Do the Official Crime Figures Tell the Full Story?

David G. Green

When the new-style crime figures were published in 2002, Mr Blunkett claimed they were the ‘most accurate measure’ of crime ever. The report was also said to be the most comprehensive ever. But when you check the small print, it turns out that the Home Office itself thought that there were far more than the 13 million crimes discovered by the British Crime Survey for 2001/02 – perhaps four times as many.

Arriving at the true figure is not easy because police figures are notoriously unreliable, but the Home Office has made ‘best estimates’ of the extent of police under-recording of some offences. Even on the most cautious assumptions, at least another 11.3 million crimes should have been added to the 13 million acknowledged crimes originally found by the BCS in 2001/02, a total of 24.3 million. According to another Home Office research study, The Economic and Social Costs of Crime, in 2000 there were at least 60 million crimes. On these estimates, the real figure was somewhere between 24.3 million and 60 million.

The same can be said of the 2004/05 figures. The true figure is well over 35 million, as we show below.

Why the huge disparity? Has there been a cover-up? Is any of this found only in a secret report? No, there is no secret report to be found. And, if you ask Home Office officials to confirm the higher figures they do so promptly. It’s partly a case of ‘If you don’t ask, you don’t get’. And until members of the public do ask – and keep on asking – the Government has every intention of pretending that the crime problem is under control. No objective observer would say that the British Crime Survey is comprehensive when it misses out murder, sexual offences, crimes against people under 16, illegal drug use and crimes against commercial premises, including thefts of trucks, vans and shoplifting. And no independent statistician would claim that the British Crime Survey was the ‘most accurate’ measure of crime.

The central issue is the independence of the government statistical service. There can not be a proper public debate about how best to deal with crime unless the full facts are made readily available for all. But as things stand, it is simply too tempting for any political party to have control of the release of information about crime – if the next election result might depend on public perception of the Government’s effectiveness, it is not going to give easy ‘ammunition’ to opponents.

The underlying problem is that many, if not all, statistical reports are still being submitted to ministers for approval of their content and the timing of their release. In an open society, there is no justification for the involvement of party politicians in regulating public access to information. Inevitably they use their control of the flow of facts to gain advantage over their opponents.

The independence of the Bank of England provides a parallel. The Government accepted that fixing interest rates was too tempting a party-political weapon and, to its credit, it handed authority to the independent Bank of England. Similarly, the independence of the National Audit Office and the Audit Commission has been accepted. But Home Office statisticians and the Office for National Statistics (ONS) lack autonomy. The Office for National Statistics is supposed to be independent, but it too needs to be made wholly autonomous, perhaps accountable to Parliament as a whole rather than to the Government of the day, which often functions as little more than a political party with a programme and a wish to defeat its opponents by whatever means come to hand.

In the September 2002 issue of Horizons, the official publication of the Office for National Statistics (ONS), Len Cook, then the National Statistician, defended the impartiality of ONS. There have been ‘one or two rumblings recently’, he admits, about the extent to which ONS is free from political interference. The ONS, he insists, is thoroughly objective and acts with integrity under all circumstances. But does it? In the same issue of Horizons, the catchline for an article about the crime figures was: ‘Looking at recent newspaper reports, you could be forgiven for thinking that crime is spiralling out of control. But the figures behind the headlines tell a different story.’ The article goes on to attack newspapers for using headlines to sell papers and insists that, when you ‘look at the long-term picture’, crime has fallen by 22% since 1997. The chances of being a victim of crime are ‘at around their lowest since the BCS began in 1981 – so don’t have nightmares, do sleep well!’

Whilst using phrases like ‘looking at the long-term-picture’ and taking ‘a closer look at the figures’ the author of the article disregards both the long-term picture and the facts that any objective observer would see upon taking a closer look. Statistical analysis is notoriously open to interpretation and, for this very reason, we need a genuinely independent statistical service whose officials see themselves as servants of democracy, not the instruments of the party in power.

The Official Line

The Government is anxious to claim that it has got crime under control and Government press releases regularly emphasise the fall in crime since 1997. The Government is particularly anxious to encourage the public to rely on the British Crime Survey (BCS). In the press release (STAT026/2002) accompanying the 20001/02 crime figures Mr Blunkett claimed that: ‘The largest ever BCS is now widely seen as the most accurate measure of people’s experience of crime’.

This series has been discontinued because the Home Office no longer publishes one of the accompanying documents to the ‘Crime in England and Wales’ report, entitled ‘Comparing BCS and police counts of crime’. This prevents calculation of a best estimate of actual crimes against victims under the age of 16 and therefore the shortfall in the BCS figures.

The Home Office has stated that it intends to expand the BCS in 2009, introducing a new survey for under 16 year olds; it is not known when these results would be published.

Note: Civitas wishes to record its thanks to the Home Office for checking and confirming the accuracy of the comparisons between the BCS and recorded crime for 2001/02. The document Crime in England and Wales 2001/02 reported 13.0 million BCS crimes. The 2002/03 report reduced the figure to 12.6 million after ‘recalibrating’ to allow for the 2001 Census. We have left all the figures for 2001/02 as reported in the original 2001/02 Home Office document.