Google Has Not Shown That Pass-Through Effects Would Benefit Search Consumers

B. Purported Pass-Through Benefits Are Unsubstantiated And Do Not Justify Competitive Harms To Consumers In Relevant Markets

1279. Google’s payments are not passed through to consumers, and even if they were, the payments do not outweigh harm to competition in the relevant markets. Tr. 10527:12–19; 10535:21–10538:4 (Whinston (Pls. Expert)).

1. Google Did Not Demonstrate That Its Search Payments Are Passed Through To Consumers In The Form Of Lower Retail Phones Or Improved Services

1280. Google has not shown that its revenue share payments to distribution partners are passed through to consumers, whether in the form of lower device prices or improved products and services.

a) Google Has Not Shown That Payments To Apple And Android Partners Are Passed Through To Consumers

1281. Google has not established that its revenue share payments to Apple, Android OEMs, and U.S. carriers are passed through to consumers as lower retail prices for smartphones.

1282. The ISA and RSAs do not constrain how Google’s distribution partners use their revenue share payments, let alone require them to pass all or a portion of those payments through to consumers in the form of lower smartphones prices. Tr. 4944:2–4 (Braddi (Google)) (Google does “absolutely not” impose any conditions on Apple with regard to how Apple spends its revenue-share payments from Google.); Tr. 2466:11–17 (Cue (Apple)) (“[T]he money is just Apple’s to decide how to use it.”); Tr. 9565:20–9566:15 (Rosenberg (Google)) (Google does not require Android distribution partners’ to use revenue-share payments to lower phone or wireless prices.); Des. Tr. 207:8–13 (Giard (T-Mobile) Dep. ) (Google’s 2021 RSA with T-Mobile does not limit how T-Mobile uses the money); e.g., JX0033 at -797–99 (§ 4) (Apple ISA (2016 amend.)) (providing payment terms without limits on how payments are used); JX0071 at -401 (§ 4.1) (Samsung RSA (2020)) (same); JX0091 at -751 (§ 4.1) (AT&T RSA (2021)) (same); JX0093 at -498 (§ 4.1) (Verizon RSA (2021)) (same); JX0095 at -696 (§ 4.1) (T-Mobile RSA (2021)) (same).

1283. Google’s distribution partners do not earmark revenue-share payments for any particular use. Apple, for instance, does not designate the money it earns through the ISA for use in lowering retail device prices or improving services, including its preinstalled browser. Tr. 2466:11–17 (Cue (Apple)). T-Mobile similarly does not earmark Google’s revenue share and bounty payments for any specific purpose. Des. Tr. 214:18–215:1 (Giard (T-Mobile) Dep.).

1284. No OEM or carrier executive testified at trial or in designated testimony that their company takes revenue-share payments into account when setting device prices.

To the contrary, each testified either that there is no direct connection between Google’s RSA payments and the retail price of phones, or that they are not aware of any connection. Des. Tr. 215:19–216:6 (Giard (T-Mobile) Dep.) (revenue-share payments do not directly factor into the retail price of T-Mobile phones and, although T-Mobile services and offerings indirectly factor into how T-Mobile prices its devices, RSA payments are not “a driving factor” or “even the largest factor”); Des. Tr. 90:19–91:1 (Christensen (Motorola) Dep.) (disclaiming any knowledge of whether RSA payments directly lower smartphone costs).

1285. Similarly, no Google employee who testified at trial or in a designated deposition had actual or personal knowledge about specific distribution partners using revenue-share payments to lower device prices. Google employees disclaimed knowledge of any internal analyses regarding distribution partners’ use of revenue-share payments to lower retail prices, and Google’s economist testified that he did not model the issue or quantify the degree of passthrough.

Tr. 9565:1–9566:13 (Rosenberg (Google)) (disclaiming awareness of any Google analysis into Android partners’ use of revenue-share payments to lower phone or wireless plan prices); Tr. 10162:4–7 (Murphy (Def. Expert)) (Google makes no effort to track how Android partners use revenue-share payments.); id. 10162:8–13 (not aware of testimony showing a direct connection between revenue share payments and Android phone prices).

