[jdietrich]

>"When dams are built, large areas are flooded and people need to be relocated," Láng-Ritter said in a press statement. "The relocated population is usually counted precisely because dam companies pay compensation to those affected."

Sure, "we've been systematically undercounting population for decades" is a more plausible explanation than "large infrastructure projects in rural areas of underdeveloped countries are a bonanza of corruption".

[zamadatix]

Concerningly, they acknowledge accuracy of the numbers reported by the dam projects under limitations... by only highlighting the opposite, that it could be under-reported by them. I feel like I'm missing a lot for this to have been published but I'd expect such a paper about limitations of existing studies to be especially heavy on what the limitations of this new method might be.

https://www.nature.com/articles/s41467-025-56906-7#:~:text=L...

[MichaelZuo]

I’ve actually looked into this for a few developing countries, and dam resettlement figures are very likely to be inflated.

There’s no actual combination of outsiders going around to every household in every village for even a dozen days of the year to plausibly provide a third party confirmation of claimed residence for each individual.

It’s effectively villagers certifying each other that they really live there as a primary residence.

[more-nitor]

> It’s effectively villagers certifying each other that they really live there as a primary residence.

this. some smart asses would be calling all their relatives, distant-relatives, friends, etc AND claim that they all lived there. then, share the loot(?) like 50%, 30%, 20% etc

the more better approach would be compensating by property, though does have its own downsides

[epolanski]

My aunt was relocated from a dam construction and they started the census/analysis way before the project was even approved, because counting how many people you need to relocate and compensate is actually a part of the decision process.

They don't do the opposite: we build here, now let's see how many people we need to relocate.

[lurk2]

One of the more frightening things I ever learned was that if the population of China began walking in front of you at a rate of 1 person per second, it would take approximately 31 years for the entire population to pass you.

The math, roughly:

1 billion people (there are more now, but we’ll call it an even billion) / 60 seconds ≈ 16,666,666 minutes

16,666,666 / 60 minutes in an hour ≈ 277,777 hours

277,777 hours / 24 hours in a day ≈ 11,574 days

11,574 / 365 days in a year ≈ 31 years.

That’s just China. The global population would take around 220 years, with not so much time as to say hello. A bureaucracy can of course delegate census reporting such that the groups become manageable, but it puts a lot of things into perspective when you understand the sheer scale of the human population.

[sidewndr46]

How is this scary? It's like observing that I don't have enough time to write a comprehensive treatise on every German speaking philosopher of the 19th and 20th century. I obviously don't, I probably could barely complete even one at this point. My life is not any better or worse due to this.

[lurk2]

Assuming you had an interest in writing a comprehensive treatise on German philosophy, not being able to complete it in one lifetime would strike many people as frightening. It’s one of those numbers that forces you to consider your own mortality, like those charts showing how many weeks there are left assuming you live to be 80 years old.

[nradov]

Frightening how? I'll probably die in about 2000 weeks plus or minus. So what.

[VincentEvans]

Using your own contrived math - if you have just 30 helpers - you can count population of China in 1 year. Or with 219 helpers you can do the same for world population.

In US, for the 2020 decennial census, the U.S. Census Bureau hired approximately 500,000 temporary workers across the country to assist with the count.

[epolanski]

> In US, for the 2020 decennial census, the U.S. Census Bureau hired approximately 500,000 temporary workers across the country to assist with the count.

Thus..it's difficult and expensive.

[triknomeister]

US spent 13.7B on census which is 1.37B/year since a census is done per 10 year. The ROI is definitely huge. It's not expensive at all.

[kristjansson]

A dollar per dead snake, you say?

[lurk2]

I saw this happen on a Minecraft server. It was an economy simulator based out of Sweden or Denmark. The administrator was committed to Keynesian economics. One day he wants to clear a desert to build a new town. To compensate people for working there, he used a plug-in called Shop Chests, allowing users who deposited 64 sand into the chest to receive 100 units of the server’s currency. This was substantially more than the sand was worth on the open market. Since this guy was a Keynesian, he had also built a government marketplace that bought and sold all the major blocks. Rather than spending the afternoon digging, I warped to the market, bought out the entire supply of sand, warped back to the job site, and sold the sand at the higher price.

