Briefing Paper - NARUC

•

Briefing Paper

Revenue Decoupling for Natural GasUtilitiesThe National Regulatory Research Institute

April 2006Ken CostelloSenior Institute Economist

EXECUTIVE SUMMARY

High natural gas prices have provoked recent proposals to modify long-held ratemakingpractices for gas utilities. Energy conservation has emerged as an option to addressthe serious problem of consumers suffering from accelerating gas bills. With aheightened emphasis on energy conservation, gas utilities have expressed concernabout the implications of lower gas usage for their financial stability. In response tothis situation, gas utilities as well as conservationists have advocated a ratemakingmechanism generically labeled revenue decoupling (RD). From the perspective ofgasutilities, RD can prevent financial erosion from future reductions in consumption by gasconsumers. Conservationists view RD as indispensable in eliminating the disincentivefor gas utilities to promote energy conservation under standard ratemaking.

This briefing paper reviews the activities to date on the application of RD for gasutilities. Five gas utilities presently have commission-approved RD mechanisms.Several others have RD proposals pending before their state commissions. Consumergroups and others have posed several arguments in disfavor of RD. Some statecommissions have endorsed RD while others have opposed it. This paper lists thearguments on both sides together with an assessment of their merits.

This briefing paper takes a balanced perspective of RD by directing attention to boththe upside and downside of this ratemaking mechanism. It specifically analyzes theefficacy of RD in fostering prevailing regulatory and ratemaking objectives. Thepaper's primary intent is to make state commissions as well as other policymakersbetter informed on the likely outcomes of RD. While this paper concentrates on thenatural gas industry, much of its content applies equally to both the electric and waterindustries.

Contents

Impetus for Revenue Decoupling 6

Basic Structure of a RevenueDecoupling Mechanism 9

Summary and Conclusions 22

Analysis of Major Arguments 16

14Divergent Views on RevenueDecoupling

2

4

Background

Capsule of State Activities

06-06

The author appreciates the helpful connnents of Connnissioner Richard Morgan, District ofColumbiaPublic Service Commission, Robert Harding, MinnesotaPublic Utilities Commission,Bob Pauley, Indiana Utility Regulatory Commnission, James F. Wilson, LECG LLC., and Dr.Vivian Witkind Davis on earlier drafts of this paper.

High natural gas priceshaw had a profoundeffect on both local gasutilities and consumers.

Under standardratemaking, utilities havestrong motivation topromote gas sales.

Both utilities andconservationists arepromoting revenuedecoupling, which seversthe link between earningsand sales.

2

BACKGROUND

High natural gas prices since 2000have had a profound effect on bothlocal gas utilities and consumers.Utilities have witnessed consumersresponding to high prices by curtailingtheir gas use, with the consequenceof lost earnings. State commissionshave increasingly pressured, and insome instances required, gas utilitiesto become more active in promotingenergy efficiency, largely to reduceconsumers' gas bills in both theshort and long run. Consumers haveexperienced hardships from lessdiscretionary income for non-gasproducts and services, including othernecessities such as food and housingfor low-income households. To reducethe cost pressures on utilities as well asthe burden on consumers, gas utilitiesanticipate that regulators and otherpolicymakers will increasingly exertpressure on them to broaden theirfunctions and the services they offer,specifically by being more engaged inpromoting energy conservation.

Under standard ratemaking aspracticed in the vast majority of states,gas utilities have strong motivation topromote gas sales between rate cases.Owing to regulatory lag, wheneversales grow, earnings directly increasebecause ofthe prevailing rate structurethat incorporates most of a utility'sfixed costs into its volumetric charge.!Conversely, when a utility sells lessgas it recovers a smaller portion of itsfixed costs. State commissions haveendorsed this rate design, which hasa long history, largely on grounds ofequity. Over the past few years, at therequests of gas utilities in rate filings,

volumetric charges have included lessfixed costs to reduce the utility's riskfrom sales fluctuations. 2 Still, withfew exceptions, utilities' shareholdersshoulder financial harm in varyingdegrees whenever sales declinebetween rate cases.

In response to the sales repressiontriggered by market forces and theregulatory goal in several states forutilities to promote and fund energyefficiency initiatives, several gasutilities have recently proposed totheir state commissions a "tracker"mechanism that severs the linkbetween their earnings and sales. Bothgas utilities and conservationists haveaggressively fostered this ratemakingmechanism, generically labeled inthis paper as revenue decoupling(RD).3 While retaining adequateearnings is the driving motive ofgas utilities, conservationists viewrevenue decoupling as indispensablefor removing the resistance of utilitiesto promoting energy efficiency.4

These seemingly strange bedfellowsare mounting a strong charge to reshapefundamentally ratemaking practicesfor relieving the utility's risk fromlower sales between rate filing. Asevidence, in 2004 the American GasAssociation (AGA)5 and the NaturalResources Defense Council issued ajoint statement in support of RD andsubmitted it to the National Associationof Regulatory Utility Commissioners(NARUC). The Joint Statementexpressed agreement ofthe two groupson "the importance of state PublicUtility Commissions' consideration ofinnovative programs that encourageincreased total energy efficiency and

The National Regulatory Research Institute

conservation in ways that will alignthe interests of state regulators, naturalgas utility company customers, utilityshareholders and other stakeholders."6In 2005, NARUC passed a resolutionadvising state commissions to considerthe implementation of revenuedecoupling. The resolution statedthat revenue-decoupling mechanisms"may assist, especially in the shortterm, in promoting energy efficiencyand energy conservation and slowingthe rate of demand growth of naturalgas."7 As discussed below, gas utilitiesand conservationists, while generallysupportive of RD, relate differentstories about the merits of RD.

Not all gas utilities are supportiveof revenue decoupling. In theinvestigation of revenue decoupling bythe Connecticut Department of PublicUtility Control, some gas utilitiesopposed a full-sales adjustmentmechanism by arguing that it wouldremove their incentive to add newcustomers, increase their rates, andlead to a lower authorized rate ofreturn. One gas utility consultant hasargued that RD could create additionalcosts to a utility: (1) regulators couldbe under pressure to lower the utility'sauthorized rate of return, (2) in anenvironment of high gas prices, itmay be especially difficult for autility to secure regulatory approvalof a decoupling mechanism withoutmaking a commitment to energyefficiency initiatives, and (3) the utilitymay have to agree on other concessionsin rate case settlements in return forapproval of RD. 8

On the other side of the debate arethose who have raised concerns about


the desirability of RD as a ratemakingtool. These skeptics emphasize thedownside effects of "guaranteeingearnings" and the possibility of higherrates in the short term.

To date, state commissions have reacteddifferently to RD proposals, reflectingtheir ambivalence about whether thisratemaking mechanism advanceslong-held ratemaking and regulatoryobjectives. This is not so surprising,since RD has more obvious benefits toa gas utility than to its consumers andutilities vary as to their commitmentsto promoting energy efficiency.

Interest in RD also exists nationallyand regionally. Last year, the WesternGovernors' Association supported RDfor electric utilities to boost energyefficiency initiatives in addition tospecific performance incentives thatreward utilities for cost-effectiveinitiatives.9 The Energy Policy Actof 2005 (EPAct of 2005) (Section139, Subtitle C) requires the U.S.Department ofEnergy within one yearofenactment ofthe Act, in consultationwith NARUC and the NationalAssociation of State Energy Officials,to conduct a study of state and regionalpolicies that promote cost-effectiveconservation programs. The policies,among other things, should take intoconsideration methods of "removingdisincentives for [gas and electric]utilities to implement energy efficientprograms" and "ensuring appropriatereturns onenergy efficiency programs."Also at the national level, the U.S.Environmental Protection Agencyhas undertaken a Clean Energy PolicyInitiative that will review RD in termsof removing disincentives for natural

Skeptics have raisedconcerns about thedesirability ofRD as aratemaking tool; andnot all gas utilities aresupportive.

Interest in RD is at thenational, regional, andstate level.

3

The Maryland,California, Oregon,and North Carolinacommissions haveapprovedRDmechanisms.

Commissions approvedRD mechanisms because:• Gas usage per

household has beendeclining

• The commissions wantregulated utilitiesto promote energyefficiency

• With out a RDmechanism, the utilitymight not be able toearn its authorized rateofreturn.

4

gas and electric utilities to promoteenergy efficiency.lO

CAPSULE OF STATEACTIVITIES

At the time of this briefing paper,state commISSIOns had approvedRD for five gas utilities.ll Thesegas utilities are Baltimore Gas andElectric,12 Washington Gas Lightin Maryland, Southwest Gas inCalifornia,13 Northwest Natural inOregon, and Piedmont Natural Gasin North Carolina. Last August, theOregon Public Utility Commissionextended Northwest Natural's RDmechanism to four years and modifiedthe mechanism by allowing for 100percent decoupling (previously it was90 percent), excluding weather effects.14

These approvals were motivated by atleast one of two factors. The first wasevidence showing an historical declinein gas usage per household, which wasanticipated to continue in the future;the second pertains to the intent of astate commission to have the utilityaggressively and successfully promoteenergy efficiency. With the expectationof falling sales per customer, it was theapparent belief of those commissionsapproving a RD mechanism that theutility would not have a reasonableopportunity to earn its authorized ratesof return unless a RD-type mechanismwas in place. For other RD approvals,notably for Northwest Natural, thecommission acknowledged that RDshould go hand-in-hand with its intentto have the utility promote energyefficiency.

In the Piedmont Gas order, the NorthCarolina Utilities Commission saidthat:

The customer utilization tracker[CUT] represents a departurefrom the ratemaking approachtraditionally approved by theCommission, in which noone element is singled out forratemaking without considerationof other, countervailingelements.. .In this case approvingthe CUT as an experimentalrate for a limited period of timewill allow the Commission tomonitor experience under theformula-including its impacton the Company's earnings, onconservation efforts, and ontraditional ratemaking theorybefore the CUT is approved as apermanent part of Piedmont's ratestructure.IS

The Commission opined that the CUTshould benefit customers by givingthe utility "a conservation incentiveto assist residential and commercialcustomers." It also said that theRD mechanism would reduce thefrequency of future rate cases as wellas shareholder risk.16

As of early 2006, several gas utilitieshave filed RD proposals with their statecommissions in Indiana, New Jersey,Ohio, Utah, and Washington. Thegas utilities filing a RD mechanisminclude Cascade Natural Gas (WA),Puget Sound Energy (WA), PugetEnergy (WA), Questar Gas (UT),Citizens Gas and Coke Utility (IN),Vectren Energy Delivery (IN, OR),New Jersey Natural Gas, and South


Jersey Gas. In the Vectren EnergyDelivery ofOhio application, the utilityargued that its revenue-decouplingproposal would "sever or 'decouple'the traditional relationship betweensales or throughput and revenue as theessential foundation for a realignmentof interests to better advance Ohio'senergy policy." 17 The RD proposalby Questar Gas has received thesupport of the Utah Division of PublicUtilities. A witness representingthis group testified that the proposed"conservation enabling tariff" would"protect Questar's revenues fromshortfalls due to price shocks andeconomic downturns."18 In the caseof South Jersey Gas, the utility said itwould commit to a comprehensive setof programs designed to both educateand provide incentive to customersfor conserving gas, once the Boardapproves its "conservation" tariff. Asexpressed by the utility's President andCEO, "Under this pilot program, theexisting link between customer usageand cost of service recoveries will besevered. As a result, the company canaggressively and creatively encouragechanges in customer behavior thatlead to increased conservation withoutnegatively impacting out financialstability."19

In October 2005, the WashingtonUtilities and TransportationCommission closed its rulemakingdocket without action (DocketNo. UG-050369) by stating that it"believes that the wide variety ofalternative approaches to decouplingmake it more efficient to addressthese issues in the context ofspecific utility proposals included ingeneral rate case filings rather than


through a generic rulemaking." Thecommission encouraged utilities tofile a RD proposal if they deem thatsuch a proposal would "overcomedisincentives to their offering...conservation programs."

