I am well aware of what macro risk categorization is underwriter. Population models are the discrete sums of what boils down to x[n] = (BR-MR)*x[n-1] +/- migration. It feedsback on itself and leads to some interesting output when you chain in multiple populations. There have been all manner of thesis about the lorenz attractor and simple population difference equations.
I understand perfectly what you are trying to say. People around the world are migrating to cities and the difference in birth rates and death rates in the city as opposed to rural areas is going to cause the growth trend to mitigate and thus we will be able to cope. The coping mechanism will include enriching the impoverished because as I should remember only a little bit of wealth creates a proportionately high rise in standard of living.
Through are vaccination and other entrepreneurial ventures my colleagues and I have been able to raise the health standard in urban areas and health has been shown to proceed wealth so various UN en ies should continue to fund us doing so.
My 'underlying assumption' is that various censuses taken lately and the various approximations of population data --and lets keep in mind the population demographics of a place like Ethiopia or Iran are at best guesswork on their part-- have shown themselves to be wrong. Most egregiously has been India where health ministers are saying that they are not doing enough and the predictive models were wrong about rate of change of growth.
For example:
HERE is a study done detailing the inaccuracy and imprecision of projections from 15 years ago. Yes, I understand that the information is 'outdated.' That is part of the Rosling sales job entails. With his new software and dataset they are now getting it right......
The problem is that they are not. India now as it was 15 years ago proves that their modeling is wrong. Do you disagree that the population projections underprojected current growth rates?
Now I don't know if you are a claims handler --or another frontline underwriter-- or something more hands on actuarially. But there is a difference between categorizing risk and correlating said categories to claims made versus creating arbitrary populations and trying to make projection models.
At best, geographical population density models --which is what your 'urbanization' is-- are oversimplified. Also they are not some new mode of approach. They are blending together with poor understanding at the 'forces' at work as most macro models are doing.
My point is that when assholes make claims about how population growth is going to trend, they lack both the precision and the accuracy necessary to make informed policy decisions. They have shown it once again and just because you have some new asshole like Rosling claiming to have transcended 'outdated' thinking, it's bull . They still have no idea what they are talking about and I am not going to bother watching them continue to throw against the wall and maybe one day it will stick. I am not going to bother looking at the categorization du jour over and again.
You want me to look at fertility rates. I want you to look at how they choose their categories and population segmentation. Now if you would like we can talk about macro analysis versus a reductionist approach ie the difference between looking at the population of a Dehli slum or a rural area in Uganda versus a continuous temperature gradient such as used in climate science then great but you acting like I am not understanding simple macro concepts is insulting. If anything your youtube education has brought you no clue about their methodology.
For me, until population models move past poorly understood, tentatively accepted generalizations, I will forgo giving a about the Roslings of the world and their datasets. Soft science indeed.