Introduction

Energy expenses influence consumption patterns particularly among socioeconomically disadvantaged groups1,2,3,4. Many households that experience energy insecurity, defined as the inability to meet household energy needs, face a range of physical, economic and coping burdens that impair health, well-being, and wealth-building prospects5,6,7,8,9,10,11. For example, households experiencing energy insecurity often limit their energy consumption because they cannot afford to pay their energy bills, therefore not using enough heating or cooling to maintain thermal comfort12.

Energy insecurity is observed differently by race such that Black, Indigenous, and Latino households disproportionately experience associated financial burdens, inefficiencies, thermal discomfort and trade-offs between basic necessities13. Recent evidence shows that Black and Latino households pay 13–18% more on average for energy per square foot of housing compared to white households14. A robust body of research demonstrates racial disparities in energy cost burdens15; utility disconnections and power outages16,17,18 and gaps in receipt of energy efficiency and renewable energy services19,20,21,22,23,24,25,26,27,28,29.

Disproportionate exposure to energy insecurity is often attributed to income inequalities. However, income and wealth gaps are rooted in systemic and structural racism. As defined by Braveman et al., (2022) “systemic and structural racism are forms of racism that are pervasively and deeply embedded in systems, laws, written or unwritten policies, and entrenched practices and beliefs that produce, condone, and perpetuate widespread unfair treatment and oppression of people of color, with adverse health consequences”30. The synergistic effects of structural racism across various social institutions produce a range of hardships in and beyond the realm of energy31,32. Structural racism fuels persistent disparities in health, housing and neighborhood quality33,34,35,36,37. Discriminatory housing policies, in turn, foster a persistent cycle of racial residential segregation and physical disinvestment in certain communities37,38,39,40. Consequently, minoritized groups are disproportionately relegated to economically and physically underprivileged places.

The dynamics of structural racism permeate various facets of the energy supply and demand continuum25,31,41,42,43,44. The same phenomena shaping racial residential segregation has also informed the placement and quality of energy infrastructure, contributing to disparities in access to alternative energy providers, ability to affect decision-making on energy rates (that is, the cost in dollars that a utility charges a customer or “ratepayer” per unit of energy), and living in areas with heavily polluting energy facilities along with concentrated substandard housing and more intense urban heat island effects that drive up electricity consumption37,45,46,47,48,49. Emerging evidence suggests that the energy infrastructure that serves more racially diverse areas is older and less capable than the energy infrastructure that serves majority-white areas like other public and private physical investment47. Moreover, minoritized groups face longer and more frequent power outages16. These wide and varied examples of disparate racial impacts across the energy continuum suggest that there are indeed influences of structural racism at play. Yet, one understudied potential contributor to racially disparate exposures to energy insecurity are residential electricity prices that are the product of the current energy regulatory and market framework that interact with housing and community settings.

A household’s energy bill reflects energy consumption and limiting behaviors—prompting the question of how pricing for this essential household expense also contributes to disproportionate burdens by race to contain costs and maintain service. In this way, rate differences may function as an upstream contributor to energy insecurity that is not only underexplored but may itself be a byproduct of structural racism66. Figure 1 depicts various mechanistic links between structural racism and energy insecurity and shows the coalescence of macro- and micro-level factors including the unknown associations between energy pricing’s downstream effects on energy insecurity.

Fig. 1
figure 1

Conceptual framework linking structural racism to energy insecurity via (un)known pathways.

Full size image

Energy regulation, electricity rates and Racial disparities in energy insecurity

Operational and regulatory decisions inform the relationship between the energy provider and ratepayer. Residential locations dictate a household’s electricity provider. Most Americans receive electricity from investor-owned utilities (IOUs). Of the approximately 3,000 electricity providers operating in the US, differing by type, size, governance, and regulation, 168 are IOUs serving 72% of U.S. customers50. The remaining 28% of U.S. customers receive services from entities including 1,958 publicly owned utilities (POUs, or municipally owned utilities “munis”) and 812 electric cooperatives (“coops”)50. POUs’ and coops’ leadership are publicly or ratepayer-elected, respectively, and report to local governments if not directly to ratepayer constituents. State public utility commissions (PUCs) have authority over IOUs but not over POUs or coops which are regulated by statute but governed locally51,52. These utilities choose their electricity generation mix and ratepayer prices differently than IOUs. For example, coops own and operate a significant proportion of the distribution lines53,54.