1286. Google’s ordinary-course documents do not show that revenue-share payments are passed through to consumers. Notably, when analyzing potential search distribution agreements, Google does not consider how its revenue-share payments will affect smartphone prices. Tr. 10183:13–18 (Murphy (Def. Expert)) (discussing UPX1128 and agreeing that supporting lower-priced Android phones was not cited as a rationale for Android RSAs); UPX1128 at -097 (seeking approval from Google CFO for new Android RSAs, with no reference to phone prices as a rationale); UPX0161 at -053 (same); UPX0129 (August 2019 primer on the Android Commercial Agreements makes no mention of lowering retail prices as a purpose or effect of RSAs.).

Similarly, when Google considered reducing revenue share payments, there was no discussion of how the cuts would affect phone prices. UPX0293 at -425 (seeking approval from BC for Android revenue share deal that cut revenue share for Verizon from [redacted]% to [redacted]%, with no reference to effect on phone prices or output).

1287. Prof. Murphy’s claim that revenue-share payments are passed on to consumers in the form of lower retail prices on higher quality devices is based on a flawed reading of the evidence and directly contradicted by trial testimony and exhibits. Tr. 10535:21–10536:15 (Whinston (Pls. Expert)).

1288. Lacking documentary or testimonial evidence to support his claim that revenue share payments are passed on to consumers, Prof. Murphy relies on general economic theory rather than econometric analysis. Tr. 10163:20–21, 10169:11–14 (Murphy (Def. Expert)). The only data analysis Prof. Murphy provided in support of his pass-through theory was a comparison of (1) Apple’s reported profit margins for device services, which includes Google’s revenue-share payments to Apple under the ISA, to (2) Apple’s reported overall profit margins for devices. Id. 10168:9–10169:2.

Because the rising services profit margin coincided with a decreasing device profit margin, Prof. Murphy hypothesized that Apple may be passing Google’s search payments through to consumers in the form of lower device prices. Id. at Tr. 10169:3–10.

1289. Instead of conducting an econometric analysis, Prof. Murphy relied on general economic theory to prove causality between payments to Apple and iPhone prices going down. Tr. 10169:11–14 (Murphy (Def. Expert)).

1290. As a result, Prof. Murphy’s analysis of Apple’s relatively constant services margin is unreliable, potentially reflecting a wide range of unrelated factors. Tr. 10535:21– 10536:15 (Whinston (Pls.’ Expert)) (Prof. Murphy’s margin analysis was not reliable).

Apple’s services margin includes costs and revenues from various products and services besides revenue share payments from the ISA, and trends related to those products and services could explain why Apple’s services margin increased while its overall device margin stayed constant. Tr. 9711:5–9712:22 (Murphy (Def. Expert)) (“I want to make clear, I don’t want this to be misleading. I’m not just looking at default revenues here in the service margin. Service margin is going up for other reasons.

They sell other things other than search.”); Tr. 9709:19–9711:4 (Murphy (Def. Expert)) (“. . . [S]ervices would be including the default payments, but it also includes other service revenue this firm was getting . . .[.]”).

1291. Apple’s devices margin could also be affected, at least in part, by sales in recent years of lower-priced and older-model phones. Tr. 10169:19–10170:9 (Murphy (Google’s Expert)) (agreeing that Apple’s sale of lower-priced and older-model devices could account for at least part of the trend in Apple’s services margin data).

1292. Ultimately, as Prof. Murphy concedes, the trends observed in Apple’s services margin data could be mere coincidence and thus provide no support for Google’s purported justification. Tr. 9709:19–9711:4 (Murphy (Def. Expert)) (“Now, it’s also possible that . . . there was no passthrough going on here, it was just a coincidence.”).

1293. Moreover, even if there was compelling evidence of the existence of passthrough, Google introduced no evidence that measures—qualitatively or quantitatively—the amount Google’s payments have lowered retail device prices and services. Tr. 10163:17–21 (Murphy (Def. Expert)) (aware of no econometric data showing that Google’s payments to OEMs and carriers affect the price of Android devices); id. 10169:11–14 (confirming he conducted no econometric analysis to link Google’s ISA payments to Apple and iPhone prices).

b) Third-Party Browsers

1294. Google failed to establish that revenue-share payments to third-party browser companies result in more innovation and better products. Google has not identified any specific features or innovation that were introduced as a result of Google’s revenue-share payments.

