How To Read This Book
This book has one rule, and the whole thing depends on it.
Facts and opinions will never be mixed together.
Almost everything you have read about reservation — on Instagram, on WhatsApp, on news TV — mixes three completely different things and presents all of them as "the truth." A hard fact, a reasonable opinion about that fact, and a made-up claim all get said in the same confident voice.
So in this book they are colour-coded. Every time something important comes up, you will see one of these three boxes.
A number, a date, a law, a court ruling, a line from a document. You can go and verify it yourself. Both sides of the argument accept it — or they would have to stop arguing if they read it.
This is where most of the real argument lives. Two intelligent, well-meaning people look at exactly the same evidence and reach opposite conclusions. Neither is lying. When you see this box, both readings get written out properly.
This box appears for both sides. If a popular claim cannot be backed up, it gets this box, no matter who is making it or how much I might personally like it.
The other four promises
- Simple English, always. No difficult words without an explanation. No showing off. If a fifteen-year-old who hates reading cannot follow a page, that page is badly written and it is my fault, not yours.
- Both sides at full strength. Every argument in this series is written the way its smartest supporter would write it — not the stupid version that is easy to knock down. This is called steelmanning. It is the opposite of what social media does.
- My opinion will be labelled. When I give a verdict, it will sit in a box that says it is my verdict. You are free to throw it away and keep the facts.
- Sources on the page. Every number is traceable. Nothing here rests on "someone said."
If you came here hoping this book will confirm what you already believe, you are going to be uncomfortable at least four times. That is the point. A book that only tells you that you were right was not worth writing.
Why I Started Writing This
In the last week of July 2026, I was doing what most of us were doing: scrolling. And in seven days I watched an Instagram account go from nothing to nearly six million followers. Six million. That is more people than live in Singapore.
The account was called Reservation Hatao Andolan. The comment sections underneath its videos were some of the angriest writing I have ever read from Indians about other Indians. And what struck me was not the anger. Anger about jobs and exams in this country is completely understandable — I have felt it myself.
What struck me was that almost nobody arguing seemed to know what they were arguing about.
People were furious about a system whose actual rules they had never read. They were quoting history they had never checked. They were repeating numbers that were wrong — and I mean wrong on both sides. And every single person, on every side, was completely certain.
I include myself in this. When I started reading properly for this book, I found out that several things I had believed my whole life were simply not true. Some of them were things I had believed because they made my side look good.
So I decided to go back to the absolute beginning. Not to 1990. Not to 1950. To the beginning — to the oldest texts, the archaeology, the genetics, the court records, the census files, and to what the great minds on every side actually wrote rather than what people claim they wrote. And then to come forward, slowly, until we reach the argument happening on our phones right now.
This is Part 1 of fourteen. It does not settle anything. Its job is smaller and more important: to make sure that by the end of it, you actually know what the thing is. What reservation literally does. What the real numbers are. What the two sides are truly claiming, at their strongest. Everything after this is built on this foundation.
One last thing. I have tried very hard to be fair to people I disagree with. If you finish this book and cannot tell which side I am on until the chapter where I tell you, I will have done my job.
— Lovepreet Singh
July 2026
1Five Point Eight Million People in Seven Days
A young man stood up on live television and asked a question. Within a week, six million people had joined a movement.
Picture a television studio. It is July 2026. India has spent two months in the middle of the largest youth protest in a generation. Exam papers have leaked. Students have died. A movement with a strange name — the Cockroach Janta Party — has filled Jantar Mantar in Delhi and streets in dozens of cities.
On an NDTV broadcast, a NEET aspirant named Harsh Dubey stops being a polite guest. He stands up, points at the anchor, and shouts a question at the whole room:
He was, in one narrow sense, completely right. For years, mainstream Indian politics has treated reservation as a subject you do not open. Every major party supports it in public. Nobody wants the fight.
That clip went everywhere.
What happened next was faster than anything in Indian internet history
On 22 July 2026, an Instagram account appeared: @reservationhataomovement. Its bio read: "backed by citizens discarded by reservation politics, driven by merit."
It used the Harsh Dubey clip as its founding video.
Outlook India reported 5.6 million followers and 122 posts on 28 July 2026. The Quint independently reported 5.8 million as of 5 p.m. on 29 July 2026, and confirmed the account was created on 22 July. Whatever you think of the movement, this is one of the fastest civic mobilisations India has seen.
Sources: Outlook India, 28 July 2026 · The Quint, 29 July 2026
It did not call for a march. It called for a flood.
This is the part that makes 2026 different from 1990. In 1990, when India last exploded over reservation, students blocked trains and burned buses. In 2026, the instruction was different. An RHA video told followers:
No permission needed. No police lathi charge to film. No leaders to arrest. A protest that lives entirely inside the recommendation algorithms of two or three American companies.
