concurring in part and dissenting in part:
While I concur in the result reached on many of the issues in this appeal — and in fact authored, but for a number of selective changes, most of those portions of the court’s per curiam opinion with which I do not disagree — I believe the majority is patently wrong in sanctioning the enormously disparate rate design prescribed by the Commission in this proceeding. I also cannot concur in the majority’s affirmance of the “below-the-line” treatment of advertising and certain other administrative expenses. Further, I am convinced that a number of the other issues on review present much closer questions than the majority’s per curiam opinion would appear to acknowledge. I address each of my major areas of disagreement separately, devoting the bulk of this opinion to that area in which I find the error to be the most egregious: rate design.
I. Rate Design
A. Flaws in the Majority Opinion
Although I believe that the Commission's decision with respect to rate design is seriously flawed, the majority’s treatment of the issue on review is even more disturbing because of the method by which it accomplishes affirmance. The Commission’s treatment of the subject is marred by the absence of either cost or non-cost related factors in the record which could justify the discriminatory rates which were prescribed.1 Beyond that, however, the majority reaches far past the record and even beyond the Commission’s orders in a strained and defective attempt to sustain that rate structure. In doing so, the majority not only has misinterpreted, but in addition apparently intentionally has circum*1243vented, fundamental precepts both of our judicial review function and of public utility regulatory law.
To those unfamiliar with the development of this case at the agency level, the majority’s discussion of rate design may appear to constitute a plausible treatise on the various factors which the Commission, in appropriate circumstances and upon an adequate factual record, may take into account in structuring gas rates. The majority writes for pages on a number of cost and non-cost criteria — value of service, energy conservation, historic rate patterns, etc. — which in past proceedings have been invoked legitimately to justify rate disparities among customer classes. In doing so, however, the majority ascribes dispositive significance to several factors with respect to which the Commission made no factual findings and upon which it placed little or no reliance in structuring the rates.
The majority’s discussion of energy conservation and value of service is puzzling. Although the PSC mentioned that those items are among the non-cost factors which it may consider in designing rates, it did not apply those criteria to the facts of this proceeding to seek to justify the rate disparities. The Commission did invoke the concept of value of service to support the relatively high rate for interruptible customers, but the interruptible rate is not challenged by any party on review and the only “finding” made by the Commission which conceivably could be related to energy conservation as a basis for rate discrimination was the broad and unsubstantiated observation that “[t]he basic facts have not changed from Case No. 647 [in which the inherently promotional declining block structure was eliminated] to the present.” Order No. 6051, at 90. The PSC did not even purport to rely to any meaningful extent on value of service and energy conservation in setting the rates at issue.
The majority’s most conspicuous degression, however, is its heavy reliance on a “cost” factor which (1) was not even mentioned by the PSC in Orders No. 6051 and 6060, (2) is not relied upon by the PSC or by People’s Counsel in their appellate briefs, (3) is not supported by the record of these proceedings, and (4) could not, in any event, justify the rate disparities. I refer to the majority’s apparent tenet that the “rising cost of replacement gas” is a phenomenon that may be invoked to uphold any rate structure under which a utility is projected to recover disproportionately high rates of return from large-scale users of gas. It does not matter under the majority’s theory whether there is any specific evidence of such “rising costs” in the record, nor does it matter how great the rate disparities are between customer classes. The extent to which such a result-justifying technique could be taken is as alarming as the concept is unsophisticated.
It is a basic principle of administrative law that a reviewing court may not sustain an agency ruling on grounds which were not relied upon by the agency. In SEC v. Chenery Corp., 318 U.S. 80, 63 S.Ct. 454, 87 L.Ed. 626 (1943), the Supreme Court laid down the broad propositions that “the grounds upon which the administrative agency acted [must] be clearly disclosed and adequately sustained” and that “[t]he grounds upon which an administrative order must be judged are those upon which the record discloses that its action was based.” Id., at 87, 94, 63 S.Ct. at 459, 462. See 2 K. Davis, Admin. Law Treatise § 16.12 (1958) (“The first Chenery case rather clearly establishes the proposition that when an agency gives the wrong reasons for a decision of policy or law, the reviewing court will send the case back for a new determination, even though the court might have upheld the order if no reasons had been assigned.”)2 The standards for reviewing decisions of administrative agencies are, of course, more restricted than *1244those for reviewing actions of trial courts, as a trial court’s decision may be affirmed if correct in result, even though for the wrong reason. Thus, just as a reviewing court may not substitute its judgment for that of an agency in overturning the agency’s decision, neither may a court salvage an otherwise unsustainable administrative ruling by supplying supporting reasons of its own choosing.
That, however, is precisely what the majority has done. There is not a hint in either Order No. 6051 or Order No. 6060 that the PSC relied to any significant extent upon the rising cost of new gas in setting the customer charges. In fact, the Commission went to great lengths to explain why it was rejecting the uniform volumetric rate which had been proposed by People’s Counsel. The proposed volumetric rate ostensibly was based on marginal cost pricing, a pricing scheme which would take into account the costs of incremental quantities of new gas. The Commission stated:
Although exotic gas sources, such as SNG [synthetic natural gas], imported LNG [liquified natural gas], and Alaskan gas, will exceed WGL’s current average costs, their impact will be relatively restricted for years. [Order No. 6051, at 88.]
There also is evidence in the record that gas supplies from existing sources are expected to continue to improve in the near future.
The majority nevertheless concludes that “the inclusion of substantial fixed costs in the commodity charge is cost justified. The high usage customers contribute disproportionately to the exhaustion of relatively inexpensive commodity sources, and thus accelerate future increases in commodity prices for all utility customers.” Ante at 1201 (footnote omitted). The majority’s analysis is squarely at odds with the PSC’s reasoning, which simply and obviously was as follows:
We take this action essentially for equitable reasons.... At least at this time, we feel compelled to provide some degree of relief for existing customers who are in no position to respond to more cost-based rates. Equally important is the Commission’s desire to lessen the extent to which the low-volume gas user pays a higher per unit price for service. [Id., at 97.]
It thus is clear that the Commission was determined to lower customer charges for relatively low-volume gas users simply because of a perceived need for economic relief for those customers. There is nothing in the Commission’s orders indicating that it was concerned with penalizing higher-volume consumers for any increase in replacement gas costs to which they might contribute; in fact, parts of Order No. 6051 clearly refute that theory. For instance, the Commission rejected the Company’s proposal to establish a higher customer charge for high-volume residential heating customers than for lower-volume customers in the same class. Id., at 95-96. The Commission stated:
WGL concedes that there is no real difference in the customer costs based on annual consumption (above or below 100 Ccf [hundred cubic feet] per year). Therefore, we see no reason to prescribe two different customer charges. [Id., at 96.]
The majority’s willingness to supply its own rationale for upholding the PSC’s decision perhaps might be easier to understand if the new rate design actually were going to have the effect the Commission purportedly intended — i.e., rate relief for lower-income gas customers. The fact is, however, that the customer classes with the highest projected rates of return are those with probably the greatest percentage of low-income customers, namely, the commercial heating and non-heating classes, which include all group-metered apartment buildings. These classes are projected to provide WGL with returns of 15.13% and 26.11%, respectively, as compared with returns of only 4.91% and minus 8.42% for the residential heating and non-heating classes, respectively. It is ironic that the residential non-heating customers toward whom the rate relief is targeted are quite likely well-to-do customers who heat their homes with currently more costly oil or electricity. There *1245certainly is no evidence in the record, as I discuss in detail below, that lower-income consumers tend to be concentrated in the residential non-heating class.
Moreover, I think it is unfair and illogical to presume — as the majority evidently does — that commercial customers are more “guilty” than residential customers, either as individuals or as a class, of increasing the price of replacement gas. It does not make sense to penalize a customer class based on its size alone. It is fair, for example, that residents of group-metered apartments (which are included in the commercial classes) should be forced to subsidize residents of individually-metered apartments (which are included in the residential categories) simply because the group-metered building is viewed as one “large” customer while the individually-metered building consists of many “small” ones? Is it reasonable to say that 50 newly-connected residential customers, each of which uses 100 Ccf of gas per year, are less responsible for rising marginal costs than is one hotel using 5,000 Ccf per year which has been a customer for many years? Unquestionably many of WGL’s larger customers initially subscribed to gas service years ago when gas supplies were plentiful and a “declining block” rate structure was utilized; obviously it is not equitable to force those customers to shoulder the entire blame for today’s higher replacement gas costs. In other words, it is necessary to look beyond a customer’s mere label as a “large” customer to determine whether it is contributing disproportionately to rising replacement costs.
