The pollution-haven hypothesis states that increasing stringency of environmental regulations in developed nations will lead to outsourcing of polluting production to less developed nations and cause environmental and health harms in those nations. A working article from the National Bureau of Economic Research by Tanaka et al. attempts to provide empirical evidence to prove the pollution-haven hypothesis. In this case, the US tightened environmental regulations and the authors hypothesize that Mexico became a pollution haven. The research seeks to identify a causal relationship between a change in US air-quality standards for lead and A) changes in battery-recycling productivity in the US and Mexico and thereby B) changes in birth weight in infants whose mothers live near battery recycling plants in Mexico.
This study provides novel, concrete evidence of an international pollution haven effect. While similar phenomena have been documented domestically as part of the environmental justice movement, evidence on an international scale to date has failed to provide empirical evidence for international effects.
The study employs a difference-in-differences technique to estimate the treatment effect of a 2009 policy change on five response variables: US ambient lead concentrations, US exports of used acid-lead batteries (ULABs), growth of the Mexican battery-recycling industry, and incidence of low birth weights in Mexico. The authors document a decrease in ambient lead concentrations in the US after a 2009 change in National Ambient Air Quality Standards (NAAQS). Due to data constraints they cannot document a change in ambient lead concentrations in Mexico. Instead, the authors find an increase in US exports of ULABS, and an increase in value-added and gross growth in the Mexican battery recycling industry. Finally, they look at the incidence of low birth weight by mothers living within two miles of battery recycling plants in Mexico, hypothesizing that incidence will increase after the 2009 regulation. The authors claim that the incidence of low birth weight increases, although this particular finding is not strongly supported and warrants further review.