And then the other side arrived
Within days, anti-caste creators, scholars, students and cartoonists were producing rebuttals that were getting hundreds of thousands of likes of their own. One post calling the campaign "politically, historically, and statistically illiterate" crossed 52,000 likes. An illustration against manual scavenging crossed 75,000.
One line from that pile-on is worth putting in a box, because it is the sharpest single sentence produced by either side in the whole week:
The uncomfortable observation
I spent several days reading those comment sections. Thousands of them. Here is what I found, and I want to say it plainly because it applies to almost everyone, on every side.
Most people arguing did not know the rules of the thing they were arguing about.
People were confidently wrong about what percentage is reserved. About whether a reserved-category topper takes a general seat. About what the Constitution says. About what a "cut-off" even means. About whether reservation had a ten-year expiry date. About what the creamy layer is.
These are not obscure details. These are the basic operating rules. And they were being got wrong by people writing paragraphs of certainty at three in the morning.
Reading one: This is real, spontaneous frustration finally finding a voice. Government jobs are shrinking, exams are being leaked, unemployment among graduates is severe, and a generation that did everything right feels cheated. The CJP protests proved to young Indians that online anger can actually force a minister out. RHA was the second wave of the same energy.
Reading two: This is opportunism. A movement about paper leaks — a failure of the government's own exam agency — was successfully redirected within 48 hours into a fight between students and other students. The people who ran the leaky exams disappear from the story entirely. The anger stays, but the target changes.
Both readings can be true at once. Genuine frustration and convenient redirection are not opposites. That is exactly how political energy usually gets used.
Before we can judge any of this, we need to know what reservation actually is. Not what it feels like. What it is. That is the next three chapters, and they are the least exciting and most important pages in this book.
2Reservation in One Page
Forget everything you have heard. Here is the plainest possible description of what the system does.
Reservation is a rule that says: in government jobs, in government-funded colleges, and in elected seats, a fixed share of the places must go to people from listed groups.
That is it. That is the whole mechanism.
Three things follow from that one sentence, and each one matters:
- It only applies to government things. Government jobs, public universities, government-funded college seats, seats in Parliament and state assemblies. It does not apply to private companies. If you work at Infosys, TCS, Reliance, a startup, a shop, or your uncle's business, reservation has never touched your hiring. More on this in a moment, because it is enormous.
- The groups are listed, not self-declared. You cannot simply say you are Scheduled Caste. There are official lists — the Scheduled Castes list, the Scheduled Tribes list, the Other Backward Classes list — maintained by the central government and by each state. Being on the list is a legal status with paperwork.
- It is a share, not a gift. Nobody is handed a job. You still have to appear, qualify, and rank. The reservation decides which pool of candidates you are ranked against.
There is no caste reservation in private-sector hiring anywhere in India. 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 is about a small and shrinking slice of the total job market. Hold on to this. It comes back in Part 13 and it changes how the whole fight looks.
Now: the two analogies
Here is where honest writing gets hard. Every explanation of reservation smuggles in an argument through its choice of metaphor. So instead of picking one, I am going to give you both — the one each side actually uses — and let you see them side by side.
Imagine a running race. For a thousand years, one group of runners was allowed to train, wear shoes, and eat properly. Another group was legally forbidden from even entering the track, and was punished for trying.
Then one day someone fires a starting pistol and announces: from now on, everyone competes equally, may the best runner win.
That is not fairness. That is freezing an unfair result in place and calling it a fair race. Reservation is the attempt to correct a starting line that was rigged for fifty generations.
Imagine you are being wheeled into surgery. You do not ask the surgeon about their grandfather's suffering. You ask whether they are the best surgeon available.
Some jobs — doctor, engineer, pilot, judge — affect other people's lives. Filling them by any rule other than pure ability imposes the cost of historical repair onto random third parties, including poor people who never oppressed anyone.
Justice for the past should not be paid for with the future.
The race analogy assumes the starting line is still unequal today — that the disadvantage did not end in 1950. If that is false, the analogy collapses. Part 6 and Part 11 test exactly this.
The hospital analogy assumes that the exam score measures the ability that matters, and that it measures it independently of your background. If a score largely reflects who could afford eight lakh rupees of coaching, then "the best surgeon" and "the highest scorer" are not the same person. Part 11 tests exactly this.
Neither analogy is dishonest. Both are doing the argument's work before the argument starts. That is what analogies do — which is why you should never let one settle anything.
