Recession and the sneak thief
Theft from the person up 25% in England and Wales?! What is going on?
There are three possibilities, I suggest:
(1) The recession is making some people more light-fingered
(2) We are more likely to notice (or imagine) that someone's pinched stuff
(3) It is a statistical quirk
First, let's define our terms.
"Theft from the person" covers theft (including attempts) of a purse, wallet, cash etc directly from the person of the victim, but without physical force or the threat of it. We are talking pick-pockets and opportunists here - thieves who spot an unattended handbag or jacket and help themselves.
Second, let us remind ourselves how the figures are calculated.
This data in the latest quarterly crime stats come from the British Crime Survey (Crime in England and Wales: Quarterly Update to December 2008 [223KB PDF]) which interviewed 44,580 adults in England and Wales in December asking them of their experiences of crime in the previous 12 months.
It is a big and pretty robust poll and, when the replies are applied to the whole population, the latest numbers suggest there was an increase from about 573,000 "thefts from the person" in 2007 to 716,000 last year - an increase of 143,000 such crimes. It is a rise of 25% and statistically significant.
Third, let's look at the context of people's experience of crime elsewhere in the BCS figures.
Overall the BCS survey suggests that crime is stable. There's been no real change in people's experience of violence, domestic burglary, vandalism or vehicle-related thefts compared with the previous year. The risk of being a victim of crime remains historically low.
This rise in theft from the person is not reflected in other categories of BCS personal acquisitive crime (see table below) or in the comparable category of police recorded crime. "Other thefts" recorded by police actually fell 4% year on year.
Source: Home Office
If the reason for the rise is (1), the recession, then it would appear that the only group so far which has felt the economic need to cross the line into criminality is that of sneak thieves.
This may not be as crazy as it sounds. Belt-tightening plus temptation might well lead some to pinch that £20 note left on a colleague's desk.
The second possibility, (2), is that greater anxiety about money makes us more aware when something apparently goes missing. Perhaps a year ago people assumed they had spent that fiver on something and forgotten. Or they never missed it in the first place. Now, they think it must have been stolen - particularly if they like to believe they are being more careful with their cash.
Remember, the survey asks people about their "experience" of crime - it cannot determine whether people's recollections are accurate or consistent.
Lastly, it could be (3) - a statistical quirk. The fact that something is "statistically significant" does not mean that it indicates a changing trend. Even with a big survey, sometimes you get a "significant" result completely out of line with previous and later polls.
I reckon it might be a bit of all three - relatively risk-free small crimes are the ones most likely to rise when money is tight; people are more conscious of their cash and belongings and therefore quicker to notice something missing and assume theft; the 44,580 adults in England and Wales asked about their experiences just happened to include a surprising number of people who'd had their wallets and purses nicked.
Overall, I think the figures reassure us that recession-driven anarchy is still a long way off.
I'm 
© MMIX~RS~q~RS~~RS~z~RS~20~RS~)
Comments
Sign in or register to comment.
the statistics only show what most already knew, but it only shows what it wants it fails to show insurance fraud levels.
but it can be good to know there is hope and where there is hope there is progress.
Complain about this comment
Mark - maybe I'm a tightwad, but I don't really go with your suggested cause (2) - that in a recession we are more likely to notice we've been robbed! £5 taken off the desk and that kind of thing. Some of the wealthiest people are those most likely to notice such things - while some of those on modest means may be relatively casual. On PM today you seemed to be bending over backwards to explain away the figures and "side" with the Home Office, even suggesting it might be a statistical quirk despite it being a whopping 25 per cent rise rather than some marginal increase. On this logic, no opinion poll or statistic survey would seem to have any indicative value whatsoever or to have a claim to be taken seriously - it can all be explained away as some feature of sampling.
A further possibility in explaining the rise is that some of the resondents have made fraudulent insurance claims over theft - maybe such claims are increasing in the recession as a form of "victimless" crime - and they stick to their stories even in a presumably anonymous survey. Anyway, whatever the reason for the rise, your contribution was as always thoughtful and thought-provoking.
