5. Discussion
Under our policy rules for post-2005 energy efficiency, the utilities and program implementers are required to perform cost-effectiveness analyses consistent with the indicators and methodologies included in the SPM, unless we direct otherwise.21 In their Petition, the utilities claim that the SPM and Commission decisions are not clear on the economic principle for treating certain costs as transfers, but not others. They suggest that any transfer of funds made by the program administrator to a market participant (whether it is the customer installing the measure, a manufacturer, distributor or energy service provider under contract) should be treated as a transfer payment in the calculation of TRC cost-effectiveness by subtracting those payments from the cost side of the equation.
As we discussed in D.06-06-063 (and summarize above), all versions of the SPM dating back to the 1980s have been very clear on what costs are to be treated as transfers:
"The only costs that are excluded in the TRC test are those `incentives' that are to be considered and treated as transfer payments. The SPM specifically directs that such incentives are restricted to include "only dollar benefits such as rebates or rate incentives (monthly bill credits)." [footnote omitted.] The conceptual basis for ignoring transfer payments in the development of the TRC is similar to the basis for ignoring tax credits in the societal version of the test. That is, when some taxpayers receive cash transfers (in the form of a tax credit) as a result of higher taxes paid by others, economic theory suggests that those transfers be excluded when calculating the costs and benefits of the investment from the societal perspective."22
While the SPM document does not expand on economic theory in defining the INC term, only the narrow definition of transfer payments contained in the SPM preserves the fundamental objective of the TRC test, i.e., to capture the total costs of the program, including both the participants' and the utility's costs."23 This can be best understood by first examining what would happen if the INC term were not treated as a transfer payment under the TRC test formulation, and then examining what would happen if the INC term were expanded beyond the SPM definition to include the other types of program incentives suggested by the utilities.
If the TRC formulation included the PC term and all utility costs without excluding INC payments, then those dollar rebates or bill credits would be double-counted on the cost side of the equation: They would be included in the PC term (because that term is defined to represent the actual measure installation costs to the participant before any dollar rebates or bill credits) as well as in the PRC term. To avoid this double-counting, it is necessary to treat the INC payment as a "transfer" by excluding the dollar rebates or bill credits from PRC term. In this way, all the utility's costs and all the participants' costs (together comprising the total costs of the program) are reflected in the calculation, without double counting.
Now we examine a situation where dollar payments to upstream/mid-stream market actors (e.g., manufacturers or wholesalers) are made with utility program funds. The utilities recommend that these costs be treated as transfer payments by subtracting them from the TRC cost side of the equation, i.e., that the PRC term be defined to exclude any of these program costs. Under the utilities' proposed formulation of TRC costs, an upstream incentive payment to an equipment manufacturer or distributor to, for example, buy down a compact fluorescent lamp (CFL) price, would be omitted from TRC costs and treated the same as a dollar rebate to a participant who supplies a receipt to the program. This treatment would apply even without evidence that the entire $1 paid to the manufacturer resulted in the participant cost being $1 lower (or that the CFL that was bought down ended up in the hands of a customer within the utilities' service area or even within California).
The utilities' proposed definition of transfer payments would effectively ignore utility program costs in the calculation of TRC cost-effectiveness, contrary to the stated purpose of this SPM test. By definition, the PC term does not include the costs of mid-stream or up-stream incentives. Therefore, eliminating these costs from the PRC term does not address any double counting issues-rather it removes these costs from the TRC cost side of the equation entirely. Hence, treating incentives to upstream/mid-stream market actors as transfers would serve to make these programs appear more cost-effective from a TRC perspective than they really are--by understating the total costs to the utility.
