5. Distribution of GHG Emission Allowances in a Cap-and-Trade Program

If ARB determines that there will be a cap-and-trade program in California, ARB must determine how to distribute allowances to emit GHG. A GHG "allowance" is an authorization to emit a specified amount, generally one ton of CO2e of GHG emissions. At the end of a compliance period, entities with compliance obligations would be required to surrender the number of allowances equal to the amount of GHG they emitted, or meet their obligations through offsets or other flexible compliance mechanisms to the extent they are permitted. Any shortfall would subject the entity to penalties and/or other enforcement actions. Cap-and-trade market design and flexible compliance options are discussed in Section 7.

Because allowances could be traded in the cap-and-trade program, allowances would have financial value, even if distributed for free. The value would be determined by the supply of allowances, the demand to emit GHG, and the availability and cost of flexible compliance mechanisms. Because of this value, the method of allowance distribution could have a large impact on the costs to individual deliverers, retail providers, and ultimately electricity customers.

In D.08-03-018, we considered the issue of allowance distribution within the electricity sector in a multi-sector cap-and-trade program with deliverers as the point of regulation. In that decision, we recommended to ARB that "some portion of the GHG emission allowances available to the electricity sector be auctioned."³⁵ We stated further that:

An integral part of this auction recommendation is that the majority of the proceeds from the auctioning of allowances for the electricity sector should be used in ways that benefit electricity consumers in California, such as to augment investments in energy efficiency and renewable energy or to provide customer bill relief.³⁶

We determined at that time that additional record development was needed in order to allow us to make more complete recommendations on allowance distribution issues. Building on our recommendations in D.08-03-018, and with the benefit of the extensive record developed subsequent to that decision, we address in this section the following aspects of allowance allocation policy for the electricity sector in a multi-sector cap-and-trade system:

· The proper mix between auctions and administrative allocations of emission allowances to deliverers, including transitioning between the two approaches;

· Whether allowances to be auctioned should be distributed to retail providers, which would then sell their distributed allowances through the auction;

· The manner in which auction proceeds should be used for the benefit of electricity customers; and

· The manner in which administrative allocations should be made to individual deliverers and retail providers.

In Section 6, we consider allocation of allowances to CHP facilities.

While it is critically important to design auctions in a way that prevents collusion and abuse of market power, we do not make detailed recommendations to ARB regarding auction design at this time. We expect that, if ARB includes auctions in its scoping plan, detailed auction design will occur during a subsequent rulemaking process. We expect to make further recommendations to ARB regarding auction design and other remaining allocation issues as part of that process.

We recommend that the allocation process occur in steps for the electricity sector. First, ARB would determine the total number of allowances to create for each year (or other appropriate time period) for all of the sectors included in the cap-and-trade program, with the number declining over time to meet the multi-sector GHG emission reduction goals. ARB would then determine the number of allowances (or the amount of auction revenue rights if there is a multi-sector auction with the distribution of auction revenue rights) to allocate to the electricity sector. Then, the electricity sectoral allocation would be divided through a second allocation process among the relevant entities within the electricity sector. In this section, we address the allocation of allowances or auction revenues within the electricity sector. In Section 4.2 above, we address the broader determination of the amount of allowances, or auction revenue rights, to be allocated to the electricity sector.³⁷

5.1. Evaluation Criteria, Principles, and Goals

While determining in D.08-03-018 that further record development was needed to make complete recommendations to ARB regarding allowance allocation, we provided some broad direction for the more detailed recommendations on allocation policy that we make today:

In addressing allocation issues, we keep in mind that some deliverers of electricity to the California grid are also retail providers of electricity for consumers. We also recognize that allocation policy will have an impact on consumer costs. Our intent in developing additional allocation policy recommendations is to ensure that GHG emissions reductions are accomplished equitably and effectively, at the lowest cost to consumers. While we may wish to reward early actions to reduce GHG emissions in advance of 2012 when the AB 32 compliance period begins, it is not our intent to treat any market participants unfairly based on their past investments or decisions made prior to the passage of AB 32.³⁸

A staff paper on allowance allocation discussed criteria to use in evaluating allocation options based on the goals discussed in D.08-03-018. Additionally, parties were asked to comment on appropriate evaluation criteria. Based on the discussions in the staff paper and parties' comments, we believe that the following criteria and goals provide useful guidance as we evaluate the various possible allocation approaches:

· Minimize costs to consumers.

· Treat all market participants equitably and fairly.

· Support a well-functioning cap-and-trade market.

· Align incentives with the emission reduction goals of AB 32.

· Administrative simplicity.

We address each of these criteria in turn.

5.1.1. Minimize Costs to Consumers

This criterion is grounded in AB 32 (Section 38652(b)(1) and Section 38652(b)(2)³⁹) and is a key goal guiding AB 32 implementation. Several parties that propose evaluation criteria, including NRDC/UCS and PG&E, include consumer cost in their criteria. NRDC/UCS include a broad category ("Benefit consumers") that contains four subcriteria: avoid windfall profits, minimize costs/maximize benefit for consumers, benefit disadvantaged communities, and improve technology investment. The first criterion we identify focuses on the first three of these subcriteria. Morgan Stanley suggests a broad category ("[develop] a system that is of the least cost to California") that is similar.

We identify three key goals in the quest to minimize costs to consumers, which we address in turn:

Minimize increases in average retail rates and bills statewide. While the next goal considers distributional impacts, this goal seeks to allocate allowances in a manner that reduces average costs to electricity customers statewide. This goal focuses on the overall cost of the emissions reductions realized via the cap-and-trade program and on how those costs are distributed between consumers and producers of electricity.

Minimize wealth transfers among customers of different retail providers. This goal focuses on the differential impacts on retail providers of the various allocation approaches and promotes equity among electricity customers throughout California. The staff paper included a similar criterion ("Equity Among Customers of Retail Providers"), which several parties support in their comments. As we describe below, California's retail providers currently have widely differing average emissions levels. Additionally, the retail providers have varying levels of exposure to the wholesale electricity market. This goal recognizes the importance, to the extent that these characteristics are due to decisions made before AB 32, of not devising an allocation methodology that would create large transfers of wealth between customers of different retail providers.

California's generation mix differs substantially from much of the rest of the United States. Coal is the dominant source of electricity for most of the United States, while less than 10% of California's electricity is produced by coal. As a result, natural gas generation generally is the price-setting generation in California, rather than coal. Additionally, California has a larger percent of non-emitting sources than found in other parts of the United States. Over one-quarter of California's electricity is produced by non-emitting generation.

Within California, retail providers have a range of generation profiles. The majority of California's customers are served by large utilities: three investor-owned utilities (PG&E, SCE, and SDG&E/SoCalGas) and two publicly-owned utilities (LADWP and SMUD). Table 5-1 below lists the generation characteristics of retail providers in California. PG&E has the lowest average emissions rate among California's large retail providers, primarily due to its high levels of non-emitting sources. Of the five largest providers, LADWP has the highest average emissions rate due to the large amounts of coal in its generation mix. Some of the smaller publicly-owned utilities have larger percentages of coal in their generation mix. Anaheim Public Utilities, for example, serves 78% of its load with coal-generated electricity, according to the Energy Commission's 2007 Integrated Energy Policy Report.

Table 5-1

Load and Sales Data for California's Retail Providers

(Based on E3 2008 Modeling Data)

* These categories include generation by resource type that is utility-owned or under long-term contract. The Non-emitting Sources category includes generation from nuclear, large hydropower, and renewable sources.

** SDG&E Comments, June 2, 2008.

Unless great care is taken, carbon regulations inadvertently could have disparate customer impacts due to the different generation mixes. Customers of retail providers with small amounts of coal generation or large amounts of non-emitting generation in their electricity portfolio would tend to see lower price impacts due to compliance obligations under carbon regulations since the emissions levels of power serving them are lower. On the other hand, retail providers with larger amounts of coal generation or smaller amounts of non-emitting generation in their portfolio would tend to have higher rate impacts because their generation sources have higher carbon regulation compliance costs. An additional consideration is that retail providers have differing practices regarding the extent to which they own generating sources and their degree of reliance on market purchases. Customers of retail providers that obtain much of their electricity from the wholesale market would be affected by increases in wholesale prices more than would customers of retail providers that own or have long-term contracts with most of the generating assets used to serve their load. A significant focus of inquiry in this proceeding has addressed ways in which allowance allocation policies could help moderate these potential price impacts.

One important measure of potential impacts of GHG regulations on customers is the effect on the average rate levels of the various retail providers. Table 5-1 above shows current average retail rates and emission rates for retail providers in California. These rates differ significantly among the retail providers. PG&E's average retail rate is $0.14 per kWh, slightly above the average rate in California, while PG&E has the lowest average emissions rate. LADWP has the lowest retail rates among the large retail providers, with average retail rates of only $0.101 per kWh. However, LADWP has the highest average emissions rate among California's large retail providers.

One of the challenges of this proceeding is the development of allowance allocation policies that treat retail providers with such widely disparate emissions, procurement policies, and rate profiles equitably and fairly.

Avoid undue windfall profits for independent deliverers. This goal focuses on the potential for different allocation approaches to redistribute wealth from electricity consumers to independent generators and other deliverers. For the purposes of this decision, we define windfall profits as any increase in profits to deliverers that results from the establishment of an emissions cap-and-trade program and the manner in which allowances are distributed.

PG&E and several other parties support this goal. The staff paper describes how the allocation methodologies could provide differing amounts of windfall profits, which would lead to increased costs for consumers. In evaluating potential allocation methodologies, we pay close attention to the potential for windfall profits and the resulting effects on consumer costs.

Most of the allocation approaches that we have considered would increase wholesale electricity prices by an amount up to the allowance cost of the marginal generator, where allowance cost equals the market value of allowances times the number of allowances that must be surrendered for each unit of electricity from that resource. Using terminology suggested by the Market Surveillance Committee of the CAISO,⁴⁰ we distinguish two ways in which independent deliverers may obtain windfall profits due to a cap-and-trade system:

· "Allowance rents" are windfall profits obtained due to the free distribution of allowances. All deliverers that sell into the wholesale market would realize increased revenues as a result of higher wholesale electricity prices, while consumer costs would increase to the extent that individual retail providers rely on wholesale electricity purchases. Allowance rents would be a direct transfer from consumers to deliverers, with the increase in the deliverers' "producer surplus" matched by a corresponding loss in consumer surplus.

· "Clean generation rents" reflect the increase in producer surplus, and thus windfall profits, that occurs for generation with emission rates lower than the emission rate of the marginal unit that sets the wholesale market price. If the wholesale market price increases due to cap-and-trade by more than the compliance cost of other generators selling into the market, they realize clean generation rents. Conversely, if the wholesale market price increases by less than the compliance cost of other generators selling into the market, their clean generation rents would be negative.

Figure 5-1 presents a stylized example that illustrates these two types of rents for several types of independent generators selling into the wholesale electricity market.⁴¹. In this example, gas-fired combustion turbines are the marginal source of generation and set the market clearing price P₀ before the cap-and-trade system is implemented. Once cap-and-trade is in effect, the wholesale market clearing price rises to P', reflecting the allowance cost of the gas-fired combustion turbines, which remain the marginal resource.

Figure 5-1

Stylized Example of Effects of GHG Compliance Costs on Producer Surplus

As illustrated in the example in Figure 5-1, allowance costs per MWh are lower for more efficient combined cycle gas-fired plants, higher for more carbon-intensive coal-fired generation, and zero for carbon-free hydropower and nuclear facilities. If generators receive all of the allowances they need for free, they will realize allowance rents equal to (P'- P₀) on each MWh they sell into the market. These rents represent an increase in the producer surplus that was already being received by inframarginal generators. Clean generation rents would accrue to some producers even with 100% auctioning. With 100% auctioning, emitting generators would actually incur the allowance costs shown in Figure 5-1, and the producer surplus each realizes would increase or decrease depending on whether it is less or more carbon-intensive than the marginal resource. In this example, the hydroelectric, nuclear and CCGT units all receive clean generation rents because the wholesale electricity price increase exceeds their allowance cost. The reverse is true for coal-fired generators, so their producer surplus declines. There is no change in producer surplus for the gas-fired combustion turbines on the margin. The wholesale energy price increase reduces consumer surplus, but this loss may be partially compensated by distributing the auction revenues in a way that benefits retail electricity customers.

While different parties have used somewhat different terminology, we find the CAISO's terminology to be useful for our purposes. It is generally accepted that only independent deliverers would actually receive either category of windfall profits. For generation owned by or already under long-term contract to retail providers, we assume that regulators and local governments would not allow pass-through of the opportunity costs of free allowances or clean generation rents, so that for such generation only actual compliance costs would be passed on to retail customers.

SCE submits that the profits that the Market Surveillance Committee calls rents to clean generation are unavoidable, and arguably are desirable in that they create incentives to build additional low-emission generating units. It finds allowance rents to be more problematic.

While supporting a relatively quick transition to a full auction in part because of concerns about windfall profits, DRA asserts that the extent of the overall windfall would be limited, for several reasons. First, DRA states that pre-existing procurement contracts are not susceptible to generator windfalls to the extent that the generator is not able to adjust the contract price to reflect increases in wholesale market prices. Second, DRA suggests that new procurement contracts may shift the carbon risk from the generator to the utility.

WPTF asserts that the E3 GHG calculator greatly overestimates potential windfall profits by independent deliverers. First, WPTF takes issue with E3's assumption that all generation currently under contract will be procured from the market upon expiration of the contract. Second, WPTF believes that E3 overestimates the extent to which renewable facilities would sell their power through the wholesale market and thus be positioned to reap windfall profits. Upon review of WPTF's concern, we find that WPTF states incorrectly that the marginal clearing price effect modeled in the E3 calculator is the difference between the effect of allowance costs on wholesale prices and the deliverers' cost of allowances. In fact, the market clearing price effect calculated by the E3 model is the total increase in wholesale prices, which is not reduced by deliverers' compliance costs.

EPUC/CAC assert that windfall profits by independent deliverers would be limited because of qualifying facilities and other power that is sold through long-term contracts. We agree that the administrative determination of prices for qualifying facilities may reduce the potential for windfall profits for such generation. However, it seems unlikely that generators entering into bilateral contracts would forego all of their potential windfall profits in exchange for the certainty of a long-term purchase agreement. We expect that wholesale prices in new contracts will reflect, to some extent, the profits that generators would expect if they chose to sell their power through bidding into the wholesale market.

5.1.2. Treat All Market Participants Equitably and Fairly

This criterion is grounded in Section 38562(b)(1). We recognized this guidance in our statement in D.08-03-018 that, "[I]t is not our intent to treat any market participants unfairly based on their past investments or decisions made prior to the passage of AB 32." (D.08-03-018, p. 18.) We recognize that retail providers and generators have made historical investments in emitting technologies and that allowance allocation methodologies could have significant financial impacts on investors and customers that rely on these technologies. Similarly, potential impacts on retail providers that have developed procurement strategies with greater reliance on wholesale markets should be considered when assessing the desirability of different allowance allocation approaches.

We also recognize the importance of providing appropriate recognition of early actions that entities may take to reduce GHG emissions. SDG&E/SoCalGas and PG&E argue that past energy efficiency and renewable energy investments by retail providers should be reflected in the allocation of allowances or auction revenue rights. While recognizing that early actions will provide an automatic benefit by reducing compliance obligations, we also consider how the various allowance allocation methodologies would recognize early actions.

Another consideration is the extent to which an allocation methodology would provide revenues to deliverers or retail providers to help fund compliance obligations or investments in GHG emission reduction measures, or to reduce customer rate impacts. Reducing GHG emissions consistent with AB 32's goals will require long-term investments in low-emitting technologies. As we discuss in Section 5.5 below, auction revenue intended for the benefit of consumers could be used in many ways, including investments in emission reduction measures and compensation for potential increases in electricity rates. We consider the impact that various allocation options would have on providing entities with revenues that they could use in adjusting to the new GHG reduction requirements.

An important goal is to ensure that the chosen allocation approach does not have inadvertent and unfair competitive impacts. While the need for emission reductions inherently will encourage the development of lower-emitting technologies and business practices, we should take care to avoid unintended consequences that favor certain technologies or entities for reasons other than their effectiveness in helping California achieve the goals of AB 32. Some parties have expressed particular concern that no entity should have preferential access to allowances.

Finally, while we agree that there is value in recognizing the past investment and business planning decisions that entities undertook before the need to reduce GHG emissions was understood fully, equity considerations require that we recognize and encourage entities that take aggressive steps to reduce emissions. While a transition period is reasonable, equity dictates that we move to a market in which "the polluter pays."

5.1.3. Support a Well-functioning Cap-and-Trade Market

We see three aspects of potential allowance allocation approaches as being particularly important to ensure the smooth functioning of the cap-and-trade market. First is the degree to which the distribution methodology leads to accurate price signals, to guide the activities and choices of market participants.

