A nice graphical presentation of long-term trends.
25 thoughts on “In Case You Want To Feel Optimistic”
Pretty neat, eh?
Very.
I’d like to see the same evaluation of the US states.
It’s interesting that wealth is on a logarithmic scale but life expectancy on a linear scale.
Did anyone else notice the big drop in life expectancy China experienced during Mao’s Cultural Revolution? Socialism kills. I’d like to see an all-countries comparison with colors designating economic policies instead of geography.
Carl, if Aubrey De Gray gets his way I think it will be linear wealth that buys log life gains.
Uh? I don’t get that, Brock. Could you explain?
But if anything, this graph tells you how hard it is to translate wealth into lifespan. Everyone got 100x richer but only lives 3x longer! And at that, the gains in lifespan are somewhat misleading, in that the increases in life expectancy at birth are nearly all the result of young people not dying of infectious disease, and mothers not dying in childbirth.
For the real gains in ultimate longevity you have to look not at life expectancy at birth, where things like birth defects, accidents and infections take a major toll, but at something like life expectancy at age 65, where death is dominated by the ultimate killers of old age like heart disease and cancer.
Between 1900 and 2010 the US real (inflation-adjusted) GDP per capita increased by a factor of 25, but life expectancy at age 65 went up from about 78 to 85, i.e. one might say longevity increased by less than 10%.
It’s a fascinating question, I think, why this is. Why should spectacular gains in wealth be so much easier than gains in lifespan? One might argue there is something about wealth creation that allows for what Nassim Taleb might call “positive Black Swans:” it’s possible for accidental discoveries — semiconductors? Internet? thermodynamics? — to lead to spectacular gains in labor efficiency, hence wealth.
But longevity seems strangely proof against this. We have made many advances in medicine, but apparently essentially none of them are “Black Swans,” meaning there has never been a discovery that spectacularly increased lifespan. Isn’t that curious?
Mr. Pham;
I would suggest it indicates the effects of evolution, in that various systems in the human body all wear out at about the same time. Clearly it’s a waste (evolutionarily speaking) for some subsystem to last significantly longer — you could “cheapen” it without cost and boost some other aspect. As a result, to significantly increase lifespan, we must overcome systemic failure. Even “black swans” don’t really help because some other part wears out anyway.
If wealth increases faster than lifespan, at least we are better able to enjoy the time we do have.
It’s amusing that an owl is a symbol of wisdom. It so well matched to its environment that it hardly needs a brain. The wisdom is in that design.
So what about the human brain? We think pretty highly of ourselves, but in fact some animals far surpass us in certain aspects of the brain. Chimpanzees have much better memory, beyond any human, in some respects. Dolphins and bats process sonar. Wouldn’t super human abilities give us better survival to outperform the other tribes? Neanderthals had bigger brains (were they smarter?) and stronger muscles than our ancestors.
So I’m not sure what assumptions you can make about subsystems relating to life expectancy. The fact is the highest age of modern man has been roughly constant regardless of medicine.
The bounces were the intriguing part. It would be interesting to match those with the causes. It would seem the abruptness of change would help to discover the causes. I also note the size of population of China and America and their relative growth.
AOG, your theory holds the unlikely assumption that all the subsystems have by nature, and started off with, the same lifespan. For example. consider building a table out of the scraps of lumber and hardware you have in the garage. A priori, the various subsystems of your table will have substantially different lifetimes — there will be something that wears out first, and something else that lasts much different. Why should it be different for biological machines?
You could then set about systematically improving your table, of course, just as evolution systematically improved the lifespan and efficiency of species after they arise. But why would this lead to a convergance of the lifespans of the subsystems? Say you replace your cheap mild steel bolts, which rusted first, with good marine stainless steel bolts. It’s less likely that this brings the lifespan of the bolts in line with the wood, than that now the life of the bolts leaps ahead of the wood, preserving the heterogeneity of subsystem lifespans. I would think the same would happen to biological machines.
In short, if all our subsystems seem to fail at once — and they do — I would suspect a design lifetime. That is, evolution has served not to continuously and simply increase lifespan and resilience, but to actually cut us off at a certain maximum age. We are, in short, designed to die, it would seem to me, and at fairly definite age (or rather at some fixed ratio of age to environmental rigor).