In January 2006, in a report to thestate legislature20 the ConnecticutDepartment of Public Utility Controlrejected the implementation of RDin the form of sales and per customeradjustment clauses for either gas orelectric utilities. 21 The Departmentparticularly found problematic theshifting of normal business risk froma utility to consumers.22 It also saidthat market forces (namely, highgas prices) have stimulated muchenergy conservation. 23 Besides, asexpressed by the Department, gasutilities currently have a conservationadjustment mechanism, which is usedto recover from consumers both thecosts of utility-sponsored programsand lost revenue from sales foregoneattributable to those conservationprograms. Consumer groups in thestate also oppose RD in the form ofsales and per-customer adjustmentclauses, with the general argumentthat consumers would unlikely benefitwhile autility would have "guaranteed"earnings.24

The rejection or withdrawal of RDproposals has occurred in a fewstates, including Arizona, Minnesota,and Nevada. 25 For example, inearly 2006 the Arizona CorporationCommission rejected a proposal bySouthwest Gas (called a "ConservationMargin Tracker" or "CMT").26 TheCommission reasoned, "there isconflicting evidence in the record as

Gas utilities in Indiana,New Jersey, Ohio, Utah,and Washington havefiled RD proposals.

In Arizona, Minnesota,and Nevada RDproposals have beenrejected or withdrawn.

5

Increasingly energyefficient gas appliancesand home constructionare responsible for muchofthe substantial declinein natural gas usage.

to whether the recent level of decliningper customer usage will continue intothe foreseeable future, and whetherconservation efforts are the directcause of Southwest Gas' inabilityto earn its authorized return fromsuch customers." The Commissionadded, "The Company is requestingthat customers provide a guaranteedmethod of recovering authorizedrevenues, thereby virtually eliminatingthe Company's attendant risk. Neitherthe law nor public policy requires sucha result nor do we decline to adopt theCompany's CMT in this case." Thecommission opined that the issue ofdeclining usage per customer "shouldbe fully explored as part of a broaderinvestigation of usage volatility andmargin recovery."27

IMPETUS FOR REVENUEDECOUPLING

As mentioned above, two factorsexplain the heightened interest in RDfor gas utilities. Both relate to theongoing energy conservation resultingfrom the combination of high naturalgas prices and utility-funded energy

efficiency initiatives. According toan AGA study, natural gas usage perhousehold (normalized for weather)has declined by over 20 percent since1980.28 Major reasons for this include

.. .progressIve Increases In energy-efficient gas appliances and homeconstruction.29 The study predicts thisdecline will continue over the nextseveral years, although at a lower ratethan since 1980. 30 (See Figure 1.) TheAGA study, published in 2004, mayhave underestimated future declines iffor no other reason than that naturalgas prices have soared since then. Ifone assumes that consumers are priceresponsive, even in the short run, whichcoincides with past consumerbehavior,further drops in gas consumption percustomer below those projected in theAGA study may come to fruition. 31

The phenomenon of decliningusage per household over the pastseveral years, while not universal,probably has occurred for most U.S.gas utilities.32 Some gas utilitieshave experienced particularly sharpdeclines. For example, Questar Gasin Utah estimated that, adjusting for

6

Natural Gas Consumption per Household(Normalized for Weather)

175

j%' 196

1106

20 40 60 80 100 120

Normalized Consumption (Met)

Source: AGA, Forecasted Patterns in Residential Gas Consumption, 2001-2020, EA2004-04, Sept. 21,2004.

Fig. 1. Declining usage per household.


weather, its typical residential customercurrently uses about 35 percent lessnatural gas than in 1980.33

Decliningusage per customer overtimeper se does not necessarily rationalize aRD-type mechanism. 34 Its defense as aratemaking mechanism becomes moretenable, however, when base (delivery)rates do not reflect this "consumption"dynamics. 35 Common in recent ratecase filings is the argument by gasutilities that in calculating sales for therate period, historical sales should beadjusted to account for declining usage(orweather normalized over a relativelyshort time period, for example tenyears or less).36 Otherwise, utilitiesmay find it difficult to earn their predetermined authorized rate of return.37Recognizing that a commission maynot accept this adjustment in settingnew rates for various reasons, whichmore times than not has been thecase, a utility would then find RDas the only available option (exceptfor a radical change in rate design,which would likely confront strongopposition) to protect itself againstrevenue shortfalls.

Another reason for utilities' interestin RD is the growing intent of statecommissions and other groups tohave gas utilities promote energyefficiency. Largely because of highnatural gas prices, state commissionshave shown greater propensity forrequiring gas utilities to become moreactive in promoting energy efficiency.Many industry observers view energyefficiency as the most effective optionto soften gas prices over the next fewyears. 38 Gas utilities themselves seemto be supportive of energy efficiency


as long as it does not adversely affecttheir shareholders. This explainswhy some utilities have proposedRD mechanisms as a ratemakingmechanism that would protect theirshareholders if sales reductions ensuefrom successful energy efficiencyinitiatives.

Table 1 provides the fundamentalarguments in support of revenuedecoupling. These are the majorcontentions-some advanced bygas utilities while others by conservationists-that have been made instate-commission proceedings. Atfirst sight, they come across as apersuasive and coherent case for RD.As discussed in more detail below,RD can assist in promoting energyefficiency: while RD does not providethe utility with an explicit incentiveto promote energy efficiency, iteliminates the disincentive. In otherwords, if a state commission wants autility to effectively and aggressively"sell" energy efficiency, then RDor least some other mechanism (forexample, straight fixed-variable ratedesign for gas delivery39) that would notdiscourage a utility from selling lessgas should be seriously contemplated.Besides, it would seem both unfair andcounterproductive to order a utilityto promote energy efficiency whendetrimental to its shareholders.

Another argument in favor of RD isthat small changes in gas sales canhave a significant effect on earnings.Gas sales are also largely outside thecontrol of a utility in addition to beinghighly volatile from year to year, withweather as the maj or factor. Sincealmost all of a utility's short-run, non-

With recent soaringnatural gas prices, theremay befurther drops ingas consumption.

Declining use percustomer does notnecessarily callfor RDmechanisms. It becomesmore defnsible whenbase delivery rates donot account for dynamicchanges in consumption.

RD eliminates adisincentive to promoteenergy efficiency.

7

TABLE 1UNDERLYING RATIONALE FOR REVENUE DECOUPLING

Energy efficiency can benefit Given existing rate design,consumers and society in both variations in sales directlythe short and long run affect a utility's net revenues

or earningsCurrent rate design creates A small percentage change indisincentives for a utility to sales can have a significantpromote energy conservation effect on a utility's earningsGas sales are highly volatile Almost all of a gas utility'sfrom year-to-year for different short-run non-gas costs arereasons, most of which are fixed, meaning they staybeyond a utility's control virtually constant when sales

changeRate structures encourage Most of a utility's fixed costsutilities to increase sales are recovered through thebetween rate cases (which may volumetric chargebe contrary to state policies)Determining future sales is a Full recovery of fixed costscontentious issue in rate cases depends on actual gas sales

equaling sales levelscalculated in the last rate case

Consumption per residential Fixed costs recovered in thecustomer (adjusting for weather) customer charge would reducehas continuously declined over consumers' incentive tothe past 25+ years in most conserveregions of the country, and thistrend is expected to continue inthe futureSOluce. Author s construct.

One argument in favor ofRD is that small changesin gas sales can havea significant effect onearnings.

8

gas costs are invariant to changes insales,40 a mechanism such as RD thatadjusts for sales fluctuations, ratherthan cost changes, would seem valid.

As an illustration ofhow a small changein sales between rate cases can have alarge effect on a utility's earnings, letus assume that:

(1) E* =R* -FC - VC(2) L1R (L1Q x P) - L1VC = L1 E(3) (R/E) x ([L1R - L1VC]/R) = L1 E/E =

L1ROE/ROE*

Where E = earnings to common equityshareholders, R = revenues, FC = fixedcosts (exclusive of equity returns), VC= variable costs, L1Q = the change inthe quantity of sales relative to the testyear level, P = the delivered price ofgas, ROE = rate of return on equity,and * = targeted or authorized levelsfor the specified parameters. Equation

(1) assumes that common equityshareholders hold residual claimsto a utility's earnings. Equation (2)says that changes in the earnings tocommon equity shareholders equal thedifference between changes in revenueand variable costs (i.e., the change innet revenues). Equation (3) relatesthe proportional changes in earningsand the rate of return on equity to thechange in net revenues and the ratioof revenues to earnings to commonequity shareholders.

For illustration, let us assume that: (1)the utility's authorized revenues (R*)are $400 million, (2) fixed costs (FC)equal $360 million (thus, the targetedearnings to common equity holders is$40 million), (3) variable costs for gasdistribution equal zero,41 and (4) theauthorized rate ofreturn on equity (R*)equals 12 percent.42 Assuming that gassales are 1 percent less than expected


(i.e., R = $396) because of a reductionin gas use per customer, the decreasein earnings to common equity holderswould sum to $4 million, which is adecline of 10 percent [($40 million $36 million)/$40 million)]. This alsotranslates into a decrease of the pretax rate of return on equity (ROE)of 10 percent or from 12 percent to10.8 percent (i.e., 120 basis points).43In sum, the reduction of sales andrevenues by 1 percent is concomitantwith a 10 percent drop in earnings tocommon equity holders.

With the large increases in natural gasprices over the past few years, gas saleswill likely experience some (unknownat this time) downward movement,assuming other things held constant.As an example, with a 30 percentincrease in the delivered price of gasin real dollars, and assuming a shortrun price elasticity of demand equalto -0.10 (which corresponds closely toeconometric studies), gas sales wouldfall by 3 percent because of the priceincrease. Applying the relationshipsin the above illustration, this drop insales would cause pre-tax earnings tocommon equity shareholders to fall asmuch as 30 percent.