Second, market participants stress the need for some reasonable degree of predictability and certainty in the market. Market certainty would help companies plan future investments, particularly because many GHG-reducing strategies require significant long-term investments. Under a cap-and-trade program, certainty and predictability would be furthered by stable, long-term carbon prices. Additionally, it would be beneficial for entities to have some assurance regarding the level of allowances that will be available in the market and, in particular, the number of allowances that they may expect to receive. This concept is embedded in the "planning predictability" criterion that DRA proposes. We note that planning predictability will hinge on the value of allowances, not just the number available in the market or distributed to individual entities. A cap-and-trade program that would prevent or discourage allowance hoarding or other market manipulation practices would help foster accurate and more stable price signals. Third is the extent to which potential allocation methods might be vulnerable to market manipulation, a concern expressed in several parties' comments.

5.1.4. Align Incentives with the Emission Reduction Goals of AB 32

AB 32 provides guidance to the State agencies in developing GHG regulations to reduce GHG emissions. Of particular relevance in assessing allowance allocation options is the guidance in Section 38560 that regulations should "achieve the maximum technologically feasible and cost-effective greenhouse gas emission reductions." In evaluating allocation options, we consider the extent to which they provide incentives that will further the reduction of GHG emissions in California.

5.1.5. Administrative Simplicity

This criterion is included in the staff's criteria and is supported by several parties, including DRA and NRDC/UCS. In addition to improving the feasibility and ease of implementing the adopted GHG regulations, administrative simplicity would help stakeholders "reasonably predict the consequences of the program." (Staff allocation paper, p. 12.)

5.1.6. Additional Considerations

In addition to the most important criteria and goals listed above, we evaluate each allocation option to assess its desirability if California links to a regional and/or national cap-and-trade program. We recognize that future success in reducing GHG emissions will involve increasing coordination at the regional and national levels. In August 2007, several Western states (including California) and Canadian provinces established the Western Climate Initiative, an agreement to reduce GHG emissions through coordinated cap-and-trade programs. California is a full and supportive participant in the Western Climate Initiative. We also are following closely federal legislation that would establish a federal cap-and-trade program. We do not see that any of the allocation proposals considered would impede linkage with a federal or regional cap-and-trade program. Commission staff are coordinating with other Partner governments in the Western Climate Initiative to ensure that program design recommendations support the goals of the Western Climate Initiative and would contribute to a smooth transition to regional coordination and linkage.

SMUD and other parties (IEP, Dynegy) suggest that grid reliability be included as an allocation criterion, arguing that reliability was not considered adequately in the staff analysis. While grid reliability is of paramount importance, we do not find merit in these parties' arguments that allowance allocation policies could have a detrimental effect on grid reliability. Entities with a compliance obligation would be allowed to acquire allowances through auctions or from other parties. With proper design to curb the potential for market manipulation, the cost of allowances in the secondary market should reflect the supply and demand for allowances. Markets for allowances should provide generators and retail providers with appropriate price signals to guide long-term investments. Flexible compliance options, such as offsets, banking of allowances, and multi-year compliance periods, would help ease potential allowance demand spikes, as well as reduce the impact of abnormal hydropower years or other anomalies that may affect electricity generation or demand.

Some parties suggest accommodation of new entrants as a factor to consider in evaluation of the various allocation proposals. Based on the record, it appears that all allocation proposals could be structured in ways that would allow new entrants to obtain allowances equitably. By their structure, some allowance allocation approaches, in particular auctioning, would treat all deliverers equally, so that new deliverers would be on the same footing as other deliverers regarding their ability to obtain allowances. Other allocation approaches, particularly if used exclusively, may need specific provisions to accommodate the allowance needs of new entrants. For example, an approach in which allowances would be made available to deliverers in proportion to their historical emissions could, at the same time, set aside a number of allowances for new deliverers, so they would not be disadvantaged by such a general historical emissions-based approach. If an allocation approach appears desirable for other reasons, the complexity of devising and maintaining such a set-aside provision would need to be considered in deciding whether the approach should be pursued.

Finally, legal issues that parties have raised regarding allocation alternatives are addressed in Section 5.6. We do not find any convincing legal concerns with the allocation-related recommendations that we make to ARB.

5.2. Description of Allowance Distribution Options

The issue of allowance distribution is fundamentally a question of allocating the value that allowances represent. Allowance values could be distributed either by administratively allocating the actual allowances themselves or by first auctioning allowances and then distributing the resulting revenues, for example, according to a previously established structure of auction revenue rights. One party, GPI, has suggested making some or all of the allowances available for sale to deliverers at a predetermined price.

Allowances could be distributed to the entities with compliance obligations, or to other entities. In the electricity sector, allowances could be distributed to deliverers, which would have the compliance obligations under the deliverer approach that the Commissions have recommended to ARB. Allowances or auction revenues also could be distributed to retail providers on behalf of their ratepayers.

The staff paper on allowance allocation explored the impacts of several methods of allocation, including distribution to deliverers based on their historical emissions (both of in-State generation and imported electricity) during a fixed baseline period, distribution to deliverers based on the amount of electricity they currently or recently delivered to the California grid, and auctioning with allowances or auction revenues distributed to retail providers based on the retail providers' historical emissions, or on sales periodically updated to reflect more recent sales levels. The staff paper also describes various combinations of these approaches, which could be crafted to improve the extent to which various evaluation criteria are met.

We describe next the basic allowance distribution approaches that staff examined and also two other approaches suggested by parties.

5.2.1. Distribution of Allowances to Deliverers

5.2.1.1. Distributions in Proportion to Deliverers' Historical Emissions

One option would distribute allowances to deliverers in proportion to their historical emissions in a fixed prior baseline year or multi-year period. This approach is sometimes referred to as "grandfathering." Basing allocations on periodically updated emissions levels is generally not considered, because such updating would provide incentives for deliverers to increase, rather than reduce, the emissions associated with their electricity. Instead, the fixed proportion of yearly allowances that each deliverer would receive would be determined based on relative emissions during the baseline period. These fixed proportions then would be applied to the total number of allowances allocated to the electricity sector for each year to determine the number of allowances to distribute to individual deliverers. Allowances would continue to be distributed in the same proportion to individual deliverers, but deliverers would receive proportionately declining numbers of allowances each year as the overall number of allowances allocated to the electricity sector declines.

A primary drawback of historical emissions-based allowance distributions to deliverers is that there could be large windfall profits to independent generators and marketers. This approach would allow allowance rents and clean generation rents.

The expectation is that, with an historical emissions-based distribution mechanism, electricity sold through the wholesale market would reflect the full expected opportunity cost of allowances, even though deliverers were given allowances for free. This is because, if they did not operate, they would not incur compliance obligations and could sell their allowances at a profit. Because of the loss of allowance value entailed by the operation of an emitting facility, deliverers would tend to incorporate the opportunity cost of their allowances into their bids just as if the allowances had been purchased. As a result, wholesale prices would reflect the full opportunity cost of the marginal generators setting the wholesale market price. Deliverers of electricity from emitting generation resources (including deliverers from unspecified sources) would realize allowance rents because they would receive the higher wholesale electricity price while avoiding the cost of purchasing some or all of the allowances they need. Independent deliverers that receive free allowances could also reduce deliveries compared to the baseline period and sell the allowances; the resulting profits would also be considered an allowance rent. Carbon-free deliverers selling into the market also would receive the higher wholesale price without needing to purchase allowances. In this case, the resulting increase in profits would represent a clean generation rent.

These windfall profits would occur at the expense primarily of customers whose retail providers are dependent on competitive wholesale markets, which includes the investor-owned utilities and certain publicly-owned utilities. Electric service providers would be disadvantaged, to the extent they rely on the wholesale market. The windfall profits would result in wealth transfers to independent deliverers. A comparable wealth transfer would not occur for utilities that own most of their resources, because their regulatory boards presumably would prevent them from passing on the full opportunity cost of the freely received allowances to their customers.

An advantage of an historical emissions-based distribution approach is that it would avoid wealth transfers from customers of retail providers whose portfolios have higher GHG emission rates to customers of utilities with portfolios with lower GHG emission rates. Because sources that provide power to each utility are unlikely to change radically over a short time frame, the sources of power serving a retail provider's load should not be particularly short or long on allowances, particularly during the early years of an historical emissions-based approach.

Figure 5-2 provides an illustrative example of the potential effects on retail providers' rates of historical emissions-based distributions of allowances to deliverers.⁴² Recognizing that this scenario using the E3 calculator is based on only one set of modeling assumptions, we find this scenario useful because it provides a general indication of the effects that historical emissions-based distributions to deliverers could have on retail electricity rates. A comparison of the results in Figure 5-2 to results for other distribution options presented below indicates that, of the administrative allocation options we consider, historical emissions-based distributions of allowances to deliverers could have the largest impact on retail rates. While distributions on the basis of historical emissions would tend to protect retail providers like LADWP with relatively high-emitting portfolios, the large windfall profits would increase rates significantly for retail providers that are more dependent on the wholesale market.

Figure 5-2

Estimates of Effects on Average Retail Electricity Rates Due to Historical Emissions-Based Distributions of Allowances to Deliverers

($/kWh, 2008$)

To prevent new entrants with emissions from facing a competitive disadvantage relative to existing generators, an allowance set-aside or other steps would be needed to accommodate new entrants.

A shortcoming, compared to auction alternatives, is that this approach would generate no revenues to fund GHG emission reduction efforts by entities other than deliverers, or for customer bill relief. In its favor, the historical emissions-based approach would provide revenues to those deliverers with the largest compliance obligations and potentially with the most opportunity to reduce their emissions.

The extent to which historical emissions-based distributions to deliverers would recognize voluntary early actions that deliverers have taken to reduce emissions depends on the base period used in establishing the level of historical emissions to be used in determining the number of allowances each deliverer would receive. If, for example, the base period used for determining historical emissions were a period immediately prior to the enactment of AB 32, deliverers would be rewarded for any early action they take to reduce emissions after that base period. These deliverers would receive credit for their early action because their allowances would be based on their higher (pre-AB 32 enactment) historical emissions, but they would only need enough allowances to cover a level of emissions that had been reduced by the actions they took after enactment of AB 32. The receipt of the additional allowances would reward the deliverers for their voluntary early actions.

An advantage of historical emissions-based distributions to deliverers is that the number of free allowances that each deliverer would receive would be predictable.

An historical emissions-based distribution of allowances to deliverers would be relatively simple to administer. It would require administrative determinations regarding the baseline year(s). A multi-year average baseline could be used to smooth normal variations in emissions, e.g., due to varying hydro and temperature conditions and due to varying lengths of outages. Additionally, for electricity delivered from outside of California during the baseline period, the sources of generation would need to be identified and appropriate emissions factors applied to unspecified purchases. Because of the significant volume of unspecified purchases from out-of-state sources, this would entail a substantial value. The need to develop some method to set aside or otherwise provide allowances to new entrants would add administrative complexity.

The distribution of allowances in proportion to historical emissions would provide a strong incentive for deliverers to reduce emissions, since the deliverer could sell any unused allowances. A deliverer could reduce its emissions in various ways, including increases in the efficiency of its facilities, switching to lower-emitting sources, or decreasing deliveries. Since allowances would continue to be distributed in perpetuity, high-emitting facilities in particular might have an incentive to shut down in order to free up allowances to sell in the market.

5.2.1.2. Distribution in Proportion to Amount of Electricity Delivered

In this approach, allowances would be distributed to deliverers in proportion to the amount of electricity they deliver to the California grid in a specified period. This approach is often referred to as "output based." The proportions of allowances distributed to individual deliverers would be updated periodically, either annually or perhaps less frequently, to reflect relative changes in production. These updated proportions would be applied to the total number of allowances allocated to the electricity sector for the year in question to determine the number of allowances to distribute to individual deliverers.

In a pure output-based approach, the number of allowances distributed to each deliverer would be proportional to the total amount of electricity it delivers in the specified period, regardless of its emissions levels. As a variation on the output-based approach, allowances could be distributed instead in proportion to the delivery of electricity from generation with emissions. As another variation, staff suggests a fuel-differentiated approach, as explained more fully below.

Table 5-2 provides a simplified illustration of how an output-based allocation mechanism would work, along with the two variations described in the staff paper. This example assumes that the electricity sector consists of four generation sources - coal, natural gas, unspecified, and non-emitting - and that each source delivers 100 GWh to the grid. It also assumes that the total electricity sector carbon allowances equal the total sector's emissions, in tons CO2e.

Table 5-2

Illustration of Output-based Allowance Distribution Methodologies

As Table 5-2 illustrates, in a pure output-based approach, deliverers with non-emitting or relatively low-emitting generation resources would benefit relative to those with higher-emitting resources.⁴³ As a result, a pure output-based approach likely would result in large wealth transfers from customers of coal-dependent retail providers and would advantage customers of retail providers with low emissions in their electricity portfolios.

Staff and certain parties suggest variations to the output-based approach, aimed at moderating this wealth transfer. With an output-based allocation restricted to emitters, deliverers with emissions would receive a larger share of allowances than under a pure output-based allocation. As Table 5-2 illustrates, allowances would be divided among entities that deliver electricity from emitting resources (including unspecified sources) based on their portion of emitting deliveries. Because allowances would be targeted to deliverers with emissions, the wealth transfer from customers of retail providers with high levels of emitting generation would be reduced. However, there still would be wealth transfers from customers of retail providers with disproportionate amounts of coal generation to customers of largely natural gas-dependent retail providers.

With a fuel-differentiated output-based allocation, allowances would be allocated only to deliverers of electricity from emitting resources, using weighting factors based on fuel type. As illustrated in Table 5-2, the use of weighting factors would reduce, and could largely eliminate, wealth transfers from customers of coal-dependent retail providers to customers of natural gas-dependent retail providers. This reduction of wealth transfers would be accomplished by providing emitting deliveries with allocations that more closely reflect their emission levels.

Staff and certain parties argue that output-based distributions of allowances to deliverers may tend to hold down consumer costs compared to historical emissions-based distributions to deliverers, due to what they call a "market clearing price effect."⁴⁴ In an output-based approach, deliverers would have an incentive to maintain or increase sales levels, since the number of allowances they receive would depend on continued generation levels. Because of this incentive to maintain sales and generation, generators may have an incentive to not include the full value of allowances in wholesale bids or in negotiated prices in power purchase agreements. Essentially, there would be no opportunity cost for the allowances because the allocation depends on continued deliveries. If emitting sources reduce generation in order to free up and sell allowances in one period, they would lose allowances in the future period. If wholesale energy bids reflect this theorized incentive, wholesale market prices in an output-based approach would be lower than in an historical emissions-based approach. In theory, wholesale prices would increase only if, and to the extent that, the marginal generator setting the market clearing price does not receive free allowances sufficient to meet its compliance costs. Although this line of reasoning is somewhat persuasive, we note that this allocation approach has never actually been used in practice.

Staff recommends that the output-based approach, if chosen, distribute allowances only to deliveries from GHG-emitting resources, since including all generation would provide free allowances to deliverers that use non-emitting resources including nuclear, hydro, and renewable sources that do not need them. Staff recommends further that allocations be made on a fuel-differentiated basis, with more allowances provided to high emitters. In this fuel-differentiated approach, a weighting factor would allocate more allowances per MWh to deliveries from coal-fired sources. Staff states that this fuel-specific approach should be designed to produce virtually no wealth transfers among retail providers at the start of the program.

The potential effects of output-based distributions to deliverers on average retail rates depend heavily on the extent to which allowance values are reflected in wholesale market prices. The following figures provide illustrative examples of potential average rate impacts of output-based allocation approaches for the different retail providers. Because of current modeling limitations, the fuel-differentiated option has not been modeled in this proceeding. Figure 5-3 and Figure 5-4 below illustrate potential average rate impacts for retail electricity customers under a pure output-based allocation, with Figure 5-3 assuming that the full value of allowances is included in wholesale market prices while Figure 5-4 assumes that 25% of the value of allowances is included in wholesale market prices. As mentioned previously, these figures and all other figures in Section 5 assume 33% renewables, "high" levels of energy efficiency, $30/ton allowance costs, and no offsets.

Figure 5-3

Estimates of Effects on Average Retail Electricity Rates

Due to Pure Output-Based Allocation of Allowances to Deliverers,

With Inclusion of Full Value of Allowances in Wholesale Prices

($/kWh, 2008$)

Figure 5-4

Estimates of Effects on Average Retail Electricity Rates

Due to Pure Output-Based Allocation of Allowances to Deliverers,

With Inclusion of 25% of Allowance Value in Wholesale Prices

($/kWh, 2008$)

Relative to an historical emissions-based allocation (illustrated in Figure 5-2), an output-based allocation to all generation would have smaller rate impacts for retail providers with large percentages of non-emitting generation. PG&E and SCE, both with large shares of non-emitting sources, would experience lower costs with an output-based allocation to deliverers, relative to their costs with an historical emissions-based allocation to deliverers. Retail providers with relatively small amounts of non-emitting generation, such as LADWP, would experience higher rate impacts with an output-based allocation to deliverers relative to an historical emissions-based allocation. These findings apply regardless of the extent to which the value of allowances is reflected in wholesale market prices.

