Work Adaptations Insufficient to Address Growing Heat Risk for U.S. Agricultural Workers

Abstract

The over one million agricultural workers in the United States (U.S.) are amongst the populations most vulnerable to the health impacts of extreme heat. Climate change will further increase this vulnerability. Here we estimate the magnitude and spatial patterns of the growing heat exposure and health risk faced by U.S. crop workers and assess the effect of workplace adaptations on mitigating that risk. We find that the average number of days spent working in unsafe conditions will double by mid-century, and, without mitigation, triple by the end of it. Increases in rest time and the availability of climate-controlled recovery areas can eliminate this risk but could affect farm productivity, farm worker earnings, and/or labor costs much more than alternative measures. Safeguarding the health and well-being of U.S. crop workers will therefore require systemic change beyond the worker and workplace level.

Figure 1 –. Spatial distribution of agricultural workers and social vulnerability.

Number of summertime hired crop workers (MJJAS maximum of 2009-2018 average monthly values) as reported by the Bureau of Labor Statistics Quarterly Census of Employment and Wages (24). Counties outlined in dark green are in the upper quartile of the Center for Disease Control’s Social Vulnerability Index (41), indicating low community resilience to disaster events. Counties with no employment data are shown as missing values in gray.

Figure 2 –. Present-day and projected Heat Index extremes.

95^th-percentile of summertime (MJJAS) daily maximum Heat Index (°F) for a present-day observed (1979-2013), b projected with 2°C of global annual mean warming, and c projected with 4°C global annual mean warming (see Methods). Counties that contain no climate data grid centers are shown as missing values in gray. Color bar labels indicate the risk levels of the OSHA heat guidance for outdoor workers.

Figure 3 –. County and worker exposure to extreme heat levels and heat waves.

a,b 95^th-percentile of summertime (MJJAS) daily maximum Heat Index (°F) for present-day observed (gray), 2°C of global annual mean warming (orange), and 4°C of global annual mean warming (red) as a number of counties with ≥500 crop workers (%) and b cumulative number of workers (%); the percentage of c counties and d workers that experience a daily maximum HI for three or more days in a row that exceeds the present-day 95^th-percentile level – colors and distributions the same as in a,b. The numbers in each corner indicate the county- (a,c) and worker- (b,d) weighted average.

Figure 4 –. Present-day and projected worker exposure to unsafe heat levels.

a-c, number of days each summer (MJJAS) that the daily mean Heat Index exceeds the baseline Threshold Limit Value (83.4°F; see Methods) for a present-day observed (1979-2013), b projected with 2°C of global annual mean warming, and c projected with 4°C global annual mean warming (see Methods); d-f, as a-c but showing number of worker days, based on present-day crop worker employment levels (Fig. 1). Note the nonlinear color scale for the number of worker days. Counties with no employment data or that contain no climate data grid centers are shown as missing values in gray.

Figure 5 –. Effect of on-farm adaptation measures on reducing heat risk.

Median number of days per summer (MJJAS) that the average U.S. crop worker is exposed to a daily mean Heat Index higher than the TLV for their physical activity and clothing levels (values in parentheses; see Methods), for present-day observed (gray), 2°C of global annual mean warming (orange), and 4°C of global annual mean warming (red). The error bars indicate the 5^th and 95^th-percentile over the 35 years of observed and projected summers. The tested adaptation scenarios are: reduce effort (time worked/hour) from 90% to 50% (effort); reduce pace from moderate to light (pace); rest in AC instead of in the shade (rest area); wear single-layer instead of double-layer clothing (clothing); and combinations of these (see Table 1). When three or more adaptations are combined, worker exposure becomes (nearly) zero across all scenarios, so these are not plotted here.