The three words you need
You will meet these constantly. Learn them once now and the rest of the series is easy.
| Word | What it actually means |
|---|---|
| SC | Scheduled Castes. The communities historically treated as "untouchable" — kept outside villages, barred from temples, wells and schools, and assigned the work nobody else would do. Officially listed in a Presidential Order. Roughly 16.6% of India's population at the 2011 Census. |
| ST | Scheduled Tribes. India's Adivasi or indigenous communities — historically living in forests and hills, largely outside the caste system rather than at the bottom of it, and displaced by mining, dams and forestry. Roughly 8.6% at the 2011 Census. |
| OBC | Other Backward Classes. The huge middle band — communities that were not treated as untouchable, but were socially and educationally held down. Farmers, artisans, herders, weavers, potters, fishermen and many more. The single largest bloc, but we do not have a modern official count. That is why the 2027 Census matters so much. |
| EWS | Economically Weaker Sections. The newest category, created in 2019. This one is not about caste — it is for poor families among the castes that get no other reservation. In effect, a poverty quota for the general category. |
You will see this number everywhere. It comes from the Mandal Commission's 1980 report, which arrived at it by taking the last caste census — 1931 — and doing arithmetic on it. It is an estimate built on 95-year-old data. It may be roughly right. It may be badly wrong. Nobody actually knows, and anyone who tells you they know is guessing. India has not counted caste in a full census since 1931. The 2027 Census will be the first that does. Part 10 covers this.
3The Actual Numbers
Who gets how much, why the total is not what you think, and the 50% wall that cracked in 2019.
Almost every argument about reservation contains a number, and a surprising share of those numbers are wrong. So here are the real ones, for central government jobs and central educational institutions.
| Category | Share | Notes |
|---|---|---|
| Scheduled Castes (SC) | 15% | In place since the 1950s |
| Scheduled Tribes (ST) | 7.5% | In place since the 1950s |
| Other Backward Classes (OBC) | 27% | Since 1990 in jobs, 2008 in central colleges. Non-creamy-layer only. |
| Economically Weaker Sections (EWS) | 10% | Added 2019. Not caste-based. |
| Total reserved | 59.5% | Leaving 40.5% open to everyone including all of the above |
Persons with Disabilities get a further quota — 4% in government jobs under the 2016 Act, 5% in higher-education admissions — but it works differently, and that difference is the subject of the next section.
Vertical and horizontal: the distinction that ends a hundred arguments
Here is something almost nobody outside the system knows, and it clears up an enormous amount of confusion.
There are two kinds of reservation, and they stack differently.
SC, ST, OBC and EWS are vertical. Think of them as separate columns of seats. If there are 100 posts, 15 are in the SC column, 7.5 in the ST column, 27 in the OBC column, 10 in the EWS column, and 40.5 are open.
You can only stand in one vertical column.
Disability, ex-servicemen status, and in many states women, are horizontal. These do not get their own column. They cut across every column.
So a 4% disability quota means 4% of the open seats, 4% of the SC seats, 4% of the OBC seats, and so on. A disabled general-category candidate competes for the disabled slice of the general column.
This is why people who add up "15 + 7.5 + 27 + 10 + 4 + 3 + 33…" and announce that 90% of India is reserved are making a basic arithmetic error. Horizontal quotas are carved out of the existing columns, not added on top of them. The vertical total in central institutions is 59.5%.
Your state is almost certainly different
The numbers above are for central government jobs and central institutions — the UPSC, the IITs, the AIIMS, central universities. But most Indian government jobs and most college seats are state jobs and state seats. And states set their own percentages.
| State | Roughly | The situation |
|---|---|---|
| Tamil Nadu | 69% | The highest in India. Protected from court challenge by being placed in the Ninth Schedule of the Constitution in 1994. Has run at this level for three decades. |
| Most large states | 50% | The ceiling set by the Supreme Court in 1992, plus 10% EWS on top since 2019. |
| Several states | Above 50% | Various states have crossed the line for specific communities. Several such laws have been struck down; others are pending in court. |
The "keep it" side says: Tamil Nadu has run 69% reservation for thirty years and is among India's best-performing large states on health, education, industrial output and human development. If heavy reservation destroyed institutions, Tamil Nadu should be a ruin. It is not.
The "remove it" side says: Tamil Nadu was already ahead of most of India before 1970 for reasons that have nothing to do with quotas — early social reform, higher literacy, better public health, coastal trade. You cannot credit the quota for a lead that predates it, and the state's own general-category families now emigrate at high rates.
Both arguments are real. Neither has been settled. Part 11 goes at this properly with the data.
The 50% wall, and how it broke
For thirty years, Indian reservation had one hard rule that everybody knew: you cannot go above 50%.
The 10% EWS quota uses income and property tests, not caste. Which means India has already run the experiment that RHA is proposing, on 10% of seats, since 2019. Anyone arguing about whether an economic quota is possible should start by looking at how the existing one has actually worked. Part 10 does.
Why did the same political system that will not touch caste quotas find it easy to break a thirty-year constitutional ceiling within weeks, in order to create a quota for the general category? Whatever your view, the speed is worth thinking about.
4How It Actually Works
This is the chapter nobody reads. It is also the reason most of this fight is confused. Please do not skip it.
Everything so far has been the map. This is the machine. If you read only one chapter of this book, read this one — because at least four of the most common claims on both sides fall apart the moment you understand how the mechanism actually operates.
Rule 1: There are two lists, and the good candidates move up
Here is the single most misunderstood rule in Indian public recruitment.