In their Joint Petition, the utilities argue that midstream and upstream incentives "provide a cost-effective means to reducing the cost of energy efficiency measures without the additional administrative cost of providing rebates directly to the consumers." They assert that defining transfer payments narrowly to include only dollar rebates or bill credits to participating customers will motivate them to "return to direct utility rebate programs, which are less cost effective...because of higher administrative costs and lower market acceptance."24 Similarly, the utilities assert that the SPM definition of transfer payments penalizes the utility with a lower TRC if it increases efficiencies through direct-install programs.25
There is no support for these assertions given the SPM formulation of the TRC test of cost-effectiveness. As discussed above, the treatment of dollar rebates to participants as a transfer payment does not favor direct rebate programs by ignoring real program costs-it serves to ensure that such costs are not double counted in the context of the SPM definition of participant costs. The TRC test will fully capture any cost-effectiveness advantages of upstream/midstream programs under the current formulation of that test. In particular, if the costs to administer midstream and upstream incentive programs are less than the administrative costs associated with programs that provide rebates directly to the participant, then these cost savings will be fully reflected in the PRC term. If market acceptance of energy efficiency measures is higher under midstream/upstream program strategies than under rebate programs, then the savings associated with this higher level of participation will also be fully reflected in the TRC equation-on the benefits side.
Finally, if the midstream and upstream incentives are successful in reducing the retail price of energy efficiency measures to the end-users, then this benefit will also be reflected in the PC term. As DRA points out in its comments, the TRC formulation defines the participant cost (PC term) as the actual price paid by participating customers for the energy efficiency measures, and thereby includes the effect from the midstream or upstream incentive that may have acted to reduce the retail price including the possibility that the actual price may be reduced by an amount larger than the midstream or upstream incentive.
Therefore, capturing the impact of mid-stream or upstream programs on retail prices does not require any revisions to the SPM definitions or methodology, or to our determinations in D.06-06-063. Rather, it requires that the actual prices paid by these participants are reflected in the PC term, so that we can evaluate the program activity from a cost perspective and calculate a reliable net benefit for the program. Treating midstream and upstream incentives as transfers (as the utilities prefer) implicitly assumes that there is a dollar reduction in the shelf price for every dollar of cash incentive provided to upstream/midstream market actors. However, it is unreasonable to define the TRC test such that is automatically assumes a one-for-one dollar reduction in the shelf price as a result of these incentives without any scrutiny of this assumption in the context of what we can readily observe in the market.
As part of its ongoing evaluation, measurement and verification (EM&V) activities, Energy Division intends to analyze participant costs for certain measure and/or program delivery strategies where actual participant costs are suspected to be significantly different from estimated participant costs. Pursuant to the ALJ ruling on EM&V Protocols in R.01-08-018 dated September 2, 2005, the utilities are also required to track and report actual measure cost data, based on site specific installations, for all customized measures. Moreover, the utilities are authorized EM&V funding for market assessment studies, where they collect market data. Hence, there is no reason why the planning and evaluation of these program strategies from a TRC perspective cannot reflect market observations with respect to the retail shelf prices that result from upstream/midstream incentives.
Similarly, capturing the impact of program efficiencies associated with direct-install programs requires no change to the definition of transfer payments in the SPM. The utilities' assertion that restricting the definition of transfer payments to dollar rebates (or bill credits) to the participant will skew the TRC results in favor of rebate programs is simply incorrect, except in the very limited situation where the dollar rebate actually exceeds the participant's cost of installing the measure. We demonstrated this clearly with numerical examples in both D.06-06-063 and D.07-09-043. (See Attachments 2 and 3.)
Moreover, as we also discussed in D.06-06-063 and in our policy rules, we use the "dual test" of cost-effectiveness to ensure that utilities use program funds cost-efficiently and avoid excessive rebate levels. In particular, we require that the Program Administrator Cost (PAC) test is used in conjunction with the TRC test in evaluating program design options and that the portfolio as a whole must pass both tests to be eligible for funding.26
To the extent that direct-install programs increase cost efficiencies and program participation, then such benefits and efficiencies will be reflected in the TRC test using the current definitions of TRC cost components. The benefits of increased participation will be reflected in the benefits side of the equation, while the cost efficiencies will be reflected in lower total utility and participant costs. Contrary to the utilities' assertions, there is no "double counting" created when all direct-install program costs are included in the TRC equation. As discussed above, the PC term is defined in the SPM as including only those costs that participating customers incur (prior to any dollar rebates or bill credits) for the energy efficiency measure installations. Therefore, it is not necessary to redefine the PC term to recognize that participant costs under a direct- install program may be significantly lower than under traditional rebate programs.
