Don’t blame Canada

Share on FacebookShare on Google+Email this to someoneTweet about this on Twitter

The paper Eight Americas: Investigating Mortality Disparities across Races, Counties, and Race-Counties in the United States, has this fascinating map (reformatted a bit):

As you can see there is a great deal of variance in white male life expectancy in the United Sates. Compare to this map:

“American” is probably just Scotch-Irish in this case. It is noticeable it seems on this map that the countries in central Texas where Anglo ancestry is dominated by those of German origin exhibit high life expectancy.

In any case, you can actually look at the county-by-county data set from the above paper in regards to life expectancies. The minimum male life expectancy in any county is 62, with the maximum being 80.30. The median is 73.60 and the mean 73.38 (these data are ~2000). There’s a “long tail” of sparsely populated counties with low male life expectancies as evidenced by the lower mean value than the median. The standard deviation across the counties is 2.35 years.

As can be seen on the first map there is a strong geographic component to the interregional differences. Below is a chart which reports the proportion of counties in the 50 states which have a life expectancy at, or above, the Canadian national value as of the year 2000 (again, both these values are for males).

Some states obviously have very few counties. But Kentucky has 120. None of them are at the Canadian level.

Labels:

36 Comments

  1. Have not looked at paper but first guess is that this is due to many young deaths (or reduced life expectancy) as a consequence of military service.

  2. Atomic bomb was created after Einstein theory. This eight America data is found after Rushton theory.

  3. It would seem that listening to Country Music is a lethal habit.

  4. There is a huge gap in life expectancy between group 1 (Asians) and group 3 (most affluent white subgroup, “middle America”). See figures 1 and 3 in the paper. This despite the fact that per capita income is *lower* in the group 1 than in group 3. 
     
    There is some related discussion here: 
     
    http://infoproc.blogspot.com/2009/04/us-human-development-indices.html 
     
    The Asian-White gap in human development index is about the same as the White-Black/Hispanic gap!

  5. These differences in life expectancy might be due not to ancestry but to intelligence and its attendant SES, which Linda Gottfredsson has linked extensively to health outcomes. 
     
    In your article on July 15, 2008 Razib you reported that those of English and Irish ancestry actually have higher WORDSUM scores than Germans across all regions of the United States except one (New England). Intelligence differences among whites are much more significant across temperature than background, as agnostic showed in his post “Colder climates favor civilization even among Whites alone”. The correlation between White IQ and average annual temperature per state was -0.48. 
     
    This map of life expectancy looks roughly like a temperature map of the U.S. except for some exceptions in Florida and Texas (perhaps due to affluent retirees and oil men). I think its likely that temperature mediates IQ which mediates life expectancy. That would be why Canada is globally better off in life expectancy; that and its nationalized health care.

  6. Have not looked at paper but first guess is that this is due to many young deaths (or reduced life expectancy) as a consequence of military service. 
     
    if you have a guess like this it would be nice to do the math! (you can look up casualty numbers and play around with proportions) 
     
    There is a huge gap in life expectancy between group 1 (Asians) and group 3 (most affluent white subgroup, “middle America”). See figures 1 and 3 in the paper. This despite the fact that per capita income is *lower* in the group 1 than in group 3. 
     
    the dakotas aren’t particularly wealthy either. it’s more than income/wealth. 
     
     
    This map of life expectancy looks roughly like a temperature map of the U.S. except for some exceptions in Florida and Texas (perhaps due to affluent retirees and oil men). I think its likely that temperature mediates IQ which mediates life expectancy. That would be why Canada is globally better off in life expectancy; that and its nationalized health care.
     
     
    the problem with this hypothesis is appalachia. there is probably something to your thesis, but it is likely more than just climate. i guess i need to do some GIS programming to confirm.

  7. I don’t believe these self-reported Americans are “probably just Scotch-Irish” is a rock solid statement – on what do you base this claim?  
     
    The map of self-reported Scotch-Irish looks very different to the map of self-reported “Americans”.  
     
