On this page
English only · Odia translation in progress
Chapter 4: The Years That Disappear
The room is ten feet by ten feet. It is on the second floor of a three-storey building in Saheed Nagar, Bhubaneswar, one of perhaps three hundred similar rooms rented to aspirants in the blocks between Master Canteen Square and Jaydev Vihar. The rent is Rs 6,000 per month, shared between two occupants, so Rs 3,000 each. The room contains two single beds, a wooden desk, a plastic chair, a steel almirah, a ceiling fan that squeaks on its third speed, and a small shelf bolted to the wall. On that shelf: six NCERTs, four Laxmikanth editions in varying states of disintegration, Ramesh Singh’s Indian Economy, a GS Manual, three years of previous question papers bound with string, and a stack of photocopied notes from a coaching class that closed during COVID and never reopened.
On the wall above the desk, a timetable written on chart paper in blue and red ink. The structure has not changed in fourteen months:
6:00 AM — Wake. Tea from the stall downstairs (Rs 10). 6:30-9:00 AM — Current affairs + newspaper (The Hindu, editorial + economy page, sometimes Indian Express). 9:00-9:30 AM — Breakfast. Two parathas and dal from the mess across the road. Rs 40. 9:30 AM-1:00 PM — Static GS. Polity on Monday, Economy on Tuesday, History on Wednesday, Geography on Thursday, Science and Environment on Friday. Saturday: revision. Sunday: mock test. 1:00-2:00 PM — Lunch. Rice, dal, sabji, sometimes fish. Rs 50. 2:00-5:00 PM — Optional subject. Sociology, because the syllabus overlaps with GS and because the success rate for sociology optionals has been above average for the last five years. 5:00-5:30 PM — Break. Walk to Jaydev Vihar Square and back. 5:30-8:00 PM — Answer writing practice. Two questions from the previous year’s mains, timed at fifteen minutes each, then self-evaluation against model answers. 8:00-9:00 PM — Dinner. Same mess. Rs 50. 9:00-11:00 PM — Revision or weak areas. 11:00 PM — Sleep.
The calendar next to the timetable has a date circled in red: the last Sunday of May, the expected prelims date. Below it, in smaller handwriting, a number that gets updated every week: the days remaining. When I describe this room, the number reads 47.
The man at the desk is twenty-seven years old. He is from Ganjam district, from a family where his father retired as a Block Development Officer — a Group B state government post that took thirty-two years to reach. His mother is a homemaker. He has one sister, married, in Berhampur. He graduated from Utkal University in 2020 with a degree in political science. He has attempted UPSC twice. Prelims cleared once, in his second attempt, 2023. Mains: did not qualify. He is now in his third attempt cycle. His father sends Rs 12,000 per month. This covers rent, food, and coaching material. There is no margin.
His name does not matter for this chapter. What matters is that there are, at any given time, between 25 and 35 lakh people in India living some version of this life. In Bhubaneswar alone, the number is hard to estimate precisely — no census captures “aspirants” as a category — but coaching institutes, hostel registrations, and the geography of the city’s aspirant neighborhoods suggest somewhere between 15,000 and 30,000 people are preparing for civil services (UPSC/OPSC) examinations at any given time. Nationally, approximately 13 lakh people registered for UPSC Prelims in 2024. Perhaps twice that number are in some stage of preparation without registering in a given year, because not everyone who is preparing takes every exam — some are in their first year and waiting, some skipped a cycle, some are between attempts.
They live in rooms like this. They follow timetables like this. And they are in year three, or year four, or year five, of a process that has no guaranteed end.
This chapter is about what those years cost. Not the money — Chapter 2 covered the financial economics. This chapter is about the years themselves. The time. The identity. The psychological architecture that makes it impossible to leave even when leaving is the rational choice. The game theory of sunk costs applied not to investments or military campaigns but to the most intimate thing a person possesses: the years of their life.
The Time Data
Before examining the psychology, it is worth understanding the arithmetic. Because the numbers themselves tell a story that the motivational posters in coaching institute lobbies prefer to leave out.
The average UPSC aspirant spends 3 years and 3 months in active preparation before either succeeding or permanently exiting the system. This figure, compiled from surveys of both successful and unsuccessful candidates across coaching platforms, represents the mean. It obscures a distribution that is heavily right-skewed: many candidates spend two years or less before dropping out, while a significant minority spend five to seven years. For aspirants above the age of twenty-eight, the average rises to 3 years and 7 months — the older you are when you start, the longer you tend to stay, because the outside options are worse and the psychological commitment is deeper.
The attempt data tells its own story. UPSC allows six attempts for General category candidates between the ages of twenty-one and thirty-two. OBC candidates get nine attempts, with an age limit of thirty-five. SC/ST candidates face no limit on attempts and an extended age limit of thirty-seven. These are not merely administrative categories. They are the structural parameters of a commitment device — the system telling you, in effect, how many years of your life it is willing to accept.
Among candidates who successfully clear the exam, 93 percent needed more than one attempt. This single statistic is the most dangerous number in the entire system, because it contains a message that every aspirant internalizes: failure the first time is normal. Failure the second time is still within the bell curve. The third or fourth attempt is where success peaks — approximately 22 to 24 percent of successful candidates clear the exam on their third or fourth try. The average number of attempts among those who eventually succeed is 3.6.
Read those numbers carefully. They say: if you succeed, you will most likely succeed on your third or fourth attempt. They do not say: if you attempt three or four times, you will succeed. The conditional probability runs in one direction. The aspirant, understandably, reads it in the other. This is not a failure of reasoning. It is the architecture of a trap.