This is hopeful news, however, because if we die not because our various systems just wear out, higgledy-piggledy, but because the Master Switch clicks to OFF, it seems more likely we can figure out how to control the Master Switch ourselves.
ken, I believe the largest bounce is the 1918 influenza epidemic.
During the 200 year period you are talking about the dollar was stable for the first half and inflating rapidly for the second half, so it is a bad yardstick for measuring wealth. The first 100 years $400 could buy 20 oz of gold and $4,000 could buy you 200 oz of gold. Today $40,000 buys you about 28 oz of gold. Prices of everything else inflated in the second 100 years also. The first 100 years people really got richer but during the second 100 years it is mostly that dollars are worth less.
What I notice is that countries are advancing regardless of their economic theories, and that partially socialistic countries, such as modern Europe are doing fine.
This analysis may indicate that technological evolution is more important to human progress than the notions of economists and the ideologies of governing political parties (absent something as toxic as Mao’s China or Pol Pot, of course).
Rand, thanks for highlighting this video. Fascinating!
Mr. Pham;
Because biological systems are much finer grained than “cheap steel” vs. “marine steel”. So the table builder could pick to use bolts that are just good enough to last as long as the cheapest component. The table builder is selecting from pre-existing units and therefore can’t match things as well. In contrast, a biological system is completely autarkic. If the table builder had to smelt and cast his own bolts, he’d have more incentive to make them just good enough.
My theory does rest on the assumption that there is some cost to more durable subsystems. Therefore creatures who subsystems were just good enough would have an advantage.
Annoying Old Guy – Precisely. To illustrate the point, Henry Ford once commissioned a study of the mode of failure of Model T Fords being scrapped. Apparently, he discovered that kingpins (IIRC), whatever the heck they are, virtually never failed causing the breakdown of the car – and reduced specs because of that finding. No point in having one part in pristine condition when the whole machine is being scrapped.
The situation in a living organism is a bit different, as living organisms self-repair. Unfortunately, this repair takes energy and materials. It is notable that most wild animals live just long enough to get their young to reproductive age. Humans are an exception; one theory of the reason is that if granny is around to look after the kids (as opposed to infants) then both parents can get out there obtaining food. Also, older humans can pass on their knowledge to the young – which most animals can’t do nearly as well.
Another illustration; imagine a mouse that lives to 30 (unless eaten by a predator) but because of its highly efficient repairs only has two pups instead of 12. Unfortunately, mice are often eaten by predators – and if all their offspring have also been eaten then that genetic line is no more.
If a genetic mutation’s adverse effects are unlikely to strike before something else has killed the individual off, there is little selective pressure to clean out such mutations. They could thus be expected to accumulate so long as environmental factors remain steady. So, as we learn to cure or prevent one class of disease, another that was rarely a problem before comes into prominence at old age.
Because biological systems are much finer grained than “cheap steel” vs. “marine steel”.
I don’t think that matters. If it did, then the far more fine-grained and multitudinous subsystems of a computer would fail more often together, at the same time, than the fewer and simpler subsystems of a lawnmower or car. But that’s not what happens. The phenomenon is likely to be scale-invariant.
In contrast, a biological system is completely autarkic.
Not at all true. In fact, biological systems are far less likely to construct new parts de novo. As a rule, evolution only allows gradual change, which means parts intended to perform new functions must be jury-rigged together by re-purposing older parts that do something different.
Indeed, this has always been one of the major challenges to evolutionary theory: how, for example, did wings evolve? Flying subsystems are sufficiently complex that it’s inconceivable they could have arisen all at once, in one generation (parents walk, children can fly), by some incredibly rare combination of lucky mutations. But on the other hand, of what advantage would some slight movement towards wings — winglets? — give the mutant who had them? And if they don’t confer an advantage, they won’t be adapted. The usual answer is that wings evolved out of cooling fins, which are indeed useful even when very small. The point, as you see, is that “new” subsystems in evolution are almost always adaptations of and repurposing of existing subsystems.
Uh-oh, I think Vincent Cate raised a good point. The graph is not nearly as optimistic as it appears to be.
