DEPARTMENT OF JUSTICE GENERAL COUNSEL DIVISION MEMORANDUM. Legality of setting utility rates based upon the tax liability of its parent

Transcription

1 HARDY MYERS Attorney General PETER D. SHEPHERD Deputy Attorney General DEPARTMENT OF JUSTICE GENERAL COUNSEL DIVISION MEMORANDUM DATE: TO: FROM: SUBJECT: Commissioner Baum Commissioner Beyer Commissioner Savage Jason W. Jones, Assistant Attorney General Regulated Utility & Business Section Legality of setting utility rates based upon the tax liability of its parent You have asked for an opinion on the legality of changing the current Commission policy of calculating tax expenses on a stand-alone utility basis to a policy that would calculate tax expenses based upon the taxes actually paid by the utility s parent in a consolidated tax filing. 1 For the reasons set forth below, I conclude that, within the parameters outlined, the Commission has the discretion to change its policy of calculating tax expenses. Specifically, I conclude that the Commission has the discretion to choose a policy of calculating tax expenses so long as its policy is rational, including taking into account the benefits and burdens of its policy, and meets minimum constitutional requirements. Taking into account the benefits and burdens of its policy means that the benefits of consolidated tax savings are given to ratepayers (by reducing the utility s tax allowance) if the customers bore the burden of paying the deductible expenses that generated the savings. If the Commission matches the benefits and burdens in some rational manner, I conclude that the Commission s choice would meet legal requirements. As a starting point, it is important to keep in mind that ratemaking is a state legislative function that has been delegated to the Commission. As such, the Commission s decisions must be consistent with Oregon law and must pass constitutional muster. In Oregon, rates must be fair and reasonable and are set based on the utility s cost-ofservice. See generally ORS Initially, the actual taxes paid doctrine appears to represent the utility s cost-of-service because it provides that rates will contain an estimate of actual taxes paid. However, the analysis becomes problematic when a consolidated tax structure is contemplated. In that situation, the consolidated tax liability may be strikingly different than a 1 Generically, this can be referred to as the actual taxes paid doctrine. There are many possible deviants of the doctrine; however, within this memorandum it is discussed in the context of adjusting future rates to reflect estimated taxes to be actually paid by the utility or the utility s parent Court Street NE, Salem, OR Telephone: (503) Fax: (503) TTY: (503)

2 Page 2 utility s stand-alone tax liability. 2 While a utility, or other party, may fairly argue that the unregulated parent s costs do not accurately reflect the cost-of-service of the utility, I conclude that the Commission has the authority to consider the parent s tax expenses in calculating the utility s cost-of-service, but only if such authority is exercised within constitutional parameters and is based upon an application of a rational approach, which takes into account the benefits and burdens of the utility s participation in the consolidated tax group. Constitutional requirements There are two Supreme Court cases that are applicable in determining the parameters of the Commission s authority to change its policy of calculating tax expenses. If this Commission were to change its policy on calculating tax expenses, it must be done in a way that is consistent with constitutional requirements. In Federal Power Commission v. Hope Natural Gas Pipeline, 320 US 591 (1944), the Supreme Court recognized that it is the Commission s role to set rates that provide for adequate revenue to pay the reasonable expenses and reasonable capital costs of a utility and that courts review whether the end result, as opposed to a review of each expense or cost, is fair. In 2001, the Oregon Legislature put the Hope standard, which was already being applied in Oregon, into state statute when it amended ORS Subsection (1) of that statute, in relevant part, provides: The Commission shall balance the interests of the utility investor and the consumer in establishing fair and reasonable rates. Rates are fair and reasonable for purposes of this subsection if the rates provide adequate revenue both for the operating expenses of the public utility or telecommunications utility and for capital costs of the utility, with a return to the equity holder that is: (a) Commensurate with the return on investments in other enterprises having corresponding risks; and (b) Sufficient to ensure confidence in the financial integrity of the utility, allowing the utility to maintain its credit and attract capital. It is possible to imagine circumstances where the Commission s treatment of tax expenses could result in rates that violate the Hope standard. For example, if a utility s taxes are estimated at $200 million but the utility has such a bad earnings year that it pays no taxes, requiring the utility to flow-through actual tax expenses and reduce its future rates by $200 million, may create a situation that would violate the Hope standard because the utility would be unable to pay reasonable expenses and attract capital on reasonable terms. In Duquesne Light Co. v. Barasch, 488 US 299, 315 (1989), the Supreme Court reaffirmed the Hope standard and also cautioned state regulators that decisions to arbitrarily 2 In fact, this is illustrated by the PGE and Enron tax situation. PGE s taxes were estimated based upon PGE s costof-service and then sent to Enron, which filed a consolidated return that we assume resulted in dramatically lower tax liability for Enron as compared with the tax payments PGE sent to Enron.

