[00:00:01] Speaker 01: Case number 16-1354 at Health, United Parcel Service, Inc. [00:00:07] Speaker 01: Petitioner versus Postal Regulatory Commission. [00:00:10] Speaker 01: Mr. Lamkin for the petitioner, Mr. Shee for the respondent. [00:00:47] Speaker 03: Thank you, and may it please the Court. [00:00:49] Speaker 03: Jeff Lamkin, on behalf of UPS, I'd like to reserve five minutes for rebuttal. [00:00:54] Speaker 03: The Commission concedes that its approach to attribution almost certainly underestimates the inframarginal costs caused by each product. [00:01:04] Speaker 03: That's because it attributes only the minimum inframarginal cost the product could possibly have incurred. [00:01:09] Speaker 03: That minimum possible approach is inconsistent with the statute, [00:01:14] Speaker 03: It's nowhere reconciled with the express statutory purpose conceded by the commission, which is ensuring fair competition on a level playing field. [00:01:24] Speaker 04: But the statute's more complex than that. [00:01:25] Speaker 04: The statute requires for any allocated costs that they be [00:01:34] Speaker 04: through reliably identified relationships. [00:01:38] Speaker 04: And the Commission's view is they've allocated all the inframarginal costs they can. [00:01:44] Speaker 04: through a verifiable method, and that UPS's proposed system, which allocates more, rests on a series of unverifiable assumptions and requires guesswork. [00:02:00] Speaker 04: That's their theory. [00:02:02] Speaker 04: It's not that it's minimum, it's that they allocated all they could through a verifiable methodology. [00:02:09] Speaker 04: Well, Your Honor, I think the... So why does that violate the... I understand you may, you argue that that's arbitrary corporations, but why, why does that violate the statute? [00:02:20] Speaker 03: Your Honor, the statutory phrase, it has to be reliably identified causal relationships. [00:02:27] Speaker 03: And if the Commission chooses a relationship that is so narrow that it excludes a category that Congress wanted to include, then it violates the statute. [00:02:35] Speaker 04: Well, what in the statute says Congress wanted to include more than a minimum amount of inframarginal costs? [00:02:42] Speaker 03: And the statute uses the phrase that says that costs attributable includes direct and indirect costs. [00:02:49] Speaker 03: And indirect costs is a well understood term. [00:02:51] Speaker 03: It means not just if it includes costs that are [00:02:54] Speaker 03: caused by two products such that you cannot unequivocally say that it's caused by one or the other. [00:03:00] Speaker 03: If it's not unequivocal, then it's, if it is unequivocal, it's a direct cost. [00:03:05] Speaker 03: And the commission simply excludes every cost that's not unequivocally caused by one product. [00:03:11] Speaker 00: For example- It doesn't, it doesn't because it's included some inframarginal costs. [00:03:16] Speaker 03: Right, but only those inframarginal costs that are unequivocally caused by one product. [00:03:20] Speaker 03: For example, if you have two products that are both causing a cost, [00:03:24] Speaker 03: and you remove one product, the cost goes away, the commission will attribute that. [00:03:28] Speaker 03: Because it's unequivocally caused by that product. [00:03:30] Speaker 03: The product goes away, the cost goes away. [00:03:34] Speaker 03: but if you have to eliminate both products, the commission will not attribute that cost. [00:03:38] Speaker 04: Well, that's not the commission's theory. [00:03:39] Speaker 04: The commission believes it is doing that. [00:03:41] Speaker 04: I mean, take for example, suppose you have a mail truck that is carrying parcels and letters and uses 20 gallons of gas. [00:03:50] Speaker 04: Those are common costs, right? [00:03:53] Speaker 03: Correct. [00:03:54] Speaker 04: Okay. [00:03:54] Speaker 04: Well, the commission [00:03:57] Speaker 04: is attributing the amount it can verifiably allocate to, say, the parcels, say, five gallons. [00:04:04] Speaker 04: But those are still common costs. [00:04:06] Speaker 03: Actually, Your Honor, that's not correct. [00:04:08] Speaker 03: Why? [00:04:08] Speaker 03: When the postal truck actually leaves on its route. [00:04:12] Speaker 03: The only cost that gets attributed is the time it takes for the postal carrier to reach into his bag and pull out the package and put it in the mailbox. [00:04:19] Speaker 03: That gets attributed. [00:04:20] Speaker 03: If you have a postal truck that's carrying just parcels, just five pound parcels, one could think that would be attributed. [00:04:27] Speaker 04: The commission says that this methodology, quote, does not exclude all common costs, but only those without a reliably identifiable causal connection. [00:04:40] Speaker 03: The Commission, of course, doesn't... Is that false? [00:04:43] Speaker 03: It is incorrect, Your Honor. [00:04:44] Speaker 03: The Commission doesn't identify any common costs excluded, and the carrier and the City Time carrier example that Your Honor has included shows exactly that. [00:04:52] Speaker 03: If you have a truck full of five-pound parcels, a hundred of them, and there's only five-pound parcels, all that would be attributed. [00:04:59] Speaker 03: If you put one first-class letter on top, [00:05:02] Speaker 03: None of it gets attributed. [00:05:04] Speaker 03: That just doesn't make sense in special. [00:05:05] Speaker 00: That's not the way I read the record. [00:05:08] Speaker 00: And the commission, I mean, one of the things that the commission has done in response to the petition of your clients is attribute the increment of inframarginal costs that are [00:05:27] Speaker 00: at the right end of the curve, because it says those, we can reliably attribute to the bringing on or not of this product class. [00:05:40] Speaker 03: And those are shared costs. [00:05:42] Speaker 03: Those on the right end of the curve, they are not going to be shared costs. [00:05:45] Speaker 03: They're not joint costs in the sense because they go away and they can be unequivocally. [00:05:50] Speaker 00: You're mixing apples and oranges there. [00:05:51] Speaker 00: They go away. [00:05:52] Speaker 00: That portion of the shared cost goes away when that product is taken offline. [00:05:58] Speaker 00: But that doesn't answer what the underlying character of those costs is. [00:06:02] Speaker 03: Yes. [00:06:02] Speaker 03: So they may rely on shared inputs. [00:06:05] Speaker 03: So if the court's referring to, for example, the idea of long distance transportation, which is cubic foot miles, those are costs that are like common inputs across a lot of products. [00:06:15] Speaker 03: And you can attribute those to individual products. [00:06:18] Speaker 03: But that's because if you pull product volume out, that cost goes down. [00:06:23] Speaker 