Billy and the Golden Gate Bridge
“The Golden Gate Bridge is practically suicide proof. Suicide from the bridge is neither possible nor probable” –Chief Engineer Joseph Strauss, 1936
Billy was an 18-year old high school student. He attended a private school, achieved a 1400 on his SAT, and was accepted on early admission for college. He did not drink, smoke, or do drugs. He told his mother, “I’m not the most popular kid, but there’s no one who doesn’t like me.” On this particular week, Billy was behind on a school paper about the Challenger disaster. He told his mom that he couldn’t stop thinking about death. His mother thought it was odd, but nobody in his family thought Billy might be depressed. The following day, Billy committed suicide by jumping from the Golden Gate Bridge. The entire student body attended his funeral.
Billy’s profile fit none of the predictive factors for adolescent suicide. When his friend was asked why he thought Billy might have killed himself that day his friend responded, “The only explanation I can find is that he never stopped to think it through. He didn’t tell anyone and he didn’t leave a note.”
Such Great Heights: Suicide Barriers and Firearms Restrictions
The Golden Gate Bridge opened up in May 18, 1937, a technological marvel marrying strength and beauty. Less than three months later, the first known suicide took place. Since then, over 1,500 people have taken their lives by jumping off the bridge, making the site one of the most popular suicide sites in the world. Calls for the construction of an anti-suicide barrier, as have been erected in other popular locations such as the Eiffel Tower, the Empire State Building, and Arroyo Seco Bridge, have been rejected by policymakers who have argued that the type of people who attempt suicide will “just go somewhere else.” Bridge or no bridge, they argue, a suicidal person will be successful.
A famous study sought to challenge this contention, tracking 515 persons who were restrained from committing suicide on the Golden Gate Bridge. Was it the case that people who failed to commit suicide simply sought recourse to other means? Definitively not. Ninety percent of people who had been prevented from jumping off the Golden Gate Bridge did not go on to commit suicide. The findings confirmed the idea that suicides are often ‘crisis-oriented’ and impulsive in nature. In many of these cases, a failed suicide lead to reflection on the preciousness of life, facilitating a renewed commitment to living. The less lethal the means of suicide, then, the higher the likelihood that this life-affirming moment of reflection will happen.
Numerous studies have confirmed the point that suicide barriers are useful deterrents: when barriers were built on the Duke Ellington Memorial Bridge in Washington, D.C., the Augusta River Bridge in Maine, the Muenster Terrace in Bern, Switzerland, and the Clifton Suspension Bridge in Bristol, England, suicide attempts on those structures dropped markedly. Total suicides in these areas dropped as well, corroborating that barriers did not simply divert suicide attempts elsewhere.
The data, in each of these cases, show that the motivations for suicide are much more complex than intuitively appealing ideas of inevitability, and are subject to very basic policy interventions. The belief that suicide always involves a meticulous planning process, one that inexorably leads to a successful suicide, is not only categorically incorrect but dangerous because it impedes the most useful strategies for preventing suicide—namely, means reduction.
The reason, of course, this is being posted on a blog about guns is that precisely the same bad arguments forwarded by policymakers to challenge the construction of suicide barriers are being used by the National Rifle Association (NRA) and conservative politicians to challenge gun regulation. And they couldn’t be more wrong.
If barriers on bridges can significantly reduce the suicide rate, why can’t barriers to accessing firearms do so well?
Debunking “They’ll Do It Anyway” – The Impulsiveness of Suicide
Study finds the impulsiveness of suicide is related to a “disability of the suicide attempters to regulate their serotonin and dopamine levels, e.g. in response to external stress” (Ryding, 2006).
Before we delve into this question, I want to make it clear precisely why ‘means reduction’ strategies are effective at reducing overall suicide rates.
An impulsive suicide is one for which there is very little preparation prior to a suicide attempt. A 2001 study, using the Beck’s Suicidal Intent Scale, examined 478 individuals who attempted suicide, and found that more than half (55%) of attempts could be classified as ‘impulsive’, while only about one-sixth (17%) of attempts were premeditated. Other studies use time criteria to temporally operationalize impulsiveness. One such study found that 40% of suicide attempt survivors contemplated suicide for less than 5 minutes before the attempt. In addition, the most recent data set from the National Violent Injury Statistics System reported that 61% of suicide victims had not indicated intent to commit suicide to friends or family members prior to the attempt.