Though rate regulations vary across states, they generally emphasize fairness (i.e., equal rates) over equity including pricing scaled by income55,56. Electricity rates for all utility types are influenced by market as well as regulatory conditions; utilities make operational decisions about how to recover costs and obtain returns by providing service. Most IOUs establish rates after providing evidence of operational costs to PUCs along with a reasonable rate of return on their investments while guaranteeing customers access to electricity at stable prices46. For example, California sets rates to cover capital improvements that may require higher rates in under-invested places (e.g. due to rurality, legacy of segregation, etc.)57.

In all cases, the rate varies by service provider and providers vary by incorporation category; consequently, different utilities charge differing energy rates across places and populations58. There are many rate structures a utility can impose, and these have vastly different equity effects59. For example, flat or fixed rates are regressive, tiered rates and demand charges penalize inefficient homes often occupied by low-income households, and time-of-use rates affect those who cannot schedule their energy use60,61,62.

Rates are most often experienced by the customer through monthly utility bills, yet whether these bills are observed to vary by a range of housing and household characteristics such as race is unclear. Several pathways may explain a possible relationship between electricity rates and race. Communities of color are more likely to have limited access to POUs that charge lower rates in comparison with private large-scale utilities; to have limited ability to affect decision-making on rates given utility governance, and to be living in areas with inferior infrastructure requiring maintenance and upgrading whose costs are passed down to ratepayers45,46,47,63,64,65. In short, past racist policy has led to contemporary variations in energy providers and services that, in turn, may lead to racial disparities in energy rates that entrench the historical racial inequities whether unintentionally or purposefully.

A descriptive examination of differences in energy rates by race is an important first step in exploring the relationship between energy prices, energy insecurity and sociodemographic factors. Therefore, the research question we address is: Do energy rates vary geographically based on the racial composition of utility service areas?

Overview of methods

In this study, we explore if utilities’ service areas and average pricing mirrors the structural racism that exists in housing market valuations. Specifically, we ask if the spatial definitions of utility service areas and associated decision-making that dictate energy pricing (i.e., electricity rates) are informed by historical cycles of disinvestment and racism. Supported by the literature on racial disparities, we hypothesize that electric service providers operating in predominantly white communities charge disparate rates compared to those operating in predominantly non-white communities.

Studying the racial implications of electricity rates among utilities is difficult because utilities do not collect or publicly report household-level data on the racial demographics of their customers, nor on their rates and total consumption by other customer classification. To address this, we developed a methodology combining publicly available data sources on electric utilities and demographics. A high level overview of our study design and approach is shown in Fig. 2 and a full description of methods can be found in the Supplemental Information section.

Fig. 2
figure 2

Conceptual sequence of research design and activities.

Full size image

Our methodological strategy involved the following steps: (1) select case study states—i.e., California, New York, and Alabama; (2) identify electric utilities within the three states that report total annual revenue and consumption, which were collected from EIA for 2017–2021; (3) estimate ratepayer demographics by aggregating American Community Survey (ACS) data within utilities’ service areas for 2019; (4) grouping utilities by type (IOU, POU, coop) and service areas’ racial composition (majority “BIPOC” as an aggregate for Black, Indigenous, Latino households and mixed-race households, and majority white households); and (5) analyze racial differences in average annual energy prices, consumption, and bills for the average customer within these utilities’ various services areas. For each state, we calculated the mean annual electricity rate, consumption, and bills associated with the type of electric utilities and racial composition of the utilities’ service area and determined their statistical relationships through parametric t-test and power analysis. This preliminary approach to connect utilities service areas with the racial distribution of the potential customer base is novel and intended to explore this relationship in the absence of energy rates, consumption and billing data that is segmented by racial demographics at the household or utility-level.

