Utility tariffs are agreements between utilities and regulators that define the contractual relationships between utilities and their customers (throughout this paper, the term “regulators” refers to the entity responsible for electric utility regulation, typically a public utility commission). Utility tariffs define rate structures and other terms for broad customer classes, such as a tariff defining terms of service for all commercial customers operating in a defined region. In some cases, utility tariffs offer choice over utility supply by allowing C&I customers to procure power from specified resources owned or under contract with the utility. Resource-specific tariffs are commonly called utility “green” tariffs because they provide access to a subset of renewable resources, though utilities can offer tariffs for specified power from non-renewable resources. Utility green tariffs function like a power purchase agreement with the utility as intermediary buying power from the generator and selling power to the C&I customer. The structure is often referred to as a “sleeved” power purchase.

Utility green tariffs have proliferated in recent years, largely in response to pressure from large C&I customers (O'Shaughnessy et al., 2021). We identified 27 distinct utility green tariffs in 16 states of our 18-state sample. Utility green tariffs can be designed as bespoke agreements to meet the unique needs of specific customers. More often, utility green tariffs are available to all C&I customers that meet specified criteria, including the 27 programs explored here. Utility green tariffs vary in terms of customer eligibility criteria. Most utility green tariffs are restricted to C&I customers with at least 1 megawatt (MW) of peak demand. Utility green tariffs generally require multi-year contractual commitments from participating customers.

Utility green tariffs also vary in terms of the resources procured under the tariffs. Twenty (20) of the tariffs include options for tariff service from new renewable energy generators, and 16 of those programs require procurement from new generation. A key distinction across utility green tariffs is whether the utility or the customer prompts the development of new resources. Resource selection and development is led solely by the utility in 6 of the 20 tariffs with new development options in our sample. In these cases, the tariffs allow customers to choose an alternative power supply but do not allow customers to choose the specific supply characteristics. These tariffs often provide a way for utilities to propose new resources and use tariffs to enroll customers onto service from those resources. For example, Georgia Power's Renewable Energy Development Initiative used a green tariff to develop a portfolio of 200 MW of new renewable resources. In contrast, eight tariffs in our sample specify that participating customers must prompt the development of new resources, and six tariffs include options for utility- or customer-driven development. For example, Idaho Power's Clean Energy Your Way tariff includes a “construction” option under which the utility offers to develop and sleeve power from projects identified by participating customers. In cases where green tariffs allow customers to procure new resources, the tariffs generally stipulate that participating customers must bear all incremental investment costs of those resources.

The Nevada Clean Transition Tariff (CTT) provides an illustrative example of an innovative utility green tariff. The CTT illustrates many common features of utility green tariffs. Like many green tariffs, the CTT is the outcome of a collaboration between a customer (in this case Google) and a specific utility (NVEnergy). The CTT defines a unique rate structure available to customers with at least 5 MW of monthly average demand. The CTT provides a pathway for customers to bring new resources online and procure supply from those resources via the utility. The CTT is unique among green tariffs in two regards. First, the CTT is designed to procure clean firm resources such as geothermal, as opposed to the variable resources of solar and wind supported by most green tariffs. Clean firm generation could be critical for deep grid decarbonization and for companies seeking to match clean generation to electricity use on an hourly basis. However, clean firm resource deployment is currently limited by the high costs of clean firm resources relative to solar and wind power generation. Utilities typically cannot obtain regulatory approval to invest in emerging clean firm technologies when cheaper alternatives are available. The CTT could solve that challenge by moving financial risks onto participating customers rather than the utility rate base. Second, the CTT is unique in its degree of cost protection for non-participants. All green tariffs include some degree of non-participant cost protection. For instance, NV Energy's existing GreenEnergy tariff requires that rates for participating customers must be at least greater than the otherwise applicable rate and that the tariff must protect non-participants. The CTT goes further by requiring participating customers to make long-term commitments equivalent to the lifetime of the clean firm generators (Wu et al., 2024), in contrast to the 1 to 5 year commitments required under the GreenEnergy tariff. Further, the CTT requires participants to post securities to guarantee contractual performance. The CTT is explicitly designed to pass all cost premiums and financial risks onto participating customers and to fully insulate non-participants from those risks (Flanagan, 2024; Wu et al., 2024). The potential use of green tariffs like the CTT for non-participant cost protection is a theme we explore in further depth in the Discussion.

