<< HELMUT AND THE HIZBIES ~ MAIN ~ OSCAR MOSQUER >>
PICK YOUR NUMBERS
Big call from Tim Lambert: “Crime and violent crime in Britain peaked in the early 90s and [have] since plummeted.” His source? The British Crime Survey, an annual affair which asks some 40,000 Brits if they’ve been encrimed during the previous year. Is it trustworthy? Depends which side you’re on:
The political discussion about crime is often a numbingly boring argument about statistics. Overall crime recorded by the police seems to have risen (so the Conservatives rely on this statistic) while crime reported by the public seems to have fallen (so Labour rely on that).
Here’s an example, from the BBC, of that Labour/Conservative divide in action:
Violent offences in England and Wales reached record levels in 2004-5 with police recording one million crimes - up 7% from the previous year ...
[Labour] Minister Hazel Blears ... told the BBC that the separate British Crime Survey, which interviews people to ask if they have been crime victims, showed a decrease in violent crime.
But [Conservative] shadow home secretary David Davis said the police figures were “further evidence that the government continues to fail on violent crime”.
I’d lean towards the official police figures myself (although jerked-around crime counting methods make comparisons problematic), mainly because they’re, you know, official police figures. The British Crime Survey is just a survey.
And one with limits that might call into doubt its value as a definitive indicator of “crime and violent crime in Britain.” For example, certain crimes aren’t included in the survey:
* Drug dealing
* Crimes committed against individuals under 16 years of age (a Home Office estimate last year suggested that 14.7 million crimes were committed against children and adults, up from the survey’s 10.8 million adults-only figure)
* Rape and other sexual offences (In 2002, 11,676 rape cases reached court; only 655 resulted in conviction)
* Crimes against business
And there’s one other crime the survey doesn’t measure, because its victims don’t tend to answer telephones:
* Murder
So Lambert’s view is based on data that excludes statistics on all of the above. Interesting; seems he’s as much a cherry picker as Lumpp and Loftus, or whoever those guys are he’s always carrying on about. Britain’s murder rate is increasing:
Britain’s murder rate for the population as a whole has almost doubled in the past 20 years, with young men from poor backgrounds by far the most likely victims.
Last year Home Office statistics recorded 833 murders in England and Wales, compared with just 565 a decade ago.
Author and academic Norman Dennis—a member of the Labour Party, as it happens—dismisses talk of a crime decline:
Part of the problem, Mr Dennis maintains, is that the Home Office, along with a lot of academic criminologists, insist on denying the most fundamental fact about crime in Britain: that it has been, and is, increasing.
“That denial is just silly,” he says. “Fifty years ago, there weren’t 400 street robberies in the whole of Britain. In 2001, there wasn’t a single month in the borough of Lambeth – that is just one London borough – in which there were fewer than 400 street robberies. Crimes against the person involving violence go up every year. There is no doubt about that at all.”
From the same link, London and New York compared:
There were more than 2,300 murders a year in New York in 1991 and well over 100,000 street robberies. London, by comparison, had 181 murders and 22,000 street robberies in that year.
Last year, there were 538 homicides in New York. That means the murder rate has decreased by a factor of five over the past 13 years. London’s murder rate has not reduced at all over the same period: there were 186 homicides in the capital last year.
More astonishing still is the comparison in the statistics for street robberies. In 2003, the last complete year for which records are available, there were just 24,334 street robberies in New York – while in London, 38,490 people were robbed in the street.
It takes some time for the significance of that statistic to sink in. New York, from having had a rate of street robbery five times that of London a decade ago, now has 14,000 fewer street robberies every year than our capital.
More on the British Crime Survey and official police figures here; worthwhile Samizdata thread here.
Simple rule of thumb in Britain…...if the figures are “official” take ‘em with a pinch of salt.
Even the Bank of England no longer believes ONS (Office Nat Statistics) data nor does The City. Where Thatcher started falsifying unemployment statistics in the 1980s Blair-Brown have gone mad, so now your official statistics are about as useful as reading up on Soviet grain harvests in 1932.
