| |
|
|
|
[5363] The main focus of the Service's proposal for Standard A Mail is on several important internal cost and rate design concerns. One is the cost support for the commercial pound rates. Another is the statutory preferred rate markup formula. Others involve cost coverage considerations, especially in terms of the effect of the proposal on mailers and competition, and the choices entailed in balancing numerous-and often competing-rate design objectives. The latter are often most evident in the level of worksharing-related savings the Service passes through to mailers and the extent to which shape-related cost differences are recognized.
[5364] The proposal, as filed, reflects a system average increase of 6.4 percent for all mail classes and services. USPS-T-32 at 36. For individual subclasses, the increases are 9.4 percent for Regular; 4.9 percent for Enhanced Carrier Route (ECR); 5.6 percent for Nonprofit; and 14.8 percent for Nonprofit Enhanced Carrier Route (Nonprofit ECR).1 USPS-T-35 at 36.
[5365] In response to persistent questions about the support for the pound rate, the Service presents a new cost approach. Based on this analysis and related considerations, it proposes reducing the Regular pound rate by 1.6 cents (from 67.7 cents to 66.1 cents), and reducing the ECR pound rate by 7.9 cents (from 66.3 cents to 58.4 cents). Based largely on other considerations, the Service proposes increasing the Nonprofit pound rate by 3 cents (from 55 cents to 58 cents), and increasing the Nonprofit ECR pound rate by 8.0 cents (from 29 cents to 37 cents). The proposed Nonprofit ECR rates are also based on an anticipated amendment to the Revenue Forgone Reform Act of 1993 (RFRA).2
[5366] Piece rates, which are assessed on all Standard A mail, increase by varying amounts for mail above and below the breakpoint. For other rate elements, the Service generally relies on updated cost studies and employs traditional rate design objectives. The proposal maintains letter/nonletter differentials at their current levels. Proposed discounts for established worksharing discounts are generally smaller, in absolute terms, than current discounts. In the automation flats category, discounts reflect about 75 percent of their current value, due to reduced cost savings and changes in the automation processing environment.
[5367] The Service proposes to increase the residual shape surcharge (which is assessed primarily on parcels3), and to apply it on a subclass-specific basis. The latter approach is designed, in part, to accommodate the proposed introduction of a 3-cent discount for prebarcoded pieces subject to the surcharge. Thus, the surcharge would increase from a now-uniform 10 cents to 15 cents in the commercial and preferred Enhanced Carrier Route subclasses and to 18 cents in Regular and Nonprofit. In addition to the parcel barcode discount, the proposal includes two other classification changes. Both have limited impact. One extends the automation letter weight limit to 3.5 ounces, thereby allowing pieces between 3.3 ounces and 3.5 ounces to qualify for lower rates. The other applies a pre-selected, rounded breakpoint of 3.3 ounces in the Standard A rate schedules. Proposals in the Special Services area extend eligibility for Delivery Confirmation, Return Receipt for Merchandise and Bulk Insurance Service to Standard A parcels.
[5368] Financial summary. Table 5-8 summarizes the financial impact of the Service's Standard A proposal, based on data in the Service's original filing. Subsequent errata (related to delivery costs) changed some of these figures, but the Service did not revise its proposed rates; instead, it indicates that the implied passthroughs (all within three percentage points of the originals) are consistent with its Standard A rate design objectives and the criteria of the Postal Reorganization Act. USPS-T-35 at 42 (revised).
[5369] Discussion. The record developed in response to the Service's Standard A proposals has been dominated by debate over the ECR pound rate reduction and ECR cost coverage. Participants' views cover a broad spectrum. They range from support for the Service's proposal as the minimum acceptable outcome on this record, to arguments that the pound rate should be kept at the current level or even increased. Reaction to the cost study varies. Some say it is adequate, especially considering the results of statistical tests on related supporting data. Others say it is seriously flawed, and that a different analysis (such as an engineering study) is required. Still others say no study is needed, on grounds that logic and common sense justify a reduction. Assessments of the impact of the reduction on mailers, their customers, and various competitors also diverge.
[5370] There is limited, but pointed, criticism of other aspects of the Service's Standard A proposal. Mailers of parcels vigorously oppose the increase in the residual shape surcharge. There is also opposition to the Service's proposal to pass through less than 100 percent of the savings associated with destination entry discounts; an assertion that the automation flats cost model understates savings; criticism of the ECR letter/flat passthrough; and comment on "address quality" issues. In addition to the cost coverage implications of the ECR pound rate, relative Standard A and First-Class Mail coverages have been questioned. However, the new prebarcoding discount is unopposed. The anticipated RFRA amendment, the 3.3 ounce prescribed breakpoint, and the automation letter rate eligibility proposal have generated little or no comment.
[5371] Commission recommendations. Recommended piece rates are shown in the accompanying rate charts. The Commission concludes that the appropriate level of the commercial pound rates is an issue requiring substantial additional analysis notwithstanding the extensive efforts of participants to explore the issue in this case. The Commission includes in this Opinion a discussion of factors relevant to designing more efficient, cost based pound rates. In the interim, it recommends more moderate changes than the Service has requested.
[5372] Specifically, the Commission recommends a Regular pound rate of 66.8 cents and an ECR pound rate of 63.8 cents. It recommends the changes in the residual shape surcharge the Service has proposed. It also recommends the proposed 3-cent discount for prebarcoded parcels. Recommended destination entry discounts are higher than the Service proposed. The Commission's recommended treatment of the letter/flat differentials varies from retaining the current percentage passthroughs to passthroughs greater than 100 percent to avoid rate shock in adversely affected rate categories. Finally, the Commission's rates reflect the following cost coverages: Regular subclass, 137.47 percent; ECR subclass, 199.45 percent; Nonprofit subclass, 107.51 percent; and Nonprofit ECR, 136.17 percent.
[5373] Although the ECR proposal has garnered most of the attention on this record, all participants have contributed substantially to the Commission's understanding and appreciation of the impact of a number of rate design decisions on users and competitors. To the extent possible, the Commission has attempted to reflect these insights in accommodating various rate design objectives.
[5374] Organization of remaining discussion. A summary of current, proposed and recommended rates follows. The discussion addresses several preliminary considerations, such as post-filing adjustments and the Commission decisions on other aspects of the Service's filing. It then reviews specific aspects of the Standard A proposal; addresses intervenors' alternatives, and presents the Commission's recommendations. Several broad issues that affect Standard A, such as the Service's proposed revenue requirement (especially the 2.5 contingency), cost methodology and cost adjustments, are addressed in other sections of this Opinion.
[5375] Standard A-formerly referred to as third-class mail-consists primarily of advertising circulars, catalogs and product samples.4 Basic eligibility provisions require that each mailing consist of at least 200 pieces or weigh at least 50 pounds. Each piece must weigh less than a pound, meet size and aspect ratio restrictions, and be submitted as part of a mailing that complies with various presentation and preparation requirements. Senders of Standard A must pay an annual bulk mail permit fee.
[5376] Recent volume and revenue figures. In fiscal year 1999, total Standard A volume amounted to 85.2 billion pieces, or more than 40 percent of all domestic mail, and generated $14.4 billion in revenue. USPS-T-6 at 107, and FY 1999 RPW Report. These results establish Standard A as the second-largest class in the U.S. postal system; only First-Class Mail generates more volume and revenue.
[5377] Subclass composition and rate structure. As a result of major restructuring in the mid-1990s, Standard A is comprised of four bulk subclasses. Two-Regular and ECR-are often referred to as the commercial subclasses. The two others-Nonprofit and Nonprofit ECR-are known as preferred subclasses, based on a longstanding statutory rate preference. This preference has changed over time in form and benefit, but is generally understood as an expression of Congressional interest in more favorable rates for mail in the preferred subclasses, relative to their commercial counterparts. While this case was pending, legislation was passed that changes the method of developing rates for the preferred subclasses.
[5378] Each Standard A subclass has the same type of rate structure. This entails a flat rate, referred to as the "minimum per piece," through a certain weight or breakpoint (now 3.3 ounces, rounded). Pieces above the breakpoint (through the 15.99ounce weight limit for the class) are assessed at both per-piece and per-pound rates.5 There are two formal shape distinctions. One is the letter/nonletter differential (referred to as the letter/flat differential below the breakpoint), which reflects cost differences between letters and all other shapes in the class, such as flats and parcels. The residual shape surcharge reflects a portion of the cost differences within the nonletter grouping, as between flats and all other shapes (generally parcels).
[5379] The established worksharing categories recognize presorting to several ZIP Code levels, certain density levels (High Density and Carrier Route) and prebarcoding of letters and flats. The Service proposes extending a prebarcoding discount to parcels in this case. Destination entry discounts (also referred to as dropshipping discounts) recognize entry at three points: bulk mail centers (BMCs), sectional center facilities (SCFs), and destination delivery units (DDUs).
[5380] Service's approach. The Service's proposal for Standard A, presented by witness Moeller, is based on the rate structure and rate design methodology that underlie current (Docket No. R97-1) rates. In general, Moeller's approach to rate design entails judgmental assessments of the statutory cost and non-cost criteria; reliance on cost and pricing support from other Postal Service witnesses; reference to a schematic commonly referred to as the presort tree; and application of a rate formula.6 For specific rates, Moeller generally begins with the passthroughs underlying current rates, with modifications made to meet the Service's rate design objectives within each subclass. USPS-T-35 at 4-5; Tr. 10/3867. These objectives include recognizing worksharing efforts through discounts; limiting the percentage increases for individual rate cells to about 10 percent; monitoring the cells that are "pushed up" to finance these limits; and limiting the reduction in the level of the established discounts, given that they have led to significant mailer investment. Additional objectives include creating (or maintaining) appropriate rate relationships, such as ensuring that the 5-digit automation letter rate is lower than the basic ECR letter rate, but higher than the basic ECR automation letter rate; avoiding rate anomalies; providing for more cost-based rates; and achieving results that are reasonable, in terms of an overall perspective. Tr. 10/3868 and 3871.
[5381] Discussion. Witness Moeller's rate design approach essentially follows the approach that has been developed and found acceptable over the past several rate cases. The Commission accepts it in most respects. In the area of worksharing discounts, the Commission shares the Service's interest in ensuring that certain important rate relationships are maintained. Thus, the recommendation includes a rate for 5digit automation letters in the Regular subclass that meets two key considerations: it is lower than the basic letter rate in ECR, but higher than the ECR automation letter rate. This relationship, according to Moeller, has led to significant, beneficial changes in mail preparation, because mailers no longer have an incentive to prepare 10-piece packages of carrier route presorted mail; instead, those with the density for ECR basic can instead choose to prepare their mail for the 5-digit automation letter category rate. Another important relationship is the "zero" passthrough of shape-related cost differences for letters and flats in ECR. The Commission's rate design incorporates this rate relationship in both the commercial and preferred subclasses. Also, Moeller states that in designing rates, he pays particular attention to maintaining or increasing the incremental differences in the dropshipping discounts (between the destination bulk mail center and destination sectional center facility), in response to concern about the relationship that was expressed in the last rate case. The Commission's recommended dropshipping discounts also reflect this interest.
[5382] The Commission's approach differs in the following respects. First, the Commission's lower average rate increases generally obviate the need for the 10 percent cap that witness Moeller employs. Second, with respect to the letter/nonletter differential, the Commission believes it is appropriate to recognize more of the reported cost difference, where this can be achieved without undue impact. Third, the Commission believes that greater recognition of destination entry savings is appropriate, and therefore recommends a higher passthrough level than the Service has proposed.
[5383] Witness Moeller relies, in part, on the cost trends in witness Daniel's weight-cost study to support his proposed pound rate reductions. This study (presented in USPST28) is based on witness McGrane's Docket No. R97-1 In-Office Cost System (IOCS) analysis, but includes changes witness Daniel says are designed to address the Commission's criticisms of the earlier effort. In brief, Daniel analyzes subclass costs in the mail processing, window service, delivery, transportation, vehicle service and "other" cost components individually by shape and, in total, over all shapes. She asserts, in general, that the results show that increasing weight results in higher total unit cost of handling mail, especially since the proportion of flats and parcels increase in heavier weight increments. She says that although the cost of handling letters tends to increase as weight increases, the costs of handling flats and parcels do not appear to increase as weight increases in the lighter weight increments, but do tend to increase in heavier weight increments. USPS-T-28 at 3.
[5384] Daniel says her cost estimates by weight increment are designed to provide a general indication of how costs are influenced by weight. Id. at 1. She cautions that they "are not necessarily intended to be an exact quantification of costs for every individual weight increment." Id. at 3. (emphasis in the original). Daniel notes that she is dealing in a complex area of causation, and has not been able to control for all variables; thus, she concludes: " . . . while it is possible to analyze the data for guidance in rate design, it is difficult, if not impossible, to isolate precisely the impact of weight on cost or identify the exact unit cost of each ounce increment for three of the major classes of mail." Id. at 4.
[5385] Worksharing discounts. Proposed presort discounts for Regular and Nonprofit are based on mail processing and delivery cost studies by witnesses Miller (USPST24) and Daniel (USPS-T-28). In general, Moeller says these studies show that presorting continues to reduce costs for the Postal Service, and therefore warrants recognition in the rate schedule. These studies also provide support for the proposed letter/nonletter differential. USPS-T-35 at 5. See also USPS-LR-I-95 and USPSLR-I-96. Witness Daniel provides separate cost estimates for ECR.
[5386] The automation discounts for letters and flats are based, respectively, on studies by Miller (USPS-T-24) and Yacobucci (USPS-T-25). Moeller says a changing automation environment for flats (discussed by witness Kingsley) and the difference in characteristics between automation and non-automation flats make estimation of the savings associated with mailer-applied barcodes more difficult in this case. Moreover, he says that Yacobucci's study finds that there is "very little benefit to the barcode alone, and illustrates that flat-shaped mail preparation and rate application may need to be reviewed as the flat processing environment evolves." USPS-T-35 at 13. Moeller says the Service is not proposing a significant change in the rate relationships between automation and non-automation flats in this proceeding, but cautions that Yacobucci's testimony "clearly indicates that the automation discounts are too large, so a reduction is warranted." Ibid. The new automation discount for prebarcoded pieces subject to the residual shape surcharge (primarily parcels) is based on witness Eggleston's updated study (in USPS-T-26) for a similar discount for Package Services.
[5387] Witness Crum (USPS-T-27) supplies cost data for the residual shape surcharge and destination entry discounts. Moeller says Crum's dropshipping study shows that savings associated with this activity have grown by 78 percent since Docket No. MC951. Id. at 10.
