Abortion = lower crime?

[barackattack]

How do you feel about that?

[natty dread]

Why would I feel anything about it? I wasn't even alive at the time.

[barackattack]

Neither was I.

[natty dread]

What's your point?

[Baron Von PWN]

Canada has slightly more police than the U.S. but also a higher crime rate.

I will also dispute this and do some basic leg work on wikipedia.

Crime rate: This appears to be a little hazy, due to differences in recording crime rates. For instance the Canada includes vehicular crimes as well as fraud which is not included in the US stats. Then there is the problem of different classifications of crimes (what in the USA is 1 crime "aggravated assault" is three different crimes in Canada).

So maybe?

Police: This seems to be false. Canada has about 195 officers per 100,000 residents while the USA has about 233 per 100,000 residents.


As for point 2. Do you have any evidence to suggest this is the case?

I decided to grab some crime stats from two states. 1 with 3 strikes and one without. For no particular reason I chose Texas and New York.

Texas (three strikes law)

Crime rate Index 1990 78.27/ 100,000 residents
2010 42.33/ 100,000

Change over ten yrs :-35.94/ 100,000
Sauce: http://www.disastercenter.com/crime/txcrime.htm


New York State (no three strikes)

Crime rate index. 1990 63.36/ 100,000 residents
2010 23.33/ 100,000

Change over 10 years -40.03/100,000

So crime has actually fallen at a more rapid rate in the non- three strikes state than the three strikes state.

To me this suggest's falling crime in the USA is a general trend and not one spurred by three strikes laws. Of course my little comparison is by no means exhaustive nor does it compare all 24 three strikes states with all non-three strikes states. However I think it sheds enough doubt for me to ask for some kind of backup to ole saxi's claims.


Anyway I would posit that abortion/contraceptives would indeed reduce crime rates by minimizing societal stressors. Fewer children to support means less poverty, less poverty results in less crime. Obviously there are other factors but I think availability of abortions and contraceptives keeps societal factors contributing to crime lower.

[barackattack]

Well according to the poll, the SNES and abortion are our clear winners.

Would it be cheaper to go to poor areas and give out PS3s or abortions?

[natty dread]

It would be cheaper to tape a picture of your face on top of the face of every poor woman. That would keep those poor bastards from spawning all over the place.

[BigBallinStalin]

Who wouldn't date Idi Amin?

He's wealthy, has clear goals, and is very driven toward achieving them.

[oVo]

So, I've read the first few chapters of Freakonomics and would like to regurgitate the information I have absorbed from this endeavour.

Crime in the US was HIGH.
And then it got EVEN HIGHER.
By the early 90s, commentators were predicting RIVERS OF BLOOD as RAPE became SPORT in the inner cities.
But then, crime started to fall (from '91 onwards).
Crime has dropped consistently, year on year. Violent crime has plummeted.

Freakonomics reckons that this is due to more abortion. The people most likely to use abortions are poor, uneducated, black, probably on drugs and poor. These people are most likely to have criminal children (Tupac, Jazzy Jeff etc.). By allowing these people abortions, you are allowing the 'thugs' and 'OGs' of the future to be neutralised before they get their hands on their first 'piece' (gun).

One of the dumbest statements ever posted here.

Is this a valid conclusion?

No, not even a decent troll.

The poll is a number of events from the year the crime recession began.

They may only be relative to the loss of IQ points incurred while reading this dribble.

[Lootifer]

My real life persona is screaming in outrage at this thread.

But no point in complaining...

Teasing out a single variable in a result statistic like crime rates is beyond retarded. As is shown by the very thin conclusions in that report linked earlier (most of it is assertion based on logic, which is fine, but terrible for making generic claims on relative magnitudes in causes).

In other words, put up (a 10 year, big budget, peer reviewed report with objectice results), or shut up.

[PLAYER57832]

In other words, put up (a 10 year, big budget, peer reviewed report with objectice results), or shut up.

Wrong criteria, but yes, the orginal post was not well put.

Studies have been done. Like most things biological/sociological/medical, the conclusions are not fully clear, but there IS a link. It is a link that is not "tasteful" to scientifically explore, though. And, its not really (as was implied above) based on race, nor specifically on income.

[Lootifer]

Oh im not saying that abortion doesnt have a genuine causal link to lower crime rates, but is it linear? is it co-dependent? is it an open loop or require iteration (ie abortion => lower crime => effect on abortion)? is there interferance? is it even possible to calcualte statistical significance? is there a risk of another unknown being actual source of effect and abortion is just a co-incidence? etc etc...

I speak from experience in the field: No systems modeller would touch this topic with a fifty foot barge pole.

No without 20+ staff and a multiyear budget for said staff. Its an IMMENSE project.