    For example, there are a lot of Scotch-Irish in the North West, why don’t they call themselves “American”? (I have an answer). Is their life-expectancy longer? (I have an answer for that, too). 
     
    And your “American” region extends well into German American ancestry regions. Your analysis is flawed. 
     
    Wealth is the most likely the main factor in life-expectancy being picked up in this data. Look how well this US income map matches the white life-expectancy graph – note the pockets of high wealth/life expectancy in the East.

  8. I don’t believe these self-reported Americans are “probably just Scotch-Irish” is a rock solid statement – on what do you base this claim?  
     
    uh, read some history. the regions where “americans “are dominant is exactly where the primary and secondary waves of scots-irish settled in the 18th to 19th century (secondary insofar as the western regions were settled from the eastern ones).* as in the china threads you’re comments kind of strike me as either ignorant or stupid too (or both), but you also are a bit too assertive for either of those for my taste (many GNXP commenters are either ignorant or stupid, but when they’re polite they’re tolerable). 
     
    Wealth is the most likely the main factor in life-expectancy being picked up in this data. Look how well this US income map matches the white life-expectancy graph – note the pockets of high wealth/life expectancy in the East.  
     
    yes. but as i noted to steve hsu above, low income areas in the upper midwest do OK. that’s one of the insights of the 8 americas paper, income can’t explain everything. 
     
    if people keep leaving annoying comments i’m just going to close this thread. 
     
    * i would qualify though that just because an area is culturally dominated by scots-irish does not mean it is demographically dominated by them. “american” subsumes british americans in general, but especially in areas historically dominated by the scots-irish subculture.

  9. Not to flood this thread box with more links to graphs, but given agnostic’s article that I mentioned early, I thought it would be fun to compare a temperature map of the United States to the life expectancy one. The result for females can be seen here
     
    Pretty striking if you ask me. One of the things it might suggest is that the global pattern towards higher IQ at greater latitude may not be only the result of evolved differences, but of the suitability of different environments to fostering intelligent people and societies.

  10. Not to flood this thread box with more links to graphs 
     
    oh, no, please be my guest! links & graphics are preferred to verbal argument (not that there’s anything wrong with the latter, but info-density is usually lower).

  11. Try comparing this to the distribution of 1. obesity rates, 2. alcohol consumption per capita, and 3. ratings of healthcare systems 
     
    This seems to correlate very strongly with obesity rates, see: http://3.bp.blogspot.com/_XDLnpGlOHFU/SSQM51KxrqI/AAAAAAAACNI/dBuLAyaLyXo/s400/obesitymap463.gif 
    (i couldn’t find one that has more refined county-by-county data) 
     
    there are some places that don’t match the trend (notice a huge splotch of unhealthiness around las vegas, nevada) but I think that that might have a lot to do with unhealthy lifestyles. 
     
    In general, I think this map says more about cultural and culinary differences in the different regions than anything else. 
     
    Fat people don’t live as long, and eating fried food and mega-sized portions doesn’t help. People in the south and midwest are notorious for being overweight, and floridians at the very least go to the beach and probably try and keep in shape. People in places like idaho, even without a lot of income or the best healthcare probably engage in more exercise and have healthier lifestyles (those are my preconceptions anyway)

  12. This is interesting if it is true: 
    http://www.websters-online-dictionary.org/coreimages/contemporary/NIAAA/spc96.gif 
     
    it seems that alcohol consumption is inversely related to life expectancy… 
     
    weird. 
     
    Also, about Nevada standing out, well, it has the highest rate of alcoholism in the country, so…

  13. I don’t see any comments about how population mobility is affecting this graph. Things like people with health problems (and thus closer to death) not continuing to live in Alaska comes to mind. 
     
    Additionally, Florida has the ‘retiree’ band which is a form of selection bias. If you retire at 65 and are healthy enough to move to Florida, you’re probably going to live longer. 
     
