Daniel's been thinking about what it's like to work inside an organization so big its workforce is literally the population of a city. Here's what he wrote to us.
During my career I've worked both in-house and as a contractor for a variety of companies, but the balance has definitely skewed toward smaller organizations. My first job was in the startup world in Ireland, and after moving to Israel I continued to spend most of my career in smaller companies. I've worked with large, publicly traded organizations too, but usually through consulting or subcontracting arrangements — not quite the same as being an employee there.
Because of that, I've often wondered what it's like to work inside a truly massive organization — one so large that if you gathered everyone together, you'd literally have the population of a city. How many companies have ever reached that scale? Has anyone kept track of the largest employer in history by headcount? I imagine it's a surprisingly difficult question because of mergers, acquisitions, and subsidiaries, but even with the strictest definition — counting only direct employees — what's the ballpark? Who holds the record, and what sorts of companies make the list?
He also wants to explore what it's actually like to be part of one of these organizations. He says as someone who's never really been a small cog in a very big wheel, there's something strangely appealing about the idea, though he can see why it wouldn't suit everyone. We spend a lot of time talking about the downsides of huge companies — bureaucracy, layers of management, feeling like just another employee — but perhaps not enough about what they get right.
His own thinking has changed over time. He loves technology and startups, but after the upheaval of the past couple of years — especially the chaos around moving apartments and constantly feeling life is in flux — he's developed a real appreciation for institutions that endure. There's something deeply appealing about joining an organization that genuinely expects people to stay for decades. Not just saying in interviews "we hope you'll be here for the long term," but actually being a place where someone could build an entire career and retire if they wanted to.
Does that world still exist? Which companies are known for that kind of longevity and employee loyalty? Is it mostly a relic of the twentieth century, or are there still employers where spending an entire working life is a realistic path? And finally, for those who have experienced it, what's it really like to be part of an organization so large its workforce rivals the population of a city? From the inside, does it feel empowering, anonymous, stable, exciting — or something else entirely?
The phrase that stuck with me is "the population of a city." Because that's not a metaphor — it's a literal headcount. So the obvious first question is, who actually holds the record? How many employers in history have ever been that big?
That's where I want to start. But I want to frame this episode as two things at once. First, the empirical question — who employs the most people, by the numbers, with all the measurement headaches that come with it. Second, the experiential question — what it's actually like inside one of these giants, and whether Daniel's instinct toward institutional stability is smarter than the startup world makes it feel.
And there's a tension worth naming up front. We spend almost all our time talking about what's wrong with mega-corps — the bureaucracy, the anonymity, the layers. We almost never ask what they get right. But Daniel's prompt does ask that, and I think it's the more interesting half of the conversation.
Agreed. So we'll go from the numbers, to the mechanics of scale, to the lived experience, and then to what this means for someone actually making a career decision. Let's start with the record books.
The record holder, under the strictest definition — counting only direct employees, no contractors, no subsidiaries treated as separate entities — is Indian Railways. It peaked at roughly one point six to one point seven million employees in the nineteen eighties and nineties.
One point six million.
That's bigger than the entire population of Luxembourg. Bigger than Iceland. It's roughly the population of Phoenix, Arizona. All working for one organization.
And that's a single national railway.
A single state-owned enterprise. It runs something like thirteen thousand passenger trains a day, carries over eight billion passengers a year — though that number's from before the pandemic — and employs people in basically every category of work you can imagine. Engineers, conductors, station masters, track maintenance crews, doctors — they run hospitals — teachers for their own schools, catering staff. It's a city in the most literal sense. It has its own police force.
Its own police force.
The Railway Protection Force. About seventy-five thousand personnel. That's a mid-sized municipal police department embedded inside a larger organization.
So when Daniel says "an organization so large its workforce rivals the population of a city," Indian Railways isn't just an example — it's the definitive one. The platonic form of the thing.
And here's where the measurement problem gets interesting. "Largest employer" sounds like a straightforward question. It's not. Do you count contractors? Because if you do, the numbers shift dramatically. Do you count subsidiaries where the parent company owns a controlling stake but the subsidiary does its own hiring? Do you count franchise employees? Walmart has about two point one million employees globally today, but if you added every person who works in a Walmart store that's actually operated by a franchisee, the number would be different — though in Walmart's case most stores are corporate-owned, so the distortion is smaller than you'd think.
What about the Chinese state-owned enterprises?
That's the real complication. China National Petroleum Corporation, Sinopec, State Grid — these are enormous. But the reporting isn't always transparent, and the line between "employee" and "state worker assigned to this entity" gets blurry. The People's Liberation Army technically employs over two million people, but we don't usually count militaries in this conversation. If we did, the US Department of Defense would be on the list too — about two point nine million when you include active duty, reserves, and civilians.