BASIC STRUCTURE OF AREVENUE DECOUPLINGMECHANISM

Revenue Decoupling: a "Tracker"

In its purest form, revenue decouplingis a "tracking" mechanism that adjustsrates and revenues whenever salesdeviate from their targeted level (i.e.,rate-year sales determined at the lastrate case). RD can also represent a


form of "attrition allowance," whichallows for rate adjustments during theperiod between general rate cases,that recognizes a utility's earningseroding during the post-test yearperiod because of billed revenuesexpecting to fall below authorizedrevenues. The underlying assumptionis that a decline in costs does not offsetdeficient revenues. In regulatoryparlance, attrition refers to a utility'sinability to earn its authorized rateof return because of a change in therevenue-cost relationship outside thetest year period.

Historically, "trackers" such as RDare justified on the basis of a threeprong threshold test: (1) the designatedactivity largely lies outside the controlof a utility, (2) variations in theoutcome of the activity have morethan a minimal effect on a utility'searnings, and (3) the actual outcomeis likely to deviate from the baselineprojections.44 When applying this testto the activity "sales," it seems tenableto apply a tracker for sales in the formof RD. As discussed above, sales arelargely external to a utility's control,and inescapable sales fluctuationscan significantly affect earnings.Unless a state commission faces legalrestrictions in implementing a "salestracker" or has a built-in policy oflimiting trackers in general, RD wouldseem to meet the regulatory thresholdfor a tracker. This should not implythat RD is necessarily in the publicinterest and defensible as a ratemakingmechanism. As argued below, RDhas some downsides (which can bepotentially damaging to customers)that a state commission should review

In its purest form, RD isa "tracking" mechanismthat adjusts rates andrevenues wheneversales deviate from theirtargeted level

Historically, trackerssuch as RD are subjectto a threshold test basedon utility control, degreeofimpact, and likelihoodofa difference betweenactual and expectedoutcomes.

RD seems to meet theregulatory thresholdfor atracker.

9

TheRD removes theinterest in promotingsales (although this isan odd; ifnot perverse,departure from thenormal corporate modusoperandi).

The distinctive featureofRD is that a utility'sacutal earnings betweenrate cases would equalthe utility's authorizedearnings.

10

in deciding upon the merits of a RDproposal.

Incentive Effect

For conservationists, RD has theprimary feature of neutralizing theutility's incentives for adjusting salesfrom a baseline level. If, for example,rate year sales are 1000 units, under apure RD mechanism the utility wouldbe indifferent to whether actual salesare 1100 units or 900 units, as it wouldbe compensated for the lower salesand required to reimburse consumersfor the higher sales. This definitelyimproves upon the situation wherethe utility has an incentive to promotesales when the commission or statepolicy perceives the utility as anenergy-service company, selling bothgas service and energy efficiency. Onthe other hand, a utility disinterestedin sales volumes is an oddity in theconfines of corporate modus operandi,and, to some observers, a perversitythat runs counter to the basic tenets ofhow markets work and private firmsmake money.

Illustration of a RevenueDecoupling Mechanism

The simplified example below showsthe basic operation ofaRD mechanism.Under a stylized RD mechanism,whenever sales deviate from a specified"baseline,"45 the utility is able to adjustits rates without having to file a formalrate case, so as to earn its authorizedearnings (assuming that the utility'sactual costs coincide with test yearcostS).46

Let us assume that: (1) the "baseline"sales determined at the most recentrate case equal ten million therms, (2)actual sales equal 9.5 million therms(or 5 percent less than the "baseline"sales), and (3) the distribution margin(or base rate) equals 30 cents per thermoWith these assumptions, the revenuesat the "baseline" sales equal $3 million(30 cents x 10 million therms), withactual revenues equaling $2.85 million(30 cents x 9.5 million therms). Thus,the revenue shortfall equals $150,000.Under revenue decoupling, rates adjustautomaticallyupwardtocompensateforthis shortfall. Specifically, it achievesthis by the following: ~Price x 9.5million therms = $150,000 or ~Price

= $150,000/9.5 million therms, whichequal 1. 579 cents (which increases thedistribution charge by 5.3 percent). Inother words, by increasing the baserate to 31.579 cents per therm (from30 cents), the utility would achieve thesame revenues of $3 million as if saleswere at the "baseline" level. Assumingthat the purchased gas cost is 70cents per therm,47 in this example theincrease in the delivered price of gasto customers would equal 1.6 percent.

The distinctive feature of RD, relativeto standard ratemaking, is that theutility's actual earnings betweenrate cases would equal, or at leastcorrespond more closely to, the utility'sauthorized earnings. Under standardratemaking, ifthe sales decline persistsas more than transitory and expectedto continue in the future, the utilitycould only incorporate this decline byfiling a rate case.

Some analysts might view RD as aspecific form of "rate design." In


TABLE 2EXPECTED OUTCOMES FROM REVENUE DECOUPLINGReduced overall risk to the Little effect on incentives forutility customer-initiated conservation

Less incentive for utility to Increased rate volatilitypromote sales, and less (although probably smalldisincentive to promote energy relative to the volatility of theefficiency gas commodity cost)Base rates inversely related to Effect similar to shiftingactual sales between rate cases recovery of fixed costs to

customer charge, except forpossible intra-class subsidyeffect

Base rates would tend to be Uncertain of the risk andhigher (as the utility's average overall economic welfare effectcost would increase, assuming on consumerslower sales), although someoffset from a possible lower costof capital)

Source: Author's construct.

standard ratemaking, rate design isthe third step in designing rates (thefirst two are revenue requirement andcost allocation). Rate design involvessetting actual billing elements (forexample, the customer charge and thevolumetric charge) to recover revenuesby customer class commensuratewith the determined costs allocatedto each class. As a rate design, RDwould allow a utility to recover thesame revenues for distribution serviceirrespective of actual sales.48 In effect,RD predetermines how much inrevenues the utility will collect fromthose customer classes subject to themechanism. This fixity of revenuesreduces the risk to a utility from underrecovering its revenues and suffering acash flow deficiency.

Expected Outcomes from RevenueDecoupling

Table 2 lists the expected outcomesfrom revenue decoupling. First, itwould obviously reduce a utility's riskfrom sales fluctuations. For a utility,


this creates more stability in revenues,cash flows and earnings. Underrevenue decoupling, for example,revenue volatility for the utility causedby a downturn in the local economy orhigher gas prices leading to fewer saleswould be less pronounced. Although autility's overall risk would seeminglydecline, exactly by how much wouldrequire a sophisticated quantitativeanalysis. In the order approvingPiedmont Gas' revenue decouplingproposal, the North Carolina UtilitiesCommission said that "Piedmontargues that there is no evidence ofreduced risk to shareholders, but theCommission disagrees on the basis ofthe Company's own case.. .In a periodof declining per-customer usage, amechanism that decouples recoveryof margin from usage, withoutrequiring the utility to file frequentrate cases or increase unpopular fixedcharges, clearly reduces shareholderrisk."49 Because of the company's RDmechanism (Rider 8), the MarylandPublic Service Commission reduced

Although a utility'soverall risk mightdecline, determininghow much would requiresophisticated quantitativeanalysis.

11

Essentially, the utilitywould become indifferentto its sales.

If a utility's customerscollectively use less gas,rates could rise. Butreduced benefits would besmall relative to realizedbenefits.

12

the authorized rate of return on equityfor Baltimore Gas and Electric by 50basis points to reflect reduced revenuerisk for the utility.

Second, revenue decoupling reducesa utility's incentive to grow itssales, or to offer new services, and,simultaneously, provides a lesserdisincentive to promote energyefficiency. Essentially the utilitybecomes indifferent to the level of itssales, assuming the utility achieves thesame earnings irrespective of actualsales. This is probably more valid inthe short term. In the longer term, autility may prefer promoting sales tothe extent it helps support new capitalexpenditures, which are rate basedand consequently add to the utility'searmngs.

Third, between rate filings revenuedecoupling would result in an inverserelationship between the utility's baserate and actual sales. For example,if sales drop because of an aggressiveeffort by the utility to promote energyconservation,underrevenuedecouplingthis would increase the base rate in theabsence of a rate filing.

Fourth, as a corollary to fewersales resulting, the utility's shortrun average cost for non-gas servicewould tend to be higher. 50 Logically,as fixed costs cover less sales, averagecost would rise. The assumption oflower sales seems valid even if theutility has no special energy-efficiencyinitiatives; the reason is that RD wouldmake the utility less motivated thanotherwise to increase its sales throughpromotional practices. Since non-gasservice reflects a fixed cost business,

any sales decline induced by revenuedecoupling would have little effecton a utility's short-run non-gas costs.This outcome is implicit under a RDmechanism, as rates adjust upwardto compensate for the utility's higheraverage cost stemming from fewersales.

Fifth, RD would probably havelittle effect on customer-initiatedenergy efficiency. 51 The benefits to acustomer from using less natural gassums to the delivered price (i.e., thebase rate plus the purchased gas costs)times the amount of gas saved. For anindividual customer consuming lessgas, RD would have a miniscule effecton a utility's rates. In other words, thepresumption here is that an individualcustomer curtailing her use of naturalgas by itself would have no visibleeffect on rates since the lost revenueto the utility would be imperceptiblerelative to total revenues. On theother hand, if a utility's customerscollectively consume less gas, thiscould cause rates to rise. In this event,the benefits to individual customerfrom energy conservation couldsomewhat decline, but even herethe reduced benefits would be smallrelative to the size of the realizedbenefits. In recent years, for manyutilities the base rate for natural gas toresidential customer has fallen to lessthan 30 percent of the total deliveredprice.52 Assuming that RD causes thebase rate to increase by 2 percent withthe base rate representing 30 percent ofthe delivered price, customers wouldsee an aggregated rate increase of 0.6percent. 53 Consequently, customerswould realize 0.6 percent less benefitsfrom energy conservation.54 As


argued above, however, an individualcustomer contemplating conservationwould unlikely take into account therate increase that might ensue.55

Sixth, revenue decoupling mightcause rates to be more volatile, asthe base rate could change annually.Nevertheless, compared (say) to thevolatility of purchased gas, any rateadjustments from a RD mechanismwould likely be minimal. Between2003 and 2005, for example, in theUnited States the residential price fornatural gas rose at an annual rate ofaround 15 percent. 56 If, as in the aboveillustration, RD caused an annualincrease of0.6 percent in the residentialprice (or 2 percent in the base rate),the overall effect on customers of aRD-driven rate adjustment appearsrelatively small.

Seventh, revenue decoupling wouldhave an effect similar to a rate designchange that shifts the recovery of allfixed costs to the customer charge.57

One variation might include a differenteffect on individual customers within acertain class. For example, assumingthat any rate adjustment affects thevolumetric charge, those customersconsuming relatively more natural gasthan other customers in the same classmight be worse off with a RD thanunder a straight fixed-variable ratedesign.