If a reserved-category candidate scores high enough to make the general cut-off on their own, without using any relaxation, they are counted in the general list — not the reserved list. Their reserved seat then goes to the next reserved candidate.
In official language they are called a Meritorious Reserved Category candidate, or selected "on own merit."
In Airports Authority of India & Ors v. Sham Krishna B & Ors (2026 INSC 69), the Supreme Court held that where a reserved-category candidate secures marks higher than the general cut-off, that candidate must be treated as qualified against an open or unreserved vacancy — and must be treated that way from the first stage of selection, not adjusted afterwards.
The rule exists precisely so that reservation adds representation instead of capping it.
What this means in practice: the sentence "he took a general seat with lower marks" describes something that, by rule, cannot happen. A general seat can only be taken by someone who cleared the general bar. What actually happens is the reverse — a reserved-category topper leaves the reserved pool and frees a seat inside it.
Rule 2: A qualifying score is not an admission score
This confusion, on its own, probably powers half the anger online.
In an exam like NEET there are two completely different numbers, and people constantly quote the first while meaning the second.
| Number | What it is |
|---|---|
| Qualifying cut-off | The minimum score to be allowed into the counselling process at all. It is a percentile floor, not a seat. In NEET 2025 this was the 50th percentile for General/EWS (686–144 marks) and the 40th percentile for OBC/SC/ST (143–113 marks). Clearing it wins you nothing except the right to apply. |
| Admission cut-off | The score at which seats actually ran out. This is the number that decides who becomes a doctor. It is far, far higher — and for the top categories it is nearly identical across groups. |
Look at what the second number actually was.
| Category | Closing marks | Closing All-India Rank |
|---|---|---|
| General | 660 | AIR 19,603 |
| OBC | 658 | AIR 20,281 |
| SC | 575 | AIR 1,05,676 |
By the final stray round the same seats closed at roughly 652 (General), 652 (OBC) and 549 (SC).
Two. Out of 720. In the final stray round it was zero — both closed at 652. The General–SC gap was real and substantial: 85 marks in Round 1. But the picture of vast, casual gaps across all reserved categories does not survive contact with the actual counselling data.
This claim, made in an RHA video and repeated thousands of times, is the emotional engine of the whole campaign. I tried to verify it and could not.
What the record shows: the lowest qualifying floor for any reserved category in NEET 2025 was 113 marks, not a negative number. And a qualifying floor is not a seat. In the all-India government MBBS pool, the very last SC seat in the final stray round went at 549 marks.
Extremely low scores can pick up seats in some private and deemed colleges in late stray rounds — but that happens across categories, including General, and it happens because those seats cost sums most families cannot pay. That is a story about money, not about caste. Marked contested until someone produces the actual allotment record.
Rule 3: The roster — how 27% becomes an actual list of posts
You cannot give 27% of one job to someone. So government departments use a 100-point roster: a numbered list from 1 to 100 where each position is pre-labelled by category, and vacancies are filled by walking down the list in order.
Two consequences follow, and each one annoys a different side:
- The percentages are not per-round, they are per-roster. In a small recruitment of four posts, you can easily get an outcome that looks nothing like 15/7.5/27. People screenshot such rounds as proof of a conspiracy in both directions. Usually it is just where the roster happened to be standing.
- Unfilled reserved posts create backlog. If no eligible candidate is found for a reserved point, that post is carried forward as a "backlog vacancy." India has accumulated large backlogs over decades. Which means the reserved share on paper and the reserved share actually working in the building are two different numbers. Part 11 has the data. It surprises people on both sides.
Rule 4: The creamy layer — the filter most people forget exists
OBC reservation does not go to all OBCs. Families above an income and status line are excluded. This is the creamy layer.
The limit was raised from ₹6 lakh to ₹8 lakh by an order of the Department of Personnel and Training dated 13 September 2017. As of 2026 no further revision has been notified, despite repeated demands from state governments and the National Commission for Backward Classes to raise it to ₹12 or ₹15 lakh for inflation. Children of senior government officers and certain professional and constitutional post-holders are excluded regardless of income.
And the EWS quota has its own filter: family income below ₹8 lakh a year, plus limits on agricultural land, residential flat size and plot size.
Reading one: ₹8 lakh a year puts a household in roughly the top few per cent of Indian incomes. Calling such a family "economically weaker" is absurd, and it shows EWS was designed to be wide enough to cover almost every general-category family that wanted it.
Reading two: the identical ₹8 lakh line applies to the OBC creamy layer. If ₹8 lakh is too generous for EWS, it is equally too generous for OBC — and the side that attacks one number while defending the other is not arguing about the number.
This is a good example of a fact that embarrasses whoever picked it up first.
Reserved toppers move into the general list. Qualifying floors are not admission scores. Percentages are delivered through a roster, not per round. And large filters — creamy layer, income tests, property tests — sit on top of all of it. Almost none of this is understood by the people arguing hardest.