Well economic systems generally boil down into fairly straight forward input and output mechanisms of supply and demand. Organisms on the other hand don’t generally follow linear distributions in terms of how they respond to input and the subsequent excretion. I think of organisms of a lattice work structure that allow materials to filter and process to sustain life giving energy. Some structures are fairly organized and process in expected manners. Others don’t follow rigid conformity in their lattice work matrix and the subsequent offset of patterns changes how materials are processed and filtered through. This is why you have some people who can eat Double Cheese burgers all day and never gain a pound till the ripe age of 85 yet others balloon up to big fat fatties and die in their sleep at 50. Plants in particular are a great illustration of how fairly well understood metabolic processes still elude our understanding of precisely why one plant will thrive off a given nutrient profile while others of the same species languish in necrotic spots and lime green pallor.
Uh-oh, I think Vincent Cate raised a good point. The graph is not nearly as optimistic as it appears to be.
The number is already adjusted for inflation.
“The number is already adjusted for inflation.”
Yes, I mean really – think about it, how many cars did the average family own 100 years ago? How many phones? How many computers? What percentage of their life was spent obtaining food to feed their family? Do you really believe people today are worse off than 100 years ago? Or that it is even close?
Where do these people come from?
Extrapolating from the graph you get that longevity is roughly the log of wealth. We already have that wealth is increasing roughly at an exponential rate. So we can project longevity as increasing linearly with time from that graph.
The problem, of course, is that extrapolation under such circumstances is notorious for giving bad answers.
The bounces were the intriguing part. It would be interesting to match those with the causes. It would seem the abruptness of change would help to discover the causes.
WWI dropped life expectancy for almost everyone (or maybe the Flu pandemic, but both were pretty big killers of people), The Great Depression moved and/or normalized most of the graph to the left, and WWII once again dropped life expectancy (note Japan, moving both down AND to the left at the end of WWII).
I’d imagine that a lot of the rest of the major individual bouncing has to do with census taking and reporting. France should be on a 7-ish-year bounce pattern, and the U.S. on a 10-year pattern until recently.
Hey, I plotted life exptancy versus lung cancer in men. Turns out life expectancy grows exponentially with the number of new lung cancer cases per 100,000 men per year.
Cool! Lung cancer causes longer life expectancy. Someone tell Joe Camel!
Very.
I’d like to see the same evaluation of the US states.
It’s interesting that wealth is on a logarithmic scale but life expectancy on a linear scale.
Did anyone else notice the big drop in life expectancy China experienced during Mao’s Cultural Revolution? Socialism kills. I’d like to see an all-countries comparison with colors designating economic policies instead of geography.
Carl, if Aubrey De Gray gets his way I think it will be linear wealth that buys log life gains.
Uh? I don’t get that, Brock. Could you explain?
But if anything, this graph tells you how hard it is to translate wealth into lifespan. Everyone got 100x richer but only lives 3x longer! And at that, the gains in lifespan are somewhat misleading, in that the increases in life expectancy at birth are nearly all the result of young people not dying of infectious disease, and mothers not dying in childbirth.
For the real gains in ultimate longevity you have to look not at life expectancy at birth, where things like birth defects, accidents and infections take a major toll, but at something like life expectancy at age 65, where death is dominated by the ultimate killers of old age like heart disease and cancer.
Between 1900 and 2010 the US real (inflation-adjusted) GDP per capita increased by a factor of 25, but life expectancy at age 65 went up from about 78 to 85, i.e. one might say longevity increased by less than 10%.
It’s a fascinating question, I think, why this is. Why should spectacular gains in wealth be so much easier than gains in lifespan? One might argue there is something about wealth creation that allows for what Nassim Taleb might call “positive Black Swans:” it’s possible for accidental discoveries — semiconductors? Internet? thermodynamics? — to lead to spectacular gains in labor efficiency, hence wealth.
But longevity seems strangely proof against this. We have made many advances in medicine, but apparently essentially none of them are “Black Swans,” meaning there has never been a discovery that spectacularly increased lifespan. Isn’t that curious?
Mr. Pham;
I would suggest it indicates the effects of evolution, in that various systems in the human body all wear out at about the same time. Clearly it’s a waste (evolutionarily speaking) for some subsystem to last significantly longer — you could “cheapen” it without cost and boost some other aspect. As a result, to significantly increase lifespan, we must overcome systemic failure. Even “black swans” don’t really help because some other part wears out anyway.
