Better information about youth suicide rates

A few weeks ago, I did a writeup on Elizabeth Hawksworth's error-ridden Washington Post "Post Everything" article about suicide.  Here's how I started that post:

 At 10:18 yesterday morning, on PostEverything, Elizabeth Hawksworth, a Toronto-based writer, posted an article making the claims that "It’s estimated that 1.5 students out of every 100 will commit suicide at some point during their college career," and that "Suicide rates among college students have increased by 200 percent since the 1950s."   

For the casual reader, both of the cited statistics might have been alarming. Thankfully, they are both wrong.  As of this morning (i.e. the next day) at 7:10 a.m., there has been no correction of either. 

Today, there is what seems to be a much more accurate piece, written by  David Finkelhor,a sociology professor and director of the Family Research Laboratory at the University of New Hampshire.  Most significantly for present purposes, Finkelhor reports that:

Among 10- to 24- year-olds, the [suicide] rate declined from 9.24 to 7.21 suicides per 100,000 people from 1991 to 2009.

So what gives?  As you probably guessed, if you read my other post, the answer is in Hawksworth's sources, which (as quoted in my previous post) reported that:

8. Youth (ages 15-24) suicide rates increased more than 200% from the 1950’s to the late 1970’s. From the late 1970’s to the mid 1990’s, suicide rates for youth remained stable and, since then, have slightly decreased.

The reference to a slight decrease is consistent with what Finkelhor says, although it's interesting that his study is ages 10-24, whereas Ms. Hawkworth's source was 15-24.  

None of this is to diminish the tragedy of suicide and the importance of suicide prevention.  I read about high school and college kids committing suicide all the time.  And my kids have known kids that have committed suicide.  I knew someone in high school who did.  So it's not all that uncommon. But it's good to know that despite everything we hear and read, it seems to be on the decline.

Back to Finkelhor.  Here are the other good statistics he cites, which are all worth keeping in mind:

·        Arrests for serious violent offenses by juveniles have dropped about 60 percent from 1994 to 2011  The juvenile arrest rate has been going down faster than the adult arrest rate over the past 10 years, and juvenile property crime is at its lowest point in 30 years.

·        The number of sexual assaults against 12- to 17-year-olds has dropped by more than half since the mid-1990s, and the number of youth arrests for sex offenses has dropped as well, according to three corroborating nationwide and statewide victim surveys.

·        Violent victimization of teenagers at school has dropped 60 percent from 1992 to 2012.

·        School homicides in the 2000's have been lower than they were in the 1990s.

·        Most surveys show that peer victimization, harassment and bullying have been abating.

·        The teen pregnancy rate is at record lows in the United States.

·        The percentage of ninth-graders who say they have had sexual intercourse has declined from 54 percent in 1991 to 47 percent in 2013, and the percentage of high schoolers who say they have had four or more sexual partners also has declined.

·        Alcohol abuse is also down, with binge drinking by 12th graders at its lowest point since 1976 , and the percentage of students who have been drunk in the past year also at a record low.  According to one survey, high school students drank and drove in 2011 at  half  the rate they did 1991.

·        56 fewer kids ran away from home in 2012 vs. 1995

·        Dropout rates among those ages 16 to 24 are at their lowest, down from 17 percent in 1968 to 6.6 percent in 2013.

I haven't checked these statistics, but they sound plausible enough, and it's good news that is worth repeating.  

I see I didn't even link to Ms. Hawksworth's Sept. 18, 2014 article in my original post.  Here's that link,  And for the record, I see that the article still hasn't been corrected -- it still reports that 1.5 out of every 100 college students commits suicide, and that college suicide rates have increased by 200% since the 1950s.  I guess the editors of the Washington Post don't read this blog.

Please don't get me wrong -- I very much approve of Ms. Hawksworth's message and the work she is doing to spread it.  I just wish she'd be more careful with her use of statistics. A Washington Post article stays around for a long time, and it will keep getting hits, and will keep misleading people, until it's fixed.

Final insight:  Conspicuously absent from Mr. Finkelhor's data is any reference to marijuana use among children.  It occurred to me that perhaps marijuana use has gone up, and that THIS is the explanation for the other good results across the board.  After all, marijuana is often easier for kids to get than alcohol (with alcohol, SOMEBODY over 21 has to be in the picture to buy it). So let's see:

• Alcohol use has gone down because marijuana is a substitute.
• Violence has gone down because marijuana makes one mellow.
• Same for victimization, harassment and bullying.
• Property crime has gone down kids are home getting stoned, not out committing crimes against property (not sure just what property crime is; maybe it's theft).
• Teen sex and pregnancy is also down, because of marijuana's effect on sex drive and sperm count.
• Kids run away less, because the marijuana keeps them happier.
• Kids commit suicide less because marijuana helps stave off depression, at least to some extent. And of course, if many child "suicides" are actually the result of auto-erotic asphyxiation, then perhaps marijuana use dampens that activity as well.

As alluded to above, marijuana in some cases may well ameliorate depression or other maladies.  It's not a substitute for anti-depressants, but if the condition is undiagnosed, then perhaps self-treatment with marijuana is better than no treatment at all. I looked at the marijuana use statistics reported by the CDC (Finkelhor's source for his alcohol data), and it's a bit unclear.  By some measures, one could say that marijuana use has decreased since its high point in the late 1990s, but by others, perhaps it has increased (it went from 31.3 to 39.9 between 1990 and 2011).  So maybe my theory needs some refinement -- while marijuana use has stayed steady (apart from the anomalously high years of 1997 and 1999 and the early, low years of 1991 and 1993, the "ever used" answer has hovered around 40%), perhaps the quality of it has gone up so much that it has positively affected the other categories.  But I'd really have to look at year-by-year info on the other statistics to try to support these conclusions.

Here's a link to the chart, but since it's important, I'll just reproduce it here (you might have to blow up the page to read it here though):

One final thought.  It's interesting how marijuana use went up during the most immoral period of the Clinton Presidency -- in 1997 and 1999, as revelations about Monica Lewinsky came out, and impeachment proceedings began.  I always thought he set a very poor moral example for kids; maybe this is one piece of real-world support for that theory.  

Let me emphasize that I am fully aware of the correlation = causation fallacy, and I recognize that much more work would need to be done to truly draw the links I am suggesting here.  But perhaps that's the sort of work somebody should be doing.