    There’s likely more effects, as well

  14. One of the things it might suggest is that the global pattern towards higher IQ at greater latitude may not be only the result of evolved differences, but of the suitability of different environments to fostering intelligent people and societies. 
     
    To add to this, perhaps aspects of general physical wellbeing such as health and longevity as well as cognitive function/IQ are influenced by the pathogen load people are subjected to. (I’m thinking of Cochran and Ewald’s pathogenic theories of disease here.) If the pathogen load is generally lower in colder and drier climes, then perhaps this might explain some of the regional differences in life expectancy and IQ.

  15. The male mortality rates in those areas are probably inflated by the popularity of motorsports, ATVs, dirt bikes, and BMX. Motorsports plus alcohol (as mentioned by other posters) isn’t a great combination if you want to live long. I live an area splotched red, and I can’t even count the number of spinal injury or head trauma stories that I’ve heard.

  16. I am from Kentucky. Oh my. Oh well, at least I wont be a burden as an elderly person.

  17. There seems to be a sharp difference somewhere near the border between Indiana and Northern Illinois. My wife grew up in a small town near the border. The white southerners in town were known as Yellowhammers, which I’m guessing is a form of woodpecker bird. And woodpecker = peckerwood = Southerner.

  18. It’s interesting that you can see the spine of the Appalachians stand out in purple amidst all the red. That might suggest that cool weather (due to altitude rather than latitude in this case) contributes to longer life expectancies even in modern American.

  19. White Canadians are typically scotch-irish as well. 
     
    Perhaps this is more accurate a map of America’s obesity epidemic? 
     
    Although we are catching up, Canadians historically are not nearly as fat as Americans.

  20. That New Scientist article talks about pre-Sapiens trends. If you could extend it to recent human evolution, then Nordic Neanderthal should have massacred the Tropical modern humans. 
     
    As for the relationship between IQ and latitude, I still don’t understand how it fits with the fact that all known independent centres of advanced civilisation (except for Northern China) were located in hot climates.

  21. John Hawks thinks the paper the New Scientist article is based on is crap: 
    http://johnhawks.net/weblog/reviews/brain/climate-brain-expansion-new-scientist-2009.html

  22. I’m not a physiologist, and I know the brain does a lot of metabolism – but can’t you increase its cooling power arbitrarily, at some cost, by means of greater vascularity?

  23. It seems that there are four general areas of lower life expectancy: the Deep South, the Ozarks, Appalachia, and counties with larger than average Native American populations. The last one is the question here. Indians do have lower life expectancy, and that does seem to explain the counties that don’t fit into the other three geographical regions, but there are other relatively Indian-dense counties (e.g., those in the Dakotas) that still have high life expectancies. 
     
    So… does this explain it, or not?

  24. Re: American Indian life expectancy: This map is for whites only, but drunk driving deaths were quite high in New Mexico, but the state has brought them down with a lot of get tough enforcement, such as tying Breathalyzers to ignition systems of convicted DUIs when they get their licenses back. 
     
    Some of the low white life expectancy in Oklahoma might be related to moderate levels of American Indian background, combined with high levels of hard-drinking Scots-Irish background.

  25. One fact that I think needs to be highlighted — and Razib, you briefly touched on this in your latest post about regionalism and life expectancy — is that Hispanics have a higher life expectancy than non-Hispanic whites. See: http://www.paho.org/English/DD/PIN/pr040609.htm 
     
    This is completely unexpected, as Hispanics are also more likely to be obese, poor, and uninsured than non-Hispanic whites. I think understanding the so-called “Hispanic Paradox” (or “Latino Paradox”) is crucial to making sense of the factors behind life expectancy in the USA.  
     
    There was also an attempt made by researchers to only gauge natural life expectancy, namely life expectancy without taking into account accidental/violent deaths (homicide, car accidents, etc.). The analysis found that the USA life expectancy was dragged down drastically by such accidental/violent deaths, much more so than most other 1st-word countries. See: http://politicalcalculations.blogspot.com/2007/09/natural-life-expectancy-in-united.html 
     
    I don’t know much about the methodology of this particular study, but investigating the “natural” life expectancy as opposed to the unadjusted life expectancy seems like an intriguing idea, at the very least. I would like to see more data on the “natural” life expectancy of individual states.