This method has some underlying assumptions and limitations, including: (1) reliance on the assumption that the racial composition of utility consumers is like that of the utility service area; (2) reliance on the average annual metric for the average customer as a marker for differences among utilities; and (3) incomplete data, as not all utilities are required to publicly report annual data. We can only report on available data, which is sparse for municipal utilities. While not intended to establish causation, this correlation study is a first-of-its-kind and represents an important advancement in observations of important energy-related racial inequities.

Results

Controlling for variation in prices between states, Table 1 compares the mean annual difference in price ($/kWh), consumption (kWh/customer), and bill ($/customer) for the average customer in predominately BIPOC service areas and predominately white areas in the three states by utility type. A t-test is used to determine if there are significant differences between means for each state.

Table 1 Mean annual difference in price, consumption, and bills for ratepayers in predominately BIPOC and predominately white services areas in case States for different utility types (2017–2021).
Full size table

For case study states where the comparison of the two groups is possible, service territories classified as majority BIPOC generally see a higher rate. The higher rate seen by predominately BIPOC areas compared to predominately white areas is statistically significant for Alabama’s POUs, California’s IOUs, and New York’s IOUs and POUs. This also holds true over time when averaging all utility types across the three states (Fig. 3).

Fig. 3
figure 3

Average energy rates in predominately white and predominately BIPOC service areas, 2017–2021, across all utility types in case states (left) and for urbanity by utility type and service area racial composition (right).

Full size image

Further, compared to electric utilities with predominately white service areas, electric utilities with predominately BIPOC service areas also had twice as many DOE-defined Justice40 disadvantaged communities in Alabama and California and six times as many disadvantaged communities census tracts in New York (Table S3). When analyzing other markers of social disadvantage, the demographic composition of utilities with predominately BIPOC service areas included a higher share of cost-burdened households, people aged 18–64 years living in poverty, low-income households, and people with less than a high school education. These utilities were also more likely to serve geographies that are more urban and have higher air pollution (Table S3).

To improve our understanding of trends over time, we plot the change in electricity price and consumption between 2017 and 2021 by utility type and the racial demographics of the service area for all electric utilities that were reported to the EIA during this time in Alabama, California, and New York (Fig. 4).

Fig. 4
figure 4

Average annual price ($/kWh) (left) and consumption (kWh/customer) (right) by utility type and service area predominate race group in the case states, 2017–2021.

Full size image

In Alabama, between 2017 and 2021 the mean annual electricity rate among POUs serving predominantly BIPOC service areas was higher by $0.011 per kilowatt hour (kWh) (or 10% more) on average in comparison with POUs serving predominantly white areas. The mean annual consumption among POUs serving predominantly BIPOC was lower by 1720 kWh (or 12% less) on average in comparison with POUs serving predominantly white service areas. We observed similarities in the mean annual bills of both groups (See SI S6). In contrast to the POU trends, we observed no consistent difference in electricity metrics between coops operating in Alabama. Since there is only one IOU in Alabama, we could not compare metrics between IOUs, but we note that Alabama Power collected the highest revenue per kWh in the state.

In California, between 2017 and 2021, IOUs with predominantly BIPOC service areas—that is, Pacific Gas & Electric (PG&E), South California Edison (SCE), and San Diego Gas & Electric (SDG&E)—had the highest mean annual electricity rate, the lowest mean annual electricity consumption, and the highest mean annual bills. In contrast, we did not observe consistent differences in electricity metrics based on race between POUs in California. Note that the price in California for IOUs seems to trend upwards from 2017 to 2021 and diverges between predominantly BIPOC and predominantly white service areas, presumably due to utility pricing changes implemented after the 2017 wildfire season. These result in the 2017 price differential of $0.01 per kWh (or 8% more) compared to the 2021 $0.06 per kWh differential (or 33% more).

