Buyers’ Cartels: Prevalence and Undercharges

This article supplements information on 24 U.S. domestic buyers’ cartels cited in a 2010 book by Blair & Harrison by assembling and analyzing a sample of 49 episodes of buyers’ cartels that have international membership or multijurisdictional price effects. It appears that such cartels comprise less than 8 percent of the total, that they are clustered in primary-products and services industries, that they employ bid rigging conduct to a greater extent than sellers’ cartels, and that undercharges average close to 20 percent of the but-for price. These quantitative characteristics are supplemented by sketches of the conduct and prosecutions of a few of the better-documented cases of buyers’ cartels.

By John M. Connor¹

 

I. INTRODUCTION

The economic theory of buyers’ price fixing is well established. If buyers are small enough in number and sufficiently cooperative, they may be able to form oligopsonies that can force down the prices of common inputs below the prices that would have reigned in a more competitive procurement market. In other words, powerful buyers can undercharge their input suppliers. Other facilitating conditions may have to be present to generate significant negative price effects. Among them are opportunities for intra-buyer communications behind a wall of secrecy, relatively atomistic suppliers, homogeneous products, geographic isolation, inelastic demand, and barriers for sellers to detect the undercharges.

Because of the economic power of effective buyers’ cartels, little or none of the lower costs need be passed on to their customers on the selling side, undercharges will widen cartel members’ operating margins and generate higher joint economic profits. Indeed, powerful business groups can in principle simultaneously raise selling prices and lower buying prices, garnering two streams of monopoly profits from each side of their businesses. Empirically, absent the detailed data typically revealed in damages cases, such mixed collusive strategies may be difficult to distinguish in practice.

Blair & Harrison (2010) is perhaps the only book-length treatment of the economics of buyer power to be written in the past couple of decades.² One of their principal themes is that monopsony and oligopsony are frequent in natural markets yet are given short shrift empirically in economics and are rarely prosecuted.³ However the empirical evidence they collected is quite modest. Restricting their purview to cases brought in U.S. courts⁴ or documented in publications by American economists, by my count they assemble a sample of 24 documented buyers’ cartels. A large share of these buyers’ cartels appear to have been organized by bidding rings comprised of firms or individuals domiciled in one urban area or nation. Moreover their sample appears to be comprised entirely of domestic U.S. buyers’ cartels. They do not present more than one or two examples of empirical buyer-power estimates.

Blair and Harrison opine that court cases alleging illegal buyers’ collusion are infrequent in part because of the mistaken belief that if buyer power forces down prices below competitive level then consumers must benefit. In fact, if buyers have the power to explicitly collude on the price of a procured input, then an undercharge⁵ is likely to be imposed on suppliers; the undercharge to suppliers is symmetric to the antitrust damages created by overcharges on buyers from sellers’ cartels (ibid. pp. 157-163). In both cases, industry output contracts from the level that would be seen in purely competitive or noncooperative oligopsonistic procurement markets and allocative inefficiency is created.

In this article, I collect and analyze information on international buyers’ cartels throughout history, where international refers to the membership composition of the cartel. The sample is larger than that of Blair and Harrison (2010) and does not overlap with theirs. Most importantly, the international sample is drawn from publications by economists that contain serious, disinterested estimates of the undercharges achieved by the buyers’ cartels.

 

II. THREE HISTORICAL EXAMPLES OF BUYERS’ PRICE-FIXING CARTELS

Case studies of historical buyers’ cartels can illuminate the cooperative conducts that will generate customer undercharges. Three are particularly well studied and documented.

The London coal cartels are among the oldest historical cartels in the empirical legal-economic literature. They affected the wholesale market for heating coal brought by coastal ships sailing from northern England (principally Newcastle) to London via the Thames River. It is certainly an old market.⁶ Records of taxes paid on “sea coal” in London go back more than 800 years to 1213 (Levy 1927: 9). From at least the 16th century, “Newcastle Coal” became the popular name for maritime coal sold in London.⁷