As DRA points out in its comments, if the direct-install program covers all of the measure and installation costs (and the participant does not have to pay anything for measure installation, including sales tax), then by definition, the PC term will be zero. However, if customer co-payments are required, the PC term will be positive to reflect the participants' share of costs for the installations. The level of the PC term will depend upon the specifics of the direct-install program (e.g, the measures involved, the incremental measure cost of those measures and the amount of participant co-payment). Therefore, it is not reasonable to re-define this term as having a "zero" value for all direct-install strategies, as the utilities suggest in their proposed language revisions to D.06-06-063. Instead, the PC term for direct-install programs should continue to reflect the participant's actual costs for the energy efficiency measures installed under the program.
The utilities also suggest that application of the SPM definitions of PC, PRC and INC terms results in more favorable TRC results if the utility pays the customer (and the customer "pays the vendor") under a direct rebate program, as compared to a direct-install program where the utility "pays the vendor" to install the measure at the customer's premises.27 However, the numerical examples presented in both D.06-06-063 (without accounting for free riders) and in D.07-09-043 (with free riders accounted for on the benefit and cost side) clearly demonstrate that this would not be the case, if all else is equal. Under either program strategy, all of the utility costs and participant costs would be properly accounted for, and the TRC results would be the same, as long as the financial incentive to the customer (either the cash rebate or the difference between the direct-install cost and what the vendor bills the customer) are equivalent and do not exceed the cost of the measures. This can be seen in the numerical examples in Attachment 2 and 3.
Therefore, we find no merit to the utilities' contention that our interpretation of the TRC calculation "will bias program choice against all programs which are not traditional utility operated customer rebate programs" and "work against all the effort over the last several years to seek the most cost effective and efficient delivery approaches including a broader base of program delivery."28 To the contrary, our directions to the utilities will ensure that the most cost-effective and efficient approaches are pursued by accounting for all of the costs to participants and the utility, and by not allowing real resource costs to go unaccounted for by inappropriately classifying them as transfer payments.
The Joint Petition also reflects the utilities' position in Phase 1 of R.06-04-010 with respect to the treatment of free riders on the TRC cost side. If we do not treat all direct-install costs and incentives to midstream/upstream market actors as transfer payments, the utilities request that we adjust all of these costs by the NTG ratio. In D.07-09-043, we rejected this position because adjusting these program costs downwards by the NTG ratio would remove real program costs (revenue requirements) from consideration and overstate TRC net benefits. Our discussion of this issue and numerical examples is presented in Attachment 3.
In support of their Joint Petition, the utilities contend that the TRC formulations used in other jurisdictions treat upstream, midstream and direct-install program costs as transfer payments, as they recommend in their Joint Petition. As an example, they refer to the most recent "Total Resource Cost Guide" adopted by the Ontario Energy Board (OEB Guide) for the local distribution companies (LDCs) in that region. However, our examination of this document leads us to conclude that the OEB Guide defines and treats transfer payments (as well as free riders) in TRC cost calculations consistent with our direction in D.06-06-063 and D.07-09-043.
In particular, the definition of "third party rebates" in the OEB Guide is explicitly limited to "any dollar discount or rebates offered to the customer" and as such, they are the only costs that are treated as transfers under the TRC test for any programs including upstream/midstream or direct-install programs.29 As with the SPM, excluding the dollar discount or rebates offered to the participating customer (irrespective of who offers them) as a TRC cost in OEB Guide avoids double counting these dollar payments in the TRC cost equation. Again, this is because of the manner in which participant costs are defined, namely, as "customer equipment costs" in the OEB Guide, irrespective of who purchases the equipment (the utility or the participating customer).30
In sum, just like the SPM, the OEB Guide defines the costs included in the TRC in a manner that removes potential double counting of the dollar rebates or discounts offered to the participating customer-but does not fail to include them altogether. As we discuss above, this approach is fully consistent with the SPM definition of the TRC test, which the OEB Guide adopts as its own:
"The TRC test is defined as a test that "measures the net costs of a demand-side management program as a resource option based on the total costs of the program, including both the participant's and the LDC's costs."31
The OEB Guide also addresses direct-install programs. These programs are instructed to treat all measure installations performed by the LDCs as program costs exactly as directed in D.06-06-063 and the Compliance Ruling. In particular, the OEB Guide identifies major categories of "conservation and demand management" (CDM) program costs that "can be expected for programs that electric LDCs in Ontario might be considering, which include: (1) development and startup, (2) promotion, (3) equipment and installation, (4) monitoring and evaluation and (5) administration". The OEB Guide then goes on to define direct-install measure cost as a program cost:
"LDC equipment and installation costs include the costs of any LDC devices needed to operate the programs such as specialized software or tools as well as any CDM measures directly installed by the LDC...."32
Finally, the OEB Guide directs that "all program costs associated with free riders must be included in the analysis," observing that "programs that have high free ridership are self-evident in the marketplace (i.e., they do not rely on a LDC promotion) and therefore are less cost-effective for the LDC to pursue since the program costs are included in the TRC calculation while the benefits are not."33
Therefore, contrary to the utilities assertions, we find that the OEB Guide is fully consistent with our definition and treatment of transfer payments for various program delivery strategies, as well as our treatment of free riders on the TRC cost side of the equation. This is particularly apparent from the numerical examples provided in the OEB Guide, which include the equipment costs (adjusted for free riders) and the LDC program costs (not adjusted for free riders) on the cost side of the equation.34
The utilities also refer to a 1994 paper published by the American Council for an Energy-Efficient Economy (ACEEE) in their discussion of "other interpretations" of the SPM, implying that this paper endorses their claim that utility program payments to parties other than the participants can be treated as transfer payments.35 However, the description and illustration of the "TRC analysis envelope" presented in the ACEEE paper leads one to exactly the opposite conclusion. An illustration of the TRC test presented in that paper is reproduced in Figure 1. Arrows pointing out of the TRC envelope represent costs under the TRC test. One can clearly see that all payments not directly paid to the participants cross out of the TRC envelope, and are therefore treated as TRC costs.
In sum, our efforts to ensure that all utility costs and participant costs are properly accounted for in calculating the TRC test is fully consistent with the "other interpretations" the utilities refer to in the Joint Petition. It is also consistent with the interpretation of the Northwest Power Planning Council (NWPPC), which oversees the energy efficiency activities of utilities that receive power from federal suppliers in the Pacific Northwest under the Northwest Power Act. The NWPPC also uses the TRC test to evaluate the cost-effectiveness of energy efficiency measures and, in doing so, requires that all costs be included in the cost-effectiveness calculation:36
"The Council has interpreted the Act's provisions to mean that in order for a conservation measure to be cost-effective the discounted present value of all the measure's benefits should be compared to the discounted present value of all its costs, regardless of who pays the costs. This interpretation was adopted in the Council's 1983 Plan and has not been modified. The reason for this interpretation is that we cannot know before hand, how much of the cost of a measure will be paid by the utility system and how much by the customer, or others. So, we look at all the reasonably quantifiable costs and benefits."37
For all of the reasons discussed above, we do not modify D.06-06-063 or suspend the Compliance Ruling, as requested in the Joint Petition. Our direction in D.06-06-063, the Compliance Ruling and D.07-09-043 ensures that the total costs associated with utility efficiency programs are captured in the TRC calculation of cost-effectiveness, as intended by the SPM.
These documents, in conjunction with the current SPM, provide clear direction on how to apply the TRC test to energy efficiency program activities. In coordination with Energy Division and its E3 calculator consultants, the utilities should jointly hold workshops to respond to the types of practical implementation questions raised by Ecology Action and to educate all program implementers on the proper classification and reporting of program costs consistent with our determinations. Once scheduled, the utilities should notify the service lists in our energy efficiency rulemaking (R.06-04-010) and the utilities' peer review group members of the workshop dates and locations.
Over the longer term, we believe that all stakeholders would benefit from the development of a fully consolidated, explanatory version of the TRC test with numerical examples for various program delivery strategies. However, we do not agree with DRA's suggestion to provide this level of explanatory detail in the energy efficiency policy rules. Instead, we direct Energy Division to update the 2001 SPM so that this document reflects the direction provided in D.06-06-063, the Compliance Ruling, D.07-09-043 and today's decision, with numerical examples for various program delivery strategies.38
For this purpose, Energy Division may utilize authorized 2006-2008 EM&V funding to contract with technical expertise in the development of the SPM update. In preparing this update, Energy Division should consider revisions to the PRC and PC cost terms that will facilitate a more consistent application of free rider adjustments throughout the SPM. For example, defining participant costs as the "customer's contribution" to the measure installation costs, i.e., "net" of the INC term (as well as net of free riders)-and fully reflecting the INC term dollar incentives in the PRC term could achieve this objective without any double counting or omission of TRC costs. Prior to posting the final SPM revisions to the Commission website, Energy Division should solicit written comments from parties to our energy efficiency rulemaking (R.06-04-010, or its successor proceeding) on its draft revisions, and may also hold workshops on those revisions as it deems necessary.
