Ministers never tire of telling us that, under Labour, crime has fallen. But how does our record compare with other countries, and what can we learn from them?
Arguably, we have little to crow about. According to the International Crime Victimisation Survey, the closest thing there is to a reliable measure of crime across countries and continents, England, Wales and Scotland had some of the highest rates of crime against individuals of any industrialised country in 1999. Northern Ireland fared better, though its long history of paramilitary-related violence and criminality complicates the picture.
At 54.5 incidents per 100 inhabitants, England and Wales topped the list for rates of victimisation by a range of crimes including burglary, car theft, sexual and physical assault. Australia came a close second, on 54.3 per 100 inhabitants. Scotland was fifth, with 41.1 incidents. Denmark and France registered lower rates (35.1 and 33.9 respectively), as did Finland (28.6).
All statistics have limitations. Comparing statistics from different countries is fraught with problems.
A brawl in a bar might be considered a serious incident in some countries, and a perfect end to a good night out in others. Sexual assaults and domestic violence are notoriously difficult to quantify; rightly viewed as an outrage in some quarters and at most the occupational hazard of being a woman in others. A population of one country might welcome researchers into the sitting-room; those of another might slam the door in their faces.
But highlighting difficulties in interpretation does not mean all conclusions are bogus. What might countries with lower victimisation rates be doing differently that we might learn from?
Portugal has a far higher proportion of police to civilians than does England, Scotland and Wales, and a much lower victimisation rate.
Those who consider this a good argument in favour of more police might want to look at Denmark and Finland first. Both countries have lower victimisation rates, fewer police and much lower rates of imprisonment. Scotland and Poland have comparable victimisation rates, but divergent policing and imprisonment rates.
No single pattern emerges from a comparison of criminal justice interventions and victimisation rates. And this suggests that there is no clear relationship between levels of crime and victimisation, and criminal justice policies. In the UK, you will find many experts emphasising just how little impact traditional criminal justice levers are likely to have on crime rates.
This point becomes clearer when we focus on particular types of crime.
The incidence of bicycle theft in Sweden, the Netherlands and Denmark is two to three times that in England, Scotland and Wales for the obvious reason that cycling and bicycle ownership are far more common in these countries.
The crime reduction lesson to draw might be to restrict or discourage bicycle ownership, though the risk to the planet of our continued reliance on the car for personal transport is clearly a greater threat than the occasional stolen bike.
To understand the difference in the figures for violent assaults and threats, we need to look elsewhere.
At between ten and 12 incidents per 100 inhabitants, England, Scotland and Wales are near the top of the league for assaults and threats. The rates for countries as diverse as Sweden, Denmark, France, Belgium and Denmark are much lower: typically four to seven incidents per 100 inhabitants.
A common factor among those countries with higher violence rates is high rates of poverty and income inequality. The internationally recognised Luxembourg Income Study calculated that 12.4 per cent of the UK population in 2000 were living in households whose income was less than half of median earnings.
In contrast, those countries with lower rates of violence tend to have a lower proportion of households living in poverty. Eight per cent of French households lived on less than half of median earnings in 2000. In Sweden, the figure was 6.5 per cent and in Finland 5.4 per cent.
That poverty, inequality and violence might be linked is hardly a dramatic conclusion. A report on children and violence in the UK, funded by the Calouste Gulbenkian Foundation and published in the mid-1990s, found a strong relationship between poverty and inequality levels on the one hand, and levels of violence on the other.
An analysis of homicide rates in England and Wales, shortly to be published by the Crime and Society Foundation, likewise finds a strong correlation between murder rates and poverty rates. But poverty and inequality do not rise or fall by accident. They are a result of macro-economic and social policy.
While poverty rates in neoliberal Britain rose during the 1980s and 1990s, they remained largely unchanged in social-democratic Sweden.
The lesson is that those countries which have pursued social and economic policies that ameliorate the corrosive impact of capitalism tend to be those whose levels of public violence are lower.