In my opinion, London Coal appears to be two radically different cartels. The London coal-buyers’ cartel began as early as 1595 and persisted on and off for about 150 years.⁸ The buyers were lightermen – wholesale coal merchants who operated river barges and purchased large quantities directly from the owners of coal-laden ships in London’s harbor. At times, they were few enough in number that they were able to manipulate the prices paid in London.⁹ Frequent complaints about high coal prices were forwarded to Parliament, and multiple parliamentary investigations confirmed that the lightermen were enjoying extraordinary financial returns at the expense of homeowners, hospitals, and the like. In 1729, a Parliamentary investigation found that ten lightermen controlled 67 percent of purchases in London, and the committee’s report specifically blamed them for 1722-1729 price increases.¹⁰

Sometime between 1690 and 1770, the locus of power in the London sea-coal market moved north and upstream the supply chain. London consumers of coal were thenceforward exploited by a sellers’ cartel of coalmine owners.¹¹ It is the Coal Gild of Northeastern England (later known as the “Newcastle Vend”) formally organized in 1770. An informal predecessor made its first recorded collusive agreement on London coal prices in 1699. This so-called Newcastle Vend kept meticulous price archives and became among the first cartels to be studied by modern scholars who were interested in applying quantitative methods to price-fixing conduct. The availability of detailed purchase records has permitted sophisticated econometric modeling of the London coal cartels. A notable pair of studies by Hausman (1980, 1984) examines the coal markets from 1699 to 1845.¹² The oldest, informal episodes (1699 to 1770) were ineffectual in raising retail prices, but 11 more formal Vend episodes from 1770 to 1845 showed overcharges that averaged about 7 percent to 8 percent. In the early 19th century when the Vend was best organized, Tan (2003: 22) estimates that five episodes during 1821-1845 resulted in coal overcharges of from 12 percent to 16 percent.¹³ Although highly unstable, the Vend did not finally collapse until 1845, 146 years after its birth. One or both London coal cartels operated for 250 years, making them the most durable cartels I have recorded.¹⁴

The English Government took many public actions in response to the coal cartels. Acts of Parliament forbidding bid rigging were passed in 1642 and 1665, but with no penalties specified. Other UK government actions came too late to affect the buyers’ cartel of lightermen.¹⁵

A second buyers’ cartel involved small numbers of UK copper smelters, most in Swansea district of So. Wales. They initially began rigging bids for ore purchased from numerous miners during 1719-1726 and later rigged bids for export copper. The first buyers’ cartel consisted of four smelters and suppressed copper ore prices during 1719-1726. A second more formal written agreement (“Associated Smelters”) endured from 1737 to 1779. One historian says the second episode was “quite effective” at lowering ore prices and raising copper prices, but no quantitative estimate is provided.

A third of the better documented historical bidding rings is a Rare Books Auction held at Ruxley Lodge in Claygate, Surrey, UK.¹⁶ The large ring of professional booksellers was able to leverage its superior knowledge of retail prices to the disadvantage of the Lodge’s recent inheritor. It suppressed purchase prices by about 400 percent, which is the oldest buyers’ cartel for which an undercharge is available. A leading economist writes:

“At an auction of rare books held at Claygate Estate, Surrey, England over 10 days in October 1919, 81 book sellers formed a bidding ring for 447 of the 641 lots of 13,600 volumes sold; one of the dealers kept a detailed diary of the bidding ring published in 1990; bid rigging was made illegal in the UK in 1927. The ring met secretly later to hold 4 ‘knockout’ auctions among themselves; prices advanced each time; total knockout sales (an undercharge yardstick) were £19,696, up from £3714 paid to the estate at the original auction” (Porter 1992: 434).

 

III. CONTEMPORARY EXAMPLES OF BUYERS’ CARTELS

Additional examples of successful buyers’ cartels have in the past few decades continued to be documented by inquisitve economists. They include the procurement of fresh canning tomatoes in California’s Central Valley (Just and Chern 1980),¹⁷ collusion by millers on the auction prices of wheat in India (Banerji and Meenakshi 2004),¹⁸ bidding rings in auctions of stamps (Asker 2010),¹⁹ and auctions of houses seized for back taxes (Kwoka 1997).²⁰ More recently, novel civil and criminal antitrust litigation has been launched against “no-poach” agreements that suppress the wages of high-tech employees, college basketball players, and “gig-economy” jobs (Delrahim 2019).²¹ For many contemporary buyers’ cartels, estimation of price effects using acceptable forensic quantitative methods of analysis is not always possible.²²

 

