Rate-base regulation causes depreciation to be reflected in cost of service both annually and cumulatively—annually through the depreciation expense component and cumulatively through the return and related income taxes component as the annual depreciation expenses accumulate in the book reserve. The cumulative component eventually overwhelms the annual component, thereby causing the initial impact of any depreciation change to reverse in just a few years. This reversal allows the treatment of depreciation (especially the removal cost component) to demonstrate whether regulation emphasizes the near-term or the long-term, thereby providing a basis for judging its fairness.
Removal cost is the expenditure involved with physical removal or safe abandonment of an asset, and is not a trivial matter, because it is not unusual for such expenditures for long-lived property to exceed the related depreciable investment amounts. Those emphasizing the near-term favor removal-cost deferral mechanisms that shift recording and recovery to future ratepayers—a process that increases the costs to be borne by ratepayers through two influences. One influence is the resulting rate-base inflation that increases the total cost of service ratepayers will bear over the life of the assets. The other influence is the inherent increase in regulatory risk, which increases the cost of capital.
The ratable (accrual) treatment of removal costs through depreciation for regulatory accounting purposes has a long history, but is challenged periodically by proposals to defer its recording and recovery. Such challenges also have a long history, but have taken on renewed vigor as a consequence of FASB Statement of Financial Accounting Standards No. 143, Accounting for Asset Retirement Obligations (SFAS 143) issued in 2001.
Various treatments of removal costs have various effects on utility ratepayers. Of all the approaches the industry uses, ratable treatment through depreciation minimizes the costs borne by ratepayers.
Challenges to ratable treatment of removal costs through depreciation tend to emphasize the fact that the annual amounts reflected in depreciation rates typically exceed the annual expenditures for removal or abandonment. There is nothing sinister about this situation, as it is the natural consequence of the accrual accounting that is specified by the FERC electric and natural gas Uniform Systems of Accounts and of the existence of growth and cost escalation. However, it has led some regulators to dictate deferral mechanisms for removal costs.
To illustrate the influence of growth and cost escalation, think of retirement dispersion in the form of frequency patterns, such as those seen in Iowa-type curves (see Figure 1). Consider a property group for which there is no growth or cost escalation, and for which the life is constant and described by the S3 dispersion pattern. The average monetary age of retirements climbs the frequency curve as the property group ages—eventually reaching maturity at the maximum frequency that occurs at a retire- ment age equal to the average service life. With no growth or cost escalation, the average age of the retirements of an ongoing mature property group stays at this maximum frequency point and at the age equal to the average service life. This causes the relationships for the scenario without growth and cost escalation, i.e., equality (see Figure 2).
This scenario is helpful for understanding the significance of property group maturity, but isn’t realistic, because life isn’t constant and growth and cost-escalation occur. Therefore, increasing maturity causes the age of retirements to climb the frequency curve (see Figure 1), but growth and cost escalation keep it from reaching an age equal to the average service life, which prevents the equality of the ages and amounts. Growth and cost escalation cause the relationships (see Figure 2) for the scenario with growth and cost escalation, i.e., inequality.
The concerns expressed in regulatory proceedings for current removal cost accruals being larger than the amounts currently being expended, place the blame on incorrectly reflecting the cost escalation. These concerns are misplaced, partly because they are limited only to this relationship, and partly because this situation is the natural consequence of growth, cost escalation, and accrual accounting. There are two aspects to concerns for the influence of cost escalation on removal costs—the rate of escalation and the period of time over which it occurs. The concern expressed is about the rate of escalation, which is the wrong concern, because: 1) depreciation analysts addressed such concerns long ago by recognizing the past rate as being the best estimate of the future rate; and 2) the influence of the period of time is greater than is the influence of the rate. The correct concern is about the period of time for escalation, but is never expressed, because the response demonstrates that depreciation rates should be increased.
Alternative treatments of removal costs have different impacts on regulated entities, ratepayers, and the economic viability of service territories. Therefore, it’s important to understand these impacts. Seven treatment scenarios are addressed here: 1) ratable accrual through straight-line depreciation; 2) recognition of cost escalation as it occurs; 3) liability (SFAS 143) treatment; 4) external funding with constant fund contributions; 5) expensing when incurred; 6) five-year amortization after being expended (generally known as the Pennsylvania method); and 7) capitalizing as a cost of the replacement asset. Straight-line depreciation matches the recording of removal costs with the pattern of usage of the associated asset, and the other scenarios defer to varying degrees the recording of these costs. The analysis is from the standpoint of the ratepayer, as it discloses what ratepayers would pay for service over the asset life under rate-base regulation (cost of service that includes return on rate base as a cost) for each scenario.