5Six Things The "Remove It" Side Gets Wrong
Stated fairly, then tested. The next chapter does exactly the same to the other side. Read both or neither.
The central figure is 59.5% reserved and 40.5% unreserved. But "unreserved" does not mean "reserved for general category" — it means open to everyone, including SC, ST, OBC and EWS candidates who make the mark. There is no such thing as a general-only seat in Indian law.
The bigger error is adding horizontal quotas — disability, ex-servicemen, women — on top of the vertical ones to reach frightening totals like 80% or 90%. Horizontal quotas are carved out of the existing columns. They add nothing to the total.
VERDICT: The arithmetic is wrong. The underlying frustration about seat scarcity is not — but the cause is the number of seats, not the split.By rule this is backwards. A reserved candidate who clears the general bar is moved into the general list and vacates their reserved seat — the "own merit" rule, restated by the Supreme Court in 2026. Nobody enters a general seat below the general bar.
And the real gaps are smaller than the folklore. General and OBC closed two marks apart for all-India government MBBS seats in 2025.
VERDICT: False as stated. The SC/ST gap is real and worth arguing about honestly — but it is a different argument from this one.This is the most repeated line in the entire debate, and it comes from a genuine misreading of the Constitution.
Article 334 did put a ten-year clock on one specific thing: reserved seats in the Lok Sabha and state legislatures — political seats. That clock has been extended by Parliament roughly every decade since.
Articles 15(4) and 16(4), which enable reservation in education and government jobs, carry no expiry date at all. They never did.
VERDICT: A real constitutional provision, applied to the wrong subject. Part 8 goes through the Constituent Assembly record line by line.This is RHA's central promise: "Remove Reservation. Remove Casteism."
The problem is chronology. Caste is at minimum many centuries old; caste-based reservation in independent India dates from 1950. Whatever caused the first cannot have been the second. Removing a 76-year-old policy cannot remove a phenomenon that predates it by a very long way.
A narrower version of the claim survives and deserves respect: state-issued caste certificates keep caste administratively alive and give people a reason to hold on to the identity. That is arguable, and it is argued properly in Part 12.
VERDICT: The strong version fails on chronology. The narrow version is a serious argument and gets a full hearing later.Everyone sits the same exam, clears a qualifying floor, passes the same university examinations, and — for doctors — obtains the same registration. Reservation affects entry, not graduation and not licensing.
The honest version of the concern is about entry-score gaps and whether they persist to the end of the course. That is a measurable question with actual published evidence, and Part 11 puts the numbers on the table. Some of it will not please the side making this claim; some of it will not please the other side either.
VERDICT: As stated, unsupported. As a question about outcomes, legitimate — and answerable with data rather than adjectives.This is the most emotionally powerful claim and the one I have the most sympathy for, because the underlying pain is completely real.
But look at the scale. Government employment is a small fraction of India's total workforce and it has been shrinking for years, while lakhs of posts sit vacant. When 25 lakh people apply for a few thousand posts, the binding constraint is that there are a few thousand posts. Redistributing them differently changes who is unemployed. It does not change how many.
VERDICT: Misdirected. The anger is justified; the address on the envelope is wrong. Part 13 argues that this may be the single most important fact in the entire debate.6Six Things The "Keep It" Side Gets Wrong
Same treatment. Same standard. If the last chapter felt fair to you, this one is going to sting — and that is how you know the last one was honest.
It is not, and defenders who say this are handing away their own best argument.
Reservation was designed as a representation measure — to break a monopoly on public power held by a small number of communities. Poverty relief is what scholarships, subsidies and welfare schemes are for.
Why the mistake is costly: once you accept "it is for the poor," you have already conceded that income is the right test, which is exactly the opposition's position. Then the only remaining question is why caste appears in it at all.
VERDICT: Wrong on its own terms, and strategically self-defeating.A great deal has changed and it is measurable — in literacy, in urban occupations, in political representation, in inter-caste contact, in who now sits in offices that were once completely closed.
Insisting nothing has changed does two bad things at once. It erases the achievement of the people who actually fought and won these gains. And it makes the argument unfalsifiable: if no amount of progress ever counts, the policy can never be reviewed, and a policy that can never be reviewed will eventually lose public consent.
VERDICT: Overstated. The defensible claim is that change has been real but deeply uneven — which is both true and much stronger.Some opponents of reservation are casteist. When a movement's own prominent voices post about "Brahmin genes," or dismiss documented historical atrocities as invented propaganda, that is not a debating position and it should be named.
But a 22-year-old who has failed an exam three times and is angry is not automatically a bigot. Treating every such person as an enemy is how you lose an argument you could have won — and it hands the other side its most effective recruitment line: "they will not even talk to you."