If wealth increases faster than lifespan, at least we are better able to enjoy the time we do have.
It’s amusing that an owl is a symbol of wisdom. It so well matched to its environment that it hardly needs a brain. The wisdom is in that design.
So what about the human brain? We think pretty highly of ourselves, but in fact some animals far surpass us in certain aspects of the brain. Chimpanzees have much better memory, beyond any human, in some respects. Dolphins and bats process sonar. Wouldn’t super human abilities give us better survival to outperform the other tribes? Neanderthals had bigger brains (were they smarter?) and stronger muscles than our ancestors.
So I’m not sure what assumptions you can make about subsystems relating to life expectancy. The fact is the highest age of modern man has been roughly constant regardless of medicine.
The bounces were the intriguing part. It would be interesting to match those with the causes. It would seem the abruptness of change would help to discover the causes. I also note the size of population of China and America and their relative growth.
AOG, your theory holds the unlikely assumption that all the subsystems have by nature, and started off with, the same lifespan. For example. consider building a table out of the scraps of lumber and hardware you have in the garage. A priori, the various subsystems of your table will have substantially different lifetimes — there will be something that wears out first, and something else that lasts much different. Why should it be different for biological machines?
You could then set about systematically improving your table, of course, just as evolution systematically improved the lifespan and efficiency of species after they arise. But why would this lead to a convergance of the lifespans of the subsystems? Say you replace your cheap mild steel bolts, which rusted first, with good marine stainless steel bolts. It’s less likely that this brings the lifespan of the bolts in line with the wood, than that now the life of the bolts leaps ahead of the wood, preserving the heterogeneity of subsystem lifespans. I would think the same would happen to biological machines.
In short, if all our subsystems seem to fail at once — and they do — I would suspect a design lifetime. That is, evolution has served not to continuously and simply increase lifespan and resilience, but to actually cut us off at a certain maximum age. We are, in short, designed to die, it would seem to me, and at fairly definite age (or rather at some fixed ratio of age to environmental rigor).
This is hopeful news, however, because if we die not because our various systems just wear out, higgledy-piggledy, but because the Master Switch clicks to OFF, it seems more likely we can figure out how to control the Master Switch ourselves.
ken, I believe the largest bounce is the 1918 influenza epidemic.
During the 200 year period you are talking about the dollar was stable for the first half and inflating rapidly for the second half, so it is a bad yardstick for measuring wealth. The first 100 years $400 could buy 20 oz of gold and $4,000 could buy you 200 oz of gold. Today $40,000 buys you about 28 oz of gold. Prices of everything else inflated in the second 100 years also. The first 100 years people really got richer but during the second 100 years it is mostly that dollars are worth less.
What I notice is that countries are advancing regardless of their economic theories, and that partially socialistic countries, such as modern Europe are doing fine.
This analysis may indicate that technological evolution is more important to human progress than the notions of economists and the ideologies of governing political parties (absent something as toxic as Mao’s China or Pol Pot, of course).
Rand, thanks for highlighting this video. Fascinating!
Mr. Pham;
Because biological systems are much finer grained than “cheap steel” vs. “marine steel”. So the table builder could pick to use bolts that are just good enough to last as long as the cheapest component. The table builder is selecting from pre-existing units and therefore can’t match things as well. In contrast, a biological system is completely autarkic. If the table builder had to smelt and cast his own bolts, he’d have more incentive to make them just good enough.
My theory does rest on the assumption that there is some cost to more durable subsystems. Therefore creatures who subsystems were just good enough would have an advantage.
Annoying Old Guy – Precisely. To illustrate the point, Henry Ford once commissioned a study of the mode of failure of Model T Fords being scrapped. Apparently, he discovered that kingpins (IIRC), whatever the heck they are, virtually never failed causing the breakdown of the car – and reduced specs because of that finding. No point in having one part in pristine condition when the whole machine is being scrapped.
The situation in a living organism is a bit different, as living organisms self-repair. Unfortunately, this repair takes energy and materials. It is notable that most wild animals live just long enough to get their young to reproductive age. Humans are an exception; one theory of the reason is that if granny is around to look after the kids (as opposed to infants) then both parents can get out there obtaining food. Also, older humans can pass on their knowledge to the young – which most animals can’t do nearly as well.