In New York, the mean annual electricity rate for Consolidated Edison, the only IOU with a predominantly BIPOC service area, was higher by $0.108 per kWh (or 71%) on average between 2017 and 2021 compared to the other five IOUs in the state. During the same time, the mean annual kWh consumption among Consolidated Edison consumers was lower by 3429 kWh. In other words, areas classified as predominantly white consume nearly double the electricity on average in comparison with the mean annual consumption of the other five IOUs operating in New York. Similarly, the mean annual electricity rate for Freeport, the only POU with a predominantly BIPOC service area, was on average $0.66 (or 200%) higher per kWh compared to the mean annual electricity rate of other 39 POUs in New York. At the same time, the mean annual consumption in Freeport was also lower by 4381 (or 35%) on average compared to the other 39 POUs in New York state. These dynamics could potentially explain why the lowest mean annual electricity bills were observed in utilities with predominantly white service areas in New York.

Discussion

We identified associations between the predominant racial makeup of utilities service areas and electricity metrics of price per kWh, consumption and bills across different utility types in Alabama, California, and New York. Our results indicate that majority BIPOC service areas are more likely to pay higher energy prices but consume less energy compared to predominantly white service areas. This finding is complementary to prior studies that identified that residents in predominantly non-white communities spend larger shares of their income on their bills despite using less energy and rely on strategies such as energy limiting behavior and vigilant conservation5,6,22,24.

The present analysis expands on that pattern by exploring how different utility types shape these disparate energy pricing schemas. In California and New York, states with multiple IOUs, we found that IOUs serving predominately BIPOC, and more urban areas have consistently higher annual electricity rates in comparison with IOUs in predominately white and less urban service areas. For example, in New York in 2019 the average annual price per kWh for customers of Consolidated Edison in New York City was $0.253, while the average annual price for the other five IOUs serving more rural parts of the state was $0.142, a 78% difference. For the same year in California, the average annual price among the three largest IOUs that serve the most urban parts of the state was $0.214, while the average annual price for smaller IOUs that serve more rural parts of the state was $0.175, a 22% difference. These findings suggest that differences in energy pricing among IOUs is associated with the racial makeup of their service area along with the scale of operation and level of urbanization in IOUs’ service areas.

Other utility types beyond IOUs are not necessarily more equal in their pricing. In Alabama, where communities of color account for a third of the population, POUs serving predominately BIPOC areas charge on average 10% more per kWh than POUs serving predominately white ratepayers. In New York, where the BIPOC population accounts for 45% of the total, the only POU that serves predominately BIPOC areas is Freeport, and its electricity rates are nearly double the average electricity rate of the 39 POUs that serve predominately white areas. In both states, POUs in predominately BIPOC areas are also more urban than the POUs serving predominately white areas. In contrast, in California, where communities of color account for 64% of the population, there were no consistent differences in energy rates between areas served by POUs. While this is an encouraging finding, access to municipal utilities remains a challenge. IOUs serve more than two-thirds of electricity customers in California, and these electricity providers charge the highest rates.

Consumption and bill differences in the service areas also surfaced in our study. Across Alabama, California, and New York, the mean annual electricity consumption among utilities serving predominately BIPOC areas was lower than the mean annual electricity consumption in utilities with a higher concentration of white residents. Consequently, the mean annual electricity bills among utilities in majority BIPOC areas were higher or comparable to the mean annual electricity bills of utilities serving predominantly white areas. This is a particular concern given the impact of higher energy prices on the energy consumption of low-moderate income households, which make up a larger share of the territory for utilities with majority-minority service areas.

Beyond race as the primary pathway, potential alternative explanations for the observed associations include: (1) higher costs for infrastructure investments, which are more spread out, outdated, facing increasingly frequent climate hazards, and in need of upgrades to meet energy transition requirements; (2) mandates to provide public purpose programs such as energy efficiency programs; (3) liability for damage caused by energy utility infrastructure (e.g., California wildfires); (4) business strategies to increase profit to shareholders, motivating investment in new projects rather than maintaining existing infrastructure; (5) compliance with state renewable energy portfolio requirements, which have increased energy purchasing costs when solar and wind are expensive; and (6) cross-subsidies with non-residential customer classes.