Eighth, whether or not revenuedecoupling would benefit customers isuncertain. Customers can benefitwhenthey face lower risks (for example, lessvolatile gas bills) or when RD inducesthe utility to spend more dollars onenergy efficiency initiatives and be


more effective in carrying out theseinitiatives. With an added incentive topromote energy efficiency, the utility'srates would tend to be higher thanunder standard ratemaking; this couldharm those consumers who have topay higher rates but do not participatein any of the utility's energy efficiencyprograms.

Defining risk in terms of customerbill variability, RD could cause riskto increase. 58 As an example, if thedelivered gas price rises by 20 percent,and assuming a price elasticity of-0.10 (or less than one in absoluteterms), customers' bills wouldincrease. With fewer sales, under RDthe utility would adjust its base rate,aggravating the burden on customers.In another example where abnormalweather occurs, RD can actuallyreduce customer risk. Specifically, ifcolder-than-normal weather occurs,customers' bills would not rise as much:under RD, bills would be lower becauseof higher actual sales translating into adownward adjustment ofbase rates. Inthis case, RD would have a hedgingtype effect from reduced volatility ofcustomers' bills. As a last example,under RD fewer sales resulting froman economic recession would causecustomers to pay higher rates at a timewhen many might be facing financialhardship.59 Overall, the direction ofthe risk change to customers fromRD depends on the source of salesfluctuations.

Overall, whether RDbenefits customers isuncertain.

Defining risk in terms ofvariability ofcustomerbills, RD could causeincreased risk.

13

In state commissionacross the country,parties have offeredvarious arguments on themerits ofRD.

Among other arguments,advocates ofRD sayrate cases, which areexpensive, would becomeless frequent.

14

DIVERGENT VIEWS ONREVENUE DECOUPLING

In state commISSIOn proceedingsacross the country, different partieshave offered several arguments on themerits of RD. These parties includesupporters, skeptics, and opponentsof revenue decoupling. This sectionpresents a list of arguments withoutjudging their merits. The list wascomplied from a review of testimony,commission orders, briefs and otherdocuments filed in either rate casesor other regulatory proceedings. Thefollowing section of this briefing paperwill closely examine several of thearguments.

The Advocates

The maj or arguments presentedby parties, usually gas utilities andconservationists, in favor of revenuedecoupling are as follows:

A small reduction in gas salescan affect significantly a utility'searmngsUnder standard ratemaking, energyefficiency initiatives harm utilityshareholders between rate cases.The accumulation of earningslosses over the period between ratecases can be significant.60

It is unfair to have a utility promoteenergy efficiency when it harmsits shareholders, as the utilityhas a fiduciary responsibility toits shareholders in maximizingreturnsRate cases, which imposesignificant costs on utilities andcommissions, would become lessfrequent over time61

Standard ratemaking steersa utility away from initiatingenergy efficiency actions, some ofwhich may be cost-effective' or, ,when forced to promote energyefficiency activities, utilitieswill do so lackadaisically. RDis therefore critical to assure thatutilities effectively carry outenergy efficiency initiatives.A utility is entitled to a reasonableopportunity to recover fully itspreviously authorized fixed costsbetween rate filings, even whenenergy efficiency initiatives andother factors adversely affectrevenues over this periodUnless state commissionsrecognize the trend of falling gasuse per customer in base rates,earnings will inevitably fall belowauthorized levels. Even if theutility is able to lower its costsbetween rate filings, it may nothave a reasonable opportunity toearn its allowed rate of return.Unless state commissions arewilling to remove fixed costsfrom the volumetric charge, RDis the only viable alternative inprotecting shareholders' interestfrom fluctuating salesRD can actually reduce risks toconsumers by suppressing gas billvolatility62

RD eliminatesamaj orcontroversialissue in rate cases, namely, thecalculation of test-year sales63


As an alternative to RD, lostrevenue adjustment (LRA)64 fromenergy efficiency initiatives wouldrequire evaluation and verificationof savings from utility-initiatedenergy conservation programs.65

Under LRA, an incentive problemarises where a utility wouldhave an incentive to maximizemeasured or reported savings butto achieve minimal actual savingsfrom energy efficiency initiatives.By stimulating energy efficiencyinitiatives, RD can benefit both gasconsumers and society in the longrun (for example, lower consumergas bills from the pursuit of theseinitiatives)66The ability of a utility to recover itsfixed costs should not hinge on itsactual sales, over which the utilityhas little control67

Full recovery of fixed costs in thecustomer charge would reduce theincentive of customers to conservesince, at the margin, customerswould save less money fromcurtailing their gas usageRD could reduce overall gasdemand, thereby placing downwardpressure on wholesale gas pricesRD is easy for state commissionsto monitorRD improves a utility's financialsituation and lowers its risk fromthe perspective of the financialcommunityRD is critical in transforming autility from a seller of least-costgas service to a provider of leastcost energy servicesRD does not affect a utility'sincentive to minimize costs andpursue operating efficiencies68


The Skeptics and Opponents

On the other side of the debate arethose who have raised several concernsabout revenue decoupling, with theirarguments enumerated below:

In theory and practice, regulationdoes not guarantee a utility toearn its authorized rate of returnbecause of increased competition,economic trends, and changesin consumption behavior (forexample, reduced sales because ofhigh prices) and technology thatmay move against the industry oran individual gas utilityExisting conditions do not warranta true-up mechanism that passeson risks to gas consumers (i.e.,extraordinary conditions do notexist)69Consumers unequivocally bearhigher risksDeclining gas use per householdmight not persist in the future toaffect significantly the ability of autility to earn its authorized rateof return. Besides, a rate case isthe proper forum for determiningwhether the decline in per customergas use will continue and, if so,how it should affect new rates.No evidence exists to support RDas necessary for the successfulimplementation of utility-fundedenergy efficiency initiatives70

Singling out revenues for "tracker"ratemaking treatment withoutconsidering deviations in actualand test-year revenue requirementsrepresents faulty ratemaking71

Skeptics say regulationdoes not guaranteea utility will earn itsauthorized rate ofreturn,including for reasons ofchanges in consumptionbehavior.

15

Among other argumentsagainst RD, opponentssay it is a blunt toolbecause it does not takeinto account reasons fordeclining sales.

Some ofthe samearguments are beingmade on both sides oftheRD debate.

Consumer groups mightsupport RD as part ofa settlement where autility agrees to specificconcessions.

16

Better, more incremental alternatives are available for addressingthe problem of reduced sales percustomerRD is a backdoor approach byutilities to recover their fixed costsRD can lower the quality of utilityservice72

RD can destabilize rates on a yearto-year basisA core function of a gas utilityshould not include the promotionof energy efficiency73RD can reduce customer incentivesfor conserving natural gas74

RD represents a blunt tool byallowing for rate adjustmentsirrespective of the reason for adecline in salesBy itself, RD does not provideany positive incentive to a utilityto promote or support energyefficiency initiativesRD will tend to increase shortterm gas prices, as the lower salesinduced by the mechanism causean upward adjustment in the baserateRD could thwart economicdevelopmenf5

RD is more difficult to administerthan alleged by proponents

ANALYSIS OF MAJORARGUMENTS

As evident from the above lists,several arguments lie on both sidesof the RD debate. Tables 3 and 4identify individual arguments fromthe standpoint of their persuasiveness("strong" or "weak") in supportingthe different positions on revenuedecoupling. The discussion belowelaborates on these assessments

in addition to making additionalobservations, some introduced previously?6

Reaction to RD

State commissions have respondeddifferently to RD, while consumergroups in general have opposed RD.

Consumer groups mostly disfavor therisk shifting aspect of RD and theexpectation that rates will increasein the short term. Some industrialgroups question whether utilitiesshould be involved at all in promotingenergy efficiency?7 While thesegroups are correct in asserting thata utility's risk would decline, it isunclear whether consumers wouldbear more risk (see above). Consumergroups might support RD if it is partof a settlement agreement where autility agrees to specific concessions.These concessions can include acommitment by the utility to spend afixed amount of money on promotingenergy efficiency, a transfer of moniesto an independent entity to administerconservation programs, and agreementby the utility to lower its authorizedearnings because of reduced risk.


TABLE 3AN ASSESSMENT OF ARGUMENTS SUPPORTING REVENUE DECOUPLING

Stmng At'guments Weak At'guments----------------------------------

Necessary, ifnot sufficient, for a Reduced risk to customersutility's financial interest in promotingeffectively energy efficiencyStrong utility profit motive under Definite benefits to customersstandard ratemaking for promoting sales(assuming energy-efficiency promotionis an explicit commission or statepolicy)Sales largely exogenous to a utility's A higher customer charge reducingcontrol significantly the incentive for customer-

initiated energy efficiencyHistorical decline in gas use per More energy efficiency causing a decline inresidential customer generally not taken the market price of gasinto account in setting new base ratesHigh sales volatility from year-to-year Absolutely necessary for aggressive utilitycausing possible significant deviations energy-efficiency initiativesfrom targeted sales and earningsShort-run delivery costs largely fixed Ease in designStrong opposition to allocating all fixed Necessary for a utility to earn its authorizedcosts to the customer charge rate of returnPotential for energy efficiency tobenefit customersReduced risk to the utility

Less contentious than a lost revenueadjustment (LRA) mechanism

Source: Author s construct.

TABLE 4AN ASSESSMENT OF ARGUMENTS IN OPPOSITION TO REVENUE

DECOUPLING---------------------------------

Strong Arguments Weak Arguments

Need to demonstrate special conditions Uncertainty over a future decline in usefor true-up recovery of revenues per customerInappropriate to single out revenues for Lower utility service qualitytrue-up adjustmentsLess likelihood of addressing rate- More price volatilitydesign problemMore certainty of utility benefits than Reduced incentive for customer-initiatedcustomer benefits energy efficiencyUpward pressure on short-term prices, Unequivocally increased customer riskas a utility's average cost for delivery islikely to increaseIncremental options should be Feasibility of alternative initiativesconsideredPossible legal/policy precedent issues Preference for lost revenue adjustment

(LRA) mechanism

Overly broad in addressing the problemat handSource. Author s construct.

The National Regulatory Research Institute 17

Throughout their history,state commissions havebeen reluctant to approveoftrackers unless specialconditions exist.

Evidence ofperformanceofRD on balance pointsto positive results.

18

"Tracker" Mechanism

State commissions should view RD asa true-up mechanism or a "tracker."

As I have noted, as a tracker RD seemsto meet the minimum criteria appliedhistorically by state commissions inother areas of a utility's operation withapproved trackers. Throughout theirhistory, state commissions generallyhave been reluctant to approve oftrackers in ratemaking unless specialconditions exist.

Compatibility with RegulatoryObjectives and RatemakingPrinciples

RD would produce outcomes that areboth compatible and incompatible withratemaking principles and generallyaccepted regulatory objectives.

Applying these criteria (for example,Bonbright's eight criteria for settingrates78

), the positive features of RDinclude: (1) increased opportunity fora utility to earn its authorized rate ofreturn,79 (2) more revenue stability,(3) removal of disincentives for autility to promote socially desirableenergy-efficiency initiatives, and (4)elimination of a major contentiousaspect of a general rate case.