The flawed nature of the majority’s reasoning further is evidenced by the projected rates of return within the commercial classes, which show that WGL will earn a significantly higher return on commercial non-heating customers than on commercial heating customers.3 Unquestionably, the larger users of gas are in the latter category; it takes considerably more gas to heat a building than to cook or provide hot water. .Thus, even if the Commission had based the rate disparities on evidence of rising replacement costs — which unquestionably it did not — the resultant rates would not achieve what the majority seems to perceive as the desired goal of recovering the greatest proportion of those costs from the largest customers.
Energy conservation is, to be sure, a valid ratemaking objective, as is the recovery of increasing costs of new gas supplies from those who are shown to be most responsible for their incurrence. However, there is a limit to what the PSC permissibly may do under the general rubric of energy conservation. Here the Commission did not even attempt to use that goal as a predicate for the glaring rate disparities, although it did mention conservation as one factor that may be taken into account. As to the allocation of costs of replacement gas, the Commission made absolutely no findings, so it is totally unwarranted for this court, on judicial review, to affirm the rate design on that basis.
The majority’s treatment of the rate design issue fails, then, in two fundamental respects, as a consequence of which the true merits of the question never are reached. First, the majority relieves the Commission of its “burden of showing fully and clearly why it has taken the particular ratemaking action.” Washington Public Interest Organization v. PSC, D.C.App., 393 A.2d 71, 75 (1978), cert. denied, 444 U.S. 926, 100 S.Ct. 265, 62 L.Ed.2d 182 (1979). There we also stated:
There are two aspects of this elaboration required of the Commission: (1) announcement of the criteria governing the rate determination, and (2) explanation of how the particular rate order reflects application of these criteria to the facts of the case. [Ibid.][4]
*1246Second, no doubt recognizing that the PSC’s decision on rate design could not be sustained under the meager rationale offered by the Commission, the majority nevertheless proceeds to sanction discriminatory rates on a theory that is unsupported in the record and which apparently did not even occur to the two-member Commissioner majority. In so doing, the majority of this division of the court flagrantly violates the principle of judicial review established long ago in SEC v. Chenery Corp., supra: “The grounds upon which an administrative order must be judged are those upon which the record discloses that its action was based.” 318 U.S. at 87, 63 S.Ct. at 459.
Such a misperception of our proper function in reviewing agency action permits the majority to sidestep the central issue in the rate design dispute, which is whether the obvious class-based rate discrimination is explained adequately by the PSC and is supported sufficiently by record evidence of legitimate cost and non-cost related considerations. Having outlined my disagreement with the majority’s approach to this case, I now explain what I believe to be the flaws in the Commission’s treatment of the subject.
B. Flaws in the Commission’s Treatment
As recognized in the per curiam opinion, the Commission departed from the traditional, inherently promotional declining block rate structure in WGL’s most recent prior rate proceeding in 1976. See In re Washington Gas Light Co., 16 P.U.R. 4th 261,278-84 (Formal Case No. 647, D.C. PSC 1976). In its place, the Commission instituted a two-part rate structure consisting of (1) a “system” charge, i.e., a fixed charge (unrelated to volume of consumption) which was designed to recover a significant portion of customer-related costs,5 and (2) a “commodity” charge, i.e., a single per-therm rate for usage which was uniform for all customers regardless of customer class or volume of consumption. See id., at 279. The Commission felt that a uniform commodity charge would encourage conservation of gas. At the same time, the system charge was presumed to allocate the recovery of customer costs equitably among customer classes.6
Although the Commission established separate system charges for each class of firm customers in the 1976 proceeding, it felt constrained to temper the effect of the sudden transition into a revolutionary new rate design. Id., at 280. Thus, despite evidence showing that customer costs were approximately the same for all residential customers, the PSC prescribed substantially unequal annual system charges of $72 for residential heating customers and $45 for residential non-heating customers.7 Id., at *1247280 & n.13. The Commission justified this discrepancy “upon pragmatic as well as theoretical considerations. When a commission ushers in a drastically new rate structure,” the PSC stated, “it has the responsibility— or at least the discretion — to temper radical changes in the historical rate patterns.” Id., at 280; accord, Apartment House Council of Metropolitan Washington, Inc. v. PSC, D.C.App., 332 A.2d 53, 57-58 (1975). Because the residential non-heating class had enjoyed low bills under the old declining block rates, the PSC saw a need to ameliorate the impact of the new rate structure on that class. It therefore felt justified in setting for residential non-heating customers an initial system charge which recovered only about two-thirds of the class’ customer costs, even though that decision meant that other customer classes would have to make up for these unrecovered costs through higher commodity charges. The Commission was careful to note, however, that the transition period would not last forever and that in the future, such a cross-subsidization of one class by another would cease to be appropriate. It thus stated in the earlier case:
Moreover, in a future general rate case, the commission will consider rate structure again. This case is only the first step in a periodic review of system charge levels. Once customers have had a fair opportunity to adjust to the new rate structure, the company can consider system charges which include more cost ingredients. [16 P.U.R. 4th at 280.]
The clear import of this passage was that in future ratemaking cases, the system charge for the residential nonheating class gradually would be raised to reflect customer costs more accurately. It is clear, however, that the new customer (formerly “system”) charges which were established in the instant case achieve precisely the opposite result.
What the Commission has done is to raise the uniform commodity charge from 21.0$ per therm to 32.59$ per therm. [Order No. 6051, at 106.] At the same time, the Commission lowered the customer charge component for residential heating customers from $72 per year to $45 per year. It lowered the customer charge for residential non-heating customers from $45 per year to $30 per year. It lowered the customer charge for small (less than 3,000 Ccf) commercial and industrial heating customers from $148.50 per year to $58.50 per year. For large commercial and industrial heating customers, the PSC left the customer charge at $148.50 per year. Finally, it left the customer charge for all non-heating commercial and industrial customers at $84 per year. The Commission thus declined to increase any class’ customer charge, despite undisputed evidence that customer costs had increased for all classes. [Order No. 6051, at 106; Order No. 6060, at 28 — 29 (Stratton, C., dissenting).]
Commissioner Stratton lodged a vigorous and well-reasoned dissent from the Commission majority’s rate design determination. His dissenting opinion, from which the petitioners on this issue understandably draw much of their support, took the position that the new customer charges — coupled with the higher commodity charge— would increase the disparities in rates of return among customer classes. Basing his calculations on test-year consumption patterns and on customer class cost-of-service studies conducted by WGL, he computed class rates of return — both existing (based on the old rates) and projected (based on the new rates) — for firm (i.e., non-interrup-tible) service, as follows:
* Includes individually metered apartments
** Includes group-metered apartments
*1248[See Order No. 6060, Charts C & D accompanying Commissioner Stratton’s dissent.] Commissioner Stratton contended (as have petitioners here) that this move toward even more unequal rates of return components — precisely the opposite direction from that in which the PSC had promised in the 1976 rate proceeding to move — is devoid of justification in the record. That being true, he asserted, the new rate design results in an impermissible cross-subsidization of the residential non-heating class (whose rate of return is projected to be a stunning negative 8.42%) by the other classes, particularly the commercial/industrial heating and non-heating classes (the rates of return for which are projected to be 15.13% and 26.11%, respectively).
Those figures reveal that WGL’s different customer classes will be making vastly unequal contributions toward the Company’s overall authorized rate of return of 9.25%. Before addressing the legality of what I perceive to be manifest rate discrimination, however, I dispose of a concern relied upon by the majority and raised by the Commission and by People’s Counsel in their appellate briefs. They challenge the validity of the projected rates of return by pointing to an admitted error in WGL’s cost allocation studies. The Company apparently overstated the costs (both capital and operating costs) associated with gas delivery mains which were allocated to certain members of the residential heating and non-heating classes. WGL assigned those customers’ costs evenly on a per-meter basis, not accounting for the fact that residential customers living in individually metered apartments are responsible for lesser main-related costs per customer than are other residential customers. The effect of a proper adjustment would have been to lower somewhat the total customer costs and the rate base for the residential non-heating class, with a concomitant increase in that class’ projected rate of return.8 None of the witnesses, however, was able to pinpoint the degree of adjustment that would be appropriate, or to estimate even roughly the degree to which the rate of return would rise.
I carefully have reviewed the relevant cost allocation data and the related testimony of WGL’s witnesses, bearing in mind that petitioners have the burden of establishing a convincing case of rate discrimination before this court may invalidate the rate design. I am well satisfied, however, that while Commissioner Stratton’s rate of return figures may be distorted slightly due to the misallocation of main-related costs, those figures — viewed in conjunction with other record evidence — are reliable and indeed establish a prima facie case of rate discrimination. A number of factors makes this conclusion inevitable.