Some states extended power choice to subsets of retail electricity customers during the initial phase of U.S. electricity market restructuring. These provisions are often known as “direct access,” in contrast to full retail competition which extends choice to all customers. Direct access provisions typically define criteria under which C&I customers can seek generation services from non-utility suppliers. Eligible customers apply for direct access from the state public utility regulator. Regulatory approval is generally conditional on a demonstration that direct access will not adversely affect non-participants. Regulated utilities remain responsible for transmission and distribution service.

Nine (9) of the 18 states in our sample have direct access provisions (see Table 2). Direct access eligibility varies substantially across the states, with peak demand thresholds defined as low as 30 kW in Oregon to as high as 100 MW in Utah. Direct access provisions may define various constraints on the ability of customers to leave utility service for alternative supplies. Common constraints include fees to switch to direct access (commonly called “exit fees”), restrictions on the ability of customers to return to utility service, and provisions to mitigate potential impacts on non-participants.

Utah provides a case study for the use of direct access as a hybrid choice policy to address increasing C&I demand. In 2025, the Utah state legislature passed Senate Bill (SB) 132 stipulating requirements for new large loads. SB 132 requires the state's investor-owned utility to conduct a technoeconomic evaluation of its ability to meet requests to serve new large loads, defined as loads expected to exceed 100 MW of demand within 5 years. SB 132 requires the utility to notify large load customers of whether, based on this evaluation, the utility can meet the customer's large load request. SB 132 stipulates that large load customers can procure service from non-utility providers if (a) the utility fails to complete the evaluation within the required 6 month timeframe or (b) the utility and customer fail to negotiate a contract within 90 days from the completion of the evaluation. Regardless of whether the customer procures service from the utility or a non-utility supplier, SB 132 requires that all incremental costs be allocated to participating customers. Direct access customers are wholly responsible for all generation costs in contracts with non-utility suppliers and responsible for any transmission and distribution costs incurred by the utility.

Utah SB 132 is unique in two regards that may provide a template for hybrid choice policies in other states. First, under SB 132, regulated utilities remain the default service provider, but the rule establishes pathways through which C&I customers can request direct access to non-utility suppliers. That process is implicitly based, via the techno economic evaluation, on the utility's ability to meet the C&I customer's request. The legislation's utility service default is distinct from direct access in other states such as Nevada, where qualifying customers can seek direct access regardless of the utility's ability to serve those customers. Second, the SB 132 size threshold of 100 MW restricts the program to a relatively small class of very large C&I customers. For context, an average-sized natural gas plant in the U.S. has a rated power capacity of 85 MW (EIA, 2024). The legislation is, at least implicitly, designed specifically to help integrate very new large loads such as data centers. Again, the high size cap will likely limit C&I power choice, in practice, as relatively few C&I customers will qualify to request direct access. Nonetheless, the large size cap may represent a practical compromise to make direct access viable in states that are unable to commit to broader options for direct access.

Community energy programs allow customers to pay special rates to procure power from specific energy projects. While community energy programs can be based on any energy resource, only community solar has achieved any meaningful scale in the United States, and we therefore focus exclusively on community solar. Community solar programs can be administered by utilities based on utility-owned assets or by third parties based on assets owned by non-utility entities. The third-party model is generally more common, with about 88% of community solar programs administered by third parties nationwide (Xu et al., 2025). However, community solar programs are evenly split in our 18-state sample between third party- and utility-administered programs, and utility-administered programs account for 91% of installed community solar capacity in the 18 states (Xu et al., 2025).

Community solar differs from utility green tariffs in that community solar may be based on third-party administered programs and projects. Another key distinction is that community solar programs are based on “subscription” models where any type of customer (including residential) can enter and exit the programs, whereas utility green tariffs involve multi-year contractual commitments and are restricted to C&I customers. Community solar also generally differs from utility green tariffs in three other regards: (1) some states allow non-utility suppliers to offer community solar products, while other states only allow utility-administered community solar, meaning that community solar can enable both types of choice (choice of supplier and choice of utility supply); (2) community solar programs typically source power from relatively small, “community scale” projects, typically meaning projects with less than 10 MW of capacity (Xu et al., 2025); and (3) community solar programs generally do not include explicit provisions to protect non-participants from rate impacts. Community solar participants may be cross-subsidized by non-participants to some degree (Haynes et al., 2020). State utility regulators are responsible for ensuring that such cross-subsidies are part of a just and reasonable rate design.