Consider the implied timeline. A student graduates at age twenty-one or twenty-two. The first year is foundation — building the knowledge base from scratch, since no undergraduate curriculum in any Indian university adequately prepares students for the UPSC syllabus. The first serious attempt comes at age twenty-three. Fails prelims. Second attempt at twenty-four. Clears prelims, fails mains. Third attempt at twenty-five. Clears prelims, clears mains, fails interview. Fourth attempt at twenty-six.
At this point, the aspirant has spent four years. They have cleared prelims twice, mains once. They know the system. They know their weak subjects. They have reason to believe they are close. The difference between clearing the interview and not clearing it could be a matter of ten or twenty marks out of 2025. A better answer in one question. A more confident response to a panel member’s provocation. They are, by any rational assessment, closer to success than they have ever been.
And this is precisely the moment when the trap’s jaws close.
The Dollar Auction
In 1971, the economist Martin Shubik described an experiment that has since become one of the most famous illustrations of irrational escalation in game theory. He called it the dollar auction.
The rules are simple. An auctioneer offers a dollar bill for sale. Bidding starts at one cent. The catch: both the highest bidder and the second-highest bidder must pay their bids. The highest bidder gets the dollar. The second-highest bidder pays and gets nothing.
Watch what happens. Bidding begins sensibly. Someone bids five cents for a dollar — an obvious bargain. Someone else bids ten cents. The escalation is rational: every bid below a dollar represents a profit for the winner. At fifty cents, the game still makes sense. At ninety cents, the winner still gains ten cents.
The inflection point arrives when the bidding reaches ninety-nine cents. The current high bid is ninety-nine cents. The second-highest bidder has bid ninety-eight cents. If the second-highest bidder drops out, they lose ninety-eight cents. If they bid one dollar, they break even (pay one dollar, receive one dollar) and the other person loses ninety-nine cents. So they bid one dollar.
Now the previous leader faces a choice. They can drop out and lose ninety-nine cents. Or they can bid $1.01 — paying more than the dollar is worth, but losing only one cent net (pay $1.01, receive $1.00, net loss one cent) instead of losing ninety-nine cents. So they bid $1.01.
And now both bidders are paying more than the prize is worth. But the logic that compels each escalation has not changed. At every stage, the cost of dropping out (losing your entire bid) exceeds the marginal cost of one more increment. The bidding can — and in Shubik’s experiments, regularly did — exceed the dollar’s value by two, three, even five times. The auction does not stop because the bidding becomes irrational. It stops because someone finally absorbs a massive loss to break the cycle.
Shubik was not studying currency auctions. He was studying Vietnam. The dollar auction was designed to model the psychology of a war in which both sides have invested too much to stop but cannot win enough to justify continuing. The United States spent $168 billion (roughly $1.4 trillion in 2024 dollars) and 58,220 American lives in a conflict that its own analysts knew, by 1967, could not be won on terms that justified the investment. The cost of withdrawal — political humiliation, the “loss” of everything already spent — consistently exceeded the marginal cost of continuing for one more year, one more escalation, one more bombing campaign. Each escalation was locally rational. The aggregate was catastrophic.
The UPSC aspirant is in a dollar auction.
After one year and Rs 1-2 lakh invested, quitting means losing everything. The year cannot be recovered. The money cannot be recovered. The knowledge gained — detailed understanding of Indian polity, economic surveys, geography, history — has almost no market value outside the exam system. No employer advertises for candidates with expertise in “Laxmikanth’s Indian Polity, 7th edition.” The investment is sunk in the strictest economic sense: it has value only if the exam is cleared.
After two years and Rs 3-5 lakh, the sunk cost is larger. The aspirant has now crossed one or more preliminary hurdles. They have proof that they are in the game. Quitting at this point means losing not just money and time but evidence of capability. The person who cleared prelims and failed mains is closer than the person who has never tried. This proximity to success — the perception of almost-there — is the gasoline that the dollar auction runs on.
After three years and Rs 5-8 lakh, the aspirant has restructured their entire life around the exam. Their social circle is composed of other aspirants. Their daily routine is the timetable on the wall. Their identity — to their family, their village, their own internal self-narrative — is “UPSC aspirant.” Quitting at year three does not mean returning to the starting line. It means starting a different race from a position that is worse than the starting line, because three years have passed and the world outside the exam has continued without them.
At each stage, the marginal cost of one more attempt — one more year of preparation, Rs 1-2 lakh more in expenses — is less than the total cost of dropping out and accepting the loss of everything invested. The logic is identical to Shubik’s bidders. It is identical to the Pentagon’s planners in 1968. It is the defining feature of the escalation of commitment: the deeper you are, the harder it is to stop, even when stopping is the only rational choice.
The critical difference between the dollar auction and the aspirant’s situation is that in Shubik’s experiment, the auctioneer is a person who can be identified and blamed. In the exam system, the auctioneer is invisible. It is the structure itself — the combination of age limits, attempt limits, probability distributions, and social expectations that create a game in which every individual decision to continue is locally rational while the aggregate outcome is the destruction of millions of years of human potential.
Identity Lock-In
There is a concept in social psychology called identity fusion — the state in which a person’s individual identity merges so completely with a group or role that the boundary between self and role dissolves. Soldiers experience it. Cult members experience it. Entrepreneurs whose companies become their identity experience it. And UPSC aspirants experience it with a completeness that researchers who study identity formation would find textbook-perfect.
The process begins subtly. In the first few months, “I am preparing for UPSC” is a description of an activity. Like “I am learning to drive” or “I am taking a course.” The self is separate from the activity. The activity could be abandoned without the self being diminished.