First, the allocation of costs pertaining to mains is an inherently imprecise endeavor. As compared with other costs incurred in delivering gas to customers (such as costs of regulators, meters, meterreading, billing, etc.), it is much more difficult to determine accurately for what percentage of the costs of the total system of mains each customer is responsible. This is true simply because many customers — and often customers in different classes — share the same mains. Hence, while WGL admittedly could have made further adjustments to its figures, it must be borne in mind that a good deal of educated quesswork naturally is involved in allocating costs associated with mains. Moreover, in computing the customer cost of mains for the residential classes, the Company took a very lenient approach.9 *1249For example, with respect to the residential heating and cooling class, less than one-third of the relevant mains were included. [Order No. 6060, at 11.]
Second, the cost data in the record make it clear that whatever the appropriate adjustment might have been, the projected rate of return for the residential non-heating class still would remain negative. As the Commission majority noted in its Final Opinion and Order, even if all costs associated with distribution mains were eliminated from the customer costs computation, the resultant customer costs still would exceed the new customer charges for the residential classes. See PSC Order No. 6060 at 11; see also Ex. WGL 7, Sch. 2, Item B, for cost figures associated with mains.10 Given the fact that residential non-heating customers use fewer therms per customer than any other class, the excess of customer costs over customer charges for this class obviously cannot be recovered through the commodity charge. When a customer class does not cover its costs through the customer and therm charges combined, a negative class rate of return necessarily results. When this happens, the utility cannot have the opportunity to earn its authorized overall rate of return unless higher-volume customers in other classes cover these excess costs through the therm charge. This is precisely the stuff of which rate discrimination is made.
Finally, even if WGL’s cost studies contain minor inaccuracies, they are the best— and indeed the only — evidence this court has to consider. They understandably were the only class-cost-of-service studies presented at the hearing. Even though one witness challenged their accuracy, neither he nor any other witness attempted to quantify the asserted misallocation. I therefore think it is not only reasonable but inescapable that the court should defer to the evidence of record, which petitioners have shown to be substantially correct. That evidence conclusively demonstrates that the PSC’s new rate design is discriminatory.
Had the majority addressed this issue correctly, the next step on review — having established that the new rate design significantly discriminates in favor of the residential non-heating class at the expense of *1250larger-volume commercial and industrial customers — would have been to determine whether the PSC advanced valid reasons for the rate disparities, and, if so, whether those justifications have adequate support in the record.
It indeed may seem peculiar that the majority opinion largely avoids quoting the Commission’s justifications for the new rate design. The reason the majority does not advert to the Commission’s reasoning, however, is clear: The PSC’s rationale simply does not square with that supplied by the majority in its desire to affirm the Commission’s actions entirely. To read the majority opinion on rate design, and then to read the Commission’s orders on the same subject, gives the impression that one is reading about two different rate proceedings. I think it is important to point out what the Commission’s reasoning actually was, so that its decision may be judged on its own merits rather than on those supplied by the majority on this appeal. The Commission explained its decision to decrease so substantially the residential non-heating class’ customer charge as follows:
This sub-class of residential customers uses gas solely for cooking and/or heating water.
Because of the very low volumes of gas which this sub-class takes relative to its customer costs, a two-part rate yields an anomoly; this sub-class pays the highest unit price for service at the same time as WGL earns the least return from this service.
In designing rates for the residential non-heating/non-cooling sub-class, the Commission faces a direct conflict between the principle of cost causation, on the one hand, and historic rate patterns and social considerations on the other hand. Recognition of historic rate patterns for this group is required. We shall accordingly reduce the customer charge for this sub-class from $3.75 to $2.50 per month. This $1.25 reduction in the customer charge maintains approximately the same relationship as before with the new lower customer charge for the residential heating and cooling sub-class.
We take this action essentially for equitable reasons. Because WGL incurs a comparatively large amount of fixed costs in order to deliver the very small volumes used by this sub-class, the Company’s service for cooking and hot water is unprofitable for WGL. Nevertheless, through past advertising and rate structure, WGL promoted the growth of this form of gas usage. As a result, today there are 51,000 cooking and hot water customers in D.C. At least at this time, we feel compelled to provide some degree of relief for existing customers who are in no position to respond to more cost-based rates. Equally important is the Commission’s desire to lessen the extent to which the low-volume gas user pays a higher per unit price for service. [Order No. 6051, at 96-97.]
On this appeal, the PSC and People’s Counsel contend that the Commission’s order amply explains the rate disparities in terms of valid non-cost based factors, particularly “equitable” considerations and the desire to preserve historic rate patterns.11 Petitioners, on the other hand, argue that the preference accorded to residential non-heating customers not only is explained inadequately by the PSC but also enjoys no independent support in the record. Petitioners draw upon Commissioner Stratton’s dissenting opinion, in which the following manifestly valid observations were made:
The purposes of rate design, once the revenue requirement is established, are, first, to establish rates that offer a reasonable prospect of generating revenues sufficient to cover the utility’s cost of service as determined by the Commission, and, second, to distribute the burden of those rates fairly among the utility’s customers. The rates established in this case most definitely fail the latter test. For what the Commission has not acknowl*1251edged is that in arbitrarily “softening the impact” of rates on small users it is increasing the impact of rates on medium and large users who pay back through the therm charge every cent and more of their customer charge reductions.
* * * * * *
I believe the Company, its customers, and a reviewing court are entitled at a minimum to a finding that fixed monthly charges based on fixed costs do have an adverse impact on small users, and, if so, how and why. Then the Commission should explain in the requisite detail why the “adverse impact” consideration should overcome all the other reasons advanced for not moving toward a one-part rate. Next the Commission should state how and why the adverse impact on only small users was taken into consideration when every other class of ratepayers is also impacted by fixed monthly charges. Finally, the Commission should justify its decision to visit additional impact on most ratepayers in order to alleviate the impact on the small user.
There is a reason that this order lacks those fundamental requisites. It is that the record in this case not only does not support this decision, but requires that the customer charges of those classes who are not now covering their customer costs through those charges be raised. [Order No. 6060, at 35-37 (Stratton, C., dissenting) (footnote omitted).]
Commissioner Stratton thus concluded that the new rates were not only discriminatory but unsupported as well. I think it is obvious, as Commissioner Stratton pointed out, that the increasedly unequal rates of return result from the PSC’s decision to lower the customer charge for most of its firm customers while raising the commodity charge. As more and more of the utility’s fixed costs (including customer costs) are recovered through the commodity charge, rather than through the customer charge, the burden of cost recovery falls disproportionately upon large-volume customers. See Order No. 6060, at 26 (Stratton, C., dissenting). All that remains to be determined is whether, contrary to Commissioner Stratton’s assertions, the rate disparities enjoy adequate record support in terms of valid cost or non-cost considerations.
As the Commission’s final order reveals, the customer charge reductions were effected “essentially for equitable reasons.” Order No. 6051, at 97. The Commission asserted that it faced “a direct conflict between the principle of cost causation, on the one hand, and historic rate patterns and social considerations on the other hand.” Ibid. It also alleged that “through past advertising and rate structure, WGL promoted the growth of this form of gas usage,” even though those customers are responsible for a comparatively large amount of fixed costs and have been unprofitable. Ibid. The Commission concluded that it was “compelled to provide some degree of relief for existing customers who are in no position to respond to more cost-based rates.” Ibid.
I turn to the “social” and “equitable” considerations cited by the PSC. Assuming, arguendo, that these factors could provide an appropriate justification for such a discriminatory rate structure, there is no evidentiary support for them in the record. While the Commission’s order suggests that the unprofitable residential non-heating customers were seduced into subscribing to gas service at low prices by WGL’s past promotional campaigns, the only record evidence of the Company’s promotional efforts is the following statement by WGL witness Smallwood:
Yes, we did have a large tonnage air conditioning promotional campaign in the mid and late ’60’s, as I guess everybody did until the crunch came in ’72.
That advertising campaign, of course, was directed to large scale users of gas in the heating and cooling sub-classes.12 As to the *1252assertion that low-volume users are, as a class, less able to respond to more cost-based rates than are other classes, there is a similar dearth of support for such a concept in the record. To the contrary, the Company’s witness Smallwood testified that although nearly 30,000 of WGL’s customers used less than 1,000 therms in the test year, no study had been made of the income distribution of that group:
We have not attempted to study whether customers using a thousand or less [therms] are low-income or fixed-income customers or whether they are well-to-do customers.