3 Page 3 switch back and forth between methodologies in a way which required investors to bear the risk of bad investments at some times while denying them the benefit of good investments at others would raise serious constitutional questions. Id. As a result, if the Commission wants to change its stand-alone policy of calculating tax expense, it must detail a rational explanation for the switch in policy. Specifically, the Commission must demonstrate why the change in policy will do a better job of estimating fair and reasonable rates than its current approach of using a utility s stand-alone tax liability to set rates in order to be consistent with administrative law requirements. See generally Market Transport, LTD. v. Mauldin, 301 Or 727 (1986). Any change in policy must be consistent with Oregon law and constitutional requirements. A policy that considers the benefits and burdens of the utility s participation in a consolidated tax group, consistent with the Hope standard, 3 would result in decision that would be consistent with Duquesne and administrative law requirements. Some background related to the historical treatment of utility tax expenses One can better follow the advice I provide in this memo by first understanding how regulators normalize the tax benefits of accelerated depreciation for ratemaking purposes. Understanding the rationale for normalization is important because the same reasoning is appropriate for the treatment of consolidated taxes. In the early 1980s, Congress enacted an amendment to the Internal Revenue Code ( Code ) that clarified what is referred to today as the consistency requirement. In the years preceding the amendment, the Federal Energy Regulatory Commission ( FERC ) began developing the law of consistency through its cases and as part of the full meaning of normalization. See Southern California Edison Co, 59 F.P.C. 341 (1972). Normalization, under the Code, provides that a utility company should not compute its expenses for ratemaking purposes using its actual or expected tax expenses. Instead, it should calculate a new regulatory tax expense that reflects regulatory tax deductions that are based on the ratemaking treatment of the underlying asset, e.g., straight-line instead of accelerated depreciation Normalization follows from the consistency requirement it matches regulatory treatment of a tax benefit with the regulatory treatment of the underlying expense item that gave rise to that benefit, thereby treating an expense and its resulting tax deduction in a consistent manner. See generally Note: Deconstructing and Reconstructing Consolidated Tax Savings for Public Utilities, 12 Va. Tax Rev. 735 (1993). In sum, to receive the accelerated depreciation benefit intended by Congress a utility must follow the consistency requirement of matching the regulatory treatment of a tax benefit with the regulatory treatment of the underlying expense item that gave rise to that benefit, regardless of which methodology it uses to estimate tax expenses. 3 As discussed above, the Hope standard requires that the Commission set rates designed to allow a utility to recover reasonable expenses and earn a reasonable return on investment that serves customers. I conclude that a rational, symmetrical approach (e.g., matching the benefits and burdens) would meet the Hope standard because it would result in fair rates consistent with Hope. Of course, that is the very point of a rational approach that considers symmetry to reach a fair result that gives the utility the opportunity to recover its reasonable expenses and earn a reasonable return on investment.