Six months in, the language shifts. “I am a UPSC aspirant.” The indefinite article matters. It signals category membership. The person has moved from doing something to being something. The activity has become an identity marker.
By the end of the first year, the identity has restructured the social world. The aspirant’s friends are other aspirants. Conversations revolve around current affairs, optional subject strategy, answer-writing technique, and the precise implications of this year’s question paper pattern. Non-aspirant friends — college classmates who took private-sector jobs, or who are in different exam pipelines — gradually fade from the social circle. Not because of hostility, but because the shared language has diverged. The aspirant lives in a world of “GS papers” and “CSAT cutoffs” and “interview boards.” The non-aspirant lives in a world of “performance reviews” and “client meetings” and “weekend plans.” The vocabularies become mutually unintelligible, and friendship follows vocabulary.
The family reinforces the identity from the outside. In Ganjam, in Puri, in Balasore, in Cuttack, in every district of Odisha where a family has sent a child to prepare, the child’s identity in the extended family’s mind is fixed: “He is preparing for IAS.” Not “he is studying” or “he is figuring things out.” He is preparing for IAS. This is how relatives introduce the topic. This is what neighbors know. This is what comes up at every wedding, every funeral, every festival gathering when the family is asked about the child.
“Ki karuchhi chhuan? UPSC?” — What is the boy doing? UPSC?
The question is asked with a tone that mixes respect, curiosity, and a faint edge of assessment. The family answers with a tone that mixes pride, anxiety, and the weight of public commitment. Every answer — “Yes, second attempt, preparing well” — is a social contract being renewed. The family has staked its narrative on this outcome. The extended network has registered the bet.
Consider what quitting means in this context. It does not mean “I have decided to pursue a different career.” It means dismantling an identity that has been constructed publicly, reinforced socially, and internalized psychologically over a period of years. The aspirant who quits must answer the question — and it will be asked, in every social interaction, for years — “Why did you stop?” And there is no answer that does not translate, in the vocabulary of the community, to “I failed.”
This is the mechanism of identity lock-in. The concept comes from technology economics, where it describes a situation in which switching costs make it prohibitively expensive to move from one platform to another, even when the alternative platform is objectively superior. A company that has built its entire infrastructure on Oracle databases cannot switch to PostgreSQL without rebuilding everything, even if PostgreSQL is free and Oracle costs millions. The switching cost is not the license fee. It is the accumulated investment in Oracle-specific skills, Oracle-specific workflows, Oracle-specific data structures. The cost of switching includes the cost of undoing everything you have already built.
The aspirant’s identity lock-in works the same way. The “platform” is the UPSC preparation ecosystem. The accumulated investment is not just money and time but the social infrastructure built around the identity: the friend group, the family narrative, the daily routine, the self-concept. Switching to a different “platform” — taking a private-sector job, starting a business, pursuing a different exam — requires not just building new skills but dismantling the entire social and psychological infrastructure that has been constructed over years. The switching cost is not the opportunity; it is the unraveling.
And the most insidious aspect of identity lock-in is that it makes the person unable to accurately evaluate alternatives. A person whose self-concept is “UPSC aspirant” processes information about non-UPSC careers through a filter that is structurally biased toward finding those alternatives inadequate. A private-sector job paying Rs 25,000 per month is not evaluated on its merits — its stability, its growth potential, its immediate cash flow, its ability to build skills that appreciate over time. It is evaluated against the imagined IAS outcome: lifetime earnings of Rs 4-8 crore, social status, housing, medical benefits, pension, marriage market premium. Against that imagined benchmark, every real alternative looks pathetic. The aspirant does not see a viable career path. They see a consolation prize. And nobody wants to trade an identity for a consolation prize.
The Social Commitment Trap
In 1955, the psychologists Morton Deutsch and Harold Gerard demonstrated a phenomenon they called the commitment effect. In a simple perceptual judgment task, participants who wrote down their answers publicly and showed them to the group were significantly more resistant to changing their answers when confronted with contradicting evidence, compared to participants who kept their answers private. The act of public commitment — of announcing your position where others could see it — changed the psychological relationship to that position. It was no longer a tentative judgment. It was a social fact.
The UPSC aspirant’s commitment is among the most public commitments a young person in Indian society can make.
It begins at home. The decision to prepare for UPSC is rarely a private one. It involves the family — because the family funds it. The father who adjusts his budget, the mother who reduces household expenses, the sister whose own education or marriage may be deferred — they are all stakeholders. The investment is collective. The announcement is collective. The expectation is collective.
It radiates outward through kinship networks. In Odisha’s tightly knit caste and community structures — and particularly in districts like Ganjam, where social prestige is calibrated with extraordinary precision to government rank and posting — the knowledge that a family has a child “preparing for IAS” confers a specific social status. It is an advance on expected prestige. The family begins to be treated, subtly, as a family that will soon have an IAS officer. Relatives calculate alliance possibilities. Neighbors adjust their respect levels. The village files the information in its collective memory.
“Amara chhuan IAS paduchhi” — Our son is studying for IAS.
The sentence is not information. It is a social position. And once that position has been claimed publicly, abandoning it carries a cost that is social rather than financial.
The paradox is this: the same community support that enables the aspirant to prepare — the family’s financial sacrifice, the relatives’ encouragement, the village’s pride — becomes the prison that prevents exit. Sociologist Robert Cialdini described this as the consistency principle: once people have publicly committed to a position, they experience intense pressure to behave consistently with that commitment, even when the commitment has become costly. The pressure is not always external. Much of it is internal — the discomfort of being seen as someone who quit, who could not follow through, who “wasted” years.
The consistency pressure operates differently on the aspirant and the family, and the difference is important.