Community solar generally requires state policies or regulatory approval. At least 23 states have legislation to enable community solar, such as rules requiring utilities to offer virtual net metering, including 7 of the states in our sample (Xu et al., 2024). Many states further promote community solar by subsidizing community solar subscriptions, typically through revenues from solar renewable energy certificates. However, community solar is common in states without enabling legislation. Indeed, Florida is by far the state leader in terms of community solar program capacity, despite a lack of any enabling policy.

Florida provides a useful case study for community solar as a hybrid choice model for C&I customers. Florida utility regulators authorized the state's investor-owned utilities to develop two of the nation's largest community solar programs. The utilities were authorized to develop 51 solar projects each with nearly 75 MW of capacity (additional regulatory restrictions would have applied for projects larger than 75 MW). Unlike community solar in most other states, these projects are owned and operated by the state's investor-owned utilities. Some community solar advocates criticize the Florida community solar model for not allowing for competitive procurement, as is typical in most community solar programs (Gheorgiu, 2020). Notwithstanding those critiques, Florida's community solar model provides a customer choice pathway in a state with otherwise limited options for C&I power choice. The relatively large size of Florida's community solar projects provide capacity to meet the demand of larger C&I customers. Outside of Florida, many states place system size limits on community solar projects (IREC, 2020), and the average community solar project outside Florida has just 2.4 MW of capacity (Xu et al., 2025). The relatively large project sizes likely facilitated substantial cost savings through economies of scale. Further, the program offered by the state's largest utility reserved 75% of program capacity for C&I customers (Gheorgiu, 2020).

All retail electricity customers can supplement their grid power supply with on-site power generation, most commonly through small-scale fuel (e.g., diesel) generators and solar photovoltaics. Small-scale, customer-sited energy resources are commonly referred to as distributed energy resources (DERs). There are no regulatory constraints on the ability of customers to use power from customer-owned distributed energy resources. However, utility regulation applies to the ability of DER system owners to deliver power to the grid and the ability of customers to buy power from third-party owned DERs. In terms of grid exports, the Public Utility Regulatory Policies Act (PURPA) requires utilities to compensate DERs at a rate that reflects the utility's avoided costs (PURPA does not apply for systems larger than 80 MW). Among our 18-state sample, 16 states require utilities to compensate excess output under specified DER capacity limits at higher rates than required under PURPA (see Table 2). The details of these state-level requirements can significantly affect the economics of DER adoption. In terms of third-party ownership, the third-party ownership model allows customers to “host” DERs that are owned by third parties such as solar system developers and banks. The DER system hosts make ongoing payments for DER system output in lieu of purchasing the system hardware. Third-party ownership appeals to many customers because it allows customers to finance DERs and shift operation and maintenance responsibilities and system risks onto the third-party owners. Utility regulation generally prohibits third-party ownership because it entails electricity sales by non-utility suppliers. However, 29 states have explicitly authorized third-party ownership and 15 states have ambiguous regulation that can support third-party ownership in certain cases (DSIRE, 2025). Only 6 states explicitly prohibit third-party ownership, including 5 states in our 18 state sample.

DER regulations generally do not contain explicit provisions to protect non-participants. DER adoption can exacerbate underlying challenges in rate design and can affect electricity prices paid by customers that do not adopt DERs (Wiser et al., 2025). Utility regulators are responsible for ensuring that price impacts on non-adopters are part of a just and reasonable rate design.