[5388] Witness Mayes provides cost coverage analysis for both commercial subclasses. USPS-T-32 at 35-36 and 38-39. Her approach, in line with past practice, assesses the Service's proposals for Regular and ECR in terms of the nine pricing criteria in 39 U.S.C. § 3622(b).
[5389] Commission reliance on underlying cost studies. The most thoroughly contested study is Daniel's weight analysis, which provides general guidance for the proposed pound rate reductions. Witness Miller's study has not been challenged. Several other analyses (such as witness Crum's studies and Daniel's ECR letter/flat differential) have been contested only in limited respects. The Commission has reviewed these studies, and finds that they are generally updates and refinements of the analyses that form the cost support for the current Standard A rate schedule. It accepts them with the following changes: use of 1999 CRA cost updates; substitution of Commission variabilities; and the Postal-Service filed errata associated with witness Daniel's delivery costs.
[5390] In related testimony, PostCom witness Glick and Time Warner witness Stralberg propose adjustments to the Yacobucci model. These are discussed in more detail in the Periodicals section of this Opinion. In brief, witness Glick identifies four errors in the Yacobucci model, and PostCom contends that these show that the Service has understated the test year cost savings for processing Standard A automation flats. Correcting for these errors (and applying them to Stralberg's revised model) yields a net reduction of 1.6 cents per piece for basic and .8 cents for the combined 3/5 digit automation pieces. Postcom does not argue that the revised costs should be substituted, but says they show the Service's proposed discounts are conservative. With the exception of Glick's assumption that the incoming secondary factor should be 70 percent, the Commission has incorporated the Glick/Stralberg adjustments in its measurement of automation flats cost savings.7
[5391] The major classification change in the Service's Standard A proposal is a new automation-related discount. This entails a 3-cent discount for prebarcoded machinable Regular and Nonprofit Standard A pieces that are subject to the residual shape surcharge.8 Pieces subject to the surcharge are those that are not letter- or flat-shaped, or pieces prepared as parcels. Witness Moeller says the proposal responds to the interest Standard A mailers expressed, in Docket No. R97-1, in a discount similar to the 3-cent barcoded parcel discount the Service proposed for Standard B in that docket. At that time, the Service objected to extending the discount to Standard A on grounds that another rate element-the residual shape surcharge-was being introduced, and the impact of a parcel barcode discount had not been taken into consideration in deciding on the amount of the proposed surcharge. Since the residual shape surcharge is now an established rate element, and since mailer-applied barcodes are expected to have some processing value, Moeller says the Service now considers it appropriate to offer this discount. USPS-T-35 at 1-2.
[5392] Cost and volume estimates. The Standard A prebarcoded parcel discount is based on the discount proposed for Standard B in this case, which reflects full passthrough of Postal Service witness Eggleston's modeled cost savings of 2.9 cents. Ibid., see also response to P.O. Information Request No. 3, Question 11 at Tr. 21/8546. The Service estimates that slightly more than one-half billion Standard A parcels will qualify for the discount (approximately 490 million in the Regular subclass and 12 million in Nonprofit). Tr. 10/3852.
[5393] Consistency with statutory classification criteria. Section 3623(c) of title 39, U.S. Code, identifies six classification criteria. Moeller reviews the proposal's consistency with these criteria, and finds that all applicable factors are satisfied. In particular, he asserts that the proposed discount enhances fairness and equity by extending to mailers of Standard A parcels pricing incentives currently offered to similarly processed Standard B parcels. USPS-T-35 at 2. He also says Postal Service processing and efficiency benefit to the extent that additional parcels become barcoded. Ibid. In connection with simplicity, Moeller notes that because there is a separate rate distinction for parcels in Standard A (the residual shape surcharge), administration of an additional rate element applying to these pieces will not be particularly complex. Ibid.
[5394] Discussion. No participant expresses opposition to this proposal, and PSA witness Zimmerman specifically includes this change among those in the Service's proposals that "make some sense" with respect to Standard A parcels. Tr. 29/14137. However, PSA takes issue with what it views as the Service's characterization of the proposal as a means of mitigating the proposed increase in the residual shape surcharge. It states: "The two have nothing to do with each other; the discount is earned and the surcharge is unwarranted." PSA Brief at 37.
[5395] The Commission agrees that a prebarcoding discount should be extended to Standard A mailers on the terms identified in the Service's proposal. It adopts witness Eggleston's study, which (as revised for Commission adjustments and updates) indicates savings of 3.0 cents per piece. The recommended discount of 3 cents reflects 100 percent passthrough. The Commission adopts witness Moeller's assessment of the proposal's consistency with the statutory criteria.
[5396] Proposal. Witness Moeller notes that the Standard A rate design is predicated on the assumption that there will be no effect on costs or revenues if the Postal Service increases the maximum weight for Standard A automation letters to 3.5 ounces.9 He expects a de minimis effect on costs and revenues, as the weight increase is relatively small; however, he says the Service will revisit this assumption after these heavier weight pieces are introduced into the mail processing environment. USPS-T-35 at 12-13. Presumably, the Service would then present its findings and any related data in a future rate case.
[5397] No witnesses, other than Moeller, have specifically addressed this matter. Although this proposal may seem minor, it touches on two significant points: the appropriate Standard A breakpoint and the changing automation environment. In reviewing this issue it must be recognized that the breakpoint is not the result of a cost analysis. Rather it is an arbitrary rate design feature. There is no reason to believe letters or flats slightly above the breakpoint have significantly different characteristics than letters or flats slightly below the breakpoint. While the breakpoint is arbitrary, a single recognized dividing line is fair.
[5398] A change in the breakpoint may be appropriate. An entirely new rate design for Standard Mail may be called for. This issue has not been explored since R801, and in subsection 4.(d.)(6) of this chapter the Commission explains that other aspects of the Standard Mail rate design also warrant study. Before allowing this de facto increase in the breakpoint for certain Standard A mailers (and thereby reducing their rates), the Commission believes a more comprehensive assessment is needed. Therefore, the Commission does not recommend the requested extension of eligibility for minimum per piece rates to automation letters between 3.3 ounces and 3.5 ounces.
[5399] The Standard A rate structure gives rise to a transitional weight, referred to as "the breakpoint," which marks the crossover from the minimum-per-piece portion of the structure to the piece-pound portion. Moeller explains that current practice, in line with the traditional approach, is to calculate a breakpoint weight to the ten-thousandths place to create a perfectly smooth transition from the minimum-per-piece rates to pound-rated rates.10 Id. at 9. He says that prior to the introduction of destination entry discounts (in 1991), this approach worked well, because there was only one transitional pound rate for all pound-rated pieces. With dropshipping discounts now established elements of the rate structure, however, Moeller notes that there are effectively as many as four pound rates in a given subclass, and 14 different pound rates in Standard A. Ibid. Thus, although the non-destination entry pound rate has continued to be used as the reference point to calculate the precise breakpoint that allows for a perfectly smooth transition, Moeller says it does not provide for a smooth transition for all destination entry pound rates.
[5400] Given this development, Moeller says the Service anticipates that a rounded breakpoint of 3.3 ounces will be used in the rate schedules when new rates are implemented. In support of this approach, he observes that "we are not sacrificing much precision, if any, by merely using the selected breakpoint of 3.3 ounces as the prescribed breakpoint; [given that] its precision has long been confined exclusively to non-destination entry categories." Id. at 9-10 (footnote omitted). Moeller also says this approach offers the advantages of practicality and simplicity. For example, he says the Service generally chooses the highest of the four breakpoint weights to use for the maximum weight of automation letters in other classes of mail, so using 3.3 ounces for all breakpoints will avoid changes in the weight limit with every rate case. In addition, he says it will eliminate expression of the breakpoint with a cumbersome figure carried to a tenthousands of an ounce decimal place. Ibid.
[5401] Commission recommendation. As witness Moeller indicates, the introduction of destination entry discounts has effectively eliminated the application of a single breakpoint to the entire Standard A subclass. Therefore, the use of a breakpoint with four decimal places, which was adopted in the interest of providing a smooth transition, has lost essentially all of its original significance. Simplicity and practicality are also valid considerations in rate administration.
[5402] The Commission finds that the proposed change is well-supported, and recommends it. To clarify this change and its application in the rate schedule, the Commission adds the following language in the footnotes to the Standard A rate schedules: "The transitional weight between the minimum per piece rate and the pound rate in Standard Mail is 3.3 ounces."
[5403] The Commission does not endorse 3.3 ounces as a permanent breakpoint for Standard A. This question-the appropriate transitional weight-should be addressed in the broad rate design inquiry the Commission believes is needed.
[5404] The Regular and ECR subclasses, which were established as a result of the Docket No. MC95-1 reclassification case, are successors to the former third-class mail Regular-Rate subclass. Moeller says mailings in the Regular subclass are sometimes viewed as demographically targeted, while those in ECR are often seen as geographically targeted. USPS-T-35 at 18. In fiscal year 1999, the combined volume of the commercial subclasses amounted to more than 71.3 billion of the 85.2 billion pieces in Standard A as a whole. The Regular subclass, with 38.5 billion pieces, generated slightly more than half of the commercial volume; ECR, with 32.8 billion pieces, generated slightly less than half. USPS-T-6 at 107. This volume (without fees) produced revenue of $12.8 billion. FY 1999 RPW Report.
[5405] The Postal Service proposes rates that reflect, on average, revenue increases of 9.4 percent for the Regular subclass. USPS-T-35 at 2. The proposal is based on the prevailing rate structure; assumes a 3.3 ounce breakpoint; applies the rate formula underlying current rates; reflects an 18-cent residual shape surcharge; and a 3-cent prebarcoding discount for machinable pieces subject to the residual surcharge. Id. at 67 and 9.
[5406] Rates. For mail below the breakpoint, the Service proposes the changes identified in Table 5-9. For mail above the breakpoint, Moeller proposes reducing the pound rate from 67.7 cents to 66.1 cents. Id. at 8. In support of this change, which he characterizes as a "modest reduction" of 1.6 cents, Moeller notes that the pound rate traditionally has been considered a proxy for a surcharge on parcels. However, he notes that the introduction of the residual shape surcharge-and the increase proposed in this case-reduce the need for the pound rate to act as a proxy for a parcel surcharge. Moeller also asserts that the proposed decrease is not inconsistent with Postal Service witness Daniel's study of the effect of weight on costs. Ibid. Corresponding piece rates for mail above the breakpoint are shown in the Table 5-9.
[5407] Letter/nonletter differential. The proposed letter/nonletter differentials in the Regular subclass are identical to those in existing rates; however, Moeller says this reflects an interest in moderating rate increases for individual rate categories, rather than a preference for a specific degree of shape recognition. He considers the resulting passthroughs (of 77 percent in basic and 64 percent in the combined 3/5 digit tier)11 reasonable, as they not only recognize a significant portion of the cost differential, but also reflect the Commission's previously expressed intentions to increase them. He also notes that if the passthroughs were larger, the percentage increase for the category receiving the highest increase in the Service's overall proposal-the minimum-per-piece rate for 3/5 digit presorted automation flats-would likely be even higher. Id. at 5-6.12
[5408] Presorting. Moeller notes that once the letter/nonletter passthroughs have been established, selection of a presort passthrough for letters is the pivotal rate design decision, because it dictates the effective passthrough for nonletters and several other tiers. Id. at 10. Relying on witness Miller's cost avoidances, Moeller attempts to justify a variety of passthroughs.
[5409] Automation letters. In the automation letters category, current rates reflect full passthrough of the measured cost avoidance. Moeller departs from this approach (in the direction of passthroughs greater than 100 percent) in all three tiers. Passthrough in basic is 110 percent (which preserves 80 percent of the existing discount); 106 percent in the 3-digit tier; and 160 percent in the 5digit automation tier. Id. at 11. Moeller cites three reasons for proposing unconventional passthroughs. First, the Regular automation letter discounts were reduced as a result of Docket No. R971, and would be further reduced if passthrough of cost avoidance estimates is limited to 100 percent. Second, this approach moderates the percentage increase for the 3-digit category which, at 9.7 percent, is above the average for the subclass and the highest increase for automation letters. Third, the proposed 3-digit tier passthrough helps achieve the desired 5-digit automation rate relationship with ECR basic. Id. at 11-12.
[5410] Automation flats. Moeller asserts that decreased savings, due to a changing automation environment and higher reported costs, would warrant a substantial reduction in the current discounts for automation flats. To mitigate the impact on mailers, he proposes maintaining them at about 75 percent of the current level. Id. at 13. This results in unconventional passthroughs of 230 percent for basic automation flats and of 500 percent for 3/5-digit automation flats. USPS-T-35, Workpaper 1.
[5411] Destination entry discounts. Moeller proposes an increase in the absolute level of the discounts, consistent with the increase in savings, but a slight reduction in the percentage passthrough relative to current rates. However, he says passthroughs-ranging from 73 to 78 percent-continue to encourage dropshipping and allow him to maintain or increase the incremental discounts. USPS-T-35 at 14-15. The proposed passthroughs are 73 percent for DBMC and 77 percent for DSCF. Moeller says that if 75 percent were used for both, the incremental differences could not be maintained. He also says a greater passthrough would result in a larger increase in the basic rates, which conflicts with the general guideline of tempering individual rate increases. Id. at 15.
[5412] Commission rate design. No participant opposes the Service's proposed minimum-per-piece and piece-pound rates in the Regular subclass. The Commission's rate design for Regular follows witness Moeller's in many respects. However, using updated costs and the Commission's costing methodology generally produces slightly larger cost differentials. This allows the Commission to adhere to a cost based methodology for presort and automation letters by using 100 percent passthroughs. To avoid undue rate increases for automation flats, and to recognize that the value of these flats will likely be higher in the test year than the Postal Service anticipates in its filing, the Commission recommends the Postal Service's proposed letter/flat passthrough and passthroughs greater than 100 percent for automation flats. This is consistent with the Service's proposal.
[5413] The Commission also recommends slightly higher passthroughs for destination entry discounts, but this is based on updated costs and the Commission's costing methodology. This recognizes more of the cost differences, and prevents the incremental differences between the dropshipping tiers from shrinking. Shrinking these differences would adversely affect mailers who dropship.
[5414] Witnesses Glick and Schick, testifying on behalf of the Association for Postal Commerce and Mail Advertising Service Association, Inc., support full passthroughs in destination entry discounts. Tr. 32/15716-18 and 15703-07. As witness Glick's testimony indicates, the Commission's stated preference is for full passthrough. However, the Commission finds-as does witness Moeller-that this interest cannot be accommodated in its entirety, due to the impact on other rates. However, the Commission does recognize more of the savings than the Service's rate design, passing through 84 percent of the savings (determined on the Commission's costs) for BMC entry, 84 percent for SCF entry, and 82 percent for DDU entry.
