7 ms·

I what some people here seemed to be missing, is this correlation isn't just lead was removed from the environment and 23 years later crime went down. It's that

I what some people here seemed to be missing, is this correlation isn't just lead was removed from the environment and 23 years later crime went down. It's that various areas has differing degrees of lead pollution and started reducing it at different times. The places that reduced lead pollution first, saw the drop in crime first, the places that reduced it later saw the drop later. The eras that initially had high levels of lead had a larger drop in crime than areas that always had a low level. And it's not just the USA either. The pattern holds pretty accurately for various nations around the world, like in Europe. Those who banned leaded fuel first, saw the drop in crime first, those who banned later, saw the drop later. And these nations had massively differing crime policies. Some nations increased prison sentences to try and deter crime, and crime went down. Some nations put a huge effort into reforming criminals, and crime went down. And some nations cut prison sentence, and crime went down.

shawkinaw 13y ago
Not to mention the fact that it makes a lot of sense. Statistics is not the only evidence supporting this theory.
- mendicantB 13y ago
  Statistics is the means of testing whether or not something is evidence. I understand what you meant, it's an intuitive conclusion, one that doesn't seem like magic. But, something making sense usually does not correlate with or serve as evidence of it being true.
  - DennisP 13y ago
    If you take a bayesian approach it affects your prior.
  - Symmetry 13y ago
    something making sense usually does not correlate with or serve as evidence of it being true. It's true that statistics is required to be sure if something is true, but it's a priori far more likely that a rise in violent crime is caused by a substance known to damage people's brains than by, say, the gravitational influence of Pluto. And we rightly demand careful confirmation when someone finds a particle traveling at a velocity greater than c than when it is found to be traveling at a velocity less than c. EDIT: And more to the point, I hope you don't find the hypothesis that violence in the 70s caused leaded gasoline in the 50s as that leaded gasoline in the 50s caused violence in the 70s, and that you would agree that one of those two hypothesis needs more evidence before we accept it. Granted, it could easily be some more complicated net of causality, but just between those two.
    - mendicantB 13y ago
      I didn't wish to bring in Bayesian approaches here so as to keep it simple. You are correct in your assertion about likeliness given a priori intuition, and I agree with the linking. But, there is a stark contrast between leveraging prior information to strengthen your estimates and stating that prior information serves as evidence or correlates with outcomes, which was my original assertion. Either way, I'm a big fan of leveraging common sense (viz Bayesian approaches) so I can't disagree.
      - Symmetry 12y ago
        If prior information didn't correlate with outcomes then it would provide no value and shouldn't be used.
charlieflowers 13y ago
Perhaps another study should be done following individuals. It seems a decent attempt at disproving the hypothesis. Take say 50 specific individuals ... actually look at where they lived each year of their lives, and estimate as accurately as possible their lead exposure. If they move cross country, "follow" them and estimate the change in lead exposure accordingly. Once you have an assessment of the lifetime lead exposure of the 50 people, see if it predicts their levels of criminal behavior or aggression.
- bjeanes 13y ago
  I'm not sure that would work. The 20 odd year lag seems to be a generational gap, not a delay in concrete effects on any specific person.
- gwern 13y ago
  Crime is rare enough that you wouldn't get any sort of conclusive result from the variations in lead level among 50 people. (Example: suppose 0.1% of people become murders in their lives, and lead is 100% responsible for every murder, and you had 2 groups of 50 people who had either tons of lead or no lead exposure at all; what will the result be? Likely, no murderers in your total group of 100 at all and certainly no difference between the two groups - despite extremely generous assumptions!)
FranOntanaya 13y ago
Aren't the politics required to enforce lead reduction the same that would tackle crime, though? I don't think there are many enviromentally-friendly corrupt parties.
- lmm 13y ago
  Huhwhat? What's corruption got to do with it?
- wisty 13y ago
  You wouldn't think there were many corrupt religious parties either, but politics doesn't always work that way.
- jtheory 13y ago
  The crime reduction reliably occurs 20 (23, from the graph?) years after the lead reduction; there's no response in the crime rate until the less-lead-exposed babies have become adults. The politicians who tackled crime (or didn't, or did whatever) 20 years ago haven't been in power most of the time since then, generally. I've read about this theory before; it's impressive and (to me, at least) pretty persuasive.
  - originofspecie 13y ago
    To paraphrase a comment I made further up, banning lead in gasoline is a binary initiative that doesn't take a long time to initiate. Cracking down on crime is a longer process.
mixmax 13y ago
If the theory is true it could also account for the fact that violent crime is more abundant per capita in urban areas compared to rural areas since lead from gasoline will obviously be more abundant in cities with lots of cars per square mile compared to rural areas with very few cars. Here's a link (2 minutes of googling..) showing the difference. http://www.victimsofcrime.org/library/crime-information-and-statistics/urban-and-rural-crime http://www.victimsofcrime.org/library/crime-information-and-...
- Timmmmmm 13y ago
