The primary question being addressed through this conference is how internal operations and regulatory policies of utilities and the regulatory agencies will be impacted in a carbon-constrained future. Setting aside for the moment the question of whether constraining carbon is the right policy direction, the question becomes "how" the carbon constraining policies can be most economically implemented. Two primary directions for the industry under a carbon constrained future are: 1) utilize technology advancements to reduce the carbon intensity of how new power supply is provided to consumers; and, 2) utilize advancements in energy efficiency ("EE") to allow consumers to generate the same level of energy 'utility' out of a reduced amount of energy consumption.1size="1"> This paper provides thoughts from the electric utility's perspective in terms of operating Demand Side Management ("DSM") programs as a product line within their portfolio with an overall goal of maximizing shareholder returns. As a key player in any push to grow EE as a method for reducing carbon, the actions of the utilities will play a central role.

EE has a long track record of playing a significant role in the energy market but the impact of utility-based Demand Side Management programs is less clear. The ACEEE brings forth data indicating that advances in energy productivity and EE have in fact played the dominant role in our energy markets since 1970. Using the relationship between energy production and total economic output as a measure of energy intensity, the ACEEE paper shows that the US economy has reduced the energy input requirement (in BTUs) per dollar of gross domestic product by roughly 50%.

"[E]nergy efficiency has contributed more value to the economy in recent decades than any conventional energy resource, meeting three-fourths of all new demand for energy services since 1970" (Laitner 2008). [E]stimates for 2008 indicate that by the end of this year U.S. energy consumption per dollar of economic output will have declined by more than 50% since 1970, from 18,000 Btus to about 8,900 in 2008. As such, current levels of energy consumption in the U.S. are only half of what they would have been if levels of energy services, the structure of the economy, and overall energy productivity had remained unchanged."2size="1">

While a portion of this savings in energy intensity comes from structural changes like the movement away from manufacturing to more of a service-based economy, much of the energy savings likely also resulted from growth in organic EE.3size="1"> The next question is what role the electric utilities have played in the overall growth of EE and what can be done to further their contribution going forward.

Data from the EIA suggest that utility-based DSM programs and investments have had modest impacts on peak demand relative to total industry demand. Since 1996, utility-based DSM programs have been generating peak demand reductions at a fairly steady rate of around 3% of total generating capacity.4size="1"> While the utility-based programs may have significantly impacted the acceptance and growth of EE programs by acting effectively as a sales channel promoting efficiency to the customer base, the overall financial roll of the utilities has remained small.

Differences in growth between total net summer capacity and DSM peak impacts are also significant. Net summer peak capacity has increased by 28.7% over the period 1996 to 2007 while DSM peak impacts have remained largely flat with a total growth of only 1.3% over the same period. These numbers also support a conclusion that the utility-based programs, have played a relatively minor role in the overall EE market.

The topic of how best to provide regulatory incentives for DSM growth has been well covered through research projects like Aligning Utility Incentives with Investment in Energy Efficiency published by the EPA in 2007. The goal for this paper over the remaining pages is to discuss how viewing DSM as a product within the utility portfolio impacts decision making and to provide a current example from Florida on how DSM programs impact utility profits under traditional ROR regulation.

DSM as a Product within Utility Product Portfolios

As noted above, DSM EE programs allow customers to receive the benefits of energy consumption at a lower level of usage. The challenge lies in getting customers to make the upfront decision to invest in EE technology in order to enjoy the benefits of lower bills in the future. It is logical to view these DSM programs as providing customers the same level of benefit from energy consumption using a different product. In this respect, DSM is a product alternative within the overall energy market that competes against a decision to simply buy more KW hours provided through fleet or purchased power generation.

Looking at DSM programs as products provides a useful paradigm for understanding incentives and behavior. Like all products, the primary decision point comes from the end user: whether it is an industrial plant or a residential customer. In the case of DSM, end users are making decisions on trading off investments in energy savings through technology in exchange for reductions in bills going forward.5size="1"> Because investor owned utilities are in business to earn profits for shareholders (among other goals), it is necessary to understand the impact of this product substitution on overall utility profits in order to help interpret the position of utility-based DSM programs relative to the larger EE industry.

A slight twist in the analysis of DSM as a product in the utility's portfolio is that utilities are really only facilitating purchases of products rather than actually providing the end product itself. A standard DSM program that provides a monetary incentive for the purchase of newer technology is really only allowing the utility and regulatory agencies (through policy) to participate in the true product transaction which is the purchase of new or substitute technology. A simple example is the case of an incentive to replace old air conditioners where the actual product transaction involves the purchase of an air conditioner.