03: And so you know that particular products are or are not using those costs. [00:06:27] Speaker 03: But when you go to attribute the costs and the way the commission does this, it only goes to the very far right of the curve, as you said, Judge Pillard, and it uses only the minimum number that could possibly be caused. [00:06:38] Speaker 03: And it never reconciles that choice with, for example, the purpose of the statute. [00:06:42] Speaker 03: The purpose of the statute, the commission concedes, is fair competition. [00:06:46] Speaker 03: Not pure economic efficiency, but fair competition. [00:06:48] Speaker 00: But it's also competition. [00:06:50] Speaker 00: And so it sounds like your position is taking us back to before the greeting card case where, I mean, you surely could, and your brief does a very able job of describing how one could take all the inframarginal costs and just [00:07:05] Speaker 00: accumulate them all and then allocate them proportionally to all the products. [00:07:09] Speaker 00: But it sounds like what the commission is doing here is trying to take advantage of competitive opportunities that the Postal Service has. [00:07:18] Speaker 00: So it's saying we have untapped capacity given what we're already doing, including with our market dominant products, and we should be thinking about how we could competitively innovate and offer additional things. [00:07:32] Speaker 03: And nobody says the Postal Service should not be able to compete. [00:07:35] Speaker 03: These are genuine costs. [00:07:36] Speaker 03: They're real costs. [00:07:37] Speaker 03: They have to be recovered somewhere. [00:07:39] Speaker 03: If they're not recovered here, they're going to have to come from market-dominant products, or the postal service is going to have to lose money, which is exactly what it's been doing for 10 years. [00:07:48] Speaker 03: It is not fair competition for UPS to be facing a statutory monopolist who gets advantages not merely by advantage of its statutory monopoly in one area, but they can lose money year after year after year because it's not recovering its costs. [00:08:01] Speaker 03: But just turning back to the example of statutory purpose, one of the things that's most problematic about the commission situation. [00:08:08] Speaker 00: I'm sorry, I missed the gist of that point. [00:08:11] Speaker 00: How would your proposal make the loss of the Postal Services losses [00:08:18] Speaker 03: better. [00:08:20] Speaker 03: Well, it simply means that the Postal Service, when it has competitive products, must cover the full costs of those competitive products in the price of each product. [00:08:28] Speaker 03: And that's not just the bare minimum at the end. [00:08:30] Speaker 04: Excuse me, I keep going back to the statute. [00:08:32] Speaker 04: That's not what the statute says. [00:08:35] Speaker 04: The statute says only cause it can be verifiably connected. [00:08:40] Speaker 04: That's correct. [00:08:41] Speaker 03: To the product. [00:08:41] Speaker 04: You keep ignoring that provision of the statute. [00:08:44] Speaker 03: Not at all, Your Honor. [00:08:45] Speaker 03: No? [00:08:46] Speaker 03: The commission has to choose certain causal relationships. [00:08:51] Speaker 03: Right. [00:08:51] Speaker 03: If it chooses causal relationships on such a narrow basis as to over-exclude, it's going to end up with too low an attribution. [00:08:58] Speaker 04: But you don't have any, you can't point to any statutory language. [00:09:03] Speaker 04: that incorporates this too low. [00:09:09] Speaker 04: In other words, the way you read the statute, it can allocate costs through a causal relationship that's verifiable unless it's too low. [00:09:25] Speaker 03: No, Your Honor, I think that... That's not in the statute. [00:09:28] Speaker 03: Well, Your Honor, I think if I could turn the court to page 1060, the appendix, that part of the opinion, that actually shows exactly what's going on here. [00:09:36] Speaker 03: The commission says at that point, any overstatement of inframarginal costs as part of those costs that are attributed to products runs contrary to prior legislative intent concerning cost attribution. [00:09:47] Speaker 03: Now one thing it doesn't say, it doesn't say why isn't under attribution equally problematic. [00:09:53] Speaker 03: In the legislative history it cites in footnote 61 in the small text you go down below, it urges the commission to reach a technically correct result placing accuracy [00:10:03] Speaker 03: above achieving a particular outcome of higher or lower attribution. [00:10:08] Speaker 03: The Commission's done exactly the opposite of legislative history here. [00:10:11] Speaker 03: It has targeted lower attribution. [00:10:14] Speaker 01: Well, can I ask you this question about the statutory language? [00:10:16] Speaker 01: So, suppose on the definition of cost attributable, the Commission reaches the conclusion, and I don't think they have reached this conclusion, just suppose that they say that indirect costs [00:10:30] Speaker 01: We'd love to include them and attribute them, but there's just no reliably identified causal way to do it. [00:10:37] Speaker 01: We just can't find one, and therefore we're keeping indirect costs out. [00:10:40] Speaker 01: The argument could be made in that situation that that's perfectly consistent with the statute. [00:10:44] Speaker 01: Even though you aim to include indirect costs, they've just determined that there's no reliably causal way to do it. [00:10:50] Speaker 01: And then the way you account for that, because I think you're right, at that point things become quite low. [00:10:56] Speaker 01: And the statute accounts for that because it doesn't deal with it in 3633A2. [00:11:02] Speaker 01: It deals with it in 3633A3 and says, well, then there's other ways that you can boost up the costs to get closer. [00:11:11] Speaker 01: But with respect to A2, you've got to have the reliable causal relationship. [00:11:15] Speaker 03: And there's two pieces to that question, Judge Srinivasan. [00:11:19] Speaker 03: The first part is, [00:11:21] Speaker 03: What the commission basically says, it looks like a statute that effectively says, you have to have reasonably sized domestic dogs and cats. [00:11:27] Speaker 03: And it says 50 pound minimum. [00:11:29] Speaker 03: That's wrong because that's going to exclude all domestic cats. [00:11:33] Speaker 03: At the very least, if it's going to choose a pound minimum that excludes all domestic cats, it's going to have to explain why it's doing that. [00:11:40] Speaker 03: And that's what's missing here. [00:11:41] Speaker 03: There's no explanation why it's going to choose the minimum as opposed to targeting accuracy. [00:11:46] Speaker 01: I thought their explanation is that they don't have the reliable causal nexus for