So the record depends on where you draw the line.
But if you draw the line at "direct civilian employees of a single organization with a unified payroll and management structure," Indian Railways at its peak is the answer. And it held that record for decades.
Let's talk about the Western champions, because those numbers are instructive in their own right.
The US Postal Service peaked around eight hundred thousand employees. General Motors peaked around eight hundred thousand in nineteen seventy-nine. GM in nineteen seventy-nine was the largest private employer in the United States, and its decline is basically the story of the end of the manufacturing-era mega-firm. By two thousand nine it was down to about two hundred thousand, and it's roughly a hundred and sixty thousand today.
Eight hundred thousand to a hundred and sixty thousand. That's not a company that shrank — that's a company that became a different kind of entity entirely.
And the modern landscape tells a different story about where mega-workforces live. Walmart, about two point one million globally today. Amazon, roughly one point five million. Both are below Indian Railways' historical peak, but they're the current private-sector giants. The shift is from manufacturing to retail and logistics as the home of the mega-workforce. It used to be that the biggest employers made things. Now they move things.
Which changes the nature of the work itself. A GM assembly line worker in nineteen seventy-nine had a fundamentally different relationship to the organization than an Amazon warehouse worker today. But we'll get to that.
Let me give you one more number that makes the scale visceral. At one point six million employees, if you assume a typical span of control — one manager for every eight to ten direct reports — you need roughly ten to twelve layers of management just to keep the structure coherent.
Ten to twelve layers.
From the line worker to the CEO, there are ten to twelve people whose job is essentially "coordinate the people below me." That's not bureaucracy as a cultural failing. That's bureaucracy as a structural necessity. You literally cannot run an organization of one point six million people without that many layers. The math doesn't work any other way.
Which means when people complain about bureaucracy in large organizations, they're often complaining about the laws of physics. You can't span-control your way out of it.
And that's the thing — the layers aren't there because someone loves paperwork. They're there because human cognition has limits. One person can effectively manage maybe ten direct reports, maybe fifteen if they're unusually skilled and the work is routine. Beyond that, you're not managing, you're just nodding while people talk at you.
So the record is Indian Railways at one point six to one point seven million. The Western peaks were USPS and GM at around eight hundred thousand. The current private-sector leaders are Walmart and Amazon, both below the historical record. And the organizational structure that makes this possible is, by mathematical necessity, about a dozen layers deep.
That's the empirical picture. But Daniel's real question isn't about the numbers. It's about what it feels like to be inside one of these things. So let's pivot.
Before we do — one thing I want to flag. When you look at the list of largest employers historically, almost all of them are either state-owned or heavily state-tethered. Indian Railways, USPS, the Chinese state enterprises, the militaries. The private-sector giants are a relatively recent phenomenon, and even they — Walmart and Amazon — exist in a regulatory environment that shapes their workforce. The "city-sized employer" was, for most of history, a government project.
The private sector didn't figure out how to manage workforces at this scale until the twentieth century, and even then it was manufacturing firms like GM and Ford that did it, and they were basically copying military organizational structures. The org chart of GM in nineteen fifty looked a lot like the US Army's.
Which maybe explains something about the culture. But okay — let's talk about the lived experience. What's it actually like inside?
The key insight, and I think this is the thing most people get wrong, is that it's not one experience. It's many. A two-hundred-person division inside a one point six million person company feels like a mid-size company. Your daily reality is the people you actually work with — your team, your manager, the three other departments you interact with regularly. The "city" is an abstraction. You don't feel the weight of one point six million people every morning. You feel the weight of maybe a hundred and fifty.
So the "small cog in a very big wheel" framing might be misleading.
It's misleading because it assumes you experience the whole wheel. You don't. You experience your spoke. And your spoke has its own culture, its own norms, its own internal politics, its own lunch routines. The organization is a container, not your daily life.
That's actually reassuring in a way I hadn't expected. The thing that sounds terrifying about a mega-employer — the anonymity, the feeling of being lost in a crowd — might not actually be the day-to-day reality.
Right. The anonymity is real in a different sense. You're anonymous to the CEO, but you were going to be anonymous to the CEO at a five-hundred-person company too. The difference is that at the mega-employer, you're also anonymous to the senior vice president three levels above you who, at a smaller company, might have known your name. The anonymity kicks in at a lower rung. But your immediate world — the people you eat lunch with, the person who approves your time off — that doesn't scale with the total headcount.
So what do these organizations actually get right? Daniel asked us to spend time on that, and I think it's worth taking seriously.
The big one is internal mobility. When you work for an organization of one point six million people, you can change jobs without changing employers. You can move from engineering to operations, from one city to another, from one division to a completely different line of business — and your benefits, your pension accrual, your institutional knowledge, your relationships, none of that resets to zero.
