Tyler Cowen cites a study that says that the rate at which life expectancy at birth (LEB) is increasing has slowed down. This does not surprise me.
Suppose that there are two outcomes. One outcome is that you live to a ripe old age. The other outcome is that you die within one year of birth. As the percentage of people who die early goes from 10 percent to 1 percent, average LEB goes up dramatically. As it goes from 1 percent to 0.1 percent, average LEB goes up more slowly. But I would not call that stagnation.
It is a one-time gain, a la R. Gordon? Have you written about Gordon’s Rise and Fall of American Growth?
Not a slowdown. US life expectancy has declined two years running. Opioids taking the blame. Unsurprisingly, federal regulations played a significant role in incentivizing over-prescription of painkillers: http://www.nationalreview.com/article/453373/opioid-crisis-medicare-incentives-made-it-grow
Life expectancy is a tool that really needs a lot of improvement to be a meaningful measure of health.
What interests me the most are the positive anomalies like England from 1840-1880.
Not sure why this result wasn’t classified as an unsurprising case of diminishing returns…
My first reaction was The Law Of Diminishing Returns hold again.
1) In 1850 – 1950 they discovered all the low hanging improvements. In reality the greatest medical advance was doctor cleanliness.
2) The number of infants dieing is a lot less. It is amazing to think a couple could have 7 children and only three or four see adulthood.
3) As a person ages their lifestyle choices makes more impact expectancy. When people died at 60, smoking had smaller impacts on life expectancy. Once smoking was under control than sugar drinks and soda.
Implicit in Arnold’s post is the assumption that the only way to change life expectancy is to fix the problem of infant deaths. Fair point, save an infant at birth from death and you’ll add maybe 75 years of life to the nations total ‘life years’ but save a 74 year old man from a heart attack and you may only add 1 additional year.
BUT only a fraction of health costs and R&D is spent on infant issues. Are we saying that we’ve already hit the upper limit on lifespan and all of our diseases the best we can hope for is to live relatively free of disease but rarely beyond 70-80 years?
We are in the long tail of a great leap of mankind. In the last 300 years, we have gained non-animal machines, new, highly formable, materials, new energy sources, new forms of energy manipulation(electricity), germ theory, antibiotics, etc. Such advances will continue iteratively, but what pray tell is the modern equivalent of learning the dependency of electricity and magnetism? The dramatic speed increase of the auto?
This is not to be pessimistic, just the low hanging fruit again. In that time, more and more of humans on earth have shifted from a demographic controlled by mortality to one controlled by fertility. Instead of 10 kids with maybe one making it, we have one or two kids with a high probability both will die of old age. Children are no longer a source of labor for the family enterprise, they are a discretionary expense.
I’ve said before, someone is going to really advance economics by figuring out how the economy will work when the population isn’t constantly growing and basic goods needs are met inexpensively. Depending on growth isn’t going to hold out for another century, assuming no widespread die off due to war or sudden disease emergence.
So yes, life expectancy growth is going to slow. So will dramatic innovation.
Indeed so. Angus Deaton makes a similar point, if I recall correctly, in his book The Great Escape.
Reading a lot about Bitcoin but I still don’t really get how to own it and use it…but one interesting aspect of it is transactions are verified by computers solving complex problems and proving to everyone else that they came up with the right answer. As an incentive, new coins are issued every ten minutes and given to the first miner. In order to maintain the 10 minutes, the algorithm will make the next problem harder if the miners solve it in less than 10 minutes, easier if it takes longer.
Innovation might be modeled along those lines. People work on ‘solving the next problem’….but of course the next problem might take ten minutes or ten decades to solve. Another difference is that ‘miners’ can work on solving the next problem or they can opt to expand the present solution to more people.
For example, one guy comes up with the idea of the wheel. The next guy has a choice, he could try to solve the next ‘wheel problem’ (say inventing the suspension system) or he could work on expanding the use of the wheel to more places. What happens if someone comes up with a great innovation? Well the marginal reward for expanding that innovation will be very high but the marginal reward for finding the next innovation will be relatively low.
In other words, do you sell more smart phones so everyone has them or do you try to outdo the iphone?
I think that model would fit what we’ve seen in history where innovations appear to come in spurts with pauses in between.
Kling’s post suggests a similar model with lifespan. One method would be to try to invent cures and innovations to extend life. Another is to work on making sure everyone is using what is already available. Both are necessary.
Complicating this is the problem that the market works but humans are irrational consumers. For example, our brains reward us with a ‘fix’ for eating sugars and the market is able to produce nearly unlimited amounts of sugar based foods for almost nothing. Result while we are working on ‘cures’ to extend lifespan we create epidemics of diabetes, obesity, heart disease and cancer that sap lifespans and demand lots of innovations just to keep from falling backwards.