In spite of the rather tangential involvement of the utility in the transaction, the roles the utility plays in managing DSM like a product in its portfolio are critical for success. Basic product management and product marketing roles including marketing to educate customers on the program, vetting and training contractors that do install work, managing the customer experience from end to end, and regulatory and financial reporting all require sufficient attention from the utility to ensure success. The underlying challenge is that companies in a market economy are going to weigh their choices on how to spend limited resources based upon their returns on investment and DSM as a product will remain underfunded relative to its potential so long as its returns lag providing more kWh to customers.

In looking at utilities as 'total energy' companies in business to provide customers energy benefits either through both kWh or EE, utilities need to determine how much attention to focus on each product in their portfolio. While the paradigm for earning shareholder returns from selling kWh are well defined, the shareholder impacts from providing more EE through DSM programs are more elusive.

DSM Programs Drive Down Earnings Under Traditional ROR Regulation

A classical theory in the economics literature of firm behavior when under ROR regulation states that firms will maximize the capital assets in their rate base in order to maximize earnings.6size="1"> In a simplified view, overall revenues and the total volume of profits for the traditional ROR regulated utility are a function of the quantity of debt and equity along with the rates of return allowed for each asset class. As the quantity of assets that are declared used and useful and admitted into the rate base increase, so does the total volume of revenues and allowed profits.7size="1">

With DSM functioning as a substitute for additional energy consumption, DSM reduces the level of peak demand and thereby reduces the level of utility investments in generating capacity. The end result is a lower rate base, reduced revenues, and a smaller total quantity of returns on assets. The concept of 'uncoupling' revenues from units of KW energy sales does not necessarily resolve this challenge as future growth in revenues associated with additional investments in fleet generation are still impacted by DSM programs.

A simple current example of a utility proposing a DSM program that provides an addition 750MW of summer peak capacity over a 10-year planning horizon can help demonstrate this point. Assuming that the utility would build a 750MW plant in year 5 absent the DSM program and would utilize existing margins and purchased power to meet power needs until the plant is built, and assuming an overall ROR of 10% for simplicity. Under this scenario, the company will experience a reduction in net income of approximately ($140M) over the 10 year period due to the avoided rate base increase.8size="1"> While there is likely a fair degree of variance around this number, the direction of the impact on earnings and the order of magnitude should be reasonable. As such, the impact on utility incentives in terms of pushing DSM programs aggressively should also be clear.

Regulatory Policy Must Level the Financial Playing Field for DSM as a Product

Assuming the case has been made that utilities should be incented to participate in the energy market as providers of both energy and energy efficiency, the question becomes how to level the financial playing field. There are many methods currently in place for controlling how utilities participate in the DSM market. Some programs provide cost neutrality for DSM expenditures while other programs allow utilities to share in the net benefits associated with the DSM program or allow utilities to earn a premium above the authorized ROR on the current rate base. While a survey of existing DSM incentive policies is has been well covered and is beyond the scope of this paper, some examples of benefits and concerns is useful.

Regulatory approaches that track total DSM program costs and recover the overall DSM expenditures on a one-for-one basis through a general rate adder that is periodically adjusted or trued-up should be avoided as these policies provide a disincentive for DSM. In this case, there is a negative financial impact on the utility. In some cases, avoiding a penalty for missing DSM goals remains the only true financial incentive. DSM programs handled in this fashion will reduce the magnitude of overall earnings over the long run while providing profit neutrality over the short run period in which the DSM outlays are recovered. This result effectively provides a negative incentive for utilities to grow DSM programs as a product in their overall energy portfolio. The utility's view of the optimized level of DSM under this approach is zero from a long run earnings perspective.

A different approach allows utilities to earn higher ROR based upon success of their DSM products in the market. There are differing levels of complexity and likelihood for stimulating DSM demand depending on how these regulatory programs are designed. An example of one form of ROR adder program to stimulate DSM implementation was built into the Florida statutes in 2008.9size="1"> Under the Florida statute, utilities will benefit from an additional 50 basis point adder on their authorized ROR if the utilities are able to show that 20% of overall demand growth in their base is covered by 'energy efficiency and conservation measures'.10size="1">

There are two important aspects of the Florida incentive to consider: 1) impacts of the Florida incentive will vary depending on the timing of the avoided generation investment; and, 2) the 50 basis point adder will put the utilities into a negative position once the avoided CAPX exceeds around 5% of the total rate base (assuming a 10% ROR). The first impact is easily shown graphically in Figure 2.11size="1"> In the case depicted here, the utility benefits from the 50 basis point adder during all 10 years and suffers the reduction in the rate base (relative to a baseline that would have included a new plant) in years 5-10. Because the size of the avoided plant addition in this hypothetical would create an avoided CAPX impact of just under 5% of existing rate base, the impacts of the DSM program under this 10-year planning view never turn negative.12size="1"> The cumulative impact is heavily affected by the assumption on which year the avoided plant would have been completed and admitted into the rate base.