DER colocation has unique benefits among the hybrid choice pathways for facilitating the grid integration of new large loads. DERs co-located with new large loads can be operated flexibly in ways that serve the load while providing grid benefits (Norris et al., 2025; Spector, 2025). For instance, battery storage infrastructure at data centers can be flexibly operated to maintain data center power supplies and provide grid ancillary services (Alaperä et al., 2018; Türker Takci et al., 2025). Utilities, regulators, and policymakers could therefore consider ways to enable DER co-location to increase the grid flexibility of new large loads (ICF, 2025). One example in our state sample is Georgia Power's Back-up Generation Solutions program. The program enables C&I customers to host DERs that are operated by the utility. The program stipulates that the utility will operate the DERs to benefit all customers. In exchange, the DER system C&I customer hosts can use the DERs for backup power during grid outages. The program allows participating customers to make a single up-front payment to host a utility-owned DER or to own the DER and receive bill credits for the DER's estimated system value. The program is designed to support relatively large DERs and is thus presumably aimed toward relatively large C&I customers. For customers that own the DERs, the program requires a minimum DER capacity of 1 MW, or an aggregation of systems with at least 250 kW. For customers that host utility-owned DERs, the minimum capacity threshold is 10 MW. By way of comparison, a typical household rooftop solar system has less than 10 kW of capacity.

Customer interest in and uptake of the Georgia Power Back-up Generation Solutions program remains unknown. Nonetheless, the program illustrates one way that utilities and grid operators could use hybrid power choice to manage increasing C&I electricity demand. The program is explicitly designed to bring new resources online that benefit all customers. Because the utility owns the DER output, the program effectively prevents potential price impacts on non-participants, thus avoiding a common concern related to DERs. Participating C&I customers are compensated in the form of increased resiliency during grid outages and bill credits based on the system value of the DERs. Other states could explore if resiliency-based models could help utilities couple new loads with new resources that insulate non-participants from electricity price impacts. We return to this point in the Discussion.

Community choice is a model where an entity representing a jurisdiction chooses a power supply on behalf of investor-owned utility customers in the jurisdiction. The entity is typically a non-profit group formed to represent the jurisdiction. Community choice is often called community choice aggregation or municipal aggregation, emphasizing that the model aggregates the demand within jurisdictions for the purposes of procuring power. Community choice policies can include measures to prevent impacts of community choice on non-participating communities. For instance, California community choice legislation required participating communities to pay fees akin to the exit fees of direct access provisions.

The term “community choice” is generally understood to imply an ability to choose an alternative non-utility supplier. Supplier-based community choice is only possible when authorized by state law and had been authorized in 10 states as of November 2025. Supplier-based community choice is not authorized in any of the 18 states in our sample, though such legislation has been explored in Arizona, Colorado, New Mexico, and Washington (LEAN Energy US, 2025). However, broadening the concept of “community choice” to mean choice over utility supply widens the sample of states where community choice is possible. Even in states without authorized community choice of supplier, cities can leverage franchise agreements with utilities to pursue some degree of community choice over utility supply, such as in the case of Boulder, CO.

Utah provides a case study in community choice in this broader sense. In the 2010s, several Utah communities explored ways to enhance community choice, including conventional community choice aggregation and municipalization. These efforts prompted the passage of the Community Renewable Energy Act (CREA) in 2019. CREA opened a brief window in 2019 allowing Utah communities to request the state's investor-owned utility to deliver 100% renewable energy service by 2030. Twenty-three Utah communities joined the program (Kunkel, 2021), including the Grand, Salt Lake, and Summit Counties, meaning that at least one-third of the state's population resides in participating communities. Participating communities agreed to bear the incremental costs required to achieve the 100% renewable energy target. Like conventional community choice aggregation, CREA allowed for an opt-out structure, where residents of participating communities are defaulted into the program but can opt out back into standard utility service (Kunkel, 2021). The key distinction between the CREA model and conventional community choice is that the Utah communities cannot seek alternative suppliers; the communities can only request alternative supplies from the utility. The CREA model may provide a useful template for a hybrid form of community choice in states that are unlikely to pass legislation to authorize supplier-based community choice.

Figure 3 depicts customer uptake of the four hybrid choice models in the 18 states from 2019 to 2024, and Figure 4 depicts uptake across models by state in 2024.