[5415] Cost coverage. Moeller notes that Postal Service witness Mayes proposes a cost coverage of 132.9 percent for Standard A Regular, which results in an average rate increase of 9.4 percent. In support of the proposed coverage, Mayes says that Regular, like the other Standard subclasses, has a relatively low intrinsic value of service (criterion 2), due to its deferability, use of ground transportation, lack of access to the collection system, and absence of free forwarding. USPS-T-32 at 35. She states that although the Service may attempt to satisfy mailer requests for delivery within a specific time frame, these typically involve advance planning and coordination by the mailer to facilitate the achievement of these requests. Ibid.
[5416] With respect to the cost floor requirement of criterion 3, Mayes notes that at projected test year after rates volumes, the $9,070 million revenue from the subclass easily exceeds its estimated incremental cost of $6,938 million. Id. at 37. She says the price elasticity for Regular (at -0.570) is higher than estimated in Docket No. R97-1, and higher than that of First-Class Letters. Ibid. However, she states that it is lower than that of ECR, and concludes that this suggests "an intermediate economic value of service." Id. at 36. She also indicates that the Service hopes that the availability of new ancillary services (Delivery Confirmation, Return Receipt, and Bulk Insurance) will slightly increase the value of service for some Standard A mailers.
[5417] With respect to criterion 4, Mayes posits that the 9.4 percent increase is above the rate of inflation, and higher than the system average of 6.4 percent. She says this results "in a noticeable, but reasonable, impact" on the users of Regular mail but, considering the 132.9 percent cost coverage over volume variable costs, also "suggests that competitors are not unfairly targeted by this increase." Ibid.
[5418] With respect to criterion 5, Mayes notes the availability of alternatives for demographically targeted advertising including special-interest magazines, cable television, and internet websites. Ibid.
[5419] Mayes states that mail within the Regular subclass has a substantial degree of mailer preparation, which must be considered by virtue of criterion 6. Id. at 36-37. With respect to criterion 7, Mayes says the rate schedule for Standard A is designed to offer a range of rates to reflect the varying ways that the mailers may choose to perform worksharing, which means that the rate schedule is not particularly simple. Id. at 37. However, she states that as the rates for Standard A only apply to bulk-entered mail, senders tend to be sophisticated users of the postal system or utilize the services of those more expert in postal matters, permitting criterion 7 considerations (generally expressed in terms of simplicity) to be manifested more in terms of creating reasonable and identifiable rate relationships, rather than a limited number of rates. Based on the foregoing evaluation, Mayes concludes that the proposed rate level is fair and equitable (criterion 1).
[5420] Relative First-Class Mail/Standard A cost coverages. No Standard A witness has specifically addressed witness Mayes' analysis; however, to finance his rate proposals for First-Class Mail, ABA/NAPM witness Clifton proposes an increase in Standard A cost coverages. Specifically, he would increase cost coverages for Standard A Regular by 9.3 percentage points (from 132.9 to 142.2) and for Standard A ECR mail by 5.6 percentage points (from 208.8 to 214.4 percent). Clifton's rationale for these adjustments draws on a review of relative cost coverages of presorted First Class Mail and Standard A since 1994 and on assertions regarding First-Class Mail's contribution to institutional costs since 1990. Tr. 26/12458-60.
[5421] Witness Prescott says Clifton's proposal is not supported. He asserts that, by definition, cost coverage for a given subclass of mail is the ratio of revenues to volume variable costs for that subclass of mail. He observes that increases in cost coverages, therefore, can occur either through an increase in revenues, a decrease in costs, or a combination of both. Prescott says the increase in First Class Mail cost coverage comes about in large part from lower costs; he says a cost coverage increase in First-Class Mail does not, as witness Clifton asserts, mean that it is being singled out for discriminatory rate increases. Tr. 44/19316. The Commission is not accepting witness Clifton's First-Class Mail proposals, so his interest in adjusting Standard A coverage is generally moot. However, the Commission notes its agreement with Prescott's observation that an increase in cost coverage can result simply from the mechanics of the ratio, and is not, in itself, indicative of discriminatory treatment.
[5422] Commission cost coverage. The Commission finds witness Mayes' analysis well reasoned. The Commission recommends rates that provide a cost coverage of 137.4 percent, on a cost basis updated for FY 1999 CRA costs and calculated according to Commission methodology.
[5423] Background. In Docket No. R97-1, the Postal Service proposed-and the Commission recommended-a uniform 10-cent surcharge for pieces that were not letter- or flat-shaped, or were prepared as parcels. The impetus for a surcharge on these "residual shape" pieces was the Commission's concern, expressed in PRC Op. MC95-1, that they were not covering their costs.13 This concern was borne out in Postal Service witness Crum's Docket No. R97-1 study, which showed an estimated cost difference of 35.2 cents (using Postal Service costs). Based on several considerations (including impact on affected mailers and fairness), the Commission agreed with the Service's proposal to pass through a fairly low percentage of the cost difference, and recommended a 10-cent surcharge. PRC Op. R97-1, paras. 5476-5489.
[5424] In this case, Moeller says witness Crum's update of his original study conclusively demonstrates that "there continues to be a measurable difference [now 65.5 cents] between the costs for flat-shaped pieces and the costs for the remaining pieces in the non-letter categories of Regular and [ECR] mail." USPS-T-35 at 6. Based on this cost finding, Moeller says the Service proposes increasing the current 10cent surcharge to 15 cents in ECR and to 18 cents in Regular. He says the increase furthers the goal of greater recognition of the cost difference, and its application on a subclass-specific basis takes into consideration the proposed 3-cent discount for prebarcoded parcels. Ibid.
[5425] Moeller says the surcharge reflects a passthrough that equates to 27.5 percent, based on the methodology the Commission used in Docket No. R97-1. He notes that this is similar to the 24 percent passthrough that underlies the current surcharge, but results in an increase because it is applied to the higher cost base that witness Crum has calculated. Moreover, he says: "Ideally, a greater passthrough would be proposed; however, in order to moderate the impact on mailers, an even greater per-piece increase - is not proposed at this time." Id. at 7.
[5426] Moeller also notes that mailers in the Regular and Nonprofit subclasses can now partially offset the 8-cent increase in the surcharge by tendering parcels that qualify for the 3-cent prebarcoding discount the Service has proposed for parcels that meet automation eligibility requirements. Id. at 6. With respect to ECR, Moeller says witness Crum demonstrates that the cost differential greatly exceeds the proposed surcharge and, unlike Regular, the pound rate does little or nothing to generate much revenue from ECR parcels relative to flats. Id. at 23-24. He also notes that parcel-shaped pieces are excluded from ECR unless they are merchandise samples, so the only surchargeable pieces in this subclass are merchandise samples. Id. at 24.
[5427] Moeller observes that the Commission's Opinion in Docket No. R97-1 acknowledged that an alternative way of considering this issue is to examine the revenue/cost relationship. He does not endorse this alternative, but says this comparison not only results in a revenue shortfall, but also fails to consider the offsetting effects of the lower pound rate and the proposed parcel barcode discount in the Regular subclass. Id. at 7. Moeller offers the further observation that even if the proposed surcharge resulted in an implicit cost coverage on surcharged pieces that slightly exceeds100 percent, the Service does not think it undesirable for Standard A parcels to make some contribution to institutional costs, especially given the proposed extension of several Special Services, such as Delivery Confirmation, to pieces subject to the residual shape surcharge, but not to "other, less contribution-challenged pieces." Id. at 7-8. Moreover, Moeller says: "[F]uture prospects for a larger residual shape surcharge seem probable even if the cost differential does not increase. The proposed passthrough was suppressed to 27.5 percent in order to moderate the rate increase on parcel mailers. A higher passthrough applied to the same cost differential in the next rate case would, by itself, result in a higher requested surcharge." Tr. 10/3845 (emphasis in original).
[5428] Volume and costs. Moeller says he does not expect the incremental increase in the surcharge to have much effect on parcel volume. Id. at 3849. With respect to costs, witness Crum identifies two major occurrences that have had opposing impacts on his results. First, an implementation decision has allowed parcels less than 1.25 inches thick to qualify for the flat automation rate if they meet all other criteria for that rate and are properly prepared. Thus, Crum says that the surcharge is not applicable to some unknown number of parcels between .75 inches and 1.25 inches thick that are prepared as automated flats. USPS-T-27 at 7-8. He says: "The logical conclusion, then, is that the pieces still subject to the surcharge will have a higher cost than those presented in this analysis and my estimate of the cost difference is conservative." Id. at 8. The other consideration is a difference between Docket No. R97-1 and this case in the way mail processing costs are calculated. Explicit econometric-based volume variability factors were part of the Service's mail processing cost presentation in that case, and Crum notes that this is not the case in this docket for effectively all of the parcel operations and some portion of the flats operations. He says this expands the cost difference between flats and parcels beyond the level that resulted under the R97-1 volume variability proposal. Ibid.
[5429] Discussion. District Photo, Inc., Mystic Color Lab and Cox Sampling (appearing jointly as DMC), MOAA, PSA, PostCom and RIAA strongly oppose the Service's residual shape surcharge proposal. DMC questions the costs that underlie the increase. DMC Brief at 12-15. RIAA contends that the Commission should recommend a surcharge for Regular mailers that does not exceed 13 cents before application of the proposed discount for prebarcoded parcels. RIAA Brief at 2. PSA contends that there is no justification for surcharge treatment, in the first instance, and that the Service's attempt to impose an 80 percent increase in this case contravenes criterion 4 of the Postal Reorganization Act (effect of increases on mailers). PSA Brief at 34-36.
[5430] PSA witness Zimmerman objects to the size of the increase and reiterates several points that were raised when the surcharge was first proposed. Among other things, he objects to the Service's "blending" into one category, both for surcharge purposes and for cost purposes, four separate subclasses of mail. Tr. 29/14137. He also says the PRA does not require that each piece of mail within a rate category fully cover attributable costs. Id. at 14164. Zimmerman further questions the "deliberate avoidance of comparing the amount of revenue per piece contributed by these pieces versus flat-shaped pieces, but constantly emphasizing the cost differences between parcel shaped pieces and non-parcel shaped pieces." Id. at 14139. He notes that to the extent a shortfall exists, the "difference is small and the remedy . . . is large." He also contends that there is some question as to the "fragility of Service's cost estimates." Id. at 14140. In particular, he says the Postal Service cannot reconcile what he refers to as "the absurd finding" that a nonprofit ECR parcel costs three times as much as a Regular parcel to process. Ibid.
[5431] RIAA, through witness Glick, presents a theoretical basis for using the revenue differences between flats and parcels in determining the residual shape surcharge. Glick notes that Moeller has used a "traditional passthrough" approach in setting the residual shape surcharge in this case. He contends that "an exact piece" comparison is a more appropriate analogy. Tr. 23/10392. Glick further states that while Crum's general analytical approach seems reasonable, he did not adjust for differences in weight. Id. at 10392-93. He says this may be reasonable for comparisons of mail of approximately the same weight, but is inappropriate here because the average Standard A commercial parcel weighs 2.5 times as much as the average Standard A flat. He contends that this amounts to double-charging parcels for weight-related costs. Id. at 10393. In the absence of reliable cost data by shape and weight increment, Glick asserts that the appropriate approach for this adjustment is to use the weight-related revenue difference between flats and parcels as a proxy. He identifies this as 20 cents per piece, based on current rates. Ibid.
[5432] With respect to revenue, RIAA says the "entirety of the revenue effect analysis performed by Postal Service witness Moeller is based on FY 1998 data." RIAA Trial Brief at 5. It says that since the surcharge did not take effect until after the conclusion of that fiscal year, these data are of little probative value. Ibid. However, it says there is nothing in the FY 1999 data that either calls into question witness Glick's explanation of why revenue differentials must be taken into account in setting the surcharge or cures the infirmities in the Service's showing. Id. at 5-6.
[5433] Postal Service's response. The Service dismisses witness Zimmerman's criticism regarding the "blending" of subclass costs, stating that the approach in the Service's proposal is consistent with the current Commission-accepted methodology for establishing a cost basis for surcharge. It further argues that witness Zimmerman offers no record basis for an alternative approach. Postal Service Brief at VII-189-190. Also, it states that witness Crum shows that even if the four subclasses were viewed independently, the proposed surcharge is still less the FY 1998 cost-revenue differential. Thus, it says attempting to tailor the surcharge for each subclass "would appear to be a fruitless exercise, since in all instances, the surcharge would not provide adequate revenue to cover the costs of residual shapes." Id. at VII-190.
[5434] The Service also rejects Zimmerman's contention that there is no requirement that all pieces within a subclass cover their own costs. It takes the position that this is not only inconsistent with the Commission's previous legal conclusions regarding the fairness and equity criterion, but also "defies logic and common sense." Ibid. The Service also says Zimmerman's criticisms fail to acknowledge the balanced nature of the Service's presentation in this docket.
[5435] With respect to witness Glick's testimony, the Service acknowledges that Glick's "exact piece" comparison may warrant further consideration, but says he has simply provided a theoretical basis for using revenue differences to determine the rate differential. Therefore, the Service says there is no record basis for adopting this approach in this case. However, it takes issue with Glick's criticism that Crum's cost study does not consider the impact of weight. In particular, the Service says that Crum's study in this proceeding reconfirms the finding in his Docket No. R97-1 analysis that in ECR, where the weight is nearly identical between parcels and flats, costs for flats and parcels differ greatly. Therefore, it argues that the Commission's findings in Docket No. R97-1 regarding the suitability of witness Crum's original study should apply with equal force in this docket. Id. at VII-193.
[5436] Commission recommendation. Several objections raised on this record were also presented and resolved in Docket No. R97-1. In essence, these include arguments that there is no cost coverage requirement below the subclass level; that costs should not be "blended"; and that other mailers have not objected to "averaged" costs. The Commission has once again considered the validity of these arguments, but finds no sound reasons to depart from its previous conclusions. In general, the Commission continues to believe that overall considerations of fairness and equity and an interest in cost-based rates overcome opponents' objections.
[5437] Given the conclusion that a residual shape surcharge is an appropriate element of the Standard A rate structure, the main question on this record is whether the amount of the increase the Service has proposed is appropriate. The Commission agrees with opponents that the increase seems large, but finds that witness Crum's results, which reflect improvements in the original analyses, appear to be generally reliable. In this regard, the Commission notes that the cost volatility DMC has pointed out is considerably muted when the costs of the individual subclasses are aggregated, as they are here. The Commission has considered whether an even lower passthrough would be appropriate, but finds that this would conflict with appropriate interests in recognizing cost differences.