  Yeah pretty sure that's because of other more obvious factors: * Greater anonymity in cities * Higher population density * Juxtaposition of poverty and wealth
  - mixmax 13y ago
    Are these factors so obvious? As an example poverty is most prevalent in rural areas according to this study: https://www.google.dk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&uact=8&ved=0CE4QFjAD&url=http%3A%2F%2Fwww.ucl.ac.uk%2Fdpu-projects%2F21st_Century%2Fmyths%2Fpdf%2520myths%2FMyth7.pdf&ei=--1UU6e7OcriywPYpIC4Aw&usg=AFQjCNEGtscaf76R3qr6UvFfRg2ZUeacBA&sig2=I7F-lTmVibFF-QyXv55YVg&bvm=bv.65058239,d.bGQ https://www.google.dk/url?sa=t&rct=j&q=&esrc=s&source=web&cd... Be careful of citing "obvious", it can be deceiving. Note that I'm not arguing that lead in gasoline proves a definite link between crime in rural vs. urban areas, just noting that there is the same correlation as stated in the article. I merely thought it was interesting, and that it might be the same effect at work due to the dilution of lead in rural areas because there are fewer cars.
    - jpatokal 13y ago
      Juxtaposition of poverty and wealth, not poverty itself. In most American cities somebody from a poor neighborhood can walk a few blocks and mug a rich person at knifepoint, whereas in the average poor rural area, there aren't really any rich people around to rob.
Flow 13y ago
Thank you for this really interesting comment. I wonder if the lead could leave the body, and if the effects it had are somewhat reversed over time, for those that lived in environments with much traffic. Do you have any idea about this?
- ekr 13y ago
  If I remember correctly (haven't read this article), there's some sort of offset between decrease in lead and crime reduction, which seems to suggest that it is infants and young children who are affected, or lead actually affects the development of some areas of the brain (the offset was explained by the fact that infants and children don't usually commit crimes, but they do once they get slightly older). So it is not lead itself, (or the presence of it in your organism) that makes you more likely to commit crimes, but having your brain development impaired by it while an infant that does. I don't have time to search for the article, but it came up with the last discussion on this topic here.
- duskwuff 13y ago
  My understanding is that lead poisoning during pregnancy and early childhood causes the brain to not develop normally, which is what likely leads to a propensity to criminal behavior. Removing lead from the body later in life would be unlikely to help, as the damage has already occurred.
achy 13y ago
It's always interesting when statistics correlate across multiple frames (time, location, etc.); however that still doesn't prove causality. I'm inclined to think that the causes behind an area being slow to remove leaded gasoline could also affect crime rates.
- edmccard 13y ago
  >I'm inclined to think that the causes behind an area being slow to remove leaded gasoline could also affect crime rates. The statistics show a 20-year lag between the banning of leaded gasoline and reduction in crime rates. So even if "unwilling to ban leaded gas" and "high crime 20 years later" could be correlated, it seems less likely that the same things that cause "willing to ban leaded gas" are also the cause of "lower crime 20 years later".
  - VLM 13y ago
    Its interesting to model the biosphere hydraulically. So lead from leaded gas drips into the barrel representing the local ecology, and the lead contaminant drips out of the biosphere at a rate that obviously must depend on local climate. Presumably lead contaminant flows into the oceans or gets buried in the ground at a rate dependent on local weather. Yet the stats show a nearly perfect and consistent shift 20 years after the addition of lead to the environment stops. That's just biologically weird, like a part of brain development doesn't depend on local lead levels, but on the first derivative of lead levels or something. It COULD make sense if lead only causes problems when inhaled. Not consumed in food, not drinking water, purely and exclusively inhaled. Even then, local dust storms and the like should totally mess with the supposed global 20 year effect latency.
    - Terr_ 13y ago
      Lead reacts with other elements! If it was as un-reactive as gold, it probably wouldn't be as toxic. Just for starters, burning leaded gasoline can involve pure Lead, Lead(II) oxide, Lead(II) bromide, and Lead(II) chloride. http://en.wikipedia.org/wiki/Tetraethyllead#Reactions http://en.wikipedia.org/wiki/Tetraethyllead#Reactions
abandonliberty 13y ago
The answer to any headline question is supposed to be 'No!' Thank you.
jrochkind1 13y ago
This reminds me very much of the study that correlated legalized abortion with a decline in crime -- I believe it also showed it wasn't just a general global trend with legalized abortion and dropping crime, but that places that legalized abortion first saw crime drop first etc. I think the conclusion is obvious: Removing lead from petrol lead to legalized abortion. Or was it vice versa? But seriously, I think we need statistical analysis showing the correlation between legalized abortion and unleaded gasoline. Perhaps it would actually show that places that legalized abortion first actually WERE correlated with places that unleaded gasoline first. It would be silly to assume causation there, even if the correlation were just as strong. Which would be a useful reminder about correlation and causation. Just because you have a correlation where causation is _plausible_ still doesn't mean causation.
dragonwriter 13y ago
What they might also be misssing -- and the article highlights this -- is that its not just a correlation analysis. There's also a fairly well-developed explanation for the causal mechanism that justifies the hypothesis -- lead has already been established to have significant effects on brain development and behavior (and been specifically linked problems in impulse control and executive functioning.)

