Effects of Electric Utility Decoupling on Energy Efficiency.

Author:Kahn-Lang, Jenya

    Under traditional regulation, energy utilities profit from marginal increases in per-customer consumption. This "throughput incentive" directly impedes energy efficiency efforts. According to the Environmental Protection Agency (EPA) (2007), increasing energy efficiency is among the most cost-effective ways to address energy security, air pollution, and climate change. End users rarely make economically efficient energy decisions independently (Gillingham, Newell, and Palmer 2009).

    Revenue decoupling, under which utility revenues are independent of sales, addresses the utility throughput incentive. Under revenue decoupling, total or per-customer revenue is fixed during the utility's standard rate case. (1) Periodic true ups, often through a customer surcharge or rebate, ensure exact recovery of this revenue during the regulatory period.

    In 2000, only four electric utilities in two states had decoupling mechanisms. By May 2014, 32 electric utilities in 13 states were fully decoupled or awaiting a final decision on full decoupling proposals. Under the Energy Independence and Security Act (2007) and the Recovery Act of 2009, to receive federal energy-efficiency funds, state regulators must consider policies, such as decoupling, that remove the throughput incentive (U.S. Congress 2007, U.S. Congress 2009). Appendix C (online) provides a full list of decoupled electric utilities.

    Many regulators prefer decoupling to other methods of eliminating throughput incentives, such as straight-fixed variable (SFV) rates. SFV rates include large mandatory fixed charges that cover all or most utility fixed costs, making variable costs commensurate with marginal utility electricity generation costs. Because they are socially regressive and provide minimal customer efficiency incentives, electric regulators rarely support SFVs (EPA 2007).

    Although governments or third parties may promote energy efficiency directly, utilities are typically better positioned to do this because of their superior information and data about customer energy use (Gillingham, Newell, and Palmer 2006). Moreover, utilities hurt by energy efficiency can use political clout to impede third-party efficiency efforts. Historically, decoupling has improved the efficacy of third-party efforts (Lazer, Weston, and Shirley 2011).

    While, by guaranteeing recovery of lost revenue, decoupling eliminates utilities' throughput incentives, it may not encourage them to promote energy efficiency through investing in demand-side management (DSM) (NARUC 2007). DSM programs aim to reduce electricity consumption or peak demand. They include practical and financial support for consumers to install energy-efficient technologies and pricing schemes that incentivize energy efficiency (Loughran and Kulick 2004, EIA 2012).

    This paper addresses whether and how decoupling promotes energy efficiency. After reviewing the literature, I present a simple model in which decoupling does not directly induce DSM investment but increases DSM spending indirectly by raising the level the regulator imposes in equilibrium. Then, using data from a panel of utilities, I show that decoupling is associated with reduced electricity demand and that this relation can be explained by decoupling's relations with DSM expenditure and efficacy. I also find evidence of a small, positive short-term association between decoupling and average retail rates (price), although including an interaction between decoupling and economic indicators renders this effect statistically insignificant.


    2.1 Theoretical Literature

    Most researchers agree that traditionally regulated utilities oppose energy efficiency investments because they profit on marginal sales (NARUC 2007, EPA 2007), although they may have incentives to reduce consumption during peak demand hours when marginal electricity production costs exceed retail prices (Brennan 2008, Zarnikau 2012).

    While decoupling eliminates profit on marginal sales, many authors argue that decoupling alone does not incentivize DSM investment (Moskovitz et al. 1994). Decoupling only addresses one of the three key utility financial barriers to DSM investment, often referred to as the "three-legged stool" (Blank and Gegax 2011); decoupling does not ensure recovery of DSM program and administrative costs or provide shareholder returns comparable to those of supply-side investments (EPA 2007). DSM investment defers or avoids supply-side investments, thereby typically reducing shareholder returns. Compared to supply-side investments, DSM investments are also riskier or too small in scale (Kihm 2009). Utilities have historically preferred large-scale investments to small-scale investments because their allowed rates of return on those investments often exceed their costs of capital. This causes an increase in investment scale to disproportionately increase shareholder profits, making utilities favor larger investments, a concept known as the Averch-Johnson effect (Kihm 2009). Combining decoupling with shareholder DSM incentives can mitigate this preference for supply-side investments (EPA 2007).

