3 ms·

This is a really interesting visual presentation of the data. Some cool observations: 1. The first five curves (1950–1990) are all in 10-year increments, so fo

by rlue 7y ago

This is a really interesting visual presentation of the data. Some cool observations: 1. The first five curves (1950–1990) are all in 10-year increments, so for instance, 1950's 20-year-old population maps very directly to 1960's 30-year-old population (hence the "rippling" effect produced by those six curves). 2. What's with the sudden pinch in the population of 20-year-olds in 2018? It tapers in very heavily from the corresponding point on the 1990 curve, and apart from the newborn population, is the only local maximum anywhere on the chart. Did I miss something in the news? Why would such a large number of kids born in 1990 be dead by now, compared to previous generations? 3. Why do curves for previous years have such pronounced staggering, while projections for the future are comparatively much smoother?

dan-robertson 7y ago
I’m also curious about 2. From my reading it looks more like these children weren’t born at all rather than that they all died. My best guess is something related to the collapse of the Soviet Union caused a decrease in birth rates.
dan-robertson 7y ago
For 3. I would guess that the “staggering” happens due to big geopolitical events causing significant changes to global birth rates, and the future lines do not show that because they can’t predict these events happening in the future, so they just try to give an average guess.
lordnacho 7y ago
> What's with the sudden pinch in the population of 20-year-olds in 2018? I'm no expert on this but I would guess the peak of China's one-child policy? Someone more knowledgeable might know more about the history of specific exemptions (city vs rural) and such. There's also implications for reporting bias, which might explain why the older data doesn't look so peaky.
ourworldindata 7y ago
Hi rlue! This is largely due to China's population dynamics. Here I've made a chart that shows the number of births (and deaths) in China over the last 69 years: https://ourworldindata.org/grapher/births-and-deaths-projected-to-2100?time=1950..2019&country=CHN https://ourworldindata.org/grapher/births-and-deaths-project... 30 and also 50 years ago China had large cohorts of newborns, these are the 'ripples' you see going through the global population pyramid. The number of births is always determined by two factors: the number of children per woman in the reproductive age bracket (called the Total Fertility Rate). And the number in the reproductive age bracket. Here is the Total Fertility Rate for China: https://ourworldindata.org/grapher/children-per-woman-un?tab=chart&time=1950..2015&country=OWID_WRL+CHN https://ourworldindata.org/grapher/children-per-woman-un?tab... The large cohort born around 50 years ago is the parent generation of the large cohort born 30 years ago. So the question is what explains the large cohort in the 1960s? That is the cohort born right before the rapid reduction of the Total Fertility Rate (from 6.4 children per woman in 1965 to half of that 15 years later as the chart above shows; this is largely before the introduction of the strict one-child-policy by the way). And it is the cohort born after one of the largest mass deaths in history. The Great Leap Forward Famine from around 1951 to 61 https://ourworldindata.org/famines#great-leap-forward-famine-china-1959-61 https://ourworldindata.org/famines#great-leap-forward-famine... Around 30 million died and as in all famines the fertility rate declined substantially, to then jump again right after the famine. Good questions! I should have mentioned them in the post. Let me know if you want to know more.
- rlue 7y ago
  Wow, wasn't expecting a response! Taken together, the data and the historical context tell a really powerful story. So the chosen UN projections for this graph don't stagger like the historic data because they assume no famines or other high-death-toll events in the given period then? On the one hand, I imagine we must have a big disaster coming our way eventually. On the other, if you factored it into your projections, you'd have to guess at the approximate time and magnitude of such an event, which I'm sure is not the business of whoever's making these projections.

by rlue 7y ago

dan-robertson 7y ago

I’m also curious about 2. From my reading it looks more like these children weren’t born at all rather than that they all died. My best guess is something related to the collapse of the Soviet Union caused a decrease in birth rates.

dan-robertson 7y ago

For 3. I would guess that the “staggering” happens due to big geopolitical events causing significant changes to global birth rates, and the future lines do not show that because they can’t predict these events happening in the future, so they just try to give an average guess.

lordnacho 7y ago

> What's with the sudden pinch in the population of 20-year-olds in 2018? I'm no expert on this but I would guess the peak of China's one-child policy? Someone more knowledgeable might know more about the history of specific exemptions (city vs rural) and such. There's also implications for reporting bias, which might explain why the older data doesn't look so peaky.

ourworldindata 7y ago

Hi rlue! This is largely due to China's population dynamics. Here I've made a chart that shows the number of births (and deaths) in China over the last 69 years: https://ourworldindata.org/grapher/births-and-deaths-projected-to-2100?time=1950..2019&country=CHN https://ourworldindata.org/grapher/births-and-deaths-project... 30 and also 50 years ago China had large cohorts of newborns, these are the 'ripples' you see going through the global population pyramid. The number of births is always determined by two factors: the number of children per woman in the reproductive age bracket (called the Total Fertility Rate). And the number in the reproductive age bracket. Here is the Total Fertility Rate for China: https://ourworldindata.org/grapher/children-per-woman-un?tab=chart&time=1950..2015&country=OWID_WRL+CHN https://ourworldindata.org/grapher/children-per-woman-un?tab... The large cohort born around 50 years ago is the parent generation of the large cohort born 30 years ago. So the question is what explains the large cohort in the 1960s? That is the cohort born right before the rapid reduction of the Total Fertility Rate (from 6.4 children per woman in 1965 to half of that 15 years later as the chart above shows; this is largely before the introduction of the strict one-child-policy by the way). And it is the cohort born after one of the largest mass deaths in history. The Great Leap Forward Famine from around 1951 to 61 https://ourworldindata.org/famines#great-leap-forward-famine-china-1959-61 https://ourworldindata.org/famines#great-leap-forward-famine... Around 30 million died and as in all famines the fertility rate declined substantially, to then jump again right after the famine. Good questions! I should have mentioned them in the post. Let me know if you want to know more.

rlue 7y ago
Wow, wasn't expecting a response! Taken together, the data and the historical context tell a really powerful story. So the chosen UN projections for this graph don't stagger like the historic data because they assume no famines or other high-death-toll events in the given period then? On the one hand, I imagine we must have a big disaster coming our way eventually. On the other, if you factored it into your projections, you'd have to guess at the approximate time and magnitude of such an event, which I'm sure is not the business of whoever's making these projections.