The Part That Is Still Moving
Every other part of this series describes something finished. This one describes something happening now, and some of it will have changed by the time you read it.
That is uncomfortable to write and it is the point. The Reservation Hatao Andolan appeared in July 2026 demanding the abolition of a system it describes as frozen, unexamined and permanent. In fact the seven years before it were the most active period of change in Indian reservation since 1990.
Since 2019 the fifty per cent ceiling has been broken, a caste-blind quota has been created and upheld by a single vote, the Scheduled Caste category has been split for the first time, two states have already implemented the split, and India has begun counting caste again after ninety-five years.
And there is one more thing, which is the largest of all and which almost nobody in the July argument has mentioned.
In April 2026, three months before the Reservation Hatao Andolan existed, Parliament sat in special session to consider raising the Lok Sabha from 543 seats to 850. Reserved seats under Articles 330 and 332 are proportional. Which means the single biggest expansion of political reservation in Indian history is currently before Parliament — and the national argument is about medical college cut-offs.
Chapter 8 covers it. I think it is the most important chapter in this part, and I would be surprised if one reader in a hundred has heard about it.
Two warnings. First, this part contains figures that are current as of July 2026 and are moving; where something is under litigation or before Parliament, I say so. Second, I have tried to give the Reservation Hatao Andolan a fair hearing on its own strongest ground — Chapter 10 — because in this landscape they have one very good point and it is not the one they lead with.
— Lovepreet Singh
July 2026
1The Ceiling Breaks
A limit that had stood for twenty-seven years was removed by a constitutional amendment that passed in two days.
In January 2019, Parliament passed the 103rd Constitutional Amendment, inserting Articles 15(6) and 16(6) and creating a 10% reservation for Economically Weaker Sections among those not covered by any existing quota.
The Bill was introduced and passed by the Lok Sabha on 8 January 2019 and by the Rajya Sabha on 9 January 2019. It received assent within days.
An amendment to the Constitution of India, breaching a ceiling the Supreme Court had established in 1992 and traced in Part 8 to Ambedkar's own words in 1948, was through both Houses in roughly forty-eight hours.
Compare the timelines elsewhere in this series. Mandal waited nine and a half years on a shelf. The Balakrishnan Commission on Dalit converts, from Part 4, has been extended repeatedly since 2022 and has still not reported.
What EWS actually is
| Test | Threshold |
|---|---|
| Who is eligible | Those not covered by SC, ST or OBC reservation — in practice, the general category |
| Family income | Below ₹8 lakh a year |
| Agricultural land | Below 5 acres |
| Residential flat | Below 1,000 square feet |
| Residential plot | Below 100 square yards in a notified municipality; below 200 square yards elsewhere |
SC 15% + ST 7.5% + OBC 27% + EWS 10% = 59.5%, leaving 40.5% open.
The Indra Sawhney ceiling had held since 1992. It no longer does.
The first is about what EWS proves. India now runs a caste-blind, income-tested quota, and has done since 2019. The Reservation Hatao Andolan's central proposal — replace caste with economic criteria — is not hypothetical. Ten per cent of central seats already work that way, and Chapter 3 examines how.
The second is about the exclusion. EWS is available only to people who get no other reservation. A poor Dalit family and a poor Brahmin family with identical incomes are treated differently: one is eligible, one is not.
Defenders say this is obvious — the excluded groups have their own provision. Critics say a poverty quota that excludes the poorest communities in the country is not a poverty quota. That disagreement is the whole of Chapter 2, and it split the Supreme Court down the middle.
2Three To Two
The 2022 judgment upholding EWS. Both dissents, and the warnings the majority itself issued.
In November 2022, a five-judge bench of the Supreme Court decided Janhit Abhiyan v. Union of India. EWS survived by three votes to two.
The majority
1. Reservation based purely on economic criteria is permissible. The Constitution does not confine special provision to social and educational backwardness.
2. Excluding SCs, STs and OBCs from EWS does not violate equality, because those groups already have their own reservation. Including them would give them a double benefit.
3. The 50% ceiling is not inviolable. It applies to reservation under Articles 15(4) and 16(4), and does not bind a separate provision created by a new amendment.
