A policy design perspective on electricity rates

Abstract

Electricity rates affect citizens’ energy-related behaviors. In the past and present, studies of the behavioral influences of the designs of these rates have heavily focused on the structure and effects of pricing mechanisms. This price-centric perspective has, this paper demonstrates, led other crucial aspects of rate designs to be overlooked despite their potential implication for more efficient and equitable citizen energy access and climate change mitigation. To explore the composition of rate designs beyond the oft-examined pricing component, this study develops and utilizes a novel application of policy design scholarship to more comprehensively examine the rate designs that facilitate residential electricity provision and shape energy consumption behaviors. Through this approach, multiple design elements and more granular sub-elements are identified in contemporary rate designs. Furthermore, emergent properties of rate design diversity and complexity are discerned as design features occur individually and combinatorically. Ultimately, this study not only tractably expands the concept of rate designs, but also advances a method for examining the policy instruments through which public service provision occurs.

Keywords:

• Policy design
• rate design
• public service provision
• electricity access
• energy behavior

1. Introduction

Alongside approaches such as rebates, rate designs have become instruments for the achievement of policy and social goals such as more equitable energy access or increased uptake of renewable energy technologies (Carley 2009; Convery, Mohlin, and Spiller 2017; Faruqui 2022). For example, in states where net metering policies (i.e. policies requiring utilities to allow and compensation customer-generation of electricity) are implemented, the presence of rates for net metering customers are influential on solar photovoltaic panel deployment as residents engage with this policy (Darghouth et al. 2016). Another notable example of rate designs as instruments deployed for policy purposes are rates for low-income households offered through the California Alternate Rates for Energy (CARE) Program which serve as instruments for policy responses to address inequity in residential energy access (Sherwin and Azevedo 2020).

Past research has underscored the ramifications of rate design for addressing important policy issues such as rates for specific household energy behaviors (e.g. electric vehicle (EV) usage) or for facilitating energy access among residents (e.g. inequitable burdens on low-income residents) (e.g. Berg and Drehobl 2018; Carley 2009; Convery, Mohlin, and Spiller 2017). Yet, much of existing rate design scholarship has primarily focused on the possible effects of rates’ cost-causal price mechanisms (i.e. pricing varying by the marginal cost of electricity delivery). A knowledge gap currently exists, however, in understanding whether other design aspects can be identified alongside the oft-examined pricing component and, if these features are present, how they occur.

The present study, therefore, looks around the pricing aspect for other identifiable compositional elements of rate designs. To do so, it draws on another “design” perspective – policy design scholarship - and its study of the composite aspects of policies that shape citizen behaviors. In intersecting the ideas of rate and policy design, this paper also demonstrates a novel application of a policy design typology toward the tractable identification and investigation of policy instruments, in this case, the electricity rate designs offered by investor-owned utilities (IOUs) in the United States (US). I focus on IOUs due to residential US electricity provision being predominantly provided by IOUs (Lindstrom and Hoff 2019) and to examine if, and how, IOUs are responding to policy considerations despite their private ownership (Curley et al. 2021). The remainder of this article reviews relevant existing literature, describes how rate designs were collected, coded, and examined, and discusses key takeaways from this research.

2. Understanding electricity rate designs as policy instruments

2.1. Conceptual origins of rate design

Bonbright’s (1961) seminal work on public utility ratemaking suggests electricity rates should be designed to be understandable and acceptable to citizens and feasible for utilities to implement. In one of the earliest examinations of rate designs, Solomon (1970) studied the influences of regulatory oversight on rate prices and utilities’ rates of return. Bonbright, Danielsen, and Kamerschen (1988) later leant support to such inquiry, conceptualizing rates as negotiated compromises of prices between multiple parties including utilities, regulators, interest groups, and citizens. Ultimately, in early characterizations of rate designs, “rate” and “price” were largely interchangeable insofar as the design of a rate was the price charged by the utility and faced by the customer. More recently, the concept of rate design has been revisited as researchers explore rate designs as ways to facilitate energy consumption behaviors that address critical problems like climate change, electric grid stress, and energy insecurity. These present-day considerations are reviewed next.

