Facts, Lies and Gun Crime statistics

At the risk of getting right up Dizzy’s nose (yet again) it’s worth pointing out that unlike some bloggers I personally make a point of never taking statistical information provided by politicians, the media (or in some cases, other bloggers) at face value without first checking the original source information.

Call me Mr Picky, but I prefer to deal in the facts and not in conjecture, misinterpretation and propaganda, and if I’ve learned one this over the years it’s that the reporting of statistical information is rather akin to a game of ‘Chinese Whispers’ – the more people it’s been through before it gets to you, the more likely it is to have been overlaid with bullshit before it arrives.

So, on those occasions where Iain Dale starts ‘bigging up’ Dizzy with words to the effect of:

THIS post demonstrates why Dizzy is a must read blog.

And one follows the link to find that the starting point for Dizzy’s commentary is a graph lifted off the BBC’s website and third-hand report in the Times about the second-hand use of statistical information by a politician; ‘Basher’ Davis in this case…

Given that the BBC’s own graph showed the reality of firearms crimes (excluding airguns) whilst maintaining that gun crime “overall” is down, it really doesn’t surprise me that David Davis has written to Jacqui Smith pointing out that the Government is basically lying about the state of firearms crimes, and in particular firearms homicide.

The Home Office’s own statistics show that gun-related killings and injuries (excluding airguns) has increased for a factor of four since 1998. David Davis letter to Jacqui Smith quite rightly points out that to therefore say that gun crime is down (simply based on a reduction between this year and last, is both “inaccurate and misleading”.

…then one knows immediately that only sensible thing to do is check the facts before going any further.

The figures at issue here are those specifically relating to, as Dizzy notes, ‘gun-related killings and injuries (excluding air guns), which appear to show a fourfold increase since the total number of injuries since 1998/9:

THE government was accused yesterday of covering up the full extent of the gun crime epidemic sweeping Britain, after official figures showed that gun-related killings and injuries had risen more than fourfold since 1998.

The Home Office figures – which exclude crimes involving air weapons – show the number of deaths and injuries caused by gun attacks in England and Wales soared from 864 in 1998-99 to 3,821 in 2005-06. That means that more than 10 people are injured or killed in a gun attack every day.

The Times has obtained its statistics from ‘Homicides, Firearms Offences and Intimate Violence 2005/6’, a Home Office statistical bulletin published as a supplementary volume to the main crime statistics for England and Wales, and are derived specifically from ‘Table 2b Crimes recorded by the police in England and Wales in which firearms were used by degree of injury’ (pp 36), which is given below (excluding data on injuries from air-guns).














































% increase






And as is obvious, what the Times is quoting as its ‘more than fourfold’ increase in gun related killings and injuries is derived from the total figures for all ‘injuries’ (342%), which does indeed show a more than fourfold rise in the total number of recorded injuries. However, as is also apparent, there is marked disparity in the degree to which this increase is spread across the three different categories of injury – due to a fall in recorded fatalities since 2001/2 there is no actual percentage change between the figures for 1998/9 and those for 2005/6. Serious injuries are up significantly (194% or almost threefold) over the period, while the biggest increase has been in ‘slight’ injuries, which are up by 405%.

To make sense of this data, one needs more than just the numbers – one also needs to understand how injuries are classified into the three given categories.

‘Fatalities’ is, of course, self-explanatory.

‘Serious’ injuries are defined by the Home Office as ones which necessitate detention in hospital, or which involve fractures, concussion, severe general shock, penetration by a bullet or multiple gunshot wounds.

And if you’ve been paying attention then you may have realized, from the inclusion of fractures, concussion and severe general shock in the list of serious injuries, that there’s a little more to firearms injuries than just getting shot; the definition of injury in use by the Home Office also includes injuries stemming from the use of a firearm as a blunt instrument and, in the category of ‘slight’ injuries, instances where a victim has simply been threatened with a firearm (or what appears to them, at the time, to be a firearm) even where they otherwise suffer no physical injury; and it is in the ‘slight injury’ category that there has been the greatest increase since 1998. It’s also the case, as we’ll see in a moment, that a firearm is not necessarily a gun.