everything that's not included. [00:11:52] Speaker 03: And the reason they don't think there's a reliable causal nexus is because you don't know the ordering of operations, which comes first and which comes last. [00:12:00] Speaker 03: But that's exactly true. [00:12:02] Speaker 00: That's not the only reason. [00:12:03] Speaker 00: They also, because they don't operate without the market-dominant products, they don't have data on what the actual inframarginal costs are at the very left side of the curve, right, closer to zero. [00:12:17] Speaker 03: So I should distinguish two things here very quickly. [00:12:19] Speaker 03: One is whether the commission has justified its choice, its chosen numbers, versus why it rejected UPSs. [00:12:25] Speaker 03: But even with respect to, I think, Judge Billy referring to the elasticities issue, even there, it's common ground. [00:12:32] Speaker 03: Everyone agrees you don't even need elasticities. [00:12:34] Speaker 03: You don't need that at all if, for example, it's a fully variable cost. [00:12:37] Speaker 03: or you know your fixed costs. [00:12:39] Speaker 03: But the Commission doesn't even consider for a moment the idea of at least attributing those costs where it doesn't need the elasticity of demand because you do know that it's all variable, or you don't, or you can subtract your fixed costs. [00:12:52] Speaker 03: Turning back to your question, Judge Srinivasan, the Commission doesn't say, hey, we're going to use the bare minimum under A2 because we're going to make up the difference to you under A3 for a fair attribution. [00:13:01] Speaker 03: And it doesn't say that for a reason. [00:13:03] Speaker 03: It doesn't want to bind itself to that result. [00:13:06] Speaker 03: It's done exactly the opposite historically. [00:13:08] Speaker 03: Historically, 30% of attributable costs are attributed to competitive products. [00:13:14] Speaker 03: But when you get to the proper share requirement under A3, it's 5.5%. [00:13:20] Speaker 03: That's an enormous difference. [00:13:23] Speaker 03: The problem in this case is, although we can argue back and forth about what the commission's choices were, [00:13:28] Speaker 03: At the bottom, it's very hard to find anywhere where the Commission actually addresses the language of the statute, such as the meaning of indirect costs. [00:13:35] Speaker 04: I'd be interested in hearing your answer to Judge Sridharvasan's first question, which is the Commission says we want to do what the statute says and we want to allocate indirect costs. [00:13:50] Speaker 04: detailed examination, and we can't verifiably link any of them. [00:13:53] Speaker 04: We can't meet the statutory standard reliably identifying causal relationships, so we're not going to include them. [00:14:01] Speaker 03: That would not violate the statute, right? [00:14:04] Speaker 03: I think there's a serious tension if you're reading a piece of the statute. [00:14:07] Speaker 04: No, no, no. [00:14:08] Speaker 04: The question is, would that violate the statute? [00:14:12] Speaker 03: I think it would, but I think that, because Congress specifically did indirect costs, and if it can't, the Commission can't find it. [00:14:18] Speaker 04: But it says, but the statute says, it says, the direct and indirect costs, and the verifiable language modifies indirect also. [00:14:29] Speaker 03: That's correct. [00:14:30] Speaker 04: So if they can't do it through any verifiable [00:14:35] Speaker 04: causal connection, then I don't see the basis for your argument that it would violate the statute. [00:14:41] Speaker 03: At the very least, the Commission has to look very hard and try hard to make some way to include it. [00:14:46] Speaker 04: Well, that's part of my hypothetical. [00:14:48] Speaker 04: They did everything they could. [00:14:50] Speaker 04: They really worked at it. [00:14:51] Speaker 04: They hired a lot of really expensive economists and accountants, and they just couldn't come up with a reliable system. [00:14:58] Speaker 04: It just rested on too many assumptions and guesswork. [00:15:01] Speaker 04: And so they just said, look, we can't in good conscience do this because we can't defend the methodology. [00:15:08] Speaker 04: So we're not going to include indirect costs. [00:15:10] Speaker 04: What's your argument that that would violate the statute? [00:15:13] Speaker 03: So, Your Honor, the two reasons they gave for not attributing any information... You don't want to answer my question, do you? [00:15:20] Speaker 03: You don't want to answer my question? [00:15:21] Speaker 03: I think that the rationales they gave don't make sense. [00:15:24] Speaker 03: They don't make economic sense. [00:15:25] Speaker 04: They don't make... Okay, that's your arbitrary and capricious argument. [00:15:28] Speaker 04: I get that. [00:15:28] Speaker 04: If the Commission says we can't allocate these indirect costs and that analysis is arbitrary and capricious, then you win. [00:15:37] Speaker 04: But I was operating on the assumption that their methodology is not arbitrary. [00:15:41] Speaker 04: She said it's perfectly sound, and that you don't even disagree with it. [00:15:46] Speaker 04: And they conclude, because of this methodologically sound system of trying to allocate these costs, they can't. [00:15:52] Speaker 04: So therefore, they're not going to. [00:15:54] Speaker 04: And I'd still like to know why we're not going to allocate indirect costs, why you think that would violate the statute. [00:16:00] Speaker 03: I see that hypothetically if they had scoured the earth and included everything in it, then I think in the end you might have a statute that made no sense and the agency has to do what it has to do and that would be lawful. [00:16:12] Speaker 04: So then how can it be lawful for them to exclude all indirect costs through a methodologically sound system? [00:16:23] Speaker 04: but to include some but not all through the same methodologically sound system. [00:16:29] Speaker 04: You know what I mean? [00:16:30] Speaker 04: Methodology. [00:16:31] Speaker 04: Though it's here, same hypothetical. [00:16:35] Speaker 04: Suppose you think that, suppose you agree with me that their methodology is sound, that they've come up with the best system for allocating costs, and that they can only allocate 40% of inframarginal costs, of indirect costs. [00:16:49] Speaker 03: That wouldn't violate the statute, right? [00:16:51] Speaker 03: So if they had a good system and they allocated or attributed 40% indirect costs, that would be consistent with the statute. [00:17:00] Speaker 03: But the problem we have is the commission doesn't do that has chosen a standard, which is a standard of unequivocal causation. [00:17:06] Speaker 01: And that's the word unequivocal. [00:17:08] Speaker 01: Is that coming? [00:17:09] Speaker 01: Is that your gloss