Dialogue with BBC's Mark Easton

(on crime reporting)

We sent the following email to the BBC reporter, Mark Easton, after watching his BBC1 news report on crime (27/3/07) in which huge numbers of offenders/crimes were quoted.

Dear Mark,
In your BBC1 report (Ten O'Clock News, 27/3/07) on the 100,000 people who commit 50% of all crimes, you focused on some big numbers – "60 million" crimes per year, "3.6 million" people committing crimes, etc.

With the low level of context you provided, these figures seem meaningless. If you include people who drive over the speed limit or who download pirated material, etc, no doubt you'd get even bigger figures. It doesn't tell us much.

You could have made it meaningful by showing trends. For example, the British Crime Survey shows a massive fall of 8 million crimes between 1995 and 2006 (ie total 11.1 million crimes in 2006 compared to 19.1 million in 1995).

I've received the strong impression (from watching BBC crime coverage over several years), that the BBC is interested in focusing only on perceived rises and big, shocking numbers. It seems unnecessarily alarmist. When was the last time you ran a major piece (on BBC1) on the dramatic falls in crime since the mid-1990s?

Sincerely,

Brian Dean
Media Hell
[Sent 28/3/07]

Dear Brian
Thank you for your mail. We have covered the trends identified in the British Crime Survey many times and I have personally made the point that both recorded crime and adults' experience of crime from the BCS suggests a significant fall in crime since its height in the mid-90s. I can send you numerous examples of stories I have done which reflect on the fact that our chances of being a victim of crime are, apparently, at the lowest level for 25 years.

However, the BCS is not a particularly good guide to true crime levels. It doesn't reflect the experience of anyone under 16 even though 10-15 year-olds are among the most likely to be victims of crime. It doesn't cover child abuse and paedophilia. It doesn't get to many other sex crimes and violent crime including a huge amount of domestic violence. It doesn't reflect crimes which don't have an identifible victim - for instance VAT fraud, tax evasion, drug dealing and possession, huge amounts of theft.

VAT fraud costs the economy about £5 billion a year. Shop-lifting cost the economy about £2 billion a year. We convict 80,000 shop-lifters a year. We convict about 60 VAT fraudsters.

We convict about 100,000 people for TV licence evasion a year. We convict about 50 people for tax evasion a year.

You are right about context - I could have done with a lot more time - but the real context is around the nature of crime and criminality in Britain not the massaged and highly partial figures trotted out by government Ministers.

You may be interested to know that there are a number of criminologists who argue privately that the fall in victims identified by the BCS has more to do with a change in methodology regarding the weighting of the data than it does to any real drop in crime.

Best wishes
Mark

Dear Mark,

Many thanks for addressing my points at length.

You write:

> We have covered the trends identified in the British Crime Survey
> many times and I have personally made the point that both recorded
> crime and adults' experience of crime from the BCS suggests a
> significant fall in crime since its height in the mid-90s.

I appreciate that the BBC (in its entirety) covers a wide range on crime. But I'm interested in *headline news* coverage - since it reaches large audiences (particularly BBC1). I've not seen a single BBC1 news report which *headlined* or *focused* on the large falls in official crime figures (if you know of any I'd be grateful if you could provide details). Yet many/most headline or focus on cherry-picked "rises".

This is difficult to prove without going through hours of footage with you, but it's reflected in BBC News Online *headlines* on crime, which can be analysed. Here are all the headlines on quarterly releases of official crime figures, July 2004 - January 2007:

"Violent crime figures rise by 12%" (22/7/04)
"Gun crime figures show fresh rise" (21/10/04)
"'Violent crime increases by 6%'" (25/1/05)
"Violent crime 'rise' sparks row" (21/4/05)
"Violent offences top million mark" (21/7/05)
"Violent crime shows 6% increase" (20/10/05)
"Violent crime and robbery on rise" (26/1/06)
"Robberies up 6% but crime stable" (27/4/06)
"Phones and MP3s fuel robbery rise" (20/7/06)
"Robbery continues on upward trend" (19/10/06)
"Risk of suffering crime 'rises'" (25/1/07)

Note that *all* of these headlines report "rises" in crime. In our analysis - http://www.mediahell.org/crimeheadlines.htm - we compare these headlines with the main summary points in the official crime reports. The evidence could hardly be clearer that there is systematic "cherry-picking" of crime rises in BBC headlines.

