Arvind Datar on Delimitation, Federalism & India's Constitutional Future

ELI5/TLDR

A Tamil Nadu-funded committee, headed by a retired Supreme Court judge, wrote a giant report suggesting big changes to how India runs itself. Constitutional lawyer Arvind Datar goes through it and mostly says: nice try, but the Constitution is fine, don’t rewrite it. His sharpest points are about delimitation — the redrawing of how many parliamentary seats each state gets based on population — which he argues should just be frozen forever because letting the populous north grow its share would crush the south and tear the country apart. He also wants English protected as the link language, defectors permanently barred from re-contesting, and governors given a written rulebook so they stop misbehaving.

The Full Story

This is a 43-minute conversation between Arvind Datar — one of India’s leading constitutional lawyers, currently writing the third edition of his commentary on the Constitution — and an interviewer (the veteran psephologist Dorab Sopariwala), going through a report by the Justice Kurian Joseph committee on centre-state relations, commissioned by the Tamil Nadu government. The report is sprawling, covering ten areas, and reads almost like a call for a new constitutional convention. Datar’s overall verdict sets the tone: don’t.

The constitution is absolutely fine. Some of the recommendations… are quite useful… But this will just require a few amendments to few articles. There’s no need of an overhaul of the constitution at all.

How hard should it be to amend the Constitution?

India has amended its Constitution 106 times in roughly 75 years — about one and a half amendments a year. The United States has managed 27 in over two centuries. There are three difficulty levels for amending: a simple majority, a two-thirds “special majority,” and the hardest tier that also needs ratification by the states.

Datar flags a genuine absurdity buried in this structure. Fundamental rights — free speech, equality, life — can be amended with a two-thirds majority alone, no state ratification required. But amend some minor article touching a state subject, and you need the states to sign off. So the most precious things in the document are actually easier to tamper with than the plumbing. The Supreme Court spotted this paradox back in 1965; a judge called fundamental rights “the playthings of the majority.” Datar’s fix: require a three-fourths majority to touch fundamental rights, insulating them from a government that happens to control two-thirds of Parliament.

Basic structure: vague on purpose

The “basic structure doctrine” is the idea that some features of the Constitution are so essential that even Parliament cannot amend them away. The interviewer raises the standard objection — it’s never been defined, so how can you write it down? Datar borrows the famous American line about pornography (“I know it when I see it”) and argues the vagueness is the point, not a bug.

His reasoning is grounded in a 1965 case. A judge named Mudholkar asked: if you delete the right to equality, the right to life, and the right to free speech, can what remains still be called the constitution of a republican democracy? The essential features, he said, are written right there in the preamble — liberty, equality, fraternity, social justice.

The very flexibility of this term is its strength. Like there is no definition of due process in the American constitution… Its inability of definition may be its greatest strength. It can be molded to meet the necessities of generations.

He also notes that in the 53 years since the landmark Kesavananda Bharati judgment, the court has used basic structure to strike down an amendment only five or six times — so the fear of an out-of-control judiciary is overblown.

The committee suggested two more tweaks here, and Datar splits the difference: writing “basic structure cannot be amended” into the text is unnecessary (the courts already enforce it), but requiring a constitutional amendment bill to be published three months in advance for public consultation is a good idea. He rejects a third proposal — that certain amendments expire (“sunset”) after 10 or 15 years unless re-enacted — as impractical. You can’t reliably predict which provisions will need rethinking, and a genuinely useful amendment might die simply because the politics of the day won’t allow its renewal.

Delimitation: the hot potato he wants in cold storage

This is the heart of the interview. Delimitation means redrawing the number of Lok Sabha seats each state gets, in line with population. It has been frozen since the 1971 census — Parliament currently has 550-odd seats — and the freeze is set to lift after 2026.

Here’s why it’s explosive. The southern states ran successful family-planning programs and slowed their population growth; the northern states did not. So if seats are reallocated by current population, the south loses representation and the north gains it — a reward for the north’s failure and a punishment for the south’s success. Datar puts the stakes bluntly:

If you impose delimitation now everybody in the country knows it’s going to impact the southern states very badly and it’s going to benefit the northern states. So virtually after some time you don’t need the southern states. The northern states can pass through any amendment without bothering about the southern states.

The numbers make it vivid: the southern states together have about 27 crore people and 129 seats; Uttar Pradesh alone has about 24 crore and 80 seats. Forecasts suggest UP’s population will eventually run far ahead of the south, because the south’s numbers are falling faster.

Datar’s position is to permanently freeze the count at 550, the way the US Senate is permanently fixed at 100 members regardless of how big a state grows. If the country insists on a bigger Parliament, his fallback is to scale everyone up in the same proportion — if Tamil Nadu has 40 of 550 today, it keeps the same share in a house of 700 or 1,000. That way no state can claim it was diluted.

But he questions whether a bigger Parliament does anything useful at all. The constituent assembly’s original ideal was one representative per million people, which today would mean 1,400 MPs — clearly unworkable. And more bodies don’t mean more governance:

How will 800 members going to improve the functioning of parliament? If you are not going to refer bills to the parliamentary committees, if there’s not going to be proper debates… I really don’t know what purpose will be served.