4 Page 4 The development of the consistency requirement is also applicable to the treatment of consolidated tax benefits, as it provides a rationale for matching the treatment of the tax benefit to the expense item that gives rise to the benefit. The minority approach: the actual taxes paid doctrine The cases from a minority of the states that reference the actual taxes paid doctrine were mainly decided in the early-to-mid 1980s. For reasons that I will describe in more detail below, the cases applying the actual taxes paid doctrine are both nonbinding authority in Oregon and largely unpersuasive. 4 I conclude that the cases applying the actual taxes paid doctrine are flawed and, therefore, unpersuasive for one overriding, fundamental reason in estimating the taxes of the regulated utilities the cases do not adequately consider the benefits and burdens of the consolidated group, or alternatively stated, the courts do not apply any symmetry in estimating tax expenses. During the 1980s, the Pennsylvania courts issued several decisions that discuss and apply the actual taxes paid doctrine. 5 Although the cases are difficult to understand because they are poorly reasoned, the approach of the Pennsylvania courts can be summarized as: all tax savings arising out of participation in a consolidated return must be recognized in ratemaking; otherwise, the courts would be condoning the inclusion of fictitious expenses in the rates charged to the ratepayers. Initially, this may seem logical and persuasive. However, a deeper analysis clearly demonstrates that Pennsylvania s view, while simplistically pleasing, is unpersuasive and incorrect. The Pennsylvania cases concluded that where a utility realizes tax savings because of its participation in a consolidated return, those savings must be passed on to consumers by adjusting the utility s allowance for tax expense. See Barasch v. Pa. Pub. Util. Comm., 548 A.2d 1310 (Pa. Commw. 1988). How do the Pennsylvania courts state this should be accomplished? It is accomplished by creating an effective tax rate that requires the parent to return a proportionate share of its tax savings to the utility. As discussed previously, the actual taxes paid doctrine is not a flow-through of actual taxes paid. It is a methodology that estimates tax expenses by taking into consideration that taxes are filed as a consolidated group. Thus, Pennsylvania is still estimating tax expenses, just with a different methodology. For example, Pennsylvania develops an effective tax rate to estimate the utility s proportionate share of tax liability and includes that estimate in future rates. The fact remains that it is an estimate of tax liability; it just uses different assumptions to develop the estimate. 4 It is important to note that, other than Supreme Court cases of Hope and Duquesne, none of the cases discussed are binding Oregon law and the Commission is not bound to follow them. However, such cases may be used as persuasive authority. 5 See Barasch v. Pa. Pub. Util. Comm., 507 Pa. 469, 491 A.2d 94, (Pa. 1985); Barasch v. Pa. Pub. Util. Comm., 507 Pa. 561, 493 A.2d 653 (Pa. 1985); Continental Tel. Co. of Pa. v. Pa. Pub. Util. Comm., 120 Pa. Commw. 25, 548 A.2d 344 (Pa. Commw. 1988); Barasch v. Pa. Pub. Util. Comm., 120 Pa. Commw. 292, 548 A.2d 1310 (Pa. Commw. 1988).