The dissents
On the exclusion. The core objection was that this was the first time the Court had approved a scheme built on an exclusionary principle. A quota justified by poverty that specifically bars the poorest communities in India is not a poverty measure. Being poor and being Dalit are not alternatives; the same person can be both, and the scheme tells them to choose.
On the ceiling. Once 50% is treated as breachable by creating a new provision, there is no principled stopping point. Any future government can add another category by amendment and the limit becomes decorative.
Judges in the majority separately observed that reservation should not be allowed to become a vested interest, that the policy required review, and that a time limit should be considered — echoing the framers' expectation set out in Part 8.
So of five judges hearing the case, two would have struck the scheme down and at least two of the remaining three attached explicit warnings about the direction of travel.
For the abolition side: the majority's reasoning is the single most useful thing they have. A court has now held that economic-only reservation is constitutionally permissible — reversing the position in Indra Sawhney that Part 9 described. The legal obstacle to their proposal has been removed by a judgment they rarely cite.
For the retention side: the dissent is the strongest statement anyone has made of what is wrong with replacing caste with income. Not that it is unfair to the beneficiaries of the current system, but that it does not do the thing it claims to do — it excludes the poorest.
And for everyone: a single vote. Had one judge gone the other way, EWS would not exist, the ceiling would still stand, and the argument in July 2026 would be taking place on completely different ground.
3Eight Lakh
India has been running an economic quota for seven years. This is what it has taught us — and the finding is bad for both sides.
The most useful thing about EWS is not the argument about it. It is that it is a live experiment in exactly the policy the Reservation Hatao Andolan proposes, and it has been running since 2019.
The threshold problem
₹8 lakh a year is roughly ₹67,000 a month. Critics have consistently pointed out that at that level, the overwhelming majority of Indian households — on many estimates well over ninety per cent — fall below the line.
A category defined so broadly is not identifying the economically weaker. It is identifying almost everyone who is not already covered by another quota.
When the criteria were challenged, the Supreme Court asked what study the ₹8 lakh figure was based on. The government's position was that it had adopted the same figure used for the OBC creamy layer.
A committee was then constituted to examine the criteria. It recommended retaining ₹8 lakh.
Which produces the situation already noted in Part 1 and Part 9, and it is worth stating in its final form.
And the part that repeats 2008
As with the OBC round of 2006–08 described in Part 9, central educational institutions increased total intake — by roughly a quarter — so that the absolute number of seats available to unreserved candidates would not fall.
Critics noted at the time that the expansion was not matched by a corresponding increase in faculty, in a system where a large proportion of teaching posts were already vacant.
Against the abolition case: an income-based quota does not escape the problems of a caste-based one. It still requires a category, a threshold, verification, certificates, and an administrative machine. It still generates disputes about who qualifies. And its threshold was set by copying another quota's number rather than by measuring poverty. Replacing caste with income does not produce a neutral system; it produces a differently arbitrary one.
Against the retention case: the political system created and passed a caste-blind quota in forty-eight hours, and the Supreme Court upheld it. The claim that economic criteria are constitutionally impossible is no longer true, and defenders who still make it are working from the 1992 position.
And the observation that should trouble everyone: the EWS expansion, like the 2008 expansion, worked by making the pie larger. Both times it defused the conflict. Nobody drew the general lesson, and nobody is applying it now.
4The Irony
The most powerful farming communities in India have spent a decade demanding to be officially classified as backward. That fact tells you more than any argument.
While one movement demands the abolition of reservation, several very large communities have been agitating — with road blockades, mass rallies and sustained political pressure — to be included in it.
Maharashtra, now
Passed on 20 February 2024, the Act provides 10% reservation for Marathas in state jobs and educational institutions, on the basis of findings by a commission headed by Justice Sunil Shukre.
It takes total reservation in Maharashtra to approximately 62%, well past the Indra Sawhney ceiling.
The Act is under challenge before a specially constituted bench of the Bombay High Court. Admissions and appointments made under it are being permitted for now, subject to the outcome.
