Skip

Get the lead out.
January 3, 2013 6:45 AM   Subscribe

Experts often suggest that crime resembles an epidemic. But what kind? Karl Smith, a professor of public economics and government at the University of North Carolina-Chapel Hill, has a good rule of thumb for categorizing epidemics: If it spreads along lines of communication, he says, the cause is information. Think Bieber Fever. If it travels along major transportation routes, the cause is microbial. Think influenza. If it spreads out like a fan, the cause is an insect. Think malaria. But if it's everywhere, all at once—as both the rise of crime in the '60s and '70s and the fall of crime in the '90s seemed to be—the cause is a molecule.
Long story short, we all bought lots of cars after World War II and filled them up with leaded gasoline. This lead was spewed out of tailpipes and ingested by small children, and when those children grew up they were more prone to committing violent crimes than normal children. Then, starting in the mid-70s, we all began switching to unleaded gasoline. Our kids were no longer made artificially violent by lead poisoning, and when they grew up in the mid-90s they committed fewer violent crimes. This trend continued for two decades, and it's one of the reasons that violent crime rates have dropped by half over the past 20 years and by more than that in our biggest cities.
The Impact Of Childhood Lead Exposure on Adult Crime:
The author concludes that lead's effect on violent crime may be just the tip of the iceberg. Increases in impulsivity, aggression, and ADHD can affect many other behaviors such as substance abuse, suicide, teenage pregnancy, poor academic performance, poor labor market performance, and divorce, suggesting that environmental policy can have far reaching effects on social outcomes.
But how? In The Crime Of Lead Exposure(previously):
Twenty years later, the researchers tracked down these subjects and put them in MRI machines, allowing them to measure the brain volume of participants. The researchers found that exposure to lead as a child was linked with a significant loss of brain volume in adulthood, particularly in men. Furthermore, there was a “dose-response” effect, in which the greatest brain volume loss was seen in participants with the greatest lead exposure. What’s especially tragic is that the loss of volume was concentrated in the prefrontal cortex, a part of the brain closely associated with executive function and impulse control.
Chicago Slums And The Long Shadow Of Lead Paint

We've known about the correlation between lead exposure and crime for a while - previously on MetaFilter. Even Cracked has gotten in on it.

Lead Exposure and Poverty: Have We Gotten Youth Violence All Wrong? Should we do more than just recommend people avoid lead paint and instead remove or mitigate it?

Is There Lead In Your House? Or perhaps those eggs from the community garden?

Of course, lead doesn't just poison humans. The California Condor is sick of (or from) lead in bullets.
posted by the man of twists and turns (140 comments total) 170 users marked this as a favorite

 
Awesomely timely post. I was thinking about the (potential, to me at the time) link between leaded gas and the crime rate just the other day.
posted by DU at 6:50 AM on January 3, 2013


Wow, awesomely constructed post!
posted by OmieWise at 6:59 AM on January 3, 2013 [2 favorites]


Before someone comes in with the same old thoughtlessly dismissive canard about correlation and causation, yes, we've all heard it. However, correlations when teased apart properly are pretty much how we know things - indeed more or less all the things. Of course crime and lead exposure co-correlate with poverty, where the primary means of causation for both crime and exposure to lead could come from poverty - however, the reverse also appears to actually be pretty fucking plausible. Lets think through it.
posted by Blasdelb at 7:00 AM on January 3, 2013 [25 favorites]


The Nation, from 2000. The Secret History of Lead.
posted by Devils Rancher at 7:05 AM on January 3, 2013 [6 favorites]


I live in one of the highest-lead counties in the U.S. Statewide, about 14% of students have IEPs (that is, qualify for special ed services). About 15% of students in high-poverty districts like mine qualify for IEPs, a slight bump that is fairly expected due to the effects of poverty (and particularly poor nutrition) on child development.

My district has an IEP rate of nearly 19%.

An awful lot of the difference, I am told by health officials, is the very high rates of lead exposure for poor children caused by unremediated old housing stock in the poor areas of town and industrial pollution from heavy industries located near that old housing stock. City officials are super-interested in reducing youth crime and improving schools, but only if it involves excoriating schools for failing and complaining about bad parents, not if it requires them to deal with their lead problem.
posted by Eyebrows McGee at 7:10 AM on January 3, 2013 [18 favorites]


Nice post.
posted by Tell Me No Lies at 7:11 AM on January 3, 2013


The California Condor is sick of (or from) lead in bullets.

Long since time hunting rounds went to lead-free nationwide. I regularly shoot pellet guns in my backyard, and only load lead-free pellets: I've got a 3 y.o. and the neighbor's chickens and ducks wander over regularly - little bits of lead in the yard is not acceptable. It's no longer acceptable in the wild, either.

I'm not the finest sharpshooter in the land or anything, but I can't tell much of a difference in accuracy, and it certainly has no problems punching holes in paper or knocking over cans.

Since I'm pissing off sportsmen, lead fishing weights need to go, too - lead-free split-shot and egg sinkers work more than well enough for me, and kids swim in that pond, you know?
posted by Slap*Happy at 7:13 AM on January 3, 2013 [2 favorites]


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....

And with that we have our molecule: tetraethyl lead, the gasoline additive invented by General Motors in the 1920s to prevent knocking and pinging in high-performance engines....

Like many good theories, the gasoline lead hypothesis helps explain some things we might not have realized even needed explaining. For example, murder rates have always been higher in big cities than in towns and small cities. We're so used to this that it seems unsurprising, but Nevin points out that it might actually have a surprising explanation—because big cities have lots of cars in a small area, they also had high densities of atmospheric lead during the postwar era. But as lead levels in gasoline decreased, the differences between big and small cities largely went away. And guess what? The difference in murder rates went away too. Today, homicide rates are similar in cities of all sizes. It may be that violent crime isn't an inevitable consequence of being a big city after all.
Fascinating. When you consider how much of American political culture has been driven by the rising crime rates of the 60s-80s and also by the differing rates of violence between big cities and the rest of the country, it's astonishing to imagine that all of that could be traced back to one simple technical decision by engineers at GM.
posted by enn at 7:21 AM on January 3, 2013 [13 favorites]




But Wayne LaPierre told me that the answer to our violence problem was MORE lead!
posted by The 10th Regiment of Foot at 7:25 AM on January 3, 2013 [2 favorites]


Oh Thomas Midgely, who won't you mass murder? :/
posted by sexyrobot at 7:26 AM on January 3, 2013 [7 favorites]


That. Was. Fascinating. Seriously.
posted by MuffinMan at 7:26 AM on January 3, 2013


Very interesting! Too good to be true? A quick Google on the websites of The British Society of Criminology, the American Society of Criminology, the journal Criminology and Criminal Justice and the journal Criminology indeed find nothing on gasoline/petrol lead, which is a shame. I'd like to hear the counter-arguments or a refutation. Otherwise it looks like a great opportunity to improve crime rates for relatively little cost and effort.
posted by alasdair at 7:28 AM on January 3, 2013 [2 favorites]


Previously on Metafilter.
posted by radwolf76 at 7:28 AM on January 3, 2013


Hmm, and what about the lead drinking water pipes of centuries past, I wonder?
posted by adamdschneider at 7:30 AM on January 3, 2013


lead fishing weights need to go,

Several states already ban lead weights that are small enough for birds to eat. That doesn't stop people from making their own though, which apparently happens a lot.

Interestingly the larger weights don't seem to contaminate much unless you are physically handling them.
posted by Tell Me No Lies at 7:30 AM on January 3, 2013


I was just wondering this morning about whether the fall of the Roman Empire had anything to do with its plumbing system. If lead can have far-reaching effects, what about fructose?
posted by infinitewindow at 7:30 AM on January 3, 2013


Does anyone have any info on lead polution in Brazil? Violence there is off the charts.
posted by TheyCallItPeace at 7:32 AM on January 3, 2013


The Lead In Bullets Is Creating More Pollution And Public Health Hazards Than Anyone Is Really Willing To Admit.

Incidentally, as a hunter (a lead-free hunter) I spend quite a bit of time in gun stores. While many of my conversations there center around how rapidly the assault weapons are selling "since Obama is going to take away our guns" but recently a new line of conversation has come up as more and more lead-free and "lead-safe" (less, better secured lead, but still lead) ammunition has been hitting the markets. Surprisingly, the conversation has been about how awesome the new products are! Better mushrooming, better weight retention (except in those specially made to fragment which are also better at that), better accuracy, better penetration, basically better bullets, and this is coming from the die hards!
posted by The 10th Regiment of Foot at 7:32 AM on January 3, 2013 [11 favorites]


Hmm, and what about the lead drinking water pipes of centuries past, I wonder?

People have been trying to make a case for the fall of Rome due to lead pipes since the first studies linking adult brain functions and lead contamination came out. The problem doesn't lend itself to a simple analysis...
posted by Tell Me No Lies at 7:35 AM on January 3, 2013


Long since time hunting rounds went to lead-free nationwide.

This isn't true. Steel shot is only required for waterfowl hunting. Some farmers will require it on their land, and certain public hunting grounds will require it when hunting other game, however.

Beyond that, there is no restriction on lead shot.
posted by Pogo_Fuzzybutt at 7:38 AM on January 3, 2013


American political culture

Also being led by a lead-exposed generation at the moment...
posted by stevis23 at 7:40 AM on January 3, 2013 [14 favorites]


The next step is to make Seasoned Shot mandatory.

Of course then it would only be a matter of time until cumin pollution reached catastrophic levels.
posted by Tell Me No Lies at 7:41 AM on January 3, 2013 [2 favorites]


Leaded gasoline has an interesting history.

That lead is harmful in large amounts has been known since the time of Hippocrates, that it is harmful in small amounts has been the subject of study since it was first proposed as a fuel additive in the 1920s.

And recall that lead was removed as an additive to gasoline in the mid 1970s, not because it was perceived that lead was harmful to humans, but that lead was harmful to the platinum in the newly introduced catalytic converters which were needed to reduce other forms of tailpipe emissions.

If the battle against smog resulted in less lead in the environment which has lowered the crime rate, well then those hippies must have been on to something.
posted by three blind mice at 7:43 AM on January 3, 2013 [8 favorites]


Long since time hunting rounds went to lead-free nationwide.

This isn't true.


No, it's not. I think that's why Slap*Happy was advocating for it.
posted by The 10th Regiment of Foot at 7:44 AM on January 3, 2013 [2 favorites]


Pogo_Fuzzybutt, I think you misread an idiom.
posted by Tell Me No Lies at 7:45 AM on January 3, 2013


Oh Thomas Midgely, who won't you mass murder?

Seriously - he's already accredited as being the single organism responsible for the most damage to the atmosphere, between the leaded gas and the CFCs... and his toll just keeps climbing.
posted by FatherDagon at 7:46 AM on January 3, 2013 [2 favorites]


Beyond that, there is no restriction on lead shot.

Yes, I was calling for such a restriction to be implemented nationwide, and also restrictions on lead fishing weights and lures that use lead components. (Don't get me started on plastic baits and fishing line. No-one uses them for more than a season anyways, so why on earth not switch to biodegradable options?)
posted by Slap*Happy at 7:53 AM on January 3, 2013


If it weren't for Thomas Midgely though, I'd never get my goddamn lawnmower started the first time it's needed after a long winter!
posted by The 10th Regiment of Foot at 7:57 AM on January 3, 2013


In fairness to Thomas Midgley, he did manage to prove beyond a shadow of a doubt the existence of karma:

In 1940, at the age of 51, Midgley contracted poliomyelitis, which left him severely disabled. This led him to devise an elaborate system of strings and pulleys to help others lift him from bed. This system was the eventual cause of his death when he was entangled in the ropes of this device and died of strangulation at the age of 55.
posted by Etrigan at 8:03 AM on January 3, 2013 [4 favorites]


Long since time hunting rounds went to lead-free nationwide

I use tungsten matrix shot in areas where there might be woodcutting and steel shot in other areas. The steel is tougher on barrels but I have a tribore barrel that I ordered from Italy which is specifically designed for steel. (you can't use steel in an area which might be cut because it's much harder than other kinds of shot but I prefer it because unlike tungsten matrix and various other alloys it's definitely 100% nontoxic rather than just much better than lead).

The steel shot sold here in the UK is not as good (due to current regulations on maximum allowable pressure) as some higher velocity stuff sold in the US so it's inferior to the tungsten matrix shot in terms of balistics but that will probably change as the CIP is creating a new category of regulated cartridge which will allow higher performance shot to be used in guns designed for it.

Tungsten matrix is really much better ballistically and in terms of damage to old guns not designed for harder metals but it is very expensive. It also isn't really non-toxic in the way that steel is.

For rifle rounds the choice is much easier because you can get lead free bullets that have superior ballistic characteristics. I use solid copper bullets and am quite happy with them. They're obviously more expensive but I don't think that's as much of an issue with rifle rounds as it with shotgun cartridges.
posted by atrazine at 8:04 AM on January 3, 2013 [1 favorite]


In 1937 (when the graph starts) lines of communication were pretty much anywhere there were people to anywhere there were people, so if we're taking Karl Smith's model seriously, communication could still be the case. In reality, he's observed some patterns (his observations make a certain amount of sense) but he is assuming that all things that propagate in the same manner have the same cause, or, at least, the same type of cause. In fact, there's a lovely equation that describes both the spread of tumors and of wild fire, (hint 2.71828 makes an appearance) but if you postulate that wild fire and cancer have the same cause, I'd advise you to not start working on your Nobel Prize acceptance speech just yet.

