The unequal effects of pollution on labor supply
What is this research about and why did you do it?
Air pollution is one of the largest environmental risks to health with approximately 7 million lives lost to ambient air pollution in a single year, according to the World Health Organization (WHO). A vast medical and economics literature has documented the causal negative effects of pollution on mental health, physical health, and productivity. Can workers avoid the damaging effects of air pollution? On high-pollution days workers may face a trade-off between exposure to a harmful environment and income, as performing their usual income-generating activities may increase their exposure to pollution. This trade-off is particularly acute for low-income workers, whose income is closely linked to the daily number of hours worked.
How did you answer this question?
Many cities in low- and middle-income countries frequently experience days with levels of fine particulate matter (PM2.5) pollution that substantially exceed the WHO’s guidelines. We collected daily information on PM2.5levels in Mexico City and matched it to daily working hours for are presentative sample of workers between 2005 and 2016, resulting in more than2,000,000 person-days observations. Our data contained individual information on days worked and hours worked per day during a reference week, usual income, formality status, sector of employment and demographics. The richness of the data allowed us to estimate the negative effects of extreme pollution levels on labour supply, explore whether workers compensate by re-allocating their working hours to days following high pollution and to compare the responses of high-income and low-income workers.
What did you find?
Our findings show that air pollution reduces working hours, with large effects at high levels of pollution. Using the daily number of hourswith PM2.5 above each WHO threshold, we show that the average number of hours worked decreases non-linearly with air pollution levels (see Figure 1). On days in which pollution exceeds the most extreme threshold, the number of hours worked is reduced by about 7.5%.
Notes: Coefficients and 90% confidence intervals are plotted from a regression on daily hours worked on the number of hours above the WHO air quality threshold for PM 2.5. A separate regression is run for each threshold.
We also found evidence of dynamic labour supply adjustments: daily labor supply is a function of the current day’s pollution level and of the prior five days’ pollution levels. Specifically, workers reduce their labor supply in response to same-day pollution and partially compensate for hours of work lost due to high pollution by increasing their labor supply on subsequent days.
Notes: Coefficients and 90% confidence intervals are plotted from a regression on daily hours worked on the number of hours above the highest WHO air quality threshold for PM 2.5 on the same day and on the previous 5 days.
What implications does this have for the study (research and teaching) of wealth concentration or economic inequality?
Labor markets in urban areas of- low- and middle-income countries are characterized by high levels of informality, productivity dispersion as modern sectors coexist with traditional technologies, and high levels of inequality along many dimensions. Using data on workers’ usual income, we analyse whether workers in the top 10% and the bottom10% of the income distribution respond in similar ways. First, we find that all workers, regardless of their income reduce their hours worked on days with high pollution. Second, this response becomes stronger as we move up the income distribution from low-income workers to high-income workers.
Notes: Coefficients and 90% confidence intervals are plotted from a regression on daily hours worked on the number of hours above the WHO air quality threshold for PM 2.5. A separate regression is run for workers in the top and bottom 10% income level and for each threshold.
Our findings show that low-income workers, regardless of whether they are informal or self-employed, reduce working hours on highly polluted days by substantially less than their high-income counterparts.
What are the next steps in your agenda?
Our results suggest that workers’ ability to adjust labor supply differs across the income distribution, implying that the costs of air pollution are unequally distributed across workers. This could explain, at least in part, why poorer people tend to suffer greater health consequences of extreme levels of air pollution. Our next research project explores the non-linear effects of air pollution on health to understand the effects of extreme pollution across age groups and socio-economic status.
Citation and related resources
Hoffman, B., and Rud, J. P. (2024) "The unequal effects of pollution on labor supply", Econometrica, Jul 2024, Vol 92, No. 4, pp1063-1096
Podcast, 'The unequal effects of pollution on daily labour supply', Bridget Hoffman and Juan Pablo Rud