VERDICT: True of some opponents, false as a general rule, and self-harming as a debating tactic.The Supreme Court of India disagrees. In 2024, a seven-judge bench ruled 6–1 that states may sub-classify Scheduled Castes precisely because the benefits had been captured unevenly — some SC sub-castes had advanced substantially while others had barely moved.
That verdict is, in effect, the highest court in the country accepting that elite capture within reserved categories is real. You cannot cite the Court when it suits you and call the same finding propaganda when it does not.
VERDICT: Contradicted by a 2024 Constitution Bench ruling. This is now a fact to be managed, not a claim to be denied.They are not. Backlog vacancies in reserved posts have accumulated for decades. Reserved faculty positions in central universities and premier institutes have been notoriously under-filled. And more than 13,500 SC, ST and OBC students dropped out of central universities, IITs and IIMs according to data reported in 2023.
Here is the thing though: this fact is bad for the other side too. If reserved seats sit empty, then the seats are not being taken away from anybody — and the "they took my seat" story loses its object.
VERDICT: False — and inconvenient for everyone. One of the most useful facts in this book."General" is not a class. It is everything left over after four lists. It contains genuinely wealthy families and it contains families in serious poverty — landless Brahmin priests, small Bania shopkeepers ruined by a bad year, Rajput families with a name and no money.
The strongest evidence that these families exist is that Parliament created a 10% EWS quota specifically for them in 2019, and the Supreme Court upheld it.
The serious reply — and it is serious — is that being poor is not the same as being excluded, and reservation was built to address exclusion. That is a real argument. But it has to be made, not assumed.
VERDICT: The label is being used as if it means something it does not. Poverty and exclusion overlap; they are not identical; and the difference is the whole debate.If exactly one of the last two chapters made you angry, that is worth sitting with for a minute. Both were written to the same standard, with the same tone, using the same kind of evidence.
7Ninety Seconds of History
The whole story on one timeline. Every line here becomes a full chapter later in the series.
8What Actually Happened in 2026
Cockroaches, a leaked paper, a resignation, and an explosion. Seventy-four days, in order.
To understand why the reservation argument detonated in the last week of July 2026, you have to understand what came immediately before it — because the second movement was built directly on the wreckage of the first one's victory.
A movement about leaked exam papers — an administrative failure by the state's own examination machinery — won. The minister resigned. And within twenty-four hours of that victory, the energy transferred to a movement about reservation.
Reading one: natural progression. Students who fought about one unfairness in the exam system turned to what they see as the bigger unfairness in it. The first win proved that online pressure works.
Reading two: a redirection, whether deliberate or accidental. On 24 July the conflict was students versus the state. By 27 July it was students versus other students. The examination agency, the leaks, the vacant posts and the shortage of seats vanish from the conversation entirely. If you were responsible for the exam system, you could not have designed a better outcome.
I do not think anyone needs to have planned reading two for it to be what happened. Anger looks for a target, and a target that is standing next to you is easier to reach than one in a ministry.
9What RHA Actually Wants
Their demands, in their words — then their case put at its strongest, the way their best advocate would put it.
Before criticising a position it is worth stating it accurately. Here is what the Reservation Hatao Andolan has actually published.
The four demands
| Demand | What it means in practice |
|---|---|
| Merit over Caste | Selection to public education and public employment on examination performance alone, with no caste category |
| Equal Fees | An end to category-based fee concessions in public institutions |
| Equal Age Limit | An end to category-based upper age relaxations in competitive examinations |
| Equal Cut-off | One qualifying and admission standard for all candidates |
The bigger proposal
Beyond the four demands, RHA has published a broader policy titled "Remove Reservation. Remove Casteism." Its long-term goal is stated as "One Nation. One Identity. Indian."
The proposal argues that if the aim is to end caste discrimination, the state should stop requiring or recording caste at all — no caste column on government forms, admissions or job applications — and has called for the public "showcasing" of caste to be made illegal.
The campaign describes itself as an unorganised, youth-led coalition with no political affiliation, stating that it had hoped a change of government would reform education and that this did not happen: "Our focus is on the policy, not the party."
Its slogan on X is "Educate — Agitate — Organize," alongside the Hindi line "Sari jaati ek samaan, mera bas yeh desh mahaan."
It is one of the most famous lines in Indian anti-caste history, and Ambedkar is the principal architect of the constitutional provisions that make reservation possible. A campaign to abolish reservation has adopted the slogan of the man who built it. Critics have called this a deliberate co-option; supporters say the words belong to all Indians. Either way, the fact is not in dispute.
Who is driving it
Two figures have been prominent in RHA's content: Ajeet Bharti, a right-leaning political commentator, and Anuradha Tiwari, who drew wide attention in 2024 for posting a photograph captioned "Brahmin genes."
In RHA videos, Bharti argues that caste quotas create a "creamy layer trap" and that state resources should be redirected into primary public education instead. He has also dismissed historical caste atrocities as "propaganda pushed by leftist historians to induce guilt."