5 Page 5 The Pennsylvania courts are incorrect when they state that the actual taxes doctrine prevents fictitious tax expenses. Pennsylvania is simply using a modified effective tax rate to estimate taxes. The tax expense allowed is fictitious only in the sense that it is an estimate just like many operating expenses and is not based on the taxes that the utility will actually pay to the taxing authorities. The flaw with the Pennsylvania approach is not that it follows the actual taxes paid doctrine but that it patently fails to consider aligning the benefits and burdens of the consolidated tax structure. While I have concluded that this Commission could adopt an actual taxes paid doctrine to calculate tax expenses, it must do so in a rational, symmetrical way. This is why the Pennsylvania cases are incorrect they do not even attempt to align the benefits and burdens of the tax treatment. 6 In fact, the Pennsylvania approach does not look at all the consolidated companies in the parent s corporate family. It looks only at those who lost money in recent years. It ignores those that were profitable, except for the purpose of calculating the utility s share of tax savings from the losing companies. One reviewing the Pennsylvania cases will note that the Pennsylvania Public Utility Commission attempted various methods of estimating tax expenses based upon symmetrical considerations. It is the Pennsylvania courts that rejected each of the Pennsylvania Commission s reasoned approaches and determined that it would violate Pennsylvania law to include fictional tax expenses. As we have discussed, however, Pennsylvania does include fictional estimated tax expenses when it uses the actual taxes paid approach, just without the required symmetrical considerations. In fact, it seems that the Pennsylvania courts ignored the difficult and complex, although more correct logic, of the Pennsylvania Commission and, instead, decided the issue based on a lack of understanding of cost-of-service regulation. 7 Other state approaches As noted in Commission Staff s White Paper, Staff conducted an informal survey of other states practices for calculating income taxes for setting rates. 8 The survey found that most state regulators appear to use the stand-alone method to calculate taxes but that a handful of 6 Interestingly, the Pennsylvania court noted that it agreed with the Pennsylvania Commission that the parent could eliminate any consolidated tax savings by the simple expedient of combining all of the nonregulated subsidiaries into one company. Because the combined company, in that case, would have an overall gain, as does the utility subsidiary, there would be no losses available for off-setting when the consolidated return was prepared. See Barasch v. Pa. Pub. Util. Comm., 548 A.2d at For example, the Pennsylvania courts have said that they only allow actual taxes paid to be considered in rate base. See Barasch v. Pa. Pub. Util. Comm., 507 Pa. 469 (Pa. 1985). Pennsylvania s statement is perplexing because taxes are an expense. Although taxes could impact revenue requirement or even possibly rate of return, it should not impact rate base. In Continental Tel. Co. of Pa. v. Pa. Pub. Util. Comm., 548 A.2d 344 (Pa. Commw. 1988), the court was asked to review its earlier decision in light of a private letter ruling from the IRS that concluded its decision would violate normalization requirements. The Pennsylvania court responded by concluding that the private letter ruling was not correctly reasoned. 8 See Treatment of Income Taxes In Utility Ratemaking: A White Paper Prepared for The Oregon Legislative Assembly at 11.

6 Page 6 states used some type of modified stand-alone approach in certain circumstances. 9 In some of these cases, the regulators policy for calculating taxes has survived some level of judicial review. 10 While it appears most states follow the stand-alone approach, a handful of states follow some type of modified approach. The modified approaches are all unique to their respective states, however, each approach, whether through expense or rate of return adjustments, is aimed at capturing savings from consolidated returns. One may wonder if these approaches are consistent with Oregon legal requirements. In three of the states for which we have court decisions, the courts upheld the state regulators policy decision, often with very little discussion. In Indiana, the court overturned the state regulator s decision and required the regulator to make some determination of actual tax liability based upon Indiana case law, but again with very little discussion. In addition, three of the four cases are more than twenty years old and issued during the early-1980s. This minority of states, like the Oregon Commission, has some discretion in choosing a policy of calculating tax expenses so long as the policy is rational and meets constitutional requirements. From the limited information currently available, it is not possible to determine if these states policies would be consistent with legal requirements. To the extent that these states policies do not state a rational basis for their policies, they should be viewed with suspicion. While there may be many considerations regarding whether a regulators policy will survive judicial scrutiny, such as whether