II. Lead Poisoning is the Master of Horror
“The association we observe may be one new to science or medicine and we must not dismiss it too light-heartedly as just too odd. As Sherlock Holmes advised Dr. Watson, ‘when you have eliminated the impossible, whatever remains, however improbable, must be the truth’.” (Hill, 1965)
In a 2015 article in The Atlantic, Ta-Nehisi Coates examined the causes of mass incarceration and its impact on black families, including the personal story of Odell Newton. That article featured lead poisoning in the dual role that Alfred Hitchcock often played in his classic horror films. There was a brief cameo appearance in a scene about Odell almost dying from lead poisoning as a young child, but the much larger role was behind the scenes, directing all the intertwined subplots of this article. Odell was poisoned before his grade school teacher said he should be placed in special education, before his impaired mental development became more obvious in junior high school, before he was convicted of murder at the age of 16, and before he spent 41 years in prison, where he failed in several attempts to pass the high school G.E.D. test. The cause and effect troubles of his life have been repeated in horror stories across centuries and around the world, all directed by lead poisoning.
The Atlantic later published a dissent by Hymowitz that chastised Coates for not recognizing that “black-family disruption could have some bearing on crime and incarceration”. She stated that 72% of black children are born to unwed mothers and “growing up in chaotic families … is itself highly correlated with the scourge of ghetto crime and incarceration”. Her suggested causal relationship, however, ignores an important indicator of causation: the suspected cause must precede the effect. She noted that black children were only slightly less likely than whites to grow up in two-parent homes before 1960, but “after 1960 … the family began to unravel.” If growing up in a single-parent household caused criminal behavior, then the 1960s increase in unwed birth rates should have increased crime starting in the mid-1970s, when children raised by single parents reached their teenage (juvenile offending) years. Crime and unwed birth rates increased in tandem after 1960 and have since fallen in tandem. Lead poisoning was the director of that temporal dance duet.
Coates referenced a circa-1970 memo from Daniel Patrick Moynihan acknowledging a “rather pronounced revival - in impeccably respectable circles - of the proposition that there is a difference in genetic potential” between races. In fact, those circles included The Atlantic, with its 1971 publication of “I.Q.” by Herrnstein, stating: “data on I.Q. and social-class differences show that we have been living with an inherited stratification of our society for some time.” That statement set the stage for the 1994 publication of The Bell Curve, by Herrnstein and Murray, and the angry debate over their claims that inherited IQ has a strong causal impact on incarceration, unwed births, and high school dropout rates. The Bell Curve is now ridiculed in “respectable circles” but its key findings were affirmed in a 1998 Scientific American article by Gottfredson, who had organized the 1994 treatise “Mainstream Science on Intelligence”, with 52 signatories defending The Bell Curve. The Scientific American website still displays her article, but it is mathematically impossible for the behavior risks she reported for youths with low IQ to be applicable today, after two decades of steep declines in high school dropout rates, youth incarceration, and unwed teen birth rates. The Bell Curve drama was directed by lead poisoning, and lead poisoning prevention deserves the credit for discrediting that book.
Another publication in 1994 went widely unnoticed during The Bell Curve debate over IQ and behavior. This study in the Journal of the American Medical Association reported that average blood lead levels for American children ages 1-5 fell by 77% from the late-1970s through the late-1980s, largely due to the USA phase out of leaded gasoline from the mid-1970s through the mid-1980s.
Coates highlights the significance of 1994 for a different reason, as the year when the federal government enacted a crime bill now blamed for driving up incarceration, and he rejects claims that this bill “was a purely well-intended, logical, and nonracist response to crime”. In fact, he rejects the causal relationship between crime and incarceration trends, based on a comparison of trends from 1960 to 1974, and from 1991 to 2012, and some research suggesting that mass incarceration is partly explained by an increase in the average length of prison sentences.
Coates did not mention that black male incarceration rates fell from 2001 to 2014 by 62% for ages 18-19, 51% for ages 20-24, and 46% for ages 25-29. Similarly, Hymowitz did not mention that unwed birth rates for black females fell from 1991 to 2014 by 79% for ages 15-17, 59% for ages 18-19, and 33% for ages 20-24. Few people are aware of these newsworthy trends because the news media is obsessed with the horror news genre.
During the 1994 media frenzy over The Bell Curve, as the 1994 crime bill was enacted, I was starting work on an Economic Analysis of lead paint hazard regulations for the Department of Housing and Urban Development (HUD). That analysis showed that the costs of lead hazard reduction were more than offset by the benefits from lead poisoning prevention affecting IQ, education, and lifetime earnings. My client also mentioned that we didn’t even count crime prevention benefits, suggested by recent research. In light of the strong relationship between blood lead and leaded gasoline use in the past, I wondered if there might be a relationship between crime trends and earlier gas lead trends. What I found was a stunning visual fit with a 23-year time lag, consistent with preschool lead exposure affecting the peak age of violent offending.