shawkinaw 13y ago

Not to mention the fact that it makes a lot of sense. Statistics is not the only evidence supporting this theory.

mendicantB 13y ago
Statistics is the means of testing whether or not something is evidence. I understand what you meant, it's an intuitive conclusion, one that doesn't seem like magic. But, something making sense usually does not correlate with or serve as evidence of it being true.
- DennisP 13y ago
  If you take a bayesian approach it affects your prior.
- Symmetry 13y ago
  something making sense usually does not correlate with or serve as evidence of it being true. It's true that statistics is required to be sure if something is true, but it's a priori far more likely that a rise in violent crime is caused by a substance known to damage people's brains than by, say, the gravitational influence of Pluto. And we rightly demand careful confirmation when someone finds a particle traveling at a velocity greater than c than when it is found to be traveling at a velocity less than c. EDIT: And more to the point, I hope you don't find the hypothesis that violence in the 70s caused leaded gasoline in the 50s as that leaded gasoline in the 50s caused violence in the 70s, and that you would agree that one of those two hypothesis needs more evidence before we accept it. Granted, it could easily be some more complicated net of causality, but just between those two.
  - mendicantB 13y ago
    I didn't wish to bring in Bayesian approaches here so as to keep it simple. You are correct in your assertion about likeliness given a priori intuition, and I agree with the linking. But, there is a stark contrast between leveraging prior information to strengthen your estimates and stating that prior information serves as evidence or correlates with outcomes, which was my original assertion. Either way, I'm a big fan of leveraging common sense (viz Bayesian approaches) so I can't disagree.
    - Symmetry 12y ago
      If prior information didn't correlate with outcomes then it would provide no value and shouldn't be used.