If one looks at how the proportion of injuries falling into each of the three categories has altered over time, and as the overall number of recorded injuries has increased, one finds that in 1998/9, fatalities accounted for 6% of all firearms-related injuries, serious injuries came in at 19% and slight injuries at 75%. By 2005/6, fatalities had fallen as a proportion of all firearms-related injuries to a mere 1%, serious injuries had fallen to 12% and slight injuries had increased to 86%. Statistically speaking, there may be four times more recorded firearms-related crimes resulting in an ‘injury’ today then there were eight years ago, but if you are caught up in such an incident, you’re now six times less likely to be killed and around a third less likely to be seriously injured than you were in 1998/9.

To understand why this is the case, you need to look not only at the statistics for injuries in gun-related crimes but also at how the types of weapon used in firearms offences have altered over this same period:








All Weapons

































































% increase








These figures cover all firearms offences, including possession, supply, theft and criminal damage to property, not just those in which a recordable injury occurs, but provide a useful guide to trends in firearms availability/usage that can be read in conjunction with the data on injuries to build a fair picture the overall situation.

So, from the weapons data we can see that, overall, the number of recorded firearms offences has a little more than doubled since 1998, with the number of offences increasing in all weapons categories other than for shotguns.

Handgun are the type of weapon most often used featuring in firearms offences in 2005/6, as they were in 1998/9, although as a percentage of all offences the incidence of handgun-related offences has fallen from 52% to 42% over that time. So far as other classes of weapon are concern, the percentage incidence of shotgun-related offences has fallen by half over this time, from 12% to 6% and other categories are pretty much static, but for offences involving imitation firearms, which now make up 30% of all offences, up from 11% in 1998/9.

To provide a little additional clarity, it should be noted that the vast majority of offences involving imitation firearms (85%) relate to BB guns and ‘soft’ air guns, which are not powerful enough to fall within the regulations covering traditional air-guns but which, since 2003, cannot legally be carried in a public place if they resemble a real firearm. Other ‘weapons’ included in this category are, as might be expected, conventional replica firearms and deactivated weapons. The ‘Other’ category is made up of a rather more eclectic mix of weapons ranging from the relatively harmless (unconverted starting pistols) through weapons with some damage potential (CS gas and pepper sprays, tasers) right up to, thankfully, a very small number of offences involving genuinely lethal kit – there were 133 recorded offences in 2005/6 involving machine guns, which fall into this category.

One must also take into account the fact that the manner in which these statistics were compiled has altered over the period covered by the data, due to introduction, between 2001 and 2003, of the National Crime Recording Standard, which came into force in 2002 but which some forces began to work to from 2001. Under this new standard, the police began to record all reported crimes, irrespective of whether there was any evidence to support the substance of a report – before 2001, a reported crime for which the police could find no supporting evidence, as might happen if someone phoned the police to say that they’d seen an unidentified youth with what looked, to them, like a handgun, would not be recorded in the crime figures if the police could find no evidence to confirm that the reported incident took place. This new standard required the police to record all reported offences, even if no supporting evidence was forthcoming, and resulted in a significant jump in the statistics across all recorded crime as reported offences that would not have previously been included found their way into the dataset.

To illustrate how this will have impacted on these figures, the statistical impact of the new recording standard on offences against the person, one of the categories most relevant to the statistics on firearms-related offences,  has been calculated by the Home Office to have accounted for an apparent increase in recorded offences of 23%. For 2001/2, the year in which this adjustment impacted on the figures for injuries in firearms related offences, this still left the government with an effective increase of 7-9% for fatalities and slight injuries and a worrying 38% increase in serious injuries – not good, but not as bad as the raw figures of 32% (fatalities), 61% (serious injuries) and 30% (slight injuries).

Overall, what the data actually provides is a mixed picture. The trend has been downwards in some areas since 2001/2 – fatalities (i.e. homicides) are down as are offences involving shotguns (and especially sawn-off shotguns) and, perhaps surprisingly, handgun offences, when one looks at all offences. Against that the picture in terms of serious injuries arising from firearms offences is, at best, mixed since the big increase that came with the new recording standards in 2001, and what is certainly increasing is the number of reported incidents in which someone is threatened with what appears to be a firearm, even if the massive increase in offences involving imitation weapons does tend to suggest that in many, if not most, cases, the weapon used is likely to be an imitation. Not that that is, of course, any consolation if one is on the receiving end of such a threat – it’s not as if the assailant is going to hand over their ‘gun’ to allow you to check to whether it’s a fake or not.