Complain about this comment
Whatever the conclusions drawn from these statistics, it is obvious that crime is not significantly on the wane.
The amount of unsolved crime remains unacceptably high.
The logging of crime, the questioning of suspects, the attendant police paperwork all have to be streamlined, so that our law enforcement officers spend more time on patrol, and less time in bureaucratic paper chasing.
Complain about this comment
Mr Easton,
I have just heard your conversation with Eddie Mare concerning the " crime survey report "
I could not believe my ears when I heard your theories to the evidence that "theft from the person" had risen by 25%.
You are just unbelievable!!
Can I first of all ask you to listen again to what you said.You managed to contradict yourself early-on in a fundamental way.
Additionally,Your further explanations would do credit to a Home office Minister or one of their Spin Doctors.
To suggest that, only now, people are now more attentive to their money, are more forgetful about their bank notes and therefore believe, incorrectly, that it has been stolen, more keen to report a theft is pure and unadulterated trash.You should be ashamed to call yourself a descerning reporter.
The British Crime Surveys already omit so many important statistics as to render it uninformative so don't add to the nonesense with your utterances
Give up, go and work for the Home Office as you would be very good addition to their staff a number of which are ex-BBc anyway.
John L Hoddy
Complain about this comment
The crisis is caused by our lack of congruence to the laws of Nature. And if we open up this Pandoras Box, we will enter the arena with the greatest concentration of power, strength, money, force, and armies. In other words, we will invade the place where the greatest egoism hides. It is particularly these people that must now realize that they are completely opposite to Nature - that they are unnecessary and detrimental.
Then we will come to the kind of situation that Karl Marx wrote about. Can you just imagine how many other problems will have to emerge until all of this will come to the surface?
It may seem that these problems are distant, global, and removed from us, as we ask, where are they, and what do they have to do with me? However, it is a fact that they determine our fate and our day to day lives.
http://www.laitman.com/2009/04/the-worlds-problems-are-our-problems/
Phd Michael Laitman
Complain about this comment
Well I would hail Mark Easton as one of the few independent-minded correspondents/editors working in broadcast journalism today, as well as one of the best bloggers. He certainly doesn't always parrot the Home Office line, as #4 would have us believe. Rather than always adopting the crowd-pleasing "our country's goin' to the dogs" mantra, it's good to see reason and measured analysis, even if that means occasionally saying something that doesn't present the government as the spawn of satan.
Complain about this comment
This comment was removed because the moderators found it broke the House Rules.
There is an obvious fourth possibility that occurs to me (cynic that I am):
(4) the way that the figures are recorded has changed or the way the crimes are classified has changed.
If you were a body that had been 'encouraged' to meet a target to show a fall in (what I would call) violent crime - purely for the sake of example, let'say "robbery at knifepoint" - where might you be tempted to 'hide' those figures?
Complain about this comment
This comment was removed because the moderators found it broke the House Rules.
"recession-driven anarchy is still a long way off"
Absolutely, some crime categories are up, but others are down.
Interestingly though, most of the headlines I saw yesterday were about recession 'causing' an increase in burglary - not recession 'causing' a drop in vehicle crime (if true, that would be even more remarkable - man bites dog etc).
Presumably the editors thought shock horror, crime out control, makes a better headline to sell their paper/news broadcast. But why do they think that? Is it really true that I would be less likely to buy the paper if they put some startling good news on the front?
Complain about this comment
It would be much more interesting if they surveyed how many of these victims had IPODs on or were busy talking on their phone at the time.
I have never been mugged because I stay aware of who is around me and where I am. After watching people over the last few years I'm surprised these figures are so low (maybe they suppression of them has finally broken)
Muggers must be laughing out there - most victims make it so easy for them.
I do agree with some of the other posts - a lot of it matters how the police capture the data. I think we saw it before with rape and violent assault figures being clouded.
What I can say is that it's only going to get worse - and who will be surprised really? When you have a society which is bombarded with 'obscene wealth' of film stars, pop stars, sports stars etc. and their trappings to which they are supposed to aspire - is it any wonder when young people realise they cannot achieve these heights through 'ordinary work' that they then turn to crime in order to achieve it.