IV. UNDERCHARGES BY BUYERS’ PRICE-FIXING CARTELS

The limited survey of historical examples of buyers’ cartels above illustrates that the basic characteristics of many such cartels have been documented, but going the extra step to determine market price effects was too difficult or of less interest than describing their methods of conduct. Quantitative analyses of the size of buyers’ cartels’ undercharges are hard to locate and began to be published mainly from 1980. Stocking and Watkins (1948) were pioneers in estimating undercharges; they discussed three undercharges for the 1937-1939 international Sugar Cane cartel organized by several sugar-producing and -importing nations, including the United States. More sophisticated econometric estimation with detailed transactions micro data began later. Daggett and Narasimhanm (1981) seem to be the first to have published such a study. They test an model with detailed time-series data from the domestic Titanium cartel that operated in North America during the years 1970-1976.

Drawing upon a broad, deep data set of cartels with price-fixing overcharges’ estimates, I analyze descriptive statistics of the prevalence and economic characteristics using a sample of 46 buyers’ cartels (Table 1).²³ This sample comprises 85 estimates of undercharges because three cartels had multiple episodes, and 14 cartels benefitted from the availability of multiple estimates.²⁴

Table 1. Buyers’ Cartels with Episodic Overcharges’ Estimates, 1919 – 2017
Market	Collusion Began	Collusion Ended	Episodes	Estimates
Antiques Auction, Surrey, UK (One Week)	1964	1964	1	1
Auction Houses, Fine-Art, world prices	1993	2000	1	8
Auctions, Houses in DC, U.S.	1967	1990	1	1
Auctions, Used Police Cars, NYC	1990	1991	1	2
Banks, Mexico sovereign bonds, MX & U.S.	2006	2017	1	1
Basmati rice auctions, farmer-wholesalers, Panipat, India	1999	1999	1	2
Beef in California, U.S.	1953	1970	1	1
Bitumen, Netherlands	1994	2002	1	2
Blueberries, Wild, Purchases, farmer-processor in ME, U.S.	1996	1999	1	1
Bond Underwriting, U.S.	1959	1967	1	1
Cattle auctions, cull-cow, Monroe Wis., U.S.	2000	2009	1	1
Cattle, Fed, U.S.	1994	2002	1	1
Cheese, aged, wholesale U.S. prices	1988	1993	1	1
Construction, public works, Quebec, CA	2000	2009	1	1
Dairy Processing /Milk, U.S.	1972	1975	1	1
Frozen Fish, U.S.	1986	1988	1	3
Futures Contracts, Potatoes, Maine, U.S.	1976	1976	1	1
High Tech Employees 2, Film Studios, U.S.	1986	2014	1	1
Insurance, brokers’ contingent fees, U.S.	2001	2004	1	7
Lease oil royalties, U.S.	1986	1993	1	1
Oil and Gas Rights in Michigan, U.S.	2010	2010	1	1
Petroleum, Govt. Offshore Leases, Auctions, U.S.	1954	1975	1	1
Petroleum, Military fuels in Korea, KR & U.S.	1998	2016	2	4
Private equity buyouts, U.S.	1985	2007	2	2
Procurement, several products, by U.S. Defense Department, U.S.	1960	1969	1	1
Rail industry, skilled employees, no-poach, U.S.	2009	2016	1	1
Rare books auction, Surrey, UK	1919	1919	1	1
Real estate auction, Wash. DC, U.S.	1989	1989	1	1
Round Wood Buying, Sweden	1971	1984	1	4
Scholarships, Graduate, U.S.	1958	1991	1	1
Shrimp, Wholesalers/freezers, North Sea Shrimp 2, in NL, EC	1997	2000	1	1
Stamp auctions in NY City and UK, world effect	1980	1997	1	2
Sugar Beets, procurement, farmer-processor, Midwest, U.S.	1939	1941	1	1
Sugar, Cane, world prices	1938	1939	1	4
Timber cutting rights, U.S.	1982	1990	1	4
Timber, procurement, NW U.S. Auctions	1975	1981	1	1
Titanium Metal, U.S.	1970	1976	1	1
Tobacco, leaf, procurement, auctions, U.S.	1996	2001	1	1
Tobacco, leaf, procurement, Italy	1995	2002	1	1
Tobacco, leaf, procurement, various grades in Spain, EC	1996	2000	2	9
Vanadium Ore, Colorado, U.S.	1933	1948	1	2
Wheat auctions, farmer-wholesalers, Nerala, India	1999	1999	1	1
Total 46 cartels			49	85