[PLAYER57832]

Oh im not saying that abortion doesnt have a genuine causal link to lower crime rates, but is it linear? is it co-dependent? is it an open loop or require iteration (ie abortion => lower crime => effect on abortion)? is there interferance? is it even possible to calcualte statistical significance? is there a risk of another unknown being actual source of effect and abortion is just a co-incidence? etc etc...

I speak from experience in the field: No systems modeller would touch this topic with a fifty foot barge pole.

Well, you are talking to a scientist who has designed more than one successful project. Its not quite as complicated as you indicate.
BUT.. the trick with human behavior is that almost nothing is linear... yet, that does not mean there is no cause. (think of chaos mathematics predictions). No one will touch on this much because it hints at eugenics and such. Evne so, there is a pretty direct link which I described above. I won't repeat it, though. Just think about it. Its a lot more obvious than anything said in the OP.

[Lootifer]

Heh, you dont get to design diddly though. The experiment's already been carried out and you (the analyst) have to deal with the data.

You're prob right, lack of a clear and concise analysis into the situation is likely far more due to moral politics and selection bias (of sorts).

However having read the report Barrack linked made me think that noone had bothered to do a decent job (not that that reports not decent, just incomplete) because if they had it would have been sourced in that report.

[BigBallinStalin]

Which link?

This one? http://pricetheory.uchicago.edu/levitt/ ... me2004.pdf

If you're looking for large budgets over many years with large staffs, then after aggregation, Levitt's overview of a good chunk of the literature should surpass your requirements. =P


It's been well over a year since I've read it, but many of the comments, questions, and criticisms ITT were handled pretty well by Levitt. Of course, I'd be crazy to expect most of the people ITT to value the time spent in reading it. It's so much easier to blindly stumble about and yell at each other.*


*I'm not implying you act like this. You win the award for Most Reasonable Skepticism ITT. Congrats.

[Lootifer]

Oh its a very good report like I said; i'm probably being far too anal, I just think he's been a tad too subjective in some areas.

edit: Hrmm, maybe I skimmed first time.... Ah f*ck it.

[BigBallinStalin]

Yeah go back to your energy sector. Criminology don't need your kind round here.

[Lootifer]

lol man im bad at reading...

Factor | Homocide | Violent Crime | Property Crime | Certainty of results
Strong economy 0 0 -2 High
Changing demographics 0 -2 -5 High
Better policing strategies -1 -1 -1 Low
Gun control laws 0 0 0 Medium
Concealed weapons laws 0 0 0 High
Increased usage of capital punishment -1.5 0 0 Medium
Increases in the number of police -5.5 -5.5 -5.5 Medium
Increases in the prison population -12 -12 -8 High
The decline of crack -6 -3 0 Low
Legalized abortion -10 -10 -10 Medium

[Baron Von PWN]

Which link?

This one? http://pricetheory.uchicago.edu/levitt/ ... me2004.pdf

If you're looking for large budgets over many years with large staffs, then after aggregation, Levitt's overview of a good chunk of the literature should surpass your requirements. =P


It's been well over a year since I've read it, but many of the comments, questions, and criticisms ITT were handled pretty well by Levitt. Of course, I'd be crazy to expect most of the people ITT to value the time spent in reading it. It's so much easier to blindly stumble about and yell at each other.*


*I'm not implying you act like this. You win the award for Most Reasonable Skepticism ITT. Congrats.

Some of their reasons for decreased crime seems a little dodgy. I'm talking about the More prisons reason. Here in Canada we saw a similar decrease in crime and yet we didn't increase our prison populations in a similar manner (there was steep, +20 prisoners per 100k Canadians, increase from 1980-85 and then prison populations stayed relatively stable at around 110 per 100k Canadians, from 85-99 ).

Actually the increased police conclusion seems suspect as well since in Canada we actually decreased the number of police per capita over the same period and saw similar crime reductions.

[saxitoxin]

I will also dispute this and do some basic leg work on wikipedia.

Crime rate: This appears to be a little hazy, due to differences in recording crime rates. For instance the Canada includes vehicular crimes as well as fraud which is not included in the US stats. Then there is the problem of different classifications of crimes (what in the USA is 1 crime "aggravated assault" is three different crimes in Canada).

IIRC the Nationamaster data is from the UNICJR which converts national statistics to a uniform standard. The UNICJR data is accurate for comparing nation-to-nation crime trends given the 192 different reporting standards in 192 different countries.