the Commission s decision is appealed or what the particular court s deference was to the regulators reasoning, the prudent approach to meet legal requirements is to rationally explain the policy choice, which should contain a discussion of matching the benefits and burdens. The benefits and burdens approach The actual taxes paid cases do not consider City of Charlottesville, Virginia v. FERC, 249 U.S. App. D.C. 236; 774 F.2d 1205 (1985) (City of Charlottesville), which was issued by the 9 Florida uses a modified stand-alone approach in which they conduct a parent-debt adjustment pursuant to Florida Statutes. Tennessee uses a weighted cost of debt times rate base to get the tax deductible interest expense for purposes of setting rates. Vermont recently adopted an effective tax rate policy that purportedly has not yet been applied. Virginia generally applies a stand-alone policy for the determination of income taxes but it has, is some cases, recognized consolidated tax savings through adjustments. West Virginia has a policy that includes a consolidated tax saving adjustment in rate cases. Wisconsin treats taxes on a stand-alone basis, although it has applied a different method for some small telephone companies. Connecticut uses a stand-alone policy for income taxes, but has also incorporated consolidated tax savings if there is a large tax loss caused by subsidiaries. Indiana also calculates income taxes on a stand-alone basis, but with a cost of money adjustment for interest. 10 See Off. of Util. Consumer Counselor v. Ind. Cities Water Corp., 440 N.E. 2d 14 (1982); Gen. Tel. of Fl. v. Fl. Pub. Util. Comm., 446 So. 2d 1063 (1984); Conn. Nat. Gas Corp. v. Pub. Util. Control Auth., 439 A.2d 282 (1981); GTE South Inc. v. AT&T Commun. of Va., 527 S.E. 2d 437 (2000). With the exception of Virginia, where the court upheld the regulator s decision to apply a parent company debt adjustment, the remaining cases are from the early 1980s when the discussion was ongoing about the appropriate policy for calculating taxes.

7 Page 7 United States Court of Appeals for the District of Columbia in The decision, which is authored by Judge Scalia 12, discusses the benefits and burdens test and, while not binding authority, is persuasive and consistent with Oregon law and constitutional requirements. Therefore, I conclude that if the Commission wishes to change its policy on calculation of taxes, it would be prudent to follow the benefits/burdens test. The benefits and burdens test can be simply stated as: the benefits of consolidated tax savings are given to ratepayers (by reducing the utilities tax allowance) if the customers bore the burden of paying the deductible expenses that generated the savings. See City of Charlottesville 774 F.2d at City of Charlottesville v. FERC The City of Charlottesville involved a petition to the D.C. Circuit Court for review of FERC s decision to use a stand-alone methodology to determine the tax allowances included in the cost-of-service. The City of Charlottesville contended that FERC must require the pipelines to share with their ratepayers the tax savings resulting from the use of tax losses of the system s gas supply subsidiaries in the consolidated return. The Court discussed both the actual taxes paid and stand-alone methods for calculating tax expenses. The Court concluded that the use of either method could be reasonable and found no obstacle to FERC s stand-alone method, neither in theory nor in economic reality. The Court then went on to review whether the application of the stand-alone method was reasonable as applied to the facts of that case. The Court asked, does the stand-alone method match the burdens and benefits as applied? The Court concluded that the approach to matching burdens and benefits is a matter of regulatory policy and FERC s approach of allowing deductions made possible by charges to the ratepayers is an entirely reasonable one. Although the Court reviewed the application of the benefits/burdens test to the standalone method, the benefits/burdens test is also applicable to regulators if they wish to follow an actual taxes paid method. In either situation, the regulator has discretion to pick either method so long as it is applied in a way that reflects the utility s cost-of-service, which is sometimes stated as the benefits/burdens test. Here, the utility s cost-of-service means those costs included in setting utility rates. It may include costs of the parent s activities that affect the utility, as well as costs directly related to the provision of utility service. What types of benefits and burdens might the Commission want to consider? 11 Most of the actual taxes paid cases were decided before the City of Charlottesville. The Pennsylvania Commonwealth Court declined to consider the City of Charlottesville because it was not properly raised on appeal. See Barasch v. Pa. Pub. Util. Comm., 120 Pa. Commw. at Although GTE South Inc. v. AT&T Commun. of Va., Inc., 259 Va. 338 was decided after the City of Charlottesville, it does not discuss the decision. 12 Judge Scalia is now a Supreme Court Justice. The opinion was also joined by Judge Robert Bork, who was nominated, but not confirmed, to the Supreme Court and Judge Gasch, sitting by designation from the District Court for the District of Columbia.

