Back

Public Water Quality and Birth Outcomes: Evidence from the World's Largest Nitrate Removal Facility

Semprini, J.

2025-05-07 health economics
10.1101/2025.05.06.25327112 medRxiv
Show abstract

In response to rising nitrate levels in Iowas public drinking water, the city of Des Moines built the worlds largest nitrate removal facility. The facility operates when nitrate levels exceed the regulatory threshold of 10 mg/L, incurring costs exceeding $10,000 per day. To evaluate the effect of the nitrate removal facility on birth weight and gestational age, we analyzed publicly available birth certificate microdata (1992-2004). In addition to adjusting for maternal and fraternal sociodemographic factors, our linear regression model included county-year fixed-effects to adjust for annual differences across counties and year-month fixed-effects to account for seasonal variation in birth outcomes. Unbiased identification relied on exogenous geotemporal variation in first-trimester exposure to operation of the nitrate removal facility. Operating the nitrate removal facility was associated with a 0.4-percentage-point increase in the probability of normal birth weight and was associated with a 1.6-percentage-point increase in the probability of a normal term birth. The positive associations were largest among mothers with higher risk of adverse outcomes (smoked during pregnancy, cesarean deliveries, prior preterm births, excessive weight gain, unmarried, unknown father). A post-hoc analysis suggests that this intervention may be cost effective, multiplying public investment by 3.5 dollars in health benefits. As policymakers explore how to address rising nitrate pollution, quantifying the value of nitrate reduction strategies could inform future public health interventions. Although the Des Moines nitrate removal facility appears to serve a crucial public health function and advance health equity, challenges remain for deciding how to pay for this potentially cost-effective intervention.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.