This is the single biggest factual claim being made anywhere in this argument, because if it is true, almost everything else collapses — and if it is false, the movement's foundations shift.
It is also exactly the question you asked me to investigate, and it is too important to answer in a paragraph. Part 6 of this series is devoted entirely to it — court records, land records, temple records, colonial documents, missionary accounts, travellers' accounts, modern survey measurement, and the strongest version of the "it is exaggerated" case, all in one place. No summary here. Just the evidence, later, in full.
A policy adopted as a temporary corrective has now run for three generations, and no government has ever specified what success would look like or when it would end. A remedy without an exit condition is not a remedy; it is a permanent structure.
Worse, that structure gives every Indian family a standing financial reason to prove and preserve its caste. We ask people to register caste at birth, declare it at school, prove it at admission, carry it into employment — and then express surprise that caste has not faded. A state that keeps caste on file cannot credibly ask citizens to forget it.
Meanwhile the benefit has concentrated. The Supreme Court itself found in 2024 that some sub-castes captured the gains while others got nothing — which is why sub-classification became necessary. The child of an IAS officer and the child of a landless labourer can hold the same certificate.
And the cost lands on people who did nothing. A poor general-category student in Bihar inherits no land, no capital and no coaching money, only a surname — and pays for a history he did not write.
India already runs a caste-blind quota. EWS has existed since 2019 and the Court upheld it. So the question is not whether a caste-free system is imaginable. It is: why not extend the one we already have?
10What The Critics Actually Say
Same treatment. Their objections, then their case at full strength.
The backlash to RHA came from sociologists, educators, anti-caste scholars, student organisers and a very large number of digital creators. Their objections cluster into four arguments.
1. It confuses representation with poverty
Reservation was not built as an anti-poverty scheme. It was built to break a monopoly on public power. Replacing it with an income test answers a question nobody was asking, and abandons the one it was designed for. As critics put it: reservation exists to remedy centuries of denial of education, land, resources and dignity — not to top up a bank balance.
2. It reverses the order of operations
One widely shared post put it as sharply as it can be put: what came first must go first. Caste came first; reservation came in response. Removing the response while leaving the cause in place does not produce equality — it produces the pre-1950 arrangement with better branding.
3. It uses Ambedkar's words against Ambedkar's work
The adoption of "Educate, Agitate, Organise" by a campaign to dismantle constitutional safeguards has been called a cynical use of anti-caste vocabulary against anti-caste protections.
4. The evidence that caste still bites is sitting in the same comment sections
During the same week, a post told Abhijeet Dipke — the man who had just forced a Union minister to resign — to "go back to US and clean toilets," a remark widely read as targeting his Dalit identity.
Dipke had, by any measure, achieved extraordinary success and mobility. The abuse still reached for caste. Critics used this as evidence for a specific claim: that class mobility does not switch caste discrimination off — which is precisely the claim an income-based replacement depends on being false.
And the inconvenient numbers
More than 13,500 SC, ST and OBC students dropped out of central universities, IITs and IIMs, according to data reported in 2023. Critics use this to argue that getting in is not the same as being let in — that the system delivers admission without delivering belonging.
Start with what reservation is for. It is not charity and it is not welfare. It is a structural correction to a specific historical fact: for a very long time, a small set of communities held nearly all public authority, and others were excluded by force, law and custom from literacy itself. Wealth transfers do not fix a monopoly on institutions. Presence does.
Now test the replacement. An income-only system assumes money is the barrier. But money does not stop a landlord from blocking a well, a classmate from refusing to share a room, a landlord from refusing a flat, or a comment section from telling a successful Dalit man to go clean toilets. Caste is not a bank balance. It is a door that opens for some names and not others.
Then look at delivery. If reservation were the vast handout it is described as, reserved posts would not sit vacant for decades and thousands of SC, ST and OBC students would not be dropping out of the very institutions they fought into. A system that under-delivers is not evidence that the need has ended.
Finally, look at the timing. This movement did not appear when caste violence made the news. It appeared in the week a general-category exam grievance found a microphone. The demand is not "end caste." Not one viral post asked for that. The demand is "end the remedy" — and a movement that wants the cure gone but is silent about the disease has told you what it is actually about.
11The Five Real Questions
Strip away the shouting and this entire argument reduces to five questions. Everything else is decoration.
Almost nobody online is arguing about these. They are arguing about cut-offs and slogans. But every serious position on reservation is really an answer to these five, and the rest of this series is the work of answering them properly.
If caste discrimination was systematic, violent and enforced for centuries, one set of conclusions follows. If it was mild, occasional, or largely invented later, a completely different set follows. This is a question of evidence, not opinion — and it has an answer.
→ Parts 2, 3, 4, 5 and 6
Even if the history is as bad as claimed, a remedy is justified by present conditions, not past ones. So: in 2026, does your caste still change your life outcomes after you account for income and education? This is measurable.
→ Parts 6 and 11
A real disease does not prove that a specific treatment cures it. Has reservation delivered what it promised? For whom? At what cost? And where has it failed?
