sda2.jpg

January 5, 2006

Where Did He Get Those Numbers

A reader takes issue with David Frum's claimIn a National Post article *"Canada's crime rate now 50% higher than US"...

The Canadian survey reports a rate of property victimization of 248 per 1000 households, and a violent victimization rate of 106 per 1000 persons (age 15 or older). The US survey reports a rate of property victimization of 162 per 1000 households, and a violent victimization rate of 23 per 1000 persons (age 12 and older).

[...]

Anyway, if these numbers are really comparable, the jaw-dropping comparison is between the US violent victimization rate of 23 per 1000 versus the Canadian rate of 106 per 1000 - that's more than a factor of 4 difference! For property crime alone, the Canadian rates are only 50% or so higher (248 per 1000 in Canada vs 162 in the US).


So there, Mr. Frum!

Posted by Kate at January 5, 2006 10:05 AM
TrackBacks

Friday Coffee-Break Links from Maggie's Farm
Schwartzenegger bends to the will of the people. No-one could have made a better or more earnest effort to bring Calif. into reality, but even he was not strong enough. Too bad for Calif.You will not believe what this Vermont liberal judge did. Ex-Donk. I [Read More]

Tracked on January 6, 2006 10:31 AM

Comments

I'm not sure what conclusion you're trying to draw, here, Kate. The violent crime figures would seem to confirm Frum's analysis, no?

Posted by: road_hammer at January 5, 2006 10:38 AM

To me, the comment sounds like either thinly veiled sarcasm, or she is commenting that Frum grossly underestimated the crime rate in Canada. Either way, we have no reason to be proud.

Posted by: michael at January 5, 2006 10:42 AM

No real surprise. Most states allow the citizens the Right to Keep and Bear Arms. That makes the bad boys and the government (same difference) think twice.
http://www.packing.org/
http://www.packing.org/state/images/size_400/all_united_states_phpSWVS3u_all_usa_map.gif

RJM

Posted by: RJM at January 5, 2006 11:15 AM

Yes, I was being facetious - Frum underestimated the violent crime rate differences.

Posted by: Kate at January 5, 2006 11:21 AM

The 50% figure was gleaned comparing "violent crimes" on a national model and on a percapata of 1000.

The crimes you cite such as "property crime" includes a number of codified crimes which need not be "violent: in nature...like fraud or theft...these are not neccessarily violent, let alone armed, crimes.

From the stats I have see from the FBI crime stat data and the RCMP stats...the figure of 50% higher 1000 percapata violent crimes in Canada is in the ball park.

If you do a simple comparison of just TO violent crimes comparing it to NY ..TO has the higher violent crime rate per capata.

Posted by: WL Mackenzie Redux at January 5, 2006 11:38 AM

If all citizens have arms the criminals think twice. Are you listening slimball martin?

And they can have my gun when they pry it from my cold dead fingers.

Posted by: FREE at January 5, 2006 11:55 AM

See
http://otherclub.blogspot.com/2005/12/bah-nanada.html
and
http://otherclub.blogspot.com/2006/01/causality.html
for links to statistical sources and comparability of crime rates.

Posted by: Hershblogger at January 5, 2006 1:09 PM

John Moore on CFRB 1010 AM here in Toronto called me a "pinhead that he likes to kick around the block" on air while he cut me off three different times, all while bragging about his high standards of debate. His premise for discussion and the reason that I called into his show was... that there were no easy answers for our "gun crime", and that only long range social engineering farts in the air were going to solve our problem. I brought,... perhaps we could look at what our best friend to the south has been doing so successfully and it was this stat. from Mr. Frum that I referred to that sent Moore into a bizerk tirade of anti Americanism right after he cut me off for the final time.
This is the same guy who spent 15 minutes telling Torontonians to look the other way and assured all Canadians that there is no Income Trust scandal.
We are constantly bombarded with liberal fart catchers parading about as reporters here in Toronto.
Moore BTW landed the great horror (Whoops), honour, of interviewing Fat Paulie on air.

Is this the standard payment for being a whored out liberal hack faux reporter propping up a criminal regime in the decayed dominion?
We need more Frums and less Moores,... way less Moores.