This second impact creates an interesting tension between short run and long run financial goals for the utility. A 50 basis point ROR adder dependent upon success in growing DSM programs going forward (the Florida plan) creates a short run incentive for additional returns while the longer term impacts of high DSM penetration exceeding 5% of the asset base creates a disincentive.13size="1"> Figure 3 attempts to show this dynamic with a new view of the 10-year planning horizon only now including 1800MW of existing DSM.14size="1"> As can be seen in this example, the utility continues to benefit over the first 5 years from the adder, but at a reduced rate now reflecting the rate base reductions associated with the existing DSM. When the additional 750MW of avoided fleet generation investment are added in year 5, the financial impact turns negative and quickly leads to a cummulative net negative impact. Given this dynamic, a more holistic review that considers total DSM in place as a percentage of the rate base is needed to determine whether the incentive effectively levels the playing field for the totality of DSM in the utility's base. The challenge created here is that the holistic and dynamic approach which includes analysis of the timing of past DSM implementation, the amount of avoided CAPX, assumed depreciation levels, will likely be more difficult and contentious to implement.

Go For Simplicity

Allowing utilities to classify DSM investments as regulatory assets that are included in the rate base presents a simpler approach to addressing the impacts of DSM programs on utility shareholders while incenting utilities to participate effectively in the EE market. Allowing a premium return on DSM expenditures will incent utilities to view DSM as a more profitable product and allow the industry to recognize the current cost advantages of EE. The Duke Power "Save-A-Watt" proposal is an example of this approach where the avoided CAPX in energy and capacity is amortized over the lifetime of the EE measures with the utility able to charge ratepayers a return on the un-depreciated avoided "investment". Categorizing direct investments in DSM programs and in supporting technology like AMI as regulatory assets will allow utilities and regulators the relative comfort of using the regulatory monitoring and compliance structures that have been in place for many decades.15size="1"> This level of familiarity with the rules and procedures will inherently make this regulatory approach simpler to implement.

There will, however, continue to be challenges under this approach with the biggest challenge being the assessment of whether investments are appropriate for inclusion in the rate base. Investments in DSM programs that have been thoroughly vetted and pass the standard tests for cost effectiveness and consumer impacts (e.g., RIM and TRC) should have little difficulty in entering the rate base as regulatory assets. Newer technologies like Smart Grid and AMI, which will be required to fully exploit DSM going forward, but are aimed at supporting DSM as opposed to providing the DSM 'product' itself, will have a more difficult time passing cost/benefit analysis necessary to enter the rate base. EISA 2007 begins to address uncertainty around investments in smart grid and AMI by directing states to "consider" requiring utilities to assess a wider variety of cost/benefit impacts associated with these investments.16size="1">

The case for providing higher returns on EE investments

Regulatory assets created through investments in DSM and supporting technologies like Smart Grid and AMI should enjoy a premium above existing ROR on rate base. First, if the policy goal is to move the energy market, both suppliers and customers, to DSM more quickly than status quo, then incentives should be provided to both consumers and utilities. The topic of providing incentives to customers to foster additional 'sales' of DSM is well covered. The impacts of providing incentives to utilities to generate additional DSM 'sales' and to grow DSM infrastructure is more open. While a rate adder designed to make a utility neutral on the impact of DSM on overall returns is a step in the right direction, the incentives are reduced to regulatory compliance and good will within the community and customer base.

Secondly, if it is correct that EE is the least cost form of energy in today's market, even prior to enforced carbon constraints, then the economic value to consumers and society as a whole is optimized by aggressively pursuing more EE. Under these conditions, as much new energy demand as possible should be provided by EE over new fleet generation, subject to the traditional benefit tests to ensure benefits accrue to the customer base. Programs that allow utilities to treat DSM investments as regulatory assets easily allow the benefits of the lower EE costs to be recognized in the energy market. States have begun to provide incentives to utilities over and above traditional returns through programs that share benefits between ratepayers and shareholders.17size="1">