So what we have here, in essence, is a situation in which both the government and the opposition can point to some statistical evidence which supports their chosen position, and equally there is other statistical evidence that doesn’t back their arguments up and which they will be keen to downplay. For example, the Times article includes the text of David Davis’s letter to the Home Secretary, Jacqui Smith, in which he states:

According to Home Office figures, gun crime (excluding air weapons) has almost doubled since Labour took office. The annual crime figures, released by the Home Office in July, suggest a 13% decrease on the previous year, which neglects the 18% increase in firearm homicides.

An 18% increase in firearms homicides over the 2005/6 figure of 49 amounts to an addition nine fatalities over the previous year – 58 homicides instead of 49, which is still the second lowest annual figure since 1999/2000 – and if ‘gun crime’ has ‘almost doubled’ since 1998/9 on the strength of this year’s (2006/7) figures, then it’s actually fallen over the last year as the percentage increase for 2005/6 over 1998/9 was actually 113%, more than double. It is also untrue, and deeply disingenuous of Davis, to suggest that the headline rate for the increase/decrease in all crime given in the annual crime figures (actually a 13% decrease last year) ‘neglects’ the 18% increase in firearm homicides. That figure undoubtedly includes the data for firearms homicides in its calculations but, with only 58 such offences over the year the number of such homicides, will have almost no effect whatsoever on the headline crime rate for all crimes – which should tell you everything you need to know about the actual extent to which you are at risk of being murdered with a gun of any description.

Davis goes on to make an allegation that simply doesn’t stand up when one looks at the data.

In light of this information, your claim that gun crime is down is both inaccurate and misleading. One clear fact on gun-related violence is that if you don’t count it, you won’t be able to tackle it. Your predecessors opted for spin over substance. I hope that is a path you will avoid and would be grateful for an explanation of what action you plan.

Firearms-related homicides are up by the cited 18% – nine deaths in total – but otherwise the provisional figures show that total number of firearm offences fell from just over 11,000 in 2005/6 to 9608 in 2006/7. Shotgun offences declined by another 5% over the previous year. Handgun offences fell by 11% and, having risen to 476 in 2005/6, the number of serious injuries resulting from firearms offences fell back to 413 last year.

As for biggest growth areas in recent years, ‘slight’ injuries and imitation weapons, especially BB guns and ‘soft’ air-guns, the news is, again, pretty good for the government. Slight injuries were down by 23% – and the report notes that 41% of these injuries were caused by pellets fired from BB guns, etc. Offences involving imitation weapons also fell by 24% over the previous year. (Source: Crime in England and Wales 2006/7, pp 63-64)

Overall, ‘gun crime’ is down this year on the basis of the statistical information given so far, but for the number of fatalities – and one cannot draw any inferences about those deaths without knowing the circumstances in which they occurred; nine additional homicides in gang-related

Homicides relating to to gang violence or that happen in the course of street robberies will have a very different perceptual impact on the public than firearms homicides that occurred in the course of domestic incidents. The dataset is, of course, incomplete, and may still, when published in full, provide further information about any shifting trends and patterns within the data, but if the government have been a little economical in their presentation it is only in failing to point out that the fall they’re referring to is derived from a comparison of only the most recent figure and not from long-term trend data.

Against that, Davis’s comment about ‘not counting’ gun crime appears to be a complete and utter nonsense, if not an outright lie, and a matter upon which Jacqui Smith should call him out and demand that he provide evidence of where firearms offences have not been counted in order to substantiate his allegation.

Davis’s apparent claim that the long-term trend data cited earlier is, and taken from the Home Office’s detailed report on homicides, firearms offences, etc. for 2005/6 is equally laughable.

Its published in the report in a section entitled ‘Injuries in crime involving firearms’ in a chapter entitled ‘Recorded crimes involving firearms’, the location of which is clearly indicated in the report’s table of contents. Not exactly difficult to find then.

So far as talk of ‘cover ups’ is concerned, the information in question could be said to ‘buried’ only in the sense that the majority of the general public don’t habitually read Home Office statistical reports; they rely, instead, on the press [and on opposition politicians, such as the Shadow Home Secretary] to inform them of anything of particular interest that might appear in such reports. As the 2005/6 report, about which Davis is now seemingly complaining, was published in January of this year and is freely available from the Home Office’s website – and was presumably also placed in the House of Commons library, that this information has received little or no publicity until now appears to be less a case of a cover-up or the deliberate burial of information by the government and much more a case of Davis and his research staff falling down on the job, by failing to make anything of this data until now, and crying foul to cover up their own failings, while his complaint, such as it is, amounts to little more than whinging about the government not elected to do his job for him.

