Eat Lead: Crime in America

[niky]

Thought this would be an interesting read, and topical. While many will cite the link between tough enforcement, gun control, demographics, etcetera and crime, here's an interesting twist:

Crime rose in America up to the 90's because of lead. Not the kind that comes out of a gun, but leaded gasoline.

http://www.motherjones.com/environment/2013/01/lead-crime-link-gasoline

Nevin dove in further, digging up detailed data on lead emissions and crime rates to see if the similarity of the curves was as good as it seemed. It turned out to be even better: In a 2000 paper (PDF) he concluded that if you add a lag time of 23 years, lead emissions from automobiles explain 90 percent of the variation in violent crime in America. Toddlers who ingested high levels of lead in the '40s and '50s really were more likely to become violent criminals in the '60s, '70s, and '80s.

Did America avoid the "Fall of Rome" by phasing out leaded gas? There's a correlation to teen pregnancy, too.

In states where consumption of leaded gasoline declined slowly, crime declined slowly. Where it declined quickly, crime declined quickly.

Not much else to say, but... wow. They've even found correlations beyond the United States. Interesting reads, all.

http://pic.plover.com/Nevin/Nevin2007.pdf

Of course, it's not as exciting as arguing about gun control...

 

[arora]

My mother tells me I used to chew on lead painted and trimmed wall corners as a child in the 70's. Now it's all making sense

I'm not so sure about any one single correlation tbh but I've always believed that physical environmental exposures can effect behavior. There is always the genetic aspect as well as the non physical environment to consider.

And no it's not as fun a topic as gun control

 

[Johnnypenso]

Look at enough stats and you can find correlations between almost anything. Almost impossible to determine cause and effect in any statistically significant way.

 

[PeterJB]

Hold one, were people drinking petrol or just smelling it?

 

[Exorcet]

Look at enough stats and you can find correlations between almost anything.

That does not mean they are to be dismissed.

 

[niky]

@Johnnypenso: Pays to read the entire article and the linked studies that discuss the relationship on the country level, the state level, the city level and even down to the borrough level. It's pretty exhaustive.

More:
http://www.sciencedirect.com/science/article/pii/S0160412012000566

Correlations are one thing when the causative mechanism is purely theoretical (economic factors, family factors, etcetera), but lead has been known, for a long time, to have effects in terms of mental health and development. Thus, higher levels of lead in the blood of young children do lead to mental issues in adulthood.

Here's the graphs from the end of the international paper, if you don't feel like reading through. Note that there's a nineteen year time lag because of the difference in treatment of juvenile crime and "adult" crime... which is also covered in the study.

Also, preschoolers can't shoot worth squat.

Doesn't mean it's the only answer, but the correlation factors actually fit better than poverty and demographics (percentage of young males in the population) do... which is not something I expected, honestly, when I started reading the article or the studies.

 

[ExigeEvan]

Hold one, were people drinking petrol or just smelling it?

Inhaling it, in the air after it's been combusted.

However, petrol vapour abuse has been a huge issue in Australian Aboriginal populations, so they have OPAL fuel instead.

http://en.wikipedia.org/wiki/Opal_(fuel)

 

[arora]

but lead has been known, for a long time, to have effects in terms of mental health and development. Thus, higher levels of lead in the blood of young children do lead to mental issues in adulthood.

I tend to agree with this.

 

[niky]

Vapor abuse? Kids around here use epoxy resins. Brain-melting stuff, to be sure.

 

[Johnnypenso]

That does not mean they are to be dismissed.

Also doesn't mean it should be taken at face value.

 

[Exorcet]

Also doesn't mean it should be taken at face value.

Not dismissing them doesn't mean taking them at face value.

 

[Azuremen]

Also doesn't mean it should be taken at face value.

Which no one here has done, but there certainly is a lot of interesting and worthwhile data here.

 

[Keef]

Mental effects of pollutants over time are certainly a factor in various aspects of society. Lately I've read some articles correlating cold-war era nuclear testing in the Western states to rampant disease in areas of the country effected by the fallout. Stuff like heart disease, diabetes, cancers, and even genetic predispositions to these in current and future generations are thought to have been fueled by the nuclear fallout. I don't have access to sources right now but the evidence is pretty compelling and logical.

 

[PzR Slim]

Look at enough stats and you can find correlations between almost anything. Almost impossible to determine cause and effect in any statistically significant way.

I listened to a leading UK toxicologist on the radio this morning who said he was very impressed with the work that had been done. He also said that they had carried out very good statistical analysis to negate the other factors from their findings. He concluded by saying that he needed to do more analysis on the findings and the statistical analysis to negate other factors needed more checking but his immediate conclusion was that it was very convincing.

And if you can simply, 'Look at enough stats and you can find correlations between almost anything.' How on earth have we managed to make the advancements we have in all sorts of fields by statistical analysis?

 

[daan]

so they have OPAL fuel instead.

Now known as Starburst.

 

[XS]

Look at enough stats and you can find correlations between almost anything. Almost impossible to determine cause and effect in any statistically significant way.

Chi square and correlations can indeed be drawn from two seemingly unrelated variables, especially with a large data set. However, all statistical tests are run with tests for relevance, errors, variance, and significance, the latter typically represented as "p-value". From personal experience I can tell you that tests of correlation are very reliable statistical tests and results are always backed with tests for significance, degrees of freedom, variance, and error analysis. Without going too far into it stats, studies such as this are typically statistically perfect. The stats that give the field a bad name are stats based on inaccurate or error strewn data sets, and stats done on the fly by gray journalism.

 

[niky]

What I think sells it for me is that there is a concrete and well-understood mechanism by which one influences the other, and the time delay is such that there's no argument as to which came first and is a possible causative factor. Thought it was kooky town at first when a friend from another site linked it, but the work looks solid.

 

[sumbrownkid]

This is a very clever title Niky.👍