Great Britain has not a single functioning institution and trust has been debased in so much so often that the idea of basing any kind of decision or analysis on govt numbers is like a shot in the dark
Speakin of crime, SBS is bringing out Fungus with double-hateline, this time about the U.S. soldier/“specialist” who was found guilty of hate crimes in Abu ghraib prison..kind of “Why did he do it?” etc but really just trying to keep up the momentum of the old new photos and stir up more trouble in the Global Village they keep harping on about.
It seems that there are lies, damn lies, statistics, and then the special statistics Lambert uses which belong in a universe of their own, one where the sky is a milky cream substance and people wear their shoes on their ears.
Posted by Andrea Harris, Administrator on 2006 03 05 at 11:16 AM • permalinkThere were several comments in the links on the record keeping by the British police. Apparently the Home Office started their survey in response to poor record keeping by the police, instead of trying to tighten up on that record keeping. They might have parsed the basic data set carefully, looking for ways to improve the record keeping, or at least to identify errors.
Isn’t the British police system a national force? If so, one would think that setting and enforcing nation standards would be relatively simple.
Nevertheless, the British “split personality” approach to recording, measuring, and evaluating crime statistics is not only silly, it’s counterproductive. If a group of people can’t agree on how to define and measure a problem, they can’t agree on how to solve a problem. If the original concept was to use the BCS as a means of checking police data, said concept went out the window long ago.
And then we have the all-too-obvious politically motivated cherry picking here. Granted, the system encourages such fallacious reasoning, but that doesn’t mean it should be used that way.
And using a fairly focused poll that doesn’t have the same criteria as police reporting is ludicrous. There would be some validity if the poll paralleled police reporting standards. But this way? Oi!
Add in the shifting police reporting baseline mentioned in the links, and I question the ability of anyone in Britain to describe their crime problem. This is hard enough to do with a steady base line of data and (I think) national reporting criteria (set by the FBI, I think) in the US. But at least we argue over the interpretation of the same data, or question the base line. This helps ensure a rational argument amongst the various parties.
But what’s really telling here is that the Resident Blogosphere Statistical Genius™), Dr. Tim Lambert, once again misses the need for reliable baseline data. He did so on the Lancet Report, and he does it here. A genius, you ask? If you don’t believe me, just ask Lambert.
Plus Lambert uses the term “cherry picking” in a derogative fashion, which is highly ironic, given that he is the master cherry picker.
Posted by The_Real_JeffS on 2006 03 05 at 11:33 AM • permalinkIn 2003, the last complete year for which records are available, there were just 24,334 street robberies in New York – while in London, 38,490 people were robbed in the street.
And New York has three times the population of London.
Posted by Steven Den Beste on 2006 03 05 at 12:13 PM • permalinkDidn’t that precipitous decline of violent crime in New York occur during the “fascist police-state regime” of Rudi Giuliani?
Posted by Spiny Norman on 2006 03 05 at 12:29 PM • permalinkYes, ‘official’ stats can be and often are manipulated. But so can the public sentiment Labor is relying upon for its rosier assessment—and much easier, too, if the background music provided by media reinforces a particulary point of view (yes, I know, perish the thought).
People in the U.S. routinely describe today’s economy—characterized by robust GDP growth, continuous increases in productivity and unemployment at 5%—as poor and getting worse. They routinely describe the Canadian health system as superior, Europe as more tolerant and civilized, etc.
Go figure (oops, bad pun).
Voyager, #8:
NO ! Not yet
Gotcha! And I agree with your sentiments.
Posted by The_Real_JeffS on 2006 03 05 at 12:42 PM • permalinkA simple 4-step method to minimize crime:
1. Institute corporal punishment.
2. Make prison time short but wretchedly miserable.
3. Any crime for which the sentence of more than seven years imprisonment is justified would be punishable by death.
4. No appeals process to exceed six months.Comments please. ;o)
This “survey” sounds about as rigorous as those “nobody I know voted for Nixon/Reagan/Bush” laments we used to read…
Frankly, I’d believe Ted Dalrymple and what he saw in his emergency room over all the surveys and statistics…
Posted by richard mcenroe on 2006 03 05 at 01:27 PM • permalinkHmmm.
