I taught UPSC CSE aspirants for six years, and I sat the exam five times myself. I cleared the prelims five times. I never became an officer, though I learned a lot in those first three years. I quit after the fifth attempt, and honestly I was already a year too late by then. I have seen this from both sides, as a student who sat it and as a teacher who taught economics, public administration, and organisational behaviour to the people sitting it. What follows is all of that put together. This is how it looks to me.
The exam is not a scam, and the people who clear it are not lucky. It selects, ruthlessly, for real things: discipline held over years, raw cognition, the ability to carry an enormous syllabus and produce structured answers under time. Most people who seriously decide to attempt it are already in the top 10 to 20 percent of their age cohort. These are not the country's stragglers. They are some of its most capable and most determined youth. When India had no better way to find administrators for a billion people, this exam was a reasonable instrument. It did a good job for more than half a century, and it stays fair in the one way that matters most to a poor family. It doesn't ask who you are, who your parents are, or even whether you have been a good student so far.
There was a time when chasing it was simply the right move. For my parents' and grandparents' generation, in a country with almost no private opportunity, a government job was the safest and often the only ladder out of poverty. A whole family would put its weight behind one child clearing one exam, and they were right to. The job was secure, it was respected, and it pulled everyone attached to it up a rung. They were not wrong about their world.
What has changed is everything around the exam. The opportunities outside it have exploded, and a lot of them now offer better odds for the same effort, sometimes for less. A bet that was sound two generations ago has quietly become a strange one to keep making. To still treat the government exam as the obvious pinnacle of success in 2026 is to be out of touch with the country we actually live in. The exam did not change. The world around it did, and the script in people's heads never caught up.
What I keep coming back to is where all of that talent ends up going.
Close to a million people sit the UPSC exam in a year. The final list is a few hundred to about a thousand. That is a fraction of one percent, as arithmetic, not as a figure of speech. The other ninety-nine-point-something percent don't disappear. They spend the better part of their twenties, in some cases the whole decade, inside a funnel that was always going to reject almost all of them. The twenties are the highest-energy, highest-risk-tolerance, highest-leverage years a person gets, and the country pours the top decile of them into one narrow opening.
What they are left holding is the real cost. A decade of preparation produces nothing you can show. No product, no code, no company, no body of public work. It produces a CV the private sector quietly reads as a gap. Some of the most disciplined young minds in the country spend their best years earning a credential that pays out for the few who clear and is worth almost nothing to everyone who doesn't.
I used to make my students do the actual math on this, and it didn't make me popular. The coaching costs a few lakh. The years cost a salary you never earned, maybe ten lakh a year for a capable person, and rising. So a five-year attempt doesn't really cost four lakh in fees. It costs closer to fifty lakh in income you never made, plus the compounding on it, plus the skills you didn't build and the network you never formed. Every aspirant is sitting in front of the cleanest example of opportunity cost there is, paying to study every cost except their own.
Every year the same hunger that could have built a hundred thousand companies goes into clearing a rank instead. Nothing records this loss, because there is no line item anywhere for the company that was never started by the person who spent their twenties memorising the model answer. We have the talent and the grit. We keep pointing the best of it at a queue.
The instinct is to fix the exam. More seats, more attempts, a softer cutoff. That makes it worse. A wider opening pulls more people in and holds them longer, and produces more of the same decade-shaped credential that pays off for almost no one. The thing worth changing was never inside the exam hall. It sits upstream, in what a sharp eighteen-year-old is told to do with all that hunger in the first place.
I don't want the exam gone. The country genuinely needs administrators, and for some people this is the right life. The temperament fits, the work is real, public service is a real thing to want. What sits wrong with me is its place as the default. Most of the students I taught had not really chosen it. They were there because it looks glamorous, because of the movies, because their parents pushed, because of the one uncle who talks about the service like it is larger than life. Hardly any of them had looked at the real cost before giving it years. Building could carry the same respect as a rank, and the odds are far better, because there is no list of a thousand. Building is broader than starting a company. It is research, deep science, deep tech, creating something that becomes intellectual property, solving a hard problem in whatever form it takes. The person who can hold the entire polity syllabus in their head can hold a product roadmap, or a research problem, or a hard technical build. The discipline that survives a third attempt survives a second pivot.
Picture the other version. A million builders a year. Some starting companies, some in labs, some deep in a technical problem, each chasing something specific to India, a few of them breaking through every single day. No fixed number of seats, no cap on attempts, so the learning compounds across every experiment. Even the failures pay something back, because someone who spent two years deep in a problem comes out an authority on it, and a failed builder finds work far more easily than a failed aspirant ever does. The ones who succeed go on to employ thousands, or push a field forward, or leave behind something others build on. Add it up over a decade, and a million builders a year solve more of India's problems than the few thousand officers selected over the same decade can across their whole careers. I am not saying everyone should build. I lean that way for the people who have the inclination, the appetite for risk, and a real backup, which is roughly what UPSC itself quietly demands anyway. Even our PM Modi has made this point in Parliament, that the senior officers he meets say their own children no longer want the service, they want to start something of their own.
Building is not easy, and not everyone should do it. But the people who spend a decade in the queue and never clear it were not less capable than the ones who did. They were capable the entire time. We just never built them anywhere else to go. When I quit, a year too late, I had not run out of ability. I had finally noticed the doors I'd been walking past.
And UPSC is only the part we talk about. Step back, and underneath it sits the rest of the government-exam economy: the state commissions, the staff selection tests, the banking exams, the PSU and railway recruitments, the clerk and constable and peon posts that never make the news. Here the numbers stop being lakhs and turn into crores. In 2018 the railways advertised about ninety thousand jobs, and roughly two and a quarter crore people applied. A few years before that, one state government opened 368 peon posts, the kind that ask for a Class 5 pass and the ability to ride a bicycle, and twenty-three lakh people applied, more than a lakh and a half of them graduates, two hundred and fifty-five of them PhD holders. PhDs standing in line for a peon's job, because the job came with a government stamp, and the stamp is what the whole country has been trained to want.
That is the real scale of it. Not a few hundred thousand UPSC aspirants, but tens of millions of people across every level of every exam, each handing over years they will never get back, to a system that was never going to hire most of them. Government cannot employ the crores who want a job from it. The arithmetic forbids it, and no reform changes the arithmetic. A few lakh posts a year against tens of millions of applicants is a lottery with a syllabus. The only thing that can ever absorb that many people is the companies and institutions that do not exist yet, and those have to be built by the same people now standing in the queue.
We are not short of talent in India. We are short of places for it to go that aren't a queue.
Open to hearing your thoughts.