The negative features of RD thatemerge include: (1) a potential publicacceptability problem (for example,consumers complaining that a utilitycan increase rates simply because itssales have fallen), (2) more volatileand unpredictable rates,80 (3) reducedinterestin designing more efficientrates(discussed later), (4) reduced incentive

for a utility to offer innovative newservice options and rates, (5) reducedutility promotional activities thatmight be economical, (6) introductionof another tracker mechanism toratemaking that can shift risks toconsumers, and (7) possible reducedoverall economic efficiency.

Uneven Certainty of Benefits toDifferent Parties

The benefits ofRD to gas utilities aremore definitive than the benefits toconsumers in both the short and longterm.

As discussed above, RD reduces autility's overall risk while the benefitsto consumers are less transparent andmore circular (for example, to theextent RD promotes cost-effective,utility-initiated energy efficiency,consumers may realize a benefit).When RD reduces a utility's risk,consumers can benefit in the long term,for example, from a lower authorizedrate of return. 81

Evidence of Performance

The limited evidence on the ex postperformance of RD mechanisms forgas utilities on balance points to positive results.

From the perspective of a senior staffmember oftheMarylandPublic ServiceCommission, the Baltimore and GasRD mechanism (Rider 8) has achievedthe intended goals since its inceptionover seven years ago. Specifically,it has: (1) produced more stable andpredictable revenues for the utilitybetween rate cases by accounting forrevenue "attrition" caused by declininggas use per customer, (2) reduced


the volatility of gas bills, especiallyunder cold weather conditions, and (3)allowed for the continuation of currentrate designs that provide an incentivefor consumers to conserve and thatare non-discriminatory to low usagecustomers. The staffperson also addedthat the RD mechanism is easy for theutility to administer and the commission to monitor. Overall, the staffmember concluded that the mechanismhas "[fulfilled] more regulatoryobjectives with fewer shortcomingsthan other alternatives."82

A study conducted for NorthwestNatural concluded that: (1) by reducingrevenue fluctuations, the DistributionMargin Normalization (DMN) mechanism has reduced the utility's businessand financial risks, (2) DMN marginadjustments can largely be attributedto the effect of price changes, witheconomic activity and the utility'sfunded energy efficiency efforts having a statistically insignificant effecton use per customer, (3) the utility'sfocus has shifted from marketingto promoting energy efficiency, (4)service quality did not decline, and (5)most ofthe risk reductions experiencedby the utility were eliminated ratherthan shifted to customers. Whilemaking several recommendationsfor improving Northwest Natural'sDMN mechanism (for example, fulldecoupling), the study concluded, "Thepositive effects of DMN outweigh thenegative effects."83 To date, this studyrepresents the most comprehensiveand analytical ex post investigation ofa RD mechanism for gas utilities. 84


Alternative Mechanisms

Alternatives to RD in achieving thesame objectives might be preferable,as RD is a more blunt approach thanmost alternatives.

These alternatives can include:(1) raising the customer chargeby removing fixed costs from thevolumetric charge, (2) implementingor expanding weather-normalizationadjustments, (3) implementing declining block rates, (4) using a multiyear forecast horizon in setting newrates, and (5) implementing a targetedincentive plan which allows a utility toprofit from successfully carrying outsocially desirable energy-efficiencyinitiatives. Since each of these alternatives has its own drawbacks, a statecommission might want to assesstheir desirability compared with a RDmechanism. 85

Legal/Policy Questions

For some state commissions, the legality ofRD as well as its compatibilitywith policy precedent may be anissue.

This was the case, for example, inboth Minnesota and North Carolina.In Minnesota, the state's Departmentof Commerce argued that the RD proposal by Xcel violates state statutes,which in its opinion do not providea statutory exception for a true-upmechanism that adjusts rates based onthe level of gas use per customer. InNorth Carolina, two commissionersdissented from a commission orderapproving a RD mechanism forPiedmont Natural Gas by arguing,

State commissions mightwant to assess severalalternatives to RD,although each ofthosehas its own drawbacks.

Legality ofRD may be anissue in some states.

19

In considering RD, astate commission needsto include design andimplementation, as wellas policy issues.

Is a gas utility in thebusiness ofsellingnatural gas, or, morebroadly, the busienssofselling energyservices that includeconservation?

Regulators shouldnot expect utilities toundertake pro-activelyenergy-efficiencyinitiatives that leadto deterioration ofshareholder interests.

20

among otherthings, thatthe mechanismviolates North Carolina law byreflecting retroactive ratemaking. 86

The dissenters also argued that as amatter ofpolicy state commissions haveapproved true-up mechanisms onlyunder "extraordinary circumstances,"which they contend did not applyin the case of declining gas use percustomer. In first considering a RDproposal, a state commission shouldreview its legal authority in additionto policy precedent in allowing for atrue-up between rate cases based onactual sales relative to "baseline" salesestablished at the time of the previousrate case.

Implement/Design Issues

A state commission needs to addressseveral design and implementationissues, assuming that the concept ofRD has broad stakeholder support.

These include: (1) scope of themechanism in terms of factors(for example, weather and priceelasticity) that should be included indetermining sales adjustments, (2)rate classes affected, (3) frequencyof rate adjustments, (4) the need fora rate-adjustment cap (for example,limit annual rate adjustment to 5percent ofthe base charge), (5) revenueadjustments from new customers, (6)treatment of any cost-of-capital effect,(7) pilot or permanent status,87 (8)accounting for overall utility earningsby considering cost changes over time,(9) proper forum for consideration(rate case filing, special docket), (10)accounting for quality-of-serviceeffects, (11) adjustment to specificrate components (for example, the

volumetric charge or the customercharge), and (12) true-up of pnorperiod RD adjustments.

One Rationale for RD

The merits ofRD crucially depend onthe state commission's vision ofa gasutility in promoting energy efficiencyin addition to its core function ofselling and delivering natural gas.

In considering RD, a state commission might first want to considerwhether a gas utility should be in thebusiness of selling natural gas anddelivery service or, more broadly,of selling energy services, whichinclude energy conservation. If thelatter is preferred, then RD becomesa more tenable ratemaking tool. 88 Ifnot, then a commission should assessRD in terms of the "declining gas useper customer" phenomenon. In otherwords, if a state commission requiresa gas utility to promote aggressivelyenergy efficiency, or if there is strongevidence of large benefits from utilityfunded energy efficiency initiatives,RD has definite merits as a ratemakingmechanism. 89

Regulators should not expect autility to undertake pro-activelyenergy-efficiency InItiatIves whenshareholder interests deteriorate. Acollision course leading to unintendedconsequences seems inevitable understandard ratemaking from requiringa utility, whose earnings directlyrelate to the level of sales, to play anindependent active role in reducing itssales. Furthermore, if a commissionapproves RD, it could require autility to be committed to promotingaggressively energy efficiency.9o


A Second Rationale for RD

In addition to promoting energyefficiency initiatives, RD may also bedefensible.

Specifically, three conditions wouldsupport RD. First, a state commissionbelieves that consumption per customeris likely to decline in the future.Second, the utility's ability to add newcustomers in growing sales is greatlylimited. Third, the commission'sdiscretion to recognize decliningconsumption per customer in settingbase rates is constrained because ofstatutory or commission restrictions onpro forma adjustments to include only"known and measurable changes" to ahistorical test year.91 If cost recovery,namely, recouping earnings lossesfrom an unexpected decline in sales percustomer, is the main rationale for RD,as discussed above a state commissionhas other options available to addressthis problem.

Negative Economic-EfficiencyEffects

From an economic-efficiency perspective, RD has some negativeattributes.

One as discussed earlier, stems fromthe ~xpectation that RD would causea utility to shift farther away fromthe minimum point of its short-runaverage cost curve, assuming that themechanism will result in lower salesthan otherwise. Since gas distributionis essentially a fixed cost function inthe short run, higher average costwould result from reduced sales, sinceby definition average cost equals total


cost divided by sales. In addition, apoint made above, RD makes the utilit.yindifferent to increasing sales even Ifit would be economical. For example,if the base rate is 50 cents per thermand the short-run marginal cost for gasdistribution service is zero, under aRD mechanism the utility would haveno incentive to increase sales eventhough it would clearly be economicalto do so. This outcome is contrary tothe universal corporate objective ofspreading fixed costs over additionalsales to produce greater operationalefficiencies.

Another conceivable source ofeconomic inefficiency derives fromthe presumption that a utility with aRD would have little or no incentiveto modify its rate design by removingmore of its fixed costs from thevolumetric charge, which if nothingelse would remove what some analystsconsider a major source of prevailingprice inefficiency in the utility secto~.92

With RD protecting a utility's financIalcondition from sales volatility, therewould seem to be little payoff to theutility from initiating any subsequentchange in rate design that removes fixedcosts from the volumetric charge. Inaddition, RD would tend to exacerbatepricing inefficiency by widening thegap between price and marginal costfor non-gas service whenever sales fallbelow the specified "baseline."93

With RD, there wouldbe little payoffto theutility from initiating anysubsequent change inrate design that removesfixed costs from thevolumetric charge.

21

An earnings sharingmechanism has theattractive feature oftreating symmetricallycosts and revenuedeviations.

While generallysupportive ofRD,gas utilities andconservationists telldifferent stories aboutwhy.

22

Accounting for Price Elasticity

State Commissions should recognizethe price-elasticity effect on futuregas consumption when setting newrates, or example, by liberalizingtheir interpretation of "known andmeasurable change."

As shown above, with the dramaticincrease in natural gas prices over thelast few years, and even with only asmall price-elasticity response byconsumers, the effect on a utility'searnings can be significant. 94 Becauseof the high uncertainty over the effectofthe recent natural gas price increaseson consumption, a true-up mechanismsuch as revenue decoupling may haveadded merit.

Earnings Sharing as an Alternative

Ifa state commission is concerned thata utility will not have a reasonableopportunity to earn its authorized rateofreturn, short offiling a rate case, itshould consider an earnings-sharingapproach, rather than adjusting ratessolely because of "unanticipated"sales.

Actual costs are likely to differfrom test-year revenue requirementsfor many reasons.95 It may beinappropriate, therefore, to adjustrates when actual sales deviate from"baseline" or test year sales whilenot making adjustments for expensesand other revenue-requirementcomponents of the base rate. Forexample, between rate cases all ofthe cost savings from productivityimprovements not anticipated at thetime of the last rate case, would flow toa utility's shareholders. An earnings-

sharing mechanism has the attractivefeature of treating symmetrically costsand revenue deviations. Under RD, itis conceivable for a utility to have bothits base rate adjusted upward betweenrate cases in response to a decline insales per customer and, at the sametime, earning a rate of return abovethe authorized level because of actualexpenses reduced below the test yearestimates. One of the shortcomingsof an earnings-sharing mechanism,however, is that the utility's shareholders could still suffer from lowersales to the extent they absorb a portionofthe realized earnings 10sses.96 Thus,the utility would have a disincentive,as under standard ratemaking, to fosterenergy efficiency.

