In light of new data showing China's population shrinking for the second year running, a Shanghai-based think tank has drastically changed its 2100 forecast.
A projection 75 years in the future? Yeah, that’s almost certainly bullshit. Short term predictions already have optimistic and pessimistic paths based on how we as a society react because this isn’t just a physics problem, it’s heavily reliant on culture and policy and technological advancement. And birth rates are even more reliant on culture and policy.
No, they just need to be kept in that context. We trusted science on chlorofluorocarbons impacting the ozone layer, and chose to fix it rather than let it keep going. Was the projection “wrong” because CFCs were regulated, or did we just interact with it in a practical way?
The same applies here. There’s a population issue that (as you mentioned in another comment) without other factors, will come into effect. China can fix it, or let things play out and see if the “unknowns” can fix it for them.
Except birth rates aren’t physics that will progress if left alone, they’re dominated by cultural choices that are impacted by economics and governmental policy. There’s no such thing as “without other factors”, because they’re unable to predict fundamental inputs. What’s China’s economy going to be like in 75 years? How about their food supply under climate change? Is modern day “China” even going to exist? The CCP itself is only 75 years old.
They can’t predict the inputs for 75 years, let alone their feedback into birth choices. It’s just a highly simplified math sim with arbitrary coefficients for the few things they try to model. Pull a different number from your ass to plug into the economy growth box or add a new function to represent widescale automation and you get whatever number you want. You can look at macro birth rate trends for a single country and think “yeah, I could fit a pretty good exponential decay line to that”, except then you look at another country that had the same birth rate in 1950 and the coefficients change. And since it’s exponential those little coefficient tweaks make a big difference 75 years later. In 1950 did anyone have any reason to think that Mongolia’s 75 year birth rate would be twice that of China? Or South Korea’s would be 60 percent?
The thing about long-term predictions (at least ones that get publicity) is that usually the goal is to change them, so few have been “proven”. No one is printing stories about how an isolated set of rocks is going to be decayed by X% due to weather, because no one cares.
Except birth rates aren’t physics that will progress if left alone, they’re dominated by cultural choices that are impacted by economics and governmental policy.
Exactly. Those are the factors that are being considered when making these predictions. If economic factors and policies are making it harder to have kids, then birth rates drop, which is what we’re seeing now. What else is going to have as much of an effect?
These predictions don’t exist to take bets on. They’re not scrying into the future. They’re just binoculars that point to where we’re going.
“Hey, here’s a possible future. Not the most likely or even the most accurate to some imagined neutral policy position, just one potential future. Or maybe even not a potential future if we’re missing a key impact.” It’s all bullshit man, with practically zero prescriptive value. One of the broadly assumed core components of birthing decisions, economics, is almost unpredictable even 20 years out, let alone 75. The simulation-based social “sciences” are just prettied-up hunches and guesswork in anything but the shortest of terms.
But I agree that they aren’t meant to be proven, because that’s a very convenient space to work in when your methodology is “I made a guess about the coefficient” combined with “what if complicated things were actually simple”. Garbage in, garbage out.
You sound really sure about your understanding of statistics and probability, and I don’t think anything I can say can impact that. I’m going to defer to the experts, but you do you I guess.
It’s not a matter of doing math well, it’s that your unknowns destroy any prediction you’re making. If you’re doing it correctly you’re expanding your error bars as you get further into the future. By 75 years out you’re all error.
That’s going to be an issue in 10-20 years. Who the fuck knows what it’s going to be like 75 years from now. We’re talking about a span of time as long as Communist China has existed. 75 years ago computers barely existed.
Social sciences and hard sciences are not the same thing. Social “sciences” are largely unfalsifiable and dominated by the “unless modified by human action” caveat that breaks predictions. You can predict very general trends like “fewer births”, but in 1949 you couldn’t look at the populations of Mongolia, China, and South Korea and know that they would have vastly different birth rates 75 years later.