There are four notable characteristics. First, the number of buyers’ cartels with overcharge estimates is a small (4.3 percent) but growing share of all such cartels. During the pasr three centuries of available data, only 5.7 percent of all reported cartel price effects were undercharges by collusive buyer groups, but that ratio had risen from practically zero before 1990 to above 8 percent after 1990. An alternative measure of prevalence is the number of undercharge estimates, which is 85 out of about 2500 (3.4 percent). Because most cartels are hidden throughout their lives and there are no other large collections focused on buyers’ cartels, it is difficult to know whether 3 percent to 8 percent more accurately reflects the proportion of buyers’ cartels to total cartels.²⁵ The rise in prevalence after 1990 suggests that either prosecutorial interests or the priorities of empirical researchers shifted about 30 or 40 years ago.²⁶

Second, buyers’ cartels are typically found in only a few industry groups. While sellers’ cartels are mainly in manufacturing of industrial inputs, buyers’ cartels are preponderantly discovered in the food, tobacco, raw materials, and services industries (Liu 2011: Table 4). This pattern is consistent with the better case studies highlighted above, products like coal, logging, canning tomatoes, and auctions for collectables.

Third, buyers’ cartels are much more likely to be penalized bid-rigging schemes than other types of cartels. Of the 83 overcharge estimates for buyers’ cartels, 59 (71 percent) are bidding rings and 69 (83 percent) were “guilty” (caught and penalized by one or more antitrust authority). Buyers’ cartels also tend to be slightly more oriented toward domestic-membership composition (46 percent versus 32 percent for all the estimates).

Fourth, the median average but-for²⁷ price effect of buyers’ cartels is 19.6 percent, which is quite injurious but 17 percent weaker than those of sellers’ cartels (Figure 1). Median undercharges of buyers’ cartels are comparable to the overcharges of the penalized and bid-rigging cartel types. Like most other overcharges, undercharges have declined slightly over time; for the 45 episodes begun since the year 1990, the median average undercharge declined to 18.2 percent, compared to the median of 22.5 percent for episodes launched in earlier years (1919 to 1989). The reasons for this decline are unsettled in the cartel-studies literature. The trend may be due to superior data, improved estimation techniques, a change in the mix of cartelized industries (specifically more service industries), the global spread of and more effective detection by antitrust enforcement, or strategic behavior of cartelists.

 

V. NOTABLE ENFORCEMENT ACTIONS AGAINST BUYERS’ CARTELS: VIGNETTES²⁸

• In 328-326 BC, during wartime, several Athenian wholesale grain merchants formed a trade association to collude on bids made to importers of grain at Athens’s port (a buyers’ cartel). In 326 BC they were defendants at a public antitrust jury trial, where they argued that they had passed on the lower prices to citizens. However, they were convicted of illegal hoarding and sentenced to death. The prosecutor’s famous speech written by Lysias survived.
• A 1952 U.S. court decision concluded that three beet-sugar refiners had conspired to undercharge sugar-beet farmers in the 1939-1941 crop years by $0.25 /ton.
• Damages were obtained in a case of bid rigging of three of 13 English oral auctions of 340 quality-graded used 1988 Chevrolet Caprice police cars, sold by the City of New York, January 1990 to May 1991. A subsequent statistical study found that members of the alleged bidding ring reduced prices received by the City, which settled out of court.
• Concentration (the numbers of firms bidding) effected the spreads of U.S. tax-exempt bond underwriting auctions. An empirical study of 9420 bond issues during 1959-1967 found evidence of bid-rigging behavior by bond buyers (buyers’ cartel) against the seller, the U.S. Treasury, when the number of biders was few.
• Similar the case above, but for 2221 auctions for government offshore oil leases from the U.S. Department of the Interior 1954-1975. Statistical indications were found of bid-rigging behavior by smaller numbers of buyers.
• Allegations of tobacco leaf bid rigging by buyers (cigarette manufacturers) against sellers in US auctions 1996-2001 was upheld in an antitrust class action by 400,000 tobacco growers and quota holders, which was was settled by the four defendants in May 2003.
• A published 1989 study of the Swedish roundwood (timber) procurement market in 1954-1984 found evidence of oligopsonistic pricing behavior by paper buyers against forest owners in two interrelated sub markets: sawtimber and pulpwood. An EC antitrust probe was launched in 2004 but apparently passed on to the Swedish and Finnish competition authorities. No action was taken in Sweden, but three paper companies were fined by Finland in December 2006.
• A knockout-auction bidding ring of 11 dealers from the US, UK, and France colluded against other buyers in stamp auctions from about 1980 to July 1997, mostly in New York City auction houses, but also in the UK. Criminal U.S. federal convictions were imposed on some of the U.S. and UK ringleaders.
• In the United States, two California-based antitrust class actions successfully obtained substantial monetary relief for highly skilled workers injured by secret no-poach agreements. The first is the Silicon Valley High-Tech Employees case in which eight companies signed DOJ consent decrees and paid civil damages, and the other is Film Studios Software Animation Engineers in which the Disney company and three others were found guilty. The U.S. DOJ played a minor roles in investigating these buyers’ cartels.²⁹
• However, recently the head of the Antitrust Division announced in a major policy speech that the Division was launching new investigations into oligopsony in U.S. labor markets, so-called no-poach agreements (Delrahim 2019). This announcement followed the third no-poach case to be concluded by the DOJ; in 2018 two manufacturers were convicted civilly for suppressing the wages of skilled manufacturing workers of Railroad Braking Equipment. Their consent decrees were followed by a quick 2020 class-action settlement.³⁰

