Pages

Saturday, December 5, 2015

Suicide: After the financial crisis

While overall U.S. mortalities have barely tracked population growth over the past decade, the sliver of deaths resulting from intentional self-harm (suicide) has dramatically risen since the global financial crisis (GFC).  Regrettably, many of these suicides occur at this time of year.  Even as the economy shows signs of recovery, the results are obviously heterogeneous.  And the rate of suicides continues to rise for some segment of the population, as a hideous legacy of the recession.  Significantly this is coming from middle-aged Black males, who have suffered worse than most population segments in terms of losing jobs and social services.  While the media and public discourse often go over deaths from homicide (e.g., from police brutality), it is curious to note that the overall U.S. murder rate has come down over the past 10 years.  In this article we explore the data and drivers behind the suicide rate, while in a future article we will discuss the methods employed and their effectiveness.

In the pie-charts below we show the overall mortality of the total U.S. population in 2003, as well as in the last full year of known data.  This perspective provides a snapshot of mortality several years before the GFC, and several years after.  We see that the share of deaths from the leading causes (heart disease and a specific cancer) have each come down.  However suicides have jumped from 1.1% of all deaths a decade ago, to 1.4% now.  From the time series on the right, we see that the suicide rate had come down to low levels just before the recession, but then reversed all of a sudden.

And the reentry of suicide into the top 10 leading death causes now draws greater attention to this topic.  Recently Nobel Laureate Angus Deaton put out questionable research -in the Proceedings of the National Academy of Sciences- that the mortality rate among White middle-aged males was rising dramatically.  His and his wife's research were significant in collecting a handful of ill-assorted death causes (such as unintentional injury, and suicide) for the largest race segment of the population.  But their research did not explore the age-adjusted and statistical significance of this segment versus others for suicide as the lone death cause (let alone the economic connection behind them.)

So what are the statistically significant factors behind this pick-up in the suicide rate?  Exploring the Centers for Disease Control and Prevention data (as well as data from NIMH and WHO), we see that middle-aged adults are behind the rise in the suicide rate.  We see a peak among this group, who should be at the prime of their careers.  We see a second smaller peak among Millennials, who (in their 20s) should be at the prime happiness of their lives.


Additionally, the suicide rate increase is not geographically clustered in a clean and rational way, though we do see some pick up in less-populated states.  Looking at major races and gender together would also initially show a tired result.  Look at all of the steady, solid lines on the left chart (below).



We wouldn't think much of the trend among Blacks, generally, based on their suicide rates shown.  However, when segmenting further by age (e.g., those in their 50s) we see the greatest statistical separation (between before the GFC and afterwards).  For example, the suicide rate surge among Black males in their 50s (from generally low levels) is least likely to be an outcome of chance alone.  And we identify this with an orange "1" rank.  Next is White males in their 50s, in red.  Followed by White women in their 50s, in green.  There is insufficient collected suicide data for Black women in their 50s, though we can notice on this log scale that there is a modest rise in the suicide rate among Black women too, in aggregate.

Now we should note that the correlations between these segmented suicide rates, and the unemployment rates among those respective adults, are high at roughly 0.7 each!  But what makes this result slightly more causal, as opposed to merely chance, is when looking at the consistent ranked relationship between the magnitude of the economic stimulus (shock to unemployment which of course may be a variable substituting for other modern hardship factors), and the statistical significance rank of the suicide rate rise.  We can also note that we have seen a similar link currently in the United Kingdom, and in Japan in the early 1990s.

Looking now at the right chart (above), adult Black males have had the largest rise (1 in orange), followed by adult White males (2 in red), and then adult White females (3 in green).  The ranks evidently match, which would only be the case from luck, in this fraction of times:

1/3P= 16%

In other words, for this GFC, stronger economic hardship among a segment of the population has likely led to a larger rise in the suicide rates for that respective segment.  In a later article, we'll further explore the topic of suicide, including an analysis of the various methods attempted by segments of the population, and their potency.

Besides being unlawful, destroying one's gift of life is a serious enough topic that in both this (and our subsequent article), I would be remiss to not strongly advise you to seek out help, if you are generally upset and considering suicide as an option.  The International Association for Suicide Prevention is a notable starting point for people anywhere in the world, and I propose supplementing this with doing more of the things that you enjoy.

No comments:

Post a Comment