No other witness offered any testimony on the subject of income distribution by customer classes. In assuming that small users of gas tend to have lower incomes, then, the Commission acted not on the basis of reliable evidence in the record, but rather on the basis of pure, rank speculation. It just as easily might be fathomed that the typical residential non-heating customer is a relatively well-off family which uses gas for cooking and hot water, but which heats its house or apartment with now more costly oil or electricity.13
The Commission also mentioned its desire to preserve “historic rate patterns.” While preservation of historic rate patterns is a valid ratemaking objective when needed to temper a radical change in rate structure, the only relevant “historic rate pattern” in this case is a $3.75 customer charge for the residential non-heating class which was established in the 1976 rate proceeding. At that time, the customer charge recovered only about two-thirds of the class’ costs of service, but it was accompanied by the PSC’s promise to include more cost ingredients in future cases. See In re Washington Gas Light Co., supra, 16 P.U.R.4th at 280. What was then the future is now, but instead the residential non-heating class has been extended a $2.50 customer charge — a one-third reduction despite overwhelming evidence of rising customer costs. Certainly, then, the recent history of WGL’s rate structure does not support — and indeed contradicts — this action by the Commission.14
*1253The Commission and People’s Counsel nevertheless argue that the new rate design can be supported under this Court’s decision in Apartment House Council of Metropolitan Washington, Inc. v. PSC, supra. In that case, the PSC had established electric rates which resulted in somewhat unequal projected rates of return by customer class. We noted that in judging the reasonableness of the return differentials, “we first examine the cost-related factors and then the non-cost-related factors.” 332 A.2d at 56. We made it clear that “[i]t is not necessary that differences in rate of return be specifically and quantitatively supported by customer class-cost considerations.” Id., at 57. Valid non-cost-related factors — such as energy conservation and preservation of historic rate and usage patterns — may be used to help justify class return differentials. Id., at 57-58. We also subscribed to the general rule of thumb “that a rate structure must achieve the lowest price for the largest number of users.” Id., at 57, citing Philadelphia Suburban Transportation Co. v. Pennsylvania Public Utility Commission, 3 Pa.Commw.Ct. 184, 195-98, 281 A.2d 179, 186-87 (1971). Finding evidence of both cost and non-cost-related justifications in the record, we upheld the electric rate structure in Apartment House Council.
While Apartment House Council is precedent, the facts of that case are fundamentally different from the facts before us in at least two major respects. First, the disputed electric rates in Apartment House Council involved relatively minor discrepancies in rates of return. The following class returns were projected:
Residential 5.63%
General Service 7.78%
High Tension 8.27%
[332 A.2d at 55.]
In stark contrast, the projected class rates of returns in issue here (for firm service) range from 26.11% to a shockingly negative 8.42% — while WGL’s overall authorized rate of return is 9.25%.
Second, and more importantly, the court in Apartment House Council found substantial evidence in the record to support both the cost-related and the non-cost-related justifications put forth by the Commission. Here, the PSC essentially concedes that its rate design is not cost-justified, and there is no support in the record for the non-cost factors the Commission has advanced.
From a more fundamental standpoint, however, the non-cost factors relied upon by the Commission legally could not support the enormously disparate rates at issue here, even if there were evidence in the record to support them. The PSC’s enabling act provides in relevant part that:
The charge made by any such public utility for any facility or services furnished or rendered, or to be furnished or rendered, shall be reasonable, just, and non-discriminatory. Every unjust or unreasonable or discriminatory charge for such facility or service is prohibited and is hereby declared unlawful. [D.C.Code 1981, § 43-501.]
The Commission has a statutory mandate to set rates which are reasonable and non-discriminatory; nowhere in the Act is the Commission authorized to function in a contrary fashion.
The Colorado Supreme Court (sitting en banc) recently construed a virtually identical statutory provision as prohibiting that state’s public utilities commission from setting preferential rates based on social considerations. See Mountain States Legal Foundation v. Public Utilities Commission, 197 Colo. 56, 590 P.2d 495 (1979). In that case the PUC had established reduced gas rates (often called “lifeline” rates) for low-income elderly and disabled persons. The court determined that the preferential rates were invalid under the Commission’s enabling act, which forbids “unjust discriminatory rates.” The court stated:
*1254[T]he PUC has limited authority to implement a rate structure which is designed to provide financial assistance as a social policy to a narrow group of utility customers, especially where that low rate is financed by its remaining customers.
‡ * ife sjc sfc sj:
Establishing a discount gas rate plan which differentiates between economically needy individuals who receive the same service is unjustly discriminatory.
To conclude, although the PUC has been granted broad rate making powers by Article XXV of the Colorado Constitution, the PUC’s power to effect social policy through preferential rate making is restricted by statute no matter how deserving the group benefiting from the preferential rate may be. [Id., at 60, 590 P.2d at 497-98.]
Of course, the Colorado Supreme Court’s reasoning is equally applicable here, particularly given the pertinent similarity in the two jurisdictions’ utility codes. See also Public Service Co. of New Hampshire v. State, 113 N.H. 497, 509, 311 A.2d 513, 521 (1973); Note, Conservation, Lifeline Rates and Public Utility Regulatory Commissions, 19 Nat.Res.J. 411 (1979) (“Public Utilities— Public Service Commissions’ Jurisdiction and Powers Applied to Rate Structures: Public Service Commissions are limited to cost of service analysis in prescribing appropriate rate structures for public utilities. Environmental considerations and income redistribution may only be considered within the confines of this analysis”.)
The Commission in past years has recognized its proper mission, and it has acknowledged the primary role of cost-related factors in structuring rates. For example, in In re Potomac Electric Power Co., 84 P.U. R.3d 250 (D.C. PSC 1970), the PSC rejected a proposal that low-income customers be exempted from a general increase in electric rates. It then stated:
From the legal point of view, we are required by the statute to establish rates which are not unjustly discriminatory. The net result of the Consumer Council proposal is that persons with incomes higher than $5,500 per year would be paying higher rates for the same service in order to avoid an increase for those in the lower income category. The premise on which the discrimination would be based is that persons with incomes lower than $5,500 per year cannot afford any further increase. However, it could well be that persons with incomes in excess of $5,500 per year would also feel, in some cases with great justification, that they cannot afford to pay more for electricity either. We think that the standards suggested, i.e., the ability of a customer to pay the increase, is too subjective and too indefinite to provide a sound basis on which to base an obvious rate discrimination. [Id., at 255-56.]
More recently, the Commission recognized that
[b]ecause cost recovery is the ultimate test of equity among ratepayers, “the development of rate structures which conform as closely as possible to the cost standard provides the ideal goal toward which rate-making processes should tend.” [In re Chesapeake & Potomac Telephone Co., 9 P.U.R.4th 550, 571 (D.C. PSC 1975), quoting Sharfman, The Interstate Commerce Commission, Vol. III, at 325.][15]
See also Payne v. Washington Metropolitan Area Transit Commission, 134 U.S.App.D.C. 321, 335, 415 F.2d 901, 915 (1968). The preferability of cost-based rate structuring has been accepted widely by courts and regulatory agencies in other jurisdictions. See, e.g., Jager v. State, 537 P.2d 1100, 1109-10 (Alaska 1975); In re Utah Power Co. & Light Co., 22 P.U.R.4th 351, 373 (Ida*1255ho PUC 1977); Public Service Co. of Oklahoma, Opin. No. 788 (Feb. 17, 1977); In re Central Vermont Public Service Corp., 28 P.U.R.4th 469, 486 (Vt. PSB 1978); In re Virginia Electric and Power Co., 29 P.U. R.4th 65, 89 (Va. SCC 1979).
We recently upheld an electric rate structure prescribed by the Commission which was alleged to discriminate unreasonably against large scale commercial customers. In Metropolitan Washington Board of Trade v. PSC, D.C.App., 432 A.2d 343 (1981), we found no reversible error in the Commission’s decision to institute a “time-of-day” (TOD) peak load pricing scheme for commercial electric customers who consume an average of at least 1,000 kilowatt-hours monthly. Id., at 360-61. The petitioners in that case argued that it was unreasonable for the Commission to limit the new TOD pricing to large commercial customers, and that if it were to be instituted at all, it should be applied across the board. They contend that the net impact of the TOD rates would favor residential customers at the expense of the commercial class. We found the apparent rate discrimination to be justified by several legitimate mitigating factors, the likes of which are nowhere to be seen in the gas rate proceeding now before us.
First, the TOD rates are in an experimental stage, at least in this jurisdiction. As mentioned earlier, the Commission has the discretion when ushering in a radically different rate structure to temper its impact on those customers which it would affect most adversely. In addition, the Commission in Metropolitan Washington Board of Trade had noted that the costs of installing the new meters needed for TOD pricing could be borne much more feasibly by large customers than by smaller customers during the initial stages of the new pricing system. The Commission also indicated that other customer classes would be included in the future if the TOD experiment proves successful with the large electricity users.
The PSC’s electric rate decision in Metropolitan Washington Board of Trade, much like its gas rate decision in 1976, represented a commendable first step toward a new, more cost- and resource-efficient rate structure. In each case the disruption in the historic rate patterns was tempered appropriately during the transition period, but with strong indications that future rates would be structured on a more cost-realistic basis. The trouble is, now that what was then the future has arrived for WGL’s customers, the Commission has regressed toward an even less cost-based rate design.
