3 ms·
(On edit: below comment is retracted. At least for 2015 to 2016, the decline is not attributable to Simpson's paradox. See Table 4 in the CDC report linked belo
by kauffj 8y ago
(On edit: below comment is retracted. At least for 2015 to 2016, the decline is not attributable to Simpson's paradox. See Table 4 in the CDC report linked below.)
This is either incompetence or maliciousness on behalf of the CDC, WSJ, or both. Life expectancy in the US is falling, but it's only falling due to Simpson's Paradox [1].
The CDC's own data [2] shows life expectancy increasing for all genders and races except black men, which decreased marginally over the last few years.
The primary reason life expectancy is falling is due to changes in the racial composition of the population, not because of opioids or suicide.
[1]: https://en.wikipedia.org/wiki/Simpson%27s_paradox https://en.wikipedia.org/wiki/Simpson%27s_paradox
[2]: https://www.cdc.gov/nchs/data/nvsr/nvsr67/nvsr67_05.pdf https://www.cdc.gov/nchs/data/nvsr/nvsr67/nvsr67_05.pdf (see Figure 5)
- vasilipupkin 8y agoFrom the document you posted: since 1990. • In 2016 compared with 2015, life expectancy decreased for non-Hispanic white males (0.2 year), non-Hispanic black males (0.4), non-Hispanic black females (0.2), Hispanic males (0.2), and Hispanic females (0.1). Life expectancy for non-Hispanic white females remained unchanged So, no, it's not just Simpson's paradox.
- kauffj 8y agoYou're correct. I edited the parent.
- pjc50 8y agoHow does that compare to other countries with changing racial composition? What has happened to black men to cause their life expectancy to fall?
- ArrayList 8y agoA sharp rise in police killing unarmed black men, I'd wager.
- dragonwriter 8y agoAFAIK, there's been a sharp rise in people who aren't black noticing that, not a sharp rise in the underlying phenomenon.
- petermcneeley 8y agoLooking at Fig. 5 you can clearly see that peak life expectancy for Non Hispanic white males occurs around 2010 then flatlines and finally heads down.
- nonbel 8y agoNot sure why you are getting downvoted. It is certainly possible this is just an artifact of changing demographics, or even adoption of ICD-10. It could even be that birth records are becoming more reliable than in the past so people are not overstating their age as much (eg, to collect social security).
- wycy 8y agoThat's not my takeaway from Figure 5. The curve for "Non-Hispanic white male" seems to be dipping downward since 2014, and "non-Hispanic white female" appears to either also be decreasing or at best flatlining.