Comment by A_D_E_P_T | Hacker News Reader

Comment by A_D_E_P_T | Hacker News Reader

Parent Full thread

I'm probably going to get in trouble for this, but the population numbers and statistics for Africa are totally unreliable. Fertility and total population are all wrong.

The DRC is said to have 100M people, but check out satellite imaging. There's no chance -- and I mean none -- that it actually has 100M people. Unless 9-out-of-10 inhabitants live in the woods under tree cover, the actual population of the country is probably closer to 10M.

You don't have to take my word for it. Look for yourselves. And take an satellite shot of Kinshasa (reported population ~19M), rotate or mirror-image it, and then ask GPT-5 to estimate its population. Also, compare for yourself vs. a place like Shanghai. (Reportedly just 20% more populous, but also visibly denser and roughly an order of magnitude larger.)

Many other countries in the region, like Nigeria, are much the same way. The population numbers don't line up with satellite imaging.

Then there are obvious economic measures, etc.

The unavoidable conclusion is that the numbers for Africa are maximally unreliable. There are various reasons for this that we can speculate on (foreign aid dependent on population numbers, etc.), but, anyway, at least take 'em with a grain of salt.

Zoom in on Kinshasa though and you can see that it's almost entirely very densely-packed slums and shantytowns built up against each other, about 100 - 250 square meters in size, with no gardens or back yards. Slums can be very dense, Dharavi in Mumbai 2.3 square KM with a population of about a million. Manhattan has a much lower population density now than 100 years ago.

Also in all the street view pictures it looks absolutely packed - every road is gridlocked with people everywhere, but Shanghai has a lot of empty space for people despite its size. Roads have trees, they're much wider, there are a lot of open parks, office buildings etc that Kinshasa wouldn't have.

I drove right around Africa through 35 countries over three years. I drove across both Nigeria and the DRC.

There are dozens and dozens of massive cities that take hours to cross in Nigeria you’ve never heard of. Anecdotally, it’s way, way, way more populous than anything nearby. Ethiopia felt somewhat similar in parts, as did Egypt.

Can you name a few of them in Nigeria? On satellite imaging, from what I've seen, they're not so massive, and they're mostly comprised of a sprawl of 1-3 story buildings.

We can compare vs. cities that we have good numbers for. Or Chinese/Indian cities, for that matter. (After looking at Nigeria or the DRC, a quick glance at India via satellite imaging is shocking.)

That said, Egypt is very populous, there's no doubt about that one.

I drove through at least 10 cities in Nigeria I’ve never heard of that had tons of buildings over 10 stories. I just took the fastest route across, I didn’t go wandering. This was 10 years ago too.

Also remember the DRC is almost a million square miles. So it’s 1.5x Alaska.

Sigh.

Personal experience Vs. looking at pictures from the basement....

I expect it to be USA policy soon.

One is anecdotal, one is data.

Is it weird that I can't tell which one? Guangdong has 127 million people and its just the size of Oklahoma. Ok, the Chinese government could be lying about that, but I seriously doubt it.

Thank you for this. Nigerian here, and I have to say that you hit the nail squarely on the head. Population counts in Nigeria are deeply political and essentially every region/state is motivated to fake/overestimate their headcount to get a bigger chunk of the oil revenue, which is pretty much the most significant slice of gov. revenue.

But, once you dig into the figures, you realize it'd be a miracle if Nigeria has up to half of the population it claims.

Every single census that's been conducted has been marred by controversy, with states trying to buff up their populations to make their ethnicity/region look bigger and more important.

But, proxies like registered BVN (like Social Security Number, but for bank accounts) are just under 70M. Registered phone lines (~240M; each person usually owns 3-5) are similarly lackluster. Domestic demand is nothing to write home about if you run a CPG business. Zoom into a satellite view of a city that's supposed to have ~700k to 1M people and it looks like a suburb - just scanty.

Nigeria's most populous city claims to have 20M people - 2* the population of Seoul, one of the most urbanized, dense, vertical cities in the world, meanwhile, Lagos is just a sprawling slum.

Personally, I feel population counts across Africa are grossly overestimated. A good estimate would be 600-800M, but where's the fun in that when we can fearmonger about overpopulation?

Are there any proxy for population count? Like may be mobile users, I understand people use more than one.

Yes, I included that in the brackets. There's roughly ~240M registered SIMs, but unlike American phone line where you sign a contract, Nigerian SIMs are bought OTC and are quite disposable. But, they're tied to the National Identity Number which is biometric and hard/impossible to circumvent.

It's quite common to have up to 2-3-4 SIMs also. When when you account for that, you have ~60M-80M-100M phone users.

Another proxy is the Bank verification Number, which like I said, is Nigeria's Social Security Number, but for banking. Can't have a bank account without it. Even in the super-rural parts of Northern Nigeria, they're still banked, which is impossible without a BVN.

Now, there's been ~66M BVNs issued so far, according to a recent update. If we assume that covers the bulk of the adult population (remember, you can do literally nothing without a BVN) and several million teenagers, and account for the median age being 19 (that is, half the entire population is beneath that age), we can infer that Nigeria's population is close to 2* the BVN count.

My best guess is 120M to 140M max given the measurable, unfalsifiable proxies.

Hey, thanks for the inspiration de look here.

I'm strangely bullish on Nigeria. The population stats and the work y'all are doing is just amazing to me. The EkoAltantic project is so big but under the radar that I love it. I keep trying to tell my management that we need to invest in West Africa today to build up brand loyalty now.

What are you thoughts on Nigeria's future in the next 50 years or so?

This is fascinating, thank you for bringing this to the discussion!

Even better personal experience

You estimate a factor of two for over estimate, believable, but higher than I would have thought. Makes sense with the context of getting more revenue. Thank you

I caution you against the "I looked at satellite image" nonsense. Your lived experience is very valuable. Looking at a satellite image is not

I just did what you suggested and looked at Kinshasa and it looks HUGE to me. Endless rows of streets. Reminded me of Mexico City (and I just compared satellite imagery and they seem similar in that respect - and they are similar in terms of population). Big cities with few skyscrapers and tons of low rise buildings are quite common. In the west, Paris is an example of a populous city without many tall buildings.

Kinshasa's urban area (not the entire province) is generously estimated at 600 square km.

Mexico City is 1485 square km.

Mexico City's population, within that 1485 km^2 envelope, is most commonly given as ~9.5M. (It's ~20M with satellite towns and regions.)

And then consider that most of Kinshasa is comprised of buildings that are very low to the ground -- far lower than in Paris -- whereas Mexico City is in places very dense.

The arrogance of a comment like this is staggering. So you looked at some satellite photos and talked to ChatGPT and now you’re an expert on the demographics of a city? Crazy.

It's quite possible that China's population has also been significantly exaggerated. One thing that the China actually does quite well is universal childhood vaccination for certain diseases. Independent researchers looked at the number of vaccine doses ordered and found that it's way lower than the official government birth statistics would suggest.

https://www.pewresearch.org/short-reads/2022/12/05/key-facts...

Now this is an unproven "conspiracy theory" but it's entirely plausible that corrupt local officials have inflated population numbers in order to be able to embezzle more of the funds that flow in from the central government. A lot of people might exist only on paper, but it's impossible for outsiders to precisely quantify.

Reply on news.ycombinator.com