As a further example, if you look at BBC News Online on the latest quarterly figures - http://news.bbc.co.uk/1/hi/uk_politics/6297715.stm - you see the following:

"FROM THE ARCHIVE
2006: Phones fuel robbery
2005: Violent offences up 7%
2004: Violent crime rises 12%
2003: Crime fight 'being lost'
2002: Street robberies soar
2001: Violent crime on the rise
2000: Big rise in violent crime"

So whilst it may be true that the large falls in crime are covered *somewhere* in BBC reporting, it's apparent that headline coverage avoids it in favour of cherry-picked "rises" (some them are not even real rises, but only artificial - due to changed police recording procedures). It's equally apparent from the official crime reports that the large falls *should* be appearing in BBC headline coverage - assuming that the aim of such coverage is to report the official figures. The large and consistent falls in many crimes which matter to the public (burglaries, vehicle crime, etc) - and dramatic falls in overall crime over the last decade - hardly seem like minor points that you'd want to exclude from headline coverage.

Of course, there's an important distinction to be made between *reporting* the official crime figures and *challenging* the validity of those figures. To the extent that BBC coverage attempts to *report*, it shouldn't cherry-pick crime "rises" in headline coverage. And to the extent that it seeks to challenge, it shouldn't challenge only the British Crime Survey (BCS) findings of falls in crime. It should also challenge the police-recorded figures which have artificially inflated, for example, the headline violent crimes figures.

You write:
> You may be interested to know that there are a number of criminologists
> who argue privately that the fall in victims identified by the BCS has
> more to do with a change in methodology regarding the weighting of the
> data than it does to any real drop in crime.

I'd be grateful if you could supply more details. Also, why would this be a "private" matter for these criminologists? Most people I've spoken to believe that the BCS provides a far more reliable guide to crime trends than police recorded figures, since the latter are affected (sometimes dramatically) by changes in levels of reporting to the police, and in police recording practices.

I also remember a time (prior to the BCS showing falls in crime) when many politicians and reporters preferred the BCS to police figures, as the BCS includes crimes that aren't reported to the police, and therefore presents higher figures for the crimes it covers. When the BCS started showing downward *trends* in crime, it rapidly lost favour with these politicians/reporters.

Coming back to the above BBC headlines, it seems that the BCS figures are used for headlines *only* when they show rises in crime (and are excluded when they show falls). For example, the latest official crime figures (25 January 2007) show that crime is up by 1% according to BCS, but down 3% according to police figures. After the BBC headline-writers cherry-picked police figures for 10 of the 11 above headlines, they suddenly decide to go with the BCS figure for the latest headline.

The social context of this "cherry-picking" issue includes a public fear of crime which is out of proportion to the real risk of crime for most people (eg 1 in 3 elderly women fear going outside, but only 1 in 4,000 will be attacked - according to police).

The political context is of a government determined to implement freedom-eroding legislation on the basis that the threats we face (including those of crime) are "spiralling". Yet the cold evidence shows no such "spiralling". Of course, the evidence matters little if people are subjected instead to the above type of scaremongering headlines on a continual basis.

Sincerely,

Brian Dean
Media Hell

Dear Brian

[...] Journalists tend to regard their job as highlighting problems rather than reporting solutions. There is a concern that focusing on positive trends is "soft", that we must always be holding the executive to account on where systems don't work.

I have some sympathy with this view, but I agree that the BBC also has a public duty to present a fair and accurate picture, not an unduly alarmist one.

My aim is to try and change the way we report crime statistics - so that we don't pretend we can actually know whether it is going up OR down. We might be able to see trends with specific offences, - thanks to better security systems the risk of having your car stolen has clearly fallen, for example - but the claim that we can use the data available to conclude that there have been "dramatic falls in overall crime" is unsustainable I think. Neither the BCS nor the recorded crime statistics get anywhere near a measure of overall crime and it might be the people's reluctance to believe the official figures is a reflection of this.

My intention on Tuesday night was not to alarm people with big numbers but to point out that the claim of a small criminal under-class doesn't withstand scrutiny. There are no simple "lock 'em all up" solutions.