1. I believe here in the US the category of “murder” involves and includes any death that results in the intentional killing of a person.
So infanticide, murder and manslaughter would be included in the overal category of “murder”.
2. I had read somewhere that many people simply don’t bother reporting many crimes anymore. In most cases the police won’t respond for quite some time and they often are either unable or unwilling to pursue the crime. Particularly if it’s a relatively “minor” crime that involves the loss of property rather than violent injuries.
The example I remember was an American living in London who regularly had items stolen from his backyard and home, but who gave up reporting the thefts after the first half-dozen times.
So even if the official numbers are cooked, it probably is far worse than that.
Posted by memomachine on 2006 03 05 at 01:46 PM • permalinked, on point #2….too true. I had a bike stolen once, the police flat out told me that they wouldn’t come down for a report.
OTOH, when I burglarized once, I had a police officer on scene in about 10 minutes after I called. Might have been a slow day, though.
Posted by The_Real_JeffS on 2006 03 05 at 02:44 PM • permalinkHeh, Boss Hog! I hope this ain’t contagious, else this place will end up looking like Margoland…HORRORS!!!!!
Posted by The_Real_JeffS on 2006 03 05 at 04:57 PM • permalinkThat means the murder rate has decreased by a factor of five over the past 13 years.
No it hasn’t—saying something “decreased by a factor of five” has no meaning (unless you’re subtracting 5, which happens to be a factor of something).
What he means is it decreased by 80%. (Actually 77%)
Why can’t Johnny Journalist count? (jus’ a little pet peeve.)
#27:
You are of course free to pick one definition of a work and ignore all the other meanings, most of us will also accept the other meanings the dictionary gives for a word.
One of my dictionaries gives “A quantity by which a stated quantity is multiplied or divided, so as to indicate an increase or decrease in a measurement” as definition of factor. The other dictionaries don’t I admit - maybe it’s an American thing. Usage drives meaning and all that…
Journalists understand so little about statistics and how to properly report them that they actually suck understanding out of people who read their reporting, as far as I can tell.
Posted by JorgXMcKie on 2006 03 05 at 06:30 PM • permalinkI have 120 per cent faith that these journalists are interpreting these statistics correctly, expecially if it fits in with my world view.
Posted by Margos Maid on 2006 03 05 at 07:35 PM • permalinkUsage drives meaning and all that…
True enough, but as rebase indicates, “factor” does have an extremely specific meaning in mathematics, and using it interchangeably with “divisor” is much like calling a bed “table” just because they’re kinda similar and both made out of wood. Sensible usage should drive meaning - otherwise it’s just glorified gibberish.
At any rate, I can’t speak for non-US English, but the same irritating habit is quite prevalent in German-speaking countries.
(I actually have another, semi-related pet peeve about people who use “five times more” when they mean “five times as much”...“five times more” has always implied “six times as much” to my ears, but maybe I’m just thinking too much.)
I tend to believe credible surveys over police reports for 2 reasons. They can account for differences in crime definition across nations by using the same definition in international surveys. Also, they can account for unreported crimes. If police got better and more trustworthy, one might see reported crimes going up while surveyed crimes went down, and the survey would be right.
However, the BCS disagrees with not only police reports but with international surveys in the mid-1990’s.
The ICVS (International Crime Victim Survey) shows Britain having among the most violent crime for industrialized democracies. Australia is the most violent and the US is in the middle ranking for crime. This is according to the most recent released ICVS for 2000, and the BCS agrees with the ICVS for 2000.
There are a some things to keep in mind when reading that the BCS shows a decline in crime from the 1990’s to the present. It’s not quite true.
The BCS overstated British crime in the 1990’s by asking people about crimes they had encountered over an, on average, 15 month period and called this a 1 year stat. They changed their procedure so that starting with 2000, 1 year stats actually cover a 12 month period. Of course, the stats on crime went down when you moved from recording 15 months worth to recording 12 months worth.