[5438] The Commission also has considered opponents' arguments that the revenue generated by parcels should be considered. However, as the Service points out, even by this measure these parcels fail to cover cost by a considerable margin. Accordingly, this argument does little to advance parcel mailers' position. The Commission also agrees that witness Glick's "exact piece" comparison may warrant further consideration, but there is no basis for adopting that approach on this record.
[5439] The remaining question relates to the Postal Service's proposal to apply the surcharge on a subclass specific basis, in recognition of the proposed introduction of the parcel barcode discount. The Commission is recommending the proposed 3-cent parcel barcode discount; therefore, the recommended residual surcharges are 15 cents in both ECR subclasses and 18 cents in Regular and Nonprofit. The Commission considers this approach an improvement over the existing single surcharge.
[5440] The Postal Service proposes Standard A ECR rates that reflect, on average, revenue increases of 4.9 percent, with a cost coverage of 209 percent. USPS-T-32 at 38. Moeller's proposed rate design assumes retention of the ECR subclass and the established rate structure. It also reflects the Service's cost and pricing assumptions; employs the rate formula underlying current rates; includes a 3.3 breakpoint; and includes the proposed residual shape surcharge of 15 cents. Id. at 9.
[5441] Minimum per piece rates for mail below the breakpoint. Current and proposed minimum per piece rates are shown in the rate chart appearing at Table 5-10.
[5442] Piece and pound rates for mail above the breakpoint. Moeller's rate design produces increases in the piece rates for ECR mail above the breakpoint as shown in the rate chart at Table 5-10.
[5443] Moeller proposes a corresponding pound rate of 58.4 cents, which is a reduction of 7.9 cents from the current 66.3 cent rate. Id. at 10 and 19. Moeller says this proposal, which he characterizes as a "moderate reduction," addresses the Commission's prior concerns about competition and should also address objections raised by enterprises in the private sector that offer alternatives. Id. at 19-20. Specifically, he says: "The moderate reduction in the pound rate is designed to allay concerns for those that contend they may be disadvantaged by a significant reduction in the pound rate. Yet the moderate reduction acknowledges the needs of small businesses who rely on the mail, or wish to use the mail, for affordable advertising." Id. at 23.
[5444] As discussed within, the Commission agrees that a reduction in the pound rate is appropriate, but does not recommend as large a reduction as the Service proposes. Instead, it proposes 63.8 cents.
[5445] Letter/flat differential. In line with the Commission's recommendation in Docket No. R97-1, Moeller's proposed ECR design has no rate differential between basic letters and flats. Id. at 18-19. Moeller says this approach (a "zero" passthrough, in effect), coupled with rate distinctions at the other tiers, balances recognition of cost differences with the Service's interest in fostering its letter automation program. Id. at 24-25. At the other levels, Moeller proposes passthroughs of 65 percent at high density and 95 percent at saturation. Id. at 25.
[5446] Witness Haldi, sponsored by Val-Pak Direct Marketing Systems, Inc., Val-Pak Dealers' Associations, Inc., and Carol Wright Promotions, Inc., appearing jointly as Val-Pak, et al., contends that the Service's proposed ECR letter/flat differential is understated. The Commission addresses his position below. In brief, it finds that the error Haldi has identified would have a de minimis effect; therefore, it does not impact the Commission's recommendation.
[5447] Worksharing discounts. For the basic automation letter discount in ECR, Moeller proposes full passthrough of the reported cost differential in the underlying cost study. He proposes passing through 125 percent between the basic and high density tiers, and 100 percent between high density and saturation. Given the presort tree and shape passthroughs, the corresponding passthroughs for nonletters are 63 percent between basic and high density and 84 percent between high density and saturation. Id. at 25.
[5448] Moeller says these passthroughs are consistent with those underlying current rates, with the exception of the 125 percent passthrough between basic and high density letters. Id. at 25-26. He attributes the exception to the Service's interest in mitigating the effect of the "zero" shape passthrough at the basic tier; in maintaining the current per-piece rate differential between basic and high density letters; and in tempering the percentage change for the high density and saturation tiers. Moeller says that if 100 percent were used, the increase for saturation letters would be 12.3 percent, or more than double the subclass average. He says the proposed approach also allows for greater recognition of the cost difference between basic and high density flats. Id. at 26.
[5449] Destination entry. Moeller passes through 73 percent of reported savings for DMBC entry, 77 percent for DSCF entry, and 77.5 for DDU entry, for the reasons identified in the Regular subclass discussion. Id. at 26-27.
[5450] Commission rate design. The Commission accepts the basic rate design methodology that underlies the Service's ECR proposal, and recommends the proposed 15 cent residual surcharge. With respect to worksharing elements, the Commission finds that adhering to 100 percent cost passthroughs is problematic for ECR letter categories because the cost differentials have increased dramatically between FY 1998 and FY 1999, as Table 5-13 shows.
Based
on BY 1999
Growth[5451] Some of the growth in these differentials is due to the difference between the Postal Service's and the Commission's costing methodologies, but this only accounts for a small portion. Growth of these magnitudes in cost differentials makes rate design difficult, since it is generally desirable to have minimally disruptive rate changes. The Commission must also consider the possibility that the cost differences obtained in FY 1999 might shrink in the future if FY 1999 ECR letter data are anomalous, or there are operating conditions in FY 1999 that will not exist in the future or the test year. For this reason, the Commission is recommending rates based on 75 percent passthrough of the cost differences between ECR Basic letters and ECR Basic Automation letters and between ECR Basic letters and ECR High Density letters. In the future, if these costs show abrupt year-to-year changes, the Postal Service should consider using a three- to five-year moving average to smooth changes in affected rates.
[5452] In other categories, the Commission's recommended rate reflects a 100 percent passthrough for saturation letters, and for the letter/flat differential at the high density and saturation density levels. To support the Postal service's letter automation program, the Commission continues the policy of adopting a zero percent passthrough of the letter/flat cost differential for the basic density tier. By virtue of the presort tree, these passthroughs produce a 56 percent passthrough for high density flats and a 100 percent passthrough for saturation flats. The Commission's recommended rates are as cost based as possible given the underlying costs and the considerations outline above.
[5453] The most controversial issue in Standard A is the Postal Service's proposal to reduce the ECR pound rate from 66.3 cents to 58.4 cents. According to Moeller, the rationale for this reduction draws on "general guidance" from witness Daniel's cost study and several other considerations. These include: (a) an implicit coverage comparison, which shows that the reduction does not distort relative coverages for piece-rated mail and pound-rated mail; (b) alleviation of an "illogical" result under the current pound rate, where postage for two 4-ounce pieces is only three-tenths of a cent more than an 8ounce piece; (c) recognition that the pound rate is no longer needed to act as a proxy for a parcel surcharge; and (d) the conclusion that the moderate nature of the decrease addresses concerns about the impact of the proposal on mailers, their customers, and competition.
[5454] AAPS, NAA and Val-Pak, et al. oppose the proposed ECR pound rate reduction. Their opposition, expressed on brief and through the testimony of AAPS witness White, NAA witnesses Tye and Wilson, and Val-Pak, et al. witness Haldi, focuses on four main arguments. These include: Daniel's cost-weight study is unreliable or inadequate; Daniel's regression analysis is not useful, and others presented on this record add confusion to the record; Moeller's implicit cost coverage comparison is not reliable and proves little, since the Service does not use it elsewhere; and the proposal interferes with competitors, in contravention of 39 U.S.C. § 3623(b)(4) and other considerations.14
[5455] Supporters of the Service's ECR proposal include the Association of Independent Store Owners and Professionals (AISOP); the Saturation Mail Coalition (SMC); Mail Order Association of America (MOAA), the Association for Postal Commerce (PostCom) and the Direct Marketing Association, appearing jointly as MOAA, et al. Witness Crowder, on behalf of Advo, and witness Prescott, on behalf of MOAA, et al., address technical aspects of the Service's proposal and of opponents' criticisms. AISOP and SMC sponsor industry witnesses who address impact arguments.
[5456] The Postal Service sponsors witness Bozzo (USPS-RT-18) in further support of the use witness Moeller makes of Daniel's cost findings, and sponsors witness O'Hara (USPSRT19) regarding the Service's motivations for proposing a reduction in the pound rate and competition issues.
[5457] The debate over the Service's ECR pound rate proposal has consumed considerable time and effort, and has covered numerous complex, and often inter-related, topics. The Commission believes this effort has been extremely worthwhile, regardless of the position one takes on the merits of the Service's proposal or the Commission's recommendation.15 In particular, there is now a much clearer picture of the type of data and information that would materially advance the understanding of pound-rate cost causation. In the interest of obtaining and evaluating that type of data and information, the Commission calls on the Service to conduct a new analysis addressing the matters described at the conclusion of this section. This will provide the Commission and all interested participants with a sounder basis for addressing key matters left untouched in the Daniel study and witness Moeller's rate design testimony.
[5458] The broad sweep of written and oral testimony, discovery responses, library references, and other material forecloses summarizing every argument or position that has been raised. Therefore, the discussion attempts to identify basic positions and address the points most pertinent to the Commission's analysis. In the discussion of competitive impact, the SAI report-and some arguments on brief-were filed under seal. The Commission has considered this material, and finds that it can address relevant issues in this Opinion, on the public record, without interfering with the interests that prompted confidential treatment.
[5459] In essence, the following discussion indicates that the Commission finds that the Daniel study is not dispositive on the question of the appropriate pound rate. The Commission directs the Service to conduct a study to obtain the data and information needed to reach firm conclusions on this important point. At the same time, the Daniel study (as supplemented by later-filed data) has addressed some of the Commission's reservations about earlier studies. Tally thinness, for example, has been explored, and placed in proper perspective. In brief, minor technical criticisms lodged against study have not withstood scrutiny.
[5460] The Commission finds the study does not point to a clear measure of the impact of weight on the cost of ECR mail, but it does allow the Commission to have a degree of confidence that the pound rate it is recommending is not out of line with apparent cost trends. Given that the Service and the parties are urged to explore additional cost and policy issues, the Commission is not yet able to form a permanent position on the appropriate pound rate.
[5461] The Commission finds that several considerations, not directly related to the study, point to the appropriateness of a modest reduction in the ECR pound rate.16 These include (1) the demonstration that the current pound rate produces an illogical postage result, inconsistent with notions of fairness and equity and efficient postal operations; (2) the recognition that reclassification has reduced the need for the pound rate to act as a proxy; and (3) the demonstration that the pound rate "over recovers" due to shape. The first two points are largely self-evident; with the respect to the latter point, the Commission notes that witness Crowder's contention that the ECR letter/flat cost differential reflects differences due to shape and weight has merit. Unlike the Regular subclass rate differentials, which are based on special mail processing cost studies, ECR cost differentials are based directly on costs from the CRA. The mail processing costs are thus from IOCS. When comparing one CRA unit cost to another, the difference may be due to several factors related to the characteristics of the mail. By developing unit costs for each density level, the Service eliminates differences due to density level. Using the results of its destination entry cost study, the Service also eliminates differences due to dropship patterns that may differ by density level and shape. However, because the weight of letters and flats varies, the letter/flat cost differential by density level likely reflects differences in both weight and shape. As the pound rate is supposed to reflect the effect of weight on cost, passing through a substantial portion of the ECR letter/flat differential amounts to a double counting of the effect of weight. Thus, in part because the Commission is adopting a 100 percent passthrough of the letter/flat cost difference, a reduction in the pound rate for ECR is justified. The recommended reduction-of 2.5 cents-is intended to offset double counting inadvertently captured in the letter/flat rate differential.
[5462] It should also be noted that Val-Pak, et al. enhanced the record by emphasizing that the cost per pound should be marked up when it is used as an input to the rate design formula. They noted that pound costs in other subclasses are marked up in the rate design process, including First-Class, Priority, Periodicals, and Standard B. Val-Pak, et al. Brief at 23. Prior to the Val-Pak, et al. Brief, the participants were focused on only the cost per pound, either through a bounding analysis or through regression analysis. Including a markup in the pound rate is logical since all the rates should recover attributable cost plus markup. Thus, when the Postal Service proposes a pound rate and when the Commission recommends a pound rate, there is an implicit cost coverage attached to both the piece rate and pound rate for mail above the breakpoint. The importance of implicit cost coverage is discussed further below.
[5463] The Commission has evaluated the arguments participants have presented on the impact of the Service's proposal. Importantly, notwithstanding the concerns posed by the late production of a confidential internal market analysis, there is no evidence that the pound rate reduction was motivated by an improper Postal Service interest in harming competition. Instead, quite apart from the cost study itself-which was clearly a consideration for the Service-there are legitimate reasons that support the Service's interest in seeking approval of a postage rate it views as more fair. Based on the descriptions of industry practice provided on this record, the pound rate the Commission has recommended should foster competition.
[5464] Cost support. Witness Moeller acknowledges the criticisms of witness McGrane's study the Commission articulated in PRC Op. R97-1, and asserts that Daniel's new weight-cost study responds to these concerns in two ways: by improving the distribution of mail processing costs by weight increment, and by using weight, rather than pieces, as a distribution key for the elemental load time component of carrier street delivery time. Moeller observes that in the previous study, the costs of tallies where the weight of the piece is not known were distributed using the aggregate distribution of the tallies with known weight increment. He says that in the current study, Daniel performs this distribution within cost pool and IOCS activity code using the information for tallies with known weight increment. USPS-T-35 at 20, citing USPS-T-28 at 5 and 8.
[5465] The implicit cost coverage comparison. Moeller notes that cost coverages are not required below the subclass level, but says he finds them illuminating in this instance. USPS-T-35 at 20. His implicit cost coverage comparison relies on costs from Daniel's weight-cost study. Ibid. He uses Daniel's total unit attributable cost, calculated separately for piece-rated and for pound-rated ECR mail. However, he cannot use 3.3 ounces-the Standard A breakpoint-because the IOCS records weight in half-ounce increments for lightweight mail; therefore, he focuses on the unit attributable cost for ECR mail above and below two points: 3.0 ounces and 3.5 ounces. Then, to develop an implicit cost coverage, he calculates unit revenue using the breakpoint as the dividing line. Ibid. His results show that, after rates, the implicit cost coverages for piece rated and pound rated mail using 3.0 ounces as the breakpoint for costs are 215.6 percent and 216.1 percent, respectively. Similarly, using 3.5 ounces as the breakpoint for costs, the implicit cost coverage for piece rated mail is 211.5 percent and 212.6 percent for pound rated mail. Id. at 21. Given that the after-rates cost coverages for piece-rated and pound-rated ECR converge, Moeller concludes that lowering the pound rate does not distort the relative cost coverages of the two groupings. Ibid.