1295. Prof. Murphy’s data showing that Google’s revenue-share payments comprise a large percentage of Mozilla’s total revenue does not establish that those payments have made Mozilla’s Firefox browser better or more innovative.

Mozilla offers several products besides Firefox, and Mozilla uses its search revenue to finance the development of and updates to those products. Des. Tr. 45:15–17, 45:19–21 (Baker (Mozilla) Dep.) (“Our Search revenue has funded products other than Firefox, yes.”); UPX0979 at -407 (discussing Mozilla’s VPN and password manager products); UPX2080 at -123 (discussing development of the FirefoxOS).

1296. A portion of Mozilla’s revenue is also used for executive compensation. Mozilla’s Chairwoman, Mitchell Baker, earned $2.5 million in 2020 and between $3 and $5 million in 2021. Des. Tr. 292:20–293:9 (Baker (Mozilla) Dep.). Her compensation is based in part on Mozilla’s revenue. Id. 292:13–19.

2. Even If There Was Evidence Of Pass-Through, Lower Retail Prices And Improved Browsers Would Not Justify Harms To Competition

1297. More fundamentally, even if Google had demonstrated pass-through, that purported justification would not undermine Plaintiffs’ prima facie case because (1) revenue share payments would be higher in a competitive market; (2) Google has not established benefits in the relevant markets; and (3) there are less restrictive means of lowering device prices and subsidizing browser innovation.

a) Revenue Share Payments Would Be Higher In A Competitive Market

1298. Even if Google had demonstrated that distribution partners like Apple and Mozilla pass through a portion of Google’s revenue-share payments to consumers, payments to distribution partners—and the concomitant effect on device prices and products—would be larger absent the restrictive terms in Google’s contracts. Tr. 10527:12–10528:25 (Whinston (Pls. Expert)) (stronger rivals would bring more competitive pressure to the market for default arrangements).

Under less-restrictive alternatives to Google’s contracts—including ones in which Google’s contracts permit rivals to compete for some search access points on devices— rivals in the general search services market would have a better chance to gain traffic and scale and become stronger competitors.

Stronger competitors in the market would pressure Google to increase revenue share payments. Tr. 10526:12–10528:25 (Whinston (Pls. Expert)); Des. Tr. 271:8–272:2 (Baker (Mozilla) Dep.) (more competition in the search market would help Mozilla in negotiations for the Firefox default); Tr. 3504:18–3505:23 ((Nadella) (Microsoft)) (Microsoft has not won default placement on Safari, but Apple has used Bing “to bid up” the revenue share it gets from Google).

1299. Real-world evidence confirms that even the limited competition that exists in the marketplace today has increased Google’s revenue-share payments to distributors. Apple has benefited from having Bing in the general search services market because, at least in the short term, Microsoft bids up the price that Google pays for Apple’s Safari default. Tr. 3504:18– 3505:23 ((Nadella (Microsoft)); Tr. 2725:1‒20 (Parakhin (Microsoft)) (“They [Apple] use us [Bing] as a bargaining chip because the threat of them switching allows them to extract better conditions from Google . . . . It is no secret that Apple is making more money on Bing existing than Bing does.”).

In the European Union, Google’s revenue-share share payments to Android partners increased after the introduction of a choice screen allowing Android users to select their default search engine. Tr. 9499:14–9500:9 (Rosenberg (Google)) (Google increased RSA payments in Europe as a result of the European choice screen ruling); UPX0162 at -290 (Google’s first rationale for increasing RSA payments to Android partners in European countries was that the European choice screen created “opportunity for rivals to secure full Search exclusivity on devices in [European Economic Area]”); UPX0163 at -235 (Google’s willingness to increase incremental payments to non-Samsung OEM Android partners driven in part by “[c]hanges to Android’s business model in Europe, which have created opportunities for rivals to invest aggressively to win Search and Assistant access points”).

b) Google Has Not Shown That Pass-Through Effects Would Benefit Search Consumers

1300. Even if Google’s payments are to some degree passed-through to consumers in the form of lower phone prices or improved browsers, lower prices and better products benefit consumers of smartphones and browsers and not necessarily consumers of general search services or related advertising markets. Google has not shown that lower retail phone prices or improved browsers have benefited consumers in the general search market.

