Asian ‘sex gangs’ and statistical limitations

Over the last few days, proponents of the view that Jack Straw was scaremongering when he suggested, on Newsight, that there is a specific problem with some Pakistani men which leads them to target non-Muslim women have leant heavily on statistical evidence supplied by Chris Dillow as follows:

And herein lies my irritation. Straw gives us no statistics to justify his claim. Those that do exist seem to undermine his claim.

Table 5.4b of this pdf shows that, in the latest year for which we have data, Lancashire police arrested 627 people for sexual offences. 0.3% of these were Pakistanis. That’s two people. 85.5% were white British. In Lancashire, there are 1,296,900 white Brits and 45,000 Pakistanis. This means that 4.163 per 10,000 white Brits were arrested for a sex crime, compared to 0.44 Pakistanis. If you’re a journalist, you might say that the chances of being arrested for a sex crime are nine times greater if you’re white than Pakistani.

At first sight, Chris’s decision to look specifically at the arrest statistics submitted to the Ministry of Justice by Lancashire police seems an eminently reasonable one, given that this is the force responsible for policing Straw’s own constituency (Blackburn). However, a broader examination of the MoJ’s statistics on arrests for sex offences indicates, quite clearly, that the data for Lancashire is vastly unrepresentative of both the overall national picture and the picture that emerges when one looks at the three specific areas in which there have been strong allegations, arrests and/or convictions relating to the activity of predominately South Asian gangs involved in the grooming of young women with the intent of forcing them into prostitution; Greater Manchester, South Yorkshire and West Yorkshire.

As regards the statistical information that follows, the data on arrests for sexual offences by self-identified ethnic group is taken from the statistical tables the accompany the MoJ’s most recent report on Race and the Criminal Justice System, which can be accessed here, and covers both 2007/8 and 2008/9.

Calculated offending rates are derived using the most recent ONS population estimates by ethnic group (mid 2007) which, for the time being, are considered to be experimental and are given as rates per 10,000 for the male population aged 16-64.

Starting with the national picture for England and Wales, the MoJ data indicates that there were 32,016 arrests for sex offences in 2007/8 and 33,687 arrests in 2008/9.

Of those arrested in 2007/8, 23,935 (74.8%) self-identified their ethnicity as White (includes British, Irish and other), 2622 (8.2%) as Asian/Asian British (includes Bangladeshi, Indian, Pakistani and Other but excludes Chinese) of which 920 (2.9%) self-identified as Pakistani.

For 2008/9, 25940 (75.7%) self-identified as White, 2908 (8.6%) as Asian/Asian British and 1013 (3.0%) as Pakistani.

In regards to the male population of England Wales aged 16-64,  this being the demographic group that the overwhelming majority of those arrested for sex offences would belong to, the ONS mid-year data for 2007 estimates a total population of 17.562 million of which 15.442 million (87.9%) are White and 1.062 million are Asian/Asian British (6.1%) of which 306,600 (1.7%) are of Pakistani origin.

Based on these figures, the arrest rates for sex offences per 10,000 male population aged 16-64 are:

White – 15.5 per 10,000 (2007/8) and 16.5 per 10,000 (2008/9)

Asian – 24.7 per 10,000 (2007/8) and 27.4 per 10,000 (2008/9)

Pakistani – 30.0 per 10,000 (2007/8) and 33.0 per 10,000 (2008/9)

Based on the data for England and Wales, you’re a little over 1.5 times as likely to be arrest for a sex offence if you’re an Asian male aged 16-64 and around twice as likely to be arrested if you’re a Pakistani male.

As I mentioned earlier, a quick scan through the Google news archives turned up three recent instances in which gangs of predominantly Asian pimps have been either strongly alleged or definitively found to have been operating in a specific area.

In 2008, a BBC Panorama investigation included strong comments from Mohammed Shafiq, Director of the Ramadhan Foundation, in which he alleged that Asian men in Bradford, West Yorkshire, were grooming young white women for prostitution and accused the local police of being too fearful of being branded as racist to tackle the issue effectively.

The recent case which led to the conviction of five Pakistani men, which triggered the current furore over this issue, relates to offence committed in Rotherham, South Yorkshire.

Finally, the Daily Mirror reported, only yesterday, that a further nine Asian men have been arrested on suspicion of grooming young white women for prostitution, this time in Rochdale, Greater Manchester.

In two of these areas (Greater Manchester and West Yorkshire) the Asian (and Pakistani) adult male populations are larger than  than the national average for England and Wales, and both Asian men Pakistanis are significantly over-represented in the recorded arrest figures for sex offences.