If we are interested in decreasing the suicide rate (which we should be), then we ought to invest in policies that decrease the probability than these transient moments of impulsiveness will have lethal consequences.
If it were the case that suicides are the inevitable byproduct of some chronic mental or environmental issue, we should expect forecasting a future suicide may be possible. However, research has found that, because of suicide’s inherent impulsivity, statistical prediction is nearly impossible.
An American study of 4,800 veterans admitted in psychiatric facilities in Houston found that, “we do not possess any item of information or any combination of items that permit us to identify to a useful degree the particular persons who will commit suicide.” A suicide prevention researcher substantiated this claim, arguing that, because of the low base rate of suicide one “would need a test of unbelievable sensitivity and specificity to be of use.”
It should be clear, then, that there’s nothing “inevitable” about a suicide, nothing predictable about impulsiveness. To turn a blind eye to suicide based on the pretense that they’ll “just try again” demonstrates a profound ignorance of the psychology of suicide, and a callous unwillingness to consider the struggle of another human being.
The Causal Relationship Between Firearms and Suicides
The overview: Dr. David Hemenway of the Harvard School of Public Health summarized ten studies in the previous twenty years examining the relationship between gun ownership and suicide and found that “all [of them] find that firearms in the home are associated with substantially and significantly higher rates of suicide.”
Furthermore, every single case–control study done in the United States has found the presence of a firearm is a strong risk factor for suicide. (That’s 24 separate studies).
The meat: The latest available data on suicide rates, published by the Centers for Disease Control, shows that 38,364 suicides occurred in the United States in 2010—an average of 105 each day. This makes suicide the tenth leading cause of death for all age groups in 2010. Males are particularly at risk for suicide, given that men are four times more likely to commit suicide than females, thereby representing 79% of all suicides. Among males, firearms are the most commonly used method, being used in 56% of all suicides.
For every suicide, there were more than ten hospitalizations for injuries, which resulted in an estimated cost of $6.5 billion in medical expenses and work loss. State mental health agencies (SMHA) spent $37.3 billion in order to confront this problem. There is significant evidence to suggest that, with respect to suicide, guns increase the likelihood of a ‘successful’ attempt—this is referred to as an ‘instrumentality’ effect. If this is the case, then it’s not unreasonable to think that the overall firearm ownership rate in an area increases the suicide rate because less efficacious means would be substituted for guns. An alternative mechanism by which gun availability might influence suicides is by increasing the number of suicide attempts—that is, a suicidal person may avoid committing suicide in the absence of gun availability because they are concerned about failing. In addition, it may be the case that the presence of a weapon primes aggressive thoughts which may encourage suicidal impulses in an otherwise stable person.
Several studies have investigated the effectiveness of gun regulation laws in attenuating the suicide rate, with many arguing that gun restrictions should be one of the key elements in suicide reduction. Even though this is well known among public health officials, gun debates are often mired in the question of how gun regulation affects homicide and crime rates, but there is very little discussion in policy circles about the enormous effect gun regulation could have on suicide rates. This is true despite the fact that “more guns = more suicides” is the closest we will ever get to an incontrovertible statement in the gun debate.
Two theoretical caveats problematize the assessment of causation, and should be mentioned before delving into the literature on this topic. First, though a correlation exists between gun ownership per capita and suicide, there may be a self-selection bias. That is, that the types of people who are likely to own guns might be predisposed to suicidal behavior. It could also be the case that some exogenous change in the environment decreased suicide rates and gun ownership concomitantly. A more liberal Congress, for example, might implement both gun regulation policies and suicide prevention policies, producing the illusion of causation between these two variables. Both of these issues have been dealt with through statistical techniques and robust research methods, and the data clearly show that the direction of causality runs from guns to suicides.
A 2000 paper by Ludwig and Cook estimated whether declines in suicides over the period 1985-1997 were associated with the passage of the Brady Handgun Violence Prevention Act. The Brady Act required that federally licensed firearms dealers perform a background check and implement a five-day waiting period prior to the sale of a handgun. Eighteen states and the District of Colombia already satisfied Brady requirements, while the other thirty-four states required more stringent procedures. Therefore, researchers observed a natural experiment in which the states that already met Brady requirements were considered a ‘control’ group, and were compared against states in which dealers and law enforcement officials had yet to implement requirements. The study found that the legislation produced a significant reduction in suicide rates among persons aged 55 or older, suggesting that suicidal impulses in older individuals were attenuated by the imposition of the five day waiting period, thereby decreasing the suicide rate.