Anyone projecting 75 years into the future is just making things up and doesn’t need to be covered as a news story.
Do you deny climate change on the same basis?
A projection 75 years in the future? Yeah, that’s almost certainly bullshit. Short term predictions already have optimistic and pessimistic paths based on how we as a society react because this isn’t just a physics problem, it’s heavily reliant on culture and policy and technological advancement. And birth rates are even more reliant on culture and policy.
No, they just need to be kept in that context. We trusted science on chlorofluorocarbons impacting the ozone layer, and chose to fix it rather than let it keep going. Was the projection “wrong” because CFCs were regulated, or did we just interact with it in a practical way?
The same applies here. There’s a population issue that (as you mentioned in another comment) without other factors, will come into effect. China can fix it, or let things play out and see if the “unknowns” can fix it for them.
Except birth rates aren’t physics that will progress if left alone, they’re dominated by cultural choices that are impacted by economics and governmental policy. There’s no such thing as “without other factors”, because they’re unable to predict fundamental inputs. What’s China’s economy going to be like in 75 years? How about their food supply under climate change? Is modern day “China” even going to exist? The CCP itself is only 75 years old.
They can’t predict the inputs for 75 years, let alone their feedback into birth choices. It’s just a highly simplified math sim with arbitrary coefficients for the few things they try to model. Pull a different number from your ass to plug into the economy growth box or add a new function to represent widescale automation and you get whatever number you want. You can look at macro birth rate trends for a single country and think “yeah, I could fit a pretty good exponential decay line to that”, except then you look at another country that had the same birth rate in 1950 and the coefficients change. And since it’s exponential those little coefficient tweaks make a big difference 75 years later. In 1950 did anyone have any reason to think that Mongolia’s 75 year birth rate would be twice that of China? Or South Korea’s would be 60 percent?
The thing about long-term predictions (at least ones that get publicity) is that usually the goal is to change them, so few have been “proven”. No one is printing stories about how an isolated set of rocks is going to be decayed by X% due to weather, because no one cares.
Exactly. Those are the factors that are being considered when making these predictions. If economic factors and policies are making it harder to have kids, then birth rates drop, which is what we’re seeing now. What else is going to have as much of an effect?
These predictions don’t exist to take bets on. They’re not scrying into the future. They’re just binoculars that point to where we’re going.
“Hey, here’s a possible future. Not the most likely or even the most accurate to some imagined neutral policy position, just one potential future. Or maybe even not a potential future if we’re missing a key impact.” It’s all bullshit man, with practically zero prescriptive value. One of the broadly assumed core components of birthing decisions, economics, is almost unpredictable even 20 years out, let alone 75. The simulation-based social “sciences” are just prettied-up hunches and guesswork in anything but the shortest of terms.
But I agree that they aren’t meant to be proven, because that’s a very convenient space to work in when your methodology is “I made a guess about the coefficient” combined with “what if complicated things were actually simple”. Garbage in, garbage out.
You sound really sure about your understanding of statistics and probability, and I don’t think anything I can say can impact that. I’m going to defer to the experts, but you do you I guess.
Time series forecasting is a pretty well-established statistical method though.
It’s not a matter of doing math well, it’s that your unknowns destroy any prediction you’re making. If you’re doing it correctly you’re expanding your error bars as you get further into the future. By 75 years out you’re all error.
Removed by mod
That’s going to be an issue in 10-20 years. Who the fuck knows what it’s going to be like 75 years from now. We’re talking about a span of time as long as Communist China has existed. 75 years ago computers barely existed.
Removed by mod
Social sciences and hard sciences are not the same thing. Social “sciences” are largely unfalsifiable and dominated by the “unless modified by human action” caveat that breaks predictions. You can predict very general trends like “fewer births”, but in 1949 you couldn’t look at the populations of Mongolia, China, and South Korea and know that they would have vastly different birth rates 75 years later.
U mad?