Cause and effect - cause and effect
Complain about this comment
#7
You have this spot on. My brother who's in the Met told me how a workman had his tools stolen from a hut on a building site, which by definition is burgulary. After calling the Burgulary squad, it was redefined as a 'Walk in Theft' as they had no chance of solving the crime. It seems that the police are indirectly encouraged to reclassify crimes so they can meet Home Office targets. As well as 'Theft from the person' (or Robbery as it's otherwise known possibly!) could we have the stats for 'Walk in Thefts' please?
Complain about this comment
This comment was removed because the moderators found it broke the House Rules.
Very interesting story as always Mark, but I believe there’s a problem with the govt stats here. I’ve looked through the documents, as well as the 2007-8 technical report vol 1 located at: [Unsuitable/Broken URL removed by Moderator]and I believe there is actually no significant increase in ‘theft from person’ at all; the statisticians have failed to analyse and report the figures correctly. Let me explain...
Roughly 1 in 1000 people in this country were surveyed. Hence the numbers in the tables above (in your article) are roughly the actual numbers of people talked to by the researchers (i.e. the “000s” as written at the top of the table is just a theoretical extrapolation to the population). This means that in reality only about 143 more people out of about 46,000 reported ‘theft from person’ in 2008 than 2007. This is an increase of less than 0.3 of ONE percent (in a sample of 46,000!). In statistical terms this is far from significant. I don't know how they messed this up but they may have extrapolated the population BEFORE doing the tests (which is not valid mathematically, and gives the wrong result). The 25% increase figure you mention is generated because of the small numbers involved, and is misleading as it also neglects the rest of the sample!
Additionally (and this is less clear from the doc) the technical report (vol 1) talks of confidence intervals representing 95% (1 in 20) but there are FIVE categories to apply these to, so actually that instantly increases the possibility of a statistical error to 1 in 4. The report however is not as clear on this. It’s pretty badly written actually, so I may be wrong here.
Basically this means the rise in theft from person IS just normal statistical variation (and not much of it actually), NOT significant as claimed in the govt report. Although the report is not very clear (and doesn’t give the SPSS files), it’s fairly certain that there is no real change at all in ‘theft from person’.
Am I wrong? Happy to be so, could anyone else take a look??
Complain about this comment
Mark
I've been checking whether I've been robbed
I looked at the details of my small private pension. You are right - my tax has gone up. I hope nobody wastes my contribution - like giving it to the pension of a "retired" banker.
Thanks Gordon.
Complain about this comment
I think the statisticians are wrong (just been through the tech report vol 1 in detail). The increase in 'theft from person' (2007 - 2008) is about 0.003 across the sample (i.e a third of one percent). It is not statistically significant either (despite what they claim). I explained fully but my explanation was referred to the moderator and I can't be bothered to re-type it.
Complain about this comment
The stats are just as good or bad as the people that collect them and the purpose. The police and others such tally them for their self interests and self righteousness to get more public funds for the benefit of the system. Except for individual victims all the rest is figured in added up front by entities like supermarkets, insurance companies etc as cost of doing business and figured into the price the consumer pays.
These are just penny stuff, the bigger issues are real high end cheating, stealing, use,abuse, exploitation by Government workers, politicians and the such that submit expenses for per diem extra. In the US the great example of Ms. Palin the Governor of Alaska charging some 310 days per diem, when she slept in her own bed, ate her own food and yet billed the Alaska taxpayers per diem. UK great example of all the MP submitting expense of a second home during the Parliament session, when some body like Jacqui Smith, is living with her sister and calls it her permanent resident.
Complain about this comment
Good topic as always Mark. My earlier comment has got stuck in moderation, so,
I think this is a phantom effect caused by statistical error and we're all on a wild goose chase.
The sample was about 44,500 (about 1 in 1000 UK adults), so to get the real numbers just forget the "000s" bit on the table.
In 2007 about 573 out of 44,500 people said they'd experienced 'theft from person'.