The real correlation / causation issue with this is the lack of dwell time in the onset. It looks more to me like crime causes leaded gasoline, at least from 1937-1970. I would argue that other factors that appeared everywhere all at once and follow the growth of 20th century urbanization would show a very similar correlation. Anyone up for graphing assembly line jobs?
posted by Kid Charlemagne at 8:05 AM on January 3, 2013 [1 favorite]


Although it does seem reasonable that unleaded gasoline might be a bigger culprit if lead is to be blamed, lead paint has been a huge problem, especially in older cities like those in New England, for quite awhile.

When I was living in Rhode Island a few years back, I had the privilege to sit in on the lower court trial in this case, in which a unanimous jury eventually found three large paint manufacturers liable for knowing the dangerous effects of lead paint and continuing to sell it, causing major public health issues in cities across the state. The verdict was unfortunately overturned, largely because of its derivation under public nuisance law and causation issues in my unsophisticated understanding, but was for the briefest of moments a very exciting thing to have happen, and a real thrill to watch. One of the four companies did settle the case out of court before the trial, though, so the state was able to secure some money (not enough) for lead remediation work.

Despite the unlikely potential for suing companies about this in court, however, I will say that watching the trial firmly convinced me that those companies were aware of the public health catastrophe in their products, and chose not to address them. There really is, and has been for decades, a tremendous amount of evidence that the harm of exposure to lead by young children or pregnant women is severe and was, in large part, preventable.
posted by likeatoaster at 8:11 AM on January 3, 2013 [1 favorite]


Kid Charlemagne, which particular graph are you referring to? I am looking at the ones in Nevin's paper and there are a number of them, some start in 1941, some in 1876, but I don't see anything starting in 1937.

The state-specific correlations mentioned in the Mother Jones link and published in this NBER paper would seem to address some of your criticisms; it seems hard to see how states' specific lead-removal timelines could be correlated with any other broader metrics of urbanization that might be causing the reduction in crime.
posted by enn at 8:14 AM on January 3, 2013 [2 favorites]


It looks more to me like crime causes leaded gasoline, at least from 1937-1970.

You may note that the crime data is from 20 years after the lead data, so the crime would have to retroactively cause leaded gas.
posted by The 10th Regiment of Foot at 8:19 AM on January 3, 2013 [5 favorites]


In fact, there's a lovely equation that describes both the spread of tumors and of wild fire, (hint 2.71828 makes an appearance) but if you postulate that wild fire and cancer have the same cause, I'd advise you to not start working on your Nobel Prize acceptance speech just yet.

I don't think you've understood the quote about "causes". They are not literally The Cause. It's the nature of the cause. Is it something that has to travel from place to place? If so, is that physical or not? If it doesn't have to travel, is it pervasive? Etc.

In that sense, wild fire and cancer DO have the same cause. "It starts $SOMEHOW and spreads by a little piece of it getting into an unaffected area and managing to keep going."
posted by DU at 8:19 AM on January 3, 2013 [7 favorites]


What I really like about this story is all the right wing politicians who wanted to "get tough on crime" would of course never have agreed to any environmental regulations.
posted by DU at 8:23 AM on January 3, 2013 [5 favorites]


"New research finds Pb is the hidden villain behind violent crime"

There's always an evil mod, isn't there?
posted by mr_crash_davis at 8:33 AM on January 3, 2013 [16 favorites]


Hmm, and what about the lead drinking water pipes of centuries past, I wonder?

Apparently cold water pipes aren't such a problem in most areas as mineral deposits quickly form and prevent the lead from contacting the water.

I was just wondering this morning about whether the fall of the Roman Empire had anything to do with its plumbing system. If lead can have far-reaching effects, what about fructose?

Well fructose is not a potent neurotoxin, so quite limited I imagine.
posted by atrazine at 8:37 AM on January 3, 2013 [5 favorites]


What I really like about this story is all the right wing politicians who wanted to "get tough on crime" would of course never have agreed to any environmental regulations.

That's the thing about the lead-free ammo story as well, at first there was resistance to the regulatory mandates in places like California, it was those damn eco-nuts imposing their socialism on something that was simply impossible to replace. Then Obama came along and all the wing-nuts saw it as a conspiracy to further restrict gun ownership. But now that the actual regulatory writing has been on the wall and the industry can see that restrictions (or law suits) are coming some day in the near future, they've actually done research and development and come up with products superior to their lead conterparts that they can market for a premium. Ignoring the initial impetus being impending regulation - its as if one of those libertarian dreams of market forces driving change has actually come to fruition!
posted by The 10th Regiment of Foot at 8:40 AM on January 3, 2013


Sounds interesting and somewhat plausible, though an immediate confounding variable that jumps out at me is the-very-depths-of-desperation poverty associated with living in places that simply aren't fit to live in... and non-remediated houses would fit that bill.
posted by Slackermagee at 8:54 AM on January 3, 2013 [1 favorite]


Led Zeppelin I - released in 1969.

Coincidence?
posted by gottabefunky at 9:05 AM on January 3, 2013 [1 favorite]


Down with molecules!
posted by Mister_A at 9:12 AM on January 3, 2013


Also, I've seen lead fingered as the culprit in mental illness and societal decline in the Roman Empire - it was used to line vessels used in wine production I believe.
posted by Mister_A at 9:15 AM on January 3, 2013


MOLECULAR GASTRONOMY IS KILLING OUR CHILDREN!!!





Very cool post, BTW
posted by It's Raining Florence Henderson at 9:16 AM on January 3, 2013 [2 favorites]


Interesting fact: leaded gasoline is still used in the aviation industry. There's a lot of hand wringing over what to do about Avgas 100LL which is used by a lot of smaller airplane engines and older aircraft. Changing to an unleaded fuel mix is largely impossible given the requirements of changing out the engine, but new aircraft that use unleaded fuels cost incredible amounts of money.
posted by fremen at 9:47 AM on January 3, 2013


I first ran across this argument a few years ago and dismissed it as moonbat conspiracy theorizing (i.e., "The nation that controls tungsten will control the universe!"), as it was just too damn tidy and simple. After all, it's not as though there aren't a hundred theories as to why crime dropped starting in the early nineties: the impact of Roe v. Wade being felt, the two-decade decline in price of cocaine and cocaine-derived products lessening the violence of the drug trade, the "broken window" theory, etc. Lead poisoning? Seems...dubious.

Reading this post reminds me of my grandmother's saying in such circumstances: "Well, I'll be a suck-egg mule." (I have no idea what that literally means, as I don't believe mules suck on eggs terribly often, but the idiomatic meaning is clear.)

God, I love science.
posted by Harvey Jerkwater at 9:50 AM on January 3, 2013


This is an interesting theory, but...

But there's a problem common to all of these theories: It's hard to tease out actual proof.

The important thing to remember about retroactive analysis is this: anybody can tie past events together into a plausible narrative (in fact, the human brain is optimized to see patterns where none exist), and there's no proof that the people doing this retroactive analysis have any talent whatsoever other than the ability to spin convincing bullshit. What is far harder is predictive analysis - the ability to predict what will happen based on an existing hypothesis. Making predictions also involves a certain degree of risk (ie, being proven wrong and mocked) which indicates how much confidence the speaker has in their own analysis. Ultimately, the ability to make predictions is what separates science from navelgazing - science allows us to predict the outcome of situations based on extrapolated data from other observable events.

Giuliani had the balls to risk his reputation by making a predictive analysis based on his "broken-window theory," taking action on this prediction, and achieving results that were almost exactly what he anticipated. I'd say that gives him far more credence than this conspiracy-theorist researcher, who stakes nothing in his analysis and makes no predictions that can be used to verify or disprove his credibility.

Obviously lead isn't good for us, but seriously - a "crime molecule"? Come on.
posted by wolfdreams01 at 10:01 AM on January 3, 2013 [1 favorite]


In fact, there's a lovely equation that describes both the spread of tumors and of wild fire, (hint 2.71828 makes an appearance) but if you postulate that wild fire and cancer have the same cause, I'd advise you to not start working on your Nobel Prize acceptance speech just yet.

I'm not quite sure if you're serious... but if you are...

Correlation does *not* mean "having behavior represented by the same equation". By that argument, since all falling bodies on the Earth follow the same equations of motion, they are all correlated.

I haven't read all the links yet - there are a lot of them - but to dismiss them because "they just have the same equation" is innumerate.

As for the correlation/causation chestnut... correlation doesn't prove causation, but it sure as heck is evidence for causation - moreover, it's a necessary condition, because lack of correlation does in fact prove lack of causation.

People have been speculating about lead piping and the fall of the Roman Empire for decades now. So far there is no proof (and there could easily be proof - for example, if we found tombs of famous individuals and discovered there was a psychotropic level of lead in their remains) but it's hard not to read of the behavior of people during the collapse and ask, "There was something deeply wrong with these people, what caused it?"
posted by lupus_yonderboy at 10:03 AM on January 3, 2013 [2 favorites]


> Giuliani had the balls to risk his reputation by making a predictive analysis based on his "broken-window theory," taking action on this prediction, and achieving results that were almost exactly what he predicted.

Actually, not at all. Did you read the article? His techniques got pretty well exactly the same results as other people using completely different techniques. That seems to be fairly strong proof that his theories were completely meaningless and his techniques ineffective, and that some other cause was responsible

> Obviously lead isn't good for us, but seriously - a "crime molecule"? Come on.

What sort of "argument" is that?! It's well-known that exposure to lead leads to cognitive impairment - the article has numerous seemingly well-crafted studies showing strong correlations - and your counter argument is "Come on." - i.e. this is so ridiculous on its face that no refutation is necessary!
posted by lupus_yonderboy at 10:11 AM on January 3, 2013 [16 favorites]


The important thing to remember about retroactive analysis is this: anybody can tie past events together into a plausible narrative (in fact, the human brain is optimized to see patterns where none exist), and there's no proof that the people doing this retroactive analysis have any talent whatsoever other than the ability to spin convincing bullshit.

I guess that's one way to think of cosmology, paleontology, and geology sure. This is not just a correlation. There is a difference between something which is just correlated and a correlation in the presence of a plausible mechanism of causation.

I'd say that gives him far more credence than this conspiracy-theorist researcher, who stakes nothing in his analysis and makes no predictions that can be used to verify or disprove his credibility.

Sure he does. We should expect similar falls in crime when lead fuel additives are phased out in the countries where they are still used.

People have been speculating about lead piping and the fall of the Roman Empire for decades now.

Romans used lead acetate (lead sugar) as a sweetener, I don't think it was the pipes that were the problem.
posted by atrazine at 10:16 AM on January 3, 2013 [8 favorites]


You may note that the crime data is from 20 years after the lead data, so the crime would have to retroactively cause leaded gas.

Damnit, the sliding scale threw me.
posted by Kid Charlemagne at 10:19 AM on January 3, 2013


Correlation does *not* mean "having behavior represented by the same equation". By that argument, since all falling bodies on the Earth follow the same equations of motion, they are all correlated.

I understand that. I'm talking about this part: " If it spreads along lines of communication, he says, the cause is information. Think Bieber Fever. If it travels along major transportation routes, the cause is microbial. Think influenza. If it spreads out like a fan, the cause is an insect. Think malaria. But if it's everywhere, all at once—as both the rise of crime in the '60s and '70s and the fall of crime in the '90s seemed to be—the cause is a molecule." Rabies would probably also have a fan like spread, for example, but is not caused by insects.

And given that all falling bodies fall due to the same fundamental cause, yes, their movement should be highly correlated.
posted by Kid Charlemagne at 10:24 AM on January 3, 2013


Predictive analysis was in fact done early in the era of leaded gasoline, in the form of concerns voiced about the health consequences as the fascinating timeline linked by Three Blind Mice shows. Sure those predictions weren't precise modern studies, and they didn't predict the specific contours of the societal effect from pervasive cognitive and neurological degradation and call it crime, and America didn't become great by letting a bunch of whiners get in the way of progress, but still.
posted by maniabug at 10:25 AM on January 3, 2013 [1 favorite]


Have some professional criminologists commented on these findings? I've read that one issue with crime stats is that they may not be comparable over a long time period. Definitions of various types of crime change, as well as the social acceptance or reject of certain crimes, the police response and the medical response (which helps decreasing the homicide rate by saving lives). For instance, some crimes that used to be underreported (i.e. rape) are now reported much more thoroughly.
posted by elgilito at 10:26 AM on January 3, 2013


What sort of "argument" is that?! It's well-known that exposure to lead leads to cognitive impairment - the article has numerous seemingly well-crafted studies showing strong correlations - and your counter argument is "Come on." - i.e. this is so ridiculous on its face that no refutation is necessary!

I see. So it's far easier for you to believe in the idea of a molecule that causes "criminal behavior" 29 years later (because apparently all people's physiology is the same, and 29 years is the magic time period when the molecule kicks in and goes into criminal overdrive regardless of other factors) as being responsible for a widespread reduction in crime over time, rather than the far more simple explanation that people share knowledge with each other, and so advances in police work which are proven successful in one area tend to get implemented in other areas as well.