If, as theorized, an output-based approach suppresses the inclusion of allowance values in wholesale prices (illustrated in Figure 5-4), the differences in rate impacts for retail providers with lower-emitting portfolios compared to those with higher-emitting portfolios could be even more pronounced. The scenario illustrated in Figure 5-4, with only 25% of the allowance value reflected in wholesale prices, indicates the possibility that lower-emitting retail providers could see rate decreases in such situations.

Figure 5-5 and Figure 5-6 below illustrate potential average rate impacts for retail providers with an output-based allocation limited to emitting generation deliverers.

Figure 5-5

Estimates of Effects on Average Retail Electricity Rates

Due to Output-Based Allocation of Allowances to Emitting Deliverers,

With Inclusion of Full Value of Allowances in Wholesale Prices

($/kWh, 2008$)

Figure 5-6

Estimates of Effects on Average Retail Electricity Rates

Due to Output-Based Allocation of Allowances to Emitting Deliverers,

With Inclusion of 25% of Allowance Value in Wholesale Prices

($/kWh, 2008$)

While average statewide rate impacts may be about the same for either a pure output-based approach or an output-based approach limited to deliverers of electricity from emitting generation resources, wealth transfers among customers of different retail providers would be moderated somewhat if the output-based allocation is limited to emitting generation deliverers, as can be seen by comparing Figure 5-5 and Figure 5-3.

A pure output-based allocation approach would provide an incentive for increasing generation from low-or non-emitting resources, to the extent that allowances would be received in excess of the number needed for such resources. At the same time, there may be an incentive to decrease production from high-emitting resources such as coal.

Output-based allocations restricted to emitters would not provide an incentive to increase generation from non-emitting sources. Under this approach, it appears that natural gas generators still would receive more allowances than they would need, particularly in the early years, and, thus, would have an incentive to increase production. Coal, on the other hand, would receive fewer allowances than it would need, which could act as an incentive for decreased coal production.

A fuel-differentiated output-based allocation could largely eliminate the incentives to increase generation from natural gas or decrease coal production, if the weighting factors approximate deliverers' emission rates.

A pure output-based allocation methodology would benefit renewable and other low-emitting generators in that they would receive free allowances that they could sell, with resulting windfall profits in the form of allowance rents. However, the variations on the output-based approach that staff considered would provide no allowances to zero-emitting generators. Generators selling into the market would be affected by the theorized characteristic that output-based methodologies might suppress the pass-through of allowance opportunity costs in market clearing prices. To the extent that occurs, clean generation rents would be less than would occur in allocation methodologies that lead to full reflection of allowance opportunity costs in the market clearing price.

An output-based approach with frequent updating would accommodate new entrants. However, to avoid a competitive advantage to existing deliverers, it may be desirable to have a small set-aside of allowances for a new entrant's first year of operation, if allowances were allocated exclusively through output-based distributions to deliverers.

Like the historical emissions-based approach, a shortcoming of an output-based distribution to deliverers is that it would not generate revenues to fund GHG emission reduction efforts by entities other than deliverers, or for customer bill relief.

If allowances were distributed to deliverers on an output basis, deliverers would obtain a benefit from any early action they had taken to increase their generating efficiency. For example, the number of allowances needed for a natural gas generator would decrease if the generator increases its efficiency, while the number of allowances it would receive would not change based on that early action.

Output-based allowance distribution approaches would not provide as much certainty for deliverers as would an historical emissions-based approach. This is because the number of allowances that an individual deliverer would receive would be determined based on its proportional share of deliveries to the grid in the previous period and therefore would depend on the output of all of the allowance-eligible deliverers. Consequently, its allocation in future periods could not be known in advance.

A pure output-based allocation approach would be fairly transparent and easy to administer, because it would provide a simple formula for allocating allowances, based on generation levels during a specified period. An output-based approach limited to emitting sources would be more complex, because the sources of the electricity would need to be identified. A fuel-differentiated approach would require development of appropriate weighting factors for each fuel type, adding some additional administrative complexity.

5.2.1.3. Distribution of Rights to Purchase Allowances at a Fixed Price

GPI asserts that giving emissions allowances away without charge would be equivalent to giving away public assets or resources and would not be in the public interest. GPI maintains that free distributions would provide a form of windfall to the recipient, whether retail sellers or generators, at the expense of electricity consumers. GPI supports the auctioning of a small fraction of allowances initially, transitioning to increased reliance on auctions as the market develops, matures, and stabilizes.

GPI submits that, to the extent that allowances are not auctioned, the proper approach is to administratively allocate to deliverers the right to purchase allowances at a pre-determined, administratively set price. GPI states that the administrative allocation to deliverers of purchasing rights for the GHG emissions allowances can be done using the same methods as have been discussed for the administrative allocation of free allowances to deliverers.

GPI asserts that its proposed approach would prevent windfalls, and would ensure that the value of emissions allowances could be applied to benefit consumers. GPI submits that its approach would provide some amount of price stabilization, at least in the early stages of the program.

GPI asserts that distribution of allowances by sales rather than without charge would provide some important market protections and benefits, including that market participants that purchase allowances rather than receive them for free would be less likely to exhibit manipulative, speculative, or hoarding behavior. It also asserts that this approach would impose greater operating costs on fossil generators, and greatly reduce the risk of windfall profits.

GPI states that the market clearing price for allowances likely would be achieved in the secondary market although the authorities "ought to be able" to set a price that is reasonably close to the market clearing price for allowances.

GPI expects that the administrative allocation of the rights to purchase allowances at a fixed price would be phased out gradually with increased auctioning.

5.2.2. Auctioning with Distributions to Retail Providers

In this approach, auctions of GHG allowances would be conducted by ARB or its agent. Deliverers, which would have the compliance obligation, would buy allowances according to anticipated need through the auction and/or in the secondary market.

With auctioning, deliverers would buy allowances (or utilize offsets or other flexible compliance options to the extent allowed) for all emitting electricity that they deliver, and would need to recover these costs. We expect that, with auctioning, wholesale electricity prices would increase to reflect allowance costs of marginal generation that sets the market clearing price. This would generally flow through to retail rates. Resourced retail providers similarly would be able to pass their allowance costs through to consumers, assuming approval by regulatory or other governing authorities.

The net effect on costs to customers and wealth transfers among customers of different retail providers would depend on how the money raised by the auction is used. If no allowances or auction revenues were distributed to retail providers, we expect that retail rates would increase statewide, with the largest increases for retail providers with generation portfolios with relatively high emission rates. Figure 5-7 illustrates potential rate impacts if allowances are auctioned without retail providers receiving any allowance value.

Figure 5-7

Estimates of Effects on Average Retail Electricity Rates of Auctions

If Retail Providers Receive No Allowances

($/kWh, 2008$)

Because of the significant rate impacts that would occur otherwise, as illustrated in Figure 5-7, we recommended in D.08-03-018 that the majority of revenues from the auctioning of allowances for the electricity sector be used for the benefit of electricity consumers. In one formulation of this approach, ARB would auction the GHG allowances and the State would receive revenues from the auction. In another formulation, ARB would distribute some or all of the allowances to retail providers and/or other entities that ARB determines should receive the value of the allowances. As discussed in Section 5.3 below, we recommend that ARB distribute allowances to retail providers, with a requirement that they then sell the allowances distributed to them through a centralized auction. This requirement would mitigate potential anti-competitive effects due to the distribution of allowances to retail providers.

Auctioning would treat all deliverers, including new entrants, equally.

Auctioning would provide a strong incentive for deliverers to reduce emissions associated with their power. In this regard, auctioning would perform on par with emissions-based allocations to deliverers and somewhat better than output-based allocations, which would provide less incentives for deliverers to shut down high-emitting plants or take other steps to reduce the emissions of the power they deliver.

An auction could be complex to develop and administer. There also would be a need to develop and implement a method for allocating allowances or auction revenue to individual retail providers. Allocating allowances or auction revenues to retail providers on a sales basis would be relatively simple, whereas an historical emissions-based approach would be somewhat more complex.

Because of the potential otherwise for large retail bill impacts, we recommend that ARB distribute all, or almost all, of the electricity sector allowances that are to be auctioned to retail providers, for the purposes of GHG emission reductions and customer bill relief. This could be done in a number of ways, including distributions in proportion to historical emissions in the retail provider's portfolio in a baseline year, or on a sales basis. We next describe these two alternatives.

5.2.2.1. Distribution in Proportion to Retail Providers' Historical Emissions

In this approach, allowances would be distributed to retail providers (for subsequent auctioning) in proportion to the historical emissions of sources and purchases used to serve each retail provider's load in a prior baseline year or multi-year period. The fixed proportions would be used to determine allowance allocations in subsequent years, with the actual amounts distributed to each retail provider depending on the total number of allowances allocated to retail providers each year. This approach is conceptually similar to distributions to deliverers on the basis of historical emissions, but the effects on average customer costs would be much less, largely due to the elimination of allowance rents to deliverers.

Figure 5-8 provides an illustrative example of the potential rate impacts for different retail providers due to a 100% auctioning approach, with all allowances distributed to retail providers in proportion to historical emissions of their portfolios.

Figure 5-8

Estimates of Effects on Average Retail Electricity Rates

Due to Allowances Distributed to Retail Providers

on the Basis of Historical Emissions

($/kWh, 2008$)

As illustrated clearly in Figure 5-8, the distribution of allowances to retail providers based on the historical emissions of their electricity portfolios would have much lower rate impacts than distributions to deliverers, and with much less variation among retail providers throughout the study period. Of course, greater variations may appear over time if individual retail providers modify their resource portfolios at different paces than assumed by E3. Larger rate impacts would also be expected if the number of allowances allocated to the electricity sector declines faster than emissions decline. While these generalizations about the potential effects of variations in resource portfolios and disparities between emission levels and available allowances also would apply to other allowance distribution approaches, we mention them in this context because of the marked similarities in modeled results for the various retail providers.

The extent to which historical emissions-based distributions to retail providers would recognize early actions that retail providers may have taken to reduce emissions would depend on the base period used.

Once the relative proportions based on the historical emissions of individual retail providers are established, retail providers would know in advance the number of allowances they could expect to receive each year. This would provide some certainty as retail providers plan for the use of auction revenues, though the auction proceeds could still vary widely depending on allowance prices.

5.2.2.2. Distribution in Proportion to Retail Providers' Sales

In this approach, allowances would be distributed to retail providers (for subsequent auctioning) in proportion to their sales during a specified period. The proportions of allowances distributed to individual retail providers would be updated periodically, to reflect relative changes in sales. This approach is conceptually similar to distributions to deliverers on the basis of output. A beneficial aspect of this approach is that it would accommodate and reflect differing growth rates in different retail providers' service territories.

Figure 5-9 provides an illustrative example of the potential rate impacts for different retail providers due to a 100% auctioning approach, with all allowances distributed to retail providers in proportion to their sales.

Figure 5-9

Estimates of Effects on Average Retail Electricity Rates

Due to Allowances Distributed to Retail Providers on the Basis of Sales

($/kWh, 2008$)

As Figure 5-9 indicates, rates would increase more for customers of retail providers with relatively high-emission portfolios and would increase less, or could even decrease, for customers of retail providers with relatively low-emission portfolios, with a resulting wealth transfer from customers of high-emitting retail providers to customers of retail providers with lower-emission portfolios.

Sales-based allocations to retail providers would provide incentives for retail providers to increase reliance on cost-effective renewables and other low-emitting generation. Some parties have argued that sales-based allocations would provide incentives for retail providers to increase sales rather than invest in energy efficiency, and that a measure of energy efficiency should be included in the sales calculation to reward early actions and to avoid incentives to increase sales. This matter is discussed in Section 5.4.3.

Compared to an historical emissions-based allocation, retail providers would have less certainty about the number of allowances they would receive, because the proportional distributions would depend on the sales of all retail providers.

5.2.3. Distribution of Allowances in Proportion to Economic Harm

SCE proposes that the allowance allocation methodology be devised to mitigate the economic harm caused by implementation of AB 32. SCE describes economic harm as the difference in an entity's economic outcome under a cap-and-trade system as opposed to business-as-usual conditions. In SCE's approach, allowances would be given to those entities that otherwise would experience economic harm due to the implementation of a GHG reduction program.

SCE asserts that this approach would be consistent with the equity guidance in AB 32 and would ensure that windfall profits are not created.

SCE submits that economic harm could occur in the electricity sector in the following situations:

· When an independent generator that sells power in a wholesale electricity market has an emissions rate that is higher than the emissions rate of the marginal generating unit that sets the market clearing price in that market. SCE submits that, in such a circumstance, the independent generator would incur emissions costs greater than the increased revenue it receives.

· When a retail provider owns generation that has GHG emissions or is responsible for the emissions costs of generation it has purchased by contract. In such a circumstance, the generation would not receive any market revenues because it directly serves load, and SCE expects that the emission costs would be recovered from the retail provider's customers, who would suffer resulting economic harm.

· When a retail provider purchases power from the wholesale electricity market but the market price has increased as a result of GHG regulation. Retail rates would be expected to increase as a result, with economic harm to customers.

· When an independent power producer has sold its output forward into the period of GHG reduction regulation without any contractual provisions to recover the new GHG costs.

If allowances are auctioned, SCE proposes that auction proceeds be distributed according to its economic harm-based methodology. SCE does not support targeting auction revenues to fund energy efficiency or renewables. It argues that the expected increases in market prices would make greater levels of energy efficiency and renewable energy projects cost-effective, and that no additional incentives would be needed. SCE points out further that, under its proposed economic harm-based allocation mechanism, a significant portion of allowances or auction revenue rights would be allocated to retail deliverers based on the economic burden of GHG regulation on their ratepayers, and would be available to mitigate increases in the revenue requirement resulting from an emissions cap. In SCE's view, the precise distribution of auction revenues by customer class should be determined by the Public Utilities Commission during an investor-owned utility's cost recovery proceedings.

5.3. Should Allowances or Auction Revenues be Distributed to Retail Providers?

With auctioning, the value of some or all of the auctioned allowances could be distributed to benefit consumers through at least two different ways:

· Direct centralized auction by ARB or its agent, with retail providers given auction revenue rights for some or all of the auctioned allowances; and

· Distribution of allowances to retail providers, with the provision that they must sell those allowances in a centralized auction undertaken by ARB or its agent, and receive the proceeds.

5.3.1. Positions of the Parties

SCPPA and PG&E prefer that allowances be distributed directly to retail providers with subsequent monetization of the allowances through an auction and a return of auction revenues in proportion to the number of allowances distributed to each retail provider. In SCPPA's view, this procedure could help to address its concerns about whether auction revenues would actually be returned to retail providers instead of being "siphoned off to other purposes." DRA expresses a similar concern that auction proceeds under the control of a State agency may be vulnerable when there are shortfalls in the State budget.

Calpine, Dynegy, WPTF, AReM, FPL, and IEP oppose distributing allowances directly to retail providers. These parties argue that such a step would raise a number of competitive fairness issues:

· Calpine is concerned that this would give control of the auction process to a certain segment of market participants, and that liquidity in the allowance market would be reduced, making it more difficult for the market to find the most cost-effective means for reducing emissions.

· Calpine states that distributing allowances to retail providers would raise market power concerns if retail provider-owned generation assets would have preferential access to allowances to the detriment of independent power producers and power marketers.

· Dynegy and IEP are concerned that retail providers could impose unreasonable conditions on allowance purchases or withhold them from the market altogether. Dynegy suggests that a retail provider could condition the availability of allowances to a supply agreement, and thus reap an unfair advantage over independent power producers. Dynegy argues further that such a system would create a price advantage for the retail providers, and would create an incentive for them to build their own generation rather than seek needed generation through competitive solicitations.

· WPTF argues that jurisdictional retail providers would have an inherent conflict of interest as the recipient of allowances because, in most instances, they also own generating resources and/or are in direct competition with independent entities for providing electricity to retail load. WPTF and AReM argue that a direct allocation of allowances to jurisdictional retail providers potentially would confer an unfair competitive advantage to utility-owned resources in procuring allowances, and create a concentration of market power.

· FPL describes that retail providers might have a competitive advantage in development of new generation projects if they have obtained the needed allowances for free.

These parties take the general position that the market structure must treat all similarly situated market participants in a non-discriminatory manner.

5.3.2. Discussion

The distribution of allowances to retail providers with the provision that they must sell those allowances in a centralized auction undertaken by ARB or its agent would satisfy both SCPPA's request for assurance that retail providers receive the anticipated revenues, and the independent providers' concerns that they not be disadvantaged due to the retail providers' access to allowance value for the benefit of retail customers.