Separately, the activist Manoj Jarange-Patil has pressed the demand that Marathas be recognised as Kunbi and therefore admitted to the existing OBC category — a demand the state government has resisted, and which OBC organisations have opposed on the ground that it would dilute their own share.
1. Reservation is understood by its opponents to be valuable. Communities do not blockade highways for a benefit they believe is worthless or stigmatising. The demand is the strongest available evidence that the thing works well enough to be worth having.
2. The categories are political, not scientific. A community can be added to a list and removed by a court within a year, as the Jats were. Whatever "backward" means administratively, it is not a stable measurement.
3. The merit argument evaporates on contact with self-interest. The communities demanding entry are, in many cases, the same social base from which anti-reservation sentiment has historically drawn. The objection is rarely to quotas as such. It is to being outside one.
4. And the real conflict is horizontal, not vertical. The fiercest resistance to Maratha inclusion has come not from upper castes but from existing OBC organisations, whose share would be diluted. Part 3 predicted exactly this: the sharpest caste conflict in India runs between adjacent rungs, not between top and bottom.
5The Ceiling Is Under Siege Everywhere
The 50% limit is not just breached at the centre. Half a dozen states have gone past it, and the courts have been striking them down one by one.
Most of this argument is conducted as though reservation is a single national policy. It is not. Most Indian government jobs and most college seats are state seats, and states set their own percentages.
And across the last few years a pattern has emerged that neither side discusses: state after state has tried to cross 50%, and court after court has said no.
Bihar's 65% was struck down in 2024 for breaching 50%. Chhattisgarh's 58% was set aside on the same ground.
Meanwhile the central total has stood at 59.5% since 2019, upheld by the Supreme Court in 2022 — because EWS was created by constitutional amendment rather than by state legislation, and the majority held that the ceiling does not bind a new provision.
The rule is now: you may exceed 50% if you can amend the Constitution, and not otherwise.
Which means the limit is not a principle about equality at all. It is a test of whether you control a two-thirds majority in Parliament.
For the abolition side: this is a much better argument than the ones they use. A ceiling that binds Bihar but not Delhi is not a constitutional safeguard; it is an artefact of who has the votes.
For the retention side: the same fact is uncomfortable in the other direction. States with genuine claims — Bihar acted after conducting its own caste survey, which is more evidence than EWS was built on — are being blocked, while a quota with an unexplained threshold sails through because it came with an amendment attached.
And notice the escape route everyone is now reaching for. Karnataka wants the Ninth Schedule. So, periodically, do others. The device Nehru's government created in 1951 to protect land reform has become the standard method of putting a reservation policy beyond judicial reach.
Part 1 noted the argument in both directions about Tamil Nadu. Here is the narrow point that survives it.
A state has run 69% reservation for three decades. Whatever one believes about causation, the predicted institutional collapse did not occur. Any confident claim that crossing 50% is catastrophic has to explain thirty years of a large state doing it.
6The Category Splits
On 1 August 2024 the Supreme Court permitted states to divide the Scheduled Caste quota. Two states have already done it.
This is the most significant structural change to Indian reservation since 1990, and it received a fraction of the attention that a viral Instagram account got two years later.
The Court upheld the constitutional validity of sub-classification within the Scheduled Castes, overruling E. V. Chinnaiah v. State of Andhra Pradesh (2004), which had held that the Scheduled Castes were a single homogeneous class that could not be divided.
States may now split the SC quota, giving the most deprived sub-castes a protected share within it.
And the part that went further
His view was that certain groups within the Scheduled Castes and Tribes have been oppressed for centuries and that state governments must identify them — and that the criteria for identifying the creamy layer among SCs and STs should differ from those used for OBCs, reflecting different circumstances.
This extends the principle from Jarnail Singh (2018), described in Part 8, from the promotion context into the general one.
What has actually happened since
The Supreme Court has formally accepted elite capture. Sub-classification exists because benefits within a reserved category concentrated unevenly. That is the argument Thomas Sowell makes comparatively, that the abolition side makes rhetorically, and that the retention side has often dismissed as bad faith. It is now a finding of a seven-judge bench.
But look at what the remedy was. Not abolition. Not reduction. Better targeting inside the existing quota. Telangana and Karnataka split the 15% rather than shrinking it.