Posted by: richfisher at January 5, 2006 1:18 PM

I’m no expert on crime stats, but I am a full-time sceptic. So I still don’t think David Frum’s crime rate numbers add up, or should be used to support what are otherwise very valid arguments to get tough on crime.

While I completely agree with Mr. Frum’s point-of-view on Canada’s pitiful anti-crime policies, the “50%” stat in his article really caught my eye as being highly questionable. So I’m glad he provided his sources to Kathy and her good reader, since it allows closer analysis.

To my mind, using those particular sources to derive and compare an aggregate “overall crime rate” (the term used in the Frum article) between the United States and Canada isn’t a good idea, especially when the results are so counterintuitive.

For example, if as Mr. Frum reported, Canada has a 50% higher “crime rate” than the U.S., that notion certainly runs counter to crime statistics based on data reported by police services. You can see the most recent for Canada here:

http://www.statcan.ca/Daily/English/050721/d050721a.htm

and for the U.S. here:

http://www.fbi.gov/ucr/cius_04/offenses_reported/index.html

For directly comparable violent crimes like “murder”, the 2004 U.S. rate per 100,000 runs almost three times higher than Canada’s (5.5 compared to 1.9), while “robbery” runs 1.5 times higher (136.7 compared to 86). Other crimes like “assault” are harder to compare, due to different definitions used.

For property crimes, direct comparisons are harder still, again due to offence definitions. But something like “motor vehicle theft” sounds close, and here Canada’s car thieves are obviously doing their bit to help support Frum’s argument. Canada has a 1.3 times higher rate than the U.S. (531 compared to 421.3). Comparing the U.S. crime of “larceny - theft” to Canada’s combined “theft over and under $5,000” may be a stretch, but the rates are close (2365.9 in the U.S. versus 2185 in Canada).

So if one is willing to accept these numbers as representing criminal activity, it suggests that the U.S. has generally higher violent crime rates than Canada, while property crime rates are comparable.

Now perhaps the fact that the rates are even close might give some Canadians pause for thought – but it’s hardly a “read that again slowly” moment.

However, instead of reported crime rates, Mr. Frum bases his analysis on “victimization rates”, that is, surveys asking people whether they believe they’ve been a victim of a crime, even if they have not reported it to police. That leads to the question of what appears to be very different reporting rates between the U.S. and Canadian sources. For example, apparently U.S. citizens report 85% of their motor vehicle theft, while Canadians report just less than 50%? (What kind of beaters are we driving up here?) One has to question why there is such a big difference between the two countries before glomming onto these stats as supporting any argument, pro or con.

But the trouble Kathy’s reader had reconciling these crime rates, and the wild comparison they obtained, shows another, larger red flag, meaning that the sources given probably aren’t really comparable at all. Look at some of the raw numbers for occurrences, no longer obscured by an uncertain denominator, and consider whether they seem reasonable.

For example, according to the two sources Mr. Frum used, the U.S. (with around ten times the population of Canada) had 501,820 robberies in 2004, while Canada had 274,000 (55%). There were 4.5 million physical assaults in the U.S., while Canada had 1.3 million (29%). And if you believe the comparison, Canada actually had more sexual assaults, at 512,000 versus 209,800 in the U.S.

These numbers just doesn’t pass the sniff test. In my opinion, it doesn’t take much to see that they aren’t comparable, and so should not be used to bolster anyone’s argument if they want to be taken seriously. (In fact, when the numbers point in the opposite direction, it may force their user onto the defensive to prove why the opposite hypothesis isn’t true. For example, if the crime rates are actually lower in Canada, maybe the Canadian approach to crime is correct…which of course it isn’t.)

Note that I am not trying to counter Mr. Frum’s conclusions. I only caution against using stats that, on their face, seem pretty questionable, and on further analysis, aren’t really valid. They have too much of an undermining effect on what are otherwise valid points.

Posted by: 2Sheds at January 5, 2006 2:19 PM

What a load of liberano CRAP!!!!

We the people are really tired of the lies and misinformation being put in front of us by slime like 2sheds. I can't wait for the day that you guys start getting put behind bar.

