Middle-aged white mortality, and CFA article

A popular front-page story in the New York Times yesterday elaborates on a gripping discovery by recent economics Nobel Laureate Angus Deaton, and Professor Case.  The mortality rate among middle-aged (45 to 54 years old) people globally is rising at a break-neck pace for just white Americans.  Regrettably, we all likely know someone who has been impacted.  Our blog covers the topic of both longevity and mortality for different important settings (here, here, here, here), so it was interesting for us to take a quick look at the study's findings.  And present the comprehensive understandings within the framework we normally look at it here, along with brief commentary.  We see that there is a statistical difference for this population segment, buttressed by a reasonable mortality difference (in escalated death count) within any broader frame.  This article does not intend to find value in producing in-depth commentary of the causal factors related to middle-aged white mortality for Americans, which are proposed for anyone interested (apparently both JAMA and NEJM were not), in the Proceedings of the National Academy of Sciences' article.

What are shown below are the mortality rates (per 100,000 Americans).  The data is current and is a few years less than the basic time period shown in the research article.  It is developed from the annual data from the U.S. Health Department and the CDC.  We see in red the mortality for Americans, ignoring race.  For example the red dashed lines -using the left axis- shows the annual mortality for all Americans has fallen 0.07% (-67/100,000) in the nearly 1.5 decades through 2013.  The major causes associated with this drop in mortality are: heart disease, and malignant neoplasms (cancer).  

The white solid line -using the right axis- demonstrates the annual mortality for 45-54 year old white Americans.  This has risen 0.005% (5/100,000) in the same nearly 1.5 decades - of course one can not stay 45-54 years old for nearly 1.5 decades so it's a puzzle why this effect only enters this time frame.  To provide important context though, the standard deviation of these annual mortality rates is nearly 600 (per 100,000, or 0.6%).  The major causes associated with this rise in mortality are: accidental injury, and intentional suicide.  The largest killers are still heart disease and cancer, but both have declined in overall share, in parellel with the trend with the overall population.  

Put differently, the mortality rates have increased relative to non-white and similarly relative to those not in the 45-54 year old range.  With the population base of U.S. citizens, this amounts to about 2 thousand additional deaths per year per age (nearly 300 thousand additional deaths for an entire cohort over say 15-years).  Out of nearly 3 million Americans who die each year from any cause (~180k in the 45-54 year age group).  Of course given the rebound since the global financial crisis (GFC) in overall mortality (see the rebound at 2009 for the red dashed line above), this clearly neutralizes somewhat the startling finding that for the smaller segment of the population (45-54 year old whites) is alone in seeing a severe pick-up in mortality.  Revisiting the chart above it is also visible that the mortality among the latter group has actually come down slightly since just before the GFC!

Now in other news, we had a top article "No winners with high-fee funds" in the Q4 CFA publication.  The article shows that across a range of index-tracking funds, we can see the impact of high fees on performance. The rationale for human advisers paying high fees is the expectation of earning greater net returns than one’s peers over time. Yet research shows that the opposite is true. The only risk variation in gross performance (and mostly by a large amount on the downside) is associated with high-fee funds. When high fees are taken into account, the comparison becomes too probabilistically implausible to ever justify them (there is barely a 25% chance that one will outperform a random low-fee fund). The modern "big data" article was shared a number of times such as by NPR radio, and Bloomberg's Ritholtz.  Additionally, now 2 of my articles with CFA are among their top 20 most read and downloaded.  

Also concerning my statistics book, Statistics Topics, in October 2015 it again breached the top of the rankings in a few locations (top couple thousand books out of millions at one point), including Quantocracy's top books bought.  And Amazon's top probability and statistics reference books, alongside eminent titles such as Freakonomics, The Signal and the Noise, and The Black Swan.