Bogus Trend-spotting: Suicide Spates

Posted: June 9, 2014 in Uncategorized
Tags: ,

In psychology, there’s something known as an “availability cascade” where something complex and not easily explainable is elucidated in a very simple and insightful way; never mind that it’s wrong. But because of the explanation’s simplicity and plausibility, everyone including political leaders and journalists latch on, creating a self-reinforcing cycle of false credibility. Few issues are as susceptible to availability cascades as “suicide spates.”

In China, we first saw this with the Taiwanese factory Foxconn in 2010, when 14 workers committed suicide allegedly due to “harsh working conditions” (the story was dusted off again last year). That year media started ticking off suicides and hundreds of reports questioned whether people could in good conscience buy Apple products. Never mind that, given the factory’s 1 million-plus employees in Mainland China, you were 16 times more likely to kill yourself if you were a typical American living in Montana that year.

Then a few weeks ago some journalists noted that four Chinese media executives had killed themselves over the previous few months. This prompted “speculation that these tragic incidents may be linked to Beijing’s ongoing anti-graft campaign,” according to one outlet. Never mind that there are more than 300,000 Chinese journalists in state-media agencies alone (not including the hundreds of private outlets), and China’s annual suicide rate is 22 9.8 per 100,000 people (compared to 12 in the United States).

Now it appears there’s another availability cascade brewing; this one suggesting that Chinese officials might be committing suicide en masse because of pressure from Xi Jinping’s corruption crackdown. There’s not an agreed upon number of suicides in this demographic. Wall Street Journal said it was 23 last year. Then this week New York Times said it was “at least 15” over the past 18 months. Either way, these are pretty paltry numbers when you consider that China has over 7 million civil servants.

Caixin actually appeared to draw the most compelling evidence that something was amiss when it said that official suicides more than doubled from 21 in 2012 to 48 in 2013. But then you notice a big flaw in the reasoning. Those numbers are based on media reports of suicides, not any actual empirical data. The article said, “An increasing number of officials are committing suicide, if media reports on the topic are any indication.” They are not.

This is the great flaw in all these supposed “suicide spates,” because it’s really only once a spate narrative has been established that people start paying attention, counting suicides and reporting them. We know for instance that there were at least 14 suicides at Foxconn in 2010 because people perked up and plugged each new death into the established storyline. But we have no idea how many there were in 2008 because this narrative didn’t yet exist to prompt reporting. This is the case with officials now. We don’t really know how many cadre suicides there were five years ago, but if anyone kills themselves now, you better believe they’re about to become a narrative-reinforcing statistic. It’s very easy to find “suicide spates” almost anywhere when you’re looking for them.

The ironic thing is that all this coverage of suicides might play a greater role in causing them than whatever supposed causal factor is being explored. It’s been well-documented that wide publicity of suicides encourages copy-cats (known as the “Werther Effect”). David Phillips, one of the most prominent sociologists to study this trend, told New York Times in 1987 that “Hearing about a suicide seems to make those who are vulnerable feel they have permission to do it.” One famous illustration was when Marilyn Monroe’s 1962 death was followed by a 12 percent nationwide spike in suicides.

This effect has been documented in China too. After the 2010 Foxconn suicides, researchers from Hong Kong and Australia found that there was actually a brief period with a slightly abnormal rapid-fire succession of suicides. This wasn’t because of harsh working conditions, but because coverage by Beijing-based media and rapid news dissemination within Foxconn influenced copycats.

Still, whether it’s factory workers, journalists or cadres, we’ll have to see dozens, if not hundreds more suicides in these groups before we can entertain the notion that something is seriously abnormal. When we’re able to pluck out several suicides with a common denominator, it’s tempting to assume a trend, grasp for convenient answers and then blow it all out of proportion. Perhaps work stress is a factor in some factory worker suicides, or anti-graft campaign pressure with some officials and journalists. But to suggest that suicides in these groups have skyrocketed in a short period due to a single factor is patently silly. Last year I spoke with Michael Phillips, a Canadian psychologist and executive director of the World Health Organization’s Suicide Prevention Center in Beijing. “[There’s an] idea that there are unique stressors that cause suicide,” he told me. “But suicide is always the result of a variety of factors – social, genetic and psychological. The press loves simple answers for complicated problems.”

[Correction: A more recent University of Hong Kong study found that China’s current suicide rate is 9.8 per 100,000 people, rather than the older 22/100,000 figure]

Comments
  1. Rob says:

    Simply brilliant. Puts nonsense where it belongs. In the corner till it behaves. Thank you.

  2. […] Availability cascade and suicide spates in China. “Caixin actually appeared to draw the most compelling evidence that something was amiss when it said that official suicides more than doubled from 21 in 2012 to 48 in 2013. But then you notice a big flaw in the reasoning. Those numbers are based on media reports of suicides, not any actual empirical data. The article said, ‘An increasing number of officials are committing suicide, if media reports on the topic are any indication.’ They are not. // This is the great flaw in all these supposed ‘suicide spates,’ because it’s really only once a spate narrative has been established that people start paying attention, counting suicides and reporting them.” (Sinostand) […]

  3. Potomacker says:

    I am always very suspicious of social research data form China, to at least the same degree as am I of economic data. There are a range of cultural and institutional factors that suppress the gathering of reliable numbers. That having been said, I agree that the media love to find trends where there are none, especially in the attempts to be the first to spot them. Good work in providing a very sensible counternarrative to this sort of morbid fascination and gawking.
    http://www.slate.com/articles/news_and_politics/press_box/2010/07/the_new_york_timesbogus_trend_factory.html

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s