charlieflowers 13y ago

Perhaps another study should be done following individuals. It seems a decent attempt at disproving the hypothesis. Take say 50 specific individuals ... actually look at where they lived each year of their lives, and estimate as accurately as possible their lead exposure. If they move cross country, "follow" them and estimate the change in lead exposure accordingly. Once you have an assessment of the lifetime lead exposure of the 50 people, see if it predicts their levels of criminal behavior or aggression.

bjeanes 13y ago
I'm not sure that would work. The 20 odd year lag seems to be a generational gap, not a delay in concrete effects on any specific person.
gwern 13y ago
Crime is rare enough that you wouldn't get any sort of conclusive result from the variations in lead level among 50 people. (Example: suppose 0.1% of people become murders in their lives, and lead is 100% responsible for every murder, and you had 2 groups of 50 people who had either tons of lead or no lead exposure at all; what will the result be? Likely, no murderers in your total group of 100 at all and certainly no difference between the two groups - despite extremely generous assumptions!)

FranOntanaya 13y ago

Aren't the politics required to enforce lead reduction the same that would tackle crime, though? I don't think there are many enviromentally-friendly corrupt parties.

lmm 13y ago
Huhwhat? What's corruption got to do with it?
wisty 13y ago
You wouldn't think there were many corrupt religious parties either, but politics doesn't always work that way.
jtheory 13y ago
The crime reduction reliably occurs 20 (23, from the graph?) years after the lead reduction; there's no response in the crime rate until the less-lead-exposed babies have become adults. The politicians who tackled crime (or didn't, or did whatever) 20 years ago haven't been in power most of the time since then, generally. I've read about this theory before; it's impressive and (to me, at least) pretty persuasive.
- originofspecie 13y ago
  To paraphrase a comment I made further up, banning lead in gasoline is a binary initiative that doesn't take a long time to initiate. Cracking down on crime is a longer process.

mixmax 13y ago

If the theory is true it could also account for the fact that violent crime is more abundant per capita in urban areas compared to rural areas since lead from gasoline will obviously be more abundant in cities with lots of cars per square mile compared to rural areas with very few cars. Here's a link (2 minutes of googling..) showing the difference. http://www.victimsofcrime.org/library/crime-information-and-statistics/urban-and-rural-crime http://www.victimsofcrime.org/library/crime-information-and-...

Timmmmmm 13y ago
Yeah pretty sure that's because of other more obvious factors: * Greater anonymity in cities * Higher population density * Juxtaposition of poverty and wealth
- mixmax 13y ago
  Are these factors so obvious? As an example poverty is most prevalent in rural areas according to this study: https://www.google.dk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=4&cad=rja&uact=8&ved=0CE4QFjAD&url=http%3A%2F%2Fwww.ucl.ac.uk%2Fdpu-projects%2F21st_Century%2Fmyths%2Fpdf%2520myths%2FMyth7.pdf&ei=--1UU6e7OcriywPYpIC4Aw&usg=AFQjCNEGtscaf76R3qr6UvFfRg2ZUeacBA&sig2=I7F-lTmVibFF-QyXv55YVg&bvm=bv.65058239,d.bGQ https://www.google.dk/url?sa=t&rct=j&q=&esrc=s&source=web&cd... Be careful of citing "obvious", it can be deceiving. Note that I'm not arguing that lead in gasoline proves a definite link between crime in rural vs. urban areas, just noting that there is the same correlation as stated in the article. I merely thought it was interesting, and that it might be the same effect at work due to the dilution of lead in rural areas because there are fewer cars.
  - jpatokal 13y ago
    Juxtaposition of poverty and wealth, not poverty itself. In most American cities somebody from a poor neighborhood can walk a few blocks and mug a rich person at knifepoint, whereas in the average poor rural area, there aren't really any rich people around to rob.