As of mid-2021, there is little evidence that the DOJ has made no-poach agreements a top priority. The fact that indictments in no-poach cases have all so far been treated as civil infractions is further evidence that the Division is not serious about this priority. And the class-action settlements have wressted compensation that is at most 20 percent of the lost wages estimated by plaintiffs’ experts.

---

1 Professor Emeritus, Purdue University.

2 A second contender is Marshall & Marx (2012), a very formal monograph of the theory of bidding rings with practically no empirical content. Marshall, Robert C. and Leslie M. Marx. The Economics of Collusion: Cartels and Bidding Rings. Cambridge: The MIT Press (2012).

3 Blair & Harrison (2010) assert that “ [buyers’ cartels are] …far more prevalent than many have recognized” (pp. 1-14). Blair, Roger D. & Jeffrey L. Harrison. Monopsony in Law and Economics. Cambridge: Cambridge University Press (2010).

4 They include some cases in which plaintiffs were denied standing or lost their cases.

5 Oddly, this term does not appear in Blair and Harrison’s book. They stick to the more rigidly formal economic jargon of a “Buyer Power Index.”

6 However, the oldest written record of an illegal buyers’ price-fixing cartel is Wholesale Grain Trading in ancient Athens, 328-324 BC (Connor 2007c: 31-34).

7 Reliable time-series price data (from the records paid for coal by London hospitals) begin around 1700. A 1699-1700 coal bid-rigging episode in London is the second-oldest overcharge estimate in Connor (2021).

8 Coal was mined in many parts of Britain, but high land transportation costs conferred a monopoly on the Vend over a wide range of delivered London prices. UK coal-cartel studies with overcharge estimates include Ashton and Sykes (1964), Levy (1927), Sweezy (1938), Hausman (1980), and Tan (2003). Newcastle coal was also shipped to France and other European ports. Sweezy, Paul M. Monopoly and Competition in the English Coal Trade 1550-1850. Cambridge: Harvard University Press (1938).

9 They were also assisted in setting prices by their trade association or guild, The Company of Lightermen. Lightermen were recognized by English laws dating from 1558 (Oxford English Dictionary online). The most egregious episodes of “engrossing,” as the practice was called at the time in English Law, occurred when the R. Thames froze.

10 There are insufficient historical records to perform overcharge estimates.

11 Mine owners who sent coal by coastal ships from Newcastle to London controlled this cartel. The number of mines was quite large at times. Output was reduced through closing smaller mines and paying the owners compensation (“side-payments”). The lightermen’s buyers’ cartel may have overlapped with the Newcastle mine-owners’ sellers’ cartel during 1700-1750.

12 Hausman, William J. A Model of the London Coal Trade in the Eighteenth Century. Quarterly Journal of Economics 94 (1980): 1-14.