Utility participation in the EE industry through DSM programs has lagged behind the growth in the EE industry in general and growth in total industry peak capacity over the last decade. It is instructive in understanding historical DSM trends and utility behavior by looking at DSM through the utility's perspective as one product in the portfolio of products that can be used to meet customers' energy needs. The topic of how to best provide incentives to utilities for DSM is well covered and there are many different approaches being utilized by regulatory agencies today. Programs that allow for cost recovery only and programs with pre-set rate adders must be viewed carefully to understand the differences between short run and long run impacts on utility shareholders. The best market-based approach to stimulate utility participation in DSM will allow utilities to seek profit maximization through DSM as a product. Combining customer safeguards through standard program benefit tests with programs that recognize EE's cost advantage in the market will help allow the industry to move toward an optimum level of EE as a choice in that way energy utility is provided to the customer base. And lastly, DSM incentive programs that utilize existing regulatory structures and rules have the best opportunity to achieve new DSM policy goals quickly. The next piece of research to be tackled is the level of incentive above current ROR that optimally reflects cost advantages that DSM currently enjoys in the market.

Next Steps for Research

Areas for continuing research in order to understand the scope of the role that utilities can play in the larger EE market include:

  • Calculating the optimal incentive level (above ROR) for utility DSM programs;
  • Calculating the total economic benefits associated with optimizing EE level; and,
  • Impacts on the larger EE market from fostering more aggressive participation by utilities.
1. Energy 'utility' refers to the overall value consumers get from energy consumption. EE as referred to here does not imply a reduction in demand for energy as an industrial input or a consumer good; that would be conservation. EE refers to a more efficient utilization of energy. Consumers get the same benefit at a lower utilization level.
2. The Size of the U.S. Energy Efficiency Market: Generating a More Complete Picture. Karen Ehrhardt-Martinez and John A. "Skip" Laitner, May 2008) Report Number E083
3. Organic EE in this context refers to efficiency gains from industry evolution around technology-based product improvements like CFLs, improved insulation, and Energy Star product lines.
4. % DSM is calculated by comparing DSM actual peak load reductions against total net summer capacity. Electric Net Summer Capacity numbers from EIA Table 8.11a. Utility DSM actual peak load reductions from EIA table 9.1.
5. Some load management and shedding programs are different in nature and more closely resemble conservation as these programs will result in direct reductions in energy usage, typically in response to approaching peak capacity constraints.
6. Averch & Johnson, Behavior of the Firm Under Regulatory Constraint, 53 Am. Econ. Rev. 1053 (1962); Wellisz, Regulation of Natural Gas Pipeline Companies: An Economic Analysis, 55 J. Pol. Econ. 30 (1963). A relatively accessible review of this theory as applied to telecommunications can be found in Albery & Sievers, The Averch-Johnson-Wellisz Model and the Telecommunications Industry, Fed. Comm. Law Jour. Vol. 40 Number 2 (1988).
7. Note that the rate of profitability or the percent returns are fixed. Only the quantity of profitability is affected as the constant percentages are applied against a larger base.
8. This estimate utilizes an estimate of $850 per KW for future construction costs in 2014, uses a 5% discount rate and applies an additional cost factor to better align impacts on earnings from changes in the rate base. The factor is derived from a current utility, is based upon actual 2008 results, and is designed to maintain the ratio between net income and a 10% ROR on total equity and debt.
9. 2008 Florida Statutes, Title XXVII, chapter 366, section 366.82, paragraph 9.
10. This DSM incentive was added to the Florida statutes in 2008 and will require regulatory review for each utility prior to implementation. Given the newness of the policy, no utilities in Florida have put the adder into place to date.
11. This example uses similar assumptions including $850 per KW of construction costs in 2014, a 10% ROR, and uses FPL's 2008 asset base drawn from its SEC 10-K filing.
12. The level of ROR adder to make DSM programs neutral under this simplified analysis is given by the following calculation ROR Adder = ((Avoided CAPXDSM /Rate Base) * ROR).
13. Could this be a clever way for the legislature to take advantage of the notoriously short-term focus of Corporate America?
14. This example is again based off of FPL and assumes rate base assets of $250 per KW reflecting lower construction costs in the past along with depreciation. This example is purely hypothetical and for exposition only.
15. I realize that some may chafe at my characterization of traditional ROR regulation as 'comfortable'. In terms of the old expression "The devil we know", both regulators and utilities may be able to more easily make progress using adjustments to existing rules rather than creating a new set of regulatory rules to follow.
16. Energy Independence and Securities Act of 2007, Pub. L. No. 110-140, -1305, 121 Stat. 1787 (2007).
17. The cost capitalization approach used in Nevada offers a 500 basis point premium for DSM investments over supply investments. For the presentation Overview of Utility Incentives, Wayne Shirley, The Regulatory Assistance Project, August 26, 2008.