No new petrol or diesel cars by 2040

[JohnD]

I would need to look deeply into the available data before I got anywhere close to being convinced.

Very fair.

Start here
http://www.nber.org/papers/w13097.pdf

This is an easy-reading article
http://www.telegraph.co.uk/news/ukn...lution-behind-the-baby-boomer-crime-wave.html

More links from here
http://lead.tulane.edu/about_us.html

 

[JohnW2]

If anyone can find any evidence suggesting that the number of crimes goes up and down according to the amount of bread consumed, I'd like to see it. .... After five years of people trying to pick holes in the leaded petrol hypothesis, can you find nothing more convincing?

It is by no means unbelievable that environmental exposure to organic lead compounds played a (probably small) part in influencing the number of violent crimes in the US. What is much harder to believe (but which seems to be the implication of the stuff you're citing) is that the changes in crime rate during the period in question were largely, perhaps even 'almost entirely', due to changes in that exposure 2-3 decades earlier, given the vast array of factors which surely must influence criminal behaviour.

As I said, it's really only the bit of that graph after lead exposure began to fall that one can really even take any notice of. One really can't attempt to draw any conclusions (in terms of causality) from the fact that violent crime increased progressively during a 33 year period and the fact that exposure to 'gasoline lead' also increased progressively during a 33-year period 23 years earlier. I'm sure you would not have to look very hard to find countless other things that showed progressive increases (or decreases) during the 1941-1976 period, but no-one appears to be suggesting that any of them were responsible for the corresponding increase in violent crime 23 years later.

By the same token, looking at the more important part of the graph beyond 1997 (and 1964), I don't think it would be hard to find plenty of things (in addition to environmental lead) which had shown progressive decreases (or increases) during 1964-1985 - so any of them could be accused of being the cause of the changes in number of violent crimes. ... not to mention changes in things in real time during the period 1973-2000, which again could be accused of being the 'cause' of the changes in violent crime.

Another thing which is a specific complication of the data we are talking about is that one is attempting to correlate something discrete (the number of violent crimes) with a surrogate for the quantitative amount of organic lead compounds in the atmosphere. If that correlation did indicate a causal relationship, it would imply that there is a large range of individual 'thresholds' for the the environmental lead level to trigger violent criminal activity two or three decades later. Although not impossible, that would perhaps be a little surprising - one might expect that it would be more likely that the ('quantitative') 'severity' of violent criminal behaviour, rather than the number of individuals committing violent crimes, would correlate with (quantitative) levels of lead in the environment (hence body).

Kind Regards, John

 

[JohnD]

If that correlation did indicate a causal relationship, it would imply that there is a large range of individual 'thresholds' for the the environmental lead level to trigger violent criminal activity two or three decades later. Although not impossible, that would perhaps be a little surprising - one might expect that it would be more likely that the ('quantitative') 'severity' of violent criminal behaviour, rather than the number of individuals committing violent crimes, would correlate with (quantitative) levels of lead in the environment (hence body).

Let's take it in pieces.

Is it believable that lead damages the brains of children and babies?

Is it believable that damage, affecting children's developing brains was highest where lead contamination was highest?

Is it believable that lead contamination was highest in city centres and districts with a lot of road traffic?

Is it believable that brain damage can lead to behavioural disorders and violence?

Is it believable that violent crime and murder rates were higher in city centres and districts with lots of road traffic?

Is it believable that violent crime rose when the brain-damaged babies and children grew to adulthood? And dropped when the supply of brain-damaged babies and children dropped?

Is it believable that this would happen, with the same time lag, in different US states and in different countries of the world, as different dates according to the date when leaded petrol was removed from sale?

Is it believable that alternative causes for the correlation have been investigated and disproved?

Is it believable that when you study the research, you will find that the above eight possibilities have already been researched and proven?

If so, we have not so much to contend over.

When a change in behaviour (e.g. criminality) occurs, it can be due to various things. The pattern and speed that it occurs gives strong clues as to the cause.