And notice where the argument has moved. The serious institutional debate in India is no longer about whether there should be reservation. It is about who inside the categories should receive it. That is a completely different question from the one being fought in comment sections, and it is the one actually being decided.
Sub-classification has been permitted for the Scheduled Castes. It has not been done nationally for the Other Backward Classes — a category of 3,743 castes, as Part 9 noted, containing communities with wildly different circumstances.
If elite capture justifies splitting a category of a few hundred, the case for examining a category of nearly four thousand is considerably stronger. Nobody at the centre has attempted it.
7The Count Returns
For the first time since 1931, India is counting caste. The field work started three months before the current argument began.
Every number in Part 9 rested on arithmetic applied to a census conducted in 1931. That is about to end.
Phase 1 — Houselisting and Housing Census: 1 April to 30 September 2026, across all 36 states and union territories. Self-enumeration was available from 1 to 15 April; house listing from 16 April.
Phase 2 — Population Enumeration: February 2027. Caste data is collected in this phase.
For snow-bound regions including Ladakh, Jammu and Kashmir, Himachal Pradesh and Uttarakhand, the reference date is 1 October 2026.
It is India's first digitally conducted census, using a mobile application, with a self-enumeration option. By 10 April 2026, more than 5.7 lakh households had used self-enumeration.
What it will settle
- The 52% question. Mandal's estimate of the OBC share has carried enormous political weight for forty-six years on 1931 data. It is about to be tested.
- The sub-classification question. Splitting a category sensibly requires knowing the size and condition of its parts. Telangana and Karnataka have had to commission their own studies. After 2027 there will be national data.
- The ceiling question. If the reserved-category population is far above the reserved share, the pressure to breach 50% becomes very hard to resist. If it is far below, the opposite.
Consider how little confident prediction there is from either camp about a dataset that will decide the terms of this argument.
The retention side is not campaigning on it — presumably because if OBCs turn out to be well under 52%, the case for 27% weakens and the demand for proportional representation becomes harder to make.
The abolition side is not campaigning against it — presumably because opposing a count looks like fearing the answer, and because if the numbers are low it helps them.
Both are waiting. Which is a reasonable thing to do, and it makes the timing of a movement demanding immediate abolition eleven months before the count somewhat strange.
Counting caste tells you how many people are in each category. It does not tell you how deprived they are, and the two are different questions.
Part 5 described what happened the last time India ranked communities officially: thousands of petitions, not to abolish the ranking but to move up within it. A caste census produces numbers, and numbers produce claims. Expect a great many of them.
8543 To 850
The largest expansion of political reservation in Indian history is before Parliament right now, and it is not being discussed in the argument about reservation.
In April 2026 — three months before the Reservation Hatao Andolan existed — Parliament held a special session on a package of bills that will reshape Indian representation more than anything since 1950.
Three interconnected bills were introduced: the Constitution (131st Amendment) Bill, 2026; the Delimitation Bill, 2026; and the Union Territories Laws (Amendment) Bill, 2026.
The central proposal is to increase the strength of the Lok Sabha from 543 seats to 850.
The exercise is tied to the 2027 census, and commentary has noted that it is unlikely to be completed before the 2029 general election — which would mean the women's reservation provided for by the 106th Amendment of 2023 does not apply to that election either.
Now do the arithmetic
Seats reserved for Scheduled Castes and Scheduled Tribes under Articles 330 and 332 are allocated in proportion to population.
1. It is the 2008 method, at constitutional scale. Part 9 identified expansion as the only demonstrated way of implementing reservation without a crisis: increase the total so the unreserved number does not fall. Delimitation does exactly that with parliamentary seats — more reserved seats and more general seats. Nobody planned it as a reservation policy. It functions as one.
2. It creates a north–south collision. Delimitation by population rewards states whose populations grew fastest and penalises those that controlled growth most successfully — which, broadly, means shifting weight from the south to the north. The southern states have been raising this for years. Combine it with a caste census and with women's reservation and you have three enormous redistributions of political power landing at once.