I wish I was able to offer you more on those criminologists who question the BCS methodology. My source (independent minded) is considering holding an academic seminar to tease some of this out at which point I may be able to say more.

Best wishes
Mark

Dear Mark,

Thanks for your reply. Earlier, you said you'd run numerous stories which reflect that "our chances of being a victim of crime are, apparently, at the lowest level for 25 years". I then challenged you to provide *headline news* examples. In the absence of such examples, would you not agree that headline BBC1 news has focused primarily on crime "rises" when reporting the official figures?

The BBC would need a very good reason to ignore trends of falls in crime in headline news coverage. Is the scepticism expressed by one or two anonymous criminologists over the official reports of falls really sufficient? (Particularly given that scepticism, publicly expressed by criminologists and others, over "rises" in violent crime hasn't stopped the BBC consistently focusing on these "rises" in headline coverage).

You write:
> [...] the claim that we can use the data available to conclude
> that there have been "dramatic falls in overall crime" is
> unsustainable I think.

Is it meaningful to compare some hypothetical, absolute measure of *all* crime (which is practically meaningless) with the view of "overall" trends as presented by the official crime figures? The BBC reports the official figures because they combine the two most comprehensive datasets available - police recorded figures and the British Crime Survey (BCS). A close reading of the official figures shows that police figures and BCS findings tend to complement and reinforce each other (eg when one takes into account changed police recording practices, etc) - including categories in which crime has fallen. Indeed, BBC1's Panorama (17/4/05) reported that the Association of Chief Police Officers acknowledged that violent crime had remained stable (or had fallen) since the late 1990s.

For the BBC to be justified in "cherry-picking" crime rises from the official crime figures (in headline news coverage), there would have to be some comprehensive crime dataset in opposition to the official data (in ways which support the BBC's focus on "rises"). To my knowledge there is no such dataset.

You write:
> Neither the BCS nor the recorded crime statistics get anywhere
> near a measure of overall crime and it might be the people's
> reluctance to believe the official figures is a reflection
> of this.

Most people are acquainted with the official crime figures only to the extent that they're reported by the media - which, as I've indicated, tends to cherry-pick crime rises. So it's unsurprising that people don't believe crime is falling in most areas covered by the official figures. After all, they've been hearing (in the media) that crime is "rising" on a continual basis for years/decades.

The perceived risk of crime, for most people, is taken from the media rather than from personal experience. Two-thirds of the people polled by 'Frontline' (a Channel 4 documentary on crime, 4/10/95) said they received most of their information on crime from TV. The reporter, Julie Flint, commented that on questioning people about their crime fears, the majority had no personal experience of crime, but they’d seen how bad things "really were" on TV. This is common sense - we're bombarded on a *daily* basis by media coverage of crime. Most people are affected directly by crime much less frequently. For example, the average house will be burgled only once every 50+ years.

You write:
> There is a concern that focusing on positive trends is "soft",
> that we must always be holding the executive to account on
> where systems don't work.

I suspect this fear of appearing "soft" is closer to the real reason for the BBC's cherry-picking than legitimate concerns with the comprehensiveness of the official figures. We're paying a huge price for this in terms of a public fear of crime which is out of all proportion to the risk of crime for most people. And we're paying a huge price in terms of the political scaremongering (in all the main political parties) which exploits media obsession with "escalating" crime.

Best wishes,

Brian
Media Hell

Notes: Police figures vs British Crime Survey (BCS)

Since 2001/2002, the official crime reports have combined police recorded crime figures with the results of the British Crime Survey (one of the world's largest and most respected crime surveys).

The first of the combined reports, Crime in England and Wales 2001/02, presents these two series of figures as being complementary and of together providing "a better picture of crime than could be obtained from either series alone". It goes on to describe the differences between police and BCS figures:

"Police statistics provide a good measure of trends in well-reported crimes, are an important indicator of police workload, and can be used for local crime pattern analysis. For the crime types it covers, the BCS [British Crime Survey] can provide a better reflection of the true extent of crime because it includes crimes that are not reported to the police. The BCS count also gives a better indication of trends in crime over time because it is unaffected by changes in levels of reporting to the police, and in police recording practices." [From the preface]

Click here for more details on the official crime figures >



 
  TOP HOME >    SITE MAP >    CONTACT >    BACK TO TOP >