Yes, "Come on" is basically the entirety of my refutation. When the other side advances an argument that completely ignores the principle of Occam's Razor, "Come on" is really all one should need to say.
posted by wolfdreams01 at 10:37 AM on January 3, 2013


Little known fact: Occam's Razor was actually made of lead. Shaving too close with Occam's Razor has a strong correlation with insanity.
posted by It's Raining Florence Henderson at 10:41 AM on January 3, 2013 [15 favorites]


So it's far easier for you to believe in the idea of a molecule that causes "criminal behavior" 29 years later (because apparently all people's physiology is the same, and 29 years is the magic time period when the molecule kicks in and goes into criminal overdrive regardless of other factors)

Logic (or Occam's Razor), or whatever you want to your way of constructing arguments, doesn't mean much if you can't tell the difference between societal trends and arguments predicated on precise timelines and precise physiological reactions. In other words, you don't seem to understand enough about how this argument was formulated to refute it.

Come on.
posted by OmieWise at 10:46 AM on January 3, 2013 [7 favorites]


You know where you can get massive amounts of data on brain damage and learning deficits in several distinct categories, and on crimes and misbehavior, along with data on lead exposure and many other possible causes and correlates for cognitive deficits and criminal behavior?

Schoolchildren.

All that data is collected. Illinois alone has 2 million school children this year, across a wide range of urban, suburban, and rural environments, in a wide variety of different housing stock of different ages; in 2009, 133,000 children in Illinois under the age of 6 were tested for lead exposure, a little over 1% of whom had clinically significant lead poisoning.

You can track that data by county, by city, by school district, by special ed program, over 20 or more years. You may not be able to show that any individual student has behavioral problems or cognitive deficits is caused by lead, but you can certainly look at an excellent educational data set of 2 million children in a huge variety of circumstances and the large quantity of public health data collected on those students, and draw some very good conclusions about environmental toxins and their impact on learning and behavior. And every state collects similar data.
posted by Eyebrows McGee at 10:51 AM on January 3, 2013 [3 favorites]


... the far more simple explanation that people share knowledge with each other, and so advances in police work which are proven successful in one area tend to get implemented in other areas as well.

This explanation makes absolutely no sense. Criminals are people too and can, and do, share knowledge. So this would predict a completely flat arms-race scenario, which is not what the data show.

Unless you want to posit an additional entity that can explain why crime got ahead for a while and then law took over. "Lead" is one candidate for that additional entity.
posted by DU at 10:52 AM on January 3, 2013 [1 favorite]


Like a guy in a $30,000 suit is going to sit down and analyze reams of crime data. Come on!
posted by dirigibleman at 10:55 AM on January 3, 2013 [10 favorites]


Rabies would probably also have a fan like spread, for example, but is not caused by insects.

In addition to the refutation I already made and you either ignored or didn't understand, there's this: The quote in the FPP is not intended to be a Law of Nature. It's a rule of thumb. "When you have characteristics X, Y and Z, look for a cause like A."
posted by DU at 10:57 AM on January 3, 2013 [1 favorite]


Yes, "Come on" is basically the entirety of my refutation. When the other side advances an argument that completely ignores the principle of Occam's Razor, "Come on" is really all one should need to say.

Occam's razor is a tool for choosing between two arguments with equal evidence. In order to reach that point you have to in fact, LOOK at the evidence presented by both arguments. Even then, sometimes the simpler argument is wrong. Occam's razor isn't built into the structure of reality.
posted by Gygesringtone at 11:06 AM on January 3, 2013 [4 favorites]


Serious question for those in the "this is just arbitrary, cherry-picked correlation and therefore not worthy of serious consideration" camp. Is every correlative study just bullshit and spin? If not, what are some of the characteristics of correlative studies that should be taken seriously?
posted by treepour at 11:18 AM on January 3, 2013 [1 favorite]


Nobody really believes the "correlation isn't causation" line in any serious, consistent way; it is purely a rhetorical device against studies whose conclusions you don't like.

Actual serious argumentation against such a study would consist of specific and well-grounded methodological complaints, showing that it doesn't really show what it thinks it does, or by offering additional observable implications of the theory and showing empirically that those implications don't hold. Wolfdreams has actually taken a step down this road by arguing that it was more effective police work, though he hasn't offered any potentially observable things that would be true if it were police work but false if lead played a role, or vice versa.
posted by ROU_Xenophobe at 11:29 AM on January 3, 2013 [8 favorites]


It is a great post, and an interesting theory. But what should we then make of the indications that some rural and suburban areas have a growing violent crime rate compared to cities now, like in this UK example? It's mentioned in here (pdf), as well, though this is an older article.
Maybe urbanites are naturally less criminal than country folks, and need some lead poisoning to get going? (joking, joking)
I would like to see this compared to other countries - here, I drove a car several times across Europe using leaded gasoline in the 90's, so it must have been widely available. As I remember it, a lot of people all over Europe drove old cars, specially after -89, where there was a literal explosion of old-car ownership, because of the opening up of the east block. But crime-rates were already falling in urban areas in some countries, while they were still rising in other places, and I don't see a 20-year lead-driven surge right now.
And why is the crime rate in Japan much lower than in the US? Concentrations of lead must have been crazy in Japan before unleaded gas? In Japan, even the suburbs are dense and intense. Are the crime rates lower in Japan today than in the -80's? Because of course it will be relative to country/culture.

As I remember it, everyone with resources moved out of the cities during the -60's and -70's. We moved out and moved back in when my step-dad went broke. The only people living in the city were broke like us. Isn't it a more obvious correlation - that children from broke and broken families commit more crime than children from well-off and well-functioning families? And they also get more bad food and more lead, and more bad environmental effects all over. And worse educations, and less job-opportunities.
When affluent people began moving back, they brought with them jobs, teachers, grocery stores, demands for healthier streets and better parks.
posted by mumimor at 11:29 AM on January 3, 2013


But what should we then make of the indications that some rural and suburban areas have a growing violent crime rate compared to cities now, like in this UK example? ... And why is the crime rate in Japan much lower than in the US? Concentrations of lead must have been crazy in Japan before unleaded gas?

No one's saying that lead is the only thing that will cause a change in crime rates.
posted by Etrigan at 11:36 AM on January 3, 2013 [1 favorite]


wolfdreams01: “Giuliani had the balls to risk his reputation by making a predictive analysis based on his ‘broken-window theory,’ taking action on this prediction, and achieving results that were almost exactly what he anticipated. I'd say that gives him far more credence than this conspiracy-theorist researcher, who stakes nothing in his analysis and makes no predictions that can be used to verify or disprove his credibility.”

That isn't precisely true. Giuliani's claim that he would be able to decrease crime was actually a lot more like Romney's claim during the campaign that he would be able to create "at least 10 million" new jobs if elected – it was based on the fact that economic projections indicated that 10 million new jobs would be created no matter what happened, so it was perfectly safe to make that prediction.

Crime had already dropped twelve percent between 1990 and 1994. Who would bet that crime wouldn't go down further? And can we really give Giuliani credit for something that predated his term in office?
posted by koeselitz at 11:36 AM on January 3, 2013 [5 favorites]


9/11 changed everything. Even the past.
posted by It's Raining Florence Henderson at 11:42 AM on January 3, 2013 [7 favorites]


lupus_yonderboy: "As for the correlation/causation chestnut... correlation doesn't prove causation, but it sure as heck is evidence for causation - moreover, it's a necessary condition, because lack of correlation does in fact prove lack of causation."

It's worse than that. Sometimes a cause can have two equal-but-opposite effects.

Take fossil fuels for instance: Burning jet fuel causes global warming by releasing greenhouse emissions into the atmosphere. However, burning jet fuel also releases particulate exhaust and vapor contrails that cause a phenomenon known as global dimming, which actually lowers temperatures on the ground.

If these two phenomena were balanced (they're not), we'd actually have no correlation between fuel consumption and global warming.

tl;dr: Science is hard.
posted by schmod at 11:42 AM on January 3, 2013 [3 favorites]


Even then, sometimes the simpler argument is wrong. Occam's razor isn't built into the structure of reality.

Epicycles and steady-state theory spring immediately to mind. Only evidence and its careful analysis can rule out a theory or bear out a hypothesis - "Come on!" doesn't cut it.
posted by Slap*Happy at 11:43 AM on January 3, 2013


wolfdreams01: “The important thing to remember about retroactive analysis is this: anybody can tie past events together into a plausible narrative (in fact, the human brain is optimized to see patterns where none exist), and there's no proof that the people doing this retroactive analysis have any talent whatsoever other than the ability to spin convincing bullshit. What is far harder is predictive analysis - the ability to predict what will happen based on an existing hypothesis. Making predictions also involves a certain degree of risk (ie, being proven wrong and mocked) which indicates how much confidence the speaker has in their own analysis. Ultimately, the ability to make predictions is what separates science from navelgazing - science allows us to predict the outcome of situations based on extrapolated data from other observable events.”

It seems worth pointing out that you are completely ignoring the second page of this essay, which begins by tossing out correlative econometric studies and investigating whether medically a link between criminal behaviors and lead poisoning can be demonstrated:
IF ECONOMETRIC STUDIES WERE ALL THERE were to the story of lead, you'd be justified in remaining skeptical no matter how good the statistics look. Even when researchers do their best—controlling for economic growth, welfare payments, race, income, education level, and everything else they can think of—it's always possible that something they haven't thought of is still lurking in the background. But there's another reason to take the lead hypothesis seriously, and it might be the most compelling one of all: Neurological research is demonstrating that lead's effects are even more appalling, more permanent, and appear at far lower levels than we ever thought.
You talk about "predictive" science, but I would argue there's no such thing where sociology is concerned. Even if there is, it's incredibly vague compared to the precision that medical science demands. And medical science indicates that this theory is true – that lead poisoning has historically been responsible for the major crime waves in the past century.
posted by koeselitz at 11:50 AM on January 3, 2013 [4 favorites]


You talk about "predictive" science, but I would argue there's no such thing where sociology is concerned.

Mr Seldon called. You go to the First Foundation, not the Second.
posted by DU at 11:55 AM on January 3, 2013 [6 favorites]


No one's saying that lead is the only thing that will cause a change in crime rates.
Well, that is basically what Karl Smith is saying
posted by mumimor at 12:00 PM on January 3, 2013 [1 favorite]


Nobody really believes the "correlation isn't causation" line in any serious, consistent way; it is purely a rhetorical device against studies whose conclusions you don't like.

Doesn't The Journal of Chemical Information and Modeling count? Statistical inference is good for a lot of things, but it is no substitute for a controlled experiment.

If somebody granulated this by city, rather than by state, so that a Yreka California and Galesburg Illinois weren't being lumped in with Chicago and Los Angels and a clear continuum was shown for various towns at various sizes with various levels of lead exposure it would be a lot more convincing, but the fundamental problem remains that there are going to be a lot of other factors that probably correlate well with the amount of leaded gasoline consumed in a given area and, ergo, will correlate about as well as lead gasoline exposure and crime.
posted by Kid Charlemagne at 12:00 PM on January 3, 2013


No one's saying that lead is the only thing that will cause a change in crime rates.
Well, that is basically what Karl Smith is saying


No. At best, Karl Smith is saying lead is the only thing that caused *these* changes in crime rates.

But I doubt he's really saying that either. He's speaking off-hand (as his "rule of thumb" thing, which has been a stumbling block to others in this thread shows). And when 90% of your variation can be explained by a single cause, then I think it's fair to call that "the only thing" vs an Apu-like "the causes of the Civil War were many and various" explanation.
posted by DU at 12:05 PM on January 3, 2013


OK, let's clarify a few things.

To begin with, you are shifting my goalposts. You seem to think that I am saying lead has nothing to do with crime. That's not what I'm saying at all. Lead causes brain damage and I am sure that is a factor, much as there are many other factors. The idea that lead would play zero role is not what I am saying at all. My objection is that the article is written in a way that suggests that lead was the primary cause of the crime drop rather than Giuliani's broken windows theory. That is the nonsensical assertion that I would like to refute.

First of all, broken windows theory has been empirically tested in psychological studies, so there is actual data to corroborate it. Any worthwhile article that tries to disprove the utility of this theory should at least try to explain why those studies are incorrect.

Second of all, only gullible people believe data provided in an article that obviously has an agenda (in this case, proving the lead theory). Intelligent and skeptical people review the actual data on their own rather than letting somebody else summarize it for them. So, let's try being intelligent skeptical people and actually do that, shall we? I'll give you a few moments to do the calculations. (Spoiler: results may surprise you!)

Now that you're done, we can all easily see the two areas where this article grossly distorts the truth. First of all, the claim that the violent crime rate (murder, forcible rape, robbery, aggravated assault) started dropping in 1990 (and so Giuliani had nothing to do with it) is technically true, but also jaw-droppingly misleading. Based on the numbers, during Mayor Dinkin's four-year term (1990-1993) the violent crime rate dropped by 8.05%. By contrast, during Giuliani's first four years in office, the violent crime rate dropped by over 36%. In statistical terms, that's a huge difference. To attempt to gloss over such a significant difference in the numbers is the height of irresponsible analysis.

The second area where this article distorts the truth is that it suggests that crime rate went down relatively evenly across all cities in America. In fact, that is incorrect as well. During Giuliani's entire time in office violent crime fell 56% in NYC but only 33% nationally - again, an extremely significant statistical difference. When the researcher who wrote this article overlooks and minimizes such vast reams of data that fly squarely in the face of his hypothesis, it's hard not to think that he has an ulterior motive.

