Even if the Nationally Determined Contributions (NDC) of the Paris Agreement are realized, global annual greenhouse gas (GHG) emissions, in 2030, are projected to be 124% higher than what is needed to limit global temperature rise to 1.5 °C (United Nations Environmental Program, 2020). These emissions occur to provide goods, services, and wealth to people around the world (Díaz et al., 2019). Yet significant economic inequality, both within and between countries, results in a powerful disconnect between the groups who reap these benefits and those that are left to deal with the harms caused by excessive GHG emissions, i.e., global climate change. Poorer and socially marginalized peoples tend to be the most impacted by climate change and other environmental degradation (Althor et al., 2016; Ash and Boyce, 2018; Diffenbaugh and Burke, 2019; Hsiang et al., 2017; Intergovernmental Panel on Climate Change, 2014; Islam and Winkel, 2017; King and Harrington, 2018; Leichenko and Silva, 2014) yet environmental change is disproportionately driven by, and for the benefit of, those with the most resources and social privilege (Feng et al., 2021; Hoegh-Guldberg et al., 2019; Moran et al., 2020; Song et al., 2019; Tessum et al., 2019; Wiedmann et al., 2015).
It is widely accepted as a basic principle of fairness that those benefiting from an activity, like the GHG emissions that drive climate change, should bear some responsibility in mitigating the damage caused by those activities. This idea was central to the creation of “loss and damage” funding at the recent United Nations Climate Change Conference (COP27). From the international community's first attempt at collective climate action, the 1992 United Nations Framework Convention on Climate Change (UNFCCC), through the 2015 Paris Agreement and COP27, climate responsibility has largely been conceptualized as national-level responsibility for emissions produced within a country's territory. However, the continued globalization of supply chains, since the UNFCCC, means that significant emissions may occur in one country to create goods and services that are exported around the globe. To account for this, an alternative consumer responsibility framework has been developed over the last few decades (Afionis et al., 2017; Bicknell et al., 1998; Büchs and Schnepf, 2013; Davis and Caldeira, 2010; Feng et al., 2021; Ghertner and Fripp, 2007; Hertwich and Peters, 2009; Jones and Kammen, 2011; Lenzen, 1998; Moran et al., 2020; Song et al., 2019; Weber and Matthews, 2008). This calculates a nation's responsibility based on emissions that occur anywhere in the world to produce the goods and services consumed within a country's territory.
Because goods and services ultimately flow to people, consumption-based emissions responsibility can be traced to individual households, including those at the top of the income distribution who have an outsized role in shaping U.S. public policy (Gilens and Page, 2014). If society is to develop effective and just climate policies it is critical to understand how emissions are distributed across the whole of U.S. society, including among those who have disproportionate power to determine climate outcomes.
Below, we present results from a highly granular time series analysis (1996–2019) of consumption-based U.S. household GHG emissions. For each year, we employ an Environmentally-Extended Multi-Region Input-Output Model (EE-MRIO) to calculate the GHG emissions embodied in 10,211 commodities across 190 countries (> 100 million inter-sectoral transfers per year). The embodied emissions in these goods and services are tracked to final-demand household-level purchasing from a mostly nationally representative1 sample of ∼14,500 U.S. households per year. Expenditures for top 1% and 0.1% households, which are under-sampled in the underlying survey data, are also estimated by constructing a synthetic dataset. Direct household emissions, such as vehicle fuel use and home heating, are also accounted for. To reveal how income inequality relates to inequality in emissions footprints, households are binned into income deciles, including a disaggregation of the top decile into the next 9% (90.0th - 99th percentile) and top 1% (99.0th - 100th percentile), and a further disaggregation of the top 1% into the next 0.9% (99.0th - 99.9th percentile) and top 0.1% (99.9th - 100th percentile) of income earners.