For the aspirant, the social cost of quitting is shame. Not the productive shame that corrects behavior, but the paralyzing shame that prevents honest evaluation. The aspirant who tells his father “I want to stop preparing” is not delivering a career update. He is delivering a confession of failure to the person who sold land to fund the preparation. The emotional valence of that conversation is not “I have made a strategic assessment of my probability of success and concluded that my expected value is higher in an alternative career path.” The emotional valence is: “I am sorry. Everything you sacrificed was for nothing.”
For the family, the social cost is more diffuse but equally powerful. They have told relatives. They have adjusted their social position. The mother has endured the questions at weddings — “How is the preparation going?” “When will we hear good news?” — with a patience that is a form of public investment. Announcing that the son has stopped preparing requires re-narrating the family’s story. The family that was going to produce an IAS officer is now the family whose son “couldn’t make it.” In the social economy of Odisha’s communities, where government jobs carry prestige that no private-sector success can match, this re-narration is expensive.
The result is a dynamic where continuing is socially easier than quitting, even when quitting is rationally optimal. The aspirant who stays in the queue for one more year faces the minor social cost of “still preparing” — a status that, while it erodes over time, is at least consistent with the original commitment. The aspirant who exits faces the immediate, concentrated social cost of “gave up.” In decision theory, this is a manifestation of loss aversion: people experience losses roughly twice as intensely as equivalent gains. The social loss of quitting is felt more acutely than the potential social gain of starting a new career, even if the new career would, over time, produce a better outcome.
And here is where the trap becomes truly binding. Every year that passes without success increases the social cost of quitting. The family has now sacrificed for two years, then three, then four. The community has been told “next year” so many times that stopping requires an explanation for all the previous years, not just the current one. The aspirant is not quitting a one-year endeavor. They are retroactively converting four years of sacrifice into waste. The social cost accumulates with time, just as the sunk financial cost does. The longer you stay, the harder it is to leave.
A family I will not identify — from a district in southern Odisha, the details changed enough to protect their privacy but not enough to distort the pattern — funded their eldest son’s UPSC preparation for six years. The father was a high school teacher. The mother worked at an anganwadi. The son graduated from a government college in 2017, began preparing in 2018, and attempted the exam every year through 2024. He cleared prelims three times. He cleared mains once. He never cleared the interview. By 2024, the family had spent approximately Rs 9 lakh — more than two years of the father’s salary. The son was twenty-nine.
In 2024, after the sixth attempt, the father sat with his son and asked: “What now?” The son said he wanted one more attempt. The father did not argue. He could not. Six years of financial and emotional investment stood between him and the word “stop.” Saying stop would mean that those six years — the reduced meals, the sister’s deferred marriage, the mother’s extra work at the anganwadi center for a monthly stipend of Rs 3,500, the loans from relatives — were all for nothing. Continuing for one more year cost Rs 1.5 lakh. Admitting that six years were wasted cost everything.
They continued.
This is not a story about an unusual family. This is the modal story. The pattern repeats in thousands of families across Odisha and millions across India. The numbers change — sometimes it is five years, sometimes seven, sometimes the family is wealthier and the financial pressure is less acute. But the social commitment trap operates identically across income levels. The wealthy family has more money to burn, but the social cost of announcing failure is, if anything, higher, because the community’s expectations scale with the family’s status.
The Age Penalty
Every labor market in the world penalizes age at entry. The penalty varies by profession, by economy, by cultural context. But it exists everywhere, and in India’s formal employment market, it is particularly brutal.
Campus recruitment — the primary mechanism by which India’s formal private sector hires at entry level — operates within a window of ages twenty-one to twenty-four. Companies visit colleges during the final year of undergraduate or postgraduate education. Students who are present, who have the right degree, who are within the expected age range, enter the pipeline. Students who are not present — because they are in a room in Saheed Nagar studying Laxmikanth — are invisible to this system.
The window is narrow and it does not reopen. An engineering graduate who misses campus placement at age twenty-two and then spends three years preparing for UPSC cannot, at age twenty-five, re-enter the campus recruitment system. The companies have moved on. The next year’s batch has graduated. The positions have been filled. The market does not wait.
Private-sector entry-level hiring outside of campus recruitment extends the window slightly — to approximately ages twenty-two to twenty-five for most sectors, and to twenty-seven or twenty-eight for certain professional services firms and the IT industry. But by age twenty-eight, the vast majority of corporate hiring doors have closed. Not officially — no HR policy states “we do not hire people over twenty-eight for entry-level positions.” Unofficially, the screening algorithms, the resume reviewers, and the hiring managers all apply the same filter: a twenty-eight-year-old applying for an entry-level position, with a three-to-five year gap on their resume explained as “UPSC preparation,” raises questions. The questions are not asked out of cruelty. They are asked because the hiring manager has a hundred other applications from twenty-two-year-olds who do not need to explain a gap.
The age penalty compounds. At twenty-two, the graduate has all doors open. At twenty-five, after three years of preparation, the campus recruitment door has closed. At twenty-eight, after six years, most formal private-sector doors have closed. At thirty, the remaining options — mid-career lateral hiring, entrepreneurship, government contracts, the informal economy — require either specific experience (which the aspirant does not have, because they were studying polity, not working) or capital (which the aspirant does not have, because it was spent on coaching).
This creates a devastating feedback loop with the sunk cost trap. The longer the aspirant stays in the queue, the fewer alternatives exist outside the queue. The fewer alternatives exist, the more rational it appears to stay. The aspirant who quits at age twenty-three has a reasonable chance of entering the private sector through off-campus placement, lateral hiring, or postgraduate education. The aspirant who quits at age twenty-eight has meaningfully fewer options. The aspirant who quits at age thirty-one may find that the only available employment resembles what they would have done at age twenty-two — but at a decade’s delay and at a lower relative position, because the twenty-two-year-olds have been building experience while they were building answer-writing skills.
