tag:blogger.com,1999:blog-264226589944705290.post8161244332836006859..comments2023-11-05T03:45:25.001-08:00Comments on God Plays Dice: life expectanciesMichael Lugohttp://www.blogger.com/profile/15671307315028242949noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-264226589944705290.post-85051921077337898402007-07-26T13:25:00.000-07:002007-07-26T13:25:00.000-07:00Reading the column carefully, I see they're talkin...Reading the column carefully, I see they're talking CDC-ish numbers, as opposed to Social Security (which do have different methodologies).<BR/><BR/>Smoking is indeed a huge factor impacting mortality. It was the life insurance companies who learned very early on the great impact smoking has on life expectancy (I don't have a table with me now, but I believe the hit regular smoking puts on adult life expectancy is about 7 years.)<BR/><BR/>Also, which I don't see particularly noted, there are noticeable differences in life expectancy in the U.S. between blacks and non-latino whites. Again, measured in years (and no, not just due to not having reliable birth certificates). I don't know what the numbers are for other races/ethnicities.<BR/><BR/>The violent death aspect is interesting, because that impacts men mostly, and men of a certain age (you can see it in the mortality tables - greatly heightened male mortality from age 15 - 25 or so, then the mortality rate drops, until about age 35 where it starts climbing again... the age 35+ is the death by heart trouble/cancer incidence increasing.) I call it the "stupid period", and it's a pretty common pattern throughout the 20th century in the U.S., even if you remove deaths from war.meephttps://www.blogger.com/profile/08893035949118989768noreply@blogger.comtag:blogger.com,1999:blog-264226589944705290.post-62529064162555330932007-07-26T13:17:00.000-07:002007-07-26T13:17:00.000-07:00Of course, you could ask the actuary what this stu...Of course, you could ask the actuary what this stuff is supposed to mean. I've actually looked at historical mortality tables and life expectancy (mainly from the U.S. Social Security office) as part of my job.<BR/><BR/>When various groups give life expectancy, they're generally not assuming mortality stays constant (say probability of 65-yr-old surviving to 66) -- many times they use a projected mortality table (and yes, there are assumptions in there). When I looked at historical mortality, I had cohort mortality tables, meaning the mortality experienced by the people born in 1920, plus some projections for those who've made it to age 80 and beyond (the tables were from 2000). You need to be careful with regards to infant mortality, too. Sometimes, to not deal with the separate issue of infant and young child mortality, they start the mortality table at age 5 in calculating life expectancy. And, with regards to infant mortality, you've got to be careful, too. Some countries require the baby to have survived birth for a few days for it to be included in stats. The standard in the U.S. is that if the baby is born alive, and it dies minutes later, it counts for infant mortality. Which is why it can be a good idea to compare life expectancy of those who have reached some non-zero age. <BR/><BR/>I have no particular reason to trust the Cuban stats. They're the same people who claim that nigh unto 99% of the Cuban population is literate, which is not credible considering mental retardation, learning disabilities, and the like. Chances are, the stats are faked. Who's going to gainsay them? It's not like there are independent life insurance companies operating in Cuba.meephttps://www.blogger.com/profile/08893035949118989768noreply@blogger.comtag:blogger.com,1999:blog-264226589944705290.post-14197669379248724892007-07-24T09:20:00.000-07:002007-07-24T09:20:00.000-07:00If we adopted a policy of killing everyone at 50, ...<I>If we adopted a policy of killing everyone at 50, population would go down but then it would grow again.</I><BR/><BR/>It would grow again because the number of births per year is also increasing. You don't just have to account for the fact that the system is growing, you have to account for the fact that it's <I>accelerating</I>.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-264226589944705290.post-52816442797220701862007-07-24T07:36:00.000-07:002007-07-24T07:36:00.000-07:00anonymous,I'm not sure how to take into account th...anonymous,<BR/><BR/>I'm not sure how to take into account that the system is growing. But I think what your computation might tell us is that if life expectancy were 49.6 years, then population would be stable.<BR/><BR/>Even that seems a bit strange, though; just because the life expectancy is shorter, that doesn't mean the population should stop growing. If we adopted a policy of killing everyone at 50, population would go down but then it would grow again.Michael Lugohttps://www.blogger.com/profile/15671307315028242949noreply@blogger.comtag:blogger.com,1999:blog-264226589944705290.post-17455121980784206462007-07-24T01:33:00.000-07:002007-07-24T01:33:00.000-07:00I can't help but wonder what Little's Law can tell...I can't help but wonder what Little's Law can tell us:<BR/>http://en.wikipedia.org/wiki/Little's_law<BR/><BR/>Paraphrasing...<BR/>The average number of things in a system, N, is equal to their average arrival rate, R, multiplied by their average time in the system, T, or:<BR/> N = R * T<BR/>or<BR/> T = N / R<BR/><BR/>It seems not ridiculous to let...<BR/><BR/>N : population<BR/>R : births/year<BR/>T : average life expectancy<BR/><BR/>N = current world population = 6,600,000,000 people<BR/><BR/>R = per www.who.int/whr/2005/media_centre/facts_en.pdfR = 133,000,000 new people / year<BR/><BR/>6600/133 = 49.6 years<BR/><BR/>I'm not sure how to account for the fact that the system is growing rather than stable.Anonymousnoreply@blogger.com