Finally, on a personal note, it would be nice in the future if you didn't accuse me of failing to do my research. Or at least next time you do so, maybe first try doing some independent analysis with a skeptical eye, instead of just believing whatever the article tells you.
posted by wolfdreams01 at 12:07 PM on January 3, 2013 [1 favorite]


No one's saying that lead is the only thing that will cause a change in crime rates.
Well, that is basically what Karl Smith is saying


There's a difference even between "This was the only thing that caused that" and "This is the only thing that will cause that."

For instance, if I throw my phone out a fifth-story window, it will likely break. That was the only thing that caused my phone to break. However, that does not mean that you could jump up and down on my phone and not break it.
posted by Etrigan at 12:09 PM on January 3, 2013 [2 favorites]


Intelligent and skeptical people review the actual data...

There's no column for lead in that data.

During Giuliani's entire time in office violent crime fell 56% in NYC but only 33% nationally - again, an extremely significant statistical difference.

Which was addressed in more than one of these links. Did you click them?
posted by DU at 12:11 PM on January 3, 2013 [3 favorites]


Finally, on a personal note, it would be nice in the future if you didn't accuse me of failing to do my research.

People accused you of not offering an argument and of not understanding what you were dismissively attempting to refute. The fact that you did a couple of Google searches after the fact does not change that.
posted by OmieWise at 12:27 PM on January 3, 2013 [2 favorites]


wolfdreams01: “The second area where this article distorts the truth is that it suggests that crime rate went down relatively evenly across all cities in America.”

This is the opposite of the truth. From the article:
If childhood lead exposure really did produce criminal behavior in adults, you'd expect that in states where consumption of leaded gasoline declined slowly, crime would decline slowly too. Conversely, in states where it declined quickly, crime would decline quickly. And that's exactly what she found.
In other words, there are state-by-state differences in the drop in crime rate, and the map perfectly to the state-by-state differences in the decline of leaded gasoline.

Again, I appreciate your skepticism. The thing is that this theory works out very well when controlling for all kinds of factors, not just on a national level but on a state-by-state level. That's a stunning level of correlation, one that we can't find with any other explanation of criminal activity. Seriously. It deserves some consideration. At least look over the paper that the article is discussing.

I get the feeling your reaction is based on your sense that "Broken Windows Theory" is infallible. It does indeed have a body of evidence behind it, but that body of evidence is not uncontroversial. I agree that it would be nice if this article spent more time deconstructing Broken Windows, but I can also see that that might be a waste; you don't have to disprove Broken Windows Theory entirely to admit that perhaps there are some other factors that need to be considered, maybe even huge factors. And I didn't even read this article as seeking to demolish Broken Windows; at most, it was a tad dismissive of its application in New York. And it is hardly the first article to do that. There's been a backlash against Broken Windows for years now.
posted by koeselitz at 12:31 PM on January 3, 2013 [7 favorites]


I guess this means I should stop licking lead bars.
posted by Damienmce at 12:31 PM on January 3, 2013


wolfdreams01: “Finally, on a personal note, it would be nice in the future if you didn't accuse me of failing to do my research. Or at least next time you do so, maybe first try doing some independent analysis with a skeptical eye, instead of just believing whatever the article tells you.”

I appreciate this point of view, and it makes a lot of sense. I studied political science in graduate school, and while I avoided sociology assiduously, I certainly became aware of just how incredibly difficult it is to find really consistent correlations that hold across the United States for a number of decades. That makes so much of criminal sociology a difficult game of sorting through the numbers in the right way.

But that's precisely why I was so impressed with this study and the findings it draws from the numbers. Really, it's incredibly striking to see that kind of correlation. Most of the time we have to dismiss nationwide statistics as too subject to mitigating factors simply because there's no way to make the numbers make sense. This time, however, they hold consistently. I really think this is worth considering.
posted by koeselitz at 12:37 PM on January 3, 2013


Romans used lead acetate (lead sugar) as a sweetener

Lead acetate was used to sweeten wines well into the 1800s.
posted by hattifattener at 12:37 PM on January 3, 2013


but seriously - a "crime molecule"? Come on.

Have you ever heard of alcohol? Nine atoms. Significant cause of violent crime. I think you even tried to ban it once.
posted by alasdair at 12:50 PM on January 3, 2013 [4 favorites]


and the map perfectly to the state-by-state differences in the decline of leaded gasoline.

Perfectly? If I somehow missed a graph with an n of 50 and an r2 of 1.0000, I'm going to cease being skeptical and assume outright fraud because that just doesn't happen in real life.
posted by Kid Charlemagne at 1:00 PM on January 3, 2013 [2 favorites]


If not, what are some of the characteristics of correlative studies that should be taken seriously?

The elasticity of violent crime with respect to childhood lead exposure is estimated to be approximately 0.8. This implies that, between 1992 and 2002, the phase-out of lead from gasoline was responsible for approximately a 56% decline in violent crime.

Well for starters, don't make claims like this while failing to account for demographics, drug-use, both legal repercussions and law enforcement, population density, poverty and wealth migration, and perhaps most importantly, a lack of granularity finer than state level.

In any case, the onus isn't on the skeptic to provide evidence, or set any sort of predetermined threshold for acceptance.
posted by clearly at 1:01 PM on January 3, 2013 [4 favorites]


Wouldn't this have some sort of relevance?
I've worked in troubled neighborhoods for a decade and I still live in one. If there is one thing I really hate, it is conservative "broken window" or similar approaches to changing the lives of the very poor. But what I hate about it/them, is the perception that one can change extremely complex situations through quick fixes. in my view, Remove shop-lifters and cure violent crime!! and Remove lead and cure violent crime!! are equally simplistic, and I refuse to believe that anyone who has an actual real daily life in any inner-city neighborhood (or rural slums) would think up anything so facile.
Yes, incarcerating half the male population has an effect, and yes, eliminating lead poisoning has an effect, but both avoid discussing the root causes, which are poverty and segregation, which always come together with bad education and lack of opportunity.
posted by mumimor at 1:12 PM on January 3, 2013 [2 favorites]


During Giuliani's entire time in office violent crime fell 56% in NYC but only 33% nationally - again, an extremely significant statistical difference.

By your description, you have 1DF, so I wouldn't bet on it.
posted by ROU_Xenophobe at 1:16 PM on January 3, 2013 [1 favorite]


Maybe the process that caused us to reconsider leaded fuel also helped cause a drop in crime.
posted by deo rei at 1:40 PM on January 3, 2013 [1 favorite]


[Folks? Please manage to not turn this into the same old super boring "one person makes the thread all about them" situation. You can email people if you need intense one-on-one conversations. This is MetaFilter, it's inappropriate here.]
posted by jessamyn at 2:03 PM on January 3, 2013


"It looks more to me like crime causes leaded gasoline, at least from 1937-1970."

You may note that the crime data is from 20 years after the lead data, so the crime would have to retroactively cause leaded gas.


That comment might have been a mistake but it's also interesting.

I'd imagine that the lag time (if it's real) isn't constant and uniform. Maybe it could be affected by puberty onset or death rates or some other thing. I dunno much about this stuff but maybe studying the lag could provide further evidence either way.
posted by PJMcPrettypants at 2:11 PM on January 3, 2013


Did you fact-check with alternative sources or are we once again simply believing whatever data the study gives us?

I'll do some digging. I'll even post the extracts, so you don't have to feel bad about not reading the links. I don't have journal access, so someone else will have to do the hard work.

The relationship between lead and crime
This study investigates the association between air-lead levels and crime rates across 2,772 U.S. counties. Data for the analysis come from the Environmental Protection Agency, the Bureau of Census, and the Federal Bureau of Investigation. Results suggest that air-lead levels have a direct effect on property and violent crime rates even after adjusting for general levels of air pollution and several structural covariates of crime. We also find that resource deprivation interacts with air-lead levels. The association between air-lead levels and crime rates-property and violent-is strongest in counties that have high levels of resource deprivation and weakest in counties that have low levels of deprivation. This interaction is consistent with arguments and evidence in the health care literature that populations most at risk of lead poisoning are least likely to get the resources required to prevent, screen, and treat the illness.
Biology and violence: from birth to adulthood
This book presents the most comprehensive study to date of the major biological, psychological and environmental predictors of criminal behavior, particularly violence, through a detailed analysis of nearly 1000 low-income black youths from their birth to early adulthood. By examining over 150 variables spanning the lives of these youths, the study concludes that both biological and environmental factors produce strong, and independent, effects on delinquency and adult crime among males and females, who are distinguished from their controls. Powerful influences on violence include behavioral disorders during youth, low school achievement, parents with a low educational level, an absent father, hyperactivity, lead poisoning, left-handedness and mixed dominance, soft neurological signs, and neurological abnormalities. Case study comparisons between the most violent males and females and their controls show that criminals evidence a higher incidence of lead poisoning, disobedience, head injury, and a history of epileptic seizures among themselves or their immediate family members.
The ATSDR's toxicological profile for lead, specifically the relevance to public health(PDF)
Despite the many factors that can potentially work against finding agreement among studies, the preponderance of the evidence indicates that lead exposure is associated with decrements in cognitive function. Meta-analyses conducted on cross-sectional studies or a combination of cross-sectional and prospective studies suggest that an IQ decline of 1–5 points is associated with an increase in PbB of 10 μg/dL. Most importantly, no threshold for the effects of lead on IQ has been identified. This has been confirmed by a series of recent studies in children that found significant inverse associations between cognitive function and PbBs [less than] 10 μg/dL.
Bone lead levels in adjudicated delinquents. A case control study
RESULTS: Cases had significantly higher mean concentrations of lead in their bones than controls (11.0+/-32.7 vs. 1.5+/-32.1 ppm). This was true for both Whites and African Americans. The unadjusted odds ratio for a lead level > or =25 vs. <2>25 ppm than controls (OR=4.0, 95% CL: 1.4-11.1).

CONCLUSION: Elevated body lead burdens, measured by bone lead concentrations, are associated with elevated risk for adjudicated delinquency.

Association of Prenatal and Childhood Blood Lead Concentrations with Criminal Arrests in Early Adulthood
Prenatal and postnatal blood lead concentrations are associated with higher rates of total arrests and/or arrests for offenses involving violence. This is the first prospective study to demonstrate an association between developmental exposure to lead and adult criminal behavior.
The relationship between lead exposure and homicide

Understanding international crime trends: The legacy of preschool lead exposure(PDF)
This study shows a very strong association between preschool blood lead and subsequent crime rate trends over several decades in the USA, Britain, Canada, France, Australia, Finland, Italy, West Germany, and New Zealand. The relationship is characterized by best-fit lags (highest R2 and t-value for blood lead) consistent with neurobehavioral damage in the first year of life and the peak age of offending for index crime, burglary, and violent crime. The impact of blood lead is also evident in age-specific arrest and incarceration trends. Regression analysis of average 1985–1994 murder rates across USA cities suggests that murder could be especially associated with more severe cases of childhood lead poisoning.
A summary of (mostly Nevin's) research: Research Summary: Childhood Lead Exposure Trends Explain International Property and Violent Crime Trends, and Differences in USA City Murder Rates

And finally: Bibliography on Childhood Lead Poisoning Science And Politics, Environmental Research Foundation

Nevin again: The Answer Is Lead Poisoning

Here's an interesting aside: Iron-Deficiency Anemia and the Cycle of Poverty among Human Immunodeficiency Virus—Infected Women in the Inner City, with some info on lead poisoning in iron-deficient people.
posted by the man of twists and turns at 2:33 PM on January 3, 2013 [13 favorites]


From the second link:

"Lead exposure does not appear to affect the murder rate though"
posted by destro at 2:34 PM on January 3, 2013 [2 favorites]


Hey, as long as we are considering this to be a valid line of scientific inquiry, I would also like to point out that the crime stats also correlate roughly to the time period when people started breakdancing, feathering their hair, and wearing just one glove as a fashion statement. Obviously their cognitive function was impaired, and we must thus take it extremely seriously when researchers suggest that the lead molecule actually causes douchiness.
posted by wolfdreams01 at 2:35 PM on January 3, 2013


From the second link:

Lead exposure does not appear to affect the murder rate though, a result the author finds “not entirely surprising” given that her analysis omitted the effects of gangs and crack and that it is possible that only substantial exposure to lead would lead to an extreme outcome like murder."

A result the author "finds not entirely surprising given that she drew her conclusions well ahead of, you know, looking at the whole picture, and besides, did you see how well the lines match up and stuff?!?"
posted by clearly at 2:53 PM on January 3, 2013 [1 favorite]


The Dunning-Kruger levels in here are unsafe for human consumption.
posted by Eyebrows McGee at 2:55 PM on January 3, 2013 [3 favorites]


Is Dunning-Kruger the new Godwin? Not saying it is or isn't happening here, but it seems like half the time people bring it up they are unwittingly referring to themselves.
posted by Pyry at 2:58 PM on January 3, 2013 [2 favorites]


Most of us go into these discussions in at least a bit over our head with the specific field we're discussing, so there will always be some level of Dunning-Kruger-ism going on.