I'm thinking of the possibility that smoking cigarettes causes lung cancer. IIRC, the original study published in the Lancet some 60 years ago was based on two things: A survey of GPs to see which of them smoked; and subsequent research to see which of them had died of lung cancer. The correlation was convincing enough that most smoking GPs stopped. The tobacco and advertising trades both spent many years, and billions of dollars, trying to fight the truth. Last I heard, they were still winning in China and some other countries where the government has priorities other than the lives and health of its citizens. In the same way, there are still countries that permit leaded petrol (the Tetraethyl lead is profitably produced and exported from the UK by an American-owned company, which is another scandal. The company has previously been convicted of bribing foreign officials in impoverished nations).

 

[charliegolf]

So it's not about cars then?

 

[JohnD]

You don't think that death and ill-heath caused by cars is about cars?

 

[John D v2.0]

really we need a control group to do it scientifically to form a link. So expose a large enough group to lead and another not to lead in order to have enough to commit crimes to be statistically significant with whatever confidence you need. Then wait 20 years, and crunch the results. If there's a statistical difference, then there's a link, if not, then it's just a correlation. Assuming your confidence level is high enough.

 

[JohnD]

If you read the research, you'll find that the idea of confirming the hypothesis by deliberately damaging the brains of a cohort of babies has already been considered and discarded on ethical grounds. The days of Tuskegee are long gone, even in the US.

However there are already surprisingly detailed data on districts where lead pollution; lead content in the blood; and even brain function and IQ, were recorded, and can be matched to school performance and criminality in later life, so studies continue. A great deal of work has already been done on this matter, as you will see if you follow some of the links and read some of the papers.

 

[charliegolf]

You don't think that death and ill-heath caused by cars is about cars?

I don't think death and ill health caused by a fuel additive no longer added is about cars, no.

 

[JohnW2]

Is it believable that lead damages the brains of children and babies?
Is it believable that damage, affecting children's developing brains was highest where lead contamination was highest?
Is it believable that lead contamination was highest in city centres and districts with a lot of road traffic?
Is it believable that brain damage can lead to behavioural disorders and violence?

Yes x 3. None of that is particularly in dispute - although I'm not sure how strong is the evidence that it specifically results in violent behaviour (other than part of 'behaviour disorders' in general).

Is it believable that violent crime and murder rates were higher in city centres and districts with lots of road traffic?

I would imagine that violent crime and murder has always been higher in city centres, even long before leaded petrol existed.

Is it believable that violent crime rose when the brain-damaged babies and children grew to adulthood? And dropped when the supply of brain-damaged babies and children dropped?

Unless you know something that I don't, you're making assumptions here by talking about "brain-damaged babies and children". Is there evidence that the perpetrators of violent crime were more 'brain-damaged' than others? In any event, as I intimated before, the fact that not all babies born in the same environment (hence subjected to the same environmental toxins) turned into violent criminals illustrates the fact that other factors, probably many other factors, are involved in determining whether a particular baby turns into a violent criminal.

As I said before, if it were merely being suggested that environmental lead exposure was probably one of the many factors which, together, gave rise to behavioural disorders (some of which might be manifested as violent criminal activity) then I would have far less concerns. However, IMO the suggestion that it is, alone, a major factor in determining whether individuals become violent criminals is something very different, and far from proven.

Is it believable that this would happen, with the same time lag, in different US states and in different countries of the world, as different dates according to the date when leaded petrol was removed from sale?

I would need to look at data but, if that is the case, I agree that it would strengthen the circumstantial evidence.

Is it believable that alternative causes for the correlation have been investigated and disproved? .... Is it believable that when you study the research, you will find that the above eight possibilities have already been researched and proven?

I have no doubt that all these things have been investigated/researched. I am far less sure that anything has been "disproved" or "proved".

When a change in behaviour (e.g. criminality) occurs, it can be due to various things. The pattern and speed that it occurs gives strong clues as to the cause.

There I totally agree with you - the available data provides 'clues' as to what might, at least partially, be responsible for changes in the incidence violent crime. However, 'having a clue' (even a 'strong clue') is very different from having proved anything.

Kind Regards, John

 

[JohnD]

I would imagine that violent crime and murder has always been higher in city centres, even long before leaded petrol existed

Yes, you might imagine that. One of the surprising results turned up by the research I linked earlier, is that murder rates in large US cities used to be higher than in small ones. Astonishingly, twenty years after the withdrawal of leaded petrol, the rate declined and is also now about the same.

