It being almost Christmas, one’s thoughts naturally turn to homicide. They were helped in this direction, at least in my case, by following a link over at The Flying Rodent’s place; he’s been moved to poetry by another blogger and I wondered what had so irritated him.
Part of the cause of his ire was a discussion about capital punishment, in which the old point about the ‘murder rate’ (I’ll explain the scare quotes in a moment) going up since the suspension of capital punishment in 1965 and its eventual abolition in 1969.
Since I know that crime statistics are notoriously difficult to interpret, I thought I’d do a little more digging. Eventually this took me to the oldest blog entry I’ve ever seen, an archived entry from 1996 by Tim Lambert of Deltoid. This gives the Homicide (or, more correctly, crimes initially recorded as Homicide) rates for England and Wales from 1857 to 1993. I’ve copied them into a spreadsheet and supplemented them with the more recent figures, which I took from Table 2.01 in the Home Office Statistical Bulletin, Violent Crime Overview, Homicide and Gun Crime 2004/2005.
The spreadsheet, should anyone wish to consult it, is here. One slight caveat to anyone consulting it; my Home Office figures give the rates for offences currently (as opposed to initially) recorded as homicide per million of population. Consequently, I had to calculate what population figure they were using for the relevant years and then use that to calculate what the rate was for offences initially thus recorded; for some years, my figures aren’t quite the same as Tim Lambert’s — I’m not quite sure why, but there’s not a great difference and I don’t think it alters the overall picture.
I’m not really trying to use the figures to argue anything, since my opposition to the death penalty isn’t particularly founded on stats, but I thought it worth making a few points.
First, homicide covers not only murder, but also manslaughter and infanticide (the last comprising between 1.86% and 5.89% of recording homicides between 1994 and 2005). The Home Office stats give a breakdown of what the crimes were eventually decided to be at court — murder, manslaughter by reason of diminished responsibility, other manslaugher and so on — from 1994 onwards (table 2.02), but since we’re dealing with the way offences were initially recorded, we don’t have that information (at least I don’t) for cases before 1994. Consequently, the bare figure for Offences Initially Recorded as Homicide doesn’t tell us anything about what the offences were eventually determined to be or, in the cases of murder, the circumstances of the murder.
This last point doesn’t affect the figures, of course, but it’s worth recalling that ‘murder’ can mean a cool, premeditated execution killing, a killing committed in the heat of the moment, and also — which people don’t always realise — the results of a serious assualt that ended up far more serious than anyone expected. Consequently, in an example provided by the Law Commision in their recent report on Murder, Manslaughter and Infanticide, (p 4) if
D intentionally punches V in the face. The punch breaks V’s nose and causes V to fall to the ground. In falling, V hits his or her head on the curb causing a massive and fatal brain haemorrhage
This would be murder if the jury decided that the harm that D intended the punch to cause (the broken nose) can be described as ‘serious’.
Parliament apparently didn’t think this was murder when they passed the Homicide Act 1957, but the Court of Appeal decided it, in fact, was very shortly afterwards (see the Law Commission Report, paras 1.26 — 1.29).
Nor, returning to the figures, do we know anything about what the recording practices were in the past. That certainly seems to vary; In 1967, 14.49% of cases initially recorded as homicide were eventually determined to have been something else, while in 2003/4 it was down to 7.68%. Does that mean they were overly suspicious in 1967 or that we’re nowadays unduly willing to go along wth the initial determination? I don’t know.
However, which things having been said, it’s certainly the case that recorded homicide rates have doubled from the 1950s and 1960s, when they were 0.7 per 100,000 of population, to 1.4 per 100,000 of population in the 1990s.
That, though, is only half the story; if we go back further in time, it seems that — for whatever reason, but it can’t be capital punishment — the rates in the 1950s and 1960s were particularly low: the mean average rates per 100,000 for each decade were
However you interpret the figures, and I’m not going to try, it’s obviously a mistake to try to infer that because the death penalty was abolished, the murder rate went up, because that leaves you at a loss to explain why it went down from the 1860s onwards. It’s always a mistake to infer causality from correlation, of course; I once saw an elegant demonstration of how you could quite closely correlate the number of television licences with the homicide rate, too.
Obviously, something as complex as homicide has a multitude of causes; as a guess, and it’s only that, it might be worth looking at the relative price of alcohol as one, but only one, factor, given the role alcohol seems to play in many crimes of violence (see Home Office stats, ‘Offence Profile’ pp 21-22).
tag: Murder Rate
, England and Wales
, Home Office Statistics