Utility green tariffs are the fastest-growing hybrid choice model in terms of customer uptake. Nationwide, utility green tariff sales grew from around 1 million MWh in 2015 to around 22 million MWh in 2024 (O'Shaughnessy et al., 2025). Utility green tariff sales account for more than 5% of all C&I sales in three states in our sample (note that Montana is among these states due to an ad-hoc utility green tariff, though the state does not yet have a confirmed utility green tariff program as reflected in Table 2). Overall, an estimated 17 million MWh were sold through utility green tariffs in the 18 states, equating to around 2% of sales across the states. Utility green tariffs account for about 2.8% of C&I sales on average in the 18-state sample, significantly higher than the average of 0.6% of sales in other states (t = 2.9). These statistics suggest that utility green tariffs provide a substitute form of choice in states without wholesale power markets or retail competition, as likewise suggested by the spatial availability of green tariffs depicted in Figure 2.

Direct access sales accounted for more than 1% of C&I sales in 2024 in four states in the sample, reaching as high as 19% of all C&I electricity sales in Nevada in 2024. Across the 18 states, about 170 C&I customers procured around 10 million MWh in 2024 through direct access, or about 1% of all C&I sales in the sample. The data suggest that direct access sales have declined over time in these 18 states in absolute and relative terms (see Figure 3). Unlike in the case of green tariffs, direct access sales are far larger outside of the 18-state sample. While non-utility suppliers account for about 1% of C&I sales in the 18-state sample, non-utility suppliers account for about 6% of C&I sales in the 17 states with wholesale power markets but without retail electricity competition, and about 69% of C&I sales in the 13 states with retail electricity competition.

Community and on-site solar are substantially smaller than the other hybrid choice models in terms of estimated C&I customer uptake. Community solar is likely a niche product for relatively small commercial customers in most states outside of Florida (see Section 3.3). Still, overall we estimate that C&I customers procured about 6 million MWh of community solar in the 18 states in 2024, or about 0.7% of all sales. As noted in Section 2, these estimates are based on uncertain assumptions around C&I customer shares of community solar capacity. Based on the range of assumptions described in Section 2 (20-80%), we estimate that C&I customer community solar procurement ranged from about 5.0 to 7.4 million MWh in 2024. On-site solar procurement is of a similar magnitude, with an estimated 1.7 million MWh of generation in the 18-state sample in 2024.

C&I customer uptake of hybrid choice models varies substantially across the 18 states in our sample. We posit and explore two potential explanations for these disparities in hybrid choice sales. First, disparities in hybrid choice sales could reflect differences in the availability of hybrid choice models across states. That hypothesis is consistent with prior research demonstrating that an expansion of C&I renewable energy procurement pathways was associated with a similar expansion in C&I renewable energy procurement (O'Shaughnessy et al., 2021). Similarly, an expansion of hybrid choice pathways within states may be associated with increased uptake of hybrid choice. The data are consistent with this hypothesis. Eight of the 18 states in the sample offer four of the five hybrid choice pathways described in Section 3 (see Table 2). The state-level average C&I customer uptake of hybrid choice in these 8 states in 2024 was about 2.4 times higher than in states with fewer hybrid choice pathways, though that difference is not statistically significant (t = 1.7). Second, the programmatic details of hybrid choice policies may affect C&I customer uptake. This hypothesis is supported by the relatively strong uptake of community solar in Florida, where the program is specifically designed for C&I customers. Further, C&I customer uptake of direct access programs is generally higher in states with lower customer eligibility thresholds. In the 4 states with eligibility thresholds no larger than 1 MW (GA, NV, OR, WA), direct access uptake in 2024 was about 7% on average, compared to 1% on average in the other direct access states, though again the difference is not statistically significant (t = 1.5). However, the difference is clear and statistically significant when considering the impacts of full retail choice on C&I customer uptake. About 70% of C&I sales were from non-utility suppliers in states with full retail choice in 2024 (see Figure 1), compared to 19% of C&I sales in states in the contiguous U.S. without full retail choice (t = 5.3).

