Number of suicides by mechanism, Australia 2012 (source data: ABS)

Today is World Suicide Prevention Day, so I’m reprising this piece published 10 September last year.


The Board of Directors of the Golden Gate Bridge in San Francisco approved expenditure of $76 million in June 2014 to erect steel suicide nets intended to discourage people leaping from the bridge to their deaths.

The Golden Gate Bridge is an internationally recognised suicide “hot-spot”. In 2013, 46 people killed themselves by jumping from the bridge, bringing the total number of suicides since the bridge opened to more than 1,600.

Barriers erected to deter people jumping off high places like look-outs and bridges are common in cities throughout the world, but they’re a controversial measure.

Suicide barriers have downsides: They can be expensive; they can compromise public access; they can limit views; and they can detract from the appearance of natural and human-made assets.

Some, like the steel net proposed for the Golden Gate Bridge, can themselves be the source of injury. It will be six metres beneath both sides of the bridge; that’s fine for minimising the visual impact compared to other alternatives like a four metre wall, but the net will have limited “give”, so anyone jumping is still likely to suffer severe injuries.

The big argument against barriers, though, is that they’re futile: individuals who’re intent on killing themselves will simply go somewhere else. It would be better, it’s argued, to put the money into measures like mental health and education services that might potentially suppress suicidal motivations.

According to ABS Statistics, jumping (or falling) is also a relatively uncommon way for people to take their own life (see exhibit). It accounted for only 4% of 2,535 suicides in Australia in 2012, compared to 54% by hanging, 23% by poisoning (including intentional drug overdoses), and 7% by firearm. (1)

But the findings of a team of researchers led by Professor Jane Pirkis, Director of the Centre for Health Policy at Melbourne University, suggest that barriers and safety nets lower the total number of suicides by jumping in a city (2).

Prof Pirkis and her team reviewed nine ‘before and after’ studies of structures erected for the purpose of reducing suicides by jumping. Six of the interventions were barriers erected on five separate bridges/viaducts and three examined the effectiveness of barriers/nets at cliffs and drops. (3)

They found the structures significantly lowered the number of suicides at these sites, reducing from a mean of 5.7 deaths per year before the intervention (436 deaths over 77 study-years) to 0.5 deaths per years afterwards (21 deaths over 46.4 study-years).

As expected, there was an increase in the number of jumping suicides at other sites in the same cities, but it was significantly lower than the decrease at the prime sites.

Following the interventions, there was an 86% reduction in jumping suicides per year at the sites in question. There was a 44% increase in jumping suicides per year at nearby sites, but the net gain was a 28% reduction in all jumping suicides per year in the study cities.

These studies can’t identify if those deterred by barriers used other means to suicide or if they travelled to other cities to jump. However, like controls on access to guns ownership, it’s thought that restricting the means of suicide buys time to allow “the individual to reconsider his or her actions, and give others the opportunity to intervene”.

2010 paper by Andrew Leigh and Christine Neill supports this theory. The authors concluded that the gun buyback program instituted in Australia in 1997 “led to a drop in the firearm suicide rates of almost 80 per cent, with no statistically significant effect on non-firearm death rates”. (4)

Pirkis et al say restricting the means may be particularly effective in circumstances:

where the individual is ambivalent about his or her wish to die (low intent) or is acting rashly (high impulsivity). Evidence from studies of survivors of jumping suicide attempts indicates that these individuals rarely go on to die by suicide…

While jumping isn’t a common way of committing suicide, the authors argue it justifies expenditure on barriers because the likelihood of dying (lethality) is high compared to more common methods. It also has a high impact on the mental wellbeing of witnesses. The capital costs of barriers are usually high, but the ongoing costs are relatively low.

There’s a similar debate around fencing rail lines to reduce suicides on the train system, where the costs of suicide are even higher. The ABS doesn’t identify the number of rail suicides in Australia but the Tracksafe Foundationclaims there are around 150 on the rail network each year (5).

These are known to cause severe trauma for some train drivers, possibly because they feel they’re an active participant. They can also cause prolonged delays that affect large parts of the rail system.

Erecting barriers presents particular problems for rail systems because of the length of track and, in particular, the number of points of access; for example, in Melbourne there over 200 stations and around 170 level crossings.

It appears that restricting the means of suicide can be effective in reducing the number of deaths in some circumstances, e.g. suicide “hot spots” like high bridges. Further, it seems the benefits justify the financial costs.

However when it comes to reducing the great majority of suicides, including many of those on the rail network, interventions that reduce the motivation to suicide seem the most plausible.

Update 24 September 2015: Blocking the means of suicide can buy time and save lives, The Conversation.


  1. The ABS records the average number of suicides in Australia by falling between 2003 and 2012 was 104. There were 112 in the most recent year, 2012. The peak was 138 in 2008 (ABS, 3303.0, Causes of Death, Australia, 2012). Males are much more likely to die by suicide than females.
  2. Jane Pirkis, Matthew J Spittal, Georgina Cox, Jo Robinson, Yee Tak, Derek Cheung, David Studdert, The effectiveness of structural interventions at suicide hotspots: a meta-analysis.
  3. The sites studied were: Grafton Bridge (Auckland, New Zealand), Clifton Suspension Bridge (Bristol, UK), Ellington Bridge (Washington, DC), Memorial Bridge (Augusta, ME), Bloor Street Viaduct (Toronto, Canada), Lawyers Head Cliff (Dunedin, New Zealand), Beachy Head (Sussex, UK), Muenster Terrace (Bern, Switzerland).
  4. They estimated effect of the gun buyback on firearm homicides was of a similar magnitude, although less precise.
  5. The Tracksafe Foundation claims there are a further 1,000 rail-related suicide attempts each year, nearly half of which result in serious injury.