Millions of U.S. water pipes still contain lead. Their corrosion can cause the metal to leach into drinking water, with long-lasting human health effects. Credit: Curie Duprat | Dreamstime.com.
Lead is a poster child for legacy pollutants: It won’t go away despite many efforts to reduce its impact on American lives. Leaded gasoline for motor vehicles began to be phased out in the 1970s, but small-aircraft fuel still contains lead, and airborne particles from past emissions can remain in the soil for years. Lead paint was banned in 1978 but is still found in old housing stock. Lead water pipes have been replaced in some areas, but an estimated 6 to 10 million lead pipes still exist today.
In parallel with ongoing lead mitigation efforts, our understanding of the metal’s impact on human capital, including health, education and productivity, also continues to expand.
“Previous research either linked higher blood lead levels with reduced learning outcomes in children and young adults or with lower cognitive abilities in older adults,” says AAE assistant professor Dustin Frye. “But we know much less about labor market outcomes for people in their 20s to 50s and wanted to estimate associations between childhood exposure to waterborne lead and adult employment.”
For this purpose, Frye and AAE affiliate faculty member Gisella Kagy, a fellow economist and assistant professor in the UW-Madison School of Human Ecology, turned to a rich source of data about Americans that has been collected every ten years since 1790: the U.S. census.
Recent technological advances allowed Frye and Kagy to follow people across space and time in multiple census waves, without revealing their identities. The researchers focused on the 1940 census because it was the first to include wages as part of labor market data. (Since few women were employed at the time, the researchers only analyzed men.) To examine the effect of childhood lead exposure on adult employment, Frye and Kagy merged the 1940 census with three previous census waves—1900, 1910 and 1920—when the 1940 adults were children.
Three pieces of data were critical for estimating childhood waterborne lead exposure: the town where the men grew up, its water chemistry and its pipe material. Chemistry matters because acidic water, with low pH and hardness levels, causes pipe corrosion and the leaching of lead into drinking water. The choice of lead—a more durable and malleable construction material than iron or steel—was mostly driven by population size, wealth and the proximity of lead mines. The health effects of lead in drinking water were unknown in the early 20th century.
The researchers obtained the material and construction year of water pipes from the 1897 Manual of American Water-Works and water chemistry data from a 1954 U.S. Geological Survey. They used geolocation and historical record linkage technology to create geographic identifiers for each respondent’s hometown and personal identifiers to track respondents and their family members across the four census waves of interest.
“We analyzed 6.2 million men of ages 20 to 50 years in 1940 who were exposed to various levels of lead-contaminated drinking water between the ages of 0 and 10,” says Kagy. “They grew up in one of roughly 1,000 towns in the 48 contiguous states for which we had complete water pipe and chemistry information.”
The researchers excluded self-reported rural residents who likely obtained their drinking water from private wells. The study’s sample size was substantially larger than earlier analyses of 7,400 adults and 5,500 U.S. Army service members.
More than 96% of the analyzed men were white; only 42% had completed high school and 3% had completed college. They reported an average of 11.4 years of education and an unemployment rate of 8%. Working men with higher exposure to waterborne lead as children, who grew up in towns with lead pipes and corrosive water, had an average 1940 income that was 3-5% lower than men with lower exposure. This corresponds to a $40 earnings gap in 1940, or roughly $1,000 in today’s dollars after adjusting for inflation.
Men with higher exposure were also nearly 5% more likely to be unemployed. If employed, they were more likely to work in lower-paying blue-collar occupations, meaning they classified themselves as laborers or operatives. Across ten different job categories, men with higher exposure consistently earned less money than those in the same category with lower exposure.
Some of the employment differences were due to lower educational attainment: Men with higher exposure had fewer years of completed education and were more likely to drop out of high school, meaning they did not meet the educational requirements of higher-paying professional positions. However, Kagy emphasizes that education alone did not fully explain the labor market outcomes.
“Higher lead levels have been associated with reduced cognitive abilities and impulse control in children and adults, and both factors play a role in the occupations people decide to pursue and the income they earn within that occupation,” says Kagy. “This is biologically plausible because environmental lead settles into bone, meaning it never leaves the body, and a ‘safe level of lead’ does not exist.”
Frye and Kagy hope to dive deeper into the reasons for the observed employment differences with more recent datasets. They have applied for grant funding to collaborate on this effort with the Federal Statistical Research Data Center at UW-Madison.
“Our study is a compelling demonstration that childhood lead exposure is associated with reduced economic wellbeing decades later, likely through a combination of educational and behavioral health channels,” says Frye. “Removing lead from gasoline and banning its use in paint were certainly steps in the right direction, but some sources of lead exposure—old homes, soil and water pipes—persist over long periods of time and require continued mitigation efforts.”