[5466] Illogical postage for heavyweight mail. Moeller also notes that under current rates, the per-piece rate for pound-rated mail is only 0.3 cent ($0.003) for pound-rated saturation nonletters. Thus, the rate for this mail nearly doubles as weight doubles. As an example, he notes that two 4-ounce ECR saturation pieces provide almost the same revenue as one 8-ounce ECR saturation piece. He observes that in both cases, the revenue is virtually the same, but says it seems illogical that the Postal Service would be that indifferent between processing and delivering two 4-ounce ECR pieces and one 8-ounce ECR piece. Id. at 21-22.
[5467] Reclassification obviates ECR pound rate's role as a proxy. Moeller further explains that the pound rate traditionally has served as a proxy for the change in shape between flats and parcels as weight per piece increases, i.e., the proportion of parcels increases and the proportion of flats decreases as weight increases. Under this approach, parcels-being heavier than flats-generate higher revenue per piece. Moeller contends that this situation was appropriate when there was one only commercial subclass because parcels in the then-combined subclass were heavier than flats. However, with the advent of separate subclasses, ECR and Regular can now be evaluated separately. His comparison of the average weight of flats to the average weight of parcels shows that they weigh about the same. Accordingly, he argues that the pound rate can be lowered, as it no longer needs to act as proxy for the change in shape mix. Id. at 22.
[5468] Moderate nature of the proposal, and moderate effects. Moeller states that witness Daniel's cost trend indicates the Service could have selected a lower pound rate but, given concerns about competition, chose a more moderate path. However, he acknowledges that regardless of the degree of the reduction, the fact that there is a reduction may lead, intuitively, to the assumption that the average revenue for pound-rated mail should, likewise, decrease. However, he emphasizes that this is not the case, because the proposed reduction in the pound rate is accompanied by a proposed increase in the corresponding per-piece rate for mail above the breakpoint. Id. at 22-23. At current rates, Moeller calculates the revenue per piece for pound-rate mail at 19.412 cents per piece. The corresponding amount under the Service's proposal is 19.472 cents per piece, a 0.26 percent increase. Id. at 23. Moreover, Moeller says that at the basic level, only pieces weighing more than 6 ounces would experience a rate decrease, and he notes that these pieces represent only 4.6 percent of volume. He concludes that the impact of the proposal is limited, and therefore moderate. Ibid.
[5469] Moeller says the moderate reduction in the pound rate is designed to allay the concerns of those that contend they may be disadvantaged by a significant reduction, yet acknowledges the needs of small businesses who rely on the mail, or wish to use the mail, for affordable advertising. Ibid.
[5470] Introduction. Contrary to witness Moeller's position, AAPS, NAA and Val-Pak, et al. contend that Daniel's cost-weight study does not provide a sufficient support for a reduction in the pound rate. Their arguments are summarized as follows: (a) witness Daniel has discredited her own study; (b) the underlying tallies are too thin, and certain data points produce anomalous results or allow other inferences that undermine the study; (c) the study is, in fact, generally unresponsive to the Commission's criticisms in Docket No. R97-1; (d) the study does not control for important variables; and (e) the use of direct tallies to evaluate the effect of weight on mail processing cost is inappropriate.
[5471] Alleged discrediting. The first argument-that Daniel has discredited her own study-is based on NAA witness Tye's reading of Daniel's testimony that " . . . while it is possible to analyze the data for guidance in rate design, it is difficult, if not impossible, to isolate precisely the impact of weight on costs or identify the exact unit cost of each ounce increment." Tr. 30/14699-14700, citing USPS-T-28 at 4. However, it seems clear, as MOAA, et al. witness Prescott observes, that Daniel is referring to the precision of her results, not to their reliability. Prescott further emphasizes-as does witness Daniel-that Moeller does not rely on specific weight interval cost estimates in Daniel's study or the related regression results to support lowering the pound rate; instead, he uses them as general guidance in selecting an appropriate rate. Tr. 44/19291-92.
[5472] Tally thinness and assertions about certain data points. Assertions that "thin tallies" and seemingly anomalous results at certain data points undermine the study have also been shown to lack merit. For example, Tye contends that the tallies underlying the cost-weight study are too thin, and a pervasive problem. In particular, he notes that only 16 mail processing tallies were recorded for the 11-13 ounces increment, and that less than 200 mail processing tallies were recorded for increments above 7 ounces. Tr. 30/14700. He contends that the thinness problem results in anomalous costs, noting that the study shows ECR parcels cost only one-third as much as Nonprofit ECR parcels, but that lightweight Nonprofit parcels cost four to seven times as much as corresponding ECR parcels. Id. at 14703.
[5473] At the same time, Tye asserts that the high unit cost for the 15-16 ounce increment is not anomalous, but the logical result of a discontinuity between Standard A and Standard B rates. Specifically, he contends that given that Standard A rates for a 15-16 ounce piece are cheaper than the corresponding Standard B rates for pieces weighing slightly more than 16 ounces, mailers of packages slightly heavier than 16 ounces have an incentive to lighten the package to qualify for Standard A rather than Standard B. Moreover, he asserts that this data point provides the only reliable unit cost, and that the costs for the other heavyweight intervals are unreliable. Tr. 30/14700, Tr. 44/19293.
[5474] In response to these assertions, Crowder characterizes Tye's criticisms as nit-picks because they focus on the upper weight brackets that constitute only a small portion of total ECR volume. She notes that nonletter volume above 8 ounces is only 1.4 percent of total volume, and that the 15-16 ounce interval represents only 0.14 percent. Moreover, she says that the tallies for ECR should be thin, as this mail does not incur much in-office processing. Tr. 44/19401. In addition, witness Prescott says that his regression analysis (discussed below) overcomes the problems Tye has identified. Id. at 44/19298.
[5475] Bozzo asserts that small volume categories should generate relatively few tallies, i.e., the tallies should be thin given the overall sample size. He says that if there is a cost measurement problem, it relates to the relative standard error of some narrowly defined weight increments, not data thinness. Id. at 19471. Bozzo asserts that when increasing the overall sample size is not practical, the solution is to "limit the reliance on individual point estimates that are subject to large sampling variation." Ibid. He notes that Moeller, consistent with this solution, uses large weight groupings for cost purposes. Id. at 19471-72. Bozzo then calculates the coefficient of variation (CV) for Moeller's cost groups, using the generalized variance formula (GVF) approach.17 Id. at 19472. The results are shown in Table 5-14
of Variation[5476] Bozzo characterizes these CVs as relatively small, noting that with the exception of Nonprofit ECR (which exceeds ten percent), the other categories have CVs of less than five percent. He says this indicates relatively small sampling error. He therefore concludes that thinness is not a problem for the current Standard A rate design. Id. at 19473-74.
[5477] Witness Prescott says that his regression analysis (discussed below) overcomes the problems Tye has identified. Tr. 44/19298.
[5478] The American Bankers Association and National Association of Presort Mailers (ABA/NAPM) jointly criticize Bozzo's use of the GVF method, saying that using this measure is not valid in the presence of bad data. ABA/NAPM Brief at 2. The Postal Service asserts that ABA/NAPM offer no citation to the record to support this claim, and says it does not withstand scrutiny. Postal Service Reply Brief at VI-71-72. The Commission agrees with the Service.
[5479] Haldi's argument regarding anomalies. Haldi states that where weight affects cost, it is reasonable to expect that the relationship would be smooth and monotonically increasing as weight per piece increases. Tr. 32/15846. To test this expectation, he computes the percentage change in unit cost from ounce increment to ounce increment. Id. at 15845. The percentage changes range from -14.4 to 805.5. Haldi believes these percentage changes are unstable showing dramatic jumps and drops and even a negative change. Id. at 15846-47. Bozzo observes that the first anomaly occurs in the 4 to 5 ounce weight increment. He adds that the 0 to 4-ounce range accounts for 99.8 percent of the letter piece volume and 99.4 percent of volume variable costs. The remaining 0.2 percent of the letter piece volume should show large sampling variation, according to Bozzo. He concludes that the problem here is not with the IOCS, but with slicing the data to thinly. Tr. 44/19469-70. On brief, Val-Pak asserts that Bozzo ignores the anomaly between the 2.5 and 3.0 ounce increments, VP-CW Brief at 50, but the Postal Service, on Reply, contends that there was a transcription error in Dr. Haldi's Table. Postal Service Reply Brief at VI-77. When corrected the anomaly disappears. After reviewing the calculations, the Commission agrees with the Service. The revised numbers do not show the anomaly.
[5480] On brief, SMC says the Commission should not be misled by looking at individual data points when assessing the Daniel's study. Rather, it says the Commission should observe the overall pattern of costs embodied in the study. SMC Brief at 25. SMC adds that if thinness were a problem, the results from study to study submitted over the years would show wildly fluctuating results. SMC notes, however, that the same pattern of cost distribution emerges from every cost-weight study. Id. at 26-27.
[5481] Whether Daniel controls for important variables. Witness Haldi cites two reasons to support his argument that Daniel has not accounted for important variables. One is that she has not controlled for certain factors that unquestionably affect cost, such as presort level and average haul. Tr. 32/15828-29. The other is that in the circumstance where she does provide an adjustment for destination entry differences, the effect is to increase weight-related costs. Ibid. In response to this assertion, witness Prescott contends that Haldi has misstated the effect of Daniel's destination entry adjustment. Rather than increasing the weight-related costs, Prescott shows that her adjustment does, in fact, decrease them. Using Daniel's data, Prescott shows that the cost per ounce is 1.55 cents before the adjustment, and 1.42 cents after the adjustment, or a .13-cent decrease. Tr. 44/19293.
[5482] On brief, SMC argues that by virtue of the supplemental data Daniel provided in response to various interrogatories, interested participants could adjust for the factors Haldi cites. These data, according to SMC, include separate costs for letters and flats; separate costs for basic versus high density and saturation combined; and costs normalized for destination entry differences by weight increment. SMC Brief at 14.
[5483] Distribution methodology: use of direct tallies for measuring the effect of weight on cost. Haldi contends that the direct tallies Daniel relies on tend to reflect individual piece handlings, and that heavyweight pieces can be handled at the same cost as lightweight pieces. Tr. 32/15832. In furtherance of this point, he explains that direct tallies are used to distribute the cost of not-handling tallies. Thus, in cases where weight is the cost driving factor, as in the instance where a mailhandler is moving empty equipment, he says that using direct tallies as the basis for distributing the cost of not-handling tallies to ounce increment will miss the causal connection between weight and cost. Id. at 15833. Haldi asserts that as the weight of individual mail pieces increases, the amount of equipment needed to move mail between processing operations increases and, accordingly, the amount of empty equipment moving between operations also increases. Id. at 15831-32. He contends this problem also extends to mixed-mail tallies. Id. at 15836. Haldi concludes that because of these deficiencies in the IOCS, the Postal Service should conduct some other type of study, such as an engineering study. Id. at 15844, 15848.
[5484] Witness Prescott asserts that Haldi assumes that equipment and personnel are already fully utilized and that all not-handling costs are driven by weight. Prescott, on the other hand, contends that if equipment and personnel are not fully utilized, more work can be performed without adding cost. He also believes that the costs of not-handling tallies are correctly distributed to weight increment, based on Postal Service witness Van-Ty-Smith's testimony that there are instances between handlings where employees are performing duties not affected by weight. These include monitoring the operation of equipment, going to pick up equipment, performing an incidental administrative task, or taking a break. Tr. 44/19300-301, citing USPS-T-17 at 13.
[5485] Crowder also believes that Haldi has overstated the effect of weight on mail processing costs, based on her contention that there is excess capacity in the system. She states that most Standard A mail does not fill the containers, which leaves excess capacity that can handle additional weight. She also contends that larger mailings are generally put in larger containers. Both of these factors, according to Crowder, contribute to scale economies in bulk handling. She concludes that this explains why Daniel's study shows that costs increase at a lesser rate than weight. Id. at 19391.
[5486] Crowder adds that this also addresses Haldi's suggestion that not-handling tallies and mixed-mail tallies should be distributed to ounce increment on the basis of weight. She contends that bulk handling scale economies mean that bulk costs do not increase in a one-to-one correspondence with piece weight or total weight. According to Crowder, these economies can be related to the direct handling tallies. For ECR, she suggests that the majority of bulk handlings reflect identical items or containers. Id. at 19394-95. Because the piece weight is identified for these handlings, the direct tallies associated with them form an appropriate distribution basis for mixed and not-handling tallies. Crowder also speculates that for ECR, the use of direct tallies for items/containers as a distribution key may overstate the effect of heavier mailings on mixed mail costs. For these reasons, Crowder believes that Daniel's distribution methodology is reasonable. Id. at 19395.
[5487] In witness Bozzo's opinion, Haldi's concerns about the distribution of not-handling and mixed-mail tallies reflect a variety of potential shortcomings that have been overcome by using MODS-based costing. Id. at 19469. He explains that the IOCS provides estimates of the proportion of labor time by activity. For example, if activity A requires more time than activity B, then there will be more tallies for activity A. Thus, if heavier mail requires more handling time than lighter mail, Bozzo says there will be more tallies for heavier mail. Id. at 19466-67. He states that mixed mail tallies contain ample information on shape (and sometimes on class) to inform cost distribution. The Postal Service's distribution keys are stratified by MODS-based cost pools and item/container type. Only direct tallies associated with each cost pool and item/container type are used for the distribution of mixed-mail cost. Id. at 19467-68. For not-handling tallies, Bozzo explains that their costs are generally distributed within the cost pool in which they reside, except for allied activities which are based the distribution of costs in the cost pools with which they are associated. Id. at 19468.
[5488] With respect to Haldi's call for an engineering study, Prescott and Crowder also agree with Daniel that the IOCS covers the broad spectrum of costs over an extended period of time. For this reason, they both conclude that IOCS data should be superior to data from an engineering study. Id. at 19290-91, 19393.
[5489] Responsiveness to previous criticisms. The question of whether Daniel's study responds to the Commission's PRC Op. R97-1 criticisms is a more complex question. On this point, witnesses Tye, White, and Haldi contend-again contrary to witness Moeller's assertions-the Daniel study does not fully respond to the Commission's criticisms. Tye says she addresses only one of the criticisms and, even then, has simply modified the distribution of elemental load cost, while ignoring the Commission's criticism concerning tally thinness. Tr. 30/14698. In response to this assertion, witness Prescott contends that Daniel has made significant improvements. In particular, he says she changed the distribution of elemental load cost from a piece basis to a weight basis. He adds that Daniel also improved the distribution of the cost associated with IOCS data observations where weight is not known by using the distribution of tallies with known weight within cost pool, activity, and subclass, rather than using the aggregate distribution of all direct tally costs with known weight. Tr. 44/19287-88. Postal Service rebuttal witness Bozzo agrees with Prescott, emphasizing that he considers Daniel's study an improvement over the previous one because, by virtue of using MODS-based costing, it recognizes the differences in the composition of handlings between direct and mixed-mail tallies. Id. at 19470.