3. And it is where the real decisions are being made. A movement with 5.8 million followers is demanding changes to exam cut-offs. Parliament is considering the composition of the legislature itself, on a timescale of years, with consequences measured in decades.
That gap between where the attention is and where the power is may be the most important observation in this book.
9The Three Places It Does Not Exist
Reservation applies to government. Three of the most powerful sectors in India are outside it — and their composition is instructive.
1. The private sector
It has been proposed and debated for decades. It has never been enacted. Since the overwhelming majority of Indian jobs are private or informal, the entire national argument concerns a small and shrinking share of employment.
And Part 6 measured what happens there. The correspondence study — identical CVs, different surnames, private-sector graduate jobs — produced ten interview calls for an upper-caste name, six for a Dalit name, three for a Muslim name.
The measured discrimination is highest in the sector where reservation has never applied.
Which damages the claim that reservation causes resentment which causes discrimination — the effect appears where the cause is absent.
And damages the claim that reservation is a sufficient remedy — it does not reach the place where most jobs are and where the gap is widest.
2. The judiciary
Appointments to the higher judiciary are made through the collegium system, and reservation does not apply. The resulting composition of the High Courts and the Supreme Court has been a subject of parliamentary questions and public comment for years, and is markedly unrepresentative of the population.
Neither side campaigns on it. The abolition side does not, because it is an unreserved institution that is visibly monopolised. The retention side does not, because the institution in question decides their cases.
3. The newsroom
Surveys of Indian media leadership have repeatedly found that senior editorial positions across print, television and digital are held overwhelmingly by people from a small number of upper-caste communities, with Dalit and Adivasi journalists almost absent at that level.
The media is not a large employer. It is the institution that decides which arguments are heard, which grievances are covered as news, and whose anger is treated as legitimate.
Consider Part 1 again. A movement demanding the abolition of reservation received extensive national coverage within days of appearing. Ask how much coverage the Balakrishnan Commission on Dalit converts has received in four years, or the Telangana sub-classification Act, or the 543-to-850 proposal.
This is not a conspiracy claim. Newsrooms cover what their staff find interesting and urgent, and staff find interesting what they and people like them experience. That is exactly why composition matters.
10Where RHA Actually Fits
Their four demands, placed against the landscape this part has described. One of them is much stronger than the rest.
Part 1 set out the Reservation Hatao Andolan's demands as published: Merit over Caste, Equal Fees, Equal Age Limit, Equal Cut-off — with the broader proposal that the state should stop recording caste at all.
Here is how each stands against what has actually been happening.
| Demand | Assessment against this part |
|---|---|
| Merit over caste | India already runs a caste-blind quota — EWS, since 2019. Chapter 3 shows what it produced: a threshold copied from another scheme, covering almost everyone eligible, requiring the same certificates and machinery. Removing caste does not remove categories. |
| Equal cut-off | Part 1 showed General and OBC closed two marks apart for all-India government MBBS seats in 2025. The gap this demand targets is largely not where it is believed to be. |
| Equal fees and age limit | Legitimate and specific. These are administrative relaxations, distinct from seat reservation, and arguing about them separately would be a more productive conversation than the one being had. |
| Stop recording caste | Directly contrary to Census 2027, which began in April 2026. It would also make the 97% manual-scavenging figure from Part 6, and the sub-classification exercises in Telangana and Karnataka, impossible to produce. |
Look at what Chapters 3, 4, 5 and 6 establish together.
The EWS threshold was set by copying another quota's number, and the government could not point to a study behind it. A community can be added to the OBC list and removed by a court within a year. Total reservation is 59.5% at the centre and about 62% in Maharashtra, both past a ceiling that was supposed to be constitutional. And the Supreme Court has now held that the benefits inside a reserved category are so unevenly distributed that the category must be split.
The honest version of the abolition case is therefore not "reservation is unjust." It is that the categories are incoherent — arbitrary at the threshold, unstable in membership, uncapped in total, and unevenly distributed inside.
That is a serious argument, it is supported by everything in this part, and it is very rarely the one that gets made.
Incoherent categories are an argument for fixing the categories. The 2024 sub-classification judgment, the Telangana Act of 2025 and the Karnataka formula of 2026 are precisely that work being done — and none of them abolished anything. Each split a quota to target it better.