The skills mismatch is total. The UPSC aspirant develops a formidable set of competencies: rapid comprehension of complex material, organized note-taking, structured writing under time pressure, broad knowledge of governance, economics, history, geography, science, and current affairs. These are genuine skills. In any rational labor market, a person who has spent three years developing the ability to synthesize a two-thousand-word essay on fiscal federalism in forty-five minutes should be highly employable.
But the market does not recognize these skills because the market has no category for them. Resume screening filters look for job titles, company names, project descriptions, technical certifications. “Prepared for UPSC CSE for three years” does not match any filter. The skills are real. The signal is absent. The aspirant is, in the language of labor economics, experiencing a signaling failure: they possess human capital that cannot be communicated through the channels the market uses to assess candidates.
The most painful dimension of the age penalty is that it is predictable and universally known — and yet it does not deter entry into the queue. Every aspirant who begins preparation at twenty-two knows, intellectually, that the campus recruitment window will close by the time they are twenty-four. Every aspirant knows that each additional year narrows the external options. The knowledge does not change the behavior, because the expected payoff of the government job — even at low probability — remains higher than the expected payoff of the narrowing alternatives. The age penalty does not defeat the sunk cost trap. It feeds it. The worse the outside options become, the more rational it feels to stay.
The Invisible Toll
This section will not deploy mental health statistics as rhetorical ammunition. The numbers deserve better than that. They represent people, and the people deserve to be seen as more than data points in an argument about the exam system’s costs.
That said, the data exists, and ignoring it would be dishonest.
In 2022, the National Crime Records Bureau reported 1,680 student suicides in India attributed to “failure in examination.” This number is almost certainly an undercount — NCRB data relies on First Information Reports, which in many cases list the proximate cause rather than the underlying driver, and families have powerful incentives to avoid having “suicide” appear on an official record. The actual figure is not knowable with precision, but researchers who study the subject believe the true number is significantly higher.
Among the NCRB categories, examination-related causes are not separated by type of examination. There is no way to extract, from the published data, how many of these deaths were linked specifically to UPSC, or to OPSC, or to SSC, or to JEE, or to NEET. The categories blur. What the data does establish is the order of magnitude: more than four people per day, every day, for an entire year, die by suicide in circumstances where examination failure was identified as a contributing factor.
Research specific to competitive exam aspirants paints a more detailed picture. A 2024 study of JEE and NEET aspirants found that 45 percent experienced anxiety or depression. This is not a marginal rate. It is close to a coin flip. The median aspirant — not the outlier, not the fragile exception — is psychologically distressed. Among long-term UPSC aspirants, the rates are less well-documented, but clinicians who work with this population in Delhi’s Mukherjee Nagar and Rajinder Nagar — the two largest UPSC preparation hubs — report patterns that include chronic anxiety, sleep disruption, social withdrawal, depressive episodes, and a specific form of obsessive ideation organized around “the next attempt.”
The aspirant population’s mental health crisis has a structure, and that structure maps precisely onto the commitment traps described above.
Anxiety tracks the approaching exam date. As prelims approach, the aspirant’s daily experience is dominated by the awareness that a single day’s performance — six hours, 200 questions — will determine whether twelve months of preparation were productive or wasted. The stakes attached to each attempt are disproportionate to the actual marginal difference in preparation between adjacent attempts. The aspirant who is 5 percent better prepared than last year faces the same binary outcome: clear or fail. The anxiety is not irrational. It is the appropriate emotional response to a system that converts continuous effort into a binary verdict.
Depression tracks the post-result period. The weeks following a failed attempt — particularly a failed mains, where the aspirant was close enough to taste it — produce a psychological pattern that clinicians describe as reactive depression: low mood, loss of interest, sleep disruption, appetite changes, and a pervasive sense of worthlessness. In a healthy psychological environment, reactive depression resolves as the person processes the loss and redirects their energy. In the aspirant’s environment, the depression does not resolve, because the response to failure is not redirection but re-commitment. The aspirant begins preparing for the next attempt before the depression from the current failure has lifted. The preparation becomes a coping mechanism — the structure of the timetable provides a scaffolding for a mind that would otherwise collapse — but it also prevents genuine recovery, because the source of the depression is not addressed. It is deferred to the next cycle.
Isolation tracks the identity lock-in. As the aspirant’s social world narrows to other aspirants, the normal sources of psychological resilience — diverse social connections, activities unrelated to the stressor, relationships that affirm the self beyond a single dimension — atrophy. The aspirant’s social media feed is aspirant content: motivational quotes, strategy discussions, cutoff analyses, success stories. Their phone calls are with other aspirants. Their definition of a “good day” is a productive study session; a “bad day” is one where concentration failed. The entire emotional landscape has been compressed into a single dimension: the exam.
The social media dimension deserves specific attention. An aspirant’s Instagram or Facebook profile, if it exists, often presents a curated image of discipline, motivation, and purpose. Motivational quotes — “Success is not final, failure is not fatal” — are shared at a rate that masks what is actually happening. The quotes are not expressions of confidence. They are incantations against despair. The aspirant who posts “One more year of hard work and the dream will be mine” at 11 PM is not making a prediction. They are trying to convince themselves. The gap between the social media projection and the lived reality — the anxiety, the doubt, the loneliness, the fear of telling parents that this year also did not work — is where the psychological damage accumulates.