The analysis demonstrates that the impact of removal costs on ratepayers varies directly with the extent of the deferral in recording such costs, and is based on $1 of removal cost, useful lives of 40 and 60 years, 40-percent composite income-tax rate, 3-percent cost-escalation rate, 50/50 debt-to-equity ratio, 7-percent debt cost and liability discount rate, 11-percent equity cost, and external fund earnings after taxes and management fees that are 0.5 percentage points less than the after-tax cost of capital of 7.6 percent. The calculations are based on the half-year convention typically utilized for depreciation studies and depreciation-rate calculations, and ignore property taxes. Assumptions made to simplify the calculations are: a single asset, removal immediately upon retirement, tax depreciation equal to book depreciation, and deferral not increasing the risk reflected in the cost of capital (see Figure 3).
The cost-of-service component for return and related income taxes in Figure 3 demonstrates the significance of rate-base regulation, because the return and taxes are the sole cause of the cost differences imposed on ratepayers, except for the External scenario. The rate bases for accrual of removal costs (Ratable, Escalate, and Liability scenarios) start at zero and eventually reach negative $1 as depreciation and accretion provisions accumulate over 40 and 60 years—Ratable on a straight-line and Escalate and Liability at ever increasing amounts. There are no rate bases for the External and Expense scenarios. The rates bases for the Amortize and Capitalize scenarios start at $1 and decrease to zero on a straight-line—amortize over five years; and capitalize over 40 and 60 years. The Ratable and Capitalize scenarios are shown to be the mirror image of each other, and the Escalate and Liability scenarios are shown to be so similar to each other that their positions on the list reverse when the life is 60 years.
A negative rate base produces negative return that is a credit for ratepayer-supplied capital, and a positive rate base produces positive return that is a charge for investor-supplied capital. The Ratable, Escalate, and Liability scenarios demonstrate that for long-lived assets the credit for ratepayer-supplied capital eventually overwhelms the annual effect of charging ratepayers for the removal cost amounts over the life of the related assets.
The six deferral scenarios can be thought of as forcing ratepayers to pay for removal costs multiple times. Each scenario charges for $1 of removal cost. Therefore, for a life of 40 years, Ratable can be viewed as charging for cost of removal once, Escalate and Liability as charging nearly twice, External as charging nearly three times, Expense and Amortize as charging more than three times, and Capitalize as charging more than four times.
These deferral scenarios do not exhaust the possibilities. For example, calculations are not made for present-value treatments other than SFAS 143 (a prepaid annuity), the most common of which is some form of sinking-fund depreciation, because there are too many ways sinking fund can be implemented for regulatory purposes. For sinking-fund depreciation, only the annual annuity amounts are considered as cost of service and the accumulations are not a deduction from rate base. For modified sinking-fund depreciation, both the annual annuity and interest amounts are considered as cost of service and the accumulations are a deduction from rate base. Therefore, the cost of service for sinking fund will be different than for modified sinking fund, and both will vary depending upon the relationship between the allowed cost of capital and the selected annuity rate and whether the annual annuity amounts are constant. The costs imposed on ratepayers by forms of present value other than SFAS 143 likely would fall in the range defined by the Liability and External scenarios, and so would be the equivalent of charging for removal costs two or three times.
Deferral scenarios damage financial viability to varying degrees by denying the entity an internal source of capital during the lifetime of removed assets, thereby increasing the entity's borrowing and the cost of capital that regulation dictates should be borne by ratepayers. These calculations ignore this additional cost, which, if recognized, would cause the extra cost of service imposed by the six deferral scenarios to increase.
An additional adverse aspect of removal cost deferral is a consequence of the fairness of regulation being a leading indicator of the business climate. Such deferrals are characteristics of regulatory decisions that emphasize near-term political considerations over long-term economic considerations. The financial community and large energy users can be expected to interpret the resulting unfairness as signaling deterioration of the business climate. The financial community might react by downgrading the securities of the entities subject to the jurisdiction of the state and of the state itself. Large energy users typically have multiple locations, and so can shift production between locations in reaction to regulatory decisions. The participation of large energy users in regulatory proceedings historically has emphasized the long-term by addressing cost-allocation (fairness) issues, rather than magnitude of cost-of-service issues. However, in recent years some large users have shifted to emphasizing the near-term in their participation in regulatory proceedings, suggesting that they may not be planning on being long-term customers.
