Wednesday, October 31, 2007

Stretching it

I am very skeptical about this. Researchers claim that violent crime is linked to consumption of lead.

Among the most startling findings:

* A pair of studies by economist Rick Nevin suggests the nation's violent-crime rate in the second half of the 20th century is closely tied to the widespread consumption of leaded gasoline.
* According to Nevin, the trend lines match almost perfectly -- Leaded-gas use climbed in the 1940s and fell in the early 1970s; 23 years later, rates for violent crime followed in near unison.
* Its gradual demise in the 1970s, he says, did more to stop violent crime among people who came of age in its wake than any social policy.


So why is violent crime more predominant amongst men? Why is it more predominant amongst lower socio-economic groups? And why is it once more on the rise in the US?

And what the heck happened HERE?

The researchers "stunning trendline" also matches the rise of welfare and fatherless families and the increasingly liberal justice approach, causes which have far more plausibility.

13 comments:

mojo said...

I know, Lindsay ... personally I blame the resurgence in violent crime, & somehow I recall it increasing in the early 70's (& certainly youth violence in NZ has risen according to the latest figures) on the La Leche league and the increase in breast feeding and the alienation caused by abrupt weaning.
One could of course expect a similar increase given stricter enforced criteria on benefits.
The lead theory was first mooted in the UK some years ago (1970's) in relation to inmates and lead poisoning ... similar correlation study. It was found then that the lead content in these chappies came from somewhat outdated plumbing in their halls of residence ... green gums, I believe.

Anonymous said...

Crime rose as the percentage of youth in the population rose and it declined as the babyboomers left behind adolescence and became adults. Much crime is related to adolescence. There are many factors involved in crime rates and it usually wrong to attribute it to any one item.

For instance the "broken window" Right tried to pretend that arresting people on petty offences reduced crime. Yet crime rates declined in other cities as well where such policies were not followed. It is dangerous to be too simplistic.

Lindsay Mitchell said...

Anon, I entirely agree that there are multiple factors at play. Some more important than others. But I do put the breakdown of the family unit pretty high on the list.

Anonymous said...

I blame the rise of all these anti social indicators on my advancing age.

Eric Crampton said...

Steve Levitt discusses the evidence on lead and crime over at the Freakonomics blog:
http://freakonomics.blogs.nytimes.com/2007/10/30/did-banning-lead-lower-crime/
Upshot: banning leaded gasoline seems to have had an effect on crime.

He cites a paper by Jessica Wolpaw Reyes that exploits variation across US states over time as different states moved more or less quickly to reduce and eliminate lead from gasoline.

This kind of study allows us to isolate the effects of lead from other things that might be going on at the same time. State-level fixed effects pull out things that are idiosyncratic to particular states; year effects pull out time trends.

Why is violent crime more predominant among men and lower socio-economic groups? Reasons unrelated to the study. All else equal, more lead exposure increases the likelihood that someone will engage in criminal activity.

Levitt is skeptical about the Nevin studies, but recommends the one I've linked above. Moving from a time series to a panel analysis really makes it a lot more convincing. For lead then to just be proxying for other things going on, like family deterioration or welfare, you'd have to say that the pattern across states and over time in those things matches the pattern of lead reduction across states and over time, which seems far more implausible than if we were just looking at a national time trend.

Eric Crampton said...

Ahem. The link I'd provided was truncated. Let's try it again. Freakonomics link.

Anonymous said...

Removal of lead from petrol, despite what you may have been told, had very little to do with the issue of lead getting into the environment. When car emission standards were being tightened, leading the people needing catalytic converters, one development company of the catalytic convertes was finding that their catalysts were being poisoned by the petrol lead, leading them to cease working. They therefore lobbied to have lead removed from petrol, citing the environmental benefits, and not making public that their motives were more or less to enable their catalytic converters to work.

The amount of lead in petrol was actually very, very low - it was added in the form of tetra-ethyl lead, in very small quantities (2mL / US gallon, of which lead makes up only a proportion of the 2 mL)

I also understand that lead is used to vulcanise rubber... and that this is the source of more lead into the environment than there was from the leaded-petrol!

Note: Therefire, the rise in violent crime with the rise in petrol consumption (if it were real) may actually reflect the increase in rubber usage from tyres, rather than the increase in petrol usage.

Anonymous said...

Removal of lead from petrol, despite what you may have been told, had very little to do with the issue of lead getting into the environment. When car emission standards were being tightened, leading the people needing catalytic converters, one development company of the catalytic convertes was finding that their catalysts were being poisoned by the petrol lead, leading them to cease working. They therefore lobbied to have lead removed from petrol, citing the environmental benefits, and not making public that their motives were more or less to enable their catalytic converters to work.

The amount of lead in petrol was actually very, very low - it was added in the form of tetra-ethyl lead, in very small quantities (2mL / US gallon, of which lead makes up only a proportion of the 2 mL)

I also understand that lead is used to vulcanise rubber... and that this is the source of more lead into the environment than there was from the leaded-petrol!

Note: Therefire, the rise in violent crime with the rise in petrol consumption (if it were real) may actually reflect the increase in rubber usage from tyres, rather than the increase in petrol usage.

Eric Crampton said...

Spam:
Regardless of the reasons for lead removal, it looks like its removal correlates with reductions in crime. I know rather little about sources of environmental lead, but Wolpers Reyes cites studies showing lead from gasoline being the largest contributor to environmental lead exposure in the early 1970s, with tightening of restrictions correlating with drops in measured blood lead levels.

I'd be rather surprised if lead used in tyre construction were as substantial a contributor to environmental lead exposure as was petrol: the lead in petrol will come out of the tailpipe while the lead in tyres will stay in the tyres, barring the occasional burn-out.

Lindsay Mitchell said...

Eric, What does this mean?

"The elasticity of violent crime with respect to lead is estimated to be approximately 0.8 " ?

I am assuming it is a measurement of how much violent crime levels are effected by ingestion of lead. Is that low?

Eric Crampton said...

Lindsay,

An elasticity of 0.8 means that a 1% decrease in the amount of lead exposure correlates with a 0.8% reduction in crime.

In general, elasticity always means the percent change in something given a one percent change in something else. In general, we say that something is relatively inelastic if the elasticity is less than one, and relatively elastic if it's greater than one.

In this case, 0.8 is pretty substantial. Given the large percentage reduction in lead exposure, she's estimating that lead reduction cut violent crime by more than half -- this is very big compared to other things. The best evidence I've ever seen on the elasticity of crime with respect to the number of police officers is that a 1% increase in police numbers results in a 0.33% reduction in crime.

mojo said...

Lindasy ... if you are beginning to ascribe to this sort of stuff I can see why some would want to arrest you.

Anonymous said...

Something else climbed in the 40s and peaked in the 1970s, and that's the greatest modern diaspora of all.. the shift of Maori from the hinterland to the cities.

You can certainly trace an upsurge in crime to this massive social upheaval.

JC