And then there is the taboo of failure. In Odisha’s social structure, where government employment is the gold standard of family achievement, the aspirant who fails is not merely someone who did not get a job. They are someone who attempted the highest ambition available and was found wanting. The community does not process this as “the odds were 0.1 percent and the outcome was statistically expected.” The community processes this as a verdict on the person. The family does not discuss it openly. The topic becomes the thing that is not mentioned — at weddings, at festivals, at family gatherings. The silence is not cruelty. It is the community’s way of managing a loss that it has no vocabulary to process. But the silence compounds the aspirant’s isolation, because the one space where the failure might be acknowledged and grieved — the family home — becomes another space where it must be hidden.
I want to be careful here about what I am and am not claiming. I am not claiming that the exam system causes mental illness. Mental health is complex, multi-factorial, and poorly served by single-cause explanations. What I am claiming is that the exam system’s structural features — the binary outcome, the annual cycle, the high stakes, the social commitment, the identity fusion, the sunk cost escalation — create conditions under which pre-existing vulnerabilities are activated and ordinary psychological distress is prevented from resolving. The system does not create the wound. It prevents the wound from healing.
The Gendered Queue
Everything described above — the sunk cost trap, the identity lock-in, the social commitment, the age penalty, the psychological toll — operates on women aspirants as well. But it operates differently, because the social architecture in which the commitment occurs is gendered, and the penalties for staying are gendered, and the penalties for leaving are gendered.
The first asymmetry is time. A woman preparing for UPSC faces the same age limits as a man in her category. But the social clock that tracks her is different. At twenty-four, a male aspirant’s family faces questions about his preparation. At twenty-four, a female aspirant’s family faces questions about her marriage. By twenty-six — after four years of preparation, which is statistically normal for eventual success — the questions are no longer polite inquiries. They are social pressure applied through the community’s most effective mechanisms: relatives who express concern, neighbors who speculate, mothers who receive “proposals” that they cannot indefinitely refuse.
“She is twenty-six and still preparing” carries a different social weight than “he is twenty-six and still preparing.” The male aspirant’s continued preparation is a sign of ambition. The female aspirant’s continued preparation is, in the community’s assessment, a risk — the risk that she is “wasting” her marriageable years on an uncertain outcome. The family that supports a son’s preparation for six years may balk at supporting a daughter’s for four, not because they love her less but because the social cost of her delayed marriage falls on the family’s reputation in ways that the son’s delayed career does not.
The marriage market itself creates a specific form of the sunk cost trap for women. A woman who has “invested” years in UPSC preparation and then enters the marriage market without having cleared the exam faces a paradox: her education and ambition, which might be assets, are often coded as liabilities. She is “overqualified” in a market that values compliance. She is “too old” in a market that penalizes delay. She has spent years in a competitive, intellectually demanding environment that has given her expectations about autonomy, partnership, and intellectual engagement that may not match what the marriage market offers. The mismatch is not her failure. It is the system’s — both the exam system that consumed her years and the marriage market that penalizes her for it.
The data is stark. Among educated women graduates in India, 39.51 percent are unemployed — nearly double the male rate. This is not because women graduates are less capable. It is because the labor market offers them fewer positions, the social structure imposes constraints on mobility and work conditions, and the exam pipeline that consumed their prime years has no exit ramp designed for women who need to transition into the labor force while simultaneously navigating marriage expectations, family obligations, and a market that structurally discriminates against older female applicants.
Women aspirants in Odisha face an additional layer. The state’s female labor force participation rate has improved — reaching 48.7 percent in recent surveys, significantly above the national average — but this aggregate masks a specific gap: educated urban women face higher unemployment than less-educated rural women, because rural women work in agriculture (where demand for labor is constant, if poorly compensated) while urban educated women seek formal employment (where positions are scarce and gendered hiring biases operate). The UPSC aspirant is by definition urban and educated. She is in the category with the worst employment outcomes.
The women’s version of the government job queue contains the same trap as the men’s, but with a shorter fuse and a heavier penalty for detonation. The sunk cost argument is the same. The identity lock-in is the same. The social commitment is the same. But the consequences of staying too long are magnified by a social system that imposes costs on women that it does not impose on men — and the consequences of leaving are worse, because the alternative market is less forgiving of the gap on the resume and the years on the clock.
I am not confident that I have fully captured the texture of this experience. The women aspirants I have read about and heard from describe something that goes beyond the structural analysis: a specific loneliness that comes from pursuing an ambition that your own family supports financially but questions socially. The father who pays for coaching but also hints that “there’s a good boy from a government family, just saying.” The mother who is proud of the preparation but also anxious about what the relatives will say. The aspirant herself, holding two contradictory messages — pursue your ambition, and also be aware that your ambition is costing your family in a currency they did not agree to spend. This tension is not captured by the game theory framework. It is the part of the experience that the framework illuminates but cannot contain.
The Economist’s Blind Spot
There is a standard way that India’s economic statistics handle the population described in this chapter. The Periodic Labour Force Survey classifies individuals by their “usual principal activity status.” A person who is not working and not seeking work is classified as “not in the labour force.” This category includes students, homemakers, the elderly, and the disabled.
The UPSC aspirant — studying full-time, not seeking employment, not employed — falls into this category. They are classified as not in the labour force. They are, by statistical definition, invisible.
This is not a conspiracy. It is the mechanical consequence of survey categories designed for a world in which “student” and “employed” are the two primary states of a working-age adult, with a brief transition between them. The categories were not designed for a world in which millions of working-age adults spend years in a state that is neither education nor employment but a holding pattern that consumes the resources of both.