2.2. Contemporary rate design considerations: efficiency, reliability, the environment, and equity

Household energy consumption exhibits peaks across daily behaviors (e.g. cooking dinner) (Sanquist et al. 2012). These spikes can lead utilities to purchase energy from power producers at premium prices or to simply fail to meet customers’ needs (Yan et al. 2018). Over time, utilities can construct more energy storage sites, or if permitted, power plants, but incur large, fixed costs for developing and operating such facilities which likely raise customers’ bills (Muratori, Schuelke-Leech, and Rizzoni 2014). To avoid these costs, utilities have deployed rate designs (e.g. “time-of-use”, or “dynamic pricing” rates) to shift energy consumption toward off-peak hours through cost-causal pricing reflecting the marginal-cost changes required to provide electricity at that time (Joskow and Wolfram 2012). Consumers, to varying extents, consume less electricity when the marginal cost is higher (i.e. high-demand hours), thus reducing contributions to aggregate demand (Faruqui and Sergici 2010; Friedman 2011).

Additionally, high-demand hours can overload transmission infrastructure, causing blackouts, especially during temperature extremes (e.g. heat waves) (Auffhammer, Baylis, and Hausman 2017). Overloaded infrastructure can put utilities at risk of lawsuits over damages caused by overwhelmed systems (e.g. wildfires started by sparking transmission lines) (Kousky et al. 2018). These risks may motivate utilities to develop rate designs promoting energy efficiency; these design decisions are often supported by state policies offsetting potential lost profit from price differentials (e.g. decoupling policies) (Cleveland, Dunning, and Heibel 2019). Additionally, to reduce uncertainty surrounding energy consumption at given times and better anticipate cumulative demand, utilities develop rates tailored for specific uses to better predict and provide power through use-tailored designs (e.g. water heating) (Faruqui et al. 2016).

Depending on the state(s) where IOUs operate, policies may require and/or subsidize rate design offerings for renewable energy technologies (RETs) (e.g. residential solar panels) (Cleveland, Dunning, and Heibel 2019; Satchwell, Mills, and Barbose 2015). Renewable portfolio standards (RPS) and net metering policies are notable types of state policies spurring the rollout of environmentally-conscious rate designs. RPS set targets or thresholds for minimum shares of renewable energy within the state’s electricity mix by a designated time, thereby requiring utilities to diversify their energy sources or else pay a penalty (Lyon and Yin 2010; Yin and Powers 2010). Similarly, net metering policies require utilities to make on-site customer-generation opportunities and associated compensation available to customers (Carley 2009; Smith, Koski, and Siddiki 2021). Rate designs for RETs have been found to increase the likelihood of residential engagement with these innovations (Carley 2009; Wolsink 2012).

While environmental policies can facilitate household energy transitions, Carley and Konisky (2020) note how RETs require costs that utilities likely integrate into customers’ bills. These policies, and the rate designs they encourage, can disproportionately affect households already spending larger shares of income on energy needs, especially those without surplus income to absorb higher bills, underscoring the need for equity-driven rate designs (Graff et al. 2021; Hernández 2016). Equity considerations are also important when rate designs involve time-varying pricing as disadvantaged households may not have the capacity to respond to price differentials and subsequently incur higher price burdens (Calver and Simcock 2021; White and Sintov 2019). Equity-focused approaches to ratemaking, therefore, are crucial as rate design considerations intersect with policy aims and implementation. Without consideration for the interplay between public policies and the rate designs they target, as states continue to enact initiatives promoting energy transition and efficiency, the risk and magnitude of the inequitable burden on energy-insecure households grows. Accordingly, some states have designed policies focused on rates customized for the energy needs of low-income households by requiring adjusted rate prices and/or subsidized participation in energy-efficiency programming (Baldwin and Felder 2019). A variety of policies have been enacted; most mandate utility-spending thresholds for low-income household programming and cost-effectiveness testing for such programs (Brown et al. 2020; Berg and Drehobl 2018).