The real problem comes when non-expert skeptics come in with enough knowledge to think they can debunk any other peer-reviewed study after reading it for five minutes. I get that the peer-review process is not foolproof, but if a large number of expert studies converge on one outcome, you need to bring your own expert sources to the table, or you need to immediately go publish your own findings yourself, because you've proven a bunch of experts wrong.
posted by tonycpsu at 3:18 PM on January 3, 2013 [4 favorites]


The Freddy-Kruger levels in here are also unsafe.
posted by It's Raining Florence Henderson at 3:20 PM on January 3, 2013 [6 favorites]


wolfdreams01: “Hey, as long as we are considering this to be a valid line of scientific inquiry, I would also like to point out that the crime stats also correlate roughly to the time period when people started breakdancing, feathering their hair, and wearing just one glove as a fashion statement. Obviously their cognitive function was impaired, and we must thus take it extremely seriously when researchers suggest that the lead molecule actually causes douchiness.”

When you say that we haven't necessarily fact-checked the studies in question, that's a fair enough point. I retain enough skepticism about all explanations of the crime drop – Broken Windows, the "perfect storm" theory, this lead theory – and I think it makes sense to keep some distance from them. There is certainly no one theory that explains all criminology trends over the past forty years. I have to say that you and I probably disagree in that I am pretty skeptical of the Broken Windows theory, and I think we've seen a lot of that kind of skepticism vindicated over the past decade or so. However, that's really neither here nor there where lead is concerned. As you've said, what matters is whether the data checks out; and until we know that it does, we can't say much.

Thanks for your thoughtful contributions to this discussion, anyway. I probably don't have much more to say, but I've enjoyed this back and forth. And I've become convinced that you're absolutely correct when you say that this needs to be fact-checked before we can really draw conclusions from it.
posted by koeselitz at 3:32 PM on January 3, 2013 [1 favorite]


This is a field where "experts" come from a very wide range of research, and don't at all publish or discuss in the same journals and conferences. Also, in spite of a lot of very interesting research, little of this reaches practice, ever. This is not medicine.
When I worked within this field, I tried to keep up with all the different approaches, but to be honest, a lot of the published and referenced work was pure BS to me. There is a huge difference between those who do university research in this field, mainly based on statistics and literature, and then people who do actual fieldwork. And then again there is a difference between sociologists who spend three weeks in a site and anthropologists or nurses who spend several years in the field.
I know all the arguments against this point of view, so don't throw it at me as if I'm stupid. I can easily see the point of meta-analysis in many cases. But there is something else at stake here, and I've heard about it from other fields as well: as social mobility becomes increasingly rare, scholars, doctors and even teachers and nurses have a lot less experience with the cultures of other social groups. So they have difficulty understanding and discussing the life-choices of for instance, teen-moms. But at least they are out there. A well-off if well-meaning economics professor living in a safe neighborhood has no idea what it means to live in a project, in an urban context. Or in a trailer home, in an area with no school choices and no jobs.
posted by mumimor at 3:33 PM on January 3, 2013 [2 favorites]


The real problem comes...

When simple answers to highly complex problems are accepted without hesitation.
I'm picturing this study being cited to give landlords a subsidy for removing lead paint from walls while budgets for community programs are slashed.

non-expert skeptics

Welcome to the Internet.

large number of expert studies*

large >= 3?

you've proven a bunch of experts wrong.

You're stating that an economist/economic consultant are "experts" on both lead poisoning and violent crime. I for one, would expect both an expert and an economist to not flippantly disregard local socio-economic conditions, demographics, or the broad political and social climate before drawing conclusions.

This isn't about being right or wrong. It's about the process and due diligence that must be undertaken before making an assertion in the name of science or research. Especially one with implications on public policy as it relates to violent crime.
posted by clearly at 4:29 PM on January 3, 2013 [2 favorites]


One reason to be skeptical is that a lot of stuff correlates with each other. Lots of things got worse in the 70s and improved in the 90s, and anyone who has done a decent amount of social science research has seen the curve in Nevin, page 1 in a lot of different places (for instance, as I recall, What's the Matter with Kansas has the big U (in the middle of the 20th century) in various guises throughout the book, with all sorts of things causally attributed to it).

A lot of things go up and down together, and there is good reason to be skeptical, even after reading the man of twists and turns's excellent mini-bibliography. The basic alternative hypothesis is that lead is a very good correlate for how poorly a child was treated by society when growing up. All the studies cited here agree on this: lead correlates with low income housing, with dense urbanism, with poor nutrition, etc (as well as with outcomes like health, mental problems, etc). Nevin, Reyes, and the others do a decent job including various plausible controls, but even so, if lead correlates with unobserved stuff like the cheapness, dilapidation, and density of housing, which in turn correlates with how well their local governments were treating their poor, then that will pick up stuff that isn't measured by the relatively coarse controls like population density, per capita income, etc.* The best studies, such as Reyes (2007), are actually cross-sectional (US states, in this case) and across time, with controls for both years and states, so that the only variance being compared is how individual states change over time in lead and crime. But even then, all we need to believe is that states that tend to fix their lead earlier or later also fix their crime earlier or later (and this isn't unreasonable; looking at her tables, the worst states for lead/crime are almost all in the south, and it's plausible that the timing for fixing both were also correlated).

Even the individual-level studies are plagued by the same problems: maybe the kids with high blood lead also had the worst environments overall, above and beyond what we have in the controls. What we really need is a good natural experiment, where some quirk of public housing, say, caused randomized variation in lead paint levels, and then we go look very closely at that specific cohort of individuals. And maybe that's been done -- I haven't done my homework on this. But until we have that, it doesn't matter whether 1 or 20 studies have been done; if one is suspicious of these results because one fears that lead might just correlate with other factors that damage children -- a common problem in social studies! -- connecting air lead to blood lead to IQ to crime is still not enough. And it's worth being skeptical, since the danger is hugely misled and wasted social policy if we incorrectly conclude that this was largely due to some simple molecule, as Drum advocates.

* For instance, in his USA Central City Murder Rate Regressions (Table 6), Nevin looks at nicely similar units and finds strong results for lead (LP). But, "When a variable is added for black percent of population, city size is still significant and LP% is not (t-value = 1.27), but the LP% coefficient still has the expected sign and retaining LP% in the model increases R2 to 69%." Ie, in this case, lead was mainly functioning as a proxy for black population.
posted by chortly at 7:10 PM on January 3, 2013 [2 favorites]


chortly: “One reason to be skeptical is that a lot of stuff correlates with each other. Lots of things got worse in the 70s and improved in the 90s, and anyone who has done a decent amount of social science research has seen the curve in Nevin, page 1 in a lot of different places (for instance, as I recall, What's the Matter with Kansas has the big U (in the middle of the 20th century) in various guises throughout the book, with all sorts of things causally attributed to it).”

Maybe we should drop the social science angle and actually talk about the medical science that the second part of the article discusses – since the article itself notes this very problem with correlation in social sciences.
posted by koeselitz at 9:03 PM on January 3, 2013


lupus_yonderboy: "As for the correlation/causation chestnut... correlation doesn't prove causation, but it sure as heck is evidence for causation - moreover, it's a necessary condition, because lack of correlation does in fact prove lack of causation."

It's worse than that. Sometimes a cause can have two equal-but-opposite effects.
Another common pitfall is Simpson's paradox where when you calculate correlation between two variables you get one one direction, but once you split up based on a third variable, the correlations switch sign.

But there is great hope for resolutions of these tricky problems, as our mathematical tools improve. Over the past two decades, Judea Pearl and others have been doing absolutely tremendous work to put causation on firm mathematical legs, directly tackling a subject that many consider taboo. For anyone with an interest and a good amount of mathematical fortitude (meaning you enjoy working out the intermediate steps on your own, and will make sure you understand concepts before proceeding), I'd highly recommend his book Causality. It explains precisely the assumptions and what types of measurements necessary in order to establish a causal relation, and therefore implicitly the conditions where it's inappropriate to claim a causal effect. It explains how to calculate the causal impact in the face of Simpson's paradox, how to evaluate counterfactual statements, and all sorts of other advanced reasoning. Standard methods for establishing causality, such as double-blind trials and interventional experiments, fall directly out of the methods in here, and the book gives greater understanding of why these tools work.

Perhaps in a few more decades, when pedagogy of causality has advanced where it can be offered to a large number of undergraduates and causality is a standard part of stats training, studies like this will have also advanced such that we can more easily trust conclusions, and have well-grounded discussions without getting too far into the weeds.
posted by Llama-Lime at 9:24 PM on January 3, 2013 [5 favorites]


My favourite correlation/causation chart of the week: Correlation between autism diagnosis and organic food sales
posted by wenat at 9:31 PM on January 3, 2013 [3 favorites]


Maybe we should drop the social science angle and actually talk about the medical science that the second part of the article discusses – since the article itself notes this very problem with correlation in social sciences.

Which article do you have in mind? Quite a number of the ones cited by the man of twists and turns combine epidemiology/social science with discussions of the connection between blood lead levels, brain matter, and behavioral problems. The question is not about whether lead harms the brain -- which has long been well-established -- but whether its effects actually caused the rise and later fall in violent crime rates. Many other things hurt kids who grew up in high-lead areas, so the causal connection between lead and brain damage does not prove that that is actually what caused crime. And as I was saying, the correlation between lead and even lead-caused brain damage and crime, on the individual level, still doesn't prove it, since it is plausible that lead levels correlate with other societal abuses. I don't know how you would answer that question using purely "medical science" without social science, though I may be missing your point.
posted by chortly at 10:03 PM on January 3, 2013


Well, I was mostly thinking of the second page of what sort of seems to be the main link in the post. I'm still kind of going over it. Medical science generally has some social components, since trials and studies are obviously over groups of people; but they tend to demand at least a bit more precision.

As I say, it summarizes a number of different studies and conclusions, and I'm still going over them, so I want to be cautious; however, if it's correct, then we can indeed say that lead is causative of a host of things that would both lessen quality of life and probably lead to increased crime rates. That doesn't prove that lead is the cause of those increased crime rates; but it does show relatively clearly that the investment involved in eradicating lead would not be wasted.

There are paragraphs in the article I still have a problem with, of course; for example:
Needless to say, not every child exposed to lead is destined for a life of crime. Everyone over the age of 40 was probably exposed to too much lead during childhood, and most of us suffered nothing more than a few points of IQ loss. But there were plenty of kids already on the margin, and millions of those kids were pushed over the edge from being merely slow or disruptive to becoming part of a nationwide epidemic of violent crime. Once you understand that, it all becomes blindingly obvious. Of course massive lead exposure among children of the postwar era led to larger numbers of violent criminals in the '60s and beyond. And of course when that lead was removed in the '70s and '80s, the children of that generation lost those artificially heightened violent tendencies.
That seems more than a little strong for the situation. I don't think it's "blindingly obvious." I think that the signs probably point to lead having a major impact on society, but society is a damned complicated thing. And there's an understanding of IQ here that I almost find ominous – the author is implying that if you're "on the margin" of being at a certain IQ, you're on the verge of becoming a violent criminal. Even if we assume that IQ is an acceptable standard measure (which is not an uncontroversial assumption, and I probably wouldn't go along with it) this is sort of an innovative way to interpret what IQ means.

In every case, however, lead needs to be eradicated. Even if people are just quietly, silently becoming more hyperactive, losing their ability to cogitate, and losing impulse control – even if there is relatively little impact on society at large (which seems doubtful to me, although I don't know for sure what that impact might be) – then we really owe it to ourselves to eradicate lead, I think.
posted by koeselitz at 10:17 PM on January 3, 2013 [2 favorites]


No disagreement about the need to eradicate lead, regardless of whether it had anything to do with the crime wave. But again, although it seems quite definitive that "lead is causative of a host of things that would ... lessen quality of life", the evidence that it "probably lead to increased crime rates" is much more uncertain. There's no way to prove this without social and epidemiological research (barring immoral experimentation), and there are good reasons to be wary of the results from these parts of the relevant papers that Drum is drawing upon. I've been following his blog posts on this topic for years now, and he's a bit more careful in this article, by mainly citing the deleterious effects of lead on brain matter and behavior. But when he says Needless to say, not every child exposed to lead is destined for a life of crime, he's jumping from the medical evidence -- lead causes brain and behavior problems -- to the sociological -- lead causes crime and the national crime wave. And that jump is much more unsupported that even his more guarded rhetoric in this most recent article warrants.

Again, I'm totally in favor of lead reduction -- clearly its effects are horrible. But Levitt's stuff on abortion shows how dangerous these jumps from mechanism to social impact can be: certain people (especially in economics and on the right) are quite happy to alight on these biological explanations because they (somewhat) absolve society of the myriad other harms done to these criminals when they were children. Because there is this plausible mechanism, and because people like surprising results, we're more likely to welcome these conclusions in a way we wouldn't if the result were about, say, building height (which also correlates strongly cross-sectionally and time-wise with later crime). But the evidence -- that lead caused the crime wave, not that it causes brain damage -- is just as weak as for the height of the projects people grew up in. Because we have a plausible mechanism, it's worth looking into -- and the work in Reyes (2007) does a good job of ruling out a lot of alternative hypotheses -- but because there is no way to answer the crime-wave part without social science, and because the policy implications are substantial, it's worth being very wary of the social-scale conclusions and the methods behind them.
posted by chortly at 11:27 PM on January 3, 2013 [2 favorites]


advances in police work which are proven successful in one area tend to get implemented in other areas as well.

