Wildfire and infant health: a geospatial approach to estimating the health impacts of wildfire smoke exposure

ABSTRACT

We estimate the effects of wildfire smoke exposure on infant health. Exposure to wildfire smoke is determined using the latitude and longitude coordinates corresponding to each infant’s home address and a fine-scaled spatial dataset of wildfire smoke plumes constructed in GIS from satellite images of the landscape. Using a difference-in-differences estimation strategy, model estimates show that exposure to wildfire smoke leads to a .034 increase in the probability of low birthweight.

KEYWORDS:

• Wildfire Smoke
• Infant Health
• Geographic Information Systems

JEL CLASSIFICATION:

• Q53
• Q54
• I12
• I18

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ Please see Appendix A.1 for a discussion of related works.

² http://www.geomac.gov/index.shtml. MTBS: http://www.mtbs.gov/.

³ http://www.nifc.gov/.

⁴ http://www.ssec.wisc.edu/.

⁵ Please see Appendix A.2 for a discussion of alternative data sources and the limitations of our methodology.

⁶ Please see Appendix A.3 for more details regarding the data utilized in our study.

⁷ As shown in Table A1, our model estimates are qualitatively unaffected by this sampling restriction.

⁸ Infants with a gestational age greater than 42 weeks are dropped (16 observations).

⁹ We use each infant’s birth date and gestational age to determine each infant’s conception date. Please refer to Currie and Rossin-Slater (2013) for a detailed discussion of the importance of this classification strategy.

¹⁰ Like Currie and Rossin-Slater (2013), we compute trimester of exposure by counting forward from conception dates to fire dates.

¹¹ We also fail to reject the null hypothesis that this coefficient equals zero at the .1 level of significance.

¹² exp(−.0396) – 1 = −.038.

¹³ https://www.ospo.noaa.gov/Products/land/hms.html.

¹⁴ The data described in this section can be accessed here: https://tinyurl.com/twya4mc.

Additional information

Funding

Support for this research was provided by the National Science Foundation, NSF DEB 1115068.