SUMMARY ANDCONCLUSIONS

Over the past few years, championsof revenue decoupling mechanismshave included conservationists andgas utilities. The federal governmenthas become more active in reviewingthis ratemaking tool for both electricand gas utilities, for example,pursuant to EPAct of 2005, and othergovernmental entities have endorsedor are examining RD as well. Whilegenerally supportive of RD, gasutilities and conservationists telldifferent stories on the desirability ofthis ratemaking mechanism.

Other stakeholders in the stateregulatory process have expressedseveral concerns with RD. StatecommISSIOns themselves haveresponded differently tothe desirabilityof RD as a ratemaking mechanism.Unquestionably, RD helps to preserve


the financial integrity of utilities andmotivate them to be less opposed topromoting energy efficiency. WhetherRD would benefit consumers is lesscertain. It is partially for this reasonthat the debate over RD mechanismsat the state level, to date, has centeredon conceptual and theoretical issues,specifically on whether RD offersconsumers any advantages overthe standard ratemaking standardapproach and is compatible withprevailing regulatory objectives.

The central point of this briefingpaper is that state commissions shouldconsider RD for those gas utilitiesrequired or pressured to promoteaggressively cost-effective, energyefficiency initiatives. RD would reducethe incentive for a utility to undertake ineffective energy-efficiencyinitiatives in addition to overcomingthe predicament of a utility inimplementing those initiatives when itwould adversely affect its shareholders.One corollary of this is that if a statecommission allows RD, it may want toconsider requiring a utility to engagein serious energy efficiency efforts byspending more-than-minimal dollarson energy efficiency and undertakingall cost-effective initiatives.

RD may also have merit if a statecommission is uncertain of future gasconsumption and believes that gasuse per customer will likely continueto decline in the future, for example,because of the elasticity effect fromhigh gas prices. The commissionmight be constrained from accountingfor this phenomenon in setting newrates because of the lack of evidencejustifying a "known and measurable"


adjustment. In this case, RD wouldact as an attrition allowance protectinga utility's earnings between ratecases from less-than-expected gasconsumption.

State commissions and consumergroups have rightly raised concernsabout some of the negative featuresof RD. One generally expressedmisgiving with RD is that, whilenecessarily beneficial to the utility inreducing its risk, it might be inimicalto consumers. Some skeptics viewRD as akin to taxing consumers forthe benefit of protecting utilities fromfinancial harm when revenues fallshort of some predetermined level.While this perception arguably is amisrepresentation, it may lie at theheart of the equivocation by statecommissions and consumer groups tothis ratemaking mechanism. At theleast, the concern with RD may requirea utility to appease the doubters bycommitting to energy efficiency or byagreeing to a downward adjustment ofits authorized rate of return on equityas compensation.

Overall, the jury is still out on how statecommissions will rule onRD proposalsin the future. If commissions viewgas utilities as purveyors of energyefficiency services, they will likelybe more receptive to a mechanismthat would keep utility shareholdersfinancially whole in addition toremoving any disincentive for a utilityto promote actively those presumablysocially desirable services. After all,if RD results in only slightly higherrates, but achieves large benefits- or at least the perception of largebenefits to consumers from utility-

The central point ofthisbriefing paper is thatstate commissions shouldconsider RD for thosegas utilities required topromote aggressivelycost-effective, energyefficiency.

23

24

funded, energy-efficiency actIvItIes- the public will look more favorablyupon the commission and utility intheir endorsement of this ratemakingmechanism.


Notes

1 This rate design initiated decades ago with a major objective of motivating gas utilities to sign up new customers and to increase theirsales. Until the dramatic rise in gas prices around 2000, public policy and utility actions were generally supportive of growing gassales. That posture no longer holds, as the emphasis in the last few years has shifted toward controlling growth in gas sales.

2 In recent years, state commissions have seen the filing of numerous rate cases by gas utilities. These filings, for some utilities the firstin over a decade, are the result of eroding profits caused by a combination of higher costs, required capacity expenditures (partly theresult of customer growth and new safety regulations), and flat demand growth. In many of these rate filings, the utility petitioned fora change in rate design, notably the shifting of fixed costs from the volumetric charge to the customer charge.

3 Gas utilities have assigned different labels to their revenue decoupling (RD) proposals: Conservation Margin Tracker, ConservationEnabling Tariff, Conservation Tariff, ConservationRider, Conservationand Usage Adjustment Tariff, Conservation Tracker Allowance,Margin per Customer Balancing Provision, Delivery Margin Normalization, Usage per Customer Tracker, Customer UtilizationTracker.

4 The financial community has also looked favorably upon revenue decoupling in reducing a utility's risk and improving its financialstability.

5 In several presentations over the past couple of years, AGA has argued that RD can: (1) reduce natural gas consumption over time,(2) lower bills to consumers, (3) increase utility-funded energy efficiency initiatives, and (4) provide a reasonable opportunity for autility to earn its authorized rate of return. Specifically, AGA has argued that RD can benefit consumers by: (1) lowering gas consumerbills over time, (2) reducing uncol-lectible bills, and (3) reducing overall gas demand that will place downward pressure on market gasprices. (See, for example, Roger Cooper, "Creating a Win/Win Natural Gas Distribution Energy Efficiency program: Recognizing andAligning Stakeholder Interests," presentation at the Indiana Utility Regulatory Commission Gas Issues Forum, July 29, 2005.)

6 See Joint Statement of the American Gas Association and the Natural Resources Defense Council, submitted to the NationalAssociation of Regulatory Utility Commissioners, July 2004.

7 See National Association of Regulatory Utility Commissioners, "Resolution on Energy Efficiency and Innovative Rate Design,"adopted Nov. 16, 2005.

8 See Russell A. Feingold, "Decoupling, Conservation, and Margin Tracking Mechanisms - An Overview," AGA Rate and RegulatoryIssues Audio Conference Series, Oct. 27, 2005.

9 As noted by Eric Hirst in a 2004 report prepared for Idaho Power Company, titled Decoupling for Idaho Power Company, "Duringthe 1990s, various forms of decoupling were deployed in Maine, New York, California, and Washington [for electric utilities]. Duringthe mid-1990s, these efforts were largely abandoned as utilities and state regulators anticipated a restructured, competitive electricityindustry, although Oregon began decoupling in the late 1990s" (at 3).

10 California is the only state that has embraced RD for electric utilities. The state initially instituted RD in the state in the 1980spartially because of evidence of a fuel oil shortage, and near doubling of oil prices and interest rates; the expectation was a declinein electric utility sales in response to these events. Incidentally, in California RD is currently under consideration for water utilitieslargely to encourage utility-funded energy conservation efforts.

11 Pacific Gas and Electric was the first gas utility subject to a RD, starting in 1978. The state commission rationalized the mechanismby the fear of gas curtailments and the expected erosion of earnings from the combination of lower sales and inverted rates. (See BruceSmith, "Revenue Sales Adjustments," presentation before the Harvard Electricity Policy Group, March 3,2005.)

12 The Baltimore Electric and Gas mechanism (Rider 8) measures test-year base rate revenues after adjusting for any change in thenumber of customers from the test-year level. The mechanism adjusts test-year revenues by accounting for the net number of customersadded since the test year. The difference between actual revenues collected and the recalibrated test-year revenues determines the rateadjustment. In effect, the mechanism is a "true-up" that accounts for customer growth as this element could offset lower per-customergas usage the mechanism is intended to capture.


13 The Southwest Gas mechanism, designed as a margin tracker, balances actual margin revenues to authorized levels. The CaliforniaPublic Utilities Commission approved this mechanism in 2004.

14 The staff of the Oregon Public Utility Commission supported Northwest Natural's original RD mechanism only after the utilityagreed to implement a service quality standard and to transfer permanently the utility's energy conservation programs to a selectedindependent entity approved by the commission.

15 North Carolina Utilities Commission, Order Approving Partial Rate Increase and Requiring Conservation Initiative, Docket Nos.G-9, Sub 499, G-21, Sub 461 and G-44, Sub 15, Nov. 3, 2005, at 23-24. Chuck W. Fleenor filed company testimony in support of CUTon April 1, 2005.

16 Ibid. 24.

17 See Vectren Energy Delivery of Ohio, Application ofVectren Energy Delivery ofOhio, Inc. for a Conservation Rider, Case No. 051444-GA-UNC, Nov. 28, 2005, 4.

18 See the testimony of Dr. George R. Compton, Docket No. 05-057-TOl, Jan. 23, 2006, at 11.

19 See Quote.com, Business Wire, "South Jersey Proposes Innovative Conservation and Usage Tariff; Measure Will Help CustomersSave Energy and Lower Heating Bills," Dec. 5,2005.

20 The legislature required the commission to investigate how decoupling the earnings of gas and electric utilities from their sales canbest promote the state's energy policy.

21 The Connecticut Department ofPublic Utility Control, DPUCInvestigation into Decoupling Energy Distribution Company Earningsfrom Sales, Docket No. 05-09-09, Draft Report, Jan. 12, 2006. In addition to sales and per customer adjustment clauses, the studydefines decoupling as also encompassing a conservation and load management adjustment clause and rate design changes that reducethe effect of sales changes on a utility's earnings.

22 The Department used the term "business risk" in explaining how a revenue-decoupling plan would reduce the risk of a utility fromsales fluctuations. Generically, business risk refers to the uncertainty linked to the operating cash flows of a business. Business riskis multi-dimensional, inclusive of both sales risk and operating risk. Some analysts may categorize this risk as "regulatory risk,"which relates to a regulator's actions that affect a utility's ability to earn a fair rate of return. For RD, regulatory risk may arise fromthe uncertainty associated with the recovery of revenue deferrals (i.e., the difference between actual revenues and targeted revenues,accumulated over time).

23 The Department found that "Past experience has shown that customer-initiated conservation by gas LDC customers can yield usagereductions of up to 3.5 percent annually in response to high prices" (Ibid. 19).

24 The Attorney General argued that a revenue decoupling mechanism would reduce customers' incentive to conserve.

25 In Minnesota, the state's Department of Commerce and Office of Attorney General challenged a RD mechanism proposal by Xcel.The utility subsequently withdrew the proposal as part of a rate case settlement. Arguments in opposition to the utility's proposalincluded: (1) it is questionable whether the absence ofRD would deprive the utility of a reasonable opportunity to earn its authorizedrate of return, (2) the proposal violates state statutes, (3) the proposal removes the incentive of a utility to forecast sales accurately ina rate case, (4) it was unsubstantiated whether gas use per customer will continue to fall in the future, and (4) under the proposal, theutility could adjust its rates upward even if declining gas use per customer is offset by an increase in the number of customers. (SeeVincent C. Chavez, Direct Testimony and Exhibit, before the Minnesota Public Utilities Commission, Feb. 11, 2005.) In 2005, theNevada Public Utilities Commission rejected a RD proposal by Southwest Gas, arguing in part that the proposal would constitute amajor change from current ratemaking practices and before it can be justified "more recognized alternatives" (such as changes in ratedesign and more frequent rate filings) should be applied to the perceived problem (i.e., reduced earnings from less-than-expected gassales) (Nevada Public Utilities Commission, Opinion, Docket No. 04-3011).