• That said, I understand it's core to the worldview of many people to imagine their nations are paradises of tranquility relative to the U.S. and so, when statistics to the contrary are presented, all sorts of caveats and exemptions and conditionals are applied to the statistics to discount them as invalid. U.S. crime rates have consistently been the lowest in the industrial world in almost every category, outside of the glaring but rare category of homicide. Colloquially, Vancouver has always struck me as horrifyingly unsafe relative to San Francisco.* A real shithole with a fanatically policed tourist core (kind-of a rug covering the dirt). So I have no observational reason to believe the data doesn't mirror street-reality.

no credits to your account

* Granted, this may be impacted by my own frame of reference as I'm well-liked in The Tenderloin where I regularly enjoy sauntering about with confidence in my short-shorts and mesh tank-top accompanied by my girlfriends enjoying the various clubs and bars and am often greeted with humorous catcalls from homosexual Latino gangsters like "Hey Saxi! Nice legs!"

As for point 2. Do you have any evidence to suggest this is the case?

I decided to grab some crime stats from two states. 1 with 3 strikes and one without. For no particular reason I chose Texas and New York.

Texas (three strikes law)

Crime rate Index 1990 78.27/ 100,000 residents
2010 42.33/ 100,000

Change over ten yrs :-35.94/ 100,000
Sauce: http://www.disastercenter.com/crime/txcrime.htm


New York State (no three strikes)

Crime rate index. 1990 63.36/ 100,000 residents
2010 23.33/ 100,000

Change over 10 years -40.03/100,000

So crime has actually fallen at a more rapid rate in the non- three strikes state than the three strikes state.

Randomly selecting two states to evaluate and casting that as a national trend is disingenuous. When I "randomly" selected two states I got opposite results as you.

Washington
1993: 312,793 / 5,225,000 --> year of adoption of habitual offender law
2003: 312,814 / 6,131,298

Oregon
1993: 174,812 / 3,032,000
2003: 180,369 / 3,564,330

[BigBallinStalin]

Which link?

This one? http://pricetheory.uchicago.edu/levitt/ ... me2004.pdf

If you're looking for large budgets over many years with large staffs, then after aggregation, Levitt's overview of a good chunk of the literature should surpass your requirements. =P


It's been well over a year since I've read it, but many of the comments, questions, and criticisms ITT were handled pretty well by Levitt. Of course, I'd be crazy to expect most of the people ITT to value the time spent in reading it. It's so much easier to blindly stumble about and yell at each other.*


*I'm not implying you act like this. You win the award for Most Reasonable Skepticism ITT. Congrats.

Some of their reasons for decreased crime seems a little dodgy. I'm talking about the More prisons reason. Here in Canada we saw a similar decrease in crime and yet we didn't increase our prison populations in a similar manner (there was steep, +20 prisoners per 100k Canadians, increase from 1980-85 and then prison populations stayed relatively stable at around 110 per 100k Canadians, from 85-99 ).

Actually the increased police conclusion seems suspect as well since in Canada we actually decreased the number of police per capita over the same period and saw similar crime reductions.

It's complex; there's very likely to be more than 2 forces which play on the crime rates of Canada.

Also, cross-country analysis becomes difficult as crimes are categorized differently, policing strategies may be different, and there's that ambiguous culture factor which is difficult to mathematically model to show its particular effects on crime.

[BigBallinStalin]

I will also dispute this and do some basic leg work on wikipedia.

Crime rate: This appears to be a little hazy, due to differences in recording crime rates. For instance the Canada includes vehicular crimes as well as fraud which is not included in the US stats. Then there is the problem of different classifications of crimes (what in the USA is 1 crime "aggravated assault" is three different crimes in Canada).

IIRC the Nationamaster data is from the UNICJR which converts national statistics to a uniform standard. The UNICJR data is accurate for comparing nation-to-nation crime trends given the 192 different reporting standards in 192 different countries.

• That said, I understand it's core to the worldview of many people to imagine their nations are paradises of tranquility relative to the U.S. and so, when statistics to the contrary are presented, all sorts of caveats and exemptions and conditionals are applied to the statistics to discount them as invalid. U.S. crime rates have consistently been the lowest in the industrial world in almost every category, outside of the glaring but rare category of homicide. Colloquially, Vancouver has always struck me as horrifyingly unsafe relative to San Francisco.* A real shithole with a fanatically policed tourist core (kind-of a rug covering the dirt). So I have no observational reason to believe the data doesn't mirror street-reality.

no credits to your account

* Granted, this may be impacted by my own frame of reference as I'm well-liked in The Tenderloin where I regularly enjoy sauntering about with confidence in my short-shorts and mesh tank-top accompanied by my girlfriends enjoying the various clubs and bars and am often greeted with humorous catcalls from homosexual Latino gangsters like "Hey Saxi! Nice legs!"

They aggregated the crime data across 192 countries???? Yeesh...

[PLAYER57832]

Heh, you dont get to design diddly though. The experiment's already been carried out and you (the analyst) have to deal with the data.

You're prob right, lack of a clear and concise analysis into the situation is likely far more due to moral politics and selection bias (of sorts).

However having read the report Barrack linked made me think that noone had bothered to do a decent job (not that that reports not decent, just incomplete) because if they had it would have been sourced in that report.