Of note, this analysis provides preliminary observational evidence advancing the hypothesis that places with different racial composition are charged different electricity rates and, specifically, that communities of color are more likely to live in areas served by electric utilities that charge higher rates. This research is of paramount importance given the escalating public interest in reducing energy burdens while equitably decarbonizing. Gaining a deeper understanding of the equity implications inherent within the current energy system and how it apportions the costs of energy provision—in this study’s case, as reflected in energy rates—is necessary for implementing and expanding utility policies with a strong evidence base. Unless more direct measures are taken to document and close existing racial gaps in the energy system, the impact of these investments may serve to reify existing inequalities that limit any benefits from the energy transformation.

In addition to the study’s broad policy relevance is the practical one regarding utility governance. For most households, the only option for an electricity provider is the single IOU that serves their area, a phenomenon that may reinforce other historic practices with disparate racial impacts such as segregationist housing policy. The ability of IOUs to operate as monopolies is part of the century-old regulatory compact, which offers ‘just and reasonable rates’ in exchange for investor security, the promise of the opportunity to earn a limited but assured reasonable return on prudent investments for providing an essential public service66. Acknowledging that IOUs are regulated differently across the US, we included in our analysis states that represent three models of energy rate regulation. In Alabama and California electricity rates are set by the regulator, and in New York electricity rates are a product of market competition under conditions set by the regulator. IOUs across these three states charge the highest rates and serve a higher share of minoritized and lower-income population than municipal utilities and rural electric cooperatives. These findings call for the attention of regulators to investigate why IOU rates are high despite the promise of improved service and rate affordability via economies of scale.

The alternatives to IOUs are POUs and coops, which provide not-for-profit energy services. Yet, not everyone has access to these types of energy providers, as seen in Alabama’s case. Further, positive alternative pilot models are also not yet showing clearly equitable benefits. For example, the growth in California’s community-choice aggregators and net-metering has created a problematic dynamic whereby early CCA and solar adopters, who are wealthier and mostly white reduce or discontinue service with IOUs—leaving a higher share of lower-income and communities of color paying to maintain legacy infrastructure. While regulators tried to address these disparities by revamping net-metering policies, the result has secured lower rates for early adopters, further perpetuating the racial gap in access to clean energy and lower energy rates. By having a better grasp of these dynamics, regulators can ensure fair energy prices as rate cases are debated. Finally, the authors hope that this study spurs a broader enterprise of research in addition to informing policy and practice. Our methodology can be replicated for any US state or other geographic location with data on the service boundaries of electric utilities, annual energy revenue and consumption metrics, and granular demographics using all publicly available data. Applying such methodology to other locations can help better elucidate the relationship between racial segregation and provision of energy services, ultimately contributing to policies to help reduce energy insecurity among BIPOC and low-income households to protect the health and well-being of all communities.

Conclusion

The association between utilities with predominately BIPOC service areas and lower electricity consumption should be further investigated to evaluate the potential link between racial disparities in energy rates, energy insecurity, and health disparities. Eliminating racial inequities in energy rates requires a thorough understanding of all the elements that cause them including how race interacts with other social risk factors. As such, future research should explore the relationship between energy rates and sociodemographic risk factors in addition to race, such as income, health status, housing tenure, housing quality, and neighborhood conditions. Additional inquiry should explore the mechanisms behind the observed association between the racial composition of utility service areas and electricity rates, as well as differences in rates between IOUs, POUs and coops. We hypothesize that urban density of the service environment might also play a role in increasing costs, but more research is needed to explain these disparities. Future research should also explore variation in energy rates between municipal utilities. Municipal utilities serving majority-BIPOC areas may face higher costs due to aging, under-invested infrastructure that requires modernization. In these areas, where inefficient housing stock is also common, higher utility rates could place cumulative financial burdens on a predominantly lower-income customer base that, as our results indicate, is doing as much as possible to lower costs by using less energy.

An important corollary to this future research agenda is the critical need for data or, more accurately, data transparency. Future research could benefit from more accessible and granular data on utility customer demographics, costs, and consumption by rate-payer segment, tier, and geography. Federal- and state-level regulatory bodies should consider mandating greater reporting and data collection among utility service providers to better assess social inequities lurking in the energy system. This study strengthens the need for more robust utility-based data to understand not only the demographic composition of the service area, but the key characteristics of their customer base.