Therefore, with due recognition of the Commission’s wide discretion in setting and structuring rates and of our limited review role, I nevertheless conclude that the PSC stepped well past the bounds of reasonableness in this proceeding. Petitioners have met their burden of showing that the new rate design is not “reasonable, just, and non-discriminatory.” D.C.Code 1981, § 43-301. The Commission failed to provide adequate explanation or evidentiary support for the obvious class cross-subsidization. The case should be remanded to the Commission with directions to restructure the rates so as to make them non-discriminatory in nature. The majority’s unpersuasive affirmance of the new rate structure is, in my view, an ominous disservice to the citizens, businesses, and governmental entities that use natural gas in the District of Columbia.
II. Treatment of Expenses Incurred for Advertising, Trade Association Membership, and Community Affairs Programs
At the suggestion of People's Counsel, the Commission — over Commissioner Strat-ton’s dissent — disallowed as above-the-line operating costs $95,000 in general advertising expenses, $56,000 in trade association dues, $130,000 expended in providing information to schools and community groups, and $35,000 in civic and business support payments. The Commission’s rationale in so doing, which is echoed by the majority here on appeal, is that “the Company failed to prove that these expenditures directly benefited the ratepayers,” and thus they should be borne by the Company’s shareholders. Ante, at 78, 87. While I recognize *1256that a number of factors recently has led a number of state commissions to adopt restrictive policies on the allowance of advertising and public affairs expenses, I do not believe that a “failure to prove” that each item of expense “directly benefits” the customers properly may be invoked as the sole basis for shifting the entire onus of those expenses to the utility’s stockholders. It is particularly unreasonable to adopt such a policy in this proceeding, in light of the Commission’s heretofore unvarying practice of recognizing all such expenses for rate-making purposes.
Commissioner Stratton aptly summed up the arbitrariness of the Commission majority’s action in this respect:
The total elimination from the cost of service of expenses for advertising, trade association dues, dues to civic and business associations, and community affairs programs is misguided overkill. Every organization, business or governmental, must communicate with its clients and customers, and the cost of doing so is a legitimate expense. The irony that those who object most strongly to the recognition of communication and outreach expenses in rates are often those whose activities make increased communication necessary only emphasizes the point. Like every expense, these must be reasonable in amount. If this Commission’s past practice of allowing all such expenses which are recorded in above the line accounts be thought too uncritical, the remedy is to correct it either through rule-making (as has been done in New York) or through a more careful analysis of these expenses in the context of the rate proceeding, but not an abrupt reversal of long-standing Commission policy. [Order No. 6060, at 54-55 (Stratton, C., dissenting).]
A. Advertising Expenses
The Commission’s decision to change its prior policy so as to disallow WGL’s advertising expenses for rate-making purposes was precipitated by the testimony of People’s Counsel’s witness Sack. He stated:
In light of the Company’s restrictions concerning connecting of new customers, I did not consider it appropriate to have the ratepayer provide funds for advertising when such advertising may be in vain.[16]
In its brief before the Commission, People’s Counsel further alleged that WGL’s advertising was “image building” in nature, intended solely to improve the utility’s stature in the community. As such, People’s Counsel argued, the Company’s advertising expenses should be borne entirely by the stockholders.
On cross-examination by WGL’s counsel, Sack admitted that he had made no effort to ascertain the actual nature of the advertising in question. He acknowledged that his characterization of the advertising as “image building” was based merely upon “speculation ... of what the descriptive titles would imply.”
In his prepared rebuttal testimony, WGL’s witness Unkle responded to Sack’s assertions as follows:
Q. Will you please discuss the nature of the advertising expenses which Mr. Sack eliminated from the cost of service.
*1257A. The $95,000 of general advertising costs included in the Administrative and General Expense relates to advertising which, in common terminology, would be described as “informational” in character. The advertising addresses two specific issues of great concern to our customers— the gas supply situation and the rising cost of gas. It is entirely proper and right that our company address these issues in communicating with customers.^17!
Despite the paucity of support in the record for People’s Counsel’s allegations, the Commission for the first time imposed upon the Company the burden to “come forward with a detailed presentation” of how its advertising directly benefited its customers. Order No. 6051, at 69. I believe it was arbitrary for the Commission ex post facto to impose such a burden upon WGL, and I do not believe it was reasonable to conclude that absent such a detailed showing the advertising in question must be presumed to benefit only the Company’s stockholders.
In 1935, the Supreme Court held that a state utilities commission’s disallowance of a gas distributor’s promotional advertising expenses was an unconstitutional usurpation of management prerogative. West Ohio Gas Co. v. PUC, 294 U.S. 63, 55 S.Ct. 316, 79 L.Ed. 761 (1935). Justice Cardozo, writing for the Court, stated:
Good faith is to be presumed on the part of the managers of a business. In the absence of a showing of inefficiency or improvidence, a court will not substitute its judgment for theirs as to the measure of a prudent outlay. The suggestion is made that there is no evidence of competition. We take judicial notice of the fact that gas is in competition with other forms of fuel, such as oil and electricity. A business never stands still. It either grows or decays. Within the limits of reason, advertising or development expenses to foster normal growth are legitimate charges upon income for rate purposes as for others. When a business disintegrates, there is damage to the stockholders, but damage also to the customers in the cost or quality of service. [Id., 294 U.S. at 72, 55 S.Ct. at 321 (citations omitted).]
West Ohio Gas has been interpreted as establishing that, once the company has made out a prima facie case that an advertising expense was incurred, the burden rests with the Commission’s staff (or any other interested party) to prove that the expense was unreasonable because of wastefulness, inefficiency, or bad faith. Boise Water Corp. v. PUC, 97 Idaho 832, -, 555 P.2d 163, 169 (1976), citing West Ohio Gas, supra, 294 U.S. at 72, 55 S.Ct. at 321; see Note, Advertising by Public Utilities as an Allowable Expense for Ratemak-ing: Assault on Management Prerogative, 13 Val.U.L.Rev. 87, 119-20 (1978). In this case WGL, relying upon the Commission’s long-established practice of approving all advertising expenses listed in the firm’s General and Administrative Expenses account, produced evidence that it had spent $95,000 on “informational” advertising which dealt with “the gas supply situation and the rising cost of gas.” This should have been sufficient to establish a prima facie showing of reasonable informational advertising expenditures, subject to proof by the Staff or People’s Counsel to the contrary. People’s Counsel’s witness made an allegation that the advertising was strictly for “image” building and that it was inappropriate as an operating expense in light of WGL’s restrictions concerning *1258the connecting of new customers.18 He admitted, however, that he had no idea what the subject matter of the ads had been. The PSC nevertheless concluded that “[fjaced with this challenge, WGL should have come forward with a detailed presentation.” Order No. 6051, at 69. In so ruling, the Commission erroneously shifted the burden of proof back to the Company upon nothing more than a bald allegation by People’s Counsel’s witness.
I also do not think that the allowance of advertising costs as an operating expense should hinge on whether the advertising is classified as “informational,” “promotional,” or “institutional.” The utility’s management must have the discretion to make decisions on advertising, and the costs of advertising should be treated as valid operating expenses so long as they are reasonable in amount and the advertising is designed to serve some legitimate utility purpose. While West Ohio Gas involved “promotional” advertising, its holding logically has been extended to cover “institutional” (i.e., goodwill or image-building) advertising as well. In New England Telephone and Telegraph Co. v. Department of Public Utilities, 360 Mass. 443, 275 N.E.2d 493 (1971), the Supreme Judicial Court of Massachusetts quoted approvingly from a New York PSC decision:
What is of concern here are advertisements which are obviously designed to project a favorable image of the company to its customers, its existing stockholders or potential investors. To the extent that such advertising fosters sound consumer relations or encourages people to invest in the company, it seems clear that the consumers, as well as the stockholders, are ultimately benefitted through the lessening of the expense of doing business. * * In conclusion, it is determined that in view of the fact management should control advertising expenditures as long as they are within the limits of reason, and so long as these expenses do not exceed what is reasonably necessary and proper in the particular case .. . there is no ground to distinguish such costs from other necessary and proper expenses. [Id., 360 Mass, at 485, 275 N.E.2d at 518, quoting In re Consolidated Edison Co. of N. Y, 41 P.U.R.3d 305 (N.Y. PSC 1962).]
More recently, the Supreme Court of Alabama wrote pointedly in addressing the same issue:
Although subject to regulation by the government, a utility, like any corporation, should be allowed to operate consistent with the free enterprise system to the extent possible.
Advertising is of vital importance to corporations in establishing and maintaining their public image, as well as in educating the consuming public. As such, it is a responsibility of the duly authorized manager of a utility to decide the type, quantity, or form of advertising which would most benefit the corporation in its continued growth. The utility has the initial right to decide the amount and type of advertisement which comports with good management practices. The function of the Alabama Public Service Commission is that of regulation, and not of management. Petition of New England Tel. & Tel. Co., 115 Vt. 494, 66 A.2d 135 (1949). The Commission should not be allowed to interfere with the proper operation of the utility as a business concern by usurping managerial prerogatives. [Alabama Power Co. v. PUC, 359 So.2d 776, 780 (Ala.1978).]