on it? [00:17:10] Speaker 01: Or is the coming [00:17:10] Speaker 03: That's from the Capitol Commission, your honor, which was one of the commissions that studied this and it appears in the legislative history. [00:17:15] Speaker 03: It's one of two definitions. [00:17:17] Speaker 03: But again, the commission doesn't tarry with the definitions here. [00:17:20] Speaker 03: It just doesn't look at the statute. [00:17:22] Speaker 03: And it feels like the commission just hasn't made the transition. [00:17:24] Speaker 01: Can I just ask you one terminological question that's bound up? [00:17:28] Speaker 01: You only have one in this case? [00:17:29] Speaker 01: Yeah, yeah. [00:17:31] Speaker 01: I have one. [00:17:31] Speaker 01: Maybe it has 17 subparts. [00:17:33] Speaker 01: That's a lot fewer than I do. [00:17:35] Speaker 01: So part of the confusion here, at least in my own mind, is that [00:17:39] Speaker 01: not that the briefs are confusing, the briefs are excellent on all scores, but is that there's four categories of IN in cost. [00:17:47] Speaker 01: There's incremental, inframarginal, institutional, and there's one more, which we're just talking about now. [00:17:56] Speaker 01: There's another one. [00:17:58] Speaker 01: Inframarginal, institutional, incremental, and then [00:18:02] Speaker 01: Indirect. [00:18:03] Speaker 01: Indirect. [00:18:03] Speaker 01: Thank you. [00:18:04] Speaker 01: Indirect, the one we're talking about. [00:18:05] Speaker 01: And they all kind of overlap to some extent. [00:18:07] Speaker 01: They don't overlap to some extent. [00:18:08] Speaker 01: So one thing I'm fixated on right now is this. [00:18:11] Speaker 01: As I understood the way you started your argument on the statute, it was that the problem with the commission's methodology is that indirect basically doesn't get captured. [00:18:21] Speaker 01: And the way you showed that is you said, look, if you look at the curve that has inframarginal [00:18:28] Speaker 01: basically none of that is getting covered. [00:18:30] Speaker 01: Maybe some tiny sliver of infomarginal is getting covered, but basically none of it is. [00:18:33] Speaker 01: But it seems to me that, as I understand the Commission's argument, that's a mismatch because they don't think indirect is infomarginal. [00:18:40] Speaker 01: In fact, they think indirect and infomarginal are mutually exclusive. [00:18:44] Speaker 01: And you're suggesting that the reason we know indirect [00:18:48] Speaker 01: isn't being captured, it's because they're capturing hardly any of inframarginal. [00:18:53] Speaker 03: Am I right in understanding that you... That's not actually our argument. [00:18:57] Speaker 03: There are indirect costs in inframarginal costs, but it's not that the term indirect is coincides with inframarginal. [00:19:04] Speaker 03: Indirect costs are costs that are caused by two products jointly. [00:19:08] Speaker 03: such that you can't unequivocally say it's caused by A or B. At that point, you have to use your best effort. [00:19:13] Speaker 00: That's what I thought. [00:19:14] Speaker 00: I thought you're saying that the commission is, as soon as something is jointly caused, it is not reliably attributable, and therefore, they're not going to count it. [00:19:27] Speaker 00: And you're saying if that's the reason why they're saying we have no reliable way to count that, that that's way too [00:19:35] Speaker 00: for want of a better term, lazy, that in fact, just because something's jointly caused doesn't, and by that token, indirect, that doesn't put it off the map. [00:19:46] Speaker 00: But I don't actually, I understand that that's your position, but I don't take them to have said that just because it's jointly caused, we can't trace the causation. [00:19:59] Speaker 00: In my view, I see some of the costs that are in the green and some of the costs that are in the red over on the right in their chart, or in your chart, I guess, are jointly caused. [00:20:12] Speaker 00: They're just only willing to look at those jointly caused costs that are in the marginal, meaning [00:20:21] Speaker 00: If we got rid of this, they would go away. [00:20:23] Speaker 00: If we add this, they come back. [00:20:25] Speaker 00: Box. [00:20:26] Speaker 03: I think even intervenors expert who developed this incremental cost test and recommended it actually conceded it does not include joint costs. [00:20:35] Speaker 03: And I would direct the court just to pages 716 to 717 of Amazon's reply. [00:20:41] Speaker 03: And to Hansar's statements at SEPJA571, I think that would probably be very enlightening. [00:20:47] Speaker 03: But if I could just turn really quickly, the one problem we really have in talking about this is the Commission hasn't confronted what the statute means. [00:20:54] Speaker 03: The Commission looks at the legislative purpose, ensuring fair competition, and then says, our test, incremental cost, is not designed to ensure fair competition. [00:21:04] Speaker 03: Then what is? [00:21:05] Speaker 03: The Commission is stepping back from that. [00:21:07] Speaker 03: It feels like the Commission has never made the transition from being the Postal Rate Commission, which recommends outcomes and disputed rate proceedings, to the Postal Regulatory Commission, which looks at the words of the statute, looks at the statutory purpose, and reconciles the choices it makes with those elements. [00:21:22] Speaker 03: If there are no further questions, I'm well beyond my rebuttal. [00:21:26] Speaker 03: Thank you very much. [00:21:39] Speaker 02: May it please the court, Mike Shi for the Postal Regulatory Commission. [00:21:44] Speaker 02: As the panels recognize, this case concerns the Commission's method of calculating what the statute terms costs attributable, which are those costs that bear a, quote, reliably identified causal relationship to a postal product. [00:21:57] Speaker 02: For more than four decades, the Commission has determined that the only such costs, apart from the minimal fixed costs that are specific to a product, are marginal costs. [00:22:06] Speaker 02: and the Supreme Court endorsed that determination in the 1983 greeting card case. [00:22:10] Speaker 02: The challenged orders are fully consistent with the Commission's long-held view of what a reliably identified causal relationship entails. [00:22:18] Speaker 02: The order simply refines the methodology the commission uses to calculate marginal costs by including some inframarginal costs not included in the previous method of calculation. [00:22:29] Speaker 02: So as a result, the commission now deems more costs to be costs attributable than it did under the previous methodology. [00:22:37] Speaker 02: And because that decision is both reasonable and lawful, the petition for review should be denied. [00:22:41] Speaker 01: So can I ask you a question about