→ Part 11
"This is imperfect" is not an argument unless you can name a superior alternative and show what happened when it was tried. Income-based? Sub-classified? Time-limited? Nothing at all? Other countries have run these experiments and the results are on the record.
→ Parts 12 and 13
This is the question neither side wants. The abolition side rarely specifies what evidence would make them keep it. The retention side rarely specifies what evidence would make them stop. A policy with no stated finish line and no stated failure condition is not a policy. It is a permanent political settlement pretending to be one.
→ Part 14
12The Five-Line Nichod
Everything in this book, compressed. If you remember nothing else, remember these.
PART 1 — GROUND ZERO
- Reservation is a share of government seats, not a gift — 59.5% reserved at the centre, 40.5% open to everyone including reserved candidates, and none of it touches the private sector where almost all Indian jobs actually are.
- A reserved candidate who clears the general bar is counted in the general list, not the reserved one — settled law, restated by the Supreme Court in 2026. In 2025 the General and OBC all-India government MBBS seats closed two marks apart.
- The "only ten years" line is a real constitutional provision applied to the wrong subject. Article 334's clock was on reserved legislative seats. Articles 15(4) and 16(4) never carried an expiry date.
- Both sides are running on bad numbers. One side inflates the reserved total by adding horizontal quotas that do not add. The other denies elite capture that the Supreme Court itself accepted in 2024.
- The real fight is five questions deep — was the injury real, is it still happening, does the medicine work, is there something better, and who decides when it ends. Nobody wins this argument on Instagram. It is won in Parts 2 to 14.
Coming next
We go back further than any WhatsApp forward goes. The Indus Valley. The oldest layer of the Rig Veda. The Purusha Sukta — the four-varna hymn — printed in full, translated, and then read line by line the way each side reads it. The Aryan question and what the genetics does and does not settle. And the argument at the centre of everything: was varna originally a description of work that later hardened, or a system of birth from the start?
Both cases. At full strength. No thumb on the scale.
A closing thought before Part 2. Reading this book will not make you neutral, and it is not supposed to. It should make you harder to fool — including by people who agree with you. That is a lower bar than wisdom and a much more useful one. — L.S.
Sources & further reading — Part 1
Every factual claim in Part 1, and where it comes from.
The 2026 movements
Outlook India, "Reservation Hatao Andolan: Why Is It Drawing Both Support And Opposition?", 28 July 2026 — follower counts, the four demands, the "One Nation. One Identity. Indian." proposal, the E20 Janata Party. · The Quint, "'End Caste Discrimination First': Critics Call Out 'Reservation Hatao Andolan'", 29 July 2026 — account creation date of 22 July, the Harsh Dubey NDTV clip, the "flood the comment sections" instruction, Ajeet Bharti and Anuradha Tiwari, the criticism, the Dipke abuse incident, and the 13,500 dropout figure. · Al Jazeera, 24–26 July 2026 — CJP talks with government, Dharmendra Pradhan's resignation on 25 July, CJP's three demands. · TIME, 22 July 2026, and Wikipedia's Cockroach Janta Party entry — CJP founded 16 May 2026 by Abhijeet Dipke.
Numbers and mechanics
Central reservation percentages and the 59.5% total — NEET and JEE 2026 category criteria as published by the examination bodies and summarised across admissions guidance. · NEET 2025 qualifying cut-offs: General/EWS 686–144 at the 50th percentile, OBC/SC/ST 143–113 at the 40th percentile. · NEET 2025 All India Quota government MBBS closing scores and ranks, Round 1 and stray round. · OBC creamy layer: raised from ₹6 lakh to ₹8 lakh by DoP&T order dated 13 September 2017; no further revision notified as of 2026, per Rajya Sabha replies.
Law
Indra Sawhney v. Union of India (1992) — 50% ceiling and creamy layer. · Janhit Abhiyan v. Union of India (2022) — EWS upheld 3–2. · State of Punjab v. Davinder Singh (2024) — SC sub-classification permitted 6–1. · Airports Authority of India & Ors v. Sham Krishna B & Ors, 2026 INSC 69 — reserved-category candidates above the general cut-off must be treated against open vacancies from the first stage. · Constitution of India, Articles 15, 16, 46, 330, 332, 334, 335, 340, 341, 342. · Census 2027 caste enumeration — field work from April 2026; Supreme Court declined petitions against it as a policy matter.
A note on what is missing
Part 1 deliberately makes no claim about whether historical caste discrimination occurred, how severe it was, or whether reservation works. Those require primary sources and datasets that belong in Parts 6 and 11, where they get proper space. Anything in this book marked CONTESTED CLAIM stays contested until evidence is produced — including claims made by people I might agree with.
If you find an error in this book, I want to know. A series built on the argument that both sides are running on bad numbers has no business running on bad numbers itself. Corrections will be carried in later parts.