After quoting with approval from West Ohio Gas, the court concluded:
Advertising, under honest, efficient, and economical management, can be an operating expense and reasonable advertising costs expended by a public utility should be allowed. [Ibid.]
See Central Maine Power Co. v. PUC, 153 Me. 228, 243—46, 136 A.2d 726, 736-37 (1957); In re New England Telephone & Telegraph Co., 115 Vt. 494, 509-12, 66 A.2d *1259135, 145-46 (1949); see also El Dorado v. PSC, 235 Ark. 812, 825, 362 S.W.2d 680, 688 (1962); Southern Bell Telephone & Telegraph Co. v. PSC, 203 Ga. 832, 49 S.E.2d 38, 65 (1948); State ex rel. North Carolina Utilities Commission v. Piedmont Natural Gas Co., 254 N.C. 536, 550-52, 119 S.E.2d 469, 479-80 (1961); Commonwealth v. Virginia Electric and Power Co., 211 Va. 758, 772, 180 S.E.2d 675, 685 (1971), aff’d In re Virginia Electric Power Co., 84 P.U.R.3d 1 (Va. SCC 1970); General Telephone Co. of Wisconsin v. PSC, 46 P.U.R.3d 1, 5 (Wis.Cir.Ct. 1962).
As the majority notes, a number of state commissions recently have begun to disallow promotional and/or institutional advertising expenses as operating costs. Some of the decisions were the result of concerns that promotional advertising was unwise in the face of then-critical energy shortages. Others simply concluded that institutional or goodwill advertising categorically is of no benefit to the consumer. I am unaware of any such opinion which even attempts to explain its departure from the principles laid out in West Ohio Gas and its progeny.19 Nevertheless, however excusable the silent abandonment of West Ohio Gas might be in other individual cases, this case presents no special circumstances which would justify such a departure.20 At the very least, the Company deserves some advance notice that the Commission is changing its longstanding policy so that it may be prepared to defend its advertising decisions in more detail.
In summary, I believe the Commission and the majority of this court have erred *1260both with respect to the allocation of the burden of proof and by drawing unsupported conclusions as to the value of the advertising involved to WGL’s consumers. I would remand this issue to the Commission with instructions to make an appropriate adjustment.
B. Trade Association Dues
For essentially the same reasons that pertain to the issue of advertising expenses, it was arbitrary and unreasonable for the Commission to have excluded from operating costs the entire $56,000 expended by WGL for trade association dues. In affirming the disallowance, the majority states that the PSC’s action was reasonable “in the absence of evidence that AGA activities benefited customers.” Ante, at 1229. I am at a loss to see how the Company can be expected to demonstrate a corresponding consumer benefit for each dollar spent on AGA dues. Such a requirement places an almost unsurmountable burden of proof upon the Company.
The Company’s management deserves considerable discretion in determining whether such membership is in the best interest of the utility. Where it appears that association membership may improve the overall efficiency of the utility’s operation, its cost should be allowed, at least in substantial part, as a legitimate cost of doing business. The uncontradicted evidence of record reflects that the AGA serves as a valuable pool of resources for its member distributors. It provides informational as well as research and development services to its members at a much lower per-member cost than would be possible if each member company were to conduct those activities itself. To the extent that the AGA engages in political or other activities which reasonably could not be expected to benefit consumers even indirectly, perhaps the Commission could apportion an appropriate part of the membership dues to the stockholders. A total exclusion of the entire membership fee from the Company’s operating expenses, however, assuredly is not justified.
The purpose of the Commission’s oversight of a utility’s operating costs is to prevent monopolistic exploitation in the form of exorbitant returns or wasteful expenditures. Insofar as a utility’s operating expenses are consistent in nature and amount with those incurred by competitive businesses, they should be allowed. It goes without saying that the concept of trade association membership is hardly unique to monopolistic enterprises. Competitive industries, businesses, and professions almost invariably are served by trade associations, and the individual members may be expected to pass the costs (and the benefits) of association membership on to their customers.
In the past, the PSC has allowed such dues as an operating expense. A number of other state commissions similarly include these costs above-the-line. See, e.g., In re Peoples Gas System, Inc., 11 P.U.R.4th 505 (Fla. PSC 1975) (abstract); In re Southwestern Bell Telephone Co., 28 P.U.R.4th 519, 537 (Kan. SCC 1979); In re Rules and Regulations Covering Rate Case Treatment of Certain Expenses, 30 P.U.R.4th 338, 343 (N.M. PSC 1979); In re Columbus and Southern Ohio Electric Co., 24 P.U.R.4th 261, 287 (Ohio PUC 1978). In my view, the Commission’s action on this expense item was arbitrary, unreasonable, and unsupported by evidence in the record. I therefore disagree with the majority on this point, too, and would remand for an appropriate adjustment.
C. Community Affairs Expenses
Again, I cannot concur in the Commission’s and the majority's treatment of community affairs expenses for essentially the same reasons that apply to the two expense items just discussed.
Despite the vague allegation by People’s Counsel’s witness Sack that WGL’s community affairs activities are “clearly to enhance the public image of the Company,” the record unquestionably reflects that the program is primarily informational in nature. WGL sends out speakers, at the invitation of schools and community groups, to *1261talk about matters relating to gas service such as fuel conservation, safety considerations, the two-part billing system, estimated billing, the gas supply situation, and the rising cost of gas. On cross-examination by WGL’s counsel, Sack admitted that his characterization of the Company’s speakers program as “one of enhancing the public image of the Company” was based solely on speculation. In prepared rebuttal testimony, WGL’s witness Unkle told the Commission:
Talking about such subjects to community groups is one important way our Company can provide information about subjects of vital concern both to us and our customers.
Mr. Sack is wrong when he says these activities are undertaken to enhance our image. That is not the purpose of our community activities. The purpose is to communicate with customers on those subjects about which they want more information. In fact, the Company is under a certain amount of community pressure to engage in such activities. Such expenses are not only proper, but they are vital.
The PSC’s action here again constitutes an unreasonable usurpation of management prerogative. Managers of a utility have the discretion, and indeed the duty, to inform their customers about such service-related concerns as safety, conservation, billing, cost-of-service, and so forth. Expenses incurred by virtue of conducting such informational activities, to the extent that they are not excessive or the result of an abuse of managerial discretion, are valid costs of doing business and should be treated as such in the ratemaking process.
The Company’s community affairs program is closely akin to the concept of informational advertising, which all parties before us and an overwhelming majority of state commissions agree is a proper above-the-line expense. See, e.g., In re Consumers Power Co., 14 P.U.R.4th 1, 29-30 (Mich. PSC 1976); In re Northern States Power Co., 6 P.U.R.4th 38, 42 (N.D. PSC 1974); In re Gas Co. of New Mexico, 21 P.U.R.4th 159, 175 (N.M. PSC 1977); In re Columbia Gas of Ohio, Inc., 25 P.U.R.4th 399 (Ohio PUC 1978); Re Utility Advertising Expenditures, 14 P.U.R.4th 578 (Ore. PUC 1976) (abstract); Pennsylvania PUC v. Peoples Natural Gas Co., 28 P.U.R.4th 180 (Pa. PUC 1978); In re Northwestern Bell Telephone Co., 15 P.U.R.4th 289 (S.D. PUC 1976); In re Wisconsin Power & Light Co., 4 P.U. R.4th 305 (Wis. PSC 1974). The Florida PSC has promulgated rules specifically providing that community activities of the type in question here are legitimate costs of service. See In re Promotional Practices of Electric Utilities, 8 P.U.R.4th 268, 276 (Fla. PSC 1975). Even the majority acknowledges informational advertising as a legitimate above-the-line expense item. See ante, at 1230. However, the majority then retreats to its familiar and transparent posture with respect to the Company’s burden of proof:
As the Company has primary control of its own records, there is no injustice in allotting to the Company the initial responsibility for demonstrating the reasonableness of above-the-line charges to its ratepayers.
The sparse record with respect to the challenged community affairs expenses consists merely of cross-allegations as to the nature and purpose of the speakers programs, unsubstantiated by documentary evidence on either side. * * * [Ante, at 1230.]
It is difficult to conceive of what further proof could reasonably be expected of the Company on an item such as this. The Company’s witness explained the nature of its speakers program, the subjects covered, and the community demand for the program. Perhaps the Commission would wish to see a verbatim transcript of each public appearance by a WGL speaker, but such a monitoring requirement hardly could improve the cost efficiency of the program for either the ratepayers or the stockholders.