Census 2027 is the evidence base that has been missing since 1931, and it arrives in eight months.
The abolition case observes correctly that the machine is badly built. What it has not established is why the response to a badly built machine is to remove it rather than to rebuild it — particularly when Part 6 established that the thing it was built for is still measurably happening.
That is the question Part 12 puts to both sides at full strength, and Part 14 answers.
11Nine Questions Nobody Asks
The blind spots in this stage of the argument.
1. Why did EWS take two days when Mandal took nine years?
A constitutional amendment breaking a twenty-seven-year-old ceiling passed both Houses on 8 and 9 January 2019. The Mandal report sat unimplemented for nine and a half.
The difference is not evidence, urgency or constitutional necessity. It is whose interests were served and how close the election was. Both facts belong in any honest account of how this policy area actually works.
2. If economic criteria are the answer, why is the threshold ₹8 lakh?
The number was adopted from the OBC creamy layer, not derived from a study of poverty, and a committee later recommended keeping it.
Anyone proposing to replace caste with income has to answer the question the government could not: on what basis is the line drawn, and by whom?
3. Why does the fiercest opposition to Maratha inclusion come from OBCs?
Because the 27% would be shared with a very large new claimant. The conflict is between adjacent groups competing for the same pool, exactly as Part 3 predicted.
The abolition side treats reservation as a conflict between reserved and unreserved. Most of the actual political conflict is inside the reserved side.
4. Why has nobody sub-classified the OBCs nationally?
The Supreme Court permitted splitting a category of a few hundred Scheduled Castes on the finding of uneven benefit. The OBC category contains 3,743 castes.
If the principle is sound, the case here is far stronger. The reason it has not happened is that the dominant communities inside the OBC category are the ones with the political power to prevent it.
5. What happens to this argument if Census 2027 surprises everyone?
Neither side is preparing publicly for either result, which suggests neither is confident. A movement demanding immediate abolition eleven months before the first caste count in ninety-five years is, at minimum, oddly timed.
6. Why is nobody discussing 543 to 850?
It is the largest expansion of reserved political representation in Indian history, it is before Parliament, and it is absent from the argument.
The most plausible explanation is that legislative seats are not what young people are anxious about. Which is understandable, and does not make it less consequential.
7. Why is measured discrimination highest where reservation does not exist?
Ten, six, three — in the private sector. Both sides find this awkward and neither raises it.
8. If reservation is stigmatising, why do communities fight to get in?
One standard argument against reservation is that it marks its beneficiaries as needing help. Marathas, Jats, Patidars, Kapus and Gujjars have spent a decade demanding that mark.
The stigma argument is worth taking seriously — Part 12 does. But it has to be reconciled with the observed behaviour of very large numbers of people.
9. Has anything actually been abolished in this period?
Yes, once — and Part 8 has it. The Anglo-Indian nomination lapsed in January 2020, quietly, by simple non-extension.
The mechanism for ending a constitutional reservation exists, has been used within living memory, and requires no movement, no amendment fight and no protest. It requires only that the affected group be small enough to have no votes.
12What The Evidence Can And Cannot Settle
The honest scorecard for Part 10.
| Question | Status | Best current answer |
|---|---|---|
| Has the 50% ceiling been breached? | Yes | 59.5% centrally; ~62% in Maharashtra |
| Is caste-blind reservation constitutional? | Yes, since 2022 | Janhit Abhiyan, upheld 3–2 |
| Has India tried an economic quota? | Yes | EWS, running since 2019 |
| Is the EWS threshold evidence-based? | No | Copied from the OBC creamy layer figure |
| Do dominant communities want in? | Yes | Marathas, Jats, Patidars, Kapus, Gujjars |
| Has the SC category been split? | Yes | Permitted 2024; Telangana 2025, Karnataka 2026 |
| Has the OBC category been split nationally? | No | 3,743 castes, one undivided quota |
| Is India counting caste again? | Yes | Census 2027; caste data in Phase 2, Feb 2027 |
| Is reserved representation expanding? | Proposed | Lok Sabha 543 → 850, before Parliament |
| Does reservation reach the private sector? | No | And that is where the measured gap is widest |
The three things nobody can tell you
- What the caste census will show. Eight months away, and it decides the terms of everything.