Demand for C&I power choice is largely driven by C&I customer interest in managing electricity costs and achieving sustainability objectives. Hybrid choice policies have helped C&I customers to achieve both objectives. In terms of C&I customer costs, the available evidence suggests power choice can help C&I customers more effectively manage electricity costs through access to more flexible rate structures (e.g., time-varying rates) and long-term contracts with alternative electricity suppliers (Borenstein and Bushnell 2015; Rose et al., 2024). In terms of sustainability objectives, in 2024, the NLR estimates that C&I customers procured around 17 million MWh of renewable energy above state-mandated levels through utility green tariffs and direct access in our 18-state sample (O'Shaughnessy et al., 2025). Those C&I customers could have otherwise procured renewable energy through other pathways that rely largely on out-of-state resources, such as by procuring unbundled renewable energy certificates or signing virtual power purchase agreements. However, the hybrid choice policies provide C&I customers with pathways to meet sustainability objectives through in-state resources, meeting the demands of many C&I customers for “local” renewable energy generation.

In 2024, we estimate that C&I customers procured about 34 million MWh through these hybrid choice pathways, representing about 4% of C&I demand in those states. C&I customer uptake of hybrid choice pathways varies substantially across the states in the 18-state sample. We estimate that sales through hybrid choice programs account for as much as 22% of C&I sales in Nevada to as little as less than 1% in South Carolina. We posit two hypotheses for these disparities in customer uptake and find suggestive evidence to support both. First, C&I customer uptake of hybrid choice correlates with the availability of different hybrid choice pathways. Different hybrid choice pathways may meet the specific needs of different customers. As a result, a greater variety of pathways would meet the needs of more customers and increase C&I customer uptake of hybrid choice models. Second, the programmatic details of hybrid choice pathways may affect customer uptake. For example, more restrictive customer eligibility criteria (e.g., larger minimum demand thresholds) will likely reduce customer uptake. We find suggestive evidence for both hypotheses. Future research could explore these and other hypotheses in further depth through more rigorous causal modeling, such as analyses of changes in C&I customer uptake before and after hybrid choice policy changes. Further, we find diverging trends in the uptake of different hybrid choice models. From 2019 to 2024, C&I customer uptake of utility green tariffs in the 18-state sample gradually increased, while customer uptake of direct access gradually declined. Future research could explore the drivers of trends in C&I customer uptake of hybrid choice across models.

In addition to helping C&I customer manage costs and achieve sustainability objectives, we posit that hybrid choice policies could mitigate adverse effects from rapidly increasing C&I electricity demand. That hypothesis stems from two observations from the hybrid choice policies summarized in Section 3. First, most hybrid customer choice programs include measures to mitigate the impacts of power choice on non-participants, meaning customers that do not or cannot choose alternative power supplies. Utility green tariffs can include contractual terms that allocate incremental costs to participating customers, and regulatory approval for direct access or community choice can be conditioned on an assessment of impacts on non-participants. Existing hybrid choice policies do not necessarily mitigate cost risks to the same extent as emerging large load tariffs (Collier and Lindemann, 2025), but such policies could be adapted for enhanced risk mitigation. The Nevada CTT, summarized in Section 3.1, may provide a useful template for ways to minimize non-participant risks in hybrid choice policies. Second, customer choice policies can be and often are designed to support the development of new resources. Utility green tariffs, in particular, are often designed to support utility procurement of new generation. We find that 16 out of 27 utility green tariffs in our geographic sample require service from new generation resources, and 4 additional tariffs include options for service from new generation. Direct access provisions can likewise be designed to support new resources, such as the Utah legislation explored in Section 3.2. As a result, hybrid choice policies can be designed to enable large C&I customers to deploy new generation resources that can facilitate integration of new large loads onto the grid, and hybrid choice policies can include safeguards to mitigate risks of cost impacts on non-participants. Many utilities already view customer power choice programs as a way to enable the development of new resources and reinforce grid reliability (ICF, 2025).