    In contrast, Zarnikau (2012) maintains that decoupled utilities inherently have incentives to invest in DSM if regulators underestimate costs, such as inflation, during true ups, although he recognizes that the opposite is also possible. DSM investments are attractive because they generate less political opposition and more positive publicity than supply-side investments. DSM positively impacts participants at a miniscule cost to many (ECW 1997). In contrast, supply-side investments, such as pollution-emitting generators, typically provide widespread benefits while severely adversely impacting concentrated communities (Bloomquist 1974). Hurley et al. (2008) and EPA (2007) also maintain that DSM investments are less risky in the long run because they reduce fuel price susceptibility.

    Finally, some opponents argue that decoupling discourages energy efficiency because customers' electricity prices increase if they reduce their usage (Graniere and Cooley 1994). However, for an individual customer, decreasing usage reduces the total electricity bill (NARUC 2007). In fact, the indirect efficiency-induced price increases encourage further efficiency (ECW 1997).

    Beyond high-level discussions of DSM shareholder incentives, none of these papers focus on the regulator's role in encouraging utility DSM investment to formally prove a relation between decoupling and DSM spending. Regulators' abilities to affect DSM investments may differ between traditionally regulated utilities and decoupled utilities. I explore this using a game-theoretic model.

    2.2 Estimated Effects

    The empirical literature on the effects of decoupling on energy efficiency is limited. Partially due to free riders (Joskow and Marron 1992), ex-post econometric analyses of energy savings from DSM programs systematically find lower savings than ex-ante predictions (Arimura et al. 2011). This inconsistency indicates the need for ex-post studies of the effects of decoupling on energy efficiency.

    However, few such studies exist. Loughran and Kulick (2004) conducted a cross-section/time-series analysis of the effect of DSM investment on electricity consumption, but they neglected the impact of decoupling. Arimura et al. (2011) included a brief analysis of the effect of decoupling on electricity consumption. Their results suggest that decoupling strengthens the demand-reducing effects of DSM spending, but the estimate is imprecise and statistically insignificant. The time period analyzed did not capture many decoupling mechanisms.

    Thus, the literature lacks 1) a game-theoretic model of strategic utility and regulator interplay that explores how regulatory regime affects DSM investment, and 2) a thorough empirical analysis of relations among decoupling, DSM investment, and electricity consumption. This study fills these gaps. The empirical analysis here adds to previous studies by including post-2006 data, isolating the effects of level and efficacy of DSM spending, recognizing the endogeneity issue that stems from including utility-level electricity prices as a right-hand-side variable, and including a vector of time fixed effects to better control for national time trends and structural breaks in electricity consumption.


    3.1 Introduction

    This section presents a repeated game between 1) a regulator (the principal) who wants to maintain consistent electricity delivery to end users, keep electricity price low, and reduce demand through utility DSM investment and 2) a utility (the agent) who maximizes profit. I assume that a higher average electricity price (i.e. a larger revenue requirement) increases shareholder returns. End-user behavior is not modeled. The environment modeled here, albeit simple, highlights the main hypotheses tested empirically.

    Assume the demand and cost functions are known and stable and that electricity consumption is a direct function of price and DSM spending:

    q = q(p,D) (1)

    where q is electricity demand, p is average electricity price, and D is DSM spending. Note that D could encompass indirect and opportunity costs of DSM programs, if applicable. For simplicity, I assume that DSM spending does not affect demand in future periods. While this assumption is strong, it is apparent that the logic of the model would hold under any finite DSM spending impact duration.

    Profit is given by:

    [pi] = pq(p,D)-f-D (2)

    [equivalent to]R(p,D)-D (3)

    where f is fixed costs and R is net revenues exclusive of DSM spending ('net revenues'). For simplicity, I assume that the per-kWh marginal utility-borne cost of output is zero, as fuel and power purchasing costs are typically passed through to ratepayers and most other utility-incurred costs do not vary on a per-kWh basis. (2)

    Under traditional regulation, p is fixed for the regulatory period. The utility makes...

To continue reading