 

[JohnD]

... I am far less sure that anything has been "disproved" or "proved".

here's an example of what I mean by "disproved." The "broken window" extreme policing hypothesis also falls down when you look closely at the crime-fall graph, because the drop started before the policing policy did, and is also found in cities that did not adopt the policy.

...

"Abortion and birth control access
The famous abortion-crime hypothesis forwarded by Freakonomics is suspect because of two reasons. The first is that it's not strictly causal i.e. cause precedes effect (crime rates were going up before Roe v. Wade) and the second is that it's not internationally expandable. For example, the Abortion Act of 1967 in the United Kingdom, all but legalizing abortion, occurred well before before Roe but the UK had a surge in crime after the United States' peak (though only by a couple years). Contrariwise, Canada experienced tighter restrictions in legalized abortions from 1969-1988 but had a crime wave similar in duration, peak, and decline of that of the United States.[17]

A more plausible hypothesis is that greater availability of birth control in general may have caused the decline in the crime rates, and as more people used birth control in the late 1960s and 1970s, fewer unwanted children were born into unfortunate circumstances. This has the advantage of cutting across all countries and being an international trend. However, there remains a very large flaw in that crime rates were considerably lower in the 1950s than they are in modern times; the surge between the 1960s and 1990s and subsequent decline remains unexplained, and if lack of family planning was truly the cause, then crime rates should have been even higher before the mass availability of birth control in the United States. This is not the case. "

 

[JohnW2]

really we need a control group to do it scientifically to form a link. So expose a large enough group to lead and another not to lead in order to have enough to commit crimes to be statistically significant with whatever confidence you need. Then wait 20 years, and crunch the results. If there's a statistical difference, then there's a link, if not, then it's just a correlation. Assuming your confidence level is high enough.

Your tongue is obviously (hopefully!) very much in your cheek but, yes, that would be the 'proper' way to prove the causal relationship!

Other than when the causal relationship is extremely strong (**), attempting to 'prove' causal relationships from observational (rather than 'interventional', such as you describe) research can be next-to-impossible.
[** if all babies exposed to environmental lead, but none of those not exposed to environmental lead, eventually turned into violent criminals, then there would be no argument. However, when the number of violent criminals is a tiny proportion of even the lead-exposed population, it becomes extremely difficult to establish a causal relationship. Many major factors other than (in addition to) exposure to lead surely must be involved in determining whether a particular baby turns into a violent criminal.

Kind Regards, John

 

[JohnD]

... the fact that not all babies born in the same environment (hence subjected to the same environmental toxins) turned into violent criminals illustrates the fact that other factors, probably many other factors, are involved in determining whether a particular baby turns into a violent criminal....

in the same way that not all smokers die of lung cancer, thereby disproving the claim that smoking causes lung cancer?

 

[JohnW2]

in the same way that not all smokers die of lung cancer, thereby disproving the claim that smoking causes lung cancer?

Maybe I'm not making my point clearly enough.

We know that environmental lead can affect brains, that the brain damage can result in behavioural disorders and that those behavioural disorders can, in some cases, be manifested by committing violent crimes. No argument about any of that. Hence no disagreement that reducing exposure to lead will probably result in some reduction in violent crime.

What I am, as yet, far less convinced about is the suggestion/implication/claim that most/all of the reduction in US violent crime seen post-1973 is a consequence of the reduction in lead exposure during the preceding 2-3 decades. In view of the above, I am pretty sure that some of that decrease was probably attributable to the reduction in lead exposure, but I am, as yet, not convinced that most/all of the decrease was due to that reason.

Smoking is rather different. A very high proportion of (some types of) lung cancer is due to smoking and, indeed, the proportion of smokers who eventually develop lung cancer is high. However, the great majority of babies exposed to lead do not turn into violent criminals and, I would suggest, the majority of violent criminals are probably not violent criminals because they have been exposed to lead (we had violent criminals long before we had leaded petrol). For those reasons, studying lead exposure is much more difficult than studying smoking.

Kind Regards, John

 

[foxhole]

I saw a research program suggesting batteries will be discarded for high capacity capacitors, they can be charged in seconds and take up less room than batteries.