2.3. Rate designs as policy instruments

Electricity rate designs facilitate access to, and provision of, a crucial public service and possess the capacity to affect residents’ consumption levels of this service. Even a generic, standard rate for all household uses and all households in a utility’s service area charging a flat price for consumed electricity is likely to affect household behaviors. For example, in response to a high electricity bill, a resident may modify how they use energy in the following month(s) by running the air conditioner less frequently or being more mindful to turn off the lights when leaving a room; these cost-reduction strategies are likely to be more drastic, potentially to the point of putting residents’ health at risk, for energy-insecure households (Carley et al. 2022). As contemporary considerations for rate designs become more prevalent, however, rates have emerged with features that utilities, policymakers, or both intend to modify facets of residents’ energy consumption behaviors whether it be how much energy they can afford (i.e. low-income programming), (Graff et al. 2021), how much they consume (e.g. time-of-use rates) (Joskow and Wolfram 2012), how energy is used (e.g. EV charging) (Faruqui et al. 2016), or how energy is sourced (e.g. net metering) (Carley 2009). Additionally, as part of facilitating vital public service provision, rate designs can contribute public benefits through the production of environmental public goods (e.g. pollution reduction through EV use) (Kotchen 2006; Wolsink 2012; Sexton and Sexton 2014) and the adequacy and reliability of energy systems, particularly where energy co-production occurs through residential RETs (Joseph 2022; Wolsink 2012, 2020).

Individually and collectively, these contemporary considerations for rate design suggest that rate designs exhibit functional similarities to public policies. Rate designs, like the policies that can spur their creation, are implemented for the fulfillment of goals promoting enhanced public service and benefit. In achieving these aims, rates, like policies, often focus on modifying the behaviors of individuals or groups (e.g. lowering energy consumption during peak times, charging EVs). Through the study of policy design, scholars have already begun to explore the composition of policies that enable or potentially inhibit their purpose(s). Given the functional parallels between policies and rates, this work proposes to utilize a policy design perspective to more tractably examine the composition of rate designs.

2.4. Studying rate designs through a policy design typology

In this research, policy design refers to the content of a policy as opposed to the process by which a policy is created (Schneider and Ingram 1997; Siddiki 2020). Schneider and Ingram (1997) develop a typology of six elements of policy design – (1) goals, the condition(s) a policy intends to achieve; (2) targets, the individuals or groups to whom a policy applies; (3) tools, the mechanisms utilized to modify target behaviors; (4) implementing agents and/or structures, the officials and infrastructure that administer tools to targets; (5) rules formalizing a policy’s implementation; and (6) underlying assumptions, rationales, and logics linking the prior elements together – that commonly constitute policy designs. Ultimately, policy designs have a behavior-altering focus that runs through these various elements. Policymakers set policy goals in relation to current and desired states of behaviors, or the outcomes of behaviors, often performed by individuals and groups who become the target populations of policy designs. They select a policy tool and corresponding agents and/or structures for the tool’s implementation based upon information and assumptions that these design features are appropriate to modify a given behavior. Finally, policymakers set rules that prescribe how a policy will be administered to a target population and its behavior(s) or that define behaviors through which targets engage with a policy.

Although not explicitly packaged in these terms, the similarities shared between policies and rates regarding behavior-modifying capability lead to identifiable alignment of elements of policy designs as identified by Schneider and Ingram (1997), particularly goals, targets, and tools, with features and considerations of rate designs. Leveraging this alignment, this research examines contemporary rate designs through an application of Schneider and Ingram’s policy design typology that examines collected rate data for the presence of similar design elements (i.e. tools, targets, and goals). To structure this inquiry, the following three research questions were asked: (1) Are other features present in rate designs other than the pricing component examined extensively in existing research? (2) What rate designs are IOUs offering citizens? and (3) Do any emergent patterns or properties appear in these rate designs? To address these questions, IOU rate data was collected and coding protocols developed to analyze their designs. Ultimately, through this application of a policy design perspective to rate designs, a novel method for tractably identifying and examining rate designs is developed and engaged; this approach is described next.