  26. “There is a huge gap in life expectancy between group 1 (Asians) and group 3 (most affluent white subgroup, “middle America”). See figures 1 and 3 in the paper. This despite the fact that per capita income is *lower* in the group 1 than in group 3.” 
     
    Yes, and Latinos in the USA have a higher life expectancy than non-Hispanic whites. Razib, I believe you touched on this in your latest post about regionalism and life expectancy. This seems to contradict the notion that income, overweight/obesity prevalence, or availability of health insurance are the driving forces behind life expectancy. See: 
     
    http://www.paho.org/English/DD/PIN/pr040609.htm

  27. This seems to contradict the notion that income, overweight/obesity prevalence, or availability of health insurance are the driving forces behind life expectancy.  
     
    i suspect that a ‘silver bullet’ model where *one* parameter is going to explain all of life expec. variation is confusing in a lot of ways. but the way public policy works, you need some simple models which can be solved via simple policies.

  28. Let me just apologize and say that I didn’t mean to post both of my last two messages —- I posted the first message, and that didn’t show up, so I then wrote a second, abbreviated one — when I tried to post that, the first post showed up instead. Then my second post showed up.  
     
    Anyway, 
     
    “i suspect that a ‘silver bullet’ model where *one* parameter is going to explain all of life expec. variation is confusing in a lot of ways. but the way public policy works, you need some simple models which can be solved via simple policies.” 
     
    I would agree with this. However, I would argue that we are not even at the stage where we even have a much in the way of a simple model to work with. I think “longevity science” is still a very young field. 
     
    “Indians do have lower life expectancy, and that does seem to explain the counties that don’t fit into the other three geographical regions, but there are other relatively Indian-dense counties (e.g., those in the Dakotas) that still have high life expectancies.” 
     
    My understanding is that Native American life expectancy is extreme low in the Dakotas, even as low as the 50s for males. See: 
     
    http://www.usatoday.com/news/health/2006-09-11-life-expectancy_x.htm

  29. America is a much more violent society, with violence usually affecting the youth. That affects life expectancy a bit. How much does that affect this outcome?

  30. Nevermind, found it. After adjusting for deaths due to car accidents and violence, America has the highest life expectancy. Interesting. 
     
    http://www.aei.org/docLib/20061017_OhsfeldtSchneiderPresentation.pdf

  31. I’m sure Razib knows this, but the term “Scots-Irish” is shorthand for people from the Borderlands. Some of them were in Ireland for 100 years or so, but for 500 years their culture had been formed by rejection of political authority, which entailed lack of civil order. High levels of violence, and a cultural toleration of violence are part of the package.

  32. After adjusting for deaths due to car accidents and violence, America has the highest life expectancy. Interesting. 
     
    Strangely enough, their adjustment has the effect of reducing some life expectancies (e.g. Japan from 78.6 to 76.0 ??) So basically they seem to “add in” some deaths that “should” be there, but aren’t? Their methodology is not exactly obvious.  
     
    Not to say they are necessarily wrong, but I’d like to see how they get these figures. 
     
    Also, comparing cancer survival rates at 5 years is not meaningful for countries that have different screening policies. Overlooking this obvious problem makes the whole presentation seem a bit suspicious.

  33. The temperature idea is really interesting, and could moderate the effect of the Dakotas being generally poor, but then how do you explain the southern tips of Texas and Florida?

  34. What was the county that got 62.0? Some military base? Indian Res? Neither?

  35. Compare this life-expectancy map with the ones shown at the URL below showing relative amounts of bad ozone: 
     
    http://earthobservatory.nasa.gov/IOTD/view.php?id=39715 
     
    best, 
     
    rv

a