Fall in crime 'too good to be true'?

Marian Fitzgerald, Visiting Professor of Criminology, University of Kent, that the 'too good to be true' fall in the Crime Survey for England and Wales requires serious consideration of its possible limitations.

By: 
Professor Marian Fitzgerald
Date: 
Monday, 28 April, 2014

The causes of crime, as Richard Garside rightly implies, are themselves far too complex for any apparent falls to be ascribed to one simple explanation. Many of those on offer for falls in violent crime observed in recent years have been speculative, ranging from the plausible (but necessarily partial) to the outright laughable.

The recent report from Cardiff University was based on data from a sample of 177 hospital emergency departments in various towns cities in England and Wales and this generated considerable discussion in the media around its finding of a significant fall in violence between 2008 and 2013. The discussion, though, been strangely silent with regard to the most obvious explanation – and the one which is best grounded empirically.

Among a range of very sophisticated models of long-term trends in crime (not only in Britain but also internationally), work in the Home Office itself has shown that these overall trends are subject to short term variations whereby violence tends to reduce at times of recession. The fact that this has been overlooked is the more surprising since the discussion has rightly focussed on violence related to the late-night economy; and it is indeed this which has the greatest influence on any figure for violence overall.

The simple fact is that when times are hard, there is a reduction in either the numbers of people who can afford to go out, get drunk and end up in a fight which, in a crowd situation, can quickly draw in others or a reduction in the frequency with which they can do so.

However, little credit has been given either to a number of sensible measures which have been taken to reduce this risk – not least of which may have been the Licensing Act of 2003 which came into force in 2005. It did so in the face of strong opposition from prophets of doom who predicted that its provisions would encourage 24 hour binge drinking.

In fact the experience of other countries already suggested that abolishing legally enforceable ‘drinking up time’ avoided large numbers of people in varying states of inebriation all being thrown out into the streets at once. Especially if the rate at which they left pubs and clubs was, so to speak, staggered over a longer period and was combined with better provision of affordable transport to get them home safely (such as all-night buses) this reduced the number of fights which typically broke out when many in the throng had then to queue for a limited number of taxis.

Alcohol pricing may have had a marginal further effect but it would be mistaken to place too much weight on this or to become complacent. Even if relatively casual drinkers can no longer afford to drink as much or as often (including many young people who appear actively to plan their collective binges), the number of hospital admissions for alcohol-related conditions continues to rise.

And while alcohol may in some circumstances increase the likelihood of violence it is by no means its sole cause. Rather, the recession may itself have increased the level of domestic violence in the widest sense of the term when the constant threat of unemployment, the inability of young people to find work at all (and thereby afford to get out of the house to let off steam), the rising cost of living and the withdrawal of benefits may cumulatively increase tensions, especially in households which were already under stress.

Nor will increasing the cost of a bottle of wine do much to reduce local pockets of serious violence associated with the drugs trade or with young people’s felt need to prove they can ‘take care of themselves’ by meeting the constant threat of violence with pre-emptive acts of violence or by ensuring they have protection from others with a well-established reputation for violence.

The danger is, of course, that these very real and often life-threatening problems risk being subordinated to the imperative for politicians to be able to claim that they have reduced crime – even if they have no explanation for how or why. Nor do they have any incentive to query the official figures they are given as long as the statistics continue to support these claims.

True, government statisticians have finally acknowledged that the fall for which both the previous and the present government have claimed credit may have been overstated – but only inasmuch as police forces, driven by politically driven demands for crime reduction had been exaggerating their claims of success. The measures some have employed to this end have now been widely and very publicly aired in the inquiry by the Public Administration Select Committee and the Office for National Statistics (ONS) has withdrawn its endorsement of the police figures, leaving the government dependent solely on the Crime Survey for England and Wales (CSEW) to provide a reliable indication of trends.

Following the acclaim that greeted the reports of a fall in violence based on hospital statistics last Wednesday, however, the government seems to have been remarkably coy about claiming credit for the fall in crime in the official figures published by ONS the following day. For violent offences account for only a minority of crime overall; yet the CSEW was now showing an unprecedented year-on-year fall of fully 15 per cent between January 2013 and the end of the year. The largest fall it had previously recorded was 10 per cent; and – even as they now stand shamed and indicted for fiddling their figures – the police service had never achieved more than 9 per cent.

The survey itself has significant limitations which have hitherto gone largely unrecognised – and of which ministers themselves may have been largely unaware. Now the CSEW is producing results which, even to them, may appear too good to be true, perhaps the time has come to explore these limitations further.