the statutory term indirect costs? [00:22:45] Speaker 01: Yes. [00:22:46] Speaker 01: in that explanation you just gave, does indirect costs fit in? [00:22:50] Speaker 01: And in particular, there was a prior methodology under which the attributed costs consisted of product-specific fixed costs plus volume variable costs. [00:23:02] Speaker 01: And then there's an increment that's been added on now where it includes not only those two, but also that share of infomarginal costs that are part of incremental costs. [00:23:11] Speaker 01: Am I right so far? [00:23:12] Speaker 01: So far, so good. [00:23:14] Speaker 01: Then there's also the statutory term indirect costs. [00:23:18] Speaker 01: And the question I have is, your position is that at least some indirect costs are being attributed. [00:23:26] Speaker 02: Yes, to the extent that there is a reliably identified causal relationship between that indirect cost and the production of a product. [00:23:32] Speaker 01: Right, so there is at least some indirect costs that are being attributed because there's a requisite causal relationship. [00:23:37] Speaker 01: Was that always the case? [00:23:39] Speaker 01: Yes. [00:23:39] Speaker 01: OK, so the addition of that sliver of inframarginal that fits with an incremental doesn't treat with indirect, because some share of indirect costs have always been in there. [00:23:51] Speaker 02: So I think that question gets to the heart of UPS's confusion about what the commission has actually done. [00:23:56] Speaker 02: So UPS's statutory argument. [00:23:58] Speaker 02: essentially attempts to equate the term indirect costs with the term inframarginal costs. [00:24:04] Speaker 02: And UPS says one must equal the other. [00:24:06] Speaker 02: But that's not how the Commission has interpreted the term indirect costs. [00:24:11] Speaker 02: So since the mid-1970s, the Commission has construed that term to refer to what it has inelegantly called piggyback costs. [00:24:18] Speaker 02: And so these are costs that do not vary with a particular product, but do vary with other costs that vary with the base product. [00:24:25] Speaker 02: So the best example of this, Your Honor, is supervisors. [00:24:27] Speaker 02: So the number of supervisors doesn't vary directly with the number of letters produced, but the number of supervisors does vary with the number of employees who actually do the line level distribution. [00:24:37] Speaker 01: And just on understanding this hypo, because I know it's in your brief, maybe it's not even a hypo, but... No, this is what we do and what we have done since the 1970s. [00:24:44] Speaker 01: The supervisors that you're talking about... [00:24:46] Speaker 01: are not supervisors that are linked to and only to one product. [00:24:50] Speaker 01: Exactly. [00:24:51] Speaker 01: They're supervisors that cut across products. [00:24:53] Speaker 02: Right. [00:24:53] Speaker 02: So the supervisor doesn't just deal with the folks who are distributing letters. [00:24:56] Speaker 02: The supervisor also deals with the folks who are sorting parcels and the folks who are sorting flats. [00:25:00] Speaker 01: So then if I can follow up with this question, which also goes to terminology and the methodological explanation you were just elucidating. [00:25:08] Speaker 01: So your view of UPS's argument is that they're conflating indirect with inframarginal. [00:25:16] Speaker 01: That's your view of theirs. [00:25:18] Speaker 01: They may have a response to that and say no, we're not, but that's your view of theirs. [00:25:21] Speaker 01: As I understand your view of your own argument, you think that inframarginal and indirect, in fact, are mutually exclusive. [00:25:29] Speaker 01: that indirect is just a different category than inframarginal because inframarginal only consists of those costs that directly vary with volume, whereas indirect, those indirect costs that are being attributed don't directly vary with volume. [00:25:45] Speaker 01: They may indirectly vary with volume through some extra step like you go from [00:25:51] Speaker 01: the volume to the line employee to the supervisor, you've got the line employee in between. [00:25:55] Speaker 02: I don't think mutually exclusive is the right answer. [00:25:57] Speaker 01: You don't think mutually exclusive is the right answer. [00:25:59] Speaker 01: They would be overlapping, no? [00:26:00] Speaker 01: They're overlapping, absolutely. [00:26:01] Speaker 01: Then your brief says, let's see, I'm looking at page 51 of your brief. [00:26:09] Speaker 01: Let me just find the right spot. [00:26:17] Speaker 01: I thought I read on, yeah, at the top of the, in the partial paragraph at the top of the page, at the end of that paragraph, the last sentence says, because indirect costs are not inframarginal, the commission's revised cost attribution method will not affect their attribution. [00:26:34] Speaker 02: I'm sorry for the confusion. [00:26:36] Speaker 02: Let me try to just explain what we meant by that admittedly opaque statement. [00:26:40] Speaker 02: So the Commission thinks of indirect costs and inframarginal costs as two different categories. [00:26:45] Speaker 02: The definition of inframarginal costs are those costs that are saved due to economies of scope and scale. [00:26:51] Speaker 02: They may be indirect. [00:26:52] Speaker 02: They may not be indirect. [00:26:54] Speaker 02: I see. [00:26:54] Speaker 02: It's just difficult to figure out how to give you an example of what an inframarginal costs are, because inframarginal costs are, by definition, determined by an equation. [00:27:02] Speaker 00: Aren't volume variable costs also potentially including direct and indirect costs? [00:27:11] Speaker 02: Yes, you're right. [00:27:13] Speaker 02: One question is, is the cost a direct or indirect cost? [00:27:17] Speaker 02: And the other question is, how is there, if there is, a causal relationship between that cost and the production of a product? [00:27:25] Speaker 02: So UPS wants to suggest that all inframarginal costs must be construed as indirect costs. [00:27:33] Speaker 02: And as I think our previous conversation illustrated, that's just an incorrect reading of the statute, or at a minimum, the statute does not unambiguously disclose that meaning. [00:27:43] Speaker 02: But even if UPS is correct, that indirect costs includes all inframarginal costs, which we've just said is not the correct interpretation of the statute. [00:27:52] Speaker 02: UPS still can't win the petition because the Commission has demonstrated that its proposed methodology has two serious analytical flaws, and it might help for me to describe those analytical flaws. [00:28:06] Speaker 01: In the course of that, can you just explain to me that statement in a brief, because indirect costs are not inframarginal? [00:28:11] Speaker 01: Is what