4 ms·
So, I work with a lot of population data for worldbuilding/stats/curiosity [0]. The number one thing when modeling really anything interesting about a setting
by Balgair 2y ago
So, I work with a lot of population data for worldbuilding/stats/curiosity [0].
The number one thing when modeling really anything interesting about a setting or story is (drum roll): Population
Pretty much everything is keyed to population [1], and I mean pretty much everything. The number of movie theaters, the tonnage of your aircraft carriers, the walking speed in your downtowns, it's all about population.
Generally, I've seen that as population rises, things get better. There are more patents per capita, better ratios of MDs to patients, more ranks in your Fire departments, etc. Yeah, there are exceptions here, but not all that many.
Most things get more efficient as you get more people [2]. Your gas stations have more customers per day, your sewer lines handle more people, your libraries have more patrons per day, etc.
So, when I see a lot of discourse that we somehow need to allow for less people to be alive, to encourage less births, well, that makes me a bit sad. Yes, yes, climate change, I agree, we need to desperately fix that. But from the data that I work with, all I see is that more people = good [3].
To me, the data suggests that we don't need to have less people, that would lead to a less vibrant and progressing world. What we need to figure out a way to grow our population without the effects on the rest of the world. And yeah, I know, impossible. But the data tells me that shrinking is not the way. Somehow, I feel that technology is really the best way forward.
[0] Honestly, I'm a weird nerd that really just likes to play with excel sheets, graphs, and wikipedia info boxes.
[1] the only thing that isn't, in my meanderings through BLS data sheets, wikipedia, etc is the number of stops on any particular subway/rail line. That's most closely correlated to population density. And, really, that's pretty much it. Just that one thing.
[2] again, not population density, it's easy enough to control for that, it's just raw numbers of people from what I have seen.
[3] again, generally. Not all cases, not all places, not all times.
- paulryanrogers 2y agoCould it also be the data is better when there are more people? So the accuracy is lower is sparser areas? Which could hide advantages to sparse living?
- Balgair 2y agoNot that I have seen. The US's BLS data is quite comprehensive. You can drill down to individual zip codes, neighborhoods, etc, some of which have very low population densities and generally low populations. I know that the French also have such detailed lists, but as my French is very poor, I'm less familiar with that data. Commonwealth countries generally are lacking in comparison in the detail of the population data they freely provide. Something about francophone countries and those like the US with a long philosophical interchange with France seems to be more 'free data' centric. Still, based on my work with such data, I don't see that a lack of data in more rural areas has much affect on the 'goodness' of things. From my work I see that it really mostly is just more people = more better.