Utilities and regulators are responding to rapid C&I demand growth by filing requests for and implementing new large-load tariffs (Collier and Lindemann, 2025; Satchwell et al., 2025). A collaboration of researchers, utilities, regulators, large-load customers, and other stakeholders identified a set of eight principles to guide the design of new large-load tariffs (Cannon and Wang, 2025). Three of the principles emphasize the importance of allocating incremental costs to new large loads and protecting non-participants. One of the principles describes the need to define eligible resources, including generation, that will be “sourced or supported via utility procurements, bilateral (i.e., between customer and project) or trilateral contracting (i.e., between, customer, utility, and project], behind-the-meter and front-of-meter co-location arrangements, or other sourcing processes” (Cannon and Wang, 2025). This principle identifies distinct types of customer choice as central elements of large load tariff design. Still, pathways for customer choice are not always clearly identified in emerging large load tariffs. We identified a sample of 28 tariffs requested or implemented in 24 states from July 2023 through October 2025. An exploratory analysis of this sample identified 10 tariffs that unambiguously identify pathways for new large loads to choose an alternative utility supply or to procure power from non-utility suppliers. Future research could further analyze emerging large load tariffs to understand how customer choice is or is not being used to integrate new large loads onto the grid.

Our review of hybrid choice polices built on a categorization based on five hybrid choice pathways. Our selection of those categories built on prior published categorizations, our own subject matter expertise, and the results of a narrative review. Nonetheless, our categorization may not comprehensively describe the hybrid choice pathways available to C&I customers in our 18-state sample. Further, there was no preexisting comprehensive database of hybrid choice policies at a national level or in our 18-state sample. While we conducted an exhaustive online search it is possible that we did not comprehensively identify existing hybrid choice pathways in our 18-state sample.

Our analysis of customer uptake of hybrid choice was based on publicly-available data. Our analysis is therefore only as comprehensive and accurate as the underlying data sources. Readers should consult the primary data sources to understand the limitations of each data collection process.

The hypothesis that hybrid choice could facilitate large-load integration requires further research. Large C&I loads with demand on the scale of whole cities remains a relatively recent phenomenon, and state responses to large-load growth are likewise nascent. New large loads are a distinct customer class in terms of their scale and their objectives. A key objective of data centers is so-called speed to power: accessing power as quickly as possible to rapidly deploy more data centers. The speed to power objective is strong enough that some new large loads are building off-grid natural gas generators to bypass lengthy grid interconnection processes (Hiller, 2025). The speed to power objective is distinct from the cost and sustainability objectives of C&I customers that can be met through existing hybrid choice policies. It remains to be seen whether and how policymakers could adapt hybrid choice policies to meet the distinct needs of new large loads. Future research could explore C&I customer uptake of these emerging hybrid choice policies and how the design and implementation of these policies can mitigate the potential impacts of increasing C&I demand. Future research could also explore how hybrid choice policies could help achieve other policy objectives such as improving grid resilience and achieving energy equity policy objectives.

Researchers and policymakers may consider ways that existing hybrid choice pathways in some states may be effectively transferred or adapted for other states. One potential research direction is the development of a transferability framework to rigorously identify how specific policies, regulations, and programs can be transferred between states.

Finally, researchers and policymakers may consider ways that distinct hybrid choice policies or other policies could be designed to simultaneously meet the distinct needs of existing C&I buyers and new large loads. Existing hybrid choice policies may continue to play key roles in enabling existing C&I customers to meet cost and sustainability objectives. In contrast to new large loads, existing C&I customers require far smaller power supplies, on the order of megawatts rather than hundreds of megawatts. The relatively smaller scale of existing C&I customers is more aligned with the sizes of typical solar and wind projects. For example, through the end of 2024, 86% of utility-scale solar projects in the United States were no larger than 100 MW (Seel et al., 2024). Further, existing C&I customers do not have the same speed-to-power constraints as some new large loads such as datacenters. Without speed-to-power constraints, existing C&I customers may be better positioned to support the development of new solar and wind resources, given that such development generally takes multiple years. As a result, existing hybrid choice policies may continue to enable C&I customers to support the development of new solar and wind resources, in particular. A distinct suite of hybrid choice or other policies may simultaneously help new large loads to meet their own distinct objectives. These policies would need to support the development of generation resources on the scale of hundreds of megawatts at a pace to meet speed-to-power objectives. Utility tariffs specifically designed for the risks of new large loads are one emerging solution. Other emerging solutions include utility green tariffs with enhanced non-participant protections (e.g., Nevada CTT), direct access for new large loads (e.g., Utah), and policies to support colocation of flexible DERs with new large loads (e.g., Georgia Power). Future research may explore emerging solutions for new large loads and how large-load strategies can complement hybrid choice policies for existing C&I customers.