5.70, then Gujarati at 4.58, Urdu at 4.19, and Kannada at 3.61. The list then runs all the way down to languages under one per cent each: Kashmiri, Nepali, Sindhi, Dogri, Konkani, Manipuri, and Bodo.
No single language is the mother tongue of a majority of the population. Even Hindi's share only grows once second- and third-language speakers are added to native speakers. Even then, the total stayed under 50 per cent in 2011. English tells the opposite story. It was the mother tongue of only 0.02 per cent of people. Yet 11 per cent more knew it as a second or third language. No language commands a majority on its own. India's linguistic diversity is not a side detail — it is the central fact any language policy here has to work around.
Historian Ramachandra Guha wrote an argument on linguistic reorganisation in the Times of India, on 1 November 2006 — and this chapter is going to ask you to do something it has not asked before: take a side, using your own state's experience, rather than wait for the chapter to hand you a verdict. That is deliberate. Not every question in this book has a printed answer at the back.
A country this vast, and this internally diverse, cannot actually be run through only two tiers of government. Some Indian states are, by population, individually as large as entire European countries — Uttar Pradesh outnumbers Russia; Maharashtra runs about level with Germany. That scale is the rationale for a third tier below the state: local government, built through decentralisation — the deliberate act of taking power away from central and state governments and handing it down to local government instead.
Decentralisation rests on a simple bet: many problems are best solved by the people who actually live with them. Local people know local problems better than a state capital ever will, judge better how to spend money and manage things efficiently for their own area, and get to participate directly in decisions that affect them — which builds the habit of democratic participation, not just the theory of it. Local government, on this bet, is the best available way to realise the democratic principle of local self-government.
India recognised the need for decentralisation early — Panchayats for villages and Municipalities for towns existed, on paper, in every state well before 1992. But they were controlled directly by state governments, elections to them were not held regularly, and they held no independent powers or resources of their own. Decentralisation, in effective terms, stayed very limited for decades despite the institutions already existing.
1992 changed that on paper, with five specific provisions. Regular elections to local government bodies became constitutionally mandatory, not optional. Seats in elected bodies, and executive-head posts, were reserved for Scheduled Castes, Scheduled Tribes, and Other Backward Classes. At least one-third of all positions were reserved for women. An independent State Election Commission was created in every state, specifically to conduct panchayat and municipal elections. And state governments were required to share some of their own powers and revenue with local government — though exactly how much, the amendment left for each state to decide for itself.
Rural local government — popularly called panchayati raj — is built up from its smallest working unit: the gram panchayat, a council for one village, or a small group of villages in some states, made up of ward members (called panch) and a president or sarpanch, every one of them directly elected by the adult population of that ward or village. The gram panchayat does not operate unsupervised — the gram sabha, made up of every voter in the village, oversees it, meeting at least two or three times a year to approve the annual budget and review the panchayat's own performance.
The rural structure does not stop at the village. Several gram panchayats group together into a panchayat samiti — also called a block or mandal — whose own members are elected by all the panchayat members across that area. All the panchayat samitis in a district together form the zilla parishad, the district-level body, whose members are mostly elected but which also includes that district's Lok Sabha MPs, MLAs, and officials from other district-level bodies; a chairperson heads it politically. Village, block, district — three tiers stacked on top of the gram sabha's own oversight, each one built from the level below it.
Urban local government mirrors the same idea in a different setting. Municipalities run in towns; big cities are constituted into municipal corporations instead. Both are controlled by elected bodies of people's representatives, exactly like their rural counterparts — a municipality's political head is its chairperson, and a municipal corporation's equivalent officer carries a different title: mayor.
Put a number on what this system has actually built, and the scale is hard to overstate. About 36 lakh elected representatives serve across India's panchayats and municipalities — more people than the entire population of many countries. Constitutional status for local government has genuinely deepened Indian democracy. It has also increased women's representation and voice inside it, in a way few other single reforms have.
But real gaps remain alongside that achievement. Gram sabhas are not held regularly. Elections to local bodies, by contrast, are held regularly and, by most accounts, enthusiastically. Most state governments have not transferred meaningful powers or adequate resources down to the local governments the 1992 amendment created. India's local government is, honestly, both things at once. It is the largest democratic experiment of its kind anywhere in the world. It is still a long way from the self-government it was built to deliver.
Decentralisation is not only an Indian question. Porto Alegre, in Brazil, ran a participatory-budgeting experiment that combined decentralisation with direct participative democracy — a body running alongside the elected municipal council let the city's roughly 13 lakh residents help decide the city's own budget directly, through ward-level and city-wide meetings that around 20,000 people attend every year. That direct participation kept spending from concentrating only in the richer neighbourhoods — buses now reach poorer colonies, and builders cannot evict slum-dwellers without resettling them first. A comparable experiment has already run in parts of Kerala, where people took part directly in planning their own locality's development.