3. Collecting and coding rate designs

All rates offered by IOUs in the United States in February 2020 were retrieved from the National Renewable Energy Laboratory’s Utility Rate Database which contains records of rate adoptions, eliminations, and changes as early as 1969. The collection point for this cross-sectional data was selected to avoid possible effects of the COVID-19 pandemic on utility decision-making. Overall, 1114 residential electricity rates were collected from 139 IOUs in 45 states. The earliest- and latest-adopted rates in the dataset were in 1969 and 2020, respectively. The average rate in the constructed dataset was adopted in 2014. Although almost 80% of contemporary rates replaced existing designs; these changes occurred, on average, within 13.2 months of the original design’s rollout. Only 26 rates (2% of the dataset) were adopted in 2019 or later thus suggesting some stickiness over time by the collected designs.

A rate design coding protocol was developed deductively from policy design concepts and inductively through an examination of the collected rates. This protocol engages three corresponding policy design elements (i.e. policy targets, tools, and goals) to derive three rate design elements (i.e. rate targets, tools, and goals). Tables 1 and 2 contain definitions and examples for policy and rate design elements, respectively.

Table 1. Rate-relevant policy design elements.

Policy design element	Definition	Examples
Policy Target(s)	Individuals or groups whose behavior is intended to be affected by a policy	Low-income households; Small businesses
Policy Tool(s)	Instruments implemented with the intent of modifying target behavior	Tax credits; Fines
Policy Goal(s)	Desired state(s) or outcome(s) that a policy intends to achieve	Emission reductions; Increased program enrollment

Table 2. Extrapolated rate design elements.

Rate design element	Definition	Examples
Rate Target(s)	Household behaviors or characteristics that a rate intends to affect	EV charging; Apartment residents
Rate Tool(s)	Pricing mechanism applied to energy consumption behavior(s)	Flat rate, TOU pricing
Rate Goal(s)	Aim(s) that a rate intends to achieve	Efficiency, equity

Additionally, rate design sub-elements, described in Table 3, were identified which allowed for an examination of rate designs extending beyond noting the presence or absence of elements. Furthermore, for rate targets, classes of sub-elements were discerned – household energy uses, residential user types, and service locations. Rate tools, meanwhile, were categorized by sub-elements for flat pricing, time-of-use (TOU) pricing, and pricing structures varying by a factor other than time. Finally, rate goals were differentiated by efficiency, environmental, and/or equity aims identified by Convery, Mohlin, and Spiller (2017) – efficiency (i.e. optimizing energy transmission and consumption), environmental (i.e. facilitating pollutant abatement and/or clean energy transition), and equity (i.e. lowering burdens of disadvantaged households). As explicit goals of were not usually described in a rate’s recorded description, an additional protocol, demonstrated in Figure 1, was developed to assign goals to nonstandard rates (i.e. rates serving any use, customer type, or location with a flat price) using criteria reported in Table 4. Figure 2 shows a goal-coding example.

Figure 1. Goal derivation example.

Figure 2. Rate coding example. “Single-family accommodation” was not coded as occupancy-specific as this designation is considered to include homes, mobile homes, and apartments.

Table 3. Rate design elements and sub-elements.

Rate design element	Rate design sub-elements
Rate Targets* *Classes are underlined*	Uses: heating, EV charging, dual fuel (combined gas-and-electric usage), energy storage Users: Occupancy-specific-housing residents (e.g. apartments); low-income households, extra-residential use households (e.g. farmers), clean-energy-program participants, energy-efficiency program participants Locations: rural or urban areas, select regions or municipalities
Rate Tools	Flat Pricing: no price change TOU Pricing: price variation depending on time of day TOU + Pricing: price variation dependent on more than time (e.g. demand)
Rate Goals	Efficiency: facilitating energy-efficient behaviors Environmental: facilitating environmentally-favorable behaviors Equity: facilitating less burdensome energy access for disadvantaged households

Table 4. Rate design goals.

Rate goal type	Inclusion criteria
Efficiency	Non-flat pricing; rates for efficiency-program participants, occupancy-specific housing, extra-residential users, and specific locations
Environmental	Rates for EV charging and clean-energy-technology households
Equity	Rates for low-income households

4. Examining IOU rate designs

4.1. Rate designs are more than just prices

As shown in Table 5, just under 75% of all collected rates contained at least one design element. Of these rates, almost half did not include the oft-examined pricing tool, but rather, engaged some other identified rate design element. Additionally, looking across all nonstandard rates, almost three-fourths contained a target use, user, location, or some combination of these classes. Table 6 displays the frequencies of each target class of which target locations were the most common. Collectively, these results demonstrate that more exists to rate designs than just the pricing tools upon which existing research has frequently focused. Furthermore, the identification of different frequencies of target classes suggests that diversity exists in what constitutes these designs.