you mean by that, because indirect costs may or may not be inframarginal? [00:28:15] Speaker 02: So, I'm so sorry, Your Honor. [00:28:18] Speaker 02: The purpose of that statement in the brief is to illustrate that the question of what is an indirect cost is very different from the question of what is an inframarginal cost. [00:28:28] Speaker 02: That's all that sentence in the brief was getting at. [00:28:30] Speaker 02: And the reason why UPS can't prevail, even assuming its reading of the statute is correct, [00:28:35] Speaker 02: is because its proposed method of allocating it for marginal costs suffers from two primary defects. [00:28:41] Speaker 02: The first defect is that UPS needs to figure out the shape of the entire cost curve, and the method it has adopted for doing that is this model by an expert named McBride who adopts the constant elasticity assumption. [00:28:54] Speaker 02: Well, the Commission explained, and this is at pages 1058 and 59 of the Joint Appendix, [00:29:00] Speaker 02: that the constant elasticity assumption is difficult to verify because who knows what the shape of the cost curve at very low volumes are. [00:29:10] Speaker 02: The Postal Service, as the commission further explained, has never experienced that level of volume. [00:29:16] Speaker 02: Not even in the days of the Pony Express, the Postal Service delivered one piece of mail. [00:29:20] Speaker 00: But I guess if you know your total costs, why does it matter? [00:29:26] Speaker 00: Because you're not really... [00:29:27] Speaker 00: extrapolating in terms of a curve, once you're no longer distinguishing between the product A, product B, and product C along the inframarginal cost curve, if you're just saying we're gonna put them all in one box, we're gonna stir them all together, and we're gonna allocate them proportionally to products, it doesn't really matter what the curve looks like because you do know total costs, and you can just proportionally allocate them to products that way. [00:29:56] Speaker 02: So the commission addressed that argument, too, which UPS did bring up. [00:29:59] Speaker 02: And the reason the commission said that that was a difficult way to consider is because the total cost category includes everything. [00:30:07] Speaker 02: It could include product-specific fixed costs. [00:30:09] Speaker 02: It could include, you know, [00:30:11] Speaker 02: just general common fixed costs. [00:30:13] Speaker 02: It can include a whole bunch of different types of costs. [00:30:16] Speaker 02: And it's very difficult to disentangle all of those costs such that the only result of the subtraction your honor is positing is the total amount of inframarginal costs incurred. [00:30:26] Speaker 02: And so that reasonable explanation of why the commission's view is that UPS's proposal is methodologically flawed, Warren. [00:30:34] Speaker 00: And you couldn't pull out all the costs that are the fixed costs that are directly attributable [00:30:40] Speaker 00: to a product and say, we're going to do those separately, and then everything that isn't, we're just going to cut that up and allocate it proportionally. [00:30:47] Speaker 00: It sort of seems like that's where their proposal is driving. [00:30:50] Speaker 02: Right. [00:30:50] Speaker 02: And, you know, Your Honor, even if you disagree with the commission as to that, as to the possibility of pulling out the various costs, there's still the second problem with UPS's methodology, which is the distribution [00:31:02] Speaker 02: of this hypothesized bucket of inframarginal costs. [00:31:06] Speaker 02: And here, UPS proposes to use the Postal Regulatory Commission's existing distribution keys. [00:31:13] Speaker 02: And distribution keys, under UPS's view, can be used to divide up proportionally all of the inframarginal costs incurred. [00:31:22] Speaker 02: Well, the problem with the use of distribution keys in this context, as the Commission explained, this is on Page's [00:31:29] Speaker 02: 1069 and 1070 of the joint appendix is that it's impossible to know where on the curve a particular product falls. [00:31:37] Speaker 02: The core assumption of the use of distribution keys is that there's no principled way to determine where along the curve any individual male piece belongs. [00:31:46] Speaker 02: So we'll just assume that it's essentially random. [00:31:49] Speaker 02: And the commission said there's no reason to assume it's essentially random. [00:31:52] Speaker 02: And there's no way to test for this because there is no such thing. [00:31:55] Speaker 00: But isn't that the nature of your [00:31:58] Speaker 00: economic disagreement. [00:32:00] Speaker 00: I mean, I take it to be they're saying, yeah, it's random. [00:32:03] Speaker 00: That's why you're summing and then proportionally dividing up. [00:32:08] Speaker 00: No problem. [00:32:09] Speaker 00: And you're also saying we don't care about ordering, but as a theoretical matter, what we're asking is [00:32:17] Speaker 00: What's the marginal contribution of this product to cost? [00:32:20] Speaker 00: And that's the question we're asking. [00:32:21] Speaker 00: So it seems like neither of you really cares about ordering. [00:32:24] Speaker 00: Your position isn't a position of ordering. [00:32:26] Speaker 00: It's a position about, theoretically, what question are we asking? [00:32:29] Speaker 00: It's always last, in your view, because what you're asking is a marginal cost question. [00:32:35] Speaker 02: No, Your Honor. [00:32:36] Speaker 02: So it might help for me to back up and just to segregate the two questions here. [00:32:40] Speaker 02: The first question is, are all inframarginal costs [00:32:44] Speaker 02: causally related to the production of a product. [00:32:46] Speaker 02: So that's question one. [00:32:47] Speaker 02: Question two is, assume the answer to that question is correct. [00:32:51] Speaker 02: Does UPS's method of attribution distribute those costs reliably? [00:32:56] Speaker 02: So far, we've been talking about the second half of that question. [00:32:59] Speaker 02: But your question, I think, goes to the first half of the analysis, which is, are all inframarginal costs caused by a particular product in a manner that requires attribution under the statute? [00:33:09] Speaker 02: And the commission addressed that. [00:33:11] Speaker 02: question two. [00:33:12] Speaker 02: And this is what I was getting at with respect to the greeting card case. [00:33:15] Speaker 02: And the Commission says it's simply not the case that all inframarginal costs must be attributed because not all inframarginal costs are caused by the production of a product. [00:33:27] Speaker 02: Rather, only those inframarginal costs that would disappear if the Postal Service stopped making that product are properly considered to be [00:33:35] Speaker 02: caused by the