Parties appear to be unified in their views that retail providers that receive allowances should be required to sell them through auction. As noted above, independent producers are concerned that allowing retail providers to use allowances that were given to them at no cost to meet compliance obligations while other entities are required to purchase allowances for their delivered electricity could have competitive consequences, including difficulties by independents in obtaining allowances and the unfair encouragement of more utility-owned generation. No party has voiced objection to the recommendation that retail providers should be required to sell at auction any allowances they receive.

We are aware of the anti-competitive concerns that the independent producers have raised regarding the distribution of allowances to retail providers. We agree that retail providers should be required to provide nondiscriminatory access to the allowances they own.

At the same time, having the retail providers rather than the State own the allowances at the time they are auctioned would simplify the auctioning and revenue distribution process, in that auction revenues would pass directly to the retail providers rather than being deposited first in State-controlled accounts and then redistributed to the retail providers through an auction revenue rights mechanism.

For these reasons, we recommend that ARB establish a centralized auction process, to be run by ARB or its agent. For the portion of allowances whose value ARB deems should be distributed to retail providers for the benefit of their customers, ARB should distribute the allowances directly to the retail providers with a requirement that they in turn sell the allowances in the centralized auction. Utility owned generation would then have the opportunity to purchase allowances on the same basis as other deliverers. Each retail provider should receive all auction revenues from the sale of the allowances that were distributed to it. ARB should establish the centralized auction with safeguards to ensure that this result is obtained. If ARB cannot design an auction that is legally separated from other State revenues, we suggest an alternate mechanism be designed.

In response to a question raised in comments on the proposed decision, we clarify that our recommendation that retail providers be required to sell the allowances they receive applies only to allowances received in their role as a retail provider, not to any allowances that a vertically-integrated entity that is both a retail provider and a deliverer may receive based on its deliveries to the grid.

5.4. Recommended Structure of Allowance Distributions in the Electricity Sector

In D.08-03-018, we determined that, if a multi-sector GHG cap-and-trade program is implemented in California, some portion of the emission allowances available to the electricity sector should be auctioned. We found, however, that additional record development was needed to allow us to make recommendations regarding the proper mix between auctions and administrative allocations of emission allowances to deliverers for the electricity sector.

As described above, the allowance distribution methods that we consider include:

· Auctioning: distribution of allowances to retail providers for subsequent auctioning;

· Distributions to deliverers, either free or at a set price;

· SCE's harm-based proposal; and

· Transitions, in particular, from mainly distributions to deliverers to greater amounts of auctioning, and from emissions-based to sales-based distributions to retail providers.

5.4.1. Positions of the Parties

5.4.1.1. Auctioning vs. Distribution to Deliverers

Most parties support initial auctioning of only a portion of allowances, either commencing immediately or within a few years after a cap-and-trade program begins, with a transition to auction larger numbers of allowances over time. As a complement to their views regarding auctioning, most parties support initial distribution of a portion of allowances to deliverers, with that portion declining as increased auctioning is phased in. Some parties support 100% auctioning from the beginning of the cap-and-trade program.

Some parties continue to argue against any auctioning. While we do not revisit our determination in D.08-03-018 that some portion of allowances should be auctioned, we consider those parties' cautions against auctioning in determining the amount of auctioning to recommend to ARB.

Low Initial Auction Levels/High Distributions to Deliverers

Some parties take the position that all allowances should be distributed to deliverers for free, with no auctioning (CMUA, Calpine, EPUC/CAC). An additional set of parties favored auctioning only a small number of allowances initially (SMUD, DRA, Dynegy, WPTF). Those parties that support no or small amounts of auctioning initially make the following arguments:

· Independent power producers would not have a guarantee of carbon cost recovery (EPUC/CAC). EPUC/CAC cite the presence of administratively determined prices, the scope of utility solicitations, and implementation of the CAISO's Market Redesign and Technology Upgrade (MRTU)⁴⁵ as factors that may affect a generator's ability to recover its carbon cost from the market.

· Independent power producers may have contracts with utilities that extend beyond 2012 for which there is no clear provision for recovery of new GHG costs. SDG&E/SoCalGas respond to this concern by suggesting that retail providers should give allowances to generators with fixed-price contracts signed prior to AB 32 that do not contemplate a GHG market.

· Auctioning may raise reliability concerns (IEP, Calpine, SMUD). Calpine argues that if third parties purchase large quantities of allowances and withhold them from the market, reliability could be threatened if insufficient allowances are available for generation to meet the load.

· Auctioning could create volatility in prices and auction revenue, making it difficult to plan effective infrastructure and programs (SMUD and CMUA). Calpine is concerned that volatility may make it difficult for generators to recover their compliance costs in the wholesale energy market.

· Uncertainty regarding allowance prices would make it difficult for entities with compliance obligations, especially publicly-owned utilities with deliverer responsibility for a significant portion of their portfolio, to plan their cash flow requirements if they must purchase allowances.

· Dynegy and SMUD assert that distribution of allowances to deliverers is needed to provide them funds for emission reduction investments.

· SCPPA raises market power and manipulation concerns about the conduct of auctioning, and general concerns about the complexity of an auction process.

Several parties favor transitioning to increased amounts of auctioning over time. DRA and WPTF submit that a transition period would provide time for deliverers to plan for compliance and make necessary adjustments to their financial plans to account for the impacts of GHG compliance obligations on their operating cash flow. DRA recommends that 25% of allowances be auctioned initially and that all allowances be auctioned by 2017. Powerex supports up to 25% auctioning initially, transitioning to 100%. These parties argue that a transition is needed for the following reasons:

· WPTF states that a transition period would enable generators to retain the resources needed for long-term investment in cleaner technologies and fuels.

· Transitioning from auctioning a small portion to auctioning a larger portion of the allowances would protect ratepayers from potential problems/market dysfunctions stemming from a sudden regulatory shift and the lack of familiarity with auctions in a regulatory context, while also ensuring adequate market liquidity for allowances.

Other parties express concern about a rapid transition to auctioning, such as the five-year transition to 100% auctioning as suggested by staff and DRA. These parties argue in favor of a slow transition to allow entities time to adjust to new market conditions. Dynegy suggests a 15-year transition to ensure that older generation needed for reliability stays online and older facilities have time to identify ways to reduce GHG emissions. Calpine recommends that a phase-in to auctions conclude around the year 2031. EPUC/CAC suggest a small two-year trial auction beginning in 2014, with future increases in auctioning phased in to avoid industry disruption. GPI supports auctioning a small fraction of allowances initially, with transitioning to increased reliance on auctions as the market develops, matures, and stabilizes.

High Initial Auction Levels/Low or No Distributions to Deliverers

Several parties (PG&E, NRDC/UCS, TURN, SCPPA, FPL, Johnson, CARE) recommend that, in the electricity sector, all or most emissions allowances be auctioned. SDG&E/SoCalGas support allocation of all allowances to retail providers, with appropriate measures to ensure that allowances are made available to the market on a non-discriminatory basis. They state that this proposal is equivalent to an auction approach with auction revenue rights allocated to retail providers, using the terminology of the staff paper.

These parties argue, variously, that auctioning would improve market liquidity (PG&E, Johnson, NRDC/UCS (joined by GPI)), reward early action (NRDC/UCS, GPI), and create a transparent price signal for the market (PG&E, Johnson). PG&E submits that retail customers will bear the ultimate costs of meeting GHG reduction goals and, therefore, should receive the value of the allowances to help mitigate their compliance costs. LADWP expresses similar views. Johnson states that whatever allocation benefits are desired could be achieved by allocating auction revenue rights, and that 100% auctioning may be simpler than a combination of auction and allocation to deliverers. NCPA argues that retail providers would have the best opportunities to mitigate carbon emissions, especially during the early years of the program.

While continuing to oppose inclusion of the electricity sector in a multi-sector cap-and-trade program, TURN states that most, if not all, allowances should be auctioned, and that it could support no more than an initial 20% allocation to deliverers based on emissions, to be phased out by 2016.

Several parties (PG&E, NRDC/UCS, GPI, TURN, SCPPA, Johnson, CARE) argue that giving allowances to deliverers would result in windfall profits to independent deliverers, with significant transfers of wealth from consumers to those deliverers. NRDC/UCS and TURN assert that most independent deliverers could recover the cost (or the opportunity cost) of allowances in their wholesale electricity prices. TURN cites information in the record that GHG emission reduction costs are likely to be much less than 50% of the value of the allowances. TURN points to a fairly low elasticity of demand for electricity, the absence of cheaper substitutes, and the lack of foreign competition as reasons why independent deliverers would be able to increase wholesale prices to recover GHG compliance costs. It states that only at certain breakpoints in allowance prices would there be a major change in the relative profitability of different production technologies. The supporters of free distributions to deliverers respond that the extent of any windfall profits would be limited, for various reasons, with DRA and WPTF arguing further that a quick transition to 100% auctioning would ensure that any windfall profits would be short-term and declining in nature.

Other

Under SCE's economic harm-based allocation proposal, deliverers and retail providers would receive allowances only to the extent that they otherwise would incur economic harm due to implementation of AB 32. SCE asserts that independent generation would incur economic harm if it sells electricity with an emissions rate higher than the emissions rate of the marginal unit that sets the market clearing price, or if it has long-term contract obligations to sell its output forward into the period of GHG regulation without contractual provisions to recover the new GHG costs. SCE submits that customers of retail providers would be harmed when a retail provider owns generation that has GHG emissions or is responsible for the emissions costs of generation it has purchased by contract, or when a retail provider purchases power at a market price that has increased as a result of GHG regulation. SCE concludes that independent generators and retail providers should receive allowances in these circumstances.

SCE asserts that, if its economic harm proposal is not adopted, capital investments made prior to AB 32 under laws and rules that did not require pricing of GHG emissions may have to be abandoned prematurely, raising questions of equitable treatment and imposing significant costs to the California economy.

5.4.1.2. Historical Emissions-based Distributions to Deliverers

Several parties (Dynegy, DRA, TURN) state that allocations to deliverers should be based on historical emissions. DRA proposes emissions-based distributions to deliverers, so that the relative proportion of free allowances allocated to each deliverer would remain constant until 2017, when all allowances would be auctioned under DRA's proposal. TURN states that it could support no more than an initial 20% allocation to deliverers based on emissions, to be eliminated by 2016. These parties offer the following arguments for historical emissions-based allocations to deliverers:

· An historical emissions-based allocation system would recognize the reliability benefits conferred by such sources, provide funding for emission reductions investments, and offset some of the expected loss of market value of emitting resources (Dynegy).

· An historical emissions-based allocation would protect the value of current resources occurred in compliance with all then-existing regulatory requirements (Dynegy).

· An historical emissions-based allocation approach would provide a predictable amount of free allowances to individual deliverers, which would be desirable from a business planning perspective (DRA).

Other parties (PG&E, SCE, NRDC/UCS) oppose historical emissions-based allowance allocations to deliverers. These parties provide the following arguments against this allocation procedure:

· An historical emissions-based approach would penalize entities that have already invested in low-GHG technologies and fuels (NRDC/UCS and Calpine).

· This approach would not provide an incentive for efficiency improvements or investments in cleaner and more-efficient generating technologies (Calpine).

· Necessary assumptions regarding emissions rates of market purchases and non-unit-specific contracts would result in an inaccurate allowance allocation (PG&E).

· Some generators would receive an unearned windfall of the allocation value (NRDC/UCS and SCE).

· An historical emissions-based allocation of allowances to deliverers would result in transfers of wealth from consumers to producers or deliverers (SCPPA).

· Clean utilities could pay twice under an emissions-based allocation: once for clean investments and a second time to generate what are more expensive emission reductions to meet the cap or obtain allowances (NRDC/UCS).

Though supporting initial allocations to deliverers based on historical emissions, DRA recognizes that an historical emissions-based allowance allocation methodology for deliverers would disadvantage customers of utilities that purchase most of their power from independent producers, relative to customers of utilities that are vertically integrated, but states that this disadvantage would be eliminated by 2017, when all allowances would be auctioned under DRA's proposal.

5.4.1.3. Output-based Distributions to Deliverers

Parties provide general comments on output-based allocation methodologies, with some also commenting on specific output-based variations, including limiting distributions to only deliverers with emitting sources, and fuel-based differentiations, as described in the staff paper.

Output-based allocations to deliverers using all or most generation types are supported by three parties (Calpine, Solar Alliance, and CRA). Solar Alliance and CRA both favor some allocation to new renewable generation, although neither comments on whether there should be allocations to deliverers using existing non-emitting sources. These parties offer the following arguments in favor of output-based allocation to deliverers:

· Output-based allocations to deliverers would reflect current market conditions and provide incentives for investment in low-GHG technologies and fuels (Calpine).

· This approach would recognize early actors since the quantity of allowances received would be based on the entity's output rather than historical emissions, and would not create perverse incentives to extend the life of dirty, inefficient generators or contracts with these generators (Calpine).

Parties that oppose an output-based allocation methodology for deliverers provide the following arguments:

· Output-based allocations would provide valuable allowances to non-emitting entities that have no need for them because they do not have a compliance obligation (Dynegy). These deliverers would already see an increase in profits as the wholesale price of power rises.

· An output-based allocation methodology might give generators the perverse incentive to increase output in order to increase their share of allowances (DRA). Calpine responds to this argument by asserting that an output-based approach would only provide incentives for cleaner technologies to increase production. Calpine asserts that the expected yearly declines in the number of allowances granted would place downward pressure on emission levels.

· This approach would create a wealth transfer from high-emitting entities to low-emitting resources (SCE, LADWP).

· An output-based approach would not help high-emitting resources receive the allowances necessary to transition to a carbon-constrained economy (SCE).

· Uncertainty regarding the level of year-to-year distributions to individual deliverers would create risk for deliverers and would make it difficult for entities to predict compliance costs (SCE and DRA).

· An output-based method for distributing allowances to deliverers should not be considered until a more robust modeling analysis of the proposal can be completed, to assess the impact of an output-based approach on bidding behavior (SCPPA).

Some parties oppose the staff proposal to limit output-based allocations to only deliverers that use emitting generation. SCE and GPI assert that this approach would result in windfall profits for natural gas generators at the expense of coal generation.

SMUD supports a fuel-differentiated output-based allocation of allowances and would include new renewables and energy efficiency after AB 32 became law, but would not grant allowances for non-emitting resources existing before passage of AB 32. SMUD asserts that this would be a simple, cost-effective method to reward early action for adding clean resources while acceptably reducing regional imbalances due to historical resource ownership. SCPPA states that a fuel-differentiated output-based allocation to emitting deliverers would merit further examination. It asserts, however, that the output-based allocation of allowances to deliverers should not be pursued without undertaking further modeling to determine whether the claimed market clearing price mitigation would actually occur.

Some parties offer arguments against fuel-differentiated output-based allocations to deliverers. These parties make the following arguments against fuel-differentiated allocations:

· Allocation to deliverers on a fuel-differentiated basis could make it more expensive for a relatively inefficient GHG gas-fired generator to run than an efficient coal-fired generator (SDG&E/SoCalGas).

· Applying a weighting factor to resources based on the fuel type would complicate an output-based allocation methodology and could be gamed (DRA).

SCE argues that an assumption that market clearing prices would not increase under an output-based approach would ignore the fact (so SCE alleges) that a marginal generating unit (which sets the market-clearing price) would not receive allowances sufficient to cover its emissions. SCE sees such a shortfall occurring in two ways. SCE contends that there would be a shortfall of allowances to emitting generators, first, if allowances are allocated to non-emitting resources and, second, because the allowance cap would decline each year. SCE maintains that generators would include these shortfalls in their bids and also would increase their bids to recover the risk uncertainty related to the number of allowances they receive. SCE also explains that, because the State's total generation fluctuates each year, the number of allowances that a deliverer would receive would vary depending on variables such as temperature and hydro levels. SCE argues further that an output-based approach would be less efficient than other approaches because entities could alter their allowance allocation through current or future behavior.

5.4.1.4. Transition from Emissions-based to Output-based Distributions for Deliverers

EPUC/CAC support a hybrid historical emissions/output-based allocation that gradually transitions to full output-based by 2020. They recommend that the output-based approach distribute allowances to deliverers based on the lower of their actual or an average emissions benchmark, and that a five-year baseline be used for output determination in the output-based approach.

5.4.1.5. Allowances for New Deliverers

EPUC/CAC submit that a new entrant reserve should be set aside for new generation, sized sufficiently to accommodate new generation needs and taking into account load growth, anticipated plant retirements, and increased efficiency from repowering. In their view, CHP and other low-carbon generation should be given priority in a new entrant reserve to recognize their efficient fuel use and carbon reduction benefits.

DRA recommends that, given the relatively short transition it proposes to 100% auction, new deliverers should purchase all of their allowances in the auction.