[5490] Witness White contends that the Commission asked for a formal study of the effect of weight on carrier street time cost, and this has not been provided. Instead, he asserts that Daniel simply assumes that route and access costs vary with pieces. He also says that Daniel admits that route cost might vary with the weight of a piece. He contends that support costs, which are distributed to weight category on the basis of the distribution of all other carrier costs, are also affected by this oversight. Tr. 22/9959. White says he does not know the exact effect of weight on costs, but concludes that it is substantial, based on his experience in the delivery business. Id. at 9961. He believes this is evidenced by his recent experience with a total market coverage (TMC) product his company delivers. Its weight per piece increased by a half ounce, thereby adding 26 pounds to each carrier's weekly load and one-half to one mile of walking distance to each route. White adds that the increased weight led to an increase in the delivery charge to cover the additional costs. Id. at 9960. Because witness White believes weight has a substantial effect on delivery costs and the Service has not studied that effect, he contends that the Commission should reject the Daniel's study. Id. at 9959.
[5491] Advo witness Crowder observes that on cross-examination, witness White concedes that, on average, his carriers carry about 15 pounds per walking loop, which amounts to about 50 percent of the available satchel capacity. According to Crowder, this undermines White's position, because the increased weight from the TMC product could be accommodated by the excess satchel capacity with minimal or no impact on cost. Tr. 44/19387-88. Using data from Docket No. 97-1, Crowder calculates that Postal Service city carriers average about 20 pounds per loop in a satchel, with a maximum allowable limit of 35 pounds. She says the Engineered Standards data base provided by Postal Service witness Raymond shows that the actual average weight in a city carrier satchel is 11.3 pounds. Id. at 19397. Crowder therefore argues that the substantial "excess satchel/weight capacity means that a marginal increase in piece weight should have no effect on the number of loops or any other activities which depend on the number of loops." Id. at 19397-8.
[5492] Val-Pak, et al. witness Haldi contends that Daniel simply assumes that elemental load cost varies with weight, rather than basing her distribution on a study showing how cost varies with weight and pieces. He also claims that using weight as a distribution key is inconsistent with CRA methodology. Tr. 32/15848-49 and Tr. 44/19475.
[5493] Witness Bozzo responds to the criticisms lodged against Daniel's distribution of access and route costs and to the contention that her weight-based keys are not consistent with the CRA methodology by computing upper and lower bounds on the costs Moeller uses in his implicit cost coverage comparison. He recomputes total attributable costs for Moeller's two weight groupings on two bases: (1) he assumes access and route costs are entirely related to pieces (which produces a lower bound, as these costs are assumed not to be pound related), and (2) he assumes these costs are entirely related to weight (for an upper bound). Bozzo concludes that the lower bound cost estimate for ECR increases Moeller's implicit cost coverages, and thereby strengthens the cost justification for reducing the pound rate. Tr. 44/19475-79.
[5494] Discussion. It is true that the Service has not provided the type of carrier study the Commission referred to in the last rate case, and the Commission continues to believe that a more comprehensive study of the cost support for the pound rate is needed. However, Daniel's analysis, especially as supplemented by material Daniel provided in response to interrogatories, is an improvement over previous efforts. It is not dispositive on the question of the appropriate pound rate, but provides a general indication of cost trends. Moreover, it helps advance the Commission's understanding of the direction a more definitive study must take.
[5495] Several regression analyses have been presented on this record. One is witness Daniel's, which was provided in a library reference accompanying testimony filed at the outset of the case. USPS-LR-I-92. In brief, she regresses weight per piece on cost per piece. Others include those performed by witnesses Prescott and Crowder.
[5496] Witness Tye acknowledges that Daniel does not endorse her regressions, but nevertheless offers several criticisms. One is that Daniel's regressions are not useful because she has included piece-rated pieces to measure the incremental cost per ounce of pound rated mail. Tr. 30/14705. Another is that her grouping of the data is not grounded in statistical theory. Based on his belief that ECR costs are not linear, Tye argues that grouping the heaviest weight increments obscures the curvature of the trend line. Ibid. Third, as noted earlier, he believes the data are too thin, which creates outliers and other anomalous results. Tr. 44/19279 and Tr. 30/14699-704. Fourth, he suggests that the regressions are not weighted to account for variation in pieces and weight from interval to interval. Tr. 30/14706.
[5497] Prescott's regression analysis. Witness Prescott contends that his revision to Daniel's regressions cures the tally thinness and the outlier problems Tye has raised. Whereas Daniel regressed weight per piece on cost per piece, Prescott regresses pieces per pound on cost per pound.18 He performs one regression for Regular and one regression for ECR. For Regular, his results show a per piece cost of 11.1 cents and a per pound cost of 52.5 cents. For ECR, his results show a per piece cost of 5.6 cents and a per pound cost of 17.6 cents. Prescott notes that the cost per pound for Regular is much larger than for ECR. Tr. 44/19281.
[5498] In evaluating these results, Prescott asserts that his data set contains no statistical outliers and that the 15-16 ounce data point is now within the normative range of the entire data set. Id. at 19283. He states that the R-squared values for Regular and ECR are 95.9 percent and 96.5 percent, respectively, and indicate the proportion of variation explained by changes in pieces per pound. He concludes that this shows "a strong relationship between changes in unit costs and changes in weight." Id. at 19282.
[5499] As noted in the discussion on tally thinness, Tye argues that the 15-16 ounce interval is more reliable than the other heavyweight intervals because it contains more tallies. Prescott agrees with Tye. He performs a regression on Daniel's data without the 15-16 ounce data point and compares the results to Daniel's regression with that data point. His results show that the variation per pound and the variation per piece are much larger when the 15-16 ounce interval is included. Prescott asserts, however, that his regressions overcome this problem. Id. at 19294-95.
[5500] Prescott also agrees with Tye that Daniel's regressions are unweighted. He notes that there is a wide variance in volume, and that this can affect the results of a regression analysis. Id. at 19295-96. However, he contends that his regressions are implicitly weighted by pounds within intervals, which substantially reduces the variance and produces results much closer to reflecting equal statistical weighting for each interval. Id. at 19297.
[5501] Crowder's regression analysis. Witness Crowder also performs regressions on Daniel's ECR data, but adjusts the costs by removing differences due to transportation and mail processing expenses avoided because of dropshipping. The data to make these adjustments was provided by witness Daniel in response to interrogatories. Crowder also adjusts the costs to reflect Haldi's heavyweight letter adjustment; to equalize the carrier in-office unit costs for high density/saturation letters and flats; and to correct rural carrier costs to reflect the figures in USPS-LR-I-95. Id. at 19378 and 19375-76, fn. 8. Her regression results show a per pound cost of 22.2 cents for basic flats and 16.5 cents for high density/saturation flats. Id. at 19378. She contends that this shows the proposed pound rate recovers much more than the corresponding cost. Id. at 19375.
[5502] Val-Pak, et al. characterize Daniel's regressions as ambiguous, limited, and confusing. VP-CW Brief at 21. In support of this characterization, Val-Pak, et al. gathers the various pound rates that can be derived from her regressions. The basis is the cost for a 16-ounce piece, which Val-Pak, et al. believe should be comparable to the pound rate. Val-Pak, et al. develop a rate for the 16-ounce piece by multiplying the cost by the Postal Service's proposed cost coverage of 132.9 percent for Regular subclass and 208.8 percent for ECR. Id. at 22-25. The results are shown in tables presented in Id. at 24 and 26. Based on this comparison, Val-Pak, et al. contend that Daniel's regressions support rates (i.e., cost times cost coverage) ranging from $0.45 to $1.51 per pound for Regular Subclass. It says comparable rates for ECR range from $0.52 to $0.93 per pound. Id. at 23. According to Val-Pak, et al. the latter range does not include the Service's proposed pound rate for a DDU saturation flat ($0.484). Id. at 25 and 26.
[5503] Val-Pak, et al. further assert that Crowder and Prescott add to the confusion by performing additional regressions on Daniel's data. Id. at 32-33, 38-40. Val-Pak, et al. adjust both witnesses' costs for cost coverage, and contend that this produces a rate for 16-ounce pieces outside the range of ECR rates developed from Daniel's regressions shown above. Id. at 32 and 39. SMC strongly contests the introduction of this material at the briefing stage. Among other things, it says Val-Pak, et al. have introduced major misunderstandings by using Daniel's unweighted regressions.
[5504] While witness Daniel's regressions were relatively basic presentations, those presented by witnesses Prescott and Crowder clearly have helped inform the record. The Commission rejects Val-Pak, et al.'s contrary assertions.
[5505] Utility of the implicit cost coverage comparison. Witness Tye attacks Moeller's implicit cost coverage argument as unreliable because Moeller does not apply the test to Regular subclass or the Nonprofit subclasses, and the Postal Service does not use it in First-Class. Tye contends that equalizing the cost coverage for piece-rated and pound-rated mail in Regular would require a decrease in the current pound rate.19 Tr. 30/14708. He also asserts that the comparison is inconsistent with the Postal Service's proposal in First-Class. Treating First-Class one-ounce mail as piece-rated and mail weighing more than one ounce as pound-rated, Tye calculates the implicit cost coverages at test year before rates as 164.7 percent and 185 percent, respectively. To equalize cost coverages, Tye says the additional ounce rate would have to decrease. He observes, however, that the Postal Service proposes to increase the additional ounce rate. Id. at 14710.
[5506] Crowder's observations. Witness Crowder notes that no witness on this record has claimed that current pound rate accurately reflects the effects of weight. She also notes that no witness has refuted Moeller's contention that it produces illogical postage. Elaborating on this position, she comments that to support the existing rate, one would have to believe that the piece-related handling cost for a saturation flat is three-tenths of a cent, with all the remaining cost solely related to weight. She says: "The notion that it costs the Postal Service only 0.3 cents more to handle two 4-ounce pieces than one 8-ounce piece is simply inconceivable." Tr. 44/19371.
[5507] Crowder adds that Haldi acknowledges that heavier weight saturation pieces pay too much in weight-related charges because the presort (density) discounts do not reflect weight-relating savings. Id. at 19372. She also notes that Haldi stated on cross-examination that if the letter-flat differential contained both shape-related and weight related cost differences, which Crowder contends it does, then full passthrough would overcharge nonletters. Id. at 19373.
[5508] Using Daniel's costs, adjusted to remove differences due to dropshipping, Crowder calculates weight related costs per pound for two cases. In the first case, she uses the total attributable costs for flats (above and below the breakpoint). She subtracts from total attributable costs, the costs of transportation, vehicle service drivers, and dropship-related mail processing, so that the remaining costs should reflect ECR mail dropshipped at the DDU. Id. at 19376. She then assumes that the costs are all weight-related and divides by pounds. In the second case, Crowder uses only the cost for flats above the breakpoint. The results are shown in Table 5-15.
Saturation Flats[5509] Crowder contends the extreme assumption shows a cost of 33.4 cents, but she argues that all the case 1 costs are excessive because some of the costs must be piece related. She believes the case 2 costs are also excessive, but she contends that they represent "the absolute upper limit on the amount of weight-related cost that should be used to develop rates" for pound-rate pieces. Ibid. Crowder provides the comparable numbers for flats not dropshipped.
Saturation Flats[5510] Crowder further states that her regressions, discussed earlier, also show that the current pound rate is excessive. She adds that the average cost per piece taken from the costs used to calculate the numbers in her tables show that costs do not rise as sharply as postage for pound-rated mail. Tr. 44/19379. In her final demonstration that the current pound rate is too high, Crowder assumes that the letter-flat cost differential for ECR is entirely weight related and divides the cost difference by pounds.20 The results are shown below.
[5511] According to Crowder, the costs above "are only a fraction of the USPS proposed pound rates." Ibid. She also argues that because the ECR letter-flat cost differential reflects weight- and shape-related differences, the high pound rate leads to a double counting of weight-related costs. Id. at 19382-83. She also says the substantial letter-flat passthrough also results in a higher per piece contribution for nonletters than letters. Id. at 19383.
[5512] Val-Pak, et al., respond that if Crowder's case 1 costs per pound are adjusted to reflect the Postal Service's proposed cost coverage for ECR and adjusted to reflect non-dropshipped mail, the resulting pound rates are higher than current rates.
Costs -
DDU Entry
208.8%
for Origin
Entry[5513] There has been considerable testimony on witness Moeller's contention-essentially a subsection 3622(b)(4) argument-that the proposed reduction in the pound rate, given its moderate nature, does not interfere with competition. Supporters of the Service's proposal sponsor AISOP witness Smith, PostCom witness Harding, and SMC witnesses Buckel, Merriman, Guiliano, and Bradpiece. Opponents sponsor NAA witness Tye, AAPS witness White and NAA witness Wilson. The Service also sponsors witness O'Hara's rebuttal testimony on this issue.
[5514] The Commission prefaces the following discussion with two observations. One is that the SAI report, an update of a previous analysis prepared for the Postal Service's marketing department, once again became the subject of controversy in an omnibus rate proceeding. The report was eventually produced and examined under seal, and participants have discussed its import in portions of their briefs filed under seal. The Commission has been able to reach conclusions on arguments related that the report without placing any portion of this Opinion under seal.
[5515] The other point is that the testimony of the industry witnesses (on both sides of the issue), considered as a whole, provides an extensive-and sometimes frank-assessment of the current state of competition in the marketplace. Much of the testimony, in fact, addresses conditions much more recent in time than the SAI report. The testimony also provides a summary of significant developments over time, in reaction to postal rate changes and reclassification. This provides a useful context for decisionmaking.
[5516] Opponents' claims. Witness Tye raises several impact issues. One is that the reduced pound rate will "continue to divert" delivery volumes from private carriers to the Postal Service. He says this is evidenced by the prospective percentage decrease in rates for pound-rate mail (with some reductions as large as 12.2 percent) and by the Postal Service's alleged "targeting" of heavier mail for rate decreases. Tr. 30/14733-34. Tye also contends that the Postal Service's success at competing with alternate delivery is indicated by Postal Service witness Tolley's estimate that 16.43 percent of the increase in ECR volume is the result of the decline in the pound rate in real terms. Further, Tye observes that the Postal Service's cross elasticity estimate of ECR with newspapers (.812) in conjunction with the Service's own-price elasticity estimate for ECR (.802) indicates that a one percent increase in the price of newspaper advertising would have roughly the same effect on ECR volume as a one percent decrease in the price of ECR. Id. at 14735. He therefore concludes that Moeller has not sufficiently considered the effect on competition. Id. at 14736.