Examine what happens over the life of a single asset (see Figure 3). In practice, depreciable property groups are ongoing, so continual additions and retirements cause the position of the age of current retirements on the frequency curves depicted by Figure 1 to become relatively static at an age less than the average service life. This situation causes a deferral to become permanent, thereby allowing only a portion of removal costs to be accrued over the life of the property. The portion of the removal-cost expenditure allowed to be accrued offsets the accumulated amount in the reserve, and the remainder decreases the reserve, which has the same impact on ratepayers as if the shortfall is treated as a construction cost of the replacement asset. Therefore, a weighting calculation using the Ratable and Capitalize scenarios can be used to determine the extra costs to be borne by ratepayers when the property is ongoing.
If the deferral mechanism causes only half of the expected removal cost to be accrued, the remainder will be as de-scribed by the Capitalize scenario, so the combination (50-percent weight to Ratable and 50-percent weight to Capitalize) would result in the extra cost being the same as for the Expense scenario. Less than half being accrued causes the extra cost to be higher than for the Expense scenario, and more than half being accrued causes the extra cost to be lower than for the Expense scenario.
A current Consumers Energy depreciation proceeding in response to a requirement that the six major energy utilities subject to Michigan jurisdiction file proceedings to address alternative treatments for removal costs, provides data that allow estimating the influence of the ongoing nature of property groups on a gas utility. In this proceeding, Consumers Energy and the Michigan Public Service Commission staff propose the Ratable scenario, an intervener proposes the Liability scenario, and another intervener proposes the Expense scenario. For the Liability scenario, the intervener’s removal cost allowance is 47 percent of what Consumers Energy and the staff propose, which leads the Ratable/Capitalize weighting to indicate an extra cost of $2.38 for each dollar of removal cost for 40-year property, as opposed to $0.85 for a single asset, and an extra cost of $3.58 for 60-year property, as opposed to $1.80 for a single asset. For the Expense scenario, the intervener’s removal cost allowance is 19 percent of what Consumers Energy and the staff propose, which leads the Ratable/Capitalize weighting to indicate an extra cost of $3.65 for each dollar of removal cost for 40-year property, as opposed to $2.25 for a single asset, and an extra cost of $5.47 for 60-year property, as opposed to $3.38 for a single asset.
These weighting calculations presume that regulatory decisions to require deferral mechanisms never are reversed. If later reversed in response to recognition of their detrimental impact on ratepayers, how much the reversal will save future ratepayers will depend upon how long the mechanism was in place and how quickly the accumulated shortfall is allowed to be recovered. For example, if reversed at the time of the removal expenditure, the Amortize scenario is substituted for the Capitalize scenario in the Consumers Energy weighting formulas, which for deferral through the Liability scenario would reduce rate-payer payments for each dollar of re-moval cost by $1.04 for 40-year prop- erty and by $1.64 for 60-year property, and for the Expense scenario would reduce ratepayer payments by $1.60 for 40-year property and by $2.51 for 60-year property.
These calculations per dollar of removal cost can be related to a real-world situation by multiplying by $3.5 billion the amounts shown here for 60-year property, because one of the interveners in the previously-mentioned Michigan proceeding asserts that Consumers Energy estimates its future removal cost expenditures for its $3 billion of depreciable gas property will be $3.5 billion, and because the other intervener asserts that Consumers estimates the average age of the property upon retirement will be 57 years.
Another aspect of deferral mechanisms in practice is that what a mechanism is named or appears to be might not be what it actually is. For example, the Pennsylvania method (the Amortize scenario) is not well understood, which leads to cash treatment (the Expense scenario) sometimes being referred to as the Pennsylvania method. Another example is that cash treatment often is accomplished by reflecting removal costs in depreciation rates in a manner that causes the rates to appear to be for accrual treatment. This appearance can lead observers to falsely believe that the Ratable scenario is being proposed. These examples demonstrate that a thorough understanding of authorized treatments of removal costs might not be obtainable from regulatory decisions, and so will require delving into the depths of proceedings.
It’s evident that how regulators treat removal costs is an indicator of the fairness of regulation. Treatments other than the Ratable scenario impose substantial extra costs on ratepayers over the life of the property from which they are served. In addition to being detrimental to ratepayers, these extra costs increase future energy prices that will make the efforts of those involved with area development more difficult, especially when an energy-intensive business is being courted.