The unemployment rate — 3.1 percent in Odisha as of 2024, according to the PLFS — does not count aspirants. The labour force participation rate does not count them. They appear nowhere in the headline economic data. GDP does not count their time as a loss. National accounts do not register the Rs 50,000-58,000 crore spent on coaching as anything other than consumption expenditure — a contribution to economic activity, technically, just as spending on cigarettes is a contribution to economic activity.
This statistical invisibility has policy consequences. A government that looks at an unemployment rate of 3.1 percent and a LFPR of 64.5 percent sees a reasonably healthy labour market. The millions of aspirants who have removed themselves from the labour force — who are neither working nor seeking work, but are consuming family resources, accumulating opportunity costs, and suffering psychological distress — do not appear as a problem that requires a policy response. They are invisible by design.
If those millions were counted — if the PLFS added a category for “preparing for competitive examinations” and measured its scale, its duration, its demographic composition, and its economic cost — the picture of India’s labour market would change dramatically. A rough calculation: if 2-3 crore working-age Indians are in active exam preparation at any given time (a conservative estimate given that UPSC, SSC, banking, railways, state public service commissions, and teaching entrance exams collectively attract this order of magnitude), and each of them is forgoing Rs 1-3 lakh per year in potential earnings, the aggregate opportunity cost is Rs 2-9 lakh crore per year. This does not include the coaching fees paid. It does not include the psychological costs. It does not include the reduced productivity of family members who subsidize the preparation.
The years that disappear do not appear in any ledger. They are not counted as unemployment because the person is not seeking work. They are not counted as education because the person is not enrolled in a degree program. They are not counted as economic loss because no survey measures the counterfactual — what these millions would have produced if they had spent those years working, building skills, starting businesses, or even just living their lives. The years simply vanish. They are absorbed by the system, converted into coaching revenue and exam fees and the ghost hours of study that produce nothing for 99.9 percent of those who invest them.
The Machine That Converts Potential Into Heat
In thermodynamics, there is a concept called waste heat — the energy that enters a system but is not converted into useful work. Every engine produces waste heat. The second law of thermodynamics guarantees it: no conversion of energy from one form to another is perfectly efficient. Some energy always degrades into heat that disperses into the environment, doing nothing, warming nothing useful, simply dissipating.
The exam system is a machine that converts human potential into waste heat.
Consider the inputs. Millions of young people enter the system with energy, intelligence, education (however imperfect), ambition, and years of life ahead of them. They bring family savings, community support, and the irreplaceable resource of time. They are, in aggregate, one of the largest concentrations of educated human capital on the planet.
Consider the useful output. Approximately 1,000 people per year are selected through UPSC CSE. A few thousand more through state public service commissions. Perhaps 50,000-100,000 through the combined machinery of SSC, banking, railways, and other central recruitment. The useful work — people placed in positions where their human capital is deployed — is a tiny fraction of the total input.
The rest is waste heat. The years spent by the 99 percent who do not succeed are not converted into anything. They are not credentials. They are not work experience. They are not transferable skills in any form the market recognizes. They are not even memories of a meaningful experience, for many aspirants — the years blend into each other, one mock test dissolving into the next, one failed attempt indistinguishable from the one before. The years simply dissipate.
A charitable reading would say: the knowledge acquired during preparation has intrinsic value. The aspirant who spent three years studying Indian polity, economics, history, and current affairs is a more informed citizen, a better-informed voter, a person with a broader understanding of how the world works. This is true. It is also irrelevant to the person’s economic situation, their psychological state, and their ability to build a life after exiting the queue.
A less charitable reading — but perhaps a more honest one — would say: the system is not designed to produce useful output for the majority. It is designed to produce useful output for the minority who are selected, and to warehouse the majority in a holding pattern that prevents them from becoming a political problem. An unemployed graduate who is visibly idle is a potential source of social unrest. An aspirant who is “preparing for the next attempt” is not idle — they are busy, purposeful, and oriented toward a goal. That the goal is statistically unattainable for most of them does not diminish its effectiveness as a mechanism of social control. The queue is a pacification device. It converts restless, potentially disruptive energy into disciplined, self-directed, and ultimately harmless preparation activity.
I should state my confidence level on that last claim. It is approximately 55-60 percent. I do not believe the exam system was consciously designed as a pacification device. I believe it evolved into one through structural logic: a system that produces few jobs but many aspirants will, over time, develop features that manage the aspirant population rather than serving it. The age limits, the attempt limits, the annual cycle, the coaching industry — these are not conspiracies. They are emergent properties of a system that has far more demand than supply and has optimized, unconsciously, for containing that demand rather than redirecting it.
The counterfactual is what makes this chapter’s argument concrete. What would those years produce if they were not consumed by the queue?
A twenty-two-year-old who spends three years building skills in a growing sector — technology, healthcare, green energy, financial services — arrives at age twenty-five with three years of work experience, industry contacts, domain expertise, and a track record. Even if the sector is imperfect, even if the job is not prestigious, the three years compound. Skills build on skills. Experience opens doors to more experience. The compounding curve is slow at first but accelerates.
A twenty-two-year-old who spends three years in the exam queue arrives at age twenty-five with three years of exam preparation, no work experience, no industry contacts, no domain expertise, and a track record that the market cannot read. The three years do not compound. They depreciate. Each additional year in the queue reduces the residual value of the years that remain, because the market discounts for both age and gap.
The difference between these two trajectories is not three years. It is the compounding that those three years would have generated over a lifetime. An investment that returns 10 percent annually for forty years — a rough analogy for the career trajectory of a working professional — converts Rs 1 into Rs 45. The same investment delayed by three years converts Rs 1 into Rs 33. The three-year delay does not cost three years of returns. It costs twelve years’ worth of compounding at the end. The exam queue does not just consume years. It destroys the value of the years that remain.