In Greater Manchester, Asian men account for an  estimated 7.6% of the local male population aged 16-64 but have accounted for 11.8% and 13.3% of arrests for sex offences for the two years (2007/8 and 8/9) covered in the data I have to hand, giving arrest rates 39.1 and 42.9 per 10,000 population against rates of 20.6 and 19.9 per 10,000 white males. For Pakistani men, the figures are even worse; they make up an estimated 3.6% of the local male population but accounted for 7.1% of arrests for sex offences in 2007/8 and 7.9% of arrrests in 2008/9, giving arrest rates of 49.4 and 52.9 per 10,000 for each of the two years.

In West Yorkshire, which has the largest Asian (and Pakistani) male population of the three areas (10.5% Asian, 6.9% Pakistani), Asian men accounted for 16.7% of arrests for sex offences in 2007/8 and 18.1% in 2008/9, giving arrest rates of 36.2 and 34.2 per 10,000 adult male population against rates of 18.0 and 17.2 per 10,000 White adult males. Pakistani adult males accounted for 12.6% and 12.0% of arrests with respective rates of 45.0 and 37.2 per 10,000 adult male population.

South Yorkshire differs from the other two areas inasmuch as the local Asian male population is much closer to the national average, in fact slightly lower overal at 4.2% against 6.1% for England and Wales, although the Pakistani adult male population is slightly higher than average (2.0% against 1.7%). Arrests for sex offences for both Asian and Pakistani males, as a percentage of the total number of arrests for these offence were also fairly close to the national average for 2007/8 at 7% for Asian males (8.2% E&W) and 3.2% for Pakistani males (2.9% E&W) although rates per 10,000 were significanly higher than the national average at 35.4 per 10,000 (Asian) against 27.4 per 10,000 for England and Wales. For this year the arrest rate for Pakistani males, 33.3 per 10,000 was comfortably in line with the national average of 30 per 10,000 given the slightly larger than average Pakistani male population living in the area. Again, however, arrests per 10,000 male population were significantly higher for Asian and Pakistani males than for White males (18.5 per 10,000)

For 2008/9, the data both a fall in the arrest rate for White males (down to 14.5 per 10,000) and a significant increase in the arrest rates for both Asian males (41.6 per 10,000) and particularly Pakistani men (44.8 per 10,000) over the previous year. With the number of arrests for sex offences in South Yorkshire having fallen, overall, by around 18% on the year this apparent rise in arrest rates for Asian indicates only that rate of arrests for white males has fallen more sharply over the year than it has for Asian and Pakistani males, both of which saw a much more modest decline in the number of arrests.

So what does all this statistical information tell us?

In truth, not very much.

It doesn’t tell us anything about the prevalence of gang-related grooming offences, which is the hot topic of the moment, or anything at all about patterns of offending within any ethnic community and whether or not these are similar or different as there’s no data, at all, on the specific types of offences for which any of these people have been arrested.

It doesn’t really tell us very much about the relative levels of sex offending within any of these communities either as there is nothing in any of this data to give us any ideas about how the numbers anf rates of arrest relate to the number of reported offences, for which there is no ethnicity data on either victims or perpetrators, or anything about relative levels of under-reporting across different communities.

As such, there is no clear way of determining whether and to what extent any of the observable diffferences in arrest rates might be a function of difference in actual offending rates, or reporting and under-reporting rates, or in the manner in which the police investigate alleged offences or prioritise and allocate resources to particular offences and types of offences.

Comparing the data from the three Metropolitan areas, and the national figures, with those from Lancashire that Chris cited, what we can say is that there’s a correlation between urbanisation and arrest rates for sex offences and, if one takes the time to look at the economic data for these areas, also a likely correlation between deprivation and arrest rates, but then that’s hardly news is it.

All this does seem to tell us is that Chris’s assertion that Jack Straw lacked any statistical evidence to back up his claim, not that I think he was necessarily basing his assertions on statistics alone, is manifestly incorrect if one looks beyond Straw immediate environs at the broader picture of arrest rates for sex offences in England and Wales and, particularly, at urban areas where there is both a large Asian/Pakistani population and high levels of socio-economic deprivation.

Do these statistics genuinely support Straw’s line of argument – superficially, perhaps, but only superficially as there really is too little detail to draw any definite conclusions from the available evidence. That said, they don’t negate his argument either, so we’re really in no position to judge the merits of Straw’s point on the basis of any of these statistics.

Where his claim about there being specific issues with some Pakistani men does stand up to scrutiny, a priori, is if one takes the view – as I did in the my previous post – that Straw was advancing rather more of a sociological than a demographic argument.

As useful as statistics are, they can’t answer every question.