Recognizing that resources are limited, and that there are other priorities to energy efficiency and related resource procurement proceedings in the coming months, we encourage Energy Division to initiate the SPM update as soon as possible, but do not establish a due date for the final SPM revisions in today's decision. Instead, we delegate to the Assigned Commissioner or ALJ in R.06-04-010 the responsibility of establishing a schedule for this task by subsequent ruling, after further consultation with Energy Division.
21 Our energy policy rules for post-2005 energy efficiency were adopted on April 25, 2005 by D.05-04-051. They are presented in Attachment 3 of that decision. See Rule IV (Cost-Effectiveness).
22 D.06-06-063, p. 66.
23 2001 SPM, p. 18, TRC chapter opening statement, emphasis added. We have also reiterated that the TRC test is intended to capture all of the program participant and utility revenue requirement costs in Commission decisions dating back to the early 1990s: "Total resource costs represent the total cost of obtaining the [demand-side management] program as a utility resource, and include both the program participants' out-of-pocket costs (i.e., customer contributions) and the utility revenue requirement costs (e.g., rebates, administrative expenses)." (Excerpt from D.92-09-080 in Attachment 4 to D.05-05-051.)
24 Joint Petition, Attachment, pp.1-2 of March 16, 2007 Letter to ALJ Gottstein.
25 Ibid., Attachment, p. 11.
26 See D.05-04-051, Attachment 3, Rule IV. Under the PAC test, the program benefits are the same as the TRC test, but costs are defined differently to include the costs incurred by the program administrator (including financial incentives or rebates paid to participants), but not the costs incurred by the participating customer.
27 See footnote 11, p. 11 of the Attachment to Joint Petition: "If the [program administrator] pays the vendor, the payment is a cost. If the [program administrator] pays the customer who pays the vendor, it is a rebate excludable from the TRC producing higher TRC results."
28 Ibid., p. 6.
29 OEB TRC Guide updates 2 October 2006, Appendix A, p. iii. See http://www.oeb.gov.on.ca/documents/cases/RP-2004-0203/cdm_trcguide_021006.pdf
30 Ibid, p. 9.
31 Ibid., p. 3. Emphasis is in original, with a footnote reference to the 2001 SPM.
32 Ibid, p. 13, Section 1.3.2, emphasis added.
33 Ibid., p. 15.
34 Ibid., pp. 21-24. A distinction between the SPM and the OEB Guide definitions is that the latter moves away from specifying who pays for the measures in the cost terms by re-naming "participant costs" as "measure costs" in the TRC equation. In either case, however, the TRC equation does not treat direct-install costs to the utility as "transfers" by eliminating them from the calculations.
35 Joint Petition, Attachment footnote 15, p. 14: "Another attempt to explicitly extend the SPM two agent (utility-participant) framework was published in 1994. See Fulmer, Mark and Bruce Biewald, "Misconceptions, Mistakes and Misnomers in DSM Cost Effectiveness Analysis" Proceedings of the ACEEE 1994 Summer Study on Energy Efficiency in Buildings, ACEEE, 1994, pp. 7.73-7.83 especially the sections on TRC starting on p. 7.78."
36 The NWPPC's focus of evaluation is at the measure level, so they do not address free rider issues.
37 From "Council Methodology" presentation by Tom Eckman and Charlie Grist (NWPPC), at the I-937 Workshop, February 23, 2007. [Emphasis added.] This statement can also be found in "Regional Policies on Cost-effectiveness of Utility Conservation," May 2005, Attachment B, Summary of Regional Policies on Cost-effectiveness of Utility Conservation. http://www.nwcouncil.org/news/2005_05/pw_ceissues.pdf.
38 As provided for under our policy rules (Rule XI), Energy Division may revise as needed the reference documents, such as the SPM, using formal or informal procedural vehicles.