A 2006 paper published by Miller and colleagues at the Harvard School of Public Health explored changes in household firearm ownership in the United States over the period 1981-2002 as it related to a decline in the suicide rate, controlling for age, unemployment, per capita alcohol consumption, and poverty rates.
Household gun ownership levels and rates of firearm and non‐firearm suicide mortality: United States, 1981–2002 (Miller et al., 2006).
The study found that, for every 10% decline in the household firearm ownership rate, firearm suicides decreased by 4.2%, and total suicides dropped by 2.5%. The decline in suicide rates was highest among children. Examining just households containing both children and firearms, every 10% decline in the percentage of households owning a firearm was related with a 8.3% drop in the suicide rate for individuals between the ages of 0-19. Gun-regulation opponents alleging that this study is merely correlation should have to point out some covariate not accounted for in this analysis that could simultaneously explain why changes in the firearm suicide rate are related to changes in firearm ownership rate, but not related to the non-firearm suicide rate (as might be the case with an exogenous environmental change). Furthermore, the largest study done to assess mental health trends in the United States over the period investigated by the paper found that there was no significant change in suicidal tendencies between 1990-2000. We therefore have evidence that, even after holding psychological tendencies constant, the presence of guns has a dramatic effect on the suicide rate.
A 2011 study published in the Journal of Health Policy related suicide rates and gun regulations using CDC data in states over the period 1994-2004. Gun regulation was operationalized through three additive indices: (1) the absence or presence of licensing requirements or prohibitions on the sale of firearms to minors; (2) firearm prohibitions based on behavioral patterns statistically linked to suicide such as alcoholism, drug problems, and mental health; (3) and an index that captured four different types of prohibitions that are related to the purchaser’s criminal history whether they be illegal immigrants, convicted felons, fugitives from justice, or those with a history of juvenile crime.
Because firearm prevalence is strongly related to firearm regulation, the study controlled for gun ownership by adding an instrumental variable—the number of hunting licenses per capita from the Fish and Wild Life Service– as a proxy for gun availability into their model. A negative binomial regression was used to estimate the effect of regulations on suicide.
The research found that prohibitions on firearm sales to minors and permit requirements were the most effective in terms of reducing suicide, with a statistically significant result even after controlling for gun prevalence and socioeconomic variables. As twelve states continue to allow the sale of long guns to minors, with the same amount requiring permits for the purchase of firearms, regulations in these areas can prove to have an immediate, significant impact on reducing suicide rates.
A 2010 study by Lubin et al. studied the effect of a policy change in Israel which reduced access to firearms for adolescents in the Israeli Defense Forces’ (IDF). Given that many IDF soldiers returned home on the weekend with their firearms, and 90% of all suicides in Israel are committed with a firearm, the mandate of the legislation required that IDF soldiers leave their weapons at the base when returning home for the weekend. Suicide rates between the years 2003-2005 were compared with the post-policy change rates of 2007-2008. The study found that the total suicide rate decreased by 40% as a direct result of the policy change, without a compensatory increase in the non-firearm suicide rate. This fact suggests that guns are somehow unique in exacerbating one’s suicidal potential—it’s not the case that a suicidal person will end his or her life irrespective of the means available, but that guns both increase the rates of impulsiveness and success.
A 2010 study by Mathieu Gagné and his colleagues examined whether stronger firearm regulations implemented in the Canadian province of Quebec in 1991 influenced suicide rates among men. In the 1990s, the rate of suicide among men in Quebec reached an all-time high, and since then has been markedly declining. The decline in suicide rates coincided with the implementation of Bill C-17 which, among other things, required firearm license applicants to provide a photograph and two references; imposed a 28-day waiting period; mandated a safety training requirement; and addressed safe storage practices. A previous study done in 2008 found no relationship between the regulations and suicide rates, but it suffered from serious methodological errors, including a prohibitively small time-frame for analysis, and statistical methods that did not take into account the lag time between the implementation of the law and behavioral changes.
Suicide rates by firearms, hanging and total suicide rates among men, Quebec, 1981–2006, according to methodological choices. (Gagné et al., 2010).