5.4.1.6. Historical Emissions-based Distributions to Retail Providers

SCPPA states that, if auctioning with the distribution of auction revenues to retail providers is undertaken, the distributions should be based on the emissions associated with each retail provider's total portfolio. It asserts that this approach would have little or no potential for creating wealth transfers among retail providers.

PG&E disagrees, arguing that an allocation methodology based on historical emissions associated with a retail provider's load would not recognize prior investments made in zero or low-carbon generation and energy efficiency. PG&E asserts that use of historical emissions associated with load would require assumptions regarding emission rates of market purchases and non-unit-specific contracts, which would result in an inaccurate allowance allocation. PG&E also contends that allowance allocation options such as those based on historical emissions or which fail to provide credit to sources or categories of sources for emissions reductions prior to implementation of AB 32 would violate the express requirement in AB 32 that sources of emissions receive credit for early actions (Section 38562(b)(3)).

SDG&E/SoCalGas argue similarly that allocation of allowances to retail providers based on emissions rather than sales would be inconsistent with the mandates of AB 32 in Sections 38562(b)(1) and (3) to "encourage early action" and give "appropriate credit for early voluntary reductions." They assert that emissions-based allocations would punish customers of retail providers that already have incurred significant costs to reduce their emissions, and would reward retail providers that have delayed reducing their emissions. They argue further that emissions-based allocations would fail to reflect the costs imposed on society by high-emission deliverers.

5.4.1.7. Sales-based Distributions to Retail Providers

PG&E supports distribution of all allowances to retail providers on the basis of sales, and suggests an updating metric such as current retail electricity sales adjusted for verified customer energy efficiency savings. PG&E supports this approach on the basis that it would recognize and encourage early action and would also encourage aggressive deployment of energy efficiency and investments in low- and zero-emissions generating technologies. PG&E states that its proposal would be equitable to retail providers with varying emissions rates, arguing that, while a utility's current emissions are one element that determines the average cost to customers, low-emitting utilities will have fewer low-cost GHG reduction opportunities and high-emitting utilities may have more lower-cost emission reduction opportunities within their own portfolio. PG&E argues further that equity goals support its proposal, asserting that those entities with high-emitting resources in their portfolio should be responsible for the cost of those emissions and that those costs should not and lawfully may not be assigned and shifted to customers who do not receive the benefits of the electricity from these higher-emitting resources.

SDG&E/SoCalGas similarly support allocation to retail providers on the basis of sales adjusted for cumulative energy efficiency savings. They state that updating allowance allocations to retail providers based on sales may introduce some inefficiency by creating incentives to increase sales, if verified energy efficiency is not included. They submit that including cumulative energy efficiency savings would reduce this potential inefficiency while accounting for higher growth in some areas.

SDG&E/SoCalGas state that mandatory GHG reduction measures would not require retail providers with a high GHG-emitting portfolio to undertake any more actions than low-emitting retail providers and argue, as a result, that it makes sense to fund the mandatory measures with allocation of allowances or auction revenue rights on a sales basis. They contend that higher-emitting retail providers have the "headroom" in rates necessary to incur costs similar to those that have been realized already by the lower-emitting retail providers in reducing their emissions. They expect that GHG-reducing strategies such as energy efficiency currently available to publicly-owned utilities are, in large part, less expensive than opportunities currently available to investor-owned utilities, because of the energy efficiency achievements already attained by investor-owned utilities.

SCE and SCPPA oppose a sales-based allocation of auction revenue rights to retail providers, because of its tendency to result in wealth transfers from more carbon-intensive retail providers to less carbon-intensive retail providers.

SCPPA states that basing retail provider allocations on net load (gross retail provider load less load served by legacy hydroelectric and nuclear resources), as suggested by staff, would mitigate somewhat the wealth transfer effect of a sales-based allocation, and that allocation to retail providers on a fuel-differentiated basis, so that there would be proportionately higher allocation of allowances or auction revenue rights to coal-served load, would further mitigate the wealth transfer.

5.4.1.8. Transition from Historical Emissions-based to Sales-based Distributions for Retail Providers

SMUD supports allocation of auction revenue rights to retail providers based on emissions initially, and sales later. SMUD supports retail providers receiving auction revenue for renewable energy and energy efficiency.

PG&E asserts that, if a sales-based distribution approach is not implemented immediately, there should be a short transition to this approach, so that all utilities are held to the same benchmark emissions rate as quickly as possible.

SCPPA opposes a transition to sales-based allocations for retail providers because of the wealth transfers that would occur. It states that such a transition would fail to recognize that various retail providers, including SCPPA members, have existing contracts with coal plants that will not expire until later years (including 2019 for the LADWP contract with the Navajo coal plant and 2027 for various SCPPA members' contracts with Intermountain Power Project). SCPPA argues that there should be, at most, a minimal transition by 2020 from an emissions-based allocation of auction revenue rights among retail providers toward a sales-based allocation.

While not making firm recommendations, NRDC/UCS suggest that auction revenue distributions to retail providers in 2012 based partly on emissions and partly on sales adjusted for verified energy savings would provide some accommodation for those carbon-intensive retail providers that need to reduce their emissions the most, but at the same time would reward and not penalize those utilities that took early actions prior to the start of the program in 2012. They recommend that the distribution approach for retail providers transition to 100% sales-based, adjusted for verified energy efficiency savings, by 2020 or earlier. In their view, this would provide long-term incentives for retail providers to reduce the overall emissions associated with serving their customers. They recommend that any sales-based distributions should use sales that are adjusted for verified energy efficiency savings, in order to provide proper incentives for emissions reductions and adherence to the State's loading order. NRDC/UCS urge the Commissions, in determining allocation policies, to focus on the equity impacts for all entities involved. They recognize that the most carbon-intensive retail providers in the State would need to make significant investments in order to clean up their systems. At the same time, they are concerned that distributions to retail providers on an emissions basis would tend to reward the dirtier utilities while penalizing the cleaner utilities; they submit that sales-based distributions would have the opposite effect.

CARE supports the staff proposal to distribute auction revenues to retail providers using a transition from an historical emissions basis to a sales basis, with the sales determination including renewables but excluding nuclear and large hydro.

5.4.2. Discussion

We determined in D.08-03-018 that some allowances allocated to the electricity sector should be auctioned. Today, we address other issues regarding the structure of allowance distributions in the electricity sector, including what portion of the allowances allocated to the electricity sector should be auctioned.

We evaluate the various alternatives for structuring allowance distributions in the electricity sector using the evaluation criteria and goals discussed in Section 5.1, as follows:

· Minimize costs to consumers.

· Treat all market participants equitably and fairly.

· Support a well-functioning cap-and-trade market.

· Align incentives with the emission reduction goals of AB 32.

· Administrative simplicity and feasibility.

We find it useful to address the allowance distribution proposals brought forward by GPI and SCE first, before turning to the other alternatives before us.

5.4.2.1. Distribution of Rights to Purchase Allowances

GPI proposes that, to the extent that allowances are not auctioned, ARB should administratively allocate to deliverers the rights to purchase allowances at a pre-determined, administratively set price. GPI's proposal is described in more detail in Section 5.2.1.3 above.

According to GPI, the allocation of purchase rights would have significant advantages over distributing free allowances. GPI states that, by granting purchase rights to entities with compliance obligations, ARB would ensure that these entities have access to the allowances they need to meet their compliance obligation. At the same time, selling these allowances at a fixed price would ensure that the State generates revenue from the allocation. GPI argues further that the sale of allowances would limit the windfall profits realized when allowances are distributed for free on an emissions basis.

We recognize the potential benefits that might be obtained by an allocation of purchase rights, as described by GPI. However, in practice, any relative benefits of this proposal would hinge on the setting of the administrative price of the allowances. Setting a "well-determined price," as GPI suggests, would determine how successful this allocation would be at limiting windfalls and generating revenue for the State.

The risks of not setting a "well-determined" price may outweigh any benefits that could be derived from this allocation method. If the administratively set price turned out to be higher than the market value of the allowances, the allocation of purchase rights at that price would provide no value to the entities with purchase rights. In such a situation, entities with purchase rights might chose not to exercise their purchase right, but instead buy allowances at market prices in the auction or secondary market. This would eliminate one of the benefits of free allocations to deliverers, that is, that free allocations would help entities avoid negative impacts due to investment and procurement decisions made prior to GHG regulation.

If the administratively set price was less than the market value of the allowances, entities with purchase rights could still derive some windfall profits from the allowances, while the State would obtain a limited share of the value of allowances for consumer purposes.

Additionally, it is not clear what relationship a "well-determined price" would have to the market price. And even if the ideal relationship were known, it is not clear what basis the State would have for administratively setting the purchase price during the initial years of the program, before experience has been gained regarding market prices.

We conclude that these risks and administrative problems make GPI's proposed method less desirable than the administrative allocation of free allowances to deliverers, to the extent that such administrative allocations are deemed appropriate.

5.4.2.2. Harm-based Distribution of Allowances

SCE asserts that the most effective way to design an equitable and low-cost cap-and-trade program is by identifying entities that would suffer economic harm under the program and allocating free allowances to such harmed entities. As described in Section 5.2.4 above, SCE identifies four types of situations in which generators or retail customers in the electricity sector could be harmed.

Some parties (SDG&E/SoCalGas and WPTF) criticize the SCE harm-based allocation approach. SDG&E/SoCalGas object to all fuel-specific allocation methods for failing to provide "near-term incentives" for high-emitting entities to reduce their emissions. WPTF argues that, because most of the specified coal in California's generation mix is utility-owned, SCE's proposal would create an unfair benefit for utilities. PG&E also opposes SCE's proposal, asserting that it would result in an ongoing inefficiency and unfairness that can create a significant cost to the economy and sustain excess profits for coal generators.

SCE's economic harm concept provides a useful perspective as we consider the various allocation proposals. The proposal that allowances should be distributed in a method that compensates for economic harm resulting from the GHG regulatory scheme has value, and is generally consistent with the equity criterion, grounded in AB 32, that we have identified and that we apply in today's decision. However, there are several shortcomings to SCE's proposal that prevent us from recommending it.

The first situation of economic harm that SCE identifies would occur if an independent generator that sells power in a wholesale electricity market has an emissions rate that is higher than the emissions rate of the marginal generating unit that sets the market clearing price in that market. While we agree in general with SCE's characterization, SCE has not suggested, and we do not readily see, how an allowance allocation mechanism could be devised that would pinpoint with any accuracy the situations and generators for which such economic harm would occur, or the amount of economic harm that would occur.

The second situation that SCE identifies is that retail rates would be expected to increase to reflect GHG costs of electricity that the retail provider either owns or is responsible for through a purchase contract. This would include, in particular, coal and other fossil resources owned by the retail provider. The third situation that SCE identifies is that retail rates would increase due to a retail provider's wholesale electricity purchases when the market price has increased as a result of GHG regulation. We agree that an equitable allocation mechanism should take into account the economic harm to consumers arising from GHG compliance obligations for such resources and market purchases.

Finally, SCE is concerned that independent producers may have long-term contracts, extending into the period of GHG regulation without contractual provisions to recover the new GHG costs.

As described in more detail below, the combined recommendations that we make to ARB regarding the appropriate allocation and distribution of allowances within the electricity sector, taken together, would achieve results generally consistent with SCE's proposal, particularly in the short term. We believe that our recommendations, however, would provide stronger incentives for deliverers and retail providers to reduce GHG emissions in the longer term than would SCE's approach. By compensating entities indefinitely, SCE's approach would not provide incentives for the long-term modifications to the resource mix that we believe are crucial to meet the goals of AB 32.

In an allocation workshop presentation, SCE suggested what it characterized as a modified version of its harm-based approach. SCE identified coal generators and ratepayers as the primary entities in the electricity sector that would be harmed by a cap-and-trade program. SCE suggested that allowances be allocated to coal generators using an historical emissions-based allocation, with remaining allowances allocated to retail providers on a sales basis. Sales would be determined net of sales from coal generation, because economic harm for this fuel source would already be addressed through the separate allocation to coal generators.

As described below, one of our recommendations to ARB is that the method of distributing allowances to retail providers transition from an historical emissions-based methodology to a sales-based methodology. With the anticipated expiration of existing coal contracts, the approach we recommend is similar to that suggested by SCE in the allocation workshop. We believe the approach we recommend is preferable, however, because it recognizes the range of past investment and procurement decisions, not just coal investments, that could cause economic harm in a GHG regulatory structure.

5.4.2.3. Comparison of Allowance Distribution Alternatives

With rejection of the GPI and SCE proposals, we now consider how the remaining allowance distribution alternatives considered in this proceeding would perform relative to the criteria and goals described in Section 5.1.

Minimization of Costs to Consumers

As we describe in Section 5.2, free distributions of allowances to deliverers in proportion to historical emissions would be the most expensive distribution option, on average, for customers, other than auctioning with no distribution of allowances to retail providers. This is due to the windfall profits in the form of allowance rents that independent deliverers would enjoy, in addition to full reflection of GHG compliance costs in market prices and the accompanying clean generation rents.

The average retail rate impacts due to free distributions to deliverers based on the amount of electricity they deliver to the California grid would depend on the extent to which the allowance value would be included in wholesale market prices. If the full allowance value was included in wholesale market rates, average retail rate increases would approach those expected with distribution to deliverers based on historical emissions. On the other hand, if no or almost no allowance value was included in wholesale market rates, average retail rate impacts would be minimal, with the possibility of average rates actually declining if distributions to deliverers were structured such that deliverers of the marginal generation that sets market prices receive allowances in excess of their compliance needs. This might happen, for example, with an emitter-only output-based allocation that leaves deliverers of coal generation short and deliverers of gas generation long on allowances.

Auctioning with distribution of all allowances to retail providers would have average statewide rate impacts resulting from reflection of full GHG compliance costs in market prices and the resulting clean generation rents. While there would be distributional effects among customers of different retail providers, the average statewide rate impacts would vary only minimally among the methods considered for distributing allowances to retail providers.

In addition to average rate impacts due to the various allowance distribution options, there would be variations in rate impacts among customers of different retail providers due to differences both in the resource mix of utility-owned or controlled resources, and in the extent to which the retail providers rely on market purchases. As our analysis in Section 5.2 indicates, auctioning with distribution of allowances to retail providers based on historical emissions would cause the least variation in rate impacts among the retail providers. Sales-based distributions to retail providers would have the largest distributional impacts among customers of different retail providers, unless and until retail providers adjust their resource mix to reduce the emissions of their portfolios.

Historical emissions-based distributions to deliverers would minimize wealth transfers from customers of retail providers with relatively high emitting portfolios to customers of retail providers with cleaner portfolios. However, there would still be distributional variations based on the degree of the retail providers' reliance on market purchases.

Fuel-differentiated output-based distributions to deliverers of electricity from emitting generation resources (including unspecified sources) would perform similarly to historical emissions-based distributions to deliverers in terms of minimizing wealth transfers based on the emissions characteristics of the retail providers' portfolios. There would still be distributional variations based on the degree of the retail providers' reliance on market purchases. On the other hand, a pure output-based distribution would provide allowance rents to independent deliverers of zero- and low-emission electricity, including those under contract to retail providers. This would result in wealth transfers from customers of retail providers with relatively high-emitting portfolios to customers of retail providers with relatively low-emitting portfolios. Limiting output-based distributions to only deliverers of electricity from emitting generation resources would moderate the allowance rents and resulting wealth transfers.

Equitable and Fair Treatment of Market Participants

One of the measures of equity is whether an allocation methodology would cause negative impacts to market participants due to investment and procurement decisions made prior to GHG regulations. For retail providers, this concept is addressed above in the discussion of wealth transfers among customers of different retail providers.

Independent deliverers are concerned about whether they would have an opportunity to recover their carbon costs. The record identifies at least two types of situations in which independent deliverers may have trouble recovering compliance costs, to the extent the costs are not mitigated through (free) allowance distributions: (1) independent deliverers with emissions rates higher than the emission rates of the marginal generator whose allowance costs are reflected in the market price, and (2) contracts that extend beyond 2011 and do not provide for recovery of carbon costs. The distribution of allowances to deliverers could help such deliverers, whereas auctioning would not.

A related concept, but with different proponents, addresses the extent to which entities that cause GHG emissions are held responsible for the compliance costs of those emissions, which has been characterized as the "polluter pays" argument.

A related equity consideration addresses the extent to which an allowance distribution method recognizes early actions that have reduced an entity's GHG emissions.

Free distributions to deliverers based on their historical emissions or fuel-differentiated output-based metrics would reduce the compliance costs of high-emitting sources. Free distributions to deliverers based on their historical emissions would reward early actions that the deliverers take after the baseline period to reduce the emissions of the electricity they deliver to the California grid, as described in Section 5.2.1.1. Distributions using output-based metrics also would also benefit deliverers that take early actions to reduce their emissions, as described in Section 5.2.1.2. Conversely, pure output-based distributions to deliverers, and sales-based distributions to retail providers would reward the development of renewable sources. As we discuss in Section 5.4.3, a sales-based distribution to retail providers could be modified to reward emission reductions due to energy efficiency. Distributions to retail providers based on their historical emissions would benefit retail providers that take early actions after the baseline period.