In 2000, Environmental Research published my peer-reviewed study on “How Lead Exposure Relates to Temporal Changes in IQ, Violent Crime, and Unwed Pregnancy”. The same journal published my 2007 study, “Understanding international crime trends: The legacy of preschool lead exposure”; and my 2009 study, “Trends in Preschool Lead Exposure, Mental Retardation, and Scholastic Achievement: Association or Causation?” My 2009 study also reported temporal shifts in incarceration by age and race, and showed that the lead poisoning research literature demonstrates all of the accepted indicators of causation. Lead exposure trends are not just correlated with important societal trends – lead poisoning caused those trends.
Coates notes that the rise and fall of crime was an “international phenomenon”, and calls this pattern a “riddle”. My 2007 study showed a consistent relationship between preschool lead exposure and crime trends in nine nations, explaining up to 91% of temporal variation in burglary rates, with an 18-year time lag; up to 89% of variation in robbery rates, with a 23-year time lag; and up to 95% of variation in overall “index” crime rates (violent plus nonviolent crimes), with a 19-year lag. Riddle solved.
Time lags that relate lead exposure and crime reflect analysis of a wide range of lags to identify the “best-fit” with the highest statistical significance. The best-fit lag in every crime category is consistent with lead-induced neurobehavioral damage in the first years of life and the peak age of offending for that category. Property crime arrests peak at ages 15 to 20 and fall sharply by age 30, consistent with the 18-year lag for burglary. Violent crime arrests peak at ages 15 to 24 and decline through age 50, consistent with the 23-year lag for robbery. Nonviolent crime is much more common than violent crime, consistent with the 19-year lag for index crime.
Coates does not mention that the rise and fall of crime occurred at different times in different nations, with seemingly inexplicable contrasts in crime rates across nations over time. In 1980, the USA index crime rate was 40% higher than Australia’s rate, but the 2001 USA index crime rate was 45% below Australia’s rate. Canada’s index crime rate was 60% higher than Britain’s rate in the early-1970s, but 20% lower in 2001. In 1974, the USA burglary rate was twice the rate in Australia and 50% higher than the rate in Britain, but the 2002 USA burglary rate was less than half of the rates in Britain and Australia. The Canadian robbery rate was five times the rate in Britain in 1962, but in 2002 the Canadian robbery rate was less than half the rate in Britain.
Crime in the USA and Canada rose earlier than in other nations because the USA and Canada accounted for the vast majority of global use of lead in gasoline before 1970, as many other nations recovered slowly from World War II. The USA and Canada phased out the use of lead in gas from the mid-1970s through the mid-1980s, but leaded gas use in many industrial nations was near its peak in the mid-1980s, delaying crime declines in those nations. Within each nation, however, crime trends have followed lead exposure trends with the same time lag.
Coates presents a graph showing USA violent crime surged after 1960, but incarceration didn’t start rising until after 1970, and he concludes that incarceration “rose independent of crime”. Hymowitz claims this lag is better explained by policy that was “slow to catch up in the 1960s as crime was on the rise”. Actually, that lag is best explained by the age groups that dominated the 1960s violent crime rise, and one key factor that neither Coates nor Hymowitz considered.
Arrests by age show that offenders under the age of 25 accounted for 77% of the total increase in violent crime arrests from 1960 to 1969. The outsized youth impact was partly due to baby boomers swelling the age groups when offending peaks, but violent crime arrest rates for youths (arrests per 100,000 in age group) also increased over the 1960s by 83% for juveniles (under age 18) and 42% for ages 18-24. The incarceration rate didn’t rise until the 1970s because first-time offenders are more likely to get probation instead of prison time, and juveniles with serious and/or multiple offenses went to juvenile detention facilities (not included in the prison data cited by Coates), and because of one other important factor affecting incarceration in the 1960s: the war in Vietnam.
There were 552,000 troops overseas in 1960, and troops overseas peaked at 1.1 million in 1968. There were 205,000 men in prison in 1960, and the number of men in prison fell to a low of 182,000 in 1968. Judges in this era were known to give young male defendants a choice between prison and “volunteering” for military service. Enlistment was the option chosen by many youths who expected to be drafted soon even if they served a short prison term. Troops overseas fell to 885,000 in 1970, 687,000 in 1971, and 507,000 in 1972. The draft ended in 1973, and the number of men in prison reached a new record high in 1974, and in every year for the next three decades.