[5517] Tye also criticizes, among other things, the fact that the proposed average rate increase is less for ECR than for Regular or Nonprofit ECR; and he observes that the pound rate has decreased in real terms because it has remained fixed since the beginning of the ECR subclass. Id. at 14736-37.
[5518] AAPS witness White argues that the reduction in Periodicals rates that resulted from the 1995 reclassification hurt the alternate delivery industry. He says his company delivered 175,000 periodicals monthly prior to reclassification, but lost this business afterwards. Tr. 22/9938-39. In his opinion, "[t]he key competitive rate when it comes to saturation advertising and our membership is the pound rate, because the shopping guides and the free publications that make up the backbone of the business for most of our members make their delivery choices with reliance almost exclusively on the pound rate, not the piece rate." Id. at 9944. He believes that lowering the pound rate will further damage his industry's ability to compete. Id. at 9940. He also speculates that although the Postal Service denies having a competitive motive, it is proposing a lower pound rate to divert delivery volumes from alternate delivery carriers to the Postal Service. Id. at 9947-48.
[5519] NAA witness Wilson suggests that newspapers compete with companies that distribute local retail advertising, rather than the Postal Service. Tr. 44/19137. He also says that since the 1995 reclassification, AAPS membership has declined from 300 in 1995 to 100. Id. at 19146. He urges the Commission to recognize that newspapers are the prime conduit for disseminating information, and that loss of advertising revenue reduces the distribution of the news. Id. at 19137. In an argument similar to witness Tye's, he contends that the Postal Service's proposal is aimed at shifting advertising from newspapers to saturation mail products. Id. at 19137-38. On brief, AAPS adds that "there is a sense of equilibrium in the market" and that lowering the pound rate will give the Postal Service an advantage. AAPS Brief at 10-11.
[5520] Supporters' views on impact. SMC witness Buckel, a consultant with a long career in the publishing field, characterizes the pound rate "as artificially contrived and excessive." Tr. 22/9915. He claims it has resulted in private delivery dominating the preprint advertising market in the New York area. Buckel states that the cost of postage for a 12 ounce piece at current rates is $406 per thousand pieces and at proposed rates is $366 per thousand pieces, compared to private delivery cost of $125 per thousand pieces. He contends that this price disparity has resulted in shared mail not being competitive in this market, and has confined shared mailers to lightweight advertising preprints. Id. at 9927. Buckel observes that this is causing shared mailers to consider shifting to private delivery. Id. at 9916.
[5521] SMC witness Bradpiece publishes a weekly mailed saturation free paper (a "Pennysaver" publication) with a circulation of 1.3 million in Maryland and northern Virginia. He also owns the Metro Community News, a weekly saturation free paper with a circulation of 290,000 that is distributed, via private carrier, in Erie and portions of Niagara counties in New York state. Tr. 44/18910. Bradpiece testifies that, in New York, lower private carrier costs warrant greater attention to the private delivery option. Id. at 18928. He also asserts that the high pound rate prevents him from competing effectively for preprint inserts. He further asserts that the pound rate effectively places a cap on his business, since he tries to avoid exceeding the breakpoint. Id. at 18918, 18941.
[5522] SMC witness Merriman publishes the Farmer and Rancher Exchange, a free distribution advertising publication with a weekly circulation of 42,000 copies. Tr. 32/15658. He claims that alternative delivery is not a viable option for his rural publication given low density (his rule of thumb is that a private carrier will not deliver on a route with four or less homes per mile). Id. at 15663,15677. Witness Merriman states that small saturation mailers like his advertising shopper will benefit from a lower pound rate. Id. at 15664. He contends that the major advertisers who use his publication are sensitive to weight-related charges because they do not face rates that are as weight sensitive from nonpostal carriers. Id. at 15664-65. He therefore believes that the Postal Service's proposal would benefit his customers and other rural adverting publications. Id. at 15664.
[5523] AISOP witness Baro is director of sales for The Flyer, a Hart-Hanks free shopper publication in South Florida. He addresses, among other things, the price sensitivity of saturation mail. He emphasizes that postage increases cause medium to large mailers to consider alternative media, while many small businesses conclude that they have to stop advertising, or reduce the size of their advertisement. Tr. 30/14377-78.
[5524] AISOP witness Smith operates the Buttercup Dairy, a full-service neighborhood grocery in Terryville (Long Island), NY. He testifies that mail advertising is essential to small businesses. Id. at 14529. He explains that he can only afford a half-page ad, while his large competitors can afford 6-to-8 ad pages. He urges the Commission consider the needs of small mailers who depend on affordable postage rates to stay in business. Id. at 14533. He believes that with the Postal Service's proposed rates he can compete with larger businesses that have more advertising choices. Id. at 14531.
[5525] PostCom, et al. witness Harding, who advises print advertisers on placement, contends that over 90 percent of his clients' print advertisements are placed in newspapers. Tr. 45/19581-82. He further states that there has been no shift from newspapers to mail over the last five years. Id. at 19585. Harding adds that the high pound rate has hurt saturation mailers' ability to compete for heavier preprints. Ibid.; SMC Brief at 42.
[5526] SMC witness Giuliano is a senior executive with Advo who has focused on postal matters since 1983. He adds that Advo began a shared mail program in the Oklahoma City market in 1991, but this initiative was affected by the Docket No. R87-1 rate increase (which he describes as a 25 percent increase), and later discontinued after Distribution Systems of Oklahoma was created as the Daily Oklahoman's private delivery arm. Tr. 44/18996. On brief, SMC contends that the newspaper industry is the dominant competitor in the print advertising market. SMC Brief at 41.
[5527] SMC argues that neither AAPS witness White nor NAA witness Tye has presented factual evidence to substantiate their claims of the adverse competitive effects of Postal Service pricing on their members. MOAA, et al. Brief at 25. AISOP agrees that AAPS and NAA have not provided any evidence of harm. AISOP Brief at 14.
[5528] MOAA, et al. contend that in the fact of the opponents' failure to produce information on the rates private alternatives charge, the Commission should invoke the adverse inference rule. It says: "`[T]he rule provides that when a party has relevant evidence within his control which he fails to produce, that failure gives rise to an inference that the evidence is unfavorable to him.'" MOAA, et al. Brief at 24, citing International Union (UAW) v. N.L.R.B., 459 F. 2d 1329, 1336 (D. C. Cir. 1972).
[5529] Postal Service witness O'Hara contends that the Service's ECR subclass proposal-both in terms of proposed cost coverages and reduced pound rate-is manifestly reasonable. He notes that both NAA and AAPS contend that the proposal is motivated in large part by a desire to divert business from newspapers and alternative delivery carriers. Tr. 46E/21935. He notes that they also represent that if the pound rate is reduced as proposed by the Postal Service, their organizations' members will suffer economic harm due to diversion of advertising from alternative media, such as newspapers, to Standard A ECR. Ibid. O'Hara says, however, that witness Tye admits that he did not review rates charged by newspapers for inserts, and claims he offers no other quantitative data to support his conclusion that volumes will shift as his testimony portends. Id. at 21936.
[5530] O'Hara notes that AAPS witness White's testimony cites ongoing commissioning of SAI research, as well as the Service's stated intent with regard to the proposed ECR pound rate reductions in earlier docket. O'Hara's assessment is that:
In effect, NAA and AAPS would have the Commission maintain ECR rates for heavier weight pieces at levels far in excess of the relationship suggested by their costs. This necessarily implies that a more affordable alternative, in the form of a more attractive rate for heavier weight ECR mail, would be denied to mailers for the sake of the protectionist self-interest of NAA's and AAPS's members, thereby restricting choice and reducing competition. Simply put, the 3622(b)(4) requirement that the Commission consider the effect on competition weighs in favor of the Postal Service's proposal, for it will enable competition to flourish in the market for high circulation advertising, to the benefit of advertisers.
[5531] Id. at 21937 (emphasis in original).
[5532] Discussion. The Commission finds no persuasive evidence on this record that a reduction in the pound rate, at the Commission's recommended level, will unduly interfere with competition. Given intense competition in the saturation market, any reduction may require competitors to adjust their practices to some degree. However, there is no evidence that a pound rate reduction will impair competitors' ability to attract or retain advertisers; in fact, there are representations on the record indicating that at least some competitors simply price below the Service's rate. The Commission's recommendation must also consider the impact on mailers (and their customers) who pay the pound rate. The Commission finds that although the reduction it is recommending is not as low as the Service proposed, it also fairly takes into consideration the interests of these participants.
[5533] Rate design for a subclass can be thought of as setting the implicit percentage markups for each rate category. The weighted average implicit markup for a subclass (weighted by cost) equals the predetermined subclass markup. Thus, rate design is a zero sum game requiring below average implicit markups to be offset by above average implicit markups. The Commission begins the rate design process assuming equal implicit markups. This is a neutral starting position which seems to be implied by § 3622(b)(1), a fair and equitable schedule. It is consistent with the Commissions general policies that the rates for each rate category be above cost; that rates reflect the costs developed in the record; and that rate design results in identifiable relationships between rate categories. Equal implicit markups, however, are only a starting place, and often may not be practicable or appropriate. The Commission frequently has good reason to depart from them in actual practice.
[5534] The Commission bases worksharing discounts on avoided costs. Basing discounts on avoided costs does not result in equal implicit markups, rather it results in equal per-piece markups. It also results in worksharing mail having higher implicit markups than mail which is not workshared and the most heavily workshared pieces (i.e. those with the largest discount) having the highest implicit markups.
[5535] This approach to worksharing discounts is called "efficient component pricing" (ECP) in the economic literature. The theory requires the discount to be 100 percent of the cost savings. The Commission tries to achieve 100 percent passthrough of the worksharing savings, but again it frequently may depart from this standard for a variety of reasons. An important virtue of ECP is that the mailer will perform the workshared activity (e.g. presort) when he can do so at a lower cost than the Postal Service. This leads to productive efficiency (i.e. the most efficient provider does the work resulting in the lowest cost to society). Because ECP also lowers the real cost of mailing, volume should increase in response to lower effective prices.
[5536] Under ECP, workshared mail makes the same unit overhead contribution as nonworkshared mail. Consequently, the Postal Service is financially indifferent to worksharing because it collects the same overhead contribution from workshared and nonworkshared mail. ECP is Pareto optimum in that some mailers are made better off, but no mailer is made worse off. In some circumstances the introduction of a workshare discount may result in deaveraging if some mailers are already doing the worksharing without a discount. If the volume of already workshared mail is small, it will have a very small impact on mailers who will not workshare. If existing worksharing is widespread, the impact will be greater. The Commission generally mitigates any impact by setting initial passthroughs at less than 100 percent. This explains why worksharing discounts are generally not opposed by mailers who don't workshare. It is largely because of ECP rate design that an extensive range of worksharing discounts has been put in place by the Postal Service over a 25-year period with relatively little opposition from mailers.
[5537] Rate design involves recognizing attributes that cause cost differences among the rate categories (i.e. distance, weight, shape or machinability). If a subclass does not have worksharing, then recognition of the cost differences would result in passing through the marked up cost differences to obtain equal implicit markups for each rate cell (i.e., the neutral starting position). When the Commission does not passthrough 100 percent of the marked up cost differences the implicit markups are not equal, and similarly when the passthrough of worksharing savings is not 100 percent then the unit markups will not be equal.
[5538] Subclass rate designs proposed by the Postal Service and recommended by the Commission frequently depart from the principle of equal implicit markups and from ECP in order to mitigate the effect of large changes in cost on affected rate payers. The Commission also departs from 100 percent passthroughs to accommodate Postal Service's policies concerning how it wants mail prepared. In addition, the Commission departs from strict cost based rates in order to avoid or mitigate anomalous relationships between subclasses.
[5539] Witness Moeller has advanced the Commission's understanding of the pound rate issue by providing us with estimates of the implicit markups for the totality of pieces above and the totality of pieces below the break point. The estimates are based on attributable costs including in-office, transportation and delivery costs. While the estimates of IOCS costs for pieces above and below the break point are statistically reliable, the Commission has not closely examined the basis upon which transportation and delivery costs are distributed. If the Commission is to make proper further use of the implicit markups in setting the pound rate, the basis for distributing transportation and delivery costs must be subject to more scrutiny.
[5540] The Commission hopes that reliable information on implicit markups may make it possible to calculate the total amount of revenue that should be obtained from pieces above and from pieces below the break point. This would be an important contribution to ensuring that intra subclass rate relationships for Standard Mail are fair and equitable. The separate issue of the best way to design rates for the pieces above and below the break point might also be addressed by studying implicit markups.
[5541] Witness Moeller's implicit markups reflect the mix of mail on either side of the break point. However, pieces above and below the break point have different worksharing profiles and different shape profiles. The Commission believes that implicit markups comparison should be adjusted for these differences.
[5542] This Commission hopes that the Postal Service and the parties can develop this information to allow the appropriate comparison of the implicit markups for pieces above and below the breakpoint. These should reflect the degree and type of worksharing in each category and the distribution of shapes within each category as well as Commission recommended passthroughs. The Commission could use this information to recommend rates that produce the appropriate amount of revenue from pieces above and below the break point.
[5543] While it is important to collect the appropriate revenue from each broad category, there is also the problem of the degree to which the current rate structure is cost based. The Commission would like to learn if statistically reliable implicit markups for subgroups of pound rated pieces can be obtained. If so, it may be possible examine alternate ways (linear and non-linear) to design rates for pound rated pieces and perhaps, for all pieces in each Standard Mail subclass.
[5544] Haldi's approach. Haldi contends that the cost of ECR letters is overstated, while the cost of ECR nonletters is understated.21 Tr. 32/15765. He explains that the problem occurs because the IOCS and RPW use different definitions of flat-shaped mail. Specifically, IOCS uses a dimensional (length, width, thickness) definition, while RPW uses the rate paid by the mailer. According to Haldi, this produces a mismatch of data by shape for costs and volumes because the IOCS letter costs include the cost of letter-shaped mail above the breakpoint, while the RPW letter revenues exclude the revenues of letter-shaped mail above the breakpoint. He notes that RPW records the revenue correctly because all Standard A mail above the breakpoint, regardless of shape, must pay the nonletter rate. Id. at 15766.
[5545] Haldi believes the IOCS costs should be adjusted to properly align shape costs and volumes. Ibid. Using Daniel's cost data, witness Ramage's information on IOCS tallies, and the assumption that the tally cost for letters between 3.0 ounce and 3.5 ounces is evenly distributed by ounce increments of 0.1, he estimates that 2.6 percent of all ECR letter tallies are assignable to the cost of nonletters. By shifting 2.6 percent of letter cost to nonletter cost, he recalculates the cost of letter and nonletters. His approach adds .29 cents to the letter-nonletter differential. Id. at 15813-19.