We also assess the extent to which allowance distribution approaches provide revenues to fund emission reductions, compliance obligations, and/or customer rate reductions. Auctions with the distribution of allowances to retail providers would provide such funds to retail providers. Distributions to deliverers based on historical emissions, or based on a fuel-differentiated output-based metric, would roughly match deliverers' compliance obligations and needs for funding emission reductions. The continued sufficiency of such funds would depend on the extent to which the number of allowances allocated to the electricity sector diverges from the sector's emissions over time. Distributions based on deliverers' output or retail providers' sales would reduce the allowances available to deliverers or retail providers with the highest compliance obligations.

As we establish in Section 5.3, retail providers that receive allowances should sell them through a centralized auction, to avoid potential competitive concerns. An important benefit of auctioning is that it would allow equal access to allowances for both established deliverers and new delivers seeking to enter the market. Auctioning with allowance distributions to retail providers based on sales would provide allowances to new retail providers on an equal basis with existing retail providers, although perhaps with a short time lag. A similar result would hold for allowance distributions to deliverers based on their output. Allowance distributions based on historical emissions of retail providers, or historical emissions of deliverers, would place new retail providers or new deliverers, respectively, at a competitive disadvantage unless appropriate set-asides were established for them.

Align Incentives with the Emission Reduction Goals of AB 32

Auctioning would provide strong incentives for all deliverers to reduce GHG emissions, in order to reduce their compliance costs. The reflection of the full cost of GHG compliance in wholesale rates would also provide incentives for retail providers to serve their customers through lower-emission means. Allowance distributions to deliverers on the basis of historical emissions would provide a stronger incentive to reduce emissions than would distributions on an output basis because the historical emissions approach would provide allowances that deliverers could sell if they reduce their emissions. Additionally, if an output-based approach results in lower wholesale market prices, as theorized, that would prompt less end-use efficiency than would the higher prices expected with historical emissions-based distributions to deliverers.

Support a Well-functioning Cap-and-Trade Market

Auctioning of allowances would improve market liquidity, which could improve the accuracy and reduce the volatility of price signals in the market.

With auctions, deliverers would have reliable access to allowances without having to rely on secondary markets, but they would not know the price they would have to pay. With free allowance distributions to deliverers, they would have a degree of certainty about the availability of some number of free allowances to help meet compliance obligations. With distributions based on historical emissions, deliverers may know the number of allowances they would receive ahead of time whereas, with distributions based on output, the number of allowances distributed to an individual deliverer would depend on its output as well as the output of other deliverers. In all distribution options, the entities that receive allowances would not know the value of the free allowances or the cost of any other allowances they may need to purchase in the secondary market.

Administrative Simplicity

Auctions could be complicated to design and implement. One concern voiced by many parties is the lack of experience with auctioning of GHG allowances in California. The various methods of distributing allowances to either retail providers (for subsequent auctioning) or to deliverers would have differing challenges but (aside from the GPI and SCE proposals which we have rejected) appear to be administratively feasible.

5.4.2.4. Conclusions

First, we consider what amount of allowances should be auctioned for the electricity sector. There are strong arguments in support of auctioning all or most allowances. Auctioning of allowances would provide market liquidity, which would improve the accuracy of price signals in the market. A centralized auction undertaken by ARB or its agent would ensure that all deliverers have equal access to allowances, and would reduce or avoid the need for a set-aside or other administrative accommodation for new entrants. We expect that, with auctioning, GHG compliance costs would be internalized in wholesale electricity prices, sending more accurate price signals that would encourage participants in the electricity sector to reduce emissions. Entities with compliance obligations would bear full financial responsibility for the emissions associated with the electricity that they deliver to the California grid. At the same time, unlike free allowance distribution to deliverers, auctioning would preclude windfall profits due to allowance rents received by independent deliverers. However, the inclusion of allowance costs in wholesale prices would allow independent deliverers of relatively low-emission electricity to earn clean generation rents. As SCE points out, such increased profits for clean generation would be expected as a normal part of a functioning market, and should help spur additional investment in clean generation technologies. For all of these reasons, we believe it is desirable to move quickly to full auctioning.

We are persuaded, however, that auctioning should be phased in, with a fairly brief transition period. We anticipate that any cap-and-trade program that ARB implements will be linked to a regional, and ideally national, market. A transition to auctioning would help protect ratepayers if problems arise as this new mechanism is implemented and experience is gained with the auctioning process. A phased approach would begin the auctioning process so that California can reap initial benefits and, at the same time, would provide some protection and stability while the cap-and-trade market develops and matures.

As another reason for phasing in auctioning, the distribution of some free allowances to deliverers would be beneficial as an interim measure. Distributing some free allowances to deliverers would reduce short-term impacts on generating resources, and would help generators adapt to the new regulatory environment. Such distributions would provide time and financial resources that deliverers may need to make necessary adjustments to their financial and investment plans to account for the impacts of GHG compliance obligations. This need for free allocations to deliverers would decline over time.

In its allocation paper, staff suggests a six-year transition to 100% auctioning. Several parties, including WPTF (recommending an 8-year transition), Dynegy (recommending 15 years), and Calpine (recommending 19 years), argue that a longer transition period is needed because of the long lead time required for new infrastructure to become operational and in order to provide more time for generators to recover their current costs and to make plans for the transition. EPUC/CAC suggest a small two-year trial beginning in 2014 with future increases phased in to avoid industry disruption.

We conclude that free allocations to deliverers should transition to an auction of 100% of allowances by 2016. By increasing auction levels over this five-year period (and recognizing the advance notice that the industry is already receiving), entities with existing high-emitting resources would have time to adjust their generation investments before they face the full cost of their emissions. At the same time, a five-year transition would ensure that any undue windfall profits to deliverers would be short-term and declining in nature, as suggested by DRA and WPTF.

We conclude that in 2012 there should be 20% auctioning and 80% free allocation of allowances to deliverers, with a transition to 100% auctioning by 2016, as shown in Table 5-3.

Table 5-3

Recommended Transition for Auctioning and
Distribution of Allowances to Deliverers

	Percentage of Allowances Sold through Auction	Percentage of Allowances Allocated to Deliverers
2012	20	80
2013	40	60
2014	60	40
2015	80	20
2016	100	0

This transition schedule would, in our judgement, allow California to gain experience with auctioning and fine-tune the auctioning structure, if needed, while ensuring that market participants receive a correct price signal regarding the cost of GHG compliance and have time to adjust their operations and investments. The knowledge that 100% auctioning would begin in a few years would give deliverers a strong incentive to move quickly to complete their preparations in a timely way.

We turn now to the manner in which allowances should be distributed to deliverers during the transition to auctioning, and also the manner in which allowances to be auctioned should be distributed to retail providers.

As discussed in Section 5.5 below, we recommend that all, or almost all, of the electricity sector allowances to be auctioned be distributed to retail providers. ARB may choose to retain a small percentage of allowances to be owned by the State in order to use the related auction revenues for various purposes consistent with AB 32, but we recommend that all auction revenues from allowances allocated to the electricity sector be used for the benefit of the electricity sector.

As the percentage of allowances distributed to deliverers phases down, the percentage distributed to retail providers would increase by comparable amounts, lacking only those allowances that ARB retains for statewide purposes.

Because of this interrelationship between distributions to deliverers and distributions to retail providers, we find it helpful to consider together the manner in which allowances should be distributed to individual deliverers and to individual retail providers. This approach makes it easier for us to ensure that the policies for distributions to deliverers and retail providers are coordinated in a manner that best meets and balances the allocation criteria and goals that we establish in Section 5.1.

The first criterion, aimed at minimizing costs to consumers, can be viewed as a subset of the second criterion regarding equitable and fair treatment of all market participants. There is no single measure of equity. We attempt to reach a reasonable balance among the competing interests and goals, so that each entity is treated fairly and each deliverer has reasonable options to ensure compliance.

Equity among customers of different retail providers would be affected by policies for distribution of allowances to both deliverers and to retail providers. The impact on customers of allowance distributions to deliverers would depend on how much of its power a retail provider owns or purchases, the emissions profile of the retail provider's electricity portfolio, and the extent to which GHG allowance cost (or opportunity cost) is reflected in market prices.

Some parties argue, on the basis of equity, that deliverers should receive allowances in proportion to their output, or similarly that retail providers should receive allowances in proportion to their sales, with several supporters of sales-based allocations requesting that the assessment of sales include a measure of energy efficiency. These parties assert that such an approach would recognize early actions appropriately and would encourage investment in low-and zero-emitting technologies. PG&E argues that its customers should benefit from its relatively low-carbon footprint and that PG&E should not be required to reduce carbon emissions as much as other retail providers that have undertaken less energy efficiency and have a more carbon-intensive resource mix.

Other parties argue that historical emissions-based allocation methods would be more equitable because they would match more closely the deliverers' compliance obligations and would help protect customers of retail providers with high-emission portfolios from economic harm. LADWP asserts that a fair allocation policy would direct allowances toward high-emitting entities with incentives to increase their low- and non-emitting resources.

In weighing the evaluation criteria, we find that a primary consideration in the early years of a cap-and-trade program is to ensure that economic harm is mitigated to the range of market participants in the electricity sector, including customers, retail providers, and deliverers. For customers and retail providers, that goal would be met through the combined policies for distributions to retail providers and distributions to deliverers. For independent producers, that goal would be met through policies for deliverer distributions. Because of the need to prevent economic harm in the short term while market participants undertake the steps necessary to align their operations to a GHG regime, we conclude that, in the early years, allowances should be allocated in a manner that reflects compliance obligations.

While always important, in the longer term greater emphasis should be placed on the provision of strong incentives for both deliverers and retail providers to reduce GHG emissions, both through reductions in the emissions profile of electricity that is delivered to the grid and procured by the retail providers, and through aggressive actions by retail providers and others to improve the efficiency with which electricity is used. While the transition to these longer-term distribution policies will be phased in, and strong programmatic measures to require energy efficiency and renewable energy gains will be in place, it is still helpful to send a clear message to all market participants that they need to make plans, commencing well before the cap-and-trade program begins, to undertake the capital investments and other changes that may be needed to protect their financial interests and customers in the longer term.

Allowance Distributions to Deliverers

For the portion of allowances distributed to deliverers, we recommend a fuel-differentiated output-based approach with distributions limited to deliverers of electricity from emitting generation resources (regardless of whether the electricity is generated inside or outside of California). This approach would provide all deliverers with allowances roughly in proportion to the amount they need.⁴⁶ The fuel-differentiated distribution of allowances to deliverers, with regular updating, would focus allowances on the deliverers that would need them most for compliance purposes, thus reducing the potential for windfall profits due to excess free allowances ("allowance rent"), compared to other output-based approaches or the historical emissions-based approach.

It has been suggested that fuel-differentiated and other output-based allocation distributions to deliverers may limit the increase in wholesale electricity prices, because they would provide generators with an incentive to maintain or increase their output. We do not know the extent to which that may be the case, although the reasoning seems somewhat persuasive. At the same time, as some parties point out, deliverers with the marginal generating units (which set the market clearing price) may or may not receive allowances sufficient to cover their compliance obligations. To the extent they do not, their allowance shortfalls would be a cost that they could be expected to include in their market bids. This amount may be considerably less than the full cost they would incur if they had to pay for all of their allowances. The theorized moderation of wholesale market prices could act to constrain consumer costs, which could be viewed as beneficial but would mute the price signal. Regardless, we do not rely on such an outcome in endorsing the fuel-differentiated output-based allocation approach for deliverers.

The fuel-differentiated output-based approach would not provide as much certainty to individual deliverers as an historical emissions-based approach regarding the number of allowances that they could expect, since a deliverer's proportional allocation would depend on both the level and fuel mix of its own deliveries and the level and fuel mix of electricity produced by other deliverers. However, in light of the limited time (four years) that we recommend for distributions to deliverers, deliverers should be able to estimate likely distribution levels adequately.

A central rationale for utilizing a fuel-differentiated output-based approach is to avoid undue economic harm to California electricity consumers whose retail providers are currently locked into a certain degree of dependence on coal. This raises the question of whether the higher weighting factor to be used in determining allowance distributions for coal-fired electricity should apply to all coal deliveries or should be restricted to only electricity from coal plants owned or under long-term contract to California retail providers. The concern is that the higher allocation rate might provide incentives for additional short-term deliveries of coal-fired electricity or for coal-fired generation that was previously sold on an unspecified basis to sell on a specified basis instead, in order to receive the higher number of allowances for coal. We recommend that the higher weighting factor be applied for all coal generation delivered to the California grid. Any generation that reports as specified coal would also have a higher per-MWh compliance obligation than unspecified power. Thus, there would be little to be gained by a short-term deliverer specifying as coal.

In order to implement a fuel-differentiated distribution to deliverers of electricity from emitting sources, additional work will be needed regarding the specific weighting factors to be used for the fuel-differentiated distributions and details on how to update the deliverer-specific output-based proportions used in the distribution process, e.g., the time period to use. A related issue that will require further consideration is whether a small number of allowances should be set aside for new deliverers' first year of operation, as described in Section 5.2.1.2 above.

If, counter to our recommendations regarding auctioning, ARB does not implement 100% auctioning by 2016, an important longer-term goal of deliverer distributions should be to provide strong incentives for GHG reductions. If ARB adopts less auctioning than we recommend (either less than 100% as the ultimate goal, or 100% phased in later than 2016), we recommend that distributions to deliverers transition toward a pure output-based approach, to be reached by 2020 if 100% auctioning is not achieved by that time. A pure output-based approach would be more effective than a fuel-differentiated approach in providing strong incentives to develop lower-emitting resources.

Distributions to Retail Providers

Following similar principles, we recommend that the allocation of allowances to retail providers (with a requirement to sell the allowances at auction) initially be in proportion to the historical emissions of the retail providers' portfolios, transitioning to a 100% sales basis by 2020. Allocating allowances to retail providers based on historical emissions in the initial years would accommodate carbon-intensive retail providers that may face relatively high compliance costs. At the same time, as emphasized by NRDC/UCS, transitioning to a sales basis would provide long-term incentives for retail providers to reduce their reliance on high-emitting generation sources.

We do not recommend at this time that the sales calculation be performed on a "net load" sales basis (excluding large hydro and nuclear), as suggested by staff.  Some parties have raised concerns that a pure sales-based approach, unadjusted to exclude large hydro and nuclear, would distribute allowances to retail providers with non-emitting legacy hydro power and nuclear generation out of proportion to the financial impact of GHG compliance on their customers.  However, we conclude that a transition to allowance allocations made in proportion to unadjusted sales by 2020 would provide strong incentives for increased reliance on all low- and non-emitting resources, including legacy generation, and would not have unacceptable impacts on customers of individual retail providers, based on existing modeling results.  Should further modeling reveal that this allocation approach would result in larger distributional impacts than estimated in this proceeding, we may revise this recommendation to ARB.

Additional work will be needed to implement our recommendations regarding distributions of allowances to retail providers, including how to calculate and update the sales-based proportions used in the distribution process as sales-based distributions are phased in and how to allocate allowances to new retail providers. As discussed in Section 5.4.3, additional work also will be needed to address whether and how allowances should be distributed for verified energy efficiency.

Summary of Recommendations

To summarize, we recommend that auctions of allowances be phased in for the electricity sector, beginning with 20% of allowances in 2012 and reaching 100% in 2016. We recommend that the allowances that are not auctioned be distributed on a fuel-differentiated output basis to deliverers of electricity from emitting generation resources (including unspecified sources). Allowances that are to be auctioned should be distributed to retail providers, with a requirement that they then sell the allowances through a centralized auction undertaken by ARB or its agent. The allowance distributions to retail providers should be made on the basis of historical emissions in 2012, transitioning to a 100% sales basis by 2020.

Figure 5-10 illustrates the potential impacts of these recommendations on the rates of individual retail providers. Because of modeling limitations, the allowance distributions to deliverers are modeled as non-fuel-differentiated output-based distributions to deliverers of electricity from emitting resources. The figure assumes that market clearing prices include 50% of the value of the allowances distributed to deliverers. If a fuel-differentiated output-based allocation to deliverers of electricity from emitting resources, which we recommend be implemented, were modeled, it would show a cost spread among retail providers in the 2012-2015 period somewhat less than indicated in Figure 5-10 with, at the extremes shown in Figure 5-10, high-coal LADWP's costs decreasing and low-coal SMUD's costs increasing somewhat.