Posted by: FREE at January 5, 2006 2:35 PM

The difference in the murder rate is less than that. The Canadian rate is the convicted rate, while the American rate includes justified homicides, both by police and citizens, and other "not guiltys".

RJM

Posted by: RJM at January 5, 2006 2:46 PM

We may not report our car thefts because we have lost our confidence in our law enforcement agencies, not because it was an old "beater".
In my experience, the Americans get more indignant and are more likely to take action against "anything" they perceive as being an infringment apon their persons and thus be more likely to report.
My impression has been that crime and and peoples proximity to each other have a direct corrolation. Shouldn't a defined crime be compared in cities of equal size?

Posted by: Al at January 5, 2006 2:57 PM

RJM, thanks for the observation. However, the way I read the StatsCan material, it looks like they use police reported occurence rates, not court reported conviction rates. They have other material on conviction stats, but of course I could be wrong.

Meanwhile, the FBI material I used says that the classification for murder "...does not include the following situations in this offense classification: deaths caused by negligence, suicide, or accident; justifiable homicides; and attempts to murder or assaults to murder, which are scored as aggravated assaults." See:

http://www.fbi.gov/ucr/cius_04/offenses_reported/violent_crime/murder.html

Posted by: 2Sheds at January 5, 2006 3:27 PM

"especially when the results are so counterintuitive"
Counterintuitive to whom ?
It doesn't matter what the truth is as long as it isn't counterintuitive to Truedeupian socialists.

Very lame analysis 2sheds especially the "lets not use that figure because it does'nt work in my version of reality, the exact same version that Mom and Dad used to elect Truedeau 40 years ago."

Posted by: richfisher at January 5, 2006 3:54 PM

If you report a smash and grab (which is a daily occurence on my 150 foot long street here in Chinatown Toronto, Jack Laydowns old riding soon to be his lifepartners Olivia Chows)the cops don't even bother coming around, and if you report it it goes on your insurance record which spells premium hike!


The crackheads who do it try to target out of province licence plates because they don't know how to report it, there are NO cops on the streets ever, just parking enforcement not-a-cops during daylight hours only.
We simply see more street sweepers to sweep up the broken glass clues, and less and less police on the steet.
I'm off to 52 division four blocks away now to give them the rifled suitcase I found last night, who's rightful owner probably does'nt think Frums statistics are as counterintuitive as antlerless er uh 2sheds.

Posted by: richfisher at January 5, 2006 4:26 PM

"However, the way I read the StatsCan material, it looks like they use police reported occurence rates, not court reported conviction rates."

Therefore, they may indeed use only court-reported convictions. We don't really know as you simply are giving your personal take on the StatsCan data. What does StatsCan say about its classifications of certain crimes as well as its sources (court-reported convictions or reported occurence)?

Although the FBI doesn't classify certain deaths as murder, you never stated if it uses court-reported convictions or reported occurences in its data.

Posted by: Jess at January 5, 2006 8:25 PM

I tried to report my stolen car to the Durham police force, and no one was there to accept the VIN number!?! I reported to the Airport Police, ( in Toronto), that my motorcycle had been rear-ended by a towlhead- the cop said that "Perhaps the driver thought you were going to run the red, so he didn't have to stop?" And: " I do not think a car did that damage to your bike- I think you took it home and took a sledgehammer to it." And when I tried to report a hit and run to the Toronto Police department, they weren't interested. I'm still here- and the next one is mine. MY Way! (Crime 'stats, my ass!)

Posted by: dave at January 5, 2006 9:37 PM

I've had my house broken into twice, my garage broken into once, and my car broken into once, my gas tank drained a couple of times, three bikes stolen, it took the police anywhere from two to six hours to respond to the break- ins. Too Busy! I was told they were unlikely to catch the culprits. Total cost to me,twelve thousand dollars. I never reported the gas or the bikes, why bother? I now have a very good alarm system and keep a gun and a shovel handy! I have no young children to worry about and now you steal from me at your own peril. Lies, damn lies, and then there are statistics!

Posted by: Greg at January 5, 2006 10:13 PM

Richfisher asked: “Counterintuitive to whom ?”