Table 5. Design element frequencies.

Design element	Share of total rates	Share of rates with design element
Tool	40.13%	53.92%
Target	53.68%	72.14%
Goal	74.42%

Table 6. Target class frequencies.

Target	Share of total rates	Share of targeted rates
Use	19.66%	36.62%
User	20.11%	37.46%
Location	26.12%	48.66%

4.2. Diversity exists among and within design elements and sub-elements

The second analysis examines frequencies of the sub-elements of design elements – and explores occurrences of these sub-components across the larger design elements – tools, targets, and goals. Table 7 reports non-flat pricing tool frequencies and indicates that most vary only by time. Table 8 displays findings regarding target sub-elements. Beginning with the use sub-elements, most targets focus on one behavior, predominantly heating or EV charging. Two-use-target rates for heating and energy storage are the third most common occurrence.

Table 7. Rate tool sub-elements.

Sub-Element(s)	Share of tooled rates
TOU Price Variation	85.91%
TOU + Price Variation	14.09%

Table 8. Rate target sub-elements.

Use Sub-Elements
Number of uses	Sub-Element(s)	Share of targeted rates
1 Use	Heating	64.84%
	EV Charging	24.66%
	Dual Fuel	1.83%
2 Uses	Heating and Energy Storage	5.94%
	Heating and Dual Fuel	1.37%
	Heating and EV Charging	1.37%
User Sub-Elements
Number of User Specifications	Sub-Element(s)	Share of Targeted Rates
1 Specification	LI Households	30.36%
	OSH Residents	17.86%
	CEP Participant	16.07%
	ERU Household	4.46%
	EEP Participant	4.46%
2 Specifications	OSH-ERU	18.30%
	CEP-ERU	1.34%
	CEP-EEP	0.89%
	OSH-CEP	0.89%
	LI-CEP	0.45%
	LI-OSH	0.45%
3 Specifications	LI-OSH-ERU	3.57%
	LI-CEP-EEP	1.80%
Location Sub-Elements
Sub-Element(s)		Share of Targeted Rates
Specific Region/Municipality		94.00%
Rural/Urban		6.00%

Acronyms for Target User Specifications are as follows: LI: Low-Income; OSH: Occupancy-Specific-Housing; CEP: Clean Energy Program; ERU: Extra-Residential Use; EEP: Energy Efficiency Program.

Additionally, Table 8 shows that utilities fine-tune rate designs’ user-targets to create rates for specific subsets of their customer bases. Rate designs with only one user specification most frequently focus solely on low-income households. Other common designs were one-user-target rates for occupation-specific-housing customers and clean-energy-program participants as well as two-user-target designs for residents meeting both criteria. An extremely small number of rate designs contained a target specifying three user-sub-elements, most intended for low-income residents in occupancy-specific buildings with extra-residential energy needs, (e.g. low-income customers in non-profit-provided housing). All other high-specification rates were designed for low-income residents enrolled in energy efficiency and clean energy programs, (e.g. initiatives facilitating energy transition among disadvantaged households). Within the location target class, two sub-element types were identified - customers living in rural or urban areas and customers in a specific regions or municipalities, who were targeted far more often.

Finally, goal sub-elements, shown in Table 9, demonstrate a combinatorial nature similar to how two or three target uses or users specifications co-existed in rates. Notably though, efficiency goals dominated this aspect of rate designs. Efficiency-only rates constituted almost three-fourths, followed by efficiency-environmental designs, and then efficiency-equity designs.

Table 9. Rate goal sub-elements.