The study found that, five years after C-17’s implementation, suicide rates among young men decreased by 11.1%. Rates declined similarly, though to a lesser extent in men aged 34-65. The study found that there was also a decrease in other methods of suicide such as hanging and gas poisoning, refuting the argument that decreases in firearm suicides would cause concomitant increases in suicides achieved by other methods. The less pronounced decline in the suicide rate for elderly populations was likely due to the fact that older men were already likely to have a securely stored firearm, and thus were relatively unaffected by C-17 specific regulations.
Another study published in 2007 in the British Journal of Psychiatry examined the effect of suicide and homicide rates in Austria before and after the implementation of firearm regulation mandated for all member states of the European Union. The regulations, which were passed in July 1997, required all firearm purchasers to specify a reason for their need of a firearm; it implemented background checks and psychological testing for all people 21 years or older wishing to obtain a handgun, semi-automatic firearm or repeating firearm; it specified safe storage procedures for firearms; and implemented a 3-day waiting period for long firearms with smooth bore and rifled barrels.
Firearm suicide rates before and after the 1997 firearm legislation (Kapusta et al., 2007).
The study looked at suicide and homicide rates 12 years prior and 8 years after the passage of these regulations. The paper found that:
“[the] firearm suicide rate decreased among women aged 20–64 years, men aged 20–64 years and men aged 65 years or older; firearm suicides as a percentage of total suicides decreased; the firearm homicide rate decreased; and the overall firearm licence rate decreased after enactment of the new law. These results hold true even when adjusting for common confounders of suicide rates such as unemployment and average alcohol consumption per capita as well as the proportion of young men in the population”
Debunking the Mark Duggan Paper: Why Guns Cause Suicide, and Not the Other Way Around
A 2002 paper by Wharton Economist Mark Duggan found that there is at least some correlation between suicidal tendencies and gun ownership that is potentially driven by the fact that gun owners may have higher-than-average suicidal tendencies. It may be the case, for example, that gun owners live in communities that place a high value on self-reliance and personal security, and are thus less likely to seek help in times of distress. If this is true, then it would complicate findings that show an association between suicide and gun ownership.
However, more recent analysis on this question found that higher rates of firearm ownership is positively associated with higher suicide rates, but there is no significant association between the firearm ownership rate and non-firearm suicide rates. If it were the case that firearm owners are more likely to commit suicide, we should see a relationship between firearm ownership and suicides by other means, but we don’t. This is true despite the fact that non-firearm methods like drugs and cutting constitute approximately 90% of all suicide attempts.
A 2012 follow-up study published in the American Journal of Epidemiology thoroughly repudiated the hypothesis put forth in the Duggan paper. The paper finds there are at least five reasons to believe that the presence of firearms is driving the rate of suicide, and not the the other way around:
- The association between firearm availability and suicide is robust to adjustments for measures of psychopathology and aggregate-level measures of suicidality such as depression, mental illness, alcoholism, poverty, unemployment, and drug abuse.
- The risk of suicide extends beyond just the gun owner to all members of a household, and lasts for years after the firearm has been purchased.
- The rates of psychiatric illness and suicidal tendencies is similar in households with and without firearms across the United States.
Multiple ecological studies have confirmed the results of individual-level studies.
- Suicide attempts are not significantly associated with firearm ownership rates. If it were the case that gun owners had stronger suicidal proclivities than non-gun owners we would expect the suicide attempt rate to be positively associated with the firearm ownership rate, but it isn’t. This means that the primary way through which firearms influence the suicide rate is by making each attempt comparatively more lethal than other methods.
The conclusion of the study is damning:
“The prevalence of household ﬁrearm ownership, which ranges from 10% to 66% across the 50 states, explains 67% of the variation in ﬁrearm suicide, 42% of the variation in overall suicide, and less than 2% of the variation in nonﬁrearm suicide. By contrast, suicide attempt rates, which range from 0.1% to 1.5%, explain less than 1% of the variation in rates of overall suicide, ﬁrearm suicide, and nonﬁrearm suicide. Indeed, suicide attempt rates are not signiﬁcantly related to suicide mortality rates overall or by method, even in crude comparisons.”
The study adds that, even small decreases in the overall firearm ownership rate will have a large impact on the suicide rate. For example, if 1% of the 22,000 people who attempted suicide in 2010 had instead substituted drugs or cutting, there would have been approximately 1,900 fewer suicide deaths.