Another illustration; imagine a mouse that lives to 30 (unless eaten by a predator) but because of its highly efficient repairs only has two pups instead of 12. Unfortunately, mice are often eaten by predators – and if all their offspring have also been eaten then that genetic line is no more.
If a genetic mutation’s adverse effects are unlikely to strike before something else has killed the individual off, there is little selective pressure to clean out such mutations. They could thus be expected to accumulate so long as environmental factors remain steady. So, as we learn to cure or prevent one class of disease, another that was rarely a problem before comes into prominence at old age.
Because biological systems are much finer grained than “cheap steel” vs. “marine steel”.
I don’t think that matters. If it did, then the far more fine-grained and multitudinous subsystems of a computer would fail more often together, at the same time, than the fewer and simpler subsystems of a lawnmower or car. But that’s not what happens. The phenomenon is likely to be scale-invariant.
In contrast, a biological system is completely autarkic.
Not at all true. In fact, biological systems are far less likely to construct new parts de novo. As a rule, evolution only allows gradual change, which means parts intended to perform new functions must be jury-rigged together by re-purposing older parts that do something different.
Indeed, this has always been one of the major challenges to evolutionary theory: how, for example, did wings evolve? Flying subsystems are sufficiently complex that it’s inconceivable they could have arisen all at once, in one generation (parents walk, children can fly), by some incredibly rare combination of lucky mutations. But on the other hand, of what advantage would some slight movement towards wings — winglets? — give the mutant who had them? And if they don’t confer an advantage, they won’t be adapted. The usual answer is that wings evolved out of cooling fins, which are indeed useful even when very small. The point, as you see, is that “new” subsystems in evolution are almost always adaptations of and repurposing of existing subsystems.
Uh-oh, I think Vincent Cate raised a good point. The graph is not nearly as optimistic as it appears to be.
Well economic systems generally boil down into fairly straight forward input and output mechanisms of supply and demand. Organisms on the other hand don’t generally follow linear distributions in terms of how they respond to input and the subsequent excretion. I think of organisms of a lattice work structure that allow materials to filter and process to sustain life giving energy. Some structures are fairly organized and process in expected manners. Others don’t follow rigid conformity in their lattice work matrix and the subsequent offset of patterns changes how materials are processed and filtered through. This is why you have some people who can eat Double Cheese burgers all day and never gain a pound till the ripe age of 85 yet others balloon up to big fat fatties and die in their sleep at 50. Plants in particular are a great illustration of how fairly well understood metabolic processes still elude our understanding of precisely why one plant will thrive off a given nutrient profile while others of the same species languish in necrotic spots and lime green pallor.
Uh-oh, I think Vincent Cate raised a good point. The graph is not nearly as optimistic as it appears to be.
The number is already adjusted for inflation.
“The number is already adjusted for inflation.”
Yes, I mean really – think about it, how many cars did the average family own 100 years ago? How many phones? How many computers? What percentage of their life was spent obtaining food to feed their family? Do you really believe people today are worse off than 100 years ago? Or that it is even close?
Where do these people come from?
Extrapolating from the graph you get that longevity is roughly the log of wealth. We already have that wealth is increasing roughly at an exponential rate. So we can project longevity as increasing linearly with time from that graph.
The problem, of course, is that extrapolation under such circumstances is notorious for giving bad answers.
The bounces were the intriguing part. It would be interesting to match those with the causes. It would seem the abruptness of change would help to discover the causes.
WWI dropped life expectancy for almost everyone (or maybe the Flu pandemic, but both were pretty big killers of people), The Great Depression moved and/or normalized most of the graph to the left, and WWII once again dropped life expectancy (note Japan, moving both down AND to the left at the end of WWII).
I’d imagine that a lot of the rest of the major individual bouncing has to do with census taking and reporting. France should be on a 7-ish-year bounce pattern, and the U.S. on a 10-year pattern until recently.
If you want to play with the data yourself, go to http://www.gapminder.org/world/.
Hey, I plotted life exptancy versus lung cancer in men. Turns out life expectancy grows exponentially with the number of new lung cancer cases per 100,000 men per year.
Cool! Lung cancer causes longer life expectancy. Someone tell Joe Camel!