Getting back to Dizzy, he advances two predictions cum arguments which he mistaken appears to think will automatically gainsay any lines of attack open to Labour bloggers, the first of which…

I can imagine what the Labour response to such an accusation will be. It will either be, as Chris Paul tried to imply in the comments here, that if you include airguns in the data set then it’s somehow not as bad (which is of course nonsense when you are being very specific about the type of crime – gun homicide has increased four fold).

…merely demonstrates that he either hasn’t read the Home Office report that Davis and the Times are referring to, or if he has, he simply hasn’t taken it in. For one thing, it is he data four all recorded injuries arising out of firearms offences that has risen fourfold since 1998/9 – give or take the caveats about how those ‘injuries’ are defined and categorised – and not firearms homicides, the actual number of which was the same in 2005/6 as it was in 1998/9.

And, of course, as I’ve shown throughout, one can advance a perfectly serviceable line of counterargument against Tory claims while excluding the data for air-guns entirely from consideration.

His second argument:

The other response, at some point, that I would expect, is for someone to say that David Davis is playing party politics with the tragic death of Rhys Jones. That it is shallow and naked opportunism. This is the stock response to anything the Tories say eventually. They’ll probably then roll out some 20 year-old statistics showing how it was much worse under the Tories, but that won’t of course be playing party politics.

…is nothing more than an obvious and poorly constructed straw man.

That opportunism is an obvious charge to level against Davis does not mean that such a charge in untrue.

He is, quite patently – and badly, when one looks at the actual evidence – seeking to exploit media interest in gun crime arising out of the aftermath of the killing of Rhys Jones to advance his party’s political agenda and mount an attack on the government’s record on gun crime. What else [sadly] should one expect from a senior opposition politician in this day and age – Davis would be perceived to be failing to do his job if he didn’t use the current media attention on gun crime to try and put one over on his opposite number in government. Shallow and naked opportunism is, I’m afraid, the order of the day when such incidents occur, whether one is dealing with a politician shilling for votes or a newspaper trying to boost circulation during the slowest sales period of the year.

And Davis is not alone is trying to exploit this whole situation for political advantage – members of the government will, in their own way, be doing much the same thing as things progress and, in that respect, neither side is any better than the other nor any more legitimate claim to automatically occupy the moral highground.

None of this takes us any closer, of course, to actually understanding why incidents, such as the senseless killing of Rhys Jones, occur or how we can prevent them in future – and for the record, my money would be on this amounting to nothing more explicable than a random and entirely senseless act of reckless stupidity by a complete dickhead who mistakenly thought it would be somehow big and clever to scare a few kids with a handgun and wound up shooting one of them in the back of the head…

…and that is anywhere close to the mark then, on the evidence of the last few days, the collective understanding and knowledge of the political classes is likely to be about as much use as a chocolate teapot when it comes to understanding such an incident, let alone trying to prevent something similar happening in future.

In response to Jones’s untimely death, the government have [as usual] announced their usual raft of half-arsed and ill-thought out ‘initiatives’ including, most absurdly, the idea of having a telephone number where people can anonymously report people who they think may be in possession of an illegal firearm. Quite how and why no one, least of all the press, has yet cottoned on to the fact that this ‘new initiative’ consists of nothing more than a description of ‘Crimestoppers’, which has been running for years, tells you just about everything you need to know about public/media gullibility in the face of a compelling looking moral panic in full flow.

Meanwhile, Cameron’s trope about a ‘broken society’ rattles on, and on, and on like a broken record, for all that its perfectly apparent that he hasn’t got the first idea of what society is, or might be other than that ‘it’s different from the state’ – duh! – and that it has [in his uninformed opinion] something vaguely to do with charities and voluntary organizations doing some of the stuff currently undertaken by the public sector. Which tells us not only that he has no concept of the nature of society, not even to the point of an explicable hypothesis, but that he also cannot manage to distinguish between the generality of the concept of society and the considerably more limited concept of ‘civil society’ – which is what all the charity/community stuff is referring to – which is no more synonymous with society, as a whole, than is the state.