Advances like what?
posted by obiwanwasabi at 12:58 AM on January 4, 2013


Given the known carcinogenic and mutagenic effects of benzene, it's a wonder that this discussion doesn't continue to examine the potential relationship between benzene and autism spectrum disorders. These disorders are strongly linked to paternal age and to large structural genomic variations (large deletions, for instance). It may eventually be of little surprise that the highest rates of autism are in places like Marin County, California, where an aging paternal population habitually commutes long distances by car into the city of San Francisco. But, perhaps there is another, stronger, correlate.

This subject has been researched recently: Residential Proximity to Freeways and Autism in the CHARGE Study. Also: Effect of paternal age seen in girls with autism.
posted by melatonic at 1:33 AM on January 4, 2013


My objection is that the article is written in a way that suggests that lead was the primary cause of the crime drop rather than Giuliani's broken windows theory. That is the nonsensical assertion that I would like to refute.

First of all, broken windows theory has been empirically tested in psychological studies, so there is actual data to corroborate it. Any worthwhile article that tries to disprove the utility of this theory should at least try to explain why those studies are incorrect.

Second of all, only gullible people believe data provided in an article that obviously has an agenda (in this case, proving the lead theory). Intelligent and skeptical people review the actual data on their own rather than letting somebody else summarize it for them. So, let's try being intelligent skeptical people and actually do that, shall we? I'll give you a few moments to do the calculations. (Spoiler: results may surprise you!)



As the article points out, the dramatic fall in crime rates was not limited to NYC. Take a look at some more data from the same web-source: the United States, the state of Illinois, and the District of Columbia. I didn't use an actual randomizing device to pick these, but I'd be willing to wager that if you pick ten random cities in the U.S. you will see the same basic pattern that you see for the U.S. as a whole: a sharp increase in violent crime rate starting in the 1960s followed by a sharp decrease in violent crime rate starting in the early 1990s.

Why is this a problem for the broken window hypothesis? Because the policing idea supported by the broken window hypothesis was not implemented nation-wide, and yet, crime fell nationwide. On the broken window hypothesis, one would expect to see decreases in places that implemented the theory and no decreases in places that did not implement the theory. But that is not what we see. What we see is a similar, dramatic decrease in crime rate regardless of policing philosophy. Hence, the 1990s gave us a nice natural experiment that tells pretty strongly against the broken window hypothesis.

Here is a good place to deploy Ockham's Razor, incidentally. The lead hypothesis has a single, simple explanation for the sharp increase in crime rates in the 1960s and the same simple explanation for the sharp fall in crime rates in the 1990s. The same explanation applies to jurisdictions that had a broken-windows style policing philosophy and in jurisdictions that did not. What explanation does the broken windows hypothesis give for the sharp increase in the 1960s? What explanation does it give for the sharp decrease in jurisdictions that didn't implement policing procedures recommended by the broken window hypothesis? Were those places just lucky? Any explanation offered on behalf of the broken windows hypothesis will be more complex than the lead hypothesis (and probably ad hoc as well). Hence, the lead theory is preferable.

(The point about corroborating theory has already been addressed, but to reiterate: broken window theory may have psychological studies to support it -- I haven't read them myself, but I'll take a look if someone wants to link a few; however, lead theory has corroboration from medical and neurological studies. I don't know about you, but prima facie, I trust the medical and neurological studies to have better ecological validity. Hence, I think that at best, the broken window hypothesis is neither differentially confirmed nor disconfirmed relative to the lead hypothesis by background research.)


What is far harder is predictive analysis - the ability to predict what will happen based on an existing hypothesis. Making predictions also involves a certain degree of risk (ie, being proven wrong and mocked) which indicates how much confidence the speaker has in their own analysis.


This just makes me scratch my head. The lead hypothesis has made -- and continues to make -- predictions. Here are a few mentioned in the article. Before seeing data, the lead hypothesis predicted that atmospheric lead concentration would be positively associated with crime rate. Having seen the data in one or two cities, the lead hypothesis predicted that the same pattern would obtain in other cities, regardless of their non-environmental policies. (Leading to the prediction that broken-window enforcing cities would have similar patterns of crime as non-broken-window enforcing cities, provided they had similar patterns of environmental lead concentration.) The lead hypothesis predicted that decrease in crime would trail decrease in atmospheric lead concentration by about a generation: 25-30 years. The lead hypothesis predicted that in places where lead was removed earlier, crime would decrease earlier. The lead hypothesis predicted that the same patterns would show up internationally. The lead hypothesis predicted that the same patterns would hold for every choice of size of statistical unit: nation, state, county, or neighborhood. Those predictions were satisfied, and they were reported in the article.

But that's not all. The lead hypothesis makes further predictions that have not yet been tested. First, there is a policy prediction, with detailed, measurable economic consequences specified in the article: holding everything else constant, if we decrease the concentration of lead in the soil and atmosphere, we will see a further decrease in violent crime. Similarly, the lead hypothesis predicts that if we do nothing at all to change environmental lead concentrations (and those concentrations do not change naturally), then we will not see a huge spike and drop in crime rates as we did in the twentieth century. In other words, the lead hypothesis denies that the spike was naturally occurring as part of a cycle or noise in the system or something like that. So, seeing an unexplained huge spike would be disconfirming for the lead hypothesis.
posted by Jonathan Livengood at 4:35 AM on January 4, 2013 [5 favorites]


My favourite correlation/causation chart of the week: Correlation between autism diagnosis and organic food sales

My god! Why do autistic people love organic food so much?!?
posted by The 10th Regiment of Foot at 4:43 AM on January 4, 2013


In other words, the lead hypothesis denies that the spike was naturally occurring as part of a cycle or noise in the system or something like that. So, seeing an unexplained huge spike would be disconfirming for the lead hypothesis.

Ask and you shall receive!

Seriously Jonathan, I admire the fact that you had the balls to put your credibility on the line, but you have totally staked out the wrong hill to die on. Frankly I'm surprised that you would say something like that when even the most basic Google search show dozens of hits for a nationwide spike in violent crime - a spike which completely fails to fit the time delay associated with the lead hypothesis.

As for your alleged "medical basis" behind the lead theory of crime, that's laughable pseudoscience. The only thing that medical science has proven is that lead causes neurological damage. Nobody is debating that, and if you think this is the bone of contention then you have either lost track of the bouncy ball or you are deliberately fabricating a straw man to fight against. It's a huge leap in the logic of causality to assume that this neurological damage causes crime, and there is zero medical evidence for it - simply this one "expert's" data correlation (which, as I have already demonstrated, was poorly done and ignores significant amounts of contradictory evidence).
posted by wolfdreams01 at 6:37 AM on January 4, 2013


The only thing that medical science has proven is that lead causes neurological damage. [...] It's a huge leap in the logic of causality to assume that this neurological damage causes crime, and there is zero medical evidence for it

I'll just leave this very simple PubMed search here.
posted by OmieWise at 6:52 AM on January 4, 2013 [1 favorite]


I'll just leave this very simple PubMed search here.

Aggression != crime. Please don't artificially conflate two separate things - it's a cheap tactic. I'm aggressive as hell, but I easily manage to avoid having a criminal record. Our society has plenty of ways to channel aggression, whether that is sports, military service, martial arts, BDSM, or even a job that requires conflict.
posted by wolfdreams01 at 7:05 AM on January 4, 2013




::rolls eyes::

Okay. So far, I've pointed out that the actual data cited in this study was grossly misleading, that statistics are glossed over in a huge way, and that trends that do not fit the hypothesis are completely ignored by the researcher. Whereas you've successfully managed to prove that I phrased one sentence in a less-than-adequate manner, even though it is completely unrelated to our overall point - which is crime.

I don't even know what to say to that. "Enjoy your victory," would be appropriate, I suppose. "You really earned it, champ."
posted by wolfdreams01 at 7:51 AM on January 4, 2013


[Folks, what we said upthread still holds true. Act decently to each other or go elsewhere.]
posted by jessamyn at 8:00 AM on January 4, 2013


Aggression != crime

No but aggression does more often lead to violence, which under most circumstances is criminal.
posted by The 10th Regiment of Foot at 8:20 AM on January 4, 2013 [1 favorite]


It's a huge leap in the logic of causality to assume that this neurological damage causes crime, and there is zero medical evidence for it

Research Links Brain Damage & Violent Crime -- USC Studies Point To Underlying Causes Of Violent Crime In Young Offenders
Mental disorders among the subjects included schizophrenia, organic brain damage and a history of head injury. PET scans measure the uptake of blood sugar (glucose) in various brain areas during the performance of simple, repetitive tasks. (Glucose is the basic fuel that powers most cell functions. The amount used is directly related to the amount of cell activity.) On average, the murderers showed significantly lower rates of glucose uptake in three areas of the brain -- the prefrontal cortex, the corpus callosum and the posterior parietal cortex. ... The prefrontal cortex is involved in the inhibition of aggressive behavior. Studies have shown that damage to the region correlates with impulsiveness and unpredictable, uncontrolled actions.
Blows to the head during development can predispose to violent criminal behaviour: rehabilitation of consequences of head injury is a measure for crime prevention
However, what differentiated the violent from the non-violent group was a history of having suffered head injuries that were never treated. Problems at school are not enough themselves to predict violent behaviour. A history of discrete neurological damage as a consequence to blows received to the head must also be present.
Alcohol, intelligence, and violent crime in young males
These interactions indicate that the prevalence of violence increases significantly at low intelligence and high alcohol consumption levels. A parallel analysis with nonviolent offending as the dependent variable failed to find significant interactions. The combination of heavy drinking and lower intelligence is associated with a synergistic surge of violent behavior.
Violence in Children and Adults: A Neurological View
Repeatedly violent behavior exists as a distinct behavioral syndrome with neurological, psychiatric, and environmental determinants. Neurological damage (electroencephalographic abnormalities, epilepsy, symptoms of psychomotor seizures, and a history of events known to predispose to brain injury) is prevalent.
posted by the man of twists and turns at 8:27 AM on January 4, 2013 [5 favorites]


Seriously Jonathan, I admire the fact that you had the balls to put your credibility on the line, but you have totally staked out the wrong hill to die on. Frankly I'm surprised that you would say something like that when even the most basic Google search show dozens of hits for a nationwide spike in violent crime - a spike which completely fails to fit the time delay associated with the lead hypothesis.

Well, there are spikes and then there are SPIKES. The articles you linked are discussing large, single-year increases on the order of 10-40% over the previous year. The spike that we're talking about with respect to U.S. crime rate was much larger and multi-year. Just to put concrete numbers on it, the violent crime rate in NYC increased 345% in the period for which you provided data (up to its peak in 1990). From 1960 to 1991, the U.S. violent crime rate increased 471%. The decrease was just as dramatic. In IL, it increased 285%, and in DC, it increased 528%.

You might notice that on the downhill side of the curve for NYC, there was a little jump from 1999 to 2000. That was nearly a twelve percent increase in violent crime. But it is clear from a graph of the data that that bump was noise. I would not be surprised to find "drastic" single-year increases or decreases -- changes in the 10-40% range -- showing up pretty regularly in the time series for other cities, states, and nations.

Apologies for not being clear enough, but when I say the lead hypothesis excludes similar spikes, I am not saying it excludes single-year increases (even large ones) like the ones you cite. I am saying it excludes 25-30 year trends an order of magnitude greater than any of the ones you cite.

As to the medical studies, are you seriously denying that the average aggressiveness of the members of a population is irrelevant to the rate of violent crime for that population?
posted by Jonathan Livengood at 8:44 AM on January 4, 2013 [2 favorites]


As to the medical studies, are you seriously denying that the average aggressiveness of the members of a population is irrelevant to the rate of violent crime for that population?

Spoken like a true pacifist! I think you may be letting your distaste for aggression blind you to its utility. There are multiple ways to express aggression, and some are in fact not only useful but necessary to a functioning society. Tell me, do you think SWAT teams do what they do for the money? Or for the sense of social obligation? Do you think soldiers only join the military because they have no other choices? In almost all of society's attack dogs (and please understand I don't call them "attack dogs" in a derogatory way - as a Taoist, I believe animals are more spiritually pure than humans) whether soldiers, police officers - you are likely to find a strong streak of aggression. Similarly, you will find this trait in many people who have the "attack dog" role, whether that is for corporations or other groups. These people are often vital to the functioning of their respective entities, and their aggression is a useful and desirable personality trait. Associating aggression with violence is an unfair stereotype, and I think it's demeaning for you to brand all these people as potentially violent without evidence.
posted by wolfdreams01 at 11:34 AM on January 4, 2013


Do you think soldiers only join the military because they have no other choices?

Wait, in this case I'd say very often they do!

Associating aggression with violence is an unfair stereotype, and I think it's demeaning for you to brand all these people as potentially violent without evidence.

Are you saying that cops and soldiers aren't violent because they take their aggressions out in other ways? Do you have any studies that back that premise up?
posted by The 10th Regiment of Foot at 11:52 AM on January 4, 2013


Associating aggression with violence is an unfair stereotype, and I think it's demeaning for you to brand all these people as potentially violent without evidence.

Projection on aisle 4759638.