The Commission also is misguided in its attempt to equate the speakers program with “forced contributions” to charity. As the record reveals, the Company’s communi*1262ty activities are not limited to certain charitable organizations, but rather are available to any group, regardless of its nature, that expresses an interest therein. The ratepayers are not being forced to contribute to selected organizations which they otherwise might not choose to support. The only selectivity involved lies in the fact that groups, rather than individuals, are the immediate beneficiaries. However, other more conventional forms of informational communication (which indisputably are proper operating expenses) probably are not equally available to all ratepayers, either. Therefore, the PSC’s finding that “only a limited number of community organizations are receiving speakers from WGL” begs the issue.
On the other hand, I do concur with the majority in affirming the disallowance of WGL’s contributions to various business and civic organizations. The Company’s direct monetary contributions to those organizations — unlike its informational speakers program — are tantamount to contributions to charitable concerns. Although even regulated companies are under a great deal of community pressure to make such donations, the question of whether ratepayers should finance this largesse is a sensitive one. While state commissions and reviewing courts are divided on the appropriateness of including civic support payments in the cost of service, see Annot., Charitable Contributions by Public Utility as Part of Operating Expenses, 59 A.L.R.3d 941 (1974), I agree that it was within the PSC’s discretion to find that WGL’s customers should not be compelled through rate payments to make contributions to organizations which they independently might not wish to support.
III. Otheh Issues and Observations
Apart from the issues which I have discussed, I join the majority in finding no reversible error in the other challenged aspects of the Commission’s orders. Nevertheless, in this, my last opinion as an active member of this court, I feel compelled to comment upon the Commission’s apparent perception of its role in setting utility rates. See also Potomac Electric Power Co. v. PSC, D.C.App., 402 A.2d 14, 27 (en banc) (Harris, J., dissenting), cert. denied, 444 U.S. 926, 100 S.Ct. 265, 62 L.Ed.2d 182 (1979).
It cannot be a coincidence that virtually every contested issue throughout the course of this proceeding was decided adversely to the Company. The Commission in recent years has shown an unmistakable predisposition to treat WGL and the city’s electric energy supplier, Potomac Electric Power Company, as some sort of miscreants who unilaterally are to blame for higher energy costs. In this case, as in others, the Commission simply has not lived up to its duty to balance the ratepayers’ interest in preventing unduly high rates with the utility shareholders’ legitimate interest in having the Company have the opportunity to earn a fair rate of return.
One must wonder why WGL’s authorized return on equity is at the bottom end of the “zone of reasonableness” in comparison with comparable gas distributors; why the Company is granted no attrition allowance despite clear evidence of consistent regulatory lag and the fact that the Company has been unable to earn anywhere near its authorized rate of return in recent years; and why the PSC denied each of the Company’s far from outlandish requests for rate relief on such items as tax normalization, consolidated tax savings, and the treatment of gains on sales of propane reserves. Moreover, one must wonder how the Commission rationally could conclude, following more than a decade of rampant inflation, cost increases, and high interest rates, that the Company’s authorized rate of return on equity should be reduced from 13.25 percent to 13 percent.
I sincerely hope the Commission is not so single-purposed as to ignore the fact that financial health is essential to the provision of reliable utility service. The regulation of natural monopolies requires give-and-take on both sides. If the utility consistently is required to do all the giving, no one can benefit in the long run and one day there may be nothing left to take.
*1263Finally, I direct attention to the fallacy in what apparently is the underlying basis for the majority’s acquiescence in all aspects of the orders under review. On p. 1194, ante, the majority takes the position that
if the Commission has fully and clearly explained what it does and why it does it, and the agency decision is supported by substantial evidence, the court, upon a finding that the Commission order is reasonable in its overall effect, should sustain the order.
It is not for me to speculate as to whether the majority genuinely believes that that unsupported assertion truly describes our standard of review. I readily state, however, that 'only by applying such a distorted concept of the proper review standard may the orders under review be sustained in all respects. No one should be misled into believing that the quoted statement may alter existing law. Utilities and their investors, no less than others, still have constitutional rights. Utilities still have common law rights which have been set forth in many decisions which need not be cited again here. Utilities also still have statutory rights. Among them is this court’s obligation to reject findings by the Commission which “are unreasonable, arbitrary, or capricious.” D.C.Code 1973, § 43-706. Also existing, as noted above, is the following statutory proscription:
The charge made by any . . . public utility for any . .. services furnished, or rendered, ... shall be reasonable, just, and nondiscriminatory. Every unjust or unreasonable or discriminatory charge for such ... service is prohibited and is hereby declared unlawful. [Id., § 43-301.]
The majority’s “see-no-error, hear-no-error” approach to this case, coupled with its purported belief that nothing matters except “the overall effect” as long as any deviant component of a rate order somehow is explained, may persuade it that total affirmance is appropriate in this case. It does not persuade me, however, nor can I conceive that it persuades even the minimally knowledgeable objective reader.
For the reasons set forth above, while I concur in the affirmance of some of the Commission’s challenged rulings, I conclude that reversible error was committed in the portions of the orders under review which I have discussed in detail. Accordingly, I respectfully dissent in part.
Associate Judge Harris was a member of this court until February 5, 1982. His opinion in this case was filed prior to his retirement on that date. Minor modifications subsequently were made to the per curiam opinion and to his opinion.
. The projected rates of return by customer class under the new rate structure are as follows:
Class of Service Rate of Return - -
Residential heating 4.91%
Residential non-heating (8.42)%
Commercial heating 15.13%
Commercial non-heating 26.11%
See Order No. 6060, Charts C & D accompanying Commissioner Stratton’s dissent. As will be discussed below, the substantial disparities result from the Commission’s inclusion of an increased- proportion of fixed, customer-related costs in the commodity or “therm” charge, and *1243a decreased proportion of such costs in the customer charge.
. The majority’s insinuation in n.18 of its opinion (pp. 1201-1202) that this standard of review is one which a court should not be expected even to try to live up to is disheartening.
. See note 1, supra.
4. I cite to the majority opinion in Washington Public Interest Organization v. PSC without enthusiasm, because I agree with the view of Judge Nebeker that the majority in that case misconstrued its review function in remanding the proceeding to the PSC. See 404 A.2d 541 (1979) (Nebeker, J., dissenting). The above-quoted passage, however, correctly reflects the *1246Commission’s burden of elaboration in rate cases.
. As the Commission has acknowledged in this case, the term “system” charge was a misnomer, because the charge was designed to recover only customer costs and not any part of the overall system’s demand costs. The Commission has corrected the error in nomenclature by renaming this component the “customer” charge. Order No. 6051, at 4, 81.
. Recovery of demand costs was allocated as part of the commodity charge, rather than the system charge. As Commissioner Stratton points out in his dissent in this proceeding, “it is reasonable to collect demand costs through the therm charge if one assumes that a portion of the revenue from each therm sold is applied toward recovery of demand costs.” PSC Order No. 6060, at 25 (Stratton, C., dissenting). This is true because higher-volume customers generally are responsible for higher demand costs than are low-volume customers. See id., at 26. The Commission did, however, leave the door open to the possibility of including some demand costs in the system charge in future rate proceedings. See In re Washington Gas Light Co., supra, 16 P.U.R.4th at 280.
.A heating customer is one who uses gas for heating and/or cooling purposes. Non-heating customers use gas solely for cooking and/or hot water. Individually-metered apartments are included in the residential class; group-metered apartment buildings are part of the commercial/industrial class.
Although the system charge (now the customer charge) was a fixed annual assessment, it was spread over the year and was payable monthly. For heating customers, however, the system charge was spread over only the nine coldest months of the year, in order to “remove *1247the incentive for heating-only customers, or those on vacations during the summer, to discontinue service during June through August.” Id., at 279-80. The same nine-month policy has been readopted in this proceeding with respect to heating customers.
. Such an adjustment would have a similar effect on the residential heating class, but only to a negligible extent. Out of approximately 87,500 billing units in that class, only about 4,000 are individually metered apartments. This constitutes less than 5% of the class. In the residential non-heating class, on the other hand, individually metered apartments comprise roughly 55% of the class (28,000 out of 51,000 billing units).
. Rather than using the full costs of all serviceable mains, the Company allocated to each customer the cost of the theoretical minimum-sized main necessary to serve that customer. WGL’s witness Smallwood explained:
We discerned the minimum size — theoretical minimum size main by doing a least squares regression analysis on the most prevalent types of mains in our system and by all of the classes. We have basically three types *1249of mains in our system: cast iron, steel and plastic, and we did a least squares regression analysis on those types of mains of the various sizes, and where the least squares trend line intercepted the Y axis at zero point, we determined that that was the minimum cost that could be associated with the theoretical minimum size main.
In other words, we don’t have any mains of zero size, but it would be unfair to assign as a customer component a two-inch main if that is our smallest because to serve only one customer you wouldn’t need a two-inch main.