product for purposes of the statute. [00:33:38] Speaker 02: So only those costs need be attributed at all. [00:33:42] Speaker 01: And that share is picked up by incremental cost. [00:33:44] Speaker 02: Yes. [00:33:45] Speaker 02: So the incremental cost test, I'm going to use a phrase that I promise I will not use, picks up the marginal, inframarginal cost as it were. [00:33:54] Speaker 02: And what I mean by that is it's the portion of inframarginal costs that goes away when a particular product is not made. [00:34:01] Speaker 02: And so that is fully consistent with the Commission's view of causation. [00:34:06] Speaker 02: And to underscore the point, the previous methodology that the Commission adopted, this is the product-specific fixed cost plus volume variable costs, did not contain any attribution of inframarginal costs at all. [00:34:18] Speaker 02: This was challenged to the Supreme Court under the functionally identical costing statute, and the Supreme Court said the Commission's understanding of causation was reasonable in Warren's deference. [00:34:29] Speaker 02: There is no difference between the Commission's understanding of causation in this context and the Commission's understanding of causation under the conceivably less accurate way of cost attribution that the Supreme Court endorsed in the greeting card case. [00:34:42] Speaker 02: If anything, as I said earlier, the commission is actually adding more costs to the pot that get attributed under the cost attributable formulation. [00:34:51] Speaker 02: And that surely is not unreasonable. [00:34:53] Speaker 00: That's very helpful. [00:34:53] Speaker 00: Now, you just said you answered no to a question that I posed. [00:34:58] Speaker 00: But then you said what we're looking for is the marginal and inframarginal costs. [00:35:02] Speaker 00: And that's exactly the point that I thought I was making, is the reason you're looking at the last [00:35:08] Speaker 00: part of the curve and you don't really care about ordering in the real world is because theoretically you are looking at the marginal, inframarginal costs that this product causes. [00:35:18] Speaker 00: Is that right? [00:35:18] Speaker 02: Yes, Your Honor. [00:35:19] Speaker 02: So I agree completely with that formulation. [00:35:21] Speaker 02: I only resisted your question due to a very technical observation of commission made in the appendix. [00:35:28] Speaker 02: And the commission said it's not the case that our calculation just goes to the very end. [00:35:33] Speaker 02: It doesn't matter where we pick from, we can pick from the front of the curve, we can pick from the back of the curve. [00:35:39] Speaker 02: It's all just mathematically identical to picking from the back of the curve, because even if you pick from the front of the curve, everything is just going to shift up. [00:35:47] Speaker 02: The objection I had was only to that part of the formulation. [00:35:53] Speaker 04: You know, I want to ask you about one of UPS's arguments about indirect costs. [00:35:59] Speaker 04: They say, you know, your brief explains the commission's understanding of what indirect cost means, but your brief does, but the commission's order does not. [00:36:11] Speaker 04: And they argue we have to remand for an explanation of that. [00:36:18] Speaker 02: So I take issue with both their characterization of the commission's order and with their proposed remedy. [00:36:22] Speaker 02: The reason why the commission didn't spend much time talking about it. [00:36:26] Speaker 02: You agree it didn't. [00:36:27] Speaker 04: No, no, no, I disagree. [00:36:28] Speaker 04: It's all in your brief, but it's not in the order. [00:36:30] Speaker 02: Well, so the order does cite the fact that the commission traditionally uses piggyback costs and for [00:36:36] Speaker 02: Since the 1970s, piggyback costs have been understood to be how the Commission figures out what indirect costs are. [00:36:44] Speaker 02: You can find a reference to the piggyback costs at, for example, footnote 14 at page 1029 of the Joint Appendix. [00:36:51] Speaker 02: And the intervener's brief, this is page 16, footnote 6, also points out several other sources, other Commission orders, where the Commission has expressly linked piggyback costs [00:37:02] Speaker 02: To indirect costs, but I think the premise of your honor's question is suppose we don't think that that's sufficient. [00:37:07] Speaker 02: Well, a remand isn't necessary because it would just serve to promote undue delay. [00:37:12] Speaker 02: And the Taurus Records case that we cite in our brief is illustrative. [00:37:16] Speaker 02: So there, there was a challenge to a DEA decision. [00:37:19] Speaker 02: And the agency's letter literally said nothing. [00:37:21] Speaker 02: But in litigation, the DEA submitted internal memoranda that wasn't previously disclosed between DEA and the Justice Department. [00:37:29] Speaker 02: Petitioner never saw those documents. [00:37:30] Speaker 02: And the court held that remand would serve no purpose because the agency could and no doubt would simply retransmit those memoranda to Petitioner. [00:37:38] Speaker 02: So here, [00:37:39] Speaker 02: It's an even stronger case against remand because the agency's reasoning is not in private memoranda. [00:37:45] Speaker 02: The agency's reasoning about what indirect costs are contained within previous agency orders. [00:37:52] Speaker 02: And as we have explained, in these circumstances, a remand would serve no function. [00:37:57] Speaker 01: So is the part of the indirect costs that are being attributed, those are coterminous with piggyback? [00:38:03] Speaker 02: I don't know the answer to that question, Your Honor. [00:38:07] Speaker 02: So I don't want to say that the only indirect costs attributed are piggyback costs. [00:38:13] Speaker 02: So at least piggyback costs. [00:38:14] Speaker 01: But at least piggyback costs. [00:38:16] Speaker 02: And that's enough to support what the Commission is doing. [00:38:21] Speaker 02: In other words, because the Commission's methodology does give meaning to indirect costs, UPS's argument that only by attributing all inframarginal costs can we give meaning to that part of the statute is incorrect. [00:38:36] Speaker 00: I just was wondering, is the commission taking steps to get better data on some of these questions, you know, fill out the category of [00:38:44] Speaker 00: of indirect costs and work on what might be further attributed to respond further to UPS's concerns. [00:38:51] Speaker 02: Yes, Your Honor. [00:38:52] Speaker 02: So I want to tackle that question with the court's leave in two ways. [00:38:55] Speaker 02: The first is to emphasize that the commission's order explains why not all indirect costs bear sufficient causal relationship to warrant attribution. [00:39:03] Speaker 02: So there's no