That's the thing the startup world can't offer. If you want a different role at a fifty-person company, your options are "convince the founders to invent a new position" or "leave."
And the benefits at scale are real. Indian Railways runs its own hospitals and schools. GM in its heyday had pensions, health insurance, tuition reimbursement — the whole package — at a level that small companies simply cannot negotiate. When you're buying health insurance for eight hundred thousand people, you get a different deal than when you're buying it for eighty.
The purchasing power of a city.
Literally. And there's another thing these organizations get right that nobody talks about — institutional memory. When you've been around for a hundred years and you've survived wars, depressions, regime changes, technological shifts, you develop a kind of... I don't know, organizational immune system. The institution knows things that no individual employee knows. Processes exist because something went catastrophically wrong in nineteen sixty-three and the fix got baked into the procedure manual and nobody remembers why the procedure exists but it's still preventing the catastrophe.
That's both the strength and the pathology, isn't it? The procedure that saved the company in nineteen sixty-three is now the reason it takes six signatures to order paper clips.
Yes. That's the trade-off. But Daniel's prompt is specifically asking us to look at the upside, and the upside is real. These are organizations built to absorb shocks. A startup dies if its Series A falls through. Indian Railways survived Partition. Literally — the country split in two and the railway kept running.
Let's talk about the longevity question, because that's where Daniel's personal situation comes in. Is the forty-year career dead? And if not, where does it still exist?
The average US employee tenure is around four years. But that number is a mean distorted by a long tail. There's a huge population of workers who stay twenty, thirty years — they're just concentrated in specific sectors. Government, utilities, healthcare, legacy manufacturing. The forty-year career isn't dead. It's just not evenly distributed.
And Indian Railways is the case study in lifetime employment. Generations of families working the same rails. You'd have a grandfather who was a station master, a father who was an engineer, a son who works in signaling. The job is practically hereditary — not officially, but culturally.
That's the world Daniel is asking about. "An organization that expects people to stay for decades." And I think what he's really describing isn't about the work itself. It's about the removal of existential career anxiety.
Say more about that.
The appeal of the forty-year career isn't "I love this job so much I want to do it forever." It's "I don't want to have to think about whether I'll have a job in three years." The apartment-moving chaos Daniel mentioned — that's not just about physical relocation. It's about the cognitive load of permanent uncertainty. The mega-employer removes that load. You know where you'll be. You know what your benefits are. You know the pension math. The future is legible.
That's a phrase I'm going to steal. The future is legible. The startup world makes the future illegible by design — it's all equity vesting cliffs and runway calculations and "we're pivoting to enterprise."
And for some people, that's energizing. For others, it's exhausting. Daniel sounds like he's in the second camp right now, and I think that's a completely rational response to the past few years.
But there's a trade-off here that nobody talks about, and we should name it. The same institutional machinery that provides stability also creates golden handcuffs.
Pension cliffs. You're at nineteen years and eleven months, and if you leave now you lose forty percent of your retirement. So you stay. Not because you want to, but because the math traps you.
And there's a subtler version. The slow atrophy of external marketability. You spend twenty years inside one organization, you become deeply valuable to that organization and increasingly unreadable to everyone else. Your skills are real, but they're expressed in the internal language of the institution. You know how to get things done inside the machine. Outside the machine, you have to translate everything, and some of it doesn't translate.
The city can become a walled garden. Safe, pleasant, well-maintained — and very hard to leave.
Which is why the question isn't "should I join a mega-employer." The question is "which one, and on what terms."
And the modern synthesis is interesting here. Companies like Walmart and Amazon are trying to build the stability of the old mega-employer without the paternalism. Amazon has career choice programs, internal training, paths upward. But the tenure data suggests it's not working the same way. The "city" of the twenty-first century is more transient than the city of the twentieth.
Amazon's median tenure is what, about a year?
Roughly, yes. It varies by role — corporate stays longer than warehouse — but the overall picture is high churn. Walmart's is longer, but still not GM-in-nineteen-seventy-nine long. The old model was "you stay for life, we take care of you." The new model is "you stay as long as it works for both of us, and we provide pathways but not guarantees."
Which might actually be healthier in some ways. The paternalistic model was stable, but it was also infantilizing. You didn't have to think about your career because the company thought about it for you. That's comforting, but it's also a kind of dependency.
And when the company broke that promise — which GM did, which lots of them did — the betrayal was total. People had organized their entire lives around a bargain that got revoked unilaterally.
So the modern version, for all its precarity, might at least be more honest.
Honest, yes. Comforting, no.