Figure 5-10

Estimates of Effects on Average Retail Electricity Rates
Due to Recommendations Regarding Auctioning and
Allowance Distributions to Deliverers and Retail Providers

($/kWh, 2008$)

While stressing that Figure 5-10 is presented for illustrative purposes only, we believe it provides a useful conceptualization of the possible effects of our recommendations to ARB.

We submit our allowance allocation recommendations to ARB as the allocation approach for the electricity sector that we find strikes a reasonable balance among the policy objectives that we have considered here. We recognize that, in contrast to our exclusive focus on the California electricity sector, ARB faces the challenge of deciding how to allocate allowances within California for a multi-sector cap-and-trade program that may be linked to a regional and/or national system. We also recognize that our modeling of the impacts of these allocation recommendations has limitations, as discussed above. Additionally, ARB will have to analyze any allocation methodologies that it considers in light of its interpretation of the specific statutory guidance in AB 32.

5.4.3. Should Allowances be Allocated to Support Emission Reduction Measures?

In this section we consider the proposals by some parties that allowances or auction revenues should be allocated as an incentive for certain activities that contribute to reducing GHG emissions. These proposals have in common the deliberate distribution of free allowances on the basis that the activities are either non-emitting (energy efficiency and renewable energy) or lower emitting than certain other sources of energy (CHP). Thus, these allocation methods would serve to encourage energy sources or measures that avoid or reduce emissions, and thus help to meet an emissions cap. Underlying these proposals is the belief that additional incentives may be needed because the GHG cap-and-trade market and other available incentives may not achieve the cap with an optimal mix of energy efficiency, renewables, and other low-carbon ways to meet energy needs.

Both the Public Utilities Commission and the Energy Commission have long supported the development of renewable energy, CHP, and energy efficiency to meet California's energy needs, and California has been a national leader in the development of these resources. All three sources have contributed substantially to reducing California's GHG emissions and, as the Energy Action Plan and ARB's Draft Scoping Plan indicate, the State is counting on all three sources to play a central role in meeting the State's future energy needs and the 2020 GHG cap. However, we are not prepared at this time to endorse any proposals to distribute free GHG allowances as an explicit incentive mechanism for these sources.

Several questions need additional analysis before we can definitively recommend any such proposals. A decision to distribute free allowances preferentially to certain activities should not be undertaken lightly, because such preferential treatment may skew the market with unintended consequences and may divert allowance value from other, potentially more valuable uses. Before we can determine whether to make this choice, two basic questions must be answered for each of these resources: (1) whether additional incentives are needed and (2) if so, whether the distribution of free GHG allowances is an effective and appropriate way of providing such incentives. The record in this proceeding has not been adequately developed to answer these questions. Below, we discuss some issues pertaining to two proposals that have been raised in this proceeding: allocation to retail providers for achieved energy efficiency and allocation to renewable energy producers for MWhs delivered. We also provide some preliminary guidance on the additional analysis required before a decision can be made. Allowance allocations to CHP are discussed in Section 6.

5.4.3.1. Energy Efficiency

Allocating allowances to retail providers on a sales basis that includes verified energy efficiency savings has been advocated by PG&E, NRDC/UCS, DRA, SMUD, and SDG&E/SoCalGas. These parties contend that any sales-based allocation of allowances to retail providers that does not include energy efficiency would deter energy efficiency savings because it would reduce the distribution of allowances to the retail provider for every megawatt-hour saved. In their view, allocating allowances for verified energy efficiency would help foster the development of feasible and cost-effective energy efficiency.

However, several questions remain about the desirability of allocating allowances on the basis of energy efficiency that have not been adequately addressed in this proceeding. SCE argues that, since generator bids are expected to internalize GHG costs, the higher energy prices in a cap-and-trade system would encourage additional energy efficiency automatically and no special treatment is necessary. AReM argues that allocating allowances to retail providers for verified energy efficiency would be unfair to ESPs. There are also uncertainties about how free allowance allocations would interact with existing energy efficiency mandates and incentives, and whether verified energy efficiency should receive allowances at the same rate as actual sales or be weighted less than actual sales. We also would want to ensure that all retail providers are held to consistent verification standards. We intend to consider these issues further, to ensure that allowance distribution policies do not impede achievement of cost-effective energy efficiency, and may make further recommendations to ARB at a later date.

5.4.3.2. Renewable Energy

Several parties support the allocation of allowances to deliverers of renewable electricity, including Solar Alliance, CRA, and SMUD. In Section 5.4.2.4. above, we recommend that deliverers of electricity from emitting generation resources receive allowances on a fuel-differentiated output basis, to be phased out by 2016. Deliverers of electricity from renewable sources that emit GHG would be eligible for such distributions, whereas deliverers of electricity from non-emitting sources would not receive allowances. In this section, we address whether there should be additional allowances distributed to or set aside for deliverers of renewable electricity to provide incentives for renewables development.

There are two issues to consider regarding the desirability of allocating allowances for deliverers of non-emitting renewable energy: the competitiveness of renewables in the market and the need for incentives for the voluntary renewables market to contribute to GHG emission reductions. We address the competitiveness concerns first.

A cap-and-trade program with an allowance allocation method that internalizes emission costs in wholesale electricity prices inherently enhances the competitiveness of renewables. Either historical emissions-based allocations to deliverers or auctioning would have this effect. However, output-based allocation to deliverers may suppress the pass-through of GHG costs in wholesale prices. To the extent that wholesale prices do not reflect GHG costs, the market would not bestow to renewables the full advantage of their lower GHG emissions. Based on the assumption that GHG costs would not be reflected fully in market prices with an output-based allocation of allowances, the Resources for the Future study of RGGI implementation attached to the staff allocation paper concluded that output-based allocations restricted to emitting sources would result in less addition of renewables than either auctioning or historical emissions-based allocations to sources.

Since we recommend that most allowances in the electricity sector be distributed initially through a fuel-differentiated output-based allocation to deliverers, an argument could be made that some complementary allocation of allowances to renewable sources may be desirable to avoid inadvertently disadvantaging those sources in the market. However, given our recommendation to rapidly transition the allocation method to 100% auctioning, any potentially deleterious effect on the competitiveness of renewables would be short-lived. This fact, coupled with the State's current, and potentially increasing, mandates for development of renewables, leads us to question whether including renewables in fuel-differentiated output-based allocations would be warranted. As discussed in Section 5.4.2.4. above, if the transition to full auctioning does not occur by 2016, we would support a transition to pure output-based allocations of allowances, which would include deliverers of renewable electricity.

The distribution of free allowances for renewables participating in the voluntary market potentially could serve another purpose. Currently, buyers in the voluntary market pay a premium for renewable electricity (or the RECs representing that electricity) for various environmental reasons: to be sustainable, to be carbon-neutral, to promote energy independence, or to contribute to reducing emissions of GHG and other pollutants. Once pollutants in the electricity sector are subject to a cap, purchases of voluntary renewables do not contribute to further reductions because the cap determines the allowable levels of emissions. In other words, once a cap is instituted, new renewables would not reduce emissions; instead, the replacement of fossil-based generation by renewables would free up allowances to be used elsewhere in the capped sectors. Solar Alliance characterizes this scenario as allowing fossil generators to free-ride on the emission reduction activities of others.

In order to allow the voluntary market to continue contributing to emission reductions, Solar Alliance recommends the creation of a set-aside of allowances for the voluntary market. Rather than sell the allowances, ARB could retire allowances from the set-aside reserve at some rate for each MWh sold (or REC retired) in the voluntary market. By this mechanism, voluntary purchases of renewable energy would reduce emissions essentially by ratcheting down the cap: ARB would retire allowances rather than issue them for use by an emitting source. Solar Alliance expresses concern that the voluntary market would collapse without a set-aside.

Currently, we do not have enough information to determine the desirability of allowance set-asides for the voluntary renewable market. We certainly do not want to damage the opportunity for voluntary contributions to GHG reductions. AB 32 directs ARB to "adopt rules and regulations...to achieve the maximum technologically feasible and cost-effective greenhouse gas emission reductions..." (Section 38560). As part this effort, AB 32 directs ARB to "identify opportunities for emission reductions measures from all verifiable and enforceable voluntary actions. . . " (Section 38561(f).) AB 32 also directs ARB to "adopt methodologies for the quantification of voluntary greenhouse gas emission reductions. . ." (Section 38571.)

While we support continuing opportunities for voluntary reductions, consistent with the cited provisions of AB 32, we do not recommend the creation of a set-aside for the voluntary market at this time. A number of questions would need to be answered about the design of the cap-and-trade market and the RPS compliance market that may include provisions for RECs. We would need to investigate the types of RECs that would count under a set-aside, including whether RECs from capped and uncapped electricity markets should count. In addition, we would need to investigate how to assign emission reduction values to the RECs that would be counted. These issues will be further complicated in a regional cap-and-trade system. For all of these reasons, we need further investigation and analysis before recommending a set-aside for the voluntary renewables market.

5.5. Use of Auction Proceeds

In supporting some amount of auctioning in D.08-03-018, we cautioned that:

As an integral part of this recommendation, we conclude that the proceeds from the auction of allowances for the electricity sector should be used primarily to benefit electricity consumers in California in some manner, in order to minimize costs of GHG emission reductions to consumers and assist with emissions reduction opportunities. Possibilities include use to augment investments in energy efficiency and renewable power or to maintain affordable electricity rates. Allocating the value of allowances and/or auction revenues primarily to benefit consumers recognizes the importance of electricity as a vital commodity. Thus, we believe that reservation of allowances or allowance value for consumers in this sector is warranted regardless of what may be done for other sectors. (D.08-03-018 at 98-99.)

We address the use of auction revenues in further detail in this section.

5.5.1. Positions of the Parties

Purposes Related to AB 32

Most parties commenting on this issue support the policy we articulated in D.08-03-018 regarding the use of auction revenues. Several parties specifically support the use of auction revenues to fund energy efficiency, renewable energy, and research and development activities, as well as to maintain affordable electricity rates. NRDC/UCS recommend further that such investments be subject to oversight and verification that the investments meet appropriate criteria, with forfeiture of the revenues to the State if a retail provider does not use the revenues in appropriate ways and within a specified time limit. Dynegy stresses its view that the expenditure of auction revenues must not advantage investor-owned utilities relative to independent power producers.

Several parties (PG&E, SDG&E/SoCalGas, SMUD, IEP, GPI, WPTF, NRDC/UCS, and FPL) support using auction revenue to support energy efficiency and renewable development programs. SMUD supports this use of auction revenue as a way to reduce electricity rates. GPI submits that all revenues raised by auctions and through its proposed direct sales of allowances to deliverers at predetermined prices should be used to invest in new, zero-emitting generating resources and efficiency, in order to benefit consumers by providing the infrastructure needed for living in a carbon-constrained world. PG&E submits that, to the extent that auction revenues are used to fund energy efficiency and renewables programs that are currently funded in utility rates, this funding source should reduce current funding needs for these programs in order to avoid double counting.

PG&E states that auction revenue could be dedicated toward utility procurement and development of carbon-free technologies, if targeted toward applied technologies most likely to benefit California's electricity consumers directly. PG&E suggests tax credits, rebates, or incentives to energy users or producers for demonstration of new technologies or applied research, but not grants or pure research, in order to focus on the development of new, commercially-available "green" technologies for the benefit of utility customers. EPUC/CAC submit that any auction revenues, whether retained in the electricity sector or employed on an economy-wide basis, should be targeted to the development and deployment of GHG reduction technologies, and that any programs encouraging technology development should be made available to all potential competitors on an equal basis. IEP asserts that, in the first five years, 50% of auction revenues should be directed to renewable investment, 30% toward clean or low-emitting alternative resources such as clean coal or low-emitting natural gas, and 20% toward energy efficiency not otherwise covered by building and appliance standards and other existing requirements.

Many parties consider supporting consumer cost reductions to be a priority. However, parties differ in their approaches to providing auction revenues to customers.

Some parties (EPUC/CAC and AReM) favor using auction revenue to reduce customer electricity rates. ICC argues for applying auction revenue to reduce the revenue requirement of retail providers in a manner that does not shift costs among customer classes.

Several parties (PG&E, WPTF, FPL, Morgan Stanley, Powerex, CARE, Dynegy, GPI, Calpine, ICC, SCE, and Powerex) recommend that the value of allowances used to mitigate customer costs be applied in a way that preserves a carbon-based price signal. Dynegy and FPL oppose the use of auction revenues for general ratepayer assistance, arguing that ratepayers should not be insulated completely from the costs of GHG reductions and that auction revenues should not be used to dampen the price signals associated with GHG costs. PG&E, WPTF, Morgan Stanley, and CARE all suggest that any direct bill reductions be designed in a way, such as periodic bill credits or refunds, that is not tied to the volume of electricity used, in order to preserve the price signal benefits of a cap-and-trade program.

SCE and SDG&E/SoCalGas submit that the distribution of allowances or auction revenue rights to retail providers should be used to mitigate increases in the revenue requirement resulting from a GHG emissions cap. SCE maintains, however, that precise distribution is best determined by the Public Utilities Commission during an investor-owned utility's cost recovery proceedings. SDG&E/SoCalGas suggest that a reduction in overall revenue requirements would retain the flexibility to use revenues to pay for existing GHG measures or to benefit one rate classification or another. They maintain that the "use it or lose it" requirement that NRDC/UCS propose would be impractical to implement, foreseeing that such an approach would be hampered by rules for carry-over spending and arguments about how much of the capital cost for rate-based investments in renewables, photovoltaics, demand response, and CHP should be counted for GHG reduction versus electricity supply.

Targeting auction revenue toward low-income households was advocated by Dynegy, TURN, PG&E, SDG&E/SoCalGas, and Powerex. While TURN continues to oppose including the electricity sector in a multi-sector cap-and-trade system, it states that it could support the use of a capped system if all, or almost all, allowances are auctioned and the proceeds allocated to retail providers to benefit lower-income customers and to offset the costs of emissions reductions in the electricity sector. NRDC/UCS would support programs that reduce costs to consumers, particularly low-income consumers, for example, by supplementing funding for existing low-income energy efficiency and bill assistance programs, and also would support providing economic opportunities for low-income and disadvantaged communities. Dynegy supports the use of auction revenues to provide assistance to low-income customers, to offset that portion of those customers' bills associated with GHG programs.

WPTF, NRDC/UCS, and FPL believe that consumer interests would be served better by dedicating a substantial portion, if not all, of the auction revenues to specific programs that develop and deploy GHG control technologies, rather than providing direct or indirect short-term rate relief.

Use for Purposes Other than AB 32

PG&E, DRA, and NRDC/UCS are concerned that use of auction revenues for purposes unrelated to AB 32 could be construed as a tax, which they say is not authorized by AB 32 and would require approval by a two-thirds vote of the Legislature. NRDC/UCS argue that deposit of auction revenues in the General Fund to be used for any purpose that is not reasonably related to the purposes of AB 32 would be considered a tax. SDG&E/SoCalGas submit likewise that placement of auction funds in the State's General Fund could conceivably be challenged as a new tax.

5.5.2. Discussion

We addressed the use of auction revenues in D.08-03-018, recommending that proceeds from the auction of allowances allocated to the electricity sector be used primarily to benefit electricity consumers, either by supporting activities that reduce GHG emissions or by reducing the rate impact to California electricity consumers. We reiterate and refine that recommendation herein.

Most parties voice support for using auction proceeds in the electricity sector for purposes related to AB 32. Almost all parties agree that a portion of the auction revenues should be spent on energy efficiency and renewables. Some also recommend that auction revenues be used to support carbon-reducing infrastructure technologies. Parties comment on whether general bill relief should be implemented in a way that mutes the price signal, and whether any bill relief should be limited to low-income consumers. Other recommendations address the following:

· The type of rate relief, e.g., to low-income ratepayers and/or through rebates rather than usage rate decreases;

· The types of investments, e.g., a preference for applied/commercially proven technologies and applied research, compared to pure research and technology development; and

· Whether ARB should adopt a "use it or lose it" policy for retail provider uses of auction revenues.

We continue to support the development of energy efficiency and renewable energy, as articulated in the Energy Action Plan 2008 Update. We believe that retail providers receiving auction revenues should be required to spend such proceeds in a manner consistent with the Energy Action Plan loading order and the goals of AB 32. To meet the goals of AB 32, California is preparing to implement the most ambitious energy efficiency programs in the world. Meeting the targets for the electricity sector outlined in ARB's Draft Scoping Plan will require significant additional expenditures on energy efficiency measures.

California investor-owned utilities currently have sufficient renewable electricity under contract and in negotiation to deliver 20% of their electricity from renewable sources soon after 2010. California's support of renewable energy through the RPS and California Solar Initiative programs demonstrate that renewables can supply a large share of California's energy needs. The Draft Scoping Plan recommends that the State adopt a mandate of 33% electricity from renewable sources by 2020. Bringing that level of new renewables online will require substantial expenditures by California electricity consumers.