Note also that although Duggan found that gun owners have stronger suicidal proclivities, he still posited that instrumentality effects influence the suicide rate. That is, Duggan noticed that the male-to-female suicide rate was systematically determined by the rate of gun ownership in such a way that it was likely that guns are having an effect on suicides:
“…there is a positive relationship between the rate of gun ownership and the male/female suicide ratio. The coefficient estimates…reveal that the suicide rates among males are more strongly related to the rate of gun ownership than are suicide rates among females. Specifically, a 10 percent increase in gun ownership is associated with a 7.9 percent increase in the male suicide rate but only a 5.2 percent increase in the female suicide rate. This difference is statistically significant…”
“Gun Control is a Band-Aid Solution”
“While gun-related violence is on a continual, steady decline, suicide deaths are being used, in part, because they are needed to ‘get the job done.’ People shout out for pity and compassion for the survivors of suicides in a call to infringe on the rights of lawful, peaceable gun owners in hopes, they say, of preventing future incidents of suicide…the gun grabbers are reaching, many of them deifying their lost loved ones before God and country as if, somehow, the dead person wasn’t responsible — that it was the “evil gun” that caused the act. And some people actually buy it, hook, line and sinker.” – Angel Shamaya
Our victim is on the ground bleeding. He has just been shot in the head by a gang member, a bullet lodged firmly in his brain. He is in agony, suffering the most excruciating pain of his life. An emergency medical team is dispatched to the scene. The victim, delirious with pain, begs for relief. The EMT responds, “How would that solve anything? Masking the pain or regulating blood flow won’t dislodge the bullet…” The victim slurs a plea “does it not matter that I’m suffering?” The EMT continues, “What we need is economic reform—we need to challenge the root causes behind gang violence to ensure this doesn’t happen again. Otherwise, you’ll just end up back in the emergency room with another bullet in your head and that’s just more the taxpayers have to…” Before the EMT can finish, the victim flat-lines.
If this scenario sounds ridiculous, it’s because it is. Obviously, it was the EMT’s responsibility to stop the pain, even if future gang violence is likely to land our victim back in the hospital. If there is an opportunity to efficiently and drastically reduce human suffering, then we have an obligation to pursue that opportunity. We have no problem accepting the fact that the EMT’s obligation is to relieve pain, and yet, when it comes to gun violence, somehow our moral intuitions are severely handicapped— gun control legislation to reduce suicides is consistently refuted with the argument “we need to treat the root causes of suicide.” How noble of these people to oppose exceedingly basic policy interventions, which can be implemented with relative ease, and that are already being run successfully virtually everywhere else in the developed world, so that gun control can be staved off for a couple more generations.
The data above clearly show that extraordinary decreases in the total suicide rate can be affected by even small reductions in the firearm ownership rate. Such efforts, then, are very clearly not a band-aid. Even if they were, I’m not convinced that the phrase should have disparaging connotations—an incredibly cheap, effective step that minimizes pain and enables day-to-day functioning should be the first solution sought to any problem.
Further, the above data should make it clear that there is no ‘root-cause’ behind suicide—there are an inestimable number of unobserved variables that may influence suicidal ideation, which society, and often the suicidal individuals themselves, do not have conscious access to. Impulsivity, almost definitionally, escapes prediction—we may not have the tools to challenge or understand the emergence of impulsiveness, but we can minimize its consequences.
There is nothing closer to an academic consensus in the gun reform debate than the argument that restricting access to firearms can drastically decrease the suicide rate. If this is the case, why is there such little discussion about suicides in top-level policy debate? The discussion always centers on homicides, mass shootings, crime, and so on, but never suicides.
I received a comment from a reader on an earlier post in which he refuted an entire section with the claim that “suicides don’t count.” I suspect that, though it’s rarely stated this explicitly, this is how many people process suicide statistics. Homicides are scary—it could be my life, after all, that someone else takes in a crowded theater. But suicides are not newsworthy—they’re distant, routine; it’s somebody else’s brain that’s receives a bullet, not mine.
Discussion about suicide should be at the forefront of gun control debates, yet it is often a footnote in meaningful policy discussion. There is a missing movement against suicide, and it reflects poorly on our nation’s priorities–it shows a cruel insensitivity to the value of human life, and a miscalibrated sense of morality which says that change is only worth having if it can benefit me.