He cites data (from the legislative-research group PRS) that roughly half of current MPs don’t speak in Parliament at all — they’re “lobby fodder” who march in and press the voting button. He also points out the hidden tail of delimitation: it would mean expanding every state assembly too, multiplying the cost for no clear gain.

The interviewer floats a clever alternative: make the Rajya Sabha (the upper house, currently the weaker chamber) as powerful as the US Senate, where every state — big or small — gets equal representation. That would let the south retain clout even as Lok Sabha numbers shift. Datar rejects it on structural grounds: India inherited the Westminster model, not the American one, and the Rajya Sabha is a “house of elders,” not a co-equal Senate. Bolting on an American fix would be a mismatch. His advice stays the same — keep the current ratio, scale proportionally if you must, and otherwise leave it alone. He’s clear this is his personal view, repeated several times.

Language: protect English, don’t force Hindi

The report wants to abandon the “one nation, one language” idea and entrench English permanently as the union’s link language and the language of the constitutional courts. Datar broadly agrees, and frames it around mobility rather than identity.

Forcing Hindi nationwide, he warns, would trigger upheaval — Tamil Nadu’s politics is built on an anti-Hindi platform, and the anti-Hindi riots of the mid-1960s showed how violent it can get. Meanwhile, on the ground, demand for English is surging: his wife runs schools through an NGO, and regional-language schools are losing students because English is the route to jobs and upward mobility. He invokes Lee Kuan Yew’s first decision as Singapore’s PM — making English the working language despite a Chinese, Malay and Tamil population — as the model.

When you make it compulsory, there’s a backlash. So let’s just leave it to the local population. Let them read English and their local language, and Hindi can be an optional subject.

There’s a real equity gap underneath this. By the 2011 census, about 15% of people in the five southern states have some knowledge of English; in the five largest northern states it’s about 3.5%. The north has been left behind in English the way the south has in Hindi, and there’s no quick fix — Datar reckons it would take decades to train enough teachers. He flags a quiet absurdity: the language of the courts is English under the Constitution, yet a district judge in the north who knows only the regional language could, in principle, rise to a High Court where judgments must be in English. His policy: let the north start now and catch up over five to ten years, with English offered as an option everywhere rather than imposed.

Elections, freebies, and defection

The report proposes a separate state election commission for each state. Datar dismisses this — there’s no reason a state-appointed body would be more independent than the central one; it might be worse. The deeper rot, he says, is two things: freebies and defection.

On freebies — election promises of free goods and cash — he notes a legal loophole. Under the law, a direct gift to a voter is a bribe that can void an election, but a party promising the same thing in its manifesto is not. He says some states now spend 40-48% of their budget on freebies, leaving nothing for capital expenditure or growth. He lost a case in the Supreme Court (Subramaniam Balaji) trying to get them banned; the Election Commission’s response — it would ban freebies only if all parties agreed — he treats as a polite never. His line: ban financial incentives outright (food security already exists, so why more handouts).

On defection, his proposed cure is the harshest part of the interview. The anti-defection law (the Tenth Schedule) keeps getting gamed — through mass resignations (as in Karnataka), through the loophole that lets two-thirds of a party defect en bloc without penalty, and through partisan Speakers who are supposed to be neutral but never are (“I’ve not seen a single case of a speaker being impartial”). Datar would slam every door:

Once you stand on party X ticket, for that 5-year term, if you want to change you just resign — and you cannot contest elections for the next remaining term of that particular assembly.

So a defector can’t switch, can’t resign-and-recontest, and can’t form a breakaway faction. He’d also close the side-door reward: a defector who resigns shouldn’t then be handed a lucrative chairmanship of some state commission (the courts have already struck down attempts to dodge the cap on ministers via “parliamentary secretary” posts). He grounds all of this in Ambedkar’s closing words to the constituent assembly — that even a good constitution will fail if worked by bad people — and the “shameful” spectacle of MLAs being herded into five-star hotels with their phones confiscated so rivals can’t poach them.

Governors: keep them, but hand them a rulebook

The report’s best idea, in Datar’s view, is a new 13th Schedule listing the dos and don’ts of a governor. He finds it paradoxical that governors — meant to be elder statesmen with gravitas — need to be told how to behave at all, like junior clerks. But governors have been openly partisan for 30-40 years, especially in opposition-ruled states. The trigger example is Tamil Nadu, where the governor sat on 12 bills for 24 months without assent. A presidential reference ruled there can be no hard time limit, but Datar argues a three-month outer limit is obviously reasonable — even the central Home Ministry has said so. A simple bill can be cleared in a day or two; even a complicated one needing a check for conflict with central law shouldn’t take more than a month.

The interviewer pushes harder: do we need governors at all? Germany’s states have no equivalent. Datar resists abolition. Strip the ceremonial role away and the governor still performs real constitutional functions — recommending President’s Rule, exercising the power of pardon, flagging state laws that conflict with central law. Remove the post and you’d put the whole architecture into jeopardy.