- Whether the Maharashtra Act will survive. Before a special bench of the Bombay High Court, with appointments and admissions proceeding subject to the outcome.
- Whether delimitation will happen on the proposed terms. The bills are introduced, not passed, and the north–south politics of it are unresolved.
Between 2019 and 2026 India broke its own reservation ceiling, built a caste-blind quota it could not justify the threshold of, split a category it had called indivisible, restarted a count it had abandoned in 1931, and put a near-doubling of its legislature before Parliament — and then a movement appeared demanding that something be done, because nothing ever changes.
13The Five-Line Nichod
Everything in this book, compressed. If you remember nothing else, remember these.
PART 10 — THE MODERN REWRITE
- The ceiling broke in two days. The 103rd Amendment passed the Lok Sabha on 8 January 2019 and the Rajya Sabha on 9 January, taking central reservation to 59.5%. The Supreme Court upheld EWS in 2022 by three votes to two — one vote from a completely different argument today.
- India has been running the economic quota for seven years, and it does not work as advertised. The ₹8 lakh threshold was copied from the OBC creamy layer with no study behind it, and covers the overwhelming majority of eligible households. Removing caste did not remove categories, certificates or arbitrariness — it relocated them.
- The most powerful farming communities in India are demanding to be called backward. Marathas (Maharashtra now at ~62%), Jats, Patidars, Kapus, Gujjars. And the fiercest resistance comes from existing OBC groups, not from upper castes — the conflict is horizontal, exactly as Part 3 predicted.
- The Scheduled Caste category has been split. Davinder Singh, 1 August 2024, 6–1, overruling Chinnaiah. Telangana implemented it in 2025; Karnataka approved a 5.25 : 5.25 : 4.5 formula in 2026. The institutional argument has moved from whether to who inside — and nobody has done the same for 3,743 OBC castes.
- And the biggest change is the one nobody is discussing. Census 2027 began counting caste in April 2026, and in the same month Parliament took up a proposal to raise the Lok Sabha from 543 to 850 seats. Reserved seats are proportional. The largest expansion of political reservation in Indian history is before Parliament while the country argues about medical college cut-offs.
Coming next
Every popular statistic from both sides, run through the same test. What "merit" actually measures, and what coaching money does to it. Dropout and performance data at the IITs and AIIMS — the real figures against the forwarded ones. Faculty and senior-post vacancies, and who is missing at the top. The composition of Group A services, the judiciary and the newsroom. Land ownership, wealth share and business ownership by caste. Mobility studies from 1950 to 2026, and how much of the change reservation can actually claim credit for.
Every chart sourced. Every chart also given its strongest criticism. And the finding both sides will dislike.
A closing thought: the gap between where the attention is and where the power is may be the most useful thing in this book. A movement with 5.8 million followers is arguing about exam cut-offs. Parliament is deciding the size and composition of the legislature. Only one of those is a decision. — L.S.
Sources & further reading — Part 10
Every factual claim in Part 10, and where it comes from.
EWS
Constitution (103rd Amendment) Act, 2019, inserting Articles 15(6) and 16(6); passed by the Lok Sabha on 8 January 2019 and the Rajya Sabha on 9 January 2019. Eligibility criteria as notified: family income below ₹8 lakh per annum, agricultural land below 5 acres, residential flat below 1,000 square feet, residential plot below 100 square yards in a notified municipality and below 200 square yards elsewhere. Central reservation total of 59.5%. · Janhit Abhiyan v. Union of India (2022), five-judge bench, EWS upheld 3–2; dissents by Chief Justice U. U. Lalit and Justice S. Ravindra Bhat on the exclusion of SCs, STs and OBCs and on the breach of the 50% ceiling; separate observations in the majority on review and on reservation not becoming a vested interest. · On the derivation of the ₹8 lakh figure, the government's position before the Court that it mirrored the OBC creamy layer limit, and the committee subsequently constituted which recommended retaining it. · Criticism of the threshold's breadth, and of seat expansion in higher education of the order of 25% without corresponding faculty recruitment, is drawn from published commentary and analysis.
