Fall in crime linked to lead in petrol/gas ...

This issue is not going away, and I guess it’s because no one has yet come up with a comprehensive explanation as to why crime has been falling everywhere for 20 years or so. I guess we’re still talking about this because the correlation is tight, and consistent across different societies, social policies and timelines:

Interesting graph on the US experience about 2/3 down the page:

The counterview is pretty compelling:

I guess he’s talking about poverty, life opportunities, etc

I notice also there’s a new scientific paper discussing high levels of lead in ancient Rome’s plumbing …
Question: But for the cross-societal, cross-policy confirmation, you’d think the correlation a coincidence … what’s going on here?

The article seems to suggest a twenty year time lag between exposure to lead, and crime. Lead levels were almost at their lowest in 1970, and crime rates almost at their highest in 1990. It would also be interesting to know if criminals have higher levels of lead in their system than others.

The correlation is interesting, but not compelling evidence of causality, as I think we agree.


Are we looking at the same graph? It pretty clearly shows that gasoline lead peaked around 1970 and violent crime peaked around 1990.

When I read this study a while ago, I seem to recall that is showed that violent criminals tended to have grown up in envioronments that had higher levels of lead contamination; e.g. old building, or locations proximal to highways. I don’t recall if they tested individuals for lead.

No, you are right. I misread the graph.


Previous thread discussing this:

Thanks for that, an interesting read.

The timeline variations, and the high vs. low (lead) State, variations paint a pretty compelling picture.

I also like the idea about decreased testosterone.

Well if the theory is that higher lead exposures causes bad decision making, the criminals wouldn’t necessarily have higher lead in their systems; non-criminals may have made bad decisions in non-criminal ways. Maybe if someone also charted bankruptcy or teen pregnancy we should expect to see correlation as well?

I thought the book Freakonomics conclusively proved that the wider availability of abortion beginning in the 1970s caused the crime rate to start dropping roughly a generation later. (Yes, I’m exaggerating.)

Isn’t there a lot of stuff we can correlate to the trend in crime? What about CFCs, for example? Or heavy metal hair bands from LA?

One reason the lead correlation is so interesting and compelling is that they looked at these rates in different countries. The correlation follows the same curve in countries that phased out lead at different times.

I would just like to point out that heavy metal hair bands from LA hit the scene in the US earlier, and it took some time for them to catch on in other countries – meaning that Motley Crue isn’t off the hook yet in my book.

and with quite different social policy ideas on policing and criminalising.

I still like the falling testosterone idea:

If you could separate out the other factors. Which would be just as difficult as for crime. I don’t know if it would be possible to separate out populations into bad-decision-makers and non-bad-decision-makers in general.

But perhaps they could compare those who went bankrupt, with criminals, and see if they had matching lead levels, or damage from lead poisoning.


Low testosterone makes a man more likely to commit crimes?! I would’ve guessed the reverse, considering what “'roid rage” does.

I think you read it wrongly.

I thought the problem with a bunch of these explanations is that it doesn’t describe what actually happened. That it isn’t the case that a younger, non-violent generation grew up and displaced their violent elder statesmen. What happened is that everyone stopped being so violent.

Here are a few more:

Changing demographics - less kids between 15 and 23 - the crime years

The availability of video games

Crime displaced to the internet

Changes in crime reporting, insurance policy terms

Mass market electronics getting cheaper, less high value stuff to steal

Better car security

Here is one that is a bit off the wall: it is all due to our police getting better at catching criminals and our politicians passing laws to put more bad people in jail so they cant commit crime.

Take your pick!

My theory is that it is the rise of the Internet. What was street crime has now moved into social media and scams. Robbing a store has now become robbing someone on ebay. A lot of this crime is simply not recorded.

Crime hasn’t gone away, it has simply moved.

I’m sure none of those things occurred to the researchers. You might want to write to them.

Many things have been suggested as leading to an increase or decrease in crime:


The chart attached to that article was violent crimes. Only your first list item would have anything to do with that, I would think. And the story talked about this happening across different countries, so it doesn’t have to do with better police or longer sentences because both those vary from country to country.

Yeah, but I thought it was pretty well established that crime is a young man’s game. Old criminals are either dead, retired, or in jail.