I can’t think of a constitution without a governor. That will be too radical a change. Instead of going down that path… put it in writing. A bill will be passed in 3 months. So just put it in writing as the committee said in the 13th schedule.

He also endorses the report’s suggestion of a single five-year term with nothing afterward, so a governor isn’t angling for the next posting.

The verdict on the report

Datar closes generously. A lot of study went into the report; it deserves credit and, more importantly, debate. This is the third such exercise after the Sarkaria and Punchhi commissions on centre-state relations. The fact that it was commissioned by an opposition state government, he says, is no reason to dismiss it — it was headed by a Supreme Court judge and considered countless submissions. His final ask: Parliament should actually debate it.

Key Takeaways

India has amended its Constitution 106 times in ~75 years (about 1.5/year); the US has 27 in over 200 years.
Constitutional amendments come in three tiers: simple majority, two-thirds special majority, and special majority plus state ratification.
A real paradox: fundamental rights can be amended with two-thirds alone (no state ratification), making them easier to alter than minor articles touching state subjects. The Supreme Court called them “the playthings of the majority” in 1965. Datar wants a three-fourths bar to protect them.
The “basic structure doctrine” holds that some features (liberty, equality, fraternity, social justice — drawn from the preamble) cannot be amended away even by Parliament. Established in the 1973 Kesavananda Bharati case.
Basic structure has been used to strike down an amendment only ~5-6 times in 53 years — courts mostly defer to Parliament.
Delimitation = redrawing each state’s share of parliamentary seats by population. Frozen since the 1971 census; freeze due to lift after 2026.
The north-south split: southern states curbed population growth, northern states didn’t. Reallocating seats by current population rewards the north and shrinks the south’s voice. Southern states (~27 cr) hold 129 seats; UP alone (~24 cr) holds 80.
Datar’s position: permanently freeze the seat count at 550 (analogy: the US Senate is fixed at 100). If expanding, scale every state up in the same proportion so no one is diluted.
The constituent assembly’s original ideal was 1 MP per million people — today that would mean ~1,400 MPs.
~Half of current MPs reportedly never speak in Parliament (“lobby fodder”); a bigger house likely means more cost, not better governance.
Why the Rajya Sabha can’t simply become a US-style Senate: India runs the Westminster model, where the upper house is a “house of elders,” not a co-equal chamber.
English literacy gap (2011 census): ~15% in the five southern states vs ~3.5% in the five largest northern states.
Datar’s language model is Lee Kuan Yew’s Singapore — adopt English as the link language for mobility, keep regional languages, make Hindi optional. Compulsion breeds backlash.
A legal loophole on freebies: a direct gift to a voter is a bribe that can void an election; an identical promise in a party manifesto is not. Some states reportedly spend 40-48% of their budget on freebies.
Anti-defection law (Tenth Schedule) is gamed via mass resignations, the two-thirds en-bloc defection exemption, and partisan Speakers. Datar’s fix: a defector can’t switch, can’t resign-and-recontest within the term, and can’t form a breakaway faction — plus no consolation chairmanships.
The report’s standout idea: a 13th Schedule codifying a governor’s dos and don’ts, with a 3-month outer limit to act on bills (Tamil Nadu’s governor sat on 12 bills for 24 months).
Datar opposes abolishing governors — they retain real functions: recommending President’s Rule, the power of pardon, and flagging state laws that conflict with central law.
Ambedkar’s closing warning to the constituent assembly: a good constitution worked by bad people will fail; a bad one worked by good people can succeed.
This is the third centre-state relations exercise after the Sarkaria and Punchhi commissions.

Claude’s Take

This is a high-signal conversation — two people who actually know the material, no posturing, no soundbites. Datar is the rare expert who’ll say “that’s not practical” to half the proposals on the table even though the report is on his side of the political fence, and that intellectual honesty is what makes it worth the 43 minutes. The interviewer is no slouch either, pushing back with real data (the PRS speaking-time numbers, the census English figures) rather than lobbing softballs.

The delimitation discussion is the keeper. It’s the most consequential federalism problem India faces in the next decade and it’s badly under-discussed precisely because it’s a no-win: the just-sounding option (one person, one vote, equal seats per capita) would hollow out the states that governed themselves best, and the stable option (freeze it) means living with permanent malapportionment. Datar’s “freeze at 550 or scale proportionally” is a clean framing of the least-bad path, and the US Senate analogy lands. His weakest moment is brushing off the Rajya Sabha-as-Senate idea a little quickly — “we chose Westminster” is a description, not an argument, and the interviewer’s point deserved more than a structural shrug.

The defection cure is bracing but I’d flag it as the one place to read skeptically: permanently barring a resigned legislator from contesting for the rest of the term is a real restriction on the right to stand for election, and the line between a corrupt defection and a genuine split of conscience is exactly the thing courts have struggled to draw. He’s diagnosing a real disease; the cure might have side effects he waves past.

Docking it slightly for the audio-transcript noise (some mangled names and numbers — “Korean Joseph” is Kurian Joseph, dates garbled) and the rushed governor section at the end. But on substance this is one of the more clarifying explainers of Indian constitutional plumbing you’ll find. Eight.