So we simply said what is the theoretical minimum size we can get, and if you go down to theoretical minimum, that is a pin hole through a steel rod, and if that is what is needed to serve, then the minimum cost would be associated with that minimum size. And so that is how we arrived at this cost and then we spread that among the customers on the basis of each customer in every class would have to bear its average cost of that main. So we just took the total cost after we determined what the theoretical minimum size total cost was and divided by the number of customers in the jurisdiction and each customer then, bore a proportionate share of that.
. It is noteworthy that when WGL’s witness Smallwood was questioned about the asserted misallocation in the costs of mains, he stated that although he could not be certain, the rate of return for the residential non-heating class “probably” would still be negative even after the misallocation was corrected. It is important to recognize that Smallwood was referring to the expected return under the Company’s proposed rate structure, not under the more discriminatory rates ultimately adopted by the PSC. Under the Company’s proposed rates, the return on the residential non-heating class was projected to be a negative 0.49%. Under the rates eventually adopted, the projected return for that class (as calculated by Commissioner Stratton) was a more dramatically deficient negative 8.42%. It goes without saying that if the return still would have remained negative under the Company's proposed rates even after an appropriate cost adjustment, as Smallwood ventured, then the return based on the rates ultimately adopted necessarily would have remained well within the negative range after the adjustment as well. The majority seems to confuse the projected rates of return computed by Commissioner Stratton with those put forth by WGL.
. There is nothing, however, in the briefs of either the PSC or People’s Counsel which suggests that the new rate structure can be upheld on the ground principally relied upon by the majority — i.e., the notion of the rising costs of gas.
. On this point, Commissioner Stratton accurately stated:
The conclusion that the Company “promoted” this business is unsupported on this record. Rather, these uneconomical customers *1252have been attached pursuant to WGL’s utility obligation. Acknowledging that they are unprofitable, the Commission has adjusted rates downward, thereby increasing their un-profitability and the burden they impose upon other ratepayers. The Commission thereby commits the classic regulatory folly upon the gas business, namely to decree uneconomically low prices for a utility service thereby insuring that it will be oversold and bring calamitous results in the future. If anyone stands fairly accused of “promoting” this business surely it is the Commission who has lowered rates to a noncompensatory level in the face of rising costs. [Order No. 6060, at 38-39 (Stratton, C., dissenting).]
. In his dissent, Commissioner Stratton commented:
What of the conclusion that low usage ratepayers “are in no position to respond to more cost-based rates.” One searches the record in vain for evidence to support this statement. We know that some utility customers have seen their utility bills rise faster than their incomes over the past five years, and they desire a subsidy of their utility usage, which a compassionate body politic and its elected representatives may provide them some day. Meanwhile, because the regulatory process affords a forum to air their plight we are invited to provide the subsidy at the expense of utility investors and other ratepayers. And PSC has done so, by this decision and by its actions in electric rate decisions in recent years. But there is no evidence that these victims of a rising cost of living are low usage consumers of gas, or that all low usage consumers of gas are similarly situated. Targeting compassionate relief in a way that reduces every bill for low gas usage benefits every customer, needy or not, on every bill for low usage at the expense of every customer, needy or not, on every bill for high usage.
The Commission has succumbed to what Alfred Kahn has called, “the almost irresistible opportunity it [economic regulation] offers to use price — typically very imprecisely and inefficiently — as an instrument for the redistribution of income.” [PSC Order No. 6060 at 39-40 (Stratton, C., dissenting) (footnotes omitted), quoting A. Kahn, Applying Economics to an Imperfect World, an article drawn from the 1978 Ely lecture before the American Economic Association, reprinted in 2 Regulation 17, 18 (November/December 1978).]
. As Commissioner Stratton noted in his dissent, the PSC apparently used the term “historic rate pattern” in a relative sense, by maintaining the existing ratio between the residential heating customer charge and the residential non-heating customer charge. See Order No. *12536060, at 38 (Stratton, C., dissenting). The Commission apparently felt it would be treating all classes fairly if the customer charge for each were reduced proportionally. I already have noted why this reasoning is fallacious: it ignores the fact that high-usage customers pay back through the increased commodity charge much more than the amount of savings they realize from a decreased customer charge. See id., at 35-36.
15. In his dissent in this case, Commissioner Stratton made the following observation:
True, there are considerations other than cost —equity, history, rate stability — that go into rate design, but these considerations have not been accorded primacy; they have been used to justify departures from cost-based rate-making. Economic regulation must commence from a basis of costs, otherwise the regulation is not “economic” but something else. [Order No. 6060, at 23 (Stratton, C., dissenting).]
16. Sack did not explain, and it is not clear from the record, exactly to what restrictions he was referring. I note, however, that in May 1978 the PSC authorized WGL to expand its annual firm service obligation by 5,000,000 therms to a maximum of 3,000 new customers. Washington Gas Light Co., Formal Case 687, Order No. 5998 (May 16, 1978). The Commission determined in that case that the Company’s projected gas supplies would be adequate to satisfy the needs of both old and new customers through the year 2000. Id., at 6. It concluded:
The prognostication for future supplies is such that, maintaining a margin for error, it is fair to conclude that the Company will not only hold its own but increase the energy available for delivery to its customers by some five percent in the fairly near term, and that this higher level of supply can reasonably be expected to continue and even rise moderately through the end of this century.
[Ibid.]
. On cross-examination, Unkle acknowledged that the Company recorded the $95,000 in disputed advertising expenses in Account 930.1 of the FERC Uniform System of Accounts. According to an explanatory note accompanying Account 930.1, properly includable in that account is “the cost of advertising activities on a local or national basis of a goodwill or institutional nature, which is primarily designed to improve the image of the utility or the industry .... ” However, Unkle disavowed any reliance by the Company on the “technical wording” of the Uniform System of Accounts, maintaining that the advertising was “informational” in nature. All parties seem to agree that “informational” advertising properly would be includa-ble in the Company’s cost of service.
. The “restrictions” to which he evidently was referring were not very restrictive at all. See n.16, supra.
. See Note, Advertising by Public Utilities as an Allowable Expense for Ratemaking: Assault on Management Prerogative, supra, in which the author criticizes the trend among state commissions toward disallowance of advertising expenses as an improper usurpation of management discretion in violation of the Supreme Court’s opinion in West Ohio Gas.
. The Supreme Court recently ruled that the New York PSC constitutionally could not ban an electric utility from advertising to promote the use of electricity. Central Hudson Gas & Electric Corp. v. PSC of New York, 447 U.S. 557, 100 S.Ct. 2343, 65 L.Ed.2d 341 (1980). The Court found that the State’s interest in energy conservation was substantial, but held that the commission had not shown that such an interest could not be protected adequately by more limited regulation of the utility’s commercial expression. Id., 447 U.S. at 569-70, 100 S.Ct. at 2353. Although the decision in Central Hudson was based on free speech considerations and thus did not deal precisely with the question presented here, the Court confirmed the continuing validity of West Ohio Gas:
Monopoly over the supply of a product provides no protection from competition with substitutes for that product. Electric utilities compete with suppliers of fuel oil and natural gas in several markets, such as those for home heating and industrial power. This Court noted the existence of interfuel competition 45 years ago, see West Ohio Gas Co. v. Public Utilities Comm’n, 294 U.S. 63, 72, 55 S.Ct. 316, 321, 79 L.Ed. 761 (1935). Each energy source continues to offer peculiar advantages and disadvantages that may influence consumer choice. For consumers in those competitive markets, advertising by utilities is just as valuable as advertising by unregulated firms.
Even in monopoly markets, the suppression of advertising reduces the information available for consumer decisions and thereby defeats the purpose of the First Amendment. The New York court’s argument appears to assume that the providers of a monopoly service or product are willing to pay for wholly ineffective advertising. Most businesses — even regulated monopolies — are unlikely to underwrite promotional advertising that is of no interest or use to consumers. Indeed, a monopoly enterprise legitimately may wish to inform the public that it has developed new services or terms of doing business. A consumer may need information to aid his decision whether or not to use the monopoly service at all, or how much of the service he should purchase. In the absence of factors that would distort the decision to advertise, we may assume that the willingness of a business to promote its products reflects a belief that consumers are interested in the advertising. [Id., 447 U.S. at 567-68, 100 S.Ct. at 2352 (footnotes omitted).]
In a footnote to the above-quoted passage, the Supreme Court added:
There may be a greater incentive for a utility to advertise if it can use promotional expenses in determining its rate of return, rather than pass those costs on solely to shareholders. That practice, however, hardly distorts the economic decision whether to advertise. Unregulated businesses pass on promotional costs to consumers, and this Court expressly approved the practice for utilities in West Ohio Gas .... [Id., 447 U.S. at 568 n.ll, 100 S.Ct. at 2352 n.ll.]