Environmental Lead and Violent Crime

mentalfloss

Prickly Curmudgeon Smiter
Jun 28, 2010
39,817
471
83
Could I have the Readers Digest version?

There are too many subfactors between lead exposure and lower crime to show the causal relationship at this point so more research is needed.

One criticism is that the symptoms of exposure would be more acute and prevalent than the lag in the graph shows. In other words, if exposure definitely was the primary cause, the drop in crime should be more immediate and it would be obvious that a crapload of people were suffering the symptoms.
 

Tonington

Hall of Fame Member
Oct 27, 2006
15,441
150
63
There are too many subfactors between lead exposure and lower crime to show the causal relationship at this point so more research is needed.

You know though, that the same arguments were made for the association between smoking and cancer rates?

A common set of criteria exist in biology for assessing if the evidence for causation is adequate. It's called the Bradford Hill criteria, a simplified version of which I'll list below:

  1. Strength of the association. How large is the effect?
  2. The consistency of the association. Has the same association been observed by others, in different populations, using a different method?
  3. Specificity. Does altering only the cause alter the effect?
  4. Temporal relationship. Does the cause precede the effect?
  5. Biological gradient. Is there a dose response?
  6. Biological plausibility. Does it make sense?
  7. Coherence. Does the evidence fit with what is known regarding the natural history and biology of the outcome?
  8. Experimental evidence. Are there any clinical studies supporting the association?
  9. Reasoning by analogy. Is the observed association supported by similar associations?

A good number of those criteria are supported by the evidence that Kevin Drum has presented. To the points you mentioned from Pinker, specifically the lag and the shape of the distribution of age of criminals, that`s fine. But nonetheless the lag is from peak to peak, and the association isn`t perfect. So I wouldn`t expect the age distribution to perfectly align anyways. Lastly, of course this hypothesis doesn`t discount other factors involved in crime, and nobody should be interpreting it that way.