Equating aggression with crime and/or violence is not a particularly controversial nor demeaning idea. No one is saying that all aggression is criminal and/or violent, but if you don't think that criminals and violent people are -- broadly speaking -- more aggressive than non-criminals and non-violent people, then you are living in a cave.
posted by Etrigan at 11:55 AM on January 4, 2013 [1 favorite]


The point is statistical. There is variation in every group. Among the more aggressive people, you will have some who channel their aggression and do not commit violent crimes. Maybe they join the military or the police (although like The 10th Regiment, I would like to see some evidence that catharsis works with respect to violence). However, you will also have some people who want to commit violent crimes. Similarly, some non-aggressives will want to commit violent crimes and some will want to go into the military, etc. However, the probability that a randomly chosen aggressive commits a violent crime (or goes into the military) is greater than the probability that a randomly chosen non-aggressive commits a violent crime (or goes into the military). Thus, if you increase the percentage of the population that is aggressive, you will increase the percentage of people in the population as a whole who want to go into the military, the police, etc. and you will increase the percentage of people in the population who want to commit violent crimes.

Illustrative example. Suppose I have a population of 100. Suppose the probability that an individual commits a violent crime (in a year) given that the individual is non-aggressive is 0.08 and the probability that an individual commits a violent crime given that the individual is aggressive is 0.1 -- slightly higher but still pretty unlikely. Suppose the population is divided into 70 non-aggressives and 30 aggressives. Then the violent crime rate will be 70*0.08 + 30*0.1 = 5.6 + 3 = 8.6 crimes per year. Then suppose the population changes to 50/50. Now the violent crime rate is 50*0.08 + 50*0.1 = 4 + 5 = 9 crimes per year. The numbers are, of course, fanciful, but the idea should be clear.

I am in no way saying that aggression is always bad or that aggressive people are necessarily going to become criminals. I am saying that an aggressive individual is more likely to commit a violent crime than is a non-aggressive individual, and therefore, in aggregate, having a larger percentage of aggressive people in the population increases the rate of violent crime.

Every human has the potential to commit a violent crime.* Humans who are more aggressive (by nature, by nurture, or by poisoning) have a higher probability of committing violent crimes. (And I'm not saying this without evidence: many links have already been posted to peer-reviewed medical literature asserting that neurological damage in general, and neurological damage caused by exposure to lead in particular, causes violent behavior, not just aggressive personality.)

If you think something in my argument here is incorrect, I would very much like for you to point it out.

* With obvious qualifiers for humans in comas and so forth.
posted by Jonathan Livengood at 1:36 PM on January 4, 2013 [1 favorite]


I think a lot of SWAT and other police paramilitary unit members do it for fun. And it's not as if all that legal aggression is an unmitigated good. Innocent people get killed or jailed; municipalities lose millions and millions of dollars in lawsuits. And that's not even counting the famous ones, like Rampart or the Riders or the NOPD.
posted by rtha at 3:13 PM on January 4, 2013 [1 favorite]


The Freakonomics people did the same thing with abortion. Choice lowers crime
posted by nikodym at 3:38 PM on January 4, 2013


Projection on aisle 4759638.

You know what? It pains me to say this, but as a logical person I must admit that this statement may have been correct. As somebody with an aggressive nature, people like me are often on the receiving end of a lot of unfair characterizations made by a society which often claims to loathe aggression while at the same time seems all too willing to exploit the results of it. So I admit that my instinctive tendency is to jump in and defend my kind when I feel like this unfair bigotry may be at play.

However, Koeselitz was a big enough person to admit the validity of my points earlier, when I pointed out the significant data errors in the study, so it's only fair for me to do the same when somebody else makes an excellent point. (And I realize that you are not deliberately attacking me - to some extent, the aggression/violence stereotype does exist for a reason, and it would be irrational to ignore that there must be at least some correllation between the two.) So now that I have had the chance to gain some perspective and consider this rationally, I must agree that - even though the data cited in the study was clearly filtered to achieve the results that the researcher wanted to see - from a purely medical perspective, the effects of lead poisoning on aggression and violence are definitely worth exploring.
posted by wolfdreams01 at 4:10 PM on January 4, 2013


I just wonder what we're going to find out was poisoning us all along in 20 or 30 years. The common guess in people I've discussed this with is High Fructose Corn Syrup but I think it's pretty clear now it's terrible.
posted by Ghostride The Whip at 9:57 PM on January 4, 2013




atrazine: "Apparently cold water pipes aren't such a problem in most areas as mineral deposits quickly form and prevent the lead from contacting the water."

True until municipalities went from using from chlorine in the water to chloramine to reduce formation of toxic reaction products. This changes redox conditions and results in release of lead (PDF) which was/is a big problem in Washington, DC. A friend of mine is a plaintiff in a class action suit against DC Water and has two adopted kids who were affected by this.
posted by exogenous at 12:54 PM on January 7, 2013


The Guardian:Yes, lead poisoning could really be a cause of violent crime
It's ridiculous – until you see the evidence. Studies between cities, states and nations show that the rise and fall in crime follows, with a roughly 20-year lag, the rise and fall in the exposure of infants to trace quantities of lead. But all that gives us is correlation: an association that could be coincidental. The Mother Jones article, which is based on several scientific papers, claimed causation.

I began by reading the papers. Do they say what the article claims? They do. Then I looked up the citations: the discussion of those papers in the scientific literature. The three whose citations I checked have been mentioned, between them, 301 times. I went through all these papers (except the handful in foreign languages), as well as dozens of others. To my astonishment, I could find just one study attacking the thesis, and this was sponsored by the Ethyl Corporation, which happens to have been a major manufacturer of the petrol additive tetraethyl lead. I found many more supporting it. Crazy as this seems, it really does look as if lead poisoning could be the major cause of the rise and fall of violent crime.
His Science Is Too Tight!:The Link Between Leaded Gasoline and Crime
Obviously, we can't take this a step further into experimental data to enhance the hypothesis. We can't expose some children to lead and not others on purpose to see the direct effects. This is the best we can do, and it's possibly quite meaningful, but perhaps not. There's no way to say with much authority one way or another at this point, not just because of the smallish sample size and the mixed results on significance. Despite an improved study design from cross-sectional studies, a cohort study is still measuring correlations, and we need more than one significant result.
Both from a followup post by Kevin Drum.

I agree with Firestone at HSITT. It really is too bad only two of these studies linking lead exposure with crime exist, and there are no links between neurological damage and violent and/or criminal behavior.

I guess this is a totally unexplored field ripe for study.
posted by the man of twists and turns at 4:21 AM on January 9, 2013 [2 favorites]


If you haven't already, you might want to take a look at the PubMed search results that OmieWise linked above, since several of them report laboratory experiments establishing a connection between lead exposure and violence in animal models, like hamsters, mice, rats, and cats.

It is true that we cannot take the large step to controlled experimental research on humans, but we have taken the intermediate step to controlled experimental research on several species of non-human mammals. Insofar as those mammals are good analogues for humans, we have experimental evidence confirming the lead hypothesis.
posted by Jonathan Livengood at 12:24 AM on January 11, 2013 [3 favorites]


NRO:Lead And Crime
But even under all these unrealistic assumptions, what this model hasn’t accounted for is that the evolution of political economy over time during these decades could systematically vary between late- and early-phase-out states. This could easily be the case, if the evolution of, say, the political economy of Sunbelt states versus Rustbelt states evolved systematically differently over a time frame of many decades, and Sunbelt states tended to phase out leaded gasoline either earlier or later than Rustbelt states. In that case, you would have to consider each state/year combination as a control rather than just each state. Think this is an obstructionist objection? Reyes herself considers this possibility significant enough that she does this analysis. The result? No statistically significant relationship between lead and violent crime.
Mother Jones:Lead And Crime: Response
But that's far from all. If Reyes' paper was the only evidence for the link between lead and violent crime, I'd agree with Manzi. It's not enough. But it's far from the only evidence. We have striking evidence at the national level, of course. We have evidence at the city level. We have evidence that merely living in a housing project near an expressway is associated with more crime. And most important, we have evidence at the international level. As long as the data is all from the United States, you can argue, as Manzi does, that there might be some systematic effect of culture or political economy that's hidden in the background and affecting the results. But it's a lot harder to say that when you find the same results in Vancouver, Montreal, and Toronto. And it's way harder to say that when you find the same results in Britain, France, and Australia over different time periods (because different countries banned leaded gasoline at different times). Sure, it's still possible that there's some systematic hidden variable affecting these results, but that would be a helluva thing, wouldn't it? You're talking about some aspect of culture, or policing tactics, or drug use, or automobile preference, or whatever, that affects country after country around the world.
posted by the man of twists and turns at 9:55 PM on January 14, 2013 [2 favorites]


Lead and Crime: Why this correlation does mean causation (PDF with embedded links)
Indicators of causation, now widely applied in public health research, were first proposed in Bradford-Hill’s 1965 treatise “The Environment and Disease: Association or Causation?” Those who dismiss research linking lead exposure and crime trends as “simply looking at associations” need to reexamine that evidence in the context of Hill’s causal criteria.
It references this paper: The Environment and Disease:Association or Causation? (PDF)
I have no wish, nor the skill, to embark upon a philosophical discussion of the meaning of 'causation'. The 'cause' of illness may be immediate and direct, it may be remote and indirect underlying the observed association. But with the aims of occupational, and alnost synonymously preventive, medicine in mind the decisive question is whether the frequency of the undesirable event B will be influenced by a change in the environmental feature A. How such a change exerts that influence may call for a great deal of research. However, before deducing 'causation' and taking action we shall not invariably have to sit around awaiting the results of that research. The whole chain may have to be unravelled or a few links may suffice. It will depend upon circumstances.
Summary here: Lead and Crime: Is Correlation Also Causation?
The key statistical issue that needs to be addressed by the correlation-never-means-causation crowd is whether they honestly believe that:

■The observed association between lead used in paint and USA murder rates from 1901 to 1960 with a time lag close to the peak age of homicide offending was a coincidence;
■The association between USA gasoline lead and violent crime from 1964-1998 with a similar time lag was another coincidence;
■The “experimental evidence” from violent crime since 1998 (including a 45% drop in the juvenile violent crime arrest rate from 1998-2011) tracking earlier trends in lead exposure is a coincidence;
■Analysis of crime in nine nations shows the same consistent relationship between lead exposure and crime trends through 2002, with statistical best-fit lags that reflect the peak age of offending for each crime category, by coincidence;
■This consistent relationship within every nation studied happens to explain otherwise bewildering changes over time in USA and Canada crime rates relative to Britain, France, and Australia, by coincidence;
■Experimental evidence from international crime trends since 2002 tracking earlier trends in lead exposure in every nation is also a coincidence.
posted by the man of twists and turns at 2:53 AM on January 30, 2013


I saw this when it came out earlier this month, but I wasn't aware of the metafilter FPP until just now. Interesting discussion. The main argument seems to be between "Skeptics" who don't think the evidence is strong enough. They bring up a lot of different points, but I think the are pretty obviously wrong.

There are a couple problems
1 - They're obviously not familiar with all the research - they keep making claims about the studies that are not correct, like saying the data is just nationwide or state by state, when in fact childhood lead exposure data goes all the way down to the individual level, with millions of data points. County and city data with thousands of data points and data for various countries around the world. All the data supports the conclusion. They say that the data only looks backwards at stuff that's already happened, but in fact the lead hypothesis has been tested in animals. Animals exposed to higher lead levels are more violent.

2 - There's no math in any of their comments and no actual counter evidence. They continue to claim that the evidence isn't strong enough, but present no evidence that it's not true. There's no mathematical analysis to show that the actual statistical results are inaccurate. If you can show statistically that A causes B with high probability, then you have to show that there is actually something wrong with the math or the data

3 - Finally, there is a problem with the form of their argument. Skepticism is fine, but if you're doing "real" science there has to be some threshold specified in order for you to consider the argument valid. This is especially problematic here because the "skeptics" - being unfamiliar with the data - keep pointing to things they think are lacking that turn out to actually be there - like county, by county, or individual data.

Put another way, the skeptics are actually positing their own hypothesis - that lead is not a major cause of crime. Like everyone they need to propose an experiment that would disprove their hypothesis, or cause them to agree that the childhood lead exposure is not a major cause of crime is not true. Without specifying that, they aren't really making a scientific argument. They're really just like the creationists who look at the evidence for evolution and claim (falsely) it's not enough, or the global warming deniers who claim (also falsely) that there isn't enough science or all the science is bad or whatever.
Anyway, seems pretty clear to me that there is a ton of solid evidence for the "childhood lead exposure causes crime" hypothesis and seeming no evidence for the "childhood lead exposure does not cause crime" hypothesis.

Also, finally, there is an "experiment" you can do, namely to spend the money to remove "legacy" lead from various places and check to see if the crime rate drops (as well as checking lead levels in kids and seeing what the future correlation ends up being). It would be an unethical experiment to leave some places contaminated as a control, but given our current political system it's likely some places won't be decontaminated at the same time.

___________________________________________________
Anyway, some responses to some off the offshoot comments from when I was reading through this:)
I guess this means I should stop licking lead bars. -- Damienmce
Should be fine so long as you don't also lick small children. (Also, if it's anything like mercury, the problem isn't the metallic form, but rather molecules containing lead that your body can easily absorb - but never get rid of)
If lead can have far-reaching effects, what about fructose? -- infinitewindow
It's interesting how so many people are just certain fructose must be bad. Sure, sugar in general is bad, but there seems to be this underlying assumption that fructose is somehow worse the sucrose.