8 Page 8 Although the Commission should consider the benefits and burdens of its policy to determine if its policy reflects the utility s cost-of-service, the Commission still has some discretion on the appropriate matching of benefits and burdens so long as it rationally explains its choices and why it finds that its choices reflect the utility s cost-of-service. Although not intended to be exhaustive, I will offer some illustrative examples of issues that the Commission may want to consider in determining the appropriate matching of benefits and burdens to determine the utility s cost-of service. Under current tax law, private companies are free to file consolidated taxes and, as a result, receive benefits as part of a consolidated group. If the Commission were to adopt a policy that passed along consolidated tax savings to ratepayers, it should explain why it is appropriate to disadvantage utilities and their nonregulated affiliates as compared to private consolidated groups. One possibility may be that the Commission can point to the regulated utility as the entity that gives rise to a benefit to the consolidated group; therefore, the ratepayers should be the ones that get the benefits. On the other hand, the Commission may consider that a policy that passes along consolidated savings to the ratepayers places the utility and its affiliates at a competitive disadvantage in the marketplace. For example, consolidated groups with regulated utilities would not be able to take full advantage of the Code s benefits, which are intended to encourage investment by providing tax benefits that make investors willing to risk more money than they otherwise would. As a result, such a policy may create disadvantages for consolidated groups with utilities and restrict their competitiveness in the marketplace. The Commission may also consider whether removing the Code s tax benefits for regulated utilities would also diminish the attractiveness of investment in the utility company. If the Code s tax benefits are removed, equity investors may look to investments that give them the full advantage of the Code s tax benefits. Theoretically, this could create a loss of willing equity investors, which could require the utility to take on additional debt, resulting in higher costs of borrowing (which could lead to higher rates). The Commission may want to consider whether the benefits of lower tax payments outweigh the possibility of higher borrowing costs. The diversification and risk entailed in a consolidated group is generally borne by the shareholders. As a result, if the loss and risk of consolidation is funded by the shareholders, not the ratepayers, then some may argue that to match the benefits and burdens, the tax savings should also belong to the shareholders. For example, consider a hypothetical in which a company, which has debt to service, holds a utility as its only asset. Under current tax law, the company would be allowed to deduct interest payments on its debt from its taxable income. In this hypothetical, if the risks of the debt are borne by the company you may consider that, because ratepayers are not responsible for any of the burdens, the company should keep its tax savings. On the other hand, if you find that ratepayers are taking the burden of the debt, then you could require the company to give ratepayers their proportionate share of the tax savings. That is, if you demonstrate that

9 Page 9 ratepayers are pay higher rates because of the company s debt, customers should get their proportionate share of tax savings resulting from interest deductions resulting from the debt. To take another practical example, consider the situation with PGE and Enron. PGE s rates contained approximately $90 million in estimated federal and state income taxes. Enron, because of substantial losses in other businesses entirely unrelated to PGE, paid no taxes. However, in this case it was Enron s other business risks that created the tax savings. PGE ratepayers did not assume any of the burdens of Enron s other business endeavors, such as a higher cost of capital. Therefore, one could conclude that because PGE s ratepayers did not take on any of the burdens that created the tax savings they should not get the windfall of benefits. On the other hand, if PGE suffered losses that lowered the tax liability of the consolidated group you could conclude that the ratepayers should receive the benefit of those lower taxes. If the ratepayers are taking the risk of the parent, ratepayers should also get the benefit of those burdens. However, if the ratepayers are not taking any risk of the parent, they should not get the benefit of tax savings resulting from the risk the parent, not the utility, incurred. The preceding examples are only intended to illustrate some of the considerations that the Commission may use when analyzing the symmetry of its tax policy. Ultimately, the Commission should consider and detail a rational explanation that considers the benefits and burdens of its decision. Conclusion The Commission currently has discretion to choose a policy of calculating tax expenses so long as the policy is rational, i.e., meets the benefits and burdens test, and meets minimum constitutional requirements.