[5546] Opposing views. Both Prescott and Crowder point out that there is no mismatch of costs and volumes in Daniel's data because the volumes for letters and nonletters reflect the IOCS shape definitions. Tr. 44/19306, 19311 and 19405. Both also explain that any adjustment would, therefore, require both the costs and volumes of heavyweight letters to be shifted to the nonletter category. Id. at 19306 and 19405. They further testify that Haldi shifts a portion of total attributable costs, rather than only mail processing and delivery costs, which both the Commission and the Postal Service use to calculate the letter-flat differential. Id. at 19305 and 19405. Crowder recalculates the letter and nonletter costs using Haldi's tally assumptions, but only mail processing and delivery cost. She shifts both the cost and volume of heavy letters from the letter category to the nonletter category. She estimates that the letter-nonletter differential would increase by .077 cents per piece. Id. at 19405.
[5547] Commission analysis. The Commission agrees with Haldi that there is a mismatch between IOCS costs and RPW volumes related to letters and nonletters. However, it also agrees with witnesses Prescott and Crowder that (1) Daniel's volumes, which are not RPW volumes, are based on the same definition of shape as used in the IOCS; (2) both the cost and volume of heavyweight letters should be shifted to the nonletter category; and, (3) only the costs of mail processing and delivery should be used in the calculation of the letter-nonletter differential.
[5548] Crowder calculates the average difference between ECR letter and nonletters in accordance with the three items above. By Crowder's calculation, the average difference would increase by only .077 cents. Rounded, this would add one-tenth of a cent to the average difference. However, for ECR rate design, Moeller uses letter-nonletter costs by density level (basic, high density, saturation), not the average difference. Thus, the add-on does not exactly align with the density levels. Further, it seems that this adjustment should apply only to IOCS data, as the problem involves the shape definition used in IOCS, but Crowder includes street time from cost segment 7 and all of rural carrier costs from cost segment 10. Ibid. Neither of these costs are developed from IOCS. Since excluding these costs from Crowder's calculation would further reduce the add-on, the Commission believes the impact on the letter-nonletter differential would be de minimis. For this reason, it is not used in the cost differential supporting the ECR letter-nonletter rates.
[5549] This does not mean that the Postal Service can ignore the problem. It extends to all four subclasses in Standard A, and directly affects rate design, which requires costs to be as accurate as possible. These mismatched costs enter the letter/nonletter differential for the Regular subclass through the CRA adjustment factor and, as described above, CRA costs by shape and density level are used for ECR rate differentials. For this reason, the Postal Service should change its IOCS recording procedures to identify Standard A mail above and below the breakpoint. Although the breakpoints currently vary across subclass from 3.2873 ounces for Nonprofit subclass to 3.3103 ounces for Nonprofit ECR, using 3.3 ounces as the reference weight in IOCS would appear to be sufficient for this purpose.
[5550] Witness Lubenow, testifying on behalf of PostCom, et al., asks the Commission to increase the automation discounts, based on mailers' extensive address quality efforts; suggests consideration of a discount for 5-digit mixed carrier route packages; and offers several other observations on rate design improvements. Tr. 29/14105-07. Lubenow's support for larger automation discounts is based on the assertion that mailers incur costs in attaining high address quality and maintaining it through regular certifications. Id. at 14105. He does not offer an economic analysis, but asserts that the relevant standard of comparison should be the depth of these discounts in the current rates. He says that other things being equal, the behavior of mailers in considering investments in address hygiene will be driven by the relative magnitude of the discounts over time. Ibid. Lubenow says the Commission should either lower the automation rates, or raise the rates used by mail with defective addresses. Id. at 14105-06.
[5551] Lubenow describes his 5-digit mixed carrier route proposal as a smaller discount than the basic carrier discount, but worth more than current 5-digit packages, as only pieces with addresses of sufficient quality to allow the carrier route code to be identified would be included. Id. at 14105.
[5552] Witness Lubenow's testimony offers an important industry perspective on the Service's automation program, and significant insights into several rate design improvements that could encourage attention to address quality. The Commission is not recommending the 5digit mixed carrier route, as important cost, operational, and other rate design considerations have not been explored on this record. Overall, however, witness Lubenow's testimony has contributed useful information on the role of address quality in postal operations. The Commission encourages the Service and the mailing industry to consider his observations as they explore possibilities for rate design improvements.
[5553] In this proceeding, the Service applies the RFRA markup formula in developing proposed Standard A Nonprofit rates, but bases proposed Nonprofit ECR rates on an anticipated amendment to the law. In areas other than markup, Moeller generally follows, for both preferred subclasses, rate design approaches that mirror those proposed for their commercial counterparts. 22 USPS-T-35 at 30-33 and 37-39. The proposed pound rate for Nonprofit is 58 cents, a 3-cent increase over the current pound rate of 55 cents. Moeller says this is consistent with the overall increase for the subclass, and avoids upward pressure on the piece rates. Id. at 30-31. The proposed Nonprofit ECR pound rate is 37 cents, 8 cents more than the current rate of 29 cents. Moeller acknowledges that this is a large increase but, like the increase for Regular, avoids upward pressure on piece rates. He also says that it is "still well below what is was prior to the implementation of Docket No. R87-1 rates. Id. at 37.
[5554] Markup considerations. Moeller notes that, as of the Service's filing, controlling provisions in RFRA set the markup for the preferred subclasses at one-half of the commercial markup. Id. at 29. He observes that the Service was able to apply this formula without undue impact in developing rates for the Nonprofit subclass, but found that adhering to the RFRA formula in Nonprofit ECR resulted in a rate increase of more than 30 percent. Id. at 36. He attributes this, in part, to the fact that recent classification reform (Docket No. MC96-2) replaced the former unitary bulk nonprofit subclass with two subclasses, which necessitates separate markups under RFRA.23 Id. at 29-30, 36.
[5555] The Postal Service's anticipated markup alternative for Nonprofit ECR. Given these results for Nonprofit ECR, Moeller says the Service's markup approach reflects an anticipated amendment to RFRA which would limit the extent to which rate increases for preferred subclasses could exceed the commercial counterpart. Specifically, it provides the option of setting the markup at less than one-half the commercial markup if necessary to keep the percentage increase within 10 percentage points of the commercial increase. Ibid.
[5556] The legislation eventually enacted (S. 2686) provides that Standard A Nonprofit rates, overall, are to be set so that the estimated average revenue per piece from each subclass is equal, as nearly as practicable, to 60 percent of the estimated average revenue per piece from the corresponding regular-rate category.24 In pertinent part, Section 1(d) adds a new 39 U.S.C. 3626(a)(6). The report accompanying S. 2686 notes: "The legislation is designed to address technical problems in the nonprofit ratesetting structure by locking in the current rate relationship between nonprofit and commercial mail." Report 106-468 at 1. The report also discusses a transitional provision in S. 2686. It says this provision is intended to mitigate the impact of these changes on regular-rate mailers in the pending rate proceeding at the Commission. Id. at 4. The report says that under this provision, the estimated reduction in revenue from Nonprofit Standard A mail caused by the enactment of the new ratemaking rules, if any, is to be treated as a reasonably assignable cost of the Postal Service to be apportioned among the various classes of mail and types of postal service in accordance with section 3622(b)(3). Id. at 4.
[5557] The new law affects the revenue requirement, but not rate design. The new law expects the Commission to recommend rate differentials consistent with the policies of the Act. Generally, this is the province of rate design. Consistent with previous Commission decisions, the Commission is relying on special cost studies that identify avoidable cost differences between rate categories as the basis for recommending discounts.
[5558] Given the timing of this amendment to RFRA, there has been essentially no opportunity for participants to discuss implementation. The Postal Service states that although the provisions of this legislation are not identical to those it had anticipated, "it is expected, if enacted, the general outcome in terms of the relationship between commercial and nonprofit rates would not differ significantly from that anticipated in the filing." Postal Service Brief at VII-205. The Service also says that nothing in the legislation should preclude the use of the special studies used to determine rate differentials, such as discounts, although depending on the final language of the bills, there could be adjustments to the overall level of the rates. Ibid.
[5559] The Commission recommends retaining the pound rate of 55 cents for Nonprofit. As discussed in some detail in the section on Standard ECR, the proper level of the pound rate is an issue that requires additional analysis. The recommended pound rates for Nonprofit ECR is 37 cents, the rate suggested by witness Moeller. Nonprofit ECR rates must increase substantially to comply with the new legislation, and it is appropriate to spread this burden among all Nonprofit ECR mailers. Therefore, pound rates are increased to the level suggested by the Service. The recommended residual shape surcharges and the prebarcoding discounts are identical to those recommended for commercial mailers as proposed by the Postal Service.
[5560] Nonprofit rate design. The Commission attempts to have rate designs as cost based as possible, so in three of the four letter categories, the Commission recommends full passthrough of cost savings. In the other category-5-digit automation letters-the Commission passes through 115 percent of cost savings. This is done to support the Service's automation program by providing a 5-digit automation rate that is between basic ECR and basic ECR automation. The Commission recommends Nonprofit flats discounts that are the same as those the Service proposed. Letter/flat differentials are also essentially similar to what the Service proposed. Thus, the passthrough patterns for both Nonprofit and Nonprofit ECR rates are similar to those in the Postal Service's proposal.
[5561] The Commission anticipates that the next proceeding will provide an opportunity for all interested participants to offer comments on rate design options under the new law.
[5562] Nonprofit ECR rate design. The Commission has adopted the discounts proposed by the Postal Service. Given the updated Nonprofit ECR costs, the Commission's passthrough percentages reflect a pattern similar to the Postal Service's, with the exception of the letter/flat passthroughs. The Postal Service's response to P.O. Information Request No. 4 shows Nonprofit ECR shape unit costs display volatility from year to year. For the years 1989 though 1991 and for 1993, 1997, and 1999, the unit mail processing cost for letters unexpectedly exceeds the unit mail processing cost for flats. For the other years in the 1989-1999 period, the unit mail processing cost for flats exceeds the corresponding unit cost for letters, as expected. Since these costs are IOCS based, the counterintuitive results for some years, including the 1999 base year used by the Commission, likely occur because Nonprofit ECR is a small-volume subclass for IOCS purposes. Thus, one might expect these volatile results over time. But these kinds of cost changes can disrupt rate design if they result only from statistical variation. Since this is likely the case for FY 1999, the Commission is passing through more than 100 percent of the letter/flat cost differentials at the high density and saturation tiers. This approach adopts the Postal Service's proposed letter/flat rate differentials and does not produce unduly burdensome rate changes.
[5563] Flats. Because the Commission is using the Postal Service's presort tree methodology, the passthroughs for letters and the letter/flat differential result in a passthrough of 44 percent for high density flats and 118 percent for saturation flats. This also results in discounts for flats equal to Postal Service's proposed discounts.
1These increases reflect revenue per piece; they do not include the effect of the increases in the residual shape surcharge.
3For convenience and readability, the term "parcels" is generally used to refer to all types of pieces that do not meet the dimensions of a letter or flat or are prepared in accordance with parcel requirements.
4The term "Standard A" is used to distinguish the component of Standard Mail that represents former third-class mail.
5This will be referred to as a "minimum per piece/piece-pound structure." Mail below the breakpoint is often referred to as piece-rated mail, while mail above the breakpoint is referred to as pound-rated mail. However, a piece rate (which varies by subclass) also applies to "pound rated" mail.
6The formula includes an algebraic equation that solves for the two piece-rate elements in the underlying structure. The inputs to the formula include discounts; a breakpoint; a target subclass cost coverage; the pound rate (for a piece that is not entered at a destination facility by a mailer); and passthroughs for various discounts. USPS-T-35 at 3-4.
8ECR and Nonprofit ECR avoid processing on the equipment that reads barcodes, so discounts for barcoding are not appropriate for these subclasses.
9Moeller says that corresponding changes to the maximum weight of non-automation letters are not contemplated. USPS-T-35 at 12-13.
11Moeller notes that the relative passthroughs differ, in part because of changes in cost methodology. Moreover, in current rates, 50 percent of the cost difference between letters and flats is recognized at the basic presort tier, and 40 percent of the difference at the 3/5 tier. USPS-T-35 at 5.
12The R97-1 and R2000-1 passthroughs are not directly comparable because of differences in calculating underlying costs and in measuring the cost difference between letters and flats. Id. at 5.
14This will generally be referred to in this discussion as part of the competitive impact or effect argument.
16No one consideration is necessarily more influential than another; each contributes to the impression that the current rate is too high.
17To distinguish between piece-rated mail and pound-rated mail, Bozzo's calculations use the costs for 0 to 3.0 ounces as a proxy for piece-rated mail and costs above 3.0 ounces as a proxy for pound-rated mail. USPS-T-35 at 1 and Tr. 44/19473, Table 2.
18The basic equation is cost = cost per piece times pieces + cost per pound times pounds. This formulation exactly matches the ECR rate structure for mail above the breakpoint weight. Daniel divided both sides of the equation by pieces, so that her regression equation is cost per piece = a fixed cost per piece plus a constant times pounds per piece. Multiplying each side of the resulting regression equation result by pieces produces regression results for the original basic equation. In contrast, Prescott divides both sides of the basic equation by pounds, so that his regression equation is cost per pound = a fixed cost per pound plus a constant times pieces per pound. As in the case of Daniel's formulation, multiplying each side of the Prescott's resulting regression equation result by pieces produces regression results for the original basic equation.
20Crowder uses Daniel's ECR costs that already exclude differences due to dropshipping. However, she adds transportation costs to develop costs for non-dropshipped basic and saturation mail. Tr. 44/19381.
22Witness Tolley's testimony indicates the two preferred subclasses accounted for approximately 10.9 billion pieces. The Nonprofit subclass contributed 8.0 pieces; 2.9 billion pieces were Nonprofit ECR. USPS-T-6 at 107-108.
23Moeller notes that application of RFRA in the reclassification case had led to significant declines for almost all rate categories of nonprofit mail. In Docket No. R97-1, Nonprofit increased 9.6 percent, and Nonprofit ECR decreased again, by 10.6 percent. USPS-T-35-29-30.
24S. 2686 was signed into law on October 27, 2000. Section 1(a) addresses regular rates. It amends RFRA by providing that regular rates are to be established by applying specified policies to the attributable costs of both the commercial and preferred subclasses.
 Postal Rate Commission http://www.prc.gov Voice: (202) 789-6800 Fax: (202) 789-6886 prc-admin@prc.gov |
| |
|
|
|