Flow 13y ago

Thank you for this really interesting comment. I wonder if the lead could leave the body, and if the effects it had are somewhat reversed over time, for those that lived in environments with much traffic. Do you have any idea about this?

ekr 13y ago
If I remember correctly (haven't read this article), there's some sort of offset between decrease in lead and crime reduction, which seems to suggest that it is infants and young children who are affected, or lead actually affects the development of some areas of the brain (the offset was explained by the fact that infants and children don't usually commit crimes, but they do once they get slightly older). So it is not lead itself, (or the presence of it in your organism) that makes you more likely to commit crimes, but having your brain development impaired by it while an infant that does. I don't have time to search for the article, but it came up with the last discussion on this topic here.
duskwuff 13y ago
My understanding is that lead poisoning during pregnancy and early childhood causes the brain to not develop normally, which is what likely leads to a propensity to criminal behavior. Removing lead from the body later in life would be unlikely to help, as the damage has already occurred.

achy 13y ago

It's always interesting when statistics correlate across multiple frames (time, location, etc.); however that still doesn't prove causality. I'm inclined to think that the causes behind an area being slow to remove leaded gasoline could also affect crime rates.

edmccard 13y ago
>I'm inclined to think that the causes behind an area being slow to remove leaded gasoline could also affect crime rates. The statistics show a 20-year lag between the banning of leaded gasoline and reduction in crime rates. So even if "unwilling to ban leaded gas" and "high crime 20 years later" could be correlated, it seems less likely that the same things that cause "willing to ban leaded gas" are also the cause of "lower crime 20 years later".
- VLM 13y ago
  Its interesting to model the biosphere hydraulically. So lead from leaded gas drips into the barrel representing the local ecology, and the lead contaminant drips out of the biosphere at a rate that obviously must depend on local climate. Presumably lead contaminant flows into the oceans or gets buried in the ground at a rate dependent on local weather. Yet the stats show a nearly perfect and consistent shift 20 years after the addition of lead to the environment stops. That's just biologically weird, like a part of brain development doesn't depend on local lead levels, but on the first derivative of lead levels or something. It COULD make sense if lead only causes problems when inhaled. Not consumed in food, not drinking water, purely and exclusively inhaled. Even then, local dust storms and the like should totally mess with the supposed global 20 year effect latency.
  - Terr_ 13y ago
    Lead reacts with other elements! If it was as un-reactive as gold, it probably wouldn't be as toxic. Just for starters, burning leaded gasoline can involve pure Lead, Lead(II) oxide, Lead(II) bromide, and Lead(II) chloride. http://en.wikipedia.org/wiki/Tetraethyllead#Reactions http://en.wikipedia.org/wiki/Tetraethyllead#Reactions

abandonliberty 13y ago

The answer to any headline question is supposed to be 'No!' Thank you.

jrochkind1 13y ago

This reminds me very much of the study that correlated legalized abortion with a decline in crime -- I believe it also showed it wasn't just a general global trend with legalized abortion and dropping crime, but that places that legalized abortion first saw crime drop first etc. I think the conclusion is obvious: Removing lead from petrol lead to legalized abortion. Or was it vice versa? But seriously, I think we need statistical analysis showing the correlation between legalized abortion and unleaded gasoline. Perhaps it would actually show that places that legalized abortion first actually WERE correlated with places that unleaded gasoline first. It would be silly to assume causation there, even if the correlation were just as strong. Which would be a useful reminder about correlation and causation. Just because you have a correlation where causation is _plausible_ still doesn't mean causation.

dragonwriter 13y ago

What they might also be misssing -- and the article highlights this -- is that its not just a correlation analysis. There's also a fairly well-developed explanation for the causal mechanism that justifies the hypothesis -- lead has already been established to have significant effects on brain development and behavior (and been specifically linked problems in impulse control and executive functioning.)