The Trap Is Not the Bug
This chapter has traced the mechanisms that keep people in the exam queue past the point of rational calculation: sunk cost escalation, identity fusion, social commitment, loss aversion, the age penalty, and the psychological toll that compounds with each passing year. The game theory framework — the dollar auction, the commitment trap, the consistency principle — illuminates these mechanisms as structural features of the system, not individual failings of the people caught within it.
But the most important insight is not about the psychology. It is about the interaction between the psychology and the structure.
The sunk cost trap would not function if there were viable exits. An aspirant who could leave the queue at age twenty-five and enter a formal-sector job paying Rs 30,000 per month — with benefits, with career progression, with dignity — would find it much easier to absorb the sunk cost of two years of preparation. The loss would sting, but it would be contained by the gain of a real alternative. The trap works because the alternative does not exist. In an economy where private-sector formal employment is scarce, where labour protections are absent, where the informal sector is the default destination for those who leave the queue, the “opportunity cost” of staying in the queue is low — because the opportunity itself barely exists.
The identity lock-in would not function if society offered multiple paths to status and self-worth. In an economy where entrepreneurship was respected, where skilled trades paid well and conferred dignity, where a young person could build a life through half a dozen different routes, the UPSC identity would be one option among many, not the only respectable game in town. The identity fuses with the exam because there is nothing else to fuse with.
The social commitment trap would not function if communities had updated their mental models to include alternatives. In a society where “my son works at a tech company in Bhubaneswar” or “my daughter started a food processing unit in Berhampur” carried social weight comparable to “my son is preparing for IAS,” the commitment pressure would dissipate. The trap works because the community’s value hierarchy has exactly one summit: the government job.
The age penalty would not function if the labour market were less rigidly structured. In economies with more fluid hiring practices, where career changes at age twenty-eight or thirty-two are normal rather than disqualifying, the cost of late exit from the queue would be manageable. The penalty works because India’s formal labour market treats career discontinuity as a character flaw rather than a life event.
Each mechanism is psychological. Each mechanism is activated by structural conditions. The psychology is real — the sunk cost fallacy is a genuine cognitive bias, identity fusion is a genuine social-psychological phenomenon, loss aversion is a well-documented feature of human decision-making. But these psychological mechanisms do not operate in a vacuum. They operate in an economy that has failed to create alternatives, a society that has failed to diversify its status hierarchies, and a labour market that has failed to accommodate the reality that millions of its potential participants will spend years in a queue before becoming available.
The sunk cost trap is not a bug in human psychology. It is the operating principle of the exam system. The system works — in the sense that it continues to function, continues to attract entrants, continues to generate coaching revenue, continues to sort a small number of candidates into government positions — precisely because the psychological traps keep people in the queue. If aspirants could rationally evaluate their odds and rationally exit when the expected value turned negative, the queue would collapse. Coaching revenue would evaporate. The social pacification function would fail. The system’s stability depends on the irrationality it produces.
What the Timetable Does Not Show
Return to the room in Saheed Nagar. The man at the desk has finished his evening answer-writing session. He has written two answers: one on the implications of the 15th Finance Commission for fiscal federalism, one on the role of civil society in strengthening democratic governance. He has timed himself at fifteen minutes each. He has compared his answers against model answers from a coaching platform. He has identified two structural weaknesses: insufficient data in the first, inadequate cross-referencing of Supreme Court judgments in the second. He has made a note to revise these areas tomorrow.
He closes the notebook. He walks to the window. The window faces the back of another building, so there is no view — just a wall, a corrugated tin shade over someone’s kitchen, and the glow of a streetlight reflected off wet concrete. It rained briefly, as it does in Bhubaneswar in the weeks before summer truly arrives. The air smells of wet dust and fried onions from the dhabas on the road below.
He is twenty-seven years old. His college classmates — the ones who took jobs after graduation, the ones who went to Bangalore, the ones who joined their father’s business, the ones who married and started families — are seven years into their adult lives. They have work experience, savings, professional networks, relationships, children in some cases. They have built, even if what they have built is modest.
He has a timetable on a wall and forty-seven days to an exam that he has a roughly 0.1 percent chance of clearing.
He does not think of himself as trapped. He thinks of himself as committed. He thinks of himself as someone who is doing the hardest possible thing — pursuing the highest possible ambition — and who will be rewarded for his persistence. The motivational frameworks that the coaching ecosystem provides — “failures are the pillars of success,” “IAS is not for the talented but for the persistent,” “your twenty-eighth birthday is the beginning, not the end” — are not cynical manipulations. They are coping strategies that the community of aspirants has evolved to manage the psychological cost of the queue. They are true enough to sustain hope and false enough to obscure the arithmetic.
The years that disappear — his years, and the years of the millions like him — do not register anywhere in the official account of what India produces, what India loses, and what India costs its young people. GDP does not subtract them. The unemployment rate does not count them. The Human Development Index does not see them. They exist in a category that economics has not named: the productive years of a human life, spent in preparation for an event that will not occur, in a system that depends on exactly this expenditure to sustain itself.
The timetable on the wall is color-coded. Blue for static subjects. Red for current affairs. Green for answer writing. It is neat, precise, and comprehensive. It accounts for every hour between 6 AM and 11 PM. It does not account for the one thing that matters most: the probability that these hours are being spent in the service of an outcome that will never arrive.
He turns off the light at 11 PM. He sets his alarm for 6 AM. Tomorrow, the timetable begins again. The number on the calendar will read 46.