need for additional data in light of that. [00:39:07] Speaker 02: understanding of causation, which is no different from that endorsed in the greeting card case. [00:39:11] Speaker 02: But if the court's concern is, you know, what's going to happen to all of the inframarginal costs that are not deemed cost attributable? [00:39:17] Speaker 02: Well, those are going to be considered, quote, institutional costs, and institutional costs are allocated according to a reasonable and appropriate sum to the various competitive products, expressly taking market considerations into account. [00:39:32] Speaker 02: And so the Commission is currently in the process of looking at the present allocation of those institutional costs to the various competitive products. [00:39:41] Speaker 02: UPS is an active participant in those discussions, and the Commission has not yet issued a final order. [00:39:47] Speaker 02: So the upshot of that... [00:39:49] Speaker 01: I'm sorry? [00:39:50] Speaker 01: That's under A3. [00:39:51] Speaker 02: Yeah, that's correct. [00:39:52] Speaker 02: That's under A3. [00:39:53] Speaker 02: And the upshot of that argument is that, you know, it's not the case, as UPS suggests, that if you rule for the commission, inframarginal costs are just going to vanish into the ether never to be allocated or assigned again. [00:40:04] Speaker 02: Rather, they're going to be put into the institutional costs bucket [00:40:06] Speaker 02: And the total in that bucket is going to be parceled out according to the Commission in a reasonable way to account for market factors to ensure that the statutorily mandated price floor is not too low as to promote undue competitive behavior. [00:40:22] Speaker 02: Unless the Court has any further questions, the petition for review should be denied. [00:40:26] Speaker 02: Thank you. [00:40:27] Speaker 04: Was Mr. Lerkin out of time? [00:40:31] Speaker 04: Okay, you can take two minutes. [00:40:36] Speaker 03: Thank you. [00:40:38] Speaker 03: Our fundamental position here is that the Commission has to justify not only the choices it rejects, but the choices it adopts. [00:40:46] Speaker 03: All this conversation would be much easier if the order here actually defined indirect costs, but it doesn't. [00:40:53] Speaker 03: At page, I guess it's 1052 of the joint appendix, the order goes through UPS's explanation of what it thinks indirect costs means, but it never provides an answer. [00:41:06] Speaker 03: Council now, before this court says, well, we think it's just piggyback costs, but the commission has so completely failed [00:41:13] Speaker 03: to construe the statutory terms, that it has to reach back for that definition to testimony from the Postal Service in the 1970s. [00:41:20] Speaker 03: That's testimony from the regulated entity, not from the regulator. [00:41:25] Speaker 03: The Postal Commission has to make the decisions that it knows what the statute means. [00:41:32] Speaker 03: It has to define it. [00:41:34] Speaker 03: And the Commission makes exactly that same mistake again when it comes to the issue of the statutory purpose. [00:41:39] Speaker 03: The commission says, we admit the statutory purpose, section 3633, its purpose is to ensure fair competition. [00:41:47] Speaker 03: But it never steps back and asks, hey, is using the minimum measure, assuming that all of the products use the smallest and cheapest elements of cost at the bottom of the curve, is that consistent with ensuring fair competition? [00:42:02] Speaker 03: Instead, the commission says, first, [00:42:05] Speaker 03: The purpose is to ensure that the Postal Service competes fairly in the provision of competitive products. [00:42:10] Speaker 03: And then it turns around and says, the purpose of the incremental cost test selected is not to ensure that the Postal Service is competing fairly in the marketplace. [00:42:18] Speaker 03: If that contradiction is not arbitrary and capricious decision making, it's hard for me to know what would be. [00:42:23] Speaker 00: Mr. Lampkin, on what basis do you conclude that the volume variable costs don't also include among them shared costs? [00:42:37] Speaker 03: We don't believe that the volume variable costs, which is marginal cost, time, product, volume, and then you divide that up according to the cost drivers. [00:42:44] Speaker 03: That we don't believe includes joint costs all either. [00:42:48] Speaker 03: Because the way you're dividing those up, the way it's done, is if it isn't unequivocally assignable to a particular product, it gets excluded. [00:42:55] Speaker 03: And the way you can figure out if it's unequivocally assignable in this test is if that product goes away, the cost goes away. [00:43:03] Speaker 00: But I thought that was sort of done en masse, and that the way you can know that a volume variable cost is attributable is that [00:43:12] Speaker 00: it's that level of cost is present even at the lowest point in the curve. [00:43:18] Speaker 00: And so you don't have to go through and say, well, is this a supervisor? [00:43:21] Speaker 00: Is this an employee? [00:43:22] Speaker 00: Is this a box? [00:43:23] Speaker 00: Is this space in a truck? [00:43:25] Speaker 00: Is this fuel? [00:43:26] Speaker 00: It could be any number of those things. [00:43:29] Speaker 00: But it's enough to know that it's accrued by any product. [00:43:36] Speaker 03: So the volume variable costs are divided up by cost components. [00:43:40] Speaker 03: And so you figure out the cost of trucking. [00:43:42] Speaker 03: And you figure out how much of that trucking is used by this particular product, 15%, 15% cost of wood. [00:43:47] Speaker 03: And you add up all of the different cost components for each product. [00:43:50] Speaker 00: So you're doing it by formula rather than by a sort of cause in the world. [00:43:56] Speaker 00: And it just seems to me that it's rife with joint costs. [00:44:01] Speaker 03: I don't think so, Your Honor, because I think that when they allocate and say, if you use up 15% of this cost driver, is the technical term they use, then you get attributed 15% of that cost driver. [00:44:10] Speaker 03: And that works, because if you stop using 15% of that cost driver, 15% of that cost driver goes away. [00:44:16] Speaker 03: But if the commission's order actually told us what it thought indirect cost means, [00:44:20] Speaker 03: And it answered what UPS put to it on page 1052 of the Joint Appendix. [00:44:25] Speaker 03: We would not be having this conversation, or at least we'd be having a very different one. [00:44:29] Speaker 03: Because the agency hasn't even explained its decisions, much less attempted to reconcile them with statutory purpose, we think the judgment must be reversed. [00:44:37] Speaker 03: Thank you. [00:44:38] Speaker 04: Thank you both. [00:44:39] Speaker 04: The case is submitted.