The reliability and robustness of a spatial microsimulation (SMS) method for providing insights to carbon tax consequences

A new article published in the Journal of the Association of Environmental and Resource Economists evaluates a nuanced approach to calculating the incidence of subnational carbon tax policies.

In “Spatial microsimulation of carbon tax incidence: An application to Washington State,” authors Nathan W. Chan and Susan Stratton Sayre argue that standard approaches for calculating the incidence of subnational carbon tax policies are prone to inaccuracy due to coarse aggregation. As an alternative, the authors evaluate a spatial microsimulation (SMS) method that generates granular household-level incidence estimates. This approach provides unique insights into the distributional consequences of carbon taxes, including across geographies.

In the paper, the authors demonstrate the SMS method for a recent carbon tax initiative in Washington State and counterfactual variations on its revenue recycling provisions. Drawing from multiple data sources, they construct a synthetic population for the state and project carbon consumption for each household. After generating distributions of household impacts at the census tract level, they investigate how counterfactual variations on the targeting provision would have altered incidence patterns across and within income groups and jurisdictions.

Comparing across counterfactuals, the authors pinpoint how different targeting provisions in revenue recycling designs will have disparate consequences for the policy package’s progressivity/regressivity and the geographic distribution of incidence. They analyze and discuss potential implications for political economy analysis of carbon taxes and benchmark the SMS approach by investigating its sensitivity and robustness to modeling choices and comparing its performance to an alternate approach for generating geographically specific estimates. SMS provides more logical aggregate measures of tax incidence and demonstrates how the modularity of SMS enables researchers to investigate the impacts of various assumptions and data. “In this way, SMS can be especially useful for analyzing political economy,” they observe.

The authors provide a data-driven perspective on targeted revenue recycling provisions and their impacts across and within income groups. “We find that a provision of Washington State’s Initiative 732 designed to mitigate regressive impacts would have provided substantial benefits for a subset of lower income households, but that many low-income households would have still experienced costs in the hundreds of dollars per year,” they write. They uncover large variations in impacts across space and note that suburbs and exurbs of major metropolitan areas face the steepest costs from the carbon tax, and significant variation in impacts exists across jurisdictions with similar population density. This challenges popular concerns that carbon taxes will most adversely affect rural households. The authors also observe substantial variation within tracts, demonstrating that standard approaches may mischaracterize household-level impacts and neglect crucial distributional consequences.

“Our SMS approach allows us to analyze impacts across groups and space under a unified framework, whereas prior work tends to analyze impacts in a single dimension,” the authors write. They propose their findings provide context for ongoing discussions surrounding carbon taxes.

With SMS, the authors conclude, “We are able to link households to local electricity sources, which permits more accurate calculations of tax incidence. Equally importantly, SMS makes it possible to leverage household-level microdata, which avoids common challenges from analyzing aggregate data and also allows for more sensible aggregation of incidence estimates.”