26 Southwest Gas proposed a "Conservation Margin Tracker" in anticipation of a continuation of the past trend of declining gas useper customer.

26 The National Regulatory Research Institute

27 Arizona Corporation Commission, Opinion and Order (Decision No. 68487), Feb. 23, 2006.

28 The published study provided no explanation how normalized consumption was calculated. While presumably AGA applied aproper methodology to normalize actual consumption, a reader of the study is unable to verify independently.

29 The AGA study estimated that almost half of the decline per household since 1997 resulted from space-heating efficiency gains.

30 The study attributes a slower decline in usage per customer in the future to the construction of larger houses and the expectation ofonly modest furnace efficiency gains. (See American Gas Association, Forecasted Patterns in Residential Gas Consumption, 20012020, EA 2004-04, Sept. 21, 2004.) Reduced consumption per customer does not imply that utilities' total gas sales to residentialcustomers will fall in the future (see, for example, Energy Information Administration, Annual Energy Outlook 2006, February 2006and other projections). Most studies expect moderate growth in total residential sales over the next several years, even in view ofa continued decline in sales per residential customer (with growth varying by state and region). This means that utilities' revenuesfrom residential sales should grow between rate cases because of the addition of new customers, notwithstanding a decline in use percustomer.

31 According to a 200S-RAND study, titled Regional Differences in the Price-Elasticity ofDemandfor Energy, the demand for naturalgas and electricity is relatively price-inelastic; in the past 20 years, this relationship has not changed significantly, as analyses performedin the 1980s showed generally the same result. Nevertheless, as the study argues, because prices were declining in real terms over mostof the period studied, inelasticity of demand may be a product of the absence ofprice increases. The study, applying data for the period1977-2004 to econometric methods, estimated the short-run price elasticity of demand for residential natural gas as equal to -0.12 andthe long-run price elasticity as -0.36. Although the results confirmed price-inelastic demand, as illustrated elsewhere in this paperconsumers' response to sharply higher gas prices can nevertheless have a significant effect on a utility's short-term earnings.

32 The implicit argument by gas utilities is that, from a revenue perspective, the growth in the number of customers does not offsetdeclining gas per customer, resulting in total actual revenues less than the test-year revenues.

33 Larger declines may be occurring in areas with a high growth of new housing construction, with the reasoning that new homes onaverage are more energy efficient and use less energy than older houses.

34 For the electric industry, where some utilities have proposed RD, the historical phenomenon of declining use per customer doesnot hold (use per customer has continuously increased over time), making less tenable the adoption of RD for this industry unless astronger case can be offered for electric utilities to promote energy efficiency than for gas utilities.

35 Options for utilities to address declining usage per customer include raising the customer charge, implementing declining blockrates, using a multi-year forecast horizon, and innovative strategies, such as revenue decoupling. These alternatives may pose political,policy or legal obstacles that make revenue decoupling the only viable and attractive alternative. For example, an adjustment made toaccount for declining usage per household would be feasible under a future test year but would be more difficult to justify under anhistorical test year.

36 Many state commissions that apply histori-cal test years in rate cases allow utilities to adjust their sales forecasts for weathernormalization. Other state commissions apply forecasted test years, with the allowance of weather-normalized sales projections.Weather-normalized sales projections arguably allow a gas utility a reasonable opportunity to meet its forecasted sales between ratecases.

37 Normally when a commission sets an authorized rate of return, it specifies a zone of reasonableness within which the commissionconsiders a rate of return to be fair. For example, a commission may rule that a reasonable rate of return lies between 9 and 11 percent,with 10 percent, the mid-point, as the authorized rate of return.

38 Most industry observers presently see the changed post-1999 market conditions as structural, with sustained effects, rather thancyclical. They expect no significant relief in natural gas prices until 2008 or later. In contrast, price spikes experienced in the 1990swere short-lived, caused largely by brief periods of unusually cold weather or regional pipeline bottlenecks. In terms of the effecton the economy, most analysts see the combination of high natural gas and oil prices slowing down short-term economic growth inaddition to exacerbating inflationary pressures.


39 Under a straight fixed-variable (SFV) rate design, a customer charge includes 100 percent of the fixed costs and the volumetric chargecontains only those costs that vary with sales. One outcome of this rate design is that whenever a utility experiences lower or highersales, the utility recovers the same amount of its fixed costs. In other words, its rate of return remains the same. Relative to alternaterate designs (with the exception of revenue decoupling), SFV reduces the utility's risk, with the recovery of fixed costs "up-front,"irrespective of actual gas sales.

40 Exceptions may include variable non-gas costs. Examples include winter high-pressure service, as the sizing of high-pressurefacilities correlates with demand under extreme weather conditions, and the foregone revenues from storage services sold off-system.(See, for example, Hethie S. Parmesano, testimony before the Illinois Commerce Commission, on behalf of NICOR Gas Company,November 2004.)

41 This assumes the fixity of all non-gas costs, thus invariant to changes in sales.

42 It is assumed that R* was determined in the last rate case and the source of equity capital is equal to $333 million ($40millionl0.12).

43 In equation (3), with I1VC equal to zero, the proportional change in earnings simplifies to the ratio of the change in revenues toearnings (I1R/E). In the illustration, the ratio R/E equals 10, which, based on an examination of rate-case filed income statements forseveral gas utilities, seems to correspond to real-world conditions.

44 RD is arguably more justified, as a true-up mechanism, than purchased gas adjustment (PGAs) clauses since a utility has somecontrol over gas-supply costs because of the ability of most utilities to choose among different supply and transportation sources, andcommercial arrangements. PGAs work similarly to RD by protecting the utility from financial losses when purchase gas costs exceedbase levels.

45 In most applications of RD, "baseline" sales account for new customers, with revenue adjustments set on a per customer basis.

46 In practice, the concept of a fair or authorized rate of return reflects a "zone of reasonableness" within which there is a highprobability that the fair rate of return lies.

47 In most states, purchased gas cost includes interstate pipeline costs.

48 Depending on how the RD mechanism is structured, actual sales may represent either total sales or per customer sales averagedacross (say) the residential class.

49 See North Carolina Utilities Commission, Order Approving Partial Rate Increase and Requiring Conservation Initiative, DocketNos. G-9, Sub 499, G-21, Sub 461 and G-44, Sub 15, Nov. 3, 2005, 24.

50 A reduced cost of capital, which might result from revenue decoupling, could offset the higher average cost.

51 "Customer-initiated" refers to customers consuming less natural gas because of higher prices or other factors that are largely marketor economy driven.

52 In 2005, for example, the city gate price of natural gas in the United States averaged $8.64 per Mcf and the residential price of gasaveraged $12.82 per Mcf. (i.e., the city gate price was about 67 percent of the residential price). See Energy Information Administration,Monthly Energy Review, March 2006, Table 9.11.

53 The 2-percent increase approximates a decline in net revenues of the same proportion (assuming minimal variable distributioncosts). If one takes the historical decline in average gas use per household of one percent, and adjusting for revenue increases from newcustomers (which occurs under most existing and proposed RD mechanisms), the 2-percent example would seem to be much higherthan expected in a real-world situation.


54 If the delivered price is (say) one dollar per therm and the customer buys a high-efficiency furnace that saves annually 50 thermsof gas usage, the customer can expect her gas bill to decline by $50. If all customers conserve by the same amount, a RD mechanismwould result in the base rate increasing (say) by 2 percent and the delivered price by 0.6 percent, the benefits to the customer from thisconservation action would decline by only 30 cents.

55 RD could induce additional customer-initiated energy efficiency to the extent that the utility more aggressively and effectivelyeducate customers on the benefits of energy efficiency.

56 Energy Information Administration, Short-Term Energy Outlook, March 7,2006, Table A4.

57 In other words, RD represents a second-best "solution" for poor rate design that includes fixed cost in the volumetric charge. Someproponents of RD view the mechanism as a form of rate design that separates the recovery of fixed costs from volumetric sales, whichis exactly what a straight fixed-variable rate design does.

58 Joseph Eto et aI. discussed how risks to consumers would be affected by revenue decoupling. (See Eto et aI., The Theory and PracticeofDecoupling, Report LBL-34555, Berkeley, CA, January 1994.)

59 In Maine, lower sales levels in the early 1990s caused by an economic recession (in the previous rate case, the sales forecast wassubstantially too high) led to substantial revenue deferrals that Central Maine Power was ultimately entitled to recover under itsrevenue decoupling mechanism ("Electric Revenue Adjustment Mechanism"). The utility did not recover these deferrals, however,because of an agreement with parties to avoid immediate rate increases during bad economic times. By the end of 1992, the revenuedeferrals had reached $52 million. Observers argued that only a small portion of this amount was the result of the utility's conservationefforts and that the vast majority of deferral, instead, resulted from the economic recession. There was the general perception thatthe revenue decoupling mechanism shielded the utility against the financial consequences of the recession, rather than providing theintended conservation incentive impact. Termination of the RD mechanism occurred in November 1993. (See, for example, MainePublic Utilities Commission, Report on Utility Incentives Mechanisms for the Promotion ofEnergy Efficiency and System Reliability,presented to the State Legislature, Feb. 1, 2004.)

60 In testimony before the Wisconsin Public Service Commission, involving Wisconsin Power and Light in Docket No. 6680-UR-1l4,Ralph Cavanagh argued that since many energy efficiency measures last ten years or more, single year impacts would multiply at leasttenfold when assessing shareholder interests. He illustrated this with an example: if total fixed costs are $80 million, with 90 percentrecovered in volumetric charge (or $72 million), a one percent decline in consumption means that $0.72 million is lost to shareholdersduring the first year; over five years, the accumulated loss to the utility would be $10.8 million ($0.72 + $0.722 + $0.723 + $0.724 +$0.725

).

61 This could occur to the extent a decline in gas use per customer helps to trigger a rate filing. A primary rationale for purchased gasadjustment clauses was that they would reduce the need for a utility to file rate cases continuously.

62 As noted above, reduced gas bill volatility would occur if abnormal weather causes the revenue adjustments.

63 Test year sales affect the required rate change to recover revenue requirements, with a pessimistic projection justifying a higher rateand an optimistic projection a lower rate. Under a RD mechanism, any bias in the projections will be trued-up over time.

64 A LRA confines itself to saleslrevenues losses from a utility's energy efficiency initiatives. In contrast, a decoupling mechanismcovers all changes in a utility's sales.

65 As expressed by one analyst, methods for verifying savings resulting from energy efficiency programs are likely to be "complex,tedious, and expensive." Another criticism of LRA is that, because of its narrowed focus, the utility would still have an incentive toincreased sales through its other activities.

66 Society may benefit from lower environmental costs associated with the reduced production, transportation, and consumption ofnatural gas.

