Railway suicides

​Railway Suicides
As a railway commuter, I have experienced multiple delays due to suicides, and enough to get some sense of their volume and patterns. The one time I was on the train that hit the person, there was an amazing flourishing of a sense of community, with people hugging the driver and conductor when we got off at Manchester Piccadilly. I talked to the conductor a little, while we were waiting, with discussion of the time patterns (Mondays, the week after Valentine’s day, the end of January) and the potential causes for this, which include the arrival of bank statements and credit card bills, broken relationships, jobs and so on.
While thinking about this, I did hunt out some statistics, only to find more of the Guardian’s failures to really grapple with numerical data. A datablog on railway suicide has some representations of data with arguments about what is going on, almost all of them wrong.
The first is a chart showing an increase in railway suicides over a 10-year period, alongside data on passenger journeys. I’m not even sure any increase isn’t just random variation: because the numbers are so small, what looks like a big percentage change is actually what we’d expect annual variation to look like. Saying that there’s been a 17% jump from 2010 to 2011 ignores the fall of a similar size from 2009 to 2010. And I’m not sure what the passenger numbers have to do with suicides: people of all sorts can go to a station or more usually a bend where the driver won’t see them to commit suicide, it’s not just a subset of passengers.
The data does show ‘Fewer suicides on the underground’, and this is argued to be ‘more indicative of lower passenger numbers than station safety’. However, passenger numbers on the underground are at the same order of magnitude as on the national rail (1.4bn v 1.6bn) so this would not explain suicide on the tube being a tenth of that on overground rail. Again, I’m not sure what passenger numbers have to do with it: suicides come from the population of humans, not from the population of train commuters or tube riders (although there could be a correlation due to confounding variables). More likely is the design of the tube system. On the national rail system, it is possible to climb a fence and get onto the track, somewhere where a train cannot stop, or to step off the end of a platform where a train is passing through. All tube trains stop at all stations, and most of the time we can’t get to the track in between stations, so attempting to commit suicide on the tube is far more difficult.
The final chart titled ‘One of the highest rates in Europe’ does not show this at all. Yes, we are near the top of the chart, with more railway suicides than all EU nations bar Germany and France. But this is because we are one of the most populous nations. ‘The UK's rail network accounted for around 7% of all suicides that took place on EU rail’ but this needs to be put into the context of having just under 9% of the EU population. A rate needs to be worked out as suicides per population. We’re mid-table here, with the Catholic countries lower than us, but many others higher.