To clear this one up, the state and civil society, not to mention the individual and a range of other collective institutions – the law, religion (unfortunately), politics and political institutions, trade unions and other group structures – are all components of society, and what matters is not how these things differ from each other and from the generality of society but whether and how they relate to each other and how those relationships are organized, managed and balanced out.

What both Cameron’s ‘broken society’ canard and Davis’s tendentious and unsustainable allegations about the government’s limited use of current gun crime statistics over the last few days demonstrate is yet more evidence to support Hannah Arendt’s characterisation of the nature of totalitarian thinking…

“Before mass leaders seize the power to fit reality to their lies, their propaganda is marked by its extreme contempt for facts as such, for in their opinion fact depends entirely on the power of the man who can fabricate it.”

And what makes this observation all the more worrying is that, much as one tries, one cannot think of any statement on the subject of law and order made by a senior politician from either of the two main parties in at least the last 15 years or more – certainly not since the Bulger murder in 1993 – that would not also support Arendt’s statement.

  • Hmmm.

    “one finds that in 1998/9, fatalities accounted for 6% of all firearms-related injuries, serious injuries came in at 19% and slight injuries at 75%. By 2005/6, fatalities had fallen as a proportion of all firearms-related injuries to a mere 1%, serious injuries had fallen to 12%”

    The NHS getting better at treating gun trauma perhaps?

  • Not gonna get up my nose Unity. For the record, the mistake I made was to not type “and injuries” .. my bad. In relation to the “straw man”, in fairness I didn’t say that the charge of political opportunism was, by definition, untrue, I merely pointed out that it’s often leveled in deliberately weak ways, and it is used as a means of closing down argument – over the past few years to quite good effect.

    I did enjoy the way that you say I was dealing in conjecture (which I wouldn’t deny of course in this case as my tagline is after all “opinionated arrogance” remember) and then proceed to make statements of certainty about David Davis’s motivation in relation to what he did. I’m curious Unity, do you have kids?

    On the plus side though you won’t be getting a video because I don’t think you’re mental. Unless of course you want a video in which I express my undying love for your sheer genius at writing in a way that makes Jane Austen look exciting? IN which case I;m up for it, but not getting naked ok?

  • Dizzy didn’t pay attention to my comment’s full contents and turning a throwaway remark in a short blog comment into almost an official line from Labour is dizzy doing what dizzy does best. Being Dizzy around statistics.

    I have kind of discovered now why he was satisfied with the strange NINE YEAR period in the BBC data. (a) It was easier than doing any research and (b) it was pretty hard to find earlier figures though I have some now.

    He also didn’t respond to questions about right wing libertarian attitude to gun ownership a la USA and on Tory Party accepting funds from gun makers.

    The other issue about the stats is that there was a step change in reporting and recording practice c 2002 which coincides with either step changes of swooshes of “improvement” in detection or “escalation” in problems.

    In fact just a different statistical base.

    Hopefully crack on and turn the handle on the stats I’ve gathered. But when I do I can save quite a bit of time thanks to the normal anal servitude of the Unity tory crunching machine.

  • “in fairness I didn

  • Katherine

    If you are curious as to whether Unity has kids, then I suggest you actually read his blog. Throwing the pseudo-accusation (that he doesn’t have kids and therefore isn’t qualified to comment) is really rather brainless.

  • alabastercodify

    “Davis would be perceived as failing to do his job if he didn’t … try to put one over on his opposite number. Shallow and naked opportunism is, I’m afraid, the order of the day”.

    I admire you dedicated research, but I fear you’ve allowed youself to forget what you’re arguing about, and as you’ve gone on you’ve got lost in the figures, so that by the end what you’re saying makes no sense.

    You’re talking about Davis saying that Jacqui Smith’s claims that gun crime had gone down were misleading because they had, on a more reliable measure, gone up four-fold.

    When you examine his “more reliable” measure, you demonstrate, convincingly, that it is flawed – the majority of incidents are “non-serious”, and a proportion (unknown) of the increase will come from new reporting methods.

    Where does that leave us? Having worked so far through your piece, it seems to leave you of the opinion that Davis has displayed “shallow and naked opportunism”.

    It makes me think “Fuck, fatalities and serious incidents have very nearly doubled? And JAcqui Smith had the nerve to claim they’d fallen? Good for Davis”.

  • Pingback: All your DNA are belonging to us()