Number of goals	Sub-Element(s)	Share of rates with identified Goals
1 Goal	Efficiency	78.29%
	Equity	3.98%
	Environmental	2.53%
2 Goals	Efficiency; Environmental	9.53%
	Efficiency; Equity	5.07%
3 Goals	Efficiency; Environmental; Equity	0.60%

The examination of design sub-elements indicates that diversity exists in the more granular aspects of rate elements utilized by IOUs, particularly in target uses and users. This design sub-element diversity highlights utilities’ deployment of use- and user-target combinations to tailor the behavior and/or customer type that a rate targets.

4.3. Compositional complexity occurs as design elements and sub-elements are combined

Looking across combinations of tools, targets, and their sub-elements (goals were not examined as combinations of these elements are already reflected in target-tool combinations), different degrees of rate design complexity are identifiable as increasing numbers of elements, and their various sub-elements, are simultaneously engaged. As each additional design element is incorporated into a rate design, the rate moves further away from its simplest form, a nonstandard rate with no target use, target user, target location, or pricing tool. Alternatively, a rate with all four of these elements present would constitute a design with more compositional complexity. To explore these differences, rates were aggregated first by their potential to exhibit particular levels of compositional complexity before examining how complexity took shape among the collected rates. This complexity analysis begins by examining rates with the potential to have all four identified elements – a target use, user, and location, as well as a pricing tool; these rates must have, at minimum, both a target use and a target user as these elements were the most infrequent.

As shown in Figure 3, only 39 of 829 rates included target uses and users, almost all being for heating purposes and most for either low-income- or occupancy-specific- households. Of this small portion, only 6 rates exhibit the highest level of compositional complexity by containing all four design elements. These 6 rates were for heating uses by low-income users in specific regions or municipalities under TOU pricing schedules. Overall, then, most of the rates that could have exhibited high compositional complexity did not as most use-and-user target combinations were accompanied by general location-targets or flat pricing tools. While some of these rates contained multi-criteria user-targets, such as low-income-occupancy-specific households, none of these rates contained multi-use targets.

Figure 3. Patterns across rates with the highest potential design complexity. Acronyms for Target User Specifications are as follows: LI: Low-Income; OSH: Occupancy-Specific Housing; CEP: Clean Energy Program; ERU: Extra-Residential Use; EEP: Energy Efficiency Program.

Descending to a lower level of compositional complexity in rate designs (i.e. rates with three possible design elements - a pricing tool, a location-target, and either a target use or user), use-targets for single and multiple household behaviors appear in the absence of user-targets. Figure 4 shows that use-targets for household heating were, once again, the most common single-use target in this type of rate design, and that these use-targets were also present in all designs that involve two-use targets (i.e. heating and energy storage, heating and EV charging, and heating and dual fuel). Single-use targets for EV charging also constituted a meaningful share of these target-use-only rate designs. Regardless of their particular target uses, however, these designs contained greater shares of location-targets and pricing tools than the previous design type.

Figure 4. Patterns across target uses, Locations, and pricing tools.

Meanwhile, among rate designs with target users but no target uses, much more variation is present among user sub-elements. In addition to a wider array of sub-elements being engaged for single-user targets (shown in Figure 5), more multi-criteria targets are present that advance two or three user specifications (described in Figure 6). Across all combinations of user-target sub-elements, low-income residents are rate designs’ most frequent focus, followed by households with extra-residential uses (e.g. farming). Roughly the same proportion of rate designs include location-targets in comparison to target-use-only rates, but the presence of pricing tools was much lower among rates that only contained user-targets.

Figure 5. Patterns across single-user targets, Locations, and pricing tools. Acronyms for Target User Specifications are as follows: LI: Low-Income; OSH: Occupancy-Specific Housing; CEP: Clean Energy Program; ERU: Extra-Residential Use; EEP: Energy Efficiency Program.

Figure 6. Patterns across multi-user targets, Locations, and pricing tools. Acronyms for Target User Specifications are as follows: LI: Low-Income; OSH: Occupancy-Specific Housing; CEP: Clean Energy Program; ERU: Extra-Residential Use; EEP: Energy Efficiency Program.

Moving to yet another level of lower compositional complexity, rate designs with the potential for location-target and pricing tool combinations are shown in Figure 7. These rates overwhelmingly targeted specific regions or municipalities within utilities’ service areas. Pricing tools were present in over a third of these rates.