Let me pull on a thread from earlier. You said the daily experience inside a mega-employer is your immediate team, not the whole organization. But there's a psychological dimension to knowing you're part of something vast. Does that feel empowering or diminishing?
I think it depends on your temperament. For some people, being part of Indian Railways — knowing you're one of one point six million people keeping the trains running across an entire subcontinent — that's a source of genuine pride. You're part of something larger than yourself, and that's not just corporate slogan. It's a felt reality. The trains actually run. Millions of people actually get where they're going. You contributed to that.
And for others?
For others, it's crushing. Your contribution is so small relative to the whole that it's hard to feel like it matters. You could do your job brilliantly or terribly and the trains would still run. The signal gets lost in the noise.
Which is the existential version of the anonymity problem. Not "nobody knows my name" but "my effort doesn't register."
And that's where management quality matters enormously. A good manager inside a mega-employer makes your work feel consequential even when it's a tiny piece of a vast machine. A bad manager lets the scale of the organization swallow your sense of agency.
The experience isn't determined by the size of the organization. It's determined by the quality of the immediate leadership, same as anywhere else.
Same as anywhere else, but with higher stakes. Because if your manager is bad at a fifty-person startup, you can walk across the room and talk to the CEO. If your manager is bad at Indian Railways, you are ten layers away from anyone who could overrule them.
That's the bureaucracy problem from the inside. The layers that exist for sound structural reasons also function as insulation for bad actors.
The grievance machinery is slow by design. Due process takes time. That's a feature when you're protecting people from arbitrary firing. It's a bug when you're trying to address a toxic manager.
Given all of this — the scale, the stability, the trade-offs, the walled garden problem — what should Daniel actually do with this information? Let's get practical.
First thing: the city-sized employer is worth a serious look, not despite its size but because of it. The internal mobility is real. The stability is real. The benefits at scale are real. The "bureaucracy" that everyone complains about is often just the price of institutional memory — and if what you want is an institution that endures, that price might be worth paying.
Second: don't look at the company-wide numbers. The average tenure at a mega-employer tells you almost nothing, because it's a bimodal distribution. You've got a huge population of short-term workers and a long tail of lifers. Ask about internal transfer rates and average tenure by department. Find out whether the division you'd be joining is part of the churn or part of the long tail.
That's a question you can actually ask in an interview. "What's the average tenure in this specific team?" If they don't know, that's a data point. If they know and it's twelve years, that's a different data point.
Third: if the forty-year career is what you want, be realistic about where it still exists. Government, utilities, healthcare, legacy industrial firms. The tech giants are not the place for it, despite their size. Amazon's scale doesn't translate to Amazon-as-lifetime-employer. The sectors that still do lifetime employment are the ones where the work itself is stable — power grids, water systems, rail networks, hospitals. Things that don't pivot.
Fourth — this is the one I think is most useful — use the city metaphor as a decision tool. When you join a mega-employer, you're not joining a company. You're immigrating. You're becoming a citizen of a polity with its own laws, its own culture, its own internal economy. The question to ask yourself is: do I want to be a citizen, or a visitor?
A visitor takes the job for two years, learns what they can, and leaves. A citizen puts down roots, learns the internal language, builds the relationships that make the institution navigable. Neither is wrong. But you should know which one you are before you sign.
If you're a visitor, the golden handcuffs won't trap you because you're not staying long enough for them to lock. If you're a citizen, the walled garden might actually be what you want — a defined, legible world where you know the rules and the rules don't change every quarter.
The future is legible.
That's the phrase, yeah.
One more thing I want to land. Daniel mentioned the chaos of moving apartments and feeling like life is in flux. That's not a small detail. That's the whole thing. The appeal of the mega-employer isn't really about the employer. It's about the desire for something solid in a world that keeps shifting under your feet.
That desire is probably going to grow, not shrink, in the next decade. AI and automation are going to reshape the largest workforces. The question of what's stable is going to become more urgent, not less.
Which brings us to an open question I want to leave hanging. As automation reshapes these giant workforces — Indian Railways is already automating signaling and ticketing, Amazon is automating warehouses — will the city-sized employer become a historical artifact? Or will the next record-holder be something we can't yet imagine? A company that employs two million people but only fifty thousand of them are human?
That's the version of the city where most of the citizens are machines. Different kind of place entirely.
For Daniel, the practical takeaway is this: the instinct toward stability is smarter than the startup world makes it feel. The forty-year career isn't dead, it's just concentrated in specific sectors. And the question isn't whether to join a giant — it's which giant, and on what terms, and with your eyes open about what you're trading for the legible future.
This has been My Weird Prompts, with thanks to our producer Hilbert Flumingtop. If you've got a weird prompt of your own, send it to show at my weird prompts dot com. We'll be back soon.
See you tomorrow.