(also, people who love to argue about this stuff always make evolutionary arguments, but if you think about it primates and apes in the wild would have had much more exposure to fructose then sucrose, since it's present in fruits, while sucrose is only present in a few plants like sugar cane they wouldn't have had much access too. Fructose consumption is very common in the animal world while sucrose in diets is actually the new thing.)
Correlation does *not* mean "having behavior represented by the same equation". By that argument, since all falling bodies on the Earth follow the same equations of motion, they are all correlated. -- lupus_yonderboy
er, technically all falling bodies do have the same cause, namely gravity :). But yeah obviously various things that seem to be governed by similar mathematical formulas, so long as the underlying processes is mathematically similar. Compound interest and bacterial colony growth, for example are both exponential processes.
As for the correlation/causation chestnut... correlation doesn't prove causation, but it sure as heck is evidence for causation - moreover, it's a necessary condition, because lack of correlation does in fact prove lack of causation. -- lupus_yonderboy
There are a couple of other things to keep in mind as well: 1) Timing - if B happens after B, B can't go back in time and cause A. That leaves only A causing B, or X causing both A and B.

The other is a plausible explanation. There's a straightforward and scientifically provable theory that lead causes brain problems that can cause increases in crime. What's the possible "X" here? Childhood exposure to environmentalism causes crime reduction as an adult, as well as the removal of lead from gasoline? Seems pretty unlikely.
My favourite correlation/causation chart of the week: Correlation between autism diagnosis and organic food sales -- wenat
Simple causal explanation for that one, paranoid hippy parents more likely to buy organic and also be paranoid about their kids developing autism, thus more likely to get them checked out.
Is Dunning-Kruger the new Godwin? Not saying it is or isn't happening here, but it seems like half the time people bring it up they are unwittingly referring to themselves.
- Pyry
Problem is, Dunning-Kruger probably does apply to, like, 90% of internet commentators.
[regarding non-lead bullets] Tungsten matrix [bullets] is really much better ballistic ally and in terms of damage to old guns not designed for harder metals but it is very expensive. It also isn't really non-toxic in the way that steel is.
I thought the discussion at the top about non-lead bullets was kind of interesting too. The thing is, lead is actually pretty expensive, and it seems like steal bullets would actually be cheaper. (Obviously it was cheaper historically, but right now lead is ~$2000 a ton while iron ore is ~$120 and scrap steel is maybe $375?)

I wonder if they couldn't coat it with some kind of plastic shell to make it easier on the insides of guns.

In other words, the lead hypothesis denies that the spike was naturally occurring as part of a cycle or noise in the system or something like that. So, seeing an unexplained huge spike would be disconfirming for the lead hypothesis. -- Jonathan Livengood
Well, not necessarily. Other causes may cause spikes in other places. Economic problems, for example. Or other pollutants.
Have some professional criminologists commented on these findings? I've read that one issue with crime stats is that they may not be comparable over a long time period. -- elgilito
From the article, no. They haven't looked at it much. I don't want to speak for the article's author, but my takeaway is that they're kind of bothered by the idea that all their theories and work over the past decades have been meaningless - nothing they've done caused the decline in crime (see Giuliani, for example - but lots of people have taken credit for the drop in crime). The other problem is, well, their field lacks a lot of scientific rigor.

I think there's a tendency in "human" sciences like criminology or economics to think outcomes are based on "moral" inputs, like how all these economists want to blame "morally weak" spending for the economic problems and prescribing "morally virtuous" austerity to fix it. Similarly criminologists want to think that bad morals cause morally bad outcomes - single parenting, culture, lack of respect for authority as causing crime rather then something that has nothing to do with anything "moral"
"Remove shop-lifters and cure violent crime!!" and "Remove lead and cure violent crime!!" are equally simplistic, and I refuse to believe that anyone who has an actual real daily life in any inner-city neighborhood (or rural slums) would think up anything so facile. ... but both avoid discussing the root causes, which are poverty and segregation -- mumimor
The problem is that there is very good evidence that lead is in fact one of the biggest "root causes" of crime in the 70s/80s/90s. Also, it's interesting you brought up "rural slums" since one of the basic things being pointed out is that the crime spice didn't affect rural areas the way it did inner cities.

Of course since then environmental lead has been drastically reduced, and while crime is lower in the US then it ever has been in history (yes, lower then the utopia of the 1940s/1950s) it still exists and would exist even with the complete removal of all lead exposure, so obviously we should continue to work on reducing poverty, etc. Plus, poverty is bad irrespective of crime.

Anyway, science doesn't care how you "feel" about the findings.

_____________________________________________________
Anyway, on to some specific responses to main argument in the thread
Giuliani had the balls to risk his reputation by making a predictive analysis based on his "broken-window theory," taking action on this prediction, and achieving results that were almost exactly what he anticipated. I'd say that gives him far more credence than this conspiracy-theorist researcher, who stakes nothing in his analysis and makes no predictions that can be used to verify or disprove his credibility. --wolfdreams01
Except that crime dropped less in NYC then other cities, including cities that didn't do much of anything in that time period. If you think about it, that makes sense as well, much less driving in NY then in other places, so lead levels would have dropped comparatively less.

Anyway, you have to understand how statistical science works. You can make a prediction about any data you haven't yet seen. So you can predict that any country that removes lead from gasoline will see a drop in crime 20 years later, then you go out and get data to see if that's correct.

You might say that you can "cheat" here by checking the data first, but you'd just be doing the experiment "for real" when you checked. You can do the same thing with any "real" experiment by simply running it first.

Finally, you can statistically analyze results to determine what the probability of the correlation simply being random is. More data points makes for more precise results. Giuliani's "experiment" with NYC isn't very useful, because A) there's only one data point and B) crime didn't drop more then the "controls" i.e. the cities that didn't implement his policies. In fact, he did worse

I see. So it's far easier for you to believe in the idea of a molecule that causes "criminal behavior" 29 years later (because apparently all people's physiology is the same, and 29 -- wolfdreams01
23 years. Also, our entire war on drugs is predicated on the idea that exposure to various molecules can cause crime. The whole anti-depressant industry is based on the idea that molecules can alter people's internal mental states, which in turn changes how they interact with other people. We know molecules can help with anxiety, stress, attention problems.

The idea of molecules altering behavior is not exactly an alien concept for most people.


This is something that holds up from the individual level, the neighborhood level, the city, the state, and the national level.
My objection is that the article is written in a way that suggests that lead was the primary cause of the crime drop rather than Giuliani's broken windows theory. That is the nonsensical assertion that I would like to refute. -- wolfdreams01
Again dude, you have to look at the math. And the math You can do the math and figure out exactly how much of the cause is lead. And, when you do that math the result is 90% - lead was responsible for 90% of the change in the crime rate. Other factors, then would be responsible for about 10%.

Since there's still lead out there, there's still removal work that should be done and we can theoretically reduce crime rates even more by getting rid of the rest of it.

Someone brought up Japan and I found this study indicating that children have unusually low lead levels in their diets (but high cadmium levels, which is bad for kidneys). this article compares various countries including Japan, but is behind a paywall. this one mentions that Japan was actually one of the first countries to get rid of lead gasoline, starting in the early 1980s, with complete removal by 1986. (Random factoid of the day: apparently the children of samurai in pre-industrial japan were exposed to a lot of lead)

Also, since we have an environmental cause to explain high crime levels in past decades, perhaps it's possible that there is some environmental factor in japan keeping crime rates low, compared to other countries? Maybe all the squid or raw fish they eat? We don't really have any way of knowing.

Spoken like a true pacifist! I think you may be letting your distaste for aggression blind you to its utility. There are multiple ways to express aggression, and some are in fact not only useful but necessary to a functioning society. Tell me, do you think SWAT teams do what they do for the money? Or for the sense of social obligation? Do you think soldiers only join the military because they have no other choices? In almost all of society's attack dogs (and please understand I don't call them "attack dogs" in a derogatory way - as a Taoist, I believe animals are more spiritually pure than humans) whether soldiers, police officers - wolfdreams01
First of all, when social scientists talk about aggression, they only mean a tendency to violence. If a person has higher aggression then they are prone to violence. That's what scientists mean when they use the term. Of course, people are capable of self control and may not act on their aggression. But they are more likely to be violent if they show high aggression. And as Jonathan Livengood has pointed out, higher aggression in a population will mean higher violent crime levels.


Secondly, if you think police should be prone to violence you have a pretty fucked up view of society, although not too surprising given your other comments - that we are at war with ourselves and police are soldiers in that war. Aggression in soldiers depends on their mission - in our recent wars we were trying to "win hearts and minds" and violent incidents involving troops. The Iraq and afghan war would have gone a lot better if we didn't have a handful of soldiers who engaged in violence against civilians. My view is that most people are good and that the job of the police is to help people - even when dealing with people who have broken the law, police should be patient and try to minimize violence as much as possible.

Thirdly, In general, I'm against war but looking at it a-morally I don't think we need soldiers to have nearly as much aggression as in the past. "modern" warfare requires we get along with civilians, and furthermore, combat is much less "intimate". Rather then skewering people with swords we're launching drones strikes from millions of miles away. The less "salient" and intimate the violence, the easier it is for people to do.

"Spiritual purity" is not a scientific concept. (And by the way, attack dogs are a product of training by humans, they are not naturally aggressive - just go on youtube and search for "Doberman and baby", "pitbull and baby", "rottwiler and baby" or whatever)
Hey, as long as we are considering this to be a valid line of scientific inquiry, I would also like to point out that the crime stats also correlate roughly to the time period when people started breakdancing, feathering their hair, and wearing just one glove as a fashion statement. Obviously their cognitive function was impaired, and we must thus take it extremely seriously when researchers suggest that the lead molecule actually causes douchiness. -- wolfdreams01
Do you have individual data for 2,772 U.S. counties? Millions of individuals? Obviously fashion trend timing differs from county to county so it wouldn't be impossible to check. Why don't you submit a grant application and get back to us with the results?
_________________________________
some responses to other 'skeptical' comments
a clear continuum was shown for various towns at various sizes with various levels of lead exposure it would be a lot more convincing, -- Kid Charlemagne


Well, there's this article, linked earlier looks at 2,772 different counties. And in addition they have millions of data points for individual children too, and the correlation still holds. People who were tested for lead levels as kids -- year later kids with higher levels were more likely to end up in trouble with the law.
Well for starters, don't make claims like this while failing to account for demographics, drug-use, both legal repercussions and law enforcement, population density, poverty and wealth migration, and perhaps most importantly, a lack of granularity finer than state level. -- clearly
We have data down to the individual level, as well as county, city, as well as data from other countries. This is a good example of 'skeptical' comments specifically complaining about the lack of data that actually exists, which they would know if they'd RTFA.
And that jump is much more unsupported that even his more guarded rhetoric in this most recent article warrants. -- chortly
The problem is there's no actual basis for saying it's "unsupported"? That's the problem I was talking about at the top of my comment the "skeptics" don't actually know how much support there is, since they haven't RTFAs and they haven't specified how much evidence, or what kind of evidence there needs to be for it to be "supported". The "support" for the "jump" between medical and sociological is that the data for society wide effects shows a clear link, and there's no plausible explanation of how crime rates could affect environmental lead levels 23 years in the past, or what third hidden cause could cause both.

If you don't think it's true, you have to be able to explain what the mathematical probability of it simply being a false correlation caused by statistical noise is.
__________________
Anyway, the thread is pretty much dead and this was kind of a huge waste of time I should have just read the whole thing and just written the summary, rather then writing replies as I read though the thread. Oh well.
posted by delmoi at 3:54 AM on January 30, 2013 [2 favorites]


There's no math in any of their comments and no actual counter evidence. They continue to claim that the evidence isn't strong enough, but present no evidence that it's not true. There's no mathematical analysis to show that the actual statistical results are inaccurate. If you can show statistically that A causes B with high probability, then you have to show that there is actually something wrong with the math or the data

Clearly you missed this comment, where I did exactly that. The skeptics are not arguing that "correlation does not equal causation" - they are arguing that the study is incorrect and that the actual data does not show a strong correlation in the first place. In other words, the author of the study grossly distorted the analysis of the data to achieve results that would support his conclusion.

Also, based on all the links you see cited, it might seem that there is a preponderance of evidence, but most of them are simply fluff pieces that (when you read fully and trace the citations) turn out to mostly be based off the original study, and don't fact-check enough to notice that the study's math is flawed.

This does not necessarily mean that the conclusion is flawed (the medical evidence connecting lead to violence is pretty strong) simply that the person conducting the study is very bad at both math and data-driven analysis.
posted by wolfdreams01 at 5:46 AM on January 30, 2013


Clearly you missed this comment, where I did exactly that.
Is that a joke?

You took the crime statistics for NYC and said it dropped faster when Giuliani took office. How does that explain the drops in crime rates in every other city in country? How does that explain data for 2700 counties? How does that explain the data for individual children who were tested for lead exposure and were more likely to get criminal records if they had higher concentrations of lead in their blood? It obviously doesn't. It's a joke.

Secondly, you didn't do what I said you needed to do. What I said you needed to do was explain why the math the researchers used was wrong.

What mathematical error are they making? If you can't explain what it is, you actually don't have a counter argument at all.
posted by delmoi at 3:45 PM on January 30, 2013 [2 favorites]


« Older Hi Stu, it is a pleasure to meet you.   |   Parakeets in the Produce Newer »


This thread has been archived and is closed to new comments



Post