The violent crime rate fell more than 50% from its 1991 peak to 2014, but Coates notes that the incarceration rate didn’t peak until 2007, and fell less than 8% from 2007 through 2012. Those trends reflect a shift in arrest rates by age that is a mirror image of arrest trends over the1960s. From 1991 to 2014, violent crime arrest rates fell 65% for juveniles and 50% for ages 18-24, but increased 21% for adults ages 50-54. The same generation that caused a youth crime surge in the 1960s has caused a large increase in the over age 50 arrest rate since 1991. Moreover, surging youth crime at the start of a crime wave, and plummeting youth arrests presaging the end of a crime wave, is a pattern that has been repeated around the world since the circa-1800 dawn of “juvenile delinquency” (trends documented in Chapter V).
Incarceration rates by age have not moved independent of arrests by age since the 1990s. From 2001 to 2014, male incarceration rates fell 62% for ages 18-19, 38% for ages 20-24, and 27% for ages 25-29, but rose 31% for ages 40-44. Data reported for older ages since 2007 show male incarceration rates rose from 2007 to 2014 by 50% for ages 50-54 and 57% for ages 55-64.
Coates faults longer sentences for mass incarceration, and Pew Trust research confirms that inmates released from state prisons in 2009 did serve longer sentences on average than those released in 1990: 2.2 years for drug crimes, up from 1.6 years in 1990; 2.3 years for property crimes, up from 1.8 years in 1990; and 5 years for violent crimes, up from 3.7 years in 1990. An increase of a half-year in average time served for nonviolent crimes can hardly explain mass incarceration. Moreover, the shift in arrests to older offenders suggests that average sentence length has increased because repeat offenders now account for a much larger share of crimes. Studies of recidivism by state prisoners released in 1983, 1994, and 2005 all report high reoffending rates, but very different age distributions for inmates released. In 1983, 36% of state prisoners released were under 25 and 9% were age 40 or older. In 2005, just 17% of prisoners released were under 25 and 32% were 40 or older.
All horror stories require some suspension of disbelief, but for this story about the horrors of lead poisoning you only need to suspend judgment until you have seen all the evidence. That evidence is presented here in the context of indicators of causation delineated by Sir Austin Bradford Hill in his 1965 treatise “The Environment and Disease: Association or Causation?” The World Health Organization has recognized Hill’s insights as a “public health classic” and a “mainstay of epidemiological textbooks and data interpretation”. The causation indicators identified by Hill are: strength, consistency, experimental evidence, dose-response, biological plausibility, time-precedence, coherence, specificity, and analogy.
 Hill, AB. (1965) The environment and disease: association or causation? Proc R Soc Med. 58:295–300
 Coates, Ta-Nehisi (October 2015) The Black Family in the Age of Mass Incarceration, The Atlantic
 Hymowitz, Kay (October 4, 2015) The Breakdown of the Black Family, The Atlantic
 Herrnstein, Richard (September, 1971) I.Q., The Atlantic
 Herrnstein, R., & Murray, C. (1994) The Bell Curve: Intelligence and Class Structure in American Life. The Free Press, New York.
 Gottfredson L. (1998) The general intelligence factor, Scientific American, 9, 24-29
 Gottfredson L. (December 13, 1994) Mainstream Science on Intelligence, The Wall Street Journal
 Pirkle, J., et.al. (1994) The decline in blood lead levels in the United States, Journal of the American Medical Association, 272, 284-91
 Bureau of Justice Statistics (2001-2014) Prisoners series
 National Center for Health Statistics (December 23, 2015) Births: Final Data for 2014
 Department of Housing and Urban Development (1999) Economic Assessment of the Final Rule on Lead-Based Paint
 Nevin, Rick (2000) How lead exposure relates to temporal changes in IQ, violent crime, and unwed pregnancy, Environmental Research, 83, 1-22.
 Nevin, Rick (2007) Understanding international crime trends: the legacy of preschool lead exposure, Environmental Research, 104, 315-336
 Nevin, Rick (2009) Trends in preschool lead exposure, mental retardation, and scholastic achievement: Association or causation? Environmental Research, 109, 301-310
 FBI Uniform Crime Reports (1960-2014), index crime arrests by age, compiled by youthfacts.org
 Kane, T. (October 27, 2004) Global U.S. Troop Deployment, 1950-2003, Heritage.org
 University at Albany (August 28, 2013) Prisoners - number and rate by sex, trend data back to 1925, Sourcebook of criminal justice statistics
 Pew Center on the States (June 2012) Time Served: The High Cost, Low Return of Longer Prison Terms, Pew Charitable Trusts
 Bureau of Justice Statistics (1989; 2002; 2014) Recidivism series
 Lucas, R. and McMichael, A. (October 2005) Association or causation: evaluating links between “environment and disease”, Bulletin of the World Health Organization