67 The argument here is that since previously a commission ruled that these fixed costs were prudent, the utility has the legal right torecover them fully.


68 The utility would still have an incentive to maximize its earnings, but under RD it would have different incentives and constraints.Specifically, since presumably increasing sales between rate cases will have no effect on earnings, the utility would instead focus onreducing costs. If a utility allows its costs to rise above test-year costs, it would jeopardize its ability to earn its authorized rate ofreturn, irrespective of the adoption of a RD mechanism.

69 This argument centers on the common perception that commission endorsement of true-up mechanisms or "trackers" requires theexistence of unusual circumstances because of the risk shifting to consumers and, in the case of cost true-ups such as purchased gasadjustment clauses, the weakened incentive of a utility to manage costs.

70 In Arizona, for example, the Residential Utility Consumer Office (RUCO) argued (in Arizona Corporation Commission Docket No.G-0l551A-04-0876) that in acknowledging the disincentive under standard ratemaking for a utility to promote energy efficiency, thecommission or a third party could be responsible for administering energy efficiency programs funded by the utility.

71 Commissioner Lorinzo Joyner's dissent in the Piedmont Gas case in North Carolina echoes this view: "The issue is whether there is ajustification for tracking and making rate adjustments based in changes in consumption without consideration of other factors affectingoverall expenses and revenues. I agree with the Attorney General that maintaining margin recovery is not sufficient justification initself and is not consistent with fundamental ratemaking principles which disfavor rate adjustments based on one element of rates'without appropriate regulatory oversight of the Company's overall expenses, sales volumes, and revenues.'" (See North CarolinaUtilities Commission, Order Approving Partial Rate Increase and Requiring Conservation Initiative, Docket Nos. G-9, Sub 499, G-21,Sub 461 and G-44, Sub 15, Nov. 3, 2005.)

72 This argument stems from the presumption that RD would remove the financial consequences for a utility from losing sales becauseof poor quality of service. For firms in most other industries, revenues are highly correlated with the quality of service offered tocustomers. Under RD, as the argument goes, the benefits to a utility from achieving a high quality of service would decline, if noteliminated.

73 Industrial customers in particular have raised this point. Other parties have questioned the role of utilities in promoting energyefficiency in light of market forces and the potential role for third parties to implement energy efficiency programs more effectivelythan the utility.

74 The logic is as follows: customers as a group would benefit less from conservation with RD as their rates would increase with utilitywide conservation, since some of the savings would flow to the utility to offset the reduction in distribution revenue collected thoughtenergy sales. Consequently, customers would receive less of the savings than they would without decoupling.

75 The reason given is simply that RD could remove the incentive of a utility to expand its sales by attracting new customers witha promotional rate or other innovative rate design, since any revenue gains may flow back to consumers in the form of lower baserates.

76 Specifically, the subsection "Expected Outcomes from Revenue Decoupling" addressed them.

77 One industrial advocate has argued that RD makes a utility indifferent to its sales volume, thus reducing its motivation to accommodatecustomer needs. He also posed the fundamental problem of a utility simultaneously selling its core service and promoting energyefficiency as conflicting, in his opinion requiring the state commission to set up additional oversight to monitor and regulate theseconflicting activities. (See Maurice Brubaker, presentation before the Harvard Electricity Policy Group, March 3,2005.)

78 The eight criteria for a desirable rate design are: (1) simplicity, understandability, public acceptability and feasibility of implementation,(2) uncontroversial as to proper interpretation, (3) effectiveness in providing the utility with adequate revenues to recover costs, (4)year-to-year revenue stability, (5) rate stability, (6) fairness among customer classes, (7) avoidance of undue price discrimination,and (8) economically efficient in giving customers proper price signals, for example, in not over-consuming a utility's service. Thesecriteria can be conflicting, with no rules of ranking offered. It is often difficult, if not impossible, to satisfy all of these criteria(for example, public acceptability and efficient pricing); almost all real-world ratemaking outcomes reflect compromises betweenthe different regulatory objectives. Bonbright also identified the four primary functions of public utility rates as capital attraction,efficiency, demand rationing, and income distribution. (See James C. Bonbright et aI., Principles ofPublic Utility Rates, 2nd Edition,Public Utilities Reports, Inc., 1988; the first edition, authored solely by Bonbright, was published in 1961.)


79 Obversely, RD would also reduce the opportunity of a utility to over-earn, which may explain why some gas utilities do not supportRD.

80 As discussed above, however, the monthly or yearly rate adjustments subject to RD would likely be small relative to the totaldelivered price of natural gas and to other factors affecting price (such as fluctuations in purchased gas costs).

81 In Oregon, the upgrading of Northwest Natural's Standard & Poor bond rating occurred shortly after the commission approval of aRD mechanism.

82 See Calvin Timmerman, "Monthly Rate or Revenue Adjustments: Regulatory Perspective," presentation at the Platts Rate Case andPricing Symposium, Oct. 25, 2004, and Calvin Timmerman, "LDC/EDC Revenue Decoupling," presentation to the 2006 MACRUCCommissioner Only Strategic Planning Session, April 3, 2006.

83 In its investigation of revenue decoupling, the Connecticut Department of Public Utility Control criticized the Northwest Naturalmechanism for being administratively burdensome and not linked to the energy conservation programs funded by the utility.

84 See Christensen Associates, A Review ofDistribution Margin Normalization as Approved by the Oregon Public Utility Commissionfor Northwest Natural, March 31, 2005.

85 For example, higher customer charges could (1) lead to customer complaints about "minimum" bills during the low-use months, and(2) disproportionately impact low income/low usage customers. A straight fixed-variable rate design for gas distribution service is rarein the United States, with Xcel in North Dakota and Atlanta Gas Light as the only gas utilities currently having this rate design. In thecase of Xcel, the company originally proposed a partial decoupling rider but was withdrawn as the parties to a settlement agreementconcurred in shifting the fixed distribution costs to a monthly basic service charge. For residential customers, the basic servicecharge increased from $5.50 to $15.68, or by 185 percent. The expectation is for the new rate design to increase summer gas bills forresidential customers and reduce their winter bills. (See North Dakota Public Service Commission, Northern States Power CompanyNatural Gas Rate Increase Application, Order Adopting Settlement, Case No. PU-04-578, June 1,2005.)

86 The majority, however, argued that the RD proposal does not violate state statute against retroactive ratemaking since it representsan approved formula as part of a utility's rate structure used to true-up an estimated rate. (See North Carolina Utilities Commission,Order Approving Partial Rate Increase and Requiring Conservation Initiative, Docket Nos. G-9, Sub 499, G-21, Sub 461 and G-44,Sub 15, Nov. 3, 2005, at 21.)

87 A pilot has the benefit of allowing for refinements of the mechanism periodically and of terminating a mechanism that exhibitsundesirable results.

88 In theory, the premise for utility involvement in energy conservation is the existence of serious market failures. Specifically, (1)incorrect pricing of natural gas and its delivery exists because of regulation or market-power conditions, (2) natural gas prices fail toreflect environmental damages and other external costs, or (3) the presence of specific information-related failures that may adverselyaffect the consumer demand for energy efficiency. Proponents of utility-funded energy efficiency initiatives argue that some or all ofthese conditions hold.

89 As an alternative to RD or in conjunction with RD, as proposed by some analysts, direct (targeted) incentives can be provided to autility in promoting energy efficiency initiatives, or responsibility for implementing utility-funded energy efficiency initiatives can beturned over to a third party.

90 In Oregon, Northwest Natural committed to promoting energy conservation and agreed to transfer the management of energyefficiency programs to a third party, Energy Trust. In addition, the utility agreed to collect a public purpose charge on gas bills. Inthe Piedmont Natural Gas case, the North Carolina Utilities Commission conditioned the approval of the RD proposal on the utilityaggressively assisting small customers in conserving on natural gas.


91 Most state commissions using an historical test year adjust test year revenues and expenses for "known and measurable" changes thatwill likely occur in the future, post-test year period. Typically, the test year is a 12-month period, with adjustments made to the extentnecessary (or permitted) based on evidence reflective of conditions during the period new rates are to be in effect. For example, proforma adjustments can take into account changes in: (1) customer additions or losses, (2) per customer usage patterns, and (3) weatherconditions or new measurements of weather normalization.

92 As identified by economists, recovering fixed costs in the volumetric charge can: (1) increase revenue and earnings variability, (2)fail to account for cost differences between or within customer classes, (3) convey improper price signals to customers, and (4) lead toinefficient use of a utility's system. On the other hand, the main goal of regulation is not to promote economic efficiency: regulationoriginated and developed before the ideas of economic efficiency and the principles of welfare economics; most enabling legislationmandates just, reasonable and fair rates, not efficient rates per se.

In addition, redesigning rates to lower the volumetric charge comes across as anti-conservation since it would lower the marginalprice of gas. Taking an example, assume that the volumetric distribution charge is 15 percent of the total volumetric charge (inclusiveof purchased gas costs), which corresponds to the actual percentage for many gas utilities. By shifting all the distribution costs toa customer charge, the volumetric charge would decline by 15 percent; this means that the marginal price of gas would also fall by15 percent. Assuming a short-run price elasticity of demand equal to -0.10, the effect would be to increase gas consumption by 1.5percent. This is about 50 percent above the average annual decline in gas consumption per residential customer over the past 20-25years (see the earlier discussion on the AGA study).

93 Specifically, declining consumption would result in a utility increasing the base rate further above short-run marginal cost, whichequals close to zero or is minimal. Besides, the pre-adjusted base rate was already above marginal cost, thereby causing RD to widenthe price-marginal cost gap.

94 By adjusting revenues for only a price elasticity effect, the utility still faces the risks associated with other factors of gas consumption,such as weather and general economic conditions.

95 While it is true that in the short term the level of sales has minimal effect on non-gas costs, these costs can easily deviate from thecosts underlying the current base rates. For example, the utility could improve its productivity or in other ways reduce its costs, priorto the next rate case, below the test-year level.

96 As an illustration, if the sharing arrangement is 50-50 (with no "dead band") and assuming a decline in sales of one percentattributable to the response of consumers to higher prices, a resulting reduction in the rate of return on equity of a 100 basis pointswould be split evenly between consumers and utility shareholders. Thus, the utility absorbs a loss of 50 basis points. Earnings-sharingmechanisms, labeled alternatively as "revenue (or return) stabilization mechanisms," are currently in place for a few gas utilities in theUnited States, including utilities located in Alabama, Louisiana, Mississippi, and South Carolina.


The National Regulatory Research Institute1080 Carmack Road

Columbus, Ohio 43210-1002Phone: 614-292-9404

Fax: 614-292-7196www.nrri.ohio-state.edu

This report was prepared by the National Regulatory Research Institute (NRRl) with funding provided by the member commissions ofthe National Association of Regulatory Utility Commissioners (NARUC). The views and opinions of the authors do not necessarilyexpress or reflect the views, opinions, or policies of the NRRI, NARUC, or NARUC member commissions.


Documents

Briefing Paper - NARUC