Demands for inclusion
Ram Singh v. Union of India (2015) — the Supreme Court quashed the inclusion of Jat communities in the central OBC list. · Patidar agitation in Gujarat from 2015; Kapu demands in Andhra Pradesh; Gujjar demands in Rajasthan. · Maharashtra State Reservation for Socially and Educationally Backward Classes (SEBC) Act, 2024, passed 20 February 2024, providing 10% reservation for Marathas on the basis of the commission headed by Justice Sunil Shukre; total state reservation of approximately 62%; the Act under challenge before a specially constituted bench of the Bombay High Court, with admissions and appointments permitted subject to the outcome. Manoj Jarange-Patil's demand for recognition of Marathas as Kunbi and its rejection by the state government; opposition from OBC organisations.
State reservation totals
Tamil Nadu's 69% reservation, placed in the Ninth Schedule in 1994. · Bihar: amendments raising reservation from 50% to 65% following the state's caste survey, struck down by the Patna High Court on 20 June 2024 on the ground that the 50% ceiling had been breached; the state appealed and on 29 July 2024 the Supreme Court declined to stay the High Court's decision. · Chhattisgarh: the High Court set aside the state's decision to raise the quota to 58%, holding reservation above 50% unconstitutional. · Karnataka: legislative resolution urging the Union government to place the state's 56% reservation in the Ninth Schedule. · Maharashtra as above.
Sub-classification
State of Punjab v. Davinder Singh, 1 August 2024, seven-judge bench, 6–1, upholding sub-classification within the Scheduled Castes and overruling E. V. Chinnaiah v. State of Andhra Pradesh (2004). Justice Gavai's holding that states must evolve a policy to identify a creamy layer among SCs and STs, with criteria differing from those used for OBCs. · Telangana Scheduled Castes (Rationalisation of Reservations) Act, 2025 — the first state to operationalise sub-categorisation, dividing the Scheduled Castes into three groups within the existing 15% quota. · Karnataka: commission headed by Justice Nagmohan Das; Cabinet approval in 2026 of an internal formula of approximately 5.25% : 5.25% : 4.5% within the 15% quota.
Census 2027 and delimitation
Census of India 2027, conducted in two phases: Houselisting and Housing Census from 1 April to 30 September 2026 across all 36 states and union territories, with self-enumeration available 1–15 April and house listing from 16 April; Population Enumeration in February 2027, in which caste data is collected. Reference date of 1 October 2026 for snow-bound regions including Ladakh, Jammu and Kashmir, Himachal Pradesh and Uttarakhand. India's first digital census, conducted by mobile application with a self-enumeration option; more than 5.72 lakh households had used self-enumeration by 10 April 2026. First comprehensive caste enumeration since 1931. · Special session of Parliament, 16–18 April 2026: the Constitution (131st Amendment) Bill, 2026, the Delimitation Bill, 2026, and the Union Territories Laws (Amendment) Bill, 2026; the proposal to increase Lok Sabha strength from 543 to 850 seats; commentary noting that the exercise is unlikely to be completed before the 2029 general election and the consequence for women's reservation under the 106th Amendment of 2023.
Where reservation does not apply
No caste reservation exists in private-sector employment in India. · Higher judicial appointments are made through the collegium system and reservation does not apply. · Surveys of Indian media leadership composition have repeatedly found senior editorial positions concentrated among a small number of upper-caste communities. · The correspondence study result (ten, six, three) is from Thorat, Attewell and Rizvi, Economic and Political Weekly, 2007, discussed in full in Part 6.
A note on what is missing — and on corrections
Part 10 describes developments up to July 2026. Several matters are live: the Maharashtra SEBC Act is under challenge; the delimitation bills are introduced but not enacted; the Balakrishnan Commission has not reported; and Census 2027's caste enumeration has not yet been conducted. Figures given for state reservation totals are approximate and vary with how horizontal quotas are counted.
Part 10 makes no claim about whether reservation works, which is Part 11. If you find an error — and in a chapter about events still in motion that is more likely than usual — I want to know.