I said from the beginning that barack's conclusions were bogus. That said, he is hinting at some things that are real and that have been legitimately studied.

BUT... the problem with good studies is that you generally get real answers. (seriously).

[Baron Von PWN]

I will also dispute this and do some basic leg work on wikipedia.

Crime rate: This appears to be a little hazy, due to differences in recording crime rates. For instance the Canada includes vehicular crimes as well as fraud which is not included in the US stats. Then there is the problem of different classifications of crimes (what in the USA is 1 crime "aggravated assault" is three different crimes in Canada).

IIRC the Nationamaster data is from the UNICJR which converts national statistics to a uniform standard. The UNICJR data is accurate for comparing nation-to-nation crime trends given the 192 different reporting standards in 192 different countries.

• That said, I understand it's core to the worldview of many people to imagine their nations are paradises of tranquility relative to the U.S. and so, when statistics to the contrary are presented, all sorts of caveats and exemptions and conditionals are applied to the statistics to discount them as invalid. U.S. crime rates have consistently been the lowest in the industrial world in almost every category, outside of the glaring but rare category of homicide. Colloquially, Vancouver has always struck me as horrifyingly unsafe relative to San Francisco.* A real shithole with a fanatically policed tourist core (kind-of a rug covering the dirt). So I have no observational reason to believe the data doesn't mirror street-reality.

no credits to your account

* Granted, this may be impacted by my own frame of reference as I'm well-liked in The Tenderloin where I regularly enjoy sauntering about with confidence in my short-shorts and mesh tank-top accompanied by my girlfriends enjoying the various clubs and bars and am often greeted with humorous catcalls from homosexual Latino gangsters like "Hey Saxi! Nice legs!"

As for point 2. Do you have any evidence to suggest this is the case?

I decided to grab some crime stats from two states. 1 with 3 strikes and one without. For no particular reason I chose Texas and New York.

Texas (three strikes law)

Crime rate Index 1990 78.27/ 100,000 residents
2010 42.33/ 100,000

Change over ten yrs :-35.94/ 100,000
Sauce: http://www.disastercenter.com/crime/txcrime.htm


New York State (no three strikes)

Crime rate index. 1990 63.36/ 100,000 residents
2010 23.33/ 100,000

Change over 10 years -40.03/100,000

So crime has actually fallen at a more rapid rate in the non- three strikes state than the three strikes state.

Randomly selecting two states to evaluate and casting that as a national trend is disingenuous. When I "randomly" selected two states I got opposite results as you.

Washington
1993: 312,793 / 5,225,000 --> year of adoption of habitual offender law
2003: 312,814 / 6,131,298

Oregon
1993: 174,812 / 3,032,000
2003: 180,369 / 3,564,330

You never stated where you got your numbers just put out the statement. If that is where you are getting your statement I will accept that Canada has more non violent crime.

On your second point with the two states. I'm not sure what you think this proves as both state's crime rates per capita have decreased at similar rates. Washington shedding 1 crime a year more than Oregon is hardly conclusive. (washington 59 per 100,00 in 93- 51 per 100 000 in 03, Oregon 57 per 100000 in 93 - 50 per 100 000 in 03). I remain unconvinced that repeat offender laws are an effective means of crime reduction and that this is not some broader national change.

[Baron Von PWN]

Which link?

This one? http://pricetheory.uchicago.edu/levitt/ ... me2004.pdf

If you're looking for large budgets over many years with large staffs, then after aggregation, Levitt's overview of a good chunk of the literature should surpass your requirements. =P


It's been well over a year since I've read it, but many of the comments, questions, and criticisms ITT were handled pretty well by Levitt. Of course, I'd be crazy to expect most of the people ITT to value the time spent in reading it. It's so much easier to blindly stumble about and yell at each other.*


*I'm not implying you act like this. You win the award for Most Reasonable Skepticism ITT. Congrats.

Some of their reasons for decreased crime seems a little dodgy. I'm talking about the More prisons reason. Here in Canada we saw a similar decrease in crime and yet we didn't increase our prison populations in a similar manner (there was steep, +20 prisoners per 100k Canadians, increase from 1980-85 and then prison populations stayed relatively stable at around 110 per 100k Canadians, from 85-99 ).

Actually the increased police conclusion seems suspect as well since in Canada we actually decreased the number of police per capita over the same period and saw similar crime reductions.

It's complex; there's very likely to be more than 2 forces which play on the crime rates of Canada.

Also, cross-country analysis becomes difficult as crimes are categorized differently, policing strategies may be different, and there's that ambiguous culture factor which is difficult to mathematically model to show its particular effects on crime.

True I don't claim my breife analysis has anything resembling the weight of the study. It's simply curious that these supposedly strong factors are absent from Canada when Canada has seen a crime reduction as well.