13 Tan, Elaine S. Economic Regulation and Rent Protection: A Study of the Coal Cartel in the First Industrial Revolution: WP03-23. Ann Arbor: Michigan Business School (October 25, 2003).

14 When railroads from the Midlands of England reached London in the early 1840s, the Newcastle owners’ transportation-cost advantage disappeared and so did the Vend.

15 The first proposed remedy in 1744 set up price controls for London coal, to be administered by three judges. In 1788, a law made any agreements among or partnerships of more than five coal buyers illegal “combinations in restraint of trade.” Whether these laws had more than fitful, short-term effects on this buyers’ cartel is doubtful; shame rarely is as effective as monetary penalties in changing illegal, profitable conduct.

16 The Lodge was expanded and renamed Ruxley Towers in the 1870s. The knockout auctions were held in rented space in the nearby town of Claygate.

17 Just, Richard E. and Wen S. Chern. Tomatoes, Technology, and Oligopsony. Bell Journal of Economics 11 (1980): 584-602.

18 Banerji, A. and J. V. Meenakshi. Buyer Collusion and Efficiency of Government Intervention in Wheat Markets in Northern India: An Asymmetric Structural Auctions Analysis. American Journal of Agricultural Economics 86 (2004): 236-253.

19 Asker, John. A Study of the Internal Organization of a Bidding Cartel: Working Paper. New York: Stern School of Business, New York University (July 2008). [See also The American Economic Review (2010): 724-762] http://pages.stern.nyu.edu/~jasker/stamps070628.pdf.

20 Kwoka, John E. The Price Effect of Bidding Conspiracies: Evidence from Real Estate “Knockouts.” Antitrust Bulletin 42 (1997): 503-516.

21 Delrahim, Makan. Remarks at the Public Workshop on Competition in labor Markets: DOJ Press Release. Washington, DC. (September 23, 2019). https://www.justice.gov/opa/speech/assistant-attorney-general-makan-delrahim-delivers-remarks-public-workshop-competition.

22 On the range of various court-approved methods, see Connor, John M. Forensic Economics Applied to Price-Fixing Overcharges. Journal of Competition Law & Economics Vol. 4, No. 1 (October 2007): 31-59.

23 An analysis of a slightly smaller sample of buyers’ cartels was begun by Liu (2011). She drew upon the 2011 edition of the Price-Fixing Overcharges collection (for details see Connor 2014). I rely on the expanded 2021 Edition of Price-Fixing Overcharges, which contains nearly 2500 quantitative overcharge estimates covering the past three centuries drawn from publications about 706 cartels with 1934 separate episodes. Liu, Jing. Buyers‘ Cartels: An Empirical Study of Prevalence and Economic Characteristics: Honor’s Thesis. West Lafayette: Department of Agricultural Economics, Purdue University (February 25, 2011).

24 Some of the episodes were studied by more than one author and occasionally a publication contained more than one equally plausible estimate by a single author.

25 As mentioned previously, Blair and Harrison (2010) have a sample of 24 buyers’ cartels, but nowhere do the authors attempt to determine the proportion of all cartels the 24 represents.

26 Cutting-edge empirical studies of many buyers’ cartels were made possible by the data available through “sunshine laws” passed by governments in the late 1970s. At the same time, antitrust class actions became more frequent, which made compensatory damages actions by small dispersed victims (farmers, consumers, local governments, and the like) feasible. Finally, testifying experts found ways to replicate their testimonial analyses of sellers’ cartels for buyers’ cartels.

27 That is, the percentages cited herein are relative to affected commerce absent collusion rather than the oft-cited proportion of total, overcharge-bloated affected commerce.

28 The sources of information on these vignettes can be found in Appendix Tables 1 and 2 in Connor, John M. Price-Fixing Overcharges: Revised 3rd Edition: SSRN Working Paper (February 24, 2014a). [http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2400780].

29 In re: High-Tech Employee Antitrust Litigation (U.S. District Court, Northern District of California 11-cv-2509) and ROBERT A. NITSCH, et al. v. DREAMWORKS ANIMATION SKG INC., et al. (U.S. District Court, Northern District of California 14-cv-04062-LHK).

30 IN RE: RAILWAY INDUSTRY EMPLOYEE NO-POACH ANTITRUST LITIGATION

(U.S. District Court, Western District of Pennsylvania Master Docket Misc. No. 18-798, MDL No. 2850).