Figure 7. Patterns across target Locations and pricing tools.

Finally, Figure 8 describes the rate designs with the least amount of potential complexity in which only the presence of one element, a pricing tool, was possible. These rates predominantly utilized TOU-pricing mechanisms. It is nonetheless interesting that this type of low-complexity rate design, not combined with any other design element or target class, exhibited the greatest involvement of non-flat pricing that varied by some dimension other than the time of day.

Figure 8. Occurrence of pricing tools.

4.4. Rates with high compositional complexity are usually found in more complex rate portfolios

To better understand the specific context within which compositionally complex rate designs occur, five types of rate designs with substantial complexity among elements and/or sub-elements were examined: (1) rates with all target classes (i.e. use, user, and location) and a pricing tool; (2) rates with a use-target, a location-target, a pricing tool, but no user-target; (3) rates with a user-target, a location-target, a pricing tool, but no use-target; (4) rates with a target use and a two-criteria user-target; and (5) rates with a target use and a user-target with three specifications. For each design type, utilities offering these rates in the dataset were identified and their entire rate portfolios subsequently examined to get a sense of how rates with complex compositions are being deployed within their larger rate portfolio. alongside those with high compositional complexity. The average number of components (i.e. target classes and/or a pricing tool) was calculated, with and without the presence of a pricing tool, to measure a rate’s compositional complexity. The average compositional complexity of an IOU rate design was found to be 1.10 components (standard deviation: 0.94) when pricing tools were included, and 0.67 component (standard deviation: 0.77) when tools were excluded. These averages are relatively unsurprising given the earlier analyses that indicated that contemporary rate designs tend to avoid combining multiple components.

Interestingly, it would appear, as shown in Table 10, that the utilities offering rate designs with high compositional complexity roll out rate portfolios that seldom contain rates consisting of no design elements. That is, the utilities in the dataset with the most compositionally complex rate designs, rarely had a generic, unspecified, flat rate. Furthermore, three of the seven studied utilities offer portfolios within which over half of the rates possess above-average compositional complexity. When pricing tools are considered, all but one of the utilities have portfolios consisting almost entirely of rates with above-average compositional complexity. Although an investigation of many more utilities’ rate portfolios is needed to explore trends in portfolio complexity, these cases provide a glimpse into how individual utilities deploy rate designs and suggests that high-complexity rate designs are not outliers in utilities’ portfolios and that such portfolios may be comprehensively more compositionally complex.

Table 10. Compositional complexity in utility rate portfolios.

Utility name	State	Rates in Portfolio	Rates with design elements	Rates with specified element complexity	Rates with above-average compositional complexity (excluding pricing tools)	Rates with above-average compositional complexity (including pricing tools)
Utilities offering rate designs with all target classes and a pricing tool
Eversource	MA	32	32	5	23	32
Pacific Gas & Electric Co.	CA	68	68	1	24	67
Utilities offering rate designs with a target use, target location, and a pricing tool
Maui Electric Company	HI	19	19	3	13	19
Utilities offering rate designs with a target user, target location, and a pricing tool
Northwestern Wisconsin Electric Co.	WI	15	15	2	11	14
Pacific Gas & Electric Co.	CA	68	68	1	24	67
Utilities offering rate designs with a target use and a two-sub-element user-target
Minnesota Power Inc.	MN	8	7	4	2	6
Utilities offering rate designs with a target use and a three-sub-element user-target
Entergy Mississippi Inc.	MS	4	2	2	0	2
Minnesota Power Inc.	MN	8	7	2	2	6

5. Discussion of findings and implications for rate and policy design

Across this study’s various analyses, some noteworthy takeaways emerge. First and foremost, this article demonstrates that policy design conceptualizations can be effectively extrapolated and applied to study policy instruments, in this case, rate designs. Through this application, rate designs are shown to be constituted by more than the pricing mechanisms on which existing scholarship has predominantly focused. The identification of rate targets and goals, as derived from studies of policy designs, in addition to pricing tools allows for a more in-depth understanding of how IOUs engage in energy provision. Moreover, this expansion of rate design conceptualization allows for a more thorough examination of how utilities facilitate electricity access for citizens.