In 2008 about 716 out of 44,500 people said they'd experience 'theft from person'
That's an increase of roughly 0.003 (143 people across the sample of 44,500 people) less than some other categories actually. Because 'theft from person' is the smallest category by far, it will experience the most apparent variation naturally year to year (that's where the illusion of a disproportional increase comes from, i.e. 25%). The statisticians in the report claim the rise is significant (to 95% confidence) but the tests used are not given, and I can't see how they reach such a conclusion unless they have misapplied the tests.
Put it another way, the number of people saying they have NOT been a victim of personal theft has gone from 43,927 to 43 784, a decrease of 0.3%; not so worrying!
The tech report and final report are not easy to make sense of and I can't find the actual numbers, but I think there's been an error.
Have I missed something perhaps?
Complain about this comment
I wish to tread very carefully here perhaps by sticking to that old favourite statistics.
I find myself wondering how these figures now compare to the rest of expanded Europe as a form of crime per head of population. Certainly pick pocketing and bag snatching are popular criminal pastimes around the Mediterranean. I wonder how far east this holds true.
Complain about this comment
I'm no statistician but I may be better than MK_Steve. When talking about the % increase in "thefts from the person" it is absolutely correct to compare the proportion in one year to the proportion in another year (raw numbers can be used if the sample is the same size in each case; if the sample sizes are different then they have to be brought to a common base). Thus if we have a sample of 1000 people and in year 1 100 are robbed (10%) and in year 2 110 people (11%) of a sample of 1000 are robbed then the number of people who have been robbed has increased by 10% (110 cf. 100). I do agree, however, that in the absence of the actual statistical tests being identified the true significance of the data and the derived information is lost. There is an old adage about "lies, damned lies and statistics"; statistics do not lie but the way they are used and interpreted can be used by the unscrupulous and the ignorant to blind the unwary and the ignorant. under any circumstances a sample of over 44000 is to be regarded as a statistically large sample and the proper derivation and interpretation of information from that data needs proper statistics, properly calculated, with clearly identified tests and an open and unbiassed discussion. MK_Steve is correct in stating that if the tests are unknown then the value of the inforamtion and interpretation is nil.
Complain about this comment
Reply to #20 Redmistrising.
You make some interesting points, but I disagree with your central premise.
Here's a fun and well known (I claim no credit) puzzle to illustrate my point(try it, before looking at the answer):
There are 2 hospitals. The largest one has 100 children born every day approximately, the smaller one has 50 approx. In each hospital they donate £1 to charity for any day where more than 60% of births are girls. After one year of doing this, which hospital has donated the most money? Try it.
Without thinking, I would say they would be about the same, wouldn't you? Wouldn't we all (if we're honest)? The answer however is the smaller hospital would donate more. The reason is because proportions vary more widely with smaller effects as the n gets smaller. Humans are known to have an intuitive blindness to this.
You claim that 44,000 is a big enough sample, but your stats comparison disregards sample size completely, making that claim irrelevent. Your method (comparison of proportions between different size groups) mean that the large sample size is irrelevent because the proportional (eg 25%) increase is based only on the dependent cases (and is mathematically blind to the rest of the sample).
Imagine the crime survey included 'murder' as follows: The result being 2 out of 45,000 people in 2007 and 4 people in 2008. That's a 100% increase in murder by your method. This is not 'incorrect' but the figure should not be compared to other proportions calculated from crime categories with much large sizes. The first graph (above) would show murder as being the most increased crime by a long way compared to the others, which is simply a factor of a small n. And it's that comparison of proportions based on different effect sizes (and negleting the even sample size) that is the problem.
In summary, proportions based on different numbers of cases are not valid for straight comparisons, and the the sample size is irrelevent when dealing only with the dependent cases in this way.
Sorry it's so long.. hope you're still awake!
Complain about this comment
I'd just like to add that there's a fundamental error in this whole 'it must be due to the recession' idea.
People often make the mistake that variable X causes variable Y, but forget to consider whether variable Y might cause variable X.
Has anyone wondered whether it isn't the fact that the recession is making ociety more prone to dishonest behaviour, but the fact that society's dishonest behaviour has caused the recession.
Consider Sir Fred, for example...
Complain about this comment
View these comments in RSS