This kind of thing happened a lot with absentee owners who would set prices for their shops lower or higher than the market price. If the owner had set up hoppers underneath the shop chest, you could effectively bankrupt him overnight. It happened a lot with minecarts, diamonds, colored wool, and things like that.

[kristjansson]

Which is why the likes of Eve Online employ real-life economists - is easy to get wrong.

Cobras aren’t really an arbitrage problem, except in the sense that the reward was (presumably) more than the cost of breeding them.

[soulofmischief]

History often rhymes because we clearly aren't teaching it effectively enough. Every six-year-old schoolkid should understand the Cobra Effect (perverse incentive) and how to recognize it around them and in their own thinking.

[pessimizer]

Are you saying that large infrastructure projects have been systematically overcounting population in order to pay more compensation? That there's an interest in making them seem more disruptive, and as if they displace more people?

If that's an obvious conclusion, I need more explanation.

[solid_fuel]

It doesn't need to be top-down corruption on the part of Dam Corp, LLC.

Think about it this way - you and your family live somewhere and are being displaced to make way for a dam, some guy in a suit comes around and says "we'll pay $1000 per resident to move you somewhere else".

Maybe your uncle lived at your house with his wife for 5 years, until they moved to the city last year. Your grandmother lived there until she died 6 months ago. So lets say it's just you, your partner, and 2 kids.

But, that's a lot of money - do you tell the man in the suit that your house has 4 residents, or maybe stretch the truth to 5 or 6 (your uncle might move back soon, after all)?

And remember, corruption often stacks - individuals might add an extra person here and there, but then the local relocation manager adds a few % to get a little extra on top, and their boss adds another few %, and so on... soon you're seeing 25% more people than actually reside there.

[Muromec]

Chekhov, "Dead souls". The scam can go both ways, you need to be local to know how it works and not benefit from it enough to snitch.

[galkk]

Gogol, not Chekhov

https://en.m.wikipedia.org/wiki/Dead_Souls

[Muromec]

I always mix those two. If it's on the funny side it's usually Chekhov

[hattmall]

My high school was Ft. Knox, armed guards, metal detectors, ID must be worn and visible at all times. No backpacks, waist band exposed. Class doors automatically lock when the bell rang, anyone in the hall would be escorted out. No hall passes, no bathroom breaks, cameras everywhere.

Except for 1 week. We had basketball tournaments and activities, movies, pizza parties, free ice cream. Wear whatever you want. No ID enforcement.

That was the week they counted attendance to determine how much funding the school got. They even bussed the "bad" kids in from the satellite campus.

[gausswho]

This is so foreign to me, as a forty-something. Why do we let schools take security theatre to such tragicomedy. It likely breeds more killing than it prevents.

[hattmall]

This wasn't so much theater. Even with all that we still had shootings, just mainly in the parking lot. This was well before Columbine and after every one joked how it wouldn't have happened here because too many students would have returned fire.

All of this security was in response to gangs and general violence of the late 80s / 90s.

[allturtles]

The issue is that the availability of money gives people an incentive to be counted, even if they don't live there.

[00N8]

No, I think the conjecture is more that people from the surrounding areas could be claiming residency in the affected area to receive the payouts, even though they normally live elsewhere.

[AnotherGoodName]

The way i see it people really do often live in multiple locations at once and bureaucracy has a hard time with that fact. Think of the itinerant workers travelling around for seasonal jobs. The census might count them in one location but it’s quite reasonable they count for the relocation too.

So it’s possibly a matter of edge cases in the wheels of bureaucracy than outright graft.

[Muromec]

Likely the manager who makes the payouts happen and not people themselves

[kmeisthax]

No, it's the people being resettled who have an incentive to overcount themselves.