Furthermore, through this examination of rates designs, a few implications emerge regarding the diversity and complexity of designs’ elements and sub-elements. First, the identification of design sub-elements suggests the rate design landscape contains compositional variation, particularly among target uses and users. IOUs’ decisions to deploy targeted rate designs for specific uses and/or users, however, are not as common as choices to offer rates with pricing tools and rates for specific service area locations. Given the grid-reliability benefits that location-specific and time-varying rates can afford utilities, it appears IOU rate-designing puts something of a premium on efficiency. Furthermore, these utilities seem to exhibit a preference for avoiding higher degrees of compositional complexity in the rates they offer, echoing Bonbright’s (1961) suggestions for rate simplicity. As the number of elements constituting rate designs increased, the frequency of such offerings decreased. Similarly, sub-element combinations such as multi-use targets, were also much less common than singular ones. Additionally, when no targets were specified, pricing tools were much more common, possibly suggesting reluctance among utilities to tie the timing of energy consumption and its marginal price to smaller subsets of their customer base and their behaviors.

The lower volume of more complex rate designs may occur due to the burdens of administering more complex sub-element combinations (e.g. the stress placed on citizens’ shoulders to understand multiple target specifications, the demands placed on utility employees to effectively manage customer needs for these specific designs). Alternatively, IOUs may not deem complex rate tailoring worthwhile as the cost (e.g. meter installation and inspection, too few eligible customers) of offering such designs may not produce adequate revenue. Thus, some complexity “threshold” may exist beyond which further design specification is impractical or unfeasible for IOUs. Additionally, in the absence of outright policy mandates for particular rate designs, utilities may forgo complex designs in favor of rolling out other programming such as rate riders for particular uses or rebates for RET purchases.

Another noteworthy finding on rate design composition is the lack of rates facilitating equitable energy transition. While rates with environmental goals are present, only 5 rates were identified as advancing RET engagement (e.g. EV charging, clean energy programs) while simultaneously being tailored for disadvantaged households. The widespread absence of more environmental designs with equity considerations suggests that rate designs are not extensively utilized for enabling clean energy transitions among disadvantaged populations, though other utility programs might be offered for this purpose.

Finally, although this article has centered around the application of a policy design typology to rate design content, it offers insights into how policy design concepts can transcend from the examination of policy designs’ content to the output content of policy designs. Siddiki (2020) delineates functional, structural, and syntactic levels of analysis within policy design content and notes that various theories and typologies can be utilized to explore intra- and inter-level complexity within and among different policy designs. This work devotes attention to the instrumental outputs of these layered policy designs. In doing so, this research introduces and demonstrates the viability of an approach for exploring the content of instruments associated with various policy designs (e.g. net metering policies). Engagement with this approach could allow scholars to more tractably examine and compare the outputs of the implementation of different policies through examinations of the content of both policies and their respective instruments. Moreover, by comprehensively identifying and analyzing differences in these types of policy-related content, researchers and practitioners can explore if, and potentially how, variation in the content of policies and/or their instruments affects public engagement with these designs (e.g. enrollment in TOU rates).

6. Conclusion

Rate design scholarship has overwhelmingly studied electricity rate through a price-centric lens, narrowly considering how these pricing tools are developed and implemented within the economic, financial, political, and policy contexts of utility decision-making and public service provision. By leveraging a policy design perspective, this study offers an expanded conceptualization and subsequently arrives at a more nuanced understanding of, the composition of rate designs. In addition to offering readers a more granular view of contemporary IOU rates that demonstrates the diversity and complexity present in these designs, this research develops an approach for examining the instruments tied to policy content.

Future utilization of this approach could afford readers with an analytical tool for the nuances of, and potential interplay between, policy design content and policy instrument content. This stream of inquiry could be useful for more comprehensively understanding the ramifications of differences between public policies and the engagement of actors (e.g. utilities, residents) with these designs. Such work will be of growing importance and interest as concerns over climate change, electric grid reliability, and energy insecurity grow, and responsive policy designs and their instruments are implemented.

Disclosure statement

No potential conflict of interest was reported by the author(s).