1301. Google argues that revenue-share payments must benefit search consumers because the total amount of searches (i.e., search output) is increasing, but that analysis is flawed. Prof. Murphy observes that output in the search market generally has increased over the years, but, as he acknowledges, his analysis does not tie that increase to phone prices. Tr. 9847:5–9848:6 (Murphy (Def. Expert)) (growth in mobile search “comes from many things,” and “I can’t tell you how much of that is due to that competition . . . [.]”). Nor does his analysis tie increased search output to browser improvements.

1302. Indeed, there is no evidence that the increase in search output over the years is related to decreased smartphone prices or improved browsers. Many external factors have driven the increase in internet searches over the years, including the growth of the internet, trends in mobile phone usage (including adoption and now ubiquity of smartphones), expanding access to broadband, and improvements in wireless telecommunications technology. Tr. 10456:17– 10458:18 (Whinston (Pls. Expert)) (“The internet has been expanding, connectivity has been improving . . . many, many things have contributed to the output going up.”).

1303. Google’s ordinary course documents do not indicate that ISAs, RSAs, and third-party browser distribution agreements increase search output. Notably, Google prepares sophisticated and detailed modeling of potential revenue share deals, and those models would be expected to factor in increases to search output if they existed.

But Google’s modeling does not incorporate any effect the deals have on search output. Tr. 10536:16–10538:4 (Whinston (Pls. Expert)) (discussing UPX1050 and noting that although Google made “very detailed calculations,” nowhere did it account for pass-through); UPX1050 at -868. For instance, when modeling different revenue share options for Apple under the ISA, Google did not project that greater payments to Apple would result in more devices being purchased or more searches being run. Tr. 10536:16–10538:4 (Whinston (Pls. Expert)).

c) There Exist Less Restrictive Means Of Lowering Devices Prices And Improving Services

1304. There are many alternatives to Google’s distribution agreements that would be less restrictive than Google’s contracts and leave a “more equal playing field in terms of [general search] distribution,” but still achieve Google’s purported pass-through benefits. Tr. 5776:22– 5778:3 (Whinston (Pls. Expert)).

1305. Google could negotiate contracts with most-favored supplier terms that continue paying out revenue-share payments, but only requiring Google to be treated on equal footing with search rivals. Contracts with those terms would permit distributors to set up choice screens, reducing competitive harms, while still providing a robust revenue stream for distributors. Tr. 10529:1–19 (Whinston (Pls. Expert)).

1306. Google could also negotiate unconditional revenue shares. Tr. 10529:1–19 (Whinston (Pls. Expert)). Under that arrangement, Google would pay distribution partners for Google queries through the device or through specific search access points while permitting distributors to work with competing search providers. Id. 10531:14–16, 10532:6–20. For example, an unconditional revenue share would permit Apple to set Google as the default search engine for most searches on Safari but set a privacy-focused GSE like DuckDuckGo as the default for private browsing modes. Id. 10529:20–10530:17.

1307. Unconditional revenue share arrangements are already common inside the general search services market. Bing, DuckDuckGo, and Yahoo pay for non-exclusive distribution in the United States. Supra ¶ 38; e.g. DX1005 at -158 (§5.1) (Microsoft’s agreement with Mozilla, under which Microsoft pays Mozilla a [redacted]% revenue share for non-exclusive “non default” distribution in the Firefox browser); DX1011 at -322 (§1 (a)) (DuckDuckGo’s agreement with Mozilla, under which DuckDuckGo pays Mozilla a [redacted]% revenue share for non-exclusive distribution in the Firefox browser); UPX0119 at -534 (showing Yahoo’s “non-exclusive” syndication deal, which allows it to “choose which provider to send each query (Bing, Gemini, or Google)”).

1308. Google itself uses unconditional revenue shares. In 2021, each of the carriers signed go-to-market agreements, or MSIAs, that incentivized the distribution of Android devices though payments tied to the number of Android devices on each carrier’s networks. Supra ¶¶ 289–291. Those agreements did not require Google Search be set as the exclusive default search engine on qualifying devices. Id.; Tr. 10166:17–10168:3 (Murphy (Def. Expert)) (acknowledging that Google separated the “search part” from the Android MSIAs).

1309. Unconditional revenue-share models are also common outside the general search market and are used in legal contingency fees, sales commissions, and CEO compensation packages. Tr. 10534:18–10535:9, 10652:7–10656:10 (Whinston (Pls. Expert)).