For these reasons, and to meet the emission reduction goals in AB 32 through a variety of means, it is critical that California's retail providers devote auction revenues toward cost-effective means of complying with AB 32. While most parties are in general agreement on this point, parties have differing options regarding the degree of oversight that should be applied to the use of the auction proceeds. Parties offer several suggestions about how the funds should used as well as what roles the Commissions and ARB should play in directing the use of those funds. Some parties appear to suggest that ARB mandate with considerable specificity the use that retail providers may make of auction revenues, whereas other parties recommend that the regulatory bodies, e.g., the Public Utilities Commission for investor-owned utilities, oversee the use of auction revenues.

We agree with parties that all auction revenues should be used for purposes related to AB 32. Such a requirement would further the goals of AB 32 and avoid the questions raised about the legality of use of auction proceeds for other purposes. In our view, the scope of permissible uses should be limited to direct steps aimed at reducing GHG emissions and also bill relief to the extent that the GHG program leads to increased utility costs and wholesale price increases. It is imperative, however, that any mechanism implemented to provide bill relief be designed so as not to dampen the price signal resulting from the cap-and-trade program.

We believe that it may be appropriate for ARB to retain a small portion of allowances for the electricity sector, to be owned by the State, in order to use the related auction revenues for statewide electricity-related purposes consistent with AB 32. With that possible exception, ARB should distribute all electricity sector allowances to be auctioned directly to retail providers, in a manner that we discuss in Section 5.4.2. The retail providers would then be required to sell the distributed allowances through a centralized auction, as we describe in Section 5.3. We recommend that all auction revenues from allowances allocated to the electricity sector, whether owned by the retail providers or resulting from the sale of allowances that ARB has retained, be used to finance investments in energy efficiency and renewable energy or for bill relief, especially for low-income customers.

Subject to this directive, the loading order and other statutory and ARB guidance, the Public Utilities Commission for load serving entities and the governing boards for publicly owned utilities should determine the appropriate use of retail providers' auction revenues. The Energy Commission should have broad review authority of publicly-owned utilities' expenditures, with the publicly-owned utilities required to demonstrate annually to the Energy Commission that their expenditures of auction revenues during the prior year were consistent with the requirements outlined herein. While we do not today adopt the "use it or lose it" approach advocated by NRDC/UCS, we recommend that ARB, in consultation with the Public Utilities Commission and the Energy Commission, specify that free distribution of allowances to each retail provider will be conditioned on a demonstration of adequate progress in complying with energy efficiency and renewable energy procurement targets established for the retail provider.

An alternative method for distributing allowance auction revenue has been proposed, in which all California residents would receive annual dividends funded by allowance auction revenues. A GHG cap-and-trade program is expected to increase the cost of energy throughout the capped sectors, and dividends would serve to mitigate the impacts of this cost increase on consumers. The dividend level could be constant for all consumers, or could be based on the proportional economic impact to consumers (with lower-income Californians perhaps receiving higher dividends), but would not be based on the level of energy used. This would preserve the price signal for consumers to reduce their energy use, since by reducing energy use they would decrease their costs without affecting their dividend. Payments would be automatic. Such an approach potentially would be similar to the annual dividends received by Alaska residents from oil revenues associated with Alaskan oil leases. While we do not recommend this approach, it may be appropriate for ARB to further explore this policy tool as part of its statewide cap-and-trade design process.

5.6. Legal Issues Related to Allowance Allocation

Several parties raise legal arguments about our recommended point of regulation for the electricity sector, the legality of auctioning allowances, and other matters covered in our prior decisions in this proceeding. These arguments have been raised previously and concern issues that have not been left open for further consideration in this decision. Accordingly, we do not discuss them here.

5.6.1. Issues of Permissibility Pursuant to AB 32

IEP argues that "[w]hile rate reduction is a worthy goal, it is not specifically authorized by AB 32 and it may conflict with the achievement of the goals [of] AB 32; for that reason, its legality is questionable." (IEP Reply Comments at 12.) IEP notes that the paramount purpose of AB 32 is to reduce the emission of GHGs, and argues that a decrease in rates may actually cause an increase in GHG emissions. (IEP Reply Comments at 13.) However, IEP views the goals of AB 32 too narrowly. It ignores, for example, the provision in Section 38561(a) requiring ARB to consult with the Public Utilities Commission and the Energy Commission concerning "the provision of reliable and affordable electrical service" (emphasis added). Furthermore, Section 38562(b)(1),(2) directs ARB to design the regulations "in a manner that is equitable" and to "[e]nsure that activities undertaken to comply with the regulations do not disproportionately impact low-income communities." Thus, the goals of AB 32 include the provision of affordable electricity service and ensuring that there is not a disproportionate impact on low-income communities. Accordingly, using auction revenues to provide bill relief to customers generally, or to low income customers who spend a larger proportion of their incomes on utility services, does further the goals of AB 32, and IEP's assertion that the legality of this use of auction revenues is questionable is without merit.

In its comments on the proposed decision, FPL argues that distributing allowances to deliverers on a fuel-differentiated output basis is biased against lower emitting resources, citing North Carolina v. EPA, 531 F.3d 836 (D.C. Cir., 2008). That case, however, provides no basis for rejecting the use of a fuel-differentiated output basis for distributing allowances under AB 32. In that case, the federal Environmental Protection Agency (EPA) had allocated nitrous oxide emission credits among states using a "fuel adjustment." "Fairness" was the EPA's only reason for adjusting the allocation of credits based on the kind of fuel used to generate electricity. The court concluded that "fairness" was not one of the factors that the EPA was authorized to consider under the federal Clean Air Act, and that in doing so the EPA had violated requirements of that statute. Here we are recommending the distribution of allowances on a fuel-differentiated output basis for reasons of equity and to help assure reasonable rates. As pointed out in the preceding paragraph, AB 32 specifically directs ARB to design the regulations "in a manner that is equitable" and to consult with the Public Utilities Commission and the Energy Commission concerning "the provision of ...affordable electrical service." Thus, allocating California GHG allowances based on the kind of fuel used to generate electricity is consistent with the authorizing statute, AB 32.

Several parties, including PG&E and NCPA, argue, without further explanation, that allocating allowances on the basis of historical emissions fails to further the goal stated in Section 38562(b)(3) to "[e]nsure that entities that have voluntarily reduced their greenhouse gas emissions prior to the implementation of this section receive appropriate credit for early voluntary reductions." We recommend that the distribution of allowances to retail providers should be made initially on the basis of historical emissions. We fail to see why this is inconsistent with the goal of giving credit for early voluntary reductions. The extent to which historical emissions-based distributions to retail providers would recognize voluntary early actions which these retail providers have taken to reduce emissions depends on the base period used. If, for example, the base period used for determining historical emissions were a period immediately prior to the enactment of AB 32, retail providers would be rewarded for any early action they take to reduce emissions after that base period. These retail providers would receive credit for their early action because their allowances would be based on their higher (pre-AB 32 enactment) historical emissions, but they would only need enough allowances to cover a level of emissions that had been reduced by the actions they have taken after enactment of AB 32. The receipt of these additional allowances would reward the retail providers for their voluntary early actions.

PG&E also argues, without citation to any particular provision of AB 32, that the only lawful method of allocating allowances is one under which the GHG compliance costs for high GHG-emitting resources must be paid by the customers who receive the electricity from those high-emitting plants. (PG&E Comments, at 28.) PG&E does not explain how this would be achieved under a deliverer point of regulation, since retail providers buy much of their electricity from others, and the market price for that electricity is set by a number of factors, such that the cost of allowances will not always be passed through. More generally, PG&E appears to argue that a "polluter pays" approach is the only lawful approach. However, there is no provision in AB 32 that requires a "polluter pays" approach. Indeed, as noted earlier in this section, AB 32 requires ARB to balance a number of goals, which sometimes may conflict. (See, e.g., Section 38562(b) and Section 38580(b).) Moreover, under the GHG regulatory system we recommend, the deliverers, not the customers of retail providers, should be considered the polluters. As the program transitions to 100% auction, deliverers will pay for all of their allowances. Thus, the polluters will be paying. The methodology for allocating free allowances to retail providers, for subsequent auction, answers a different question: who will receive the proceeds of the auction. As explained elsewhere in this decision, we have balanced the numerous goals of AB 32 and conclude that our proposal for allocating allowances to retail providers best balances those goals.

5.6.2. Commerce Clause Issues

Parties briefed the issue of whether the allowance allocation methods considered, including the methods proposed in this decision, raise concerns under the "dormant" Commerce Clause. Under the dormant Commerce Clause, a state's law or regulations may be unconstitutional if there is a differential treatment of in-state and out-of-state economic interests that benefits the former and burdens the latter. We have considered the parties' filings and conclude that allocation to deliverers using a fuel-differentiated output-based standard does not violate the Commerce Clause. We also note that this allocation methodology works within the deliverer point of regulation, which we have previously found not to be in violation of the Commerce Clause.

The allocation method we are proposing is facially neutral and does not have a discriminatory purpose or effect. In other words, allocation on a fuel-differentiated output basis does not on its face, or in effect, discriminate against interstate commerce in favor of intrastate commerce, nor is there any purpose or intent to favor intrastate commerce over interstate commerce. The allowances are allocated on a fuel-differentiated output basis alone, whether generation of the electricity occurs in California or elsewhere.

When a state law or regulation is not facially discriminatory and does not have a discriminatory purpose or effect, the courts apply the Pike balancing test. Under Pike, a state enactment "will be upheld unless the burden imposed on [interstate] commerce is clearly excessive in relation to the putative local benefits." (Pike v. Bruce Church, Inc. (1970) 397 U.S. 137,142.) Here, the burdens on interstate commerce, if any, are purely incidental to the local benefits to California of reducing GHG emissions and the impact of global warming. As detailed in D.08-03-018, the benefits to California are clear and well established.

PG&E argues that a fuel-differentiated output-based allocation methodology may create an undue impact on out-of-state generation because fuel type is a non-environmental criterion on which to base allocation, which would have a disproportionate impact on out-of-state generation. (PG&E Comments, p. 33.) PG&E appears to be arguing that there is no relationship between any burden on commerce and local benefit if a fuel-based allocation is used and that the only allocation method that is likely to survive a Commerce Clause challenge is one based solely on the GHG emissions of the regulated entity. We disagree. First, a fuel-based approach relies on an environmental criterion and has a direct relationship to the harms of GHG that AB 32 seeks to reduce. Simply put, certain fuels produce more GHG than other fuels. An allocation of allowances using a fuel-differentiated output-based criterion is a narrowly-tailored solution to a California problem and the burden on interstate commerce, if any, is purely incidental. Second, we note that under a fuel-differentiated output-based allocation coal, which is most often used in out-of-state generation, will receive a more favorable treatment than it would under a pure output-based approach.

Accordingly, we conclude that any burdens on interstate commerce that may result from the implementation of AB 32 under the allocation methods that we recommend to ARB are incidental and not excessive in relationship to the local benefits to California.

We also conclude that the fuel-differentiated output-based allocation methodology does not regulate extraterritorially in violation of the Commerce Clause. A state statute or regulation may be struck down as impermissibly extraterritorial if it regulates commerce that occurs wholly outside the state. The fuel-differentiated output-based allocation methodology is implemented through the deliverer point of regulation and does not reach over the California border and regulate commerce that occurs wholly outside the state.

Additionally, auctioning allowances would not violate the Commerce Clause. Like administrative allocation, auctioning is facially neutral and does not have a discriminatory purpose or effect, and the burden on interstate commerce, if any, is not excessive and is purely incidental to the local benefit. We recommend that auction revenues be used in a manner that will not discriminate against interstate commerce.

Lastly, we find that our recommendation to allocate allowances to retail providers for subsequent auctioning transitioning over time from being based initially on historical emissions of the retail providers' portfolios to being allocated based on sales by 2020 does not violate the Commerce Clause. It is facially neutral and does not have a discriminatory purpose or effect, and the burden on interstate commerce, if any, is not excessive and is purely incidental to the local benefit.

5.6.3. Issues Regarding the Levying of a Tax

Parties have briefed the issue of whether allowance allocation methods, including the methods proposed in this decision, raise concerns about whether they involve the levying of a tax and, therefore, would require approval by a two-thirds vote of the Legislature. Under the California Constitution, Article XIII A, Section 3, a tax can only be enacted by not less than a two-thirds vote of the Legislature. AB 32 was enacted by less than a two-thirds vote of the Legislature. We have considered the parties' filings and conclude that neither allocations nor auctions violate the California Constitution, Article XIII A, Section 3.

There is an important distinction between a tax and a regulatory fee. A regulatory fee does not require a Legislative vote of not less than two-thirds, because it is enacted under a state's traditional police power, not its taxing authority. Taxes are imposed for revenue purposes, while fees are imposed inter alia, to pay for the expenses of a regulatory program or to defray the actual or anticipated adverse effects of the payer's action. (See Sinclair Paint Co. v. State Bd. of Equal., (1997) 15 Cal. 4^th 866, 874-876.) The imposition of such "mitigating effects" fees is designed to deter the undesired conduct and to stimulate alternative behavior or products. (See id. at 877.) Fees must also "bear a reasonable relationship to those adverse effects." (See id. at 870.)

So long as any revenue generated from an allowance allocation option is used to further the purposes and goals of AB 32 and not deposited in the state's General Fund for non-AB 32 uses, and is reasonable in relationship to the adverse effects caused by the corresponding emission of GHGs, there is no levying of a tax. We recommend that all auction revenues be used for purposes related to AB 32. We urge that auction revenues not be used for General Fund purposes.

5.6.4. Other Legal Issues

LADWP argues that Article XIII, Section 19 and Article XVI, Section 6 of the State Constitution may be violated by an allowance allocation option. Article XIII, Section 19 requires that taxes or license charges be imposed on public utilities in the same manner in which they are imposed on private entities. However, LADWP has not shown that the requirement that deliverers of emitting power purchase some allowances at auction would establish "license charges" as that term is used in Article XIII, Section 19 of the State Constitution. Moreover, we recognize that Article XIII, Section 19 "does not release a utility from payments ... required by law for a special privilege.... (CA. Const. art. XIII, Section 19.) Additionally, LADWP's argument that the cost of programmatic measures is an additional tax or license charge that utilities will pay while other sectors will not, and thus is a violation of Article XII, Section 19 of the State Constitution, is unconvincing.

LADWP argues that the requirement for public entities to purchase allowances at auction violates Article XVI, Section 6 of the State Constitution. That section addresses public finances and does not allow the legislature to gift or lend public funds to private entities. LADWP fails to show how a requirement to purchase an allowance constitutes a gift.

³⁵ D.08-03-018, p. 8.

³⁶ Id., at 9.

³⁷ We recognize that ARB may develop a different method of distributing allowances for other covered sectors.

³⁸ D.08-03-018, p. 7.

³⁹ Unless indicated otherwise, citations to statutory Sections refer to California Health and Safety Code sections added by AB 32.

⁴⁰ CAISO Comments, December 3, 2007.

⁴¹ For simplicity, we assume in this example that the independent generators are the deliverers of their electricity to the grid.

⁴² All E3 scenarios in Section 5 are based on the Accelerated Policy Case, including 33% renewables and "high" levels of energy efficiency. They also assume $30/ton allowance costs and no offsets. For simplicity, E3 assumes that the number of allowances allocated to the electricity sector each year matches the level of emissions projected for that year. The E3 auction scenarios also assume that all allowances to be auctioned would be distributed to retail providers, i.e., that ARB does not retain any allowances to be auctioned with the revenues used for other purposes.

⁴³ In the example, the deliverer of zero-emission electricity would receive the same number of free allowances as the coal-based deliverer. The zero-emitting deliverer would have no compliance obligation, whereas the coal-based deliverer would have a compliance obligation twice as large as the number of allowances it received.

⁴⁴ See, Burtraw, D., Palmer, K., and Kahn, D., "Allocation of CO₂ Emissions Allowances in the Regional Greenhouse Gas Cap-and-Trade Program," Resources for the Future Discussion Paper 05-25, June 2005, attached to the April 16, 2008 staff paper on allowance allocation.

⁴⁵ EPUC/CAC submit that the MRTU "contemplates the use of several market power mitigation features that will effectively limit the ability of generators to secure recovery of their costs." They describe that MRTU prices will be subject to a system-wide cap and that MRTU will cap a supplier's bid under certain circumstances.

⁴⁶ We note that the fuel-differentiated output-based approach would provide assistance to the two categories of independent deliverers that have been identified in particular as potentially having difficulty recovering GHG compliance costs: deliverers of relatively high-emitting electricity whose emission rates and thus compliance costs may be larger than reflected in wholesale market prices, and those with existing contracts continuing into the cap-and-trade period without GHG cost recovery provisions. We note further that standard offer contract terms for electricity purchased from Qualifying Facilities are being developed in R.04-04-003/R.04-04-025, and expect that treatment of GHG compliance costs for electricity purchased through standard offers will be considered in that forum.