On Balance: "Right Enough" Numbers for Air Pollution Policy
Exposure to air pollution continues to be a major health risk, including worsening health risks related to COVID-191. Thus, accounting for these benefits of these avoided health risks is critical in the evaluation of policies that focus on improving air quality and also play an important role in the anticipated climate policies, where improving air quality should be considered as a major co-benefit. However, compared to the scrutiny that has been given to the relationship between exposure to air pollution and adverse health effects, modeling the transport and fate of air pollutants from the emission source to the ambient concentrations to which we are exposed is often given more limited consideration in the modeling chain from emissions to monetary valuation for air pollutants.
The models – known as chemical transport models (CTMs) – are the generally in the realm of atmospheric scientists, engineers, and computer modelers, requiring specialized skills and computing resources to operate and interpret the output. This may hinder aspects of the evaluation of model performance, the identification of policy designs under uncertainty, and the use, especially by groups that may not have access to these specialized resources.
Reduced complexity modeling (RCMs) of air pollutants can play an important role in enhancing transparency and usability by reducing the time and barriers to analyzing alternative policy designs and allowing for the review of the benefits of policies by outside groups and stakeholders. To facilitate the access and use of the costs of air pollution, several research groups in the US have developed approaches to capture the main features of CTMs in RCMs. RCMs link the location and important source-specific information, namely height, of key emissions to their spatially distributed, ambient concentrations using simplified representations of what is included in a CTM. By reducing the computational complexity, RCMs are then better suited for the integrated assessment from emissions to their equivalent monetary damages. Furthermore, the marginal costs per tonne of emissions by location derived from the RCMs can be easily tabulated. The air pollution social cost datasets from three leading peer reviewed RCMs can found at the Center for Air, Climate, & Energy Solutions here.
But can RCMs produce results that can be relied upon for benefits analysis? To answer this question, my colleagues and I analyzed the social cost estimates from ambient concentrations of fine particulate matter (PM2.5), which is associated with premature mortality, generated by three publicly available RCMs that cover all counties in the United States: Air Pollution Emission Experiments and Policy (AP2), Estimating Air pollution Social Impacts Using Regression (EASIUR), and the Intervention Model for Air Pollution (InMAP). We concluded that uncertainties introduced by reduced complexity models are much smaller than other uncertainties in the valuation approach and unlikely to change conclusions of a benefit-cost analysis. These findings are especially clear at the national level when summing over pollutants and regions, although there can be more meaningful differences from some harder to model pollutants, like nitrogen oxides, at averaging over smaller areas. Overall, these three models are structurally different and yet produce similar results, which generates confidence in the RCMs. With the trade-offs in complexity come some limitations, mainly that RCMs generally provide annual (or at best seasonal) averages and are less robust when used for very large changes in emissions.
RCMs also open up new opportunities for the early evaluation of new policy designs that account for uncertainty in what we know about the relationships between emissions and ambient pollutants. As the RCMs are independent models, the range of the estimates can also be understood as a measure of uncertainty in the benefits. Looking at the range of benefits from the RCMs could enhance the evaluation of different spatially or source differentiated policy designs, illuminating both overall welfare and differences in the distribution of the benefits. Another advantage of reducing the time and computational burden is that the RCMs could also strengthen the ability to consider air quality related co-benefits over different climate designs without having to run a CTM, especially in the earlier stages of policy design.
Reliable estimates of externality costs are a critical part of regulatory analysis. RCMs, by design, cannot reproduce all of the information from CTMs as there are trade-offs in developing simplified representations. CTMs will continue to represent the state of the science and be the benchmark by which RCMs should be judged. However, by reducing the time, expense, and training involved with operating CTMs, they allow anyone to conduct integrated benefits analysis. Providing tools for economic analysis to increase involvement in making decisions about trade-offs between the costs and benefits of different rules under the Clean Air Act (CAA) can be part of increasing trust in the regulatory process. Facilitating the verification of the benefits and giving access to communities to evaluate issues related to distribution and equity may help reengage a commitment to real transparency.
While this piece has been reviewed by the co-authors for the ERL piece for accuracy, the opinions in this piece are those of the author and should not be attributed to her co-authors on the 2019 paper in Environmental Research Letters.
1DOI: 10.1126/sciadv.abd4049