Well, apparently it was counterintuitive to the author, David Frum. Why else would he write, after quoting the 50% stat, “Read that again slowly -- it seems incredible, but it's true.” No one would write that unless they felt the rate was surprising and counterintuitive. And since, on closer analysis, the numbers appear flawed, perhaps that scepticism is fully justified.

Rich then suggests I'm a “Truedeupian socialist”. Odd, since I’ve always voted conservative. Me being a socialist will no doubt come as a great surprise to the many people I’ve been nagging to vote Conservative this election. I guess it means I’ll have to get rid of the big blue election sign on my lawn.

As for “my version of reality”, Rich once again makes an incorrect assumption that I’m a socialist. Not that it matters to anyone, but my version of reality is that we need more police on the street, more targeted enforcement efforts, tougher sentencing with mandatory minimum jail sentences for firearms offences, and less lenient treatment for young offenders. My version of reality is that we don’t need to make handguns more illegal, we need to adequately enforce the laws that already exist. (Rich, which parts of that do you disagree with?)

But my version of reality is also that you don’t get support for these kinds of measures by quoting unrealistic stats that opponents can easily tear to shreds, which is my main point.

In any event, the rest of Rich’s comments are just ad hominem attacks (and not even good ones), based on the unbelievably simplistic idea that if I disagree with one of Mr. Frum’s stats, I must be a closet socialist. Further, his recounting of some personal experience with minor property crimes in his own area (which is actually quite nice...I’ve spent lots of time there, day and night) doesn’t say a thing about the validity of the numbers presented.

So just to refocus, my point is only against using bad stats to support good arguments. I think it’s a poor approach that leaves us more vulnerable to opponents.

Posted by: 2Sheds at January 6, 2006 1:34 AM


So I actually took two stolen bags to 52 Division tonight, an out of towners suitcase and a security guards personal effects. Found the second one in the fifty foot walk to my car.

2Sheds
Sorry about the adhominim thing.
It was just too tempting (dropped antlers are known as sheds)

"Why else would he write, after quoting the 50% stat, “Read that again slowly"
Because it's the underlying point of his argument perhaps.

Glad to hear you are like me in trying to convince a few of my friends to vote conservative.
I can see we agree on a lot of other issues
Agreed, stats are never what they appear, Frums article is great and I believe his argument holds water.
I'm not willing to take the leaps of faith assumptions that you ask in your 20 paragraph "critique".

Keep up the good work...
And please lock everything in your trunk when you next park in China Town (were there are no blue election signs)

Posted by: richfisher at January 6, 2006 2:41 AM

Hey, you watch "Trailer Park Boys," right? That's how criminals view Canada's jails. No joke. What was it Ricky said? Ah yes: "I don't mind jail at all. Lots of liquor, three meals a day, good dope."

Posted by: Dante at January 6, 2006 6:24 AM

I am quite sure that TPB is not meant to reflect anything other than an attempt to be funny.

2sheds, your posts have been excellent, and I like to see the discourse elevated above some of the others here.

Using racist slurs to describe the people who rear-ended you is not going to convince anyone that 2sheds is wrong, only that you're a racist whose opinion no one should listen to.

Posted by: Steve R at January 6, 2006 3:43 PM

Sorry for being late with this, since people have probably moved on to other posts. But two quick comments:

Steve: many thanks for your kind words.

Rich: our different interpretations on stats aside, I applaud you and thank you for getting involved by taking those stolen bags in to 52 Division. In my experience, many people, particularly in Toronto, just look the other way, which I think is a sad commentary on our sense of duty as citizens. You are obviously someone who feels he needs to pitch in and help. If more people took action like you, maybe the crime stats (whatever they are) would go down.

Posted by: 2Sheds at January 7, 2006 11:44 AM

Fascinating to see how 2sheds useful analysis drew an immediate howl of rage from Free, who sounds like a kid who just had his favorite toy taken away. Going back to Canadian Fed. stats it seems that victimization stats have their origin in the feminist campaign to pump up unsatisfactorily low Statscan numbers on crime affecting females. And now we have a Frum climbing aboard that nasty little "special interest". They look about right posing togethner, sort of Norman Rockwell bitter enders, in front of a banner: Don't Confuse Us With Facts.

Posted by: garhane at January 8, 2006 4:42 PM
Site
Meter