Further, the BCS goes into some detail, breaking down violent crime based on the relationship between perpetrator and victim. According to BCS data, family crime (e.g. wife beating) went way down, acquaintance crime (e.g. pub brawls) went down slightly, and stranger crime (e.g. mugging) stayed the same.
So, based on BCS data, the likelihood of being mugged over a 15 month period in the 1990’s is the same as the likelihood of being mugged in a 12 month period now. That sounds like crime is going up.
#13 NY Times captured the essence of this concept some years ago when their headline proclaimed “Despite falling crime rate, prison populations soaring”. As if one is a happy accident and the other merely a result of racist, classist conservative policy.
Posted by Vanguard of the Commentariat on 2006 03 05 at 11:44 PM • permalinkIn Britain in 1921, one crime was recorded for every 370 inhabitants of England and Wales. In 2001, it was one crime for every 10 inhabitants. So there has been a twelvefold increase since 1941, especially in crimes of violence.
There are no doubt many reasons why this is so, but the welfare state must come near the top of the list.
The real reason behind the huge drop in crime all over America in the 1990’s was worked out by the brilliant “rogue ecomomist” Steven Levitt. Forget zero tolerance, or more cops on the beat, or tougher jail sentences. As Levitt proved, it was Roe vs Wade US Supreme Court 1972. What had happened to all the young criminals? They had been aborted: 18 - 20 years earlier. What was missing in 1990 was the cohort of children who stood the greatest chance of becoming criminals.
The drop in crime has been sustained ever since. Other effects were a big drop in infanticide, shotgun weddings, as well as babies available for adoption - which has led to the boom in the adoption of foreign babies. According to Levitt, Roe vs Wade led to conceptions rising by 30% but births falling by 6% indicating that abortion was being used as a means of birth control - “a crude and drastic sort of insurance policy”.
In economist speak, the drop in crime was “an unintended consequence” of the Supreme Court’s decision 18 years earlier.44 I don’t get that it matters whether Ubique’s being serious or not; the figures say what they say. Levitt himself dislikes the outcome, and complained about it in several radio interviews, plus lays out massive cautions about buying into the implication that this is the real explanation ... BUT at the end of the day he stands behind his figures.
“Freakonomics” is the title of his book; no I haven’t read it.Posted by Stoop Davy Dave on 2006 03 06 at 03:16 PM • permalinkRegarding #42, there are several problems with Levitt’s theory. If true, it should work for other nations, like the UK. The UK has seen dramatic increases in crime since abortion has been made more available. The same is true for many EU nations.
Further, crime reductions in various US cities tracks very well with when they adopted such policies as broken windows policing.
Even Levitt doesn’t claim that nothing else matters. He just claims abortion was very important in reducing crime. It’s basically a more sensational restatement of prior conventional wisdom that a reduction in the proportion of young males in society was partially responsible for the drop in crime, which seems reasonable. However, given that this drop is more pronounced in European nations which have seen vast increases in crime over the past 30 years, it can’t credibly be claimed as the main reason.
I don’t get that it matters whether Ubique’s being serious or not
I haven’t read Freakonomics either, but I’m quite sure Levitt is too good an economist to call it “the real reason” (as though it could be the only one) or imply anything like “Forget zero tolerance, or more cops on the beat, or tougher jail sentences.” Any social scientist worth their salt knows that if you’re attributing the totality of an effect to one single cause, you’re probably wrong. At any rate, as far as I understand, all that Levitt did was show a correlation. That’s a far cry from “he proved such and such”.
Hence my puzzlement as to whether #42 was supposed to be a satiric send-up of Levitt’s argument, or if Ubique simply doesn’t read very well.
48
Ouch!
Oh I hate it when they do that!Posted by Stoop Davy Dave on 2006 03 07 at 05:20 PM • permalink
Page 1 of 1 pages
Members:
Login | Register
| Member List
If only the UK would ban firearms, especially handguns, I’m sure none of this would be happening.