Published on Development Impact

Tiny plastics, big problem

This page in:
Development Impact logo

I used to think the human health costs of marine plastic were like the health benefits of flossing teeth: probably real, but unquantified. The Great Pacific Garbage Patch sounds bad for marine life, and a future ocean with more plastic than fish is worth avoiding. But people don’t live in the ocean. Potential harms remain active research areas, such as consequences from eating microplastic-filled fish, or whether plastic pollution means fewer fish available for human consumption. I reviewed marine plastic pollution three years ago and the best evidence I could find was estimates of the costs of removing plastic debris from beaches.

That’s why I was so excited to read the new working paper by Xinming Du, Shan Zhang, and Eric Zou. In “Marine Microplastics and Infant Health”, they combine satellite estimates of microplastic concentrations with variation in ocean currents. They estimate 205,800 more births that are low weight each year because of microplastic exposure. But their proposed mechanism is not the one I expected. Rather than pregnant women eating microplastic-filled fish, they think their effect occurs though inhalation of microplastic particles.

The idea of using variation in ocean currents is like obtaining identification from changes in wind direction. Some of those papers model the movement of discrete particles to calculate air pollution transportation; Du, Zhang, and Zou do the same for microplastics in the ocean. This allows them to isolate microplastics originating more than 200 km from a country’s coast, which are likely more exogenous to own-country birth weights than nearshore pollution.

The treatment data on microplastic concentrations were new to me. They key idea underpinning the data is the physical relationship between wind and ocean surface roughness. Higher wind speeds make the surface of the ocean choppier. But microplastics dampen this relationship, making the ocean surface appear smoother to satellite-based radar than we would otherwise expect given the prevailing wind speed. The data are high resolution – daily for each 0.25 degree grid cell – and available for the globe from April 2017 to September 2018. The existence of these data are yet another instance of physical scientists permitting economists to provide evidence on unanswered research questions.

The outcome data are individual birth events from 15 countries in the Americas, Africa, and Asia, and the outcome variable is an indicator that equals 1 for babies that weigh less than 2,500 grams. Limiting the sample to births within 200 km of the coast or in coastal localities, Du, Zhang, and Zou define in-utero exposure based on log microplastic concentrations in the nearest ocean grid cell.

Exposure during the second and third trimester significantly increases the probability of low birth weight events, whereas the placebo tests of exposure before conception or after birth have no effect (Figure 4). A 1% increase in microplastic exposure over an entire pregnancy increases the probability of a low birth weight event by approximately 0.02% (Table 1). This elasticity is about one order of magnitude smaller than elasticities of low birth weight incidence with respect to air pollution (Section 4.2).

The paper becomes somewhat speculative when it moves to explaining how microplastic exposure could affect birth weights. The mechanism I expected – pregnant women eating seafood contaminated with microplastics – has no empirical support, though this could be due to the authors not measuring seafood consumption at the individual level or not accounting for the geographic difference between seafood production and consumption.

Instead, the authors make an indirect argument in favor of “aerosolization”. As plastic disintegrates, it can break into pieces so small that they become airborne. This aerosolization may be more likely when evaporation rates are high, because evaporation promotes rising air. The authors proxy for aerosolization with aerosol optical depth, a commonly used measure of air pollution. They find that microplastic concentrations are positively associated with aerosol optical depth, and this relationship is stronger when evaporation rates are higher. However, they do not estimate the effect of aerosolization on low birth weights.

I’m convinced by the main result of this paper, even though the mechanisms section could be improved. Microplastics in the ocean causally increase the probability a baby is born with a low weight. This makes plastic pollution a public health problem because low birth weight predicts many negative life outcomes, from greater disease risk to lower lifetime earnings. I now believe that initiatives like the Intergovernmental Negotiating Committee on Plastic Pollution are more important for human well-being than the advice of the American Dental Association. Reducing the millions of tons of plastic that enter the ocean every year does not just reflect an aesthetic preference or a problem for marine wildlife: marine microplastic pollution significantly harms human health.


Gabriel Englander

Economist, Development Research Group, World Bank

Join the Conversation

The content of this field is kept private and will not be shown publicly
Remaining characters: 1000