Case Documents

1);l /~ I rj_,(,A/J~.L­ _1)-P-L J;JI;_A_, THE COMPETffiON TRIBUNAL

IN THE MATTER of an application by the Director of Investigation and Research for orders pursuant to section 92 of the Competition Act, R.S.C. 1985, c.C-34, as amended;

AND IN THE MATTER of the acquisition y C~M!'nli!ON TRIBUNAL Hillsdown Holdings (Canada) Limited of 56% ofr:;·r:;~fJ.'U o::: LA c;.:mrnRRENCE P the common shares of Canada Packers Inc. r R 0 l f Ae~ 23 rng1 .~ o c u I BETWEEN: REGl:.H:AR - RcGJSTRA!Rf T --OTT-A~~A, -~~\JT. -,...1#---i Applicant -and-

HILLSOOWN HOLDINGS (CANADA) LIMITED, MAPLE LEAF MILLS LIMITED, CANADA PACKERS INC. and ONTARIO RENDERING COMP ANY LIMITED

Respondents

REBUTfA L AFFIDAVIT OF DAVID D. SMITH (filed by the Applicant)

Iv J COMPETITION TRIBUNAL TRIBUNAL DE LA CONCURRENCE

File No. C..../_q I /d / No. du dossier ~. v_ )J.tl,,,t"~

et

Exhibit No 4- -5: No. de la ~e-------Filcd on ~·::Ji /if , q ~.3.S-Depcseo le ; ---Registrar A. ~~ V . : Greffier I

i7:/ :) i-CT-91/1

REBUITAL AFFIDAVIT OF DAVID D. SMITH I, David D. Smith, of the City of Alexandria, in the State of Virginia, in the United States of America, make oath and say as follows:

1. In this rebuttal affidavit I address a number of points raised by the respondents, experts in their affidavits submitted to the Tribunal on August 2, 1991, and during the examination for discovery of Joseph Kosalle on August 15, 1991.

Declining Industry 2. Several of the respondents, experts suggest that the rendering of non-captive red meat material in Ontario is a declining industry. (Trebilcock, 'Jll9-'Il20, and references therein.) These experts conclude that the merger of Rothsay and Orenco should be allowed so as to permit rationalization of production facilities. (Trebilcock, 'Il40 and Bisplinghoff, 'Il71.) 3. It is not clear that the beef industry is destined to decline. Dr. Van Duren concludes that the main factor responsible for the decline in red meat demand was the change in consumer tastes and preferences associated with health concerns. (Van Duren Affidavit, 'JI21-«]J23.) This switching away from beef, however, may not continue: Although scientific studies do not generally conclude that beef, or pork, are bad for one,s health relative to poultry, many consumers perceive this to be the case. Both the beef and pork industries continue to work towards changing this perception,

both through advertising and by further reducing the amount of fat in the product. (Van Duren Affidavit, 'Il21.)

New technologies are being developed in the livestock and meat industries to reduce the fat in the carcasses. The achievement of lower fat animals " ... will be driven by a combined consumer demand for wholesome meat with reduced fat and economic pressure to produce meat more efficiently." (Bisplinghoff Affidavit, 'Il33.)

4. Predictions about the future are, of course, uncertain. It is impossible to tell if a recent decline in a business will continue or reverse. The Tribunal, however, does not need to base its analysis and judgement regarding this merger on forecasts of future rendering activity. The competitive process, and its underlying market forces, should determine the number and configuration of competitors. For example, the collection cost savings and the administration cost savings identified by the respondents are economies of scale that could be achieved through forms of growth other than just this merger. (Response 'fl71, and Smith Affidavit, «Jl65, «Jl66, and «Jl68.) If these cost savings described by the respondents are critical to the survival of firms in this business, firms that succeed in expanding production will take advantage of these economies of scale and prosper. They will take sales away from firms that do not achieve these cost savings. This process would provide consumers with the benefits of competitive pricing to the extent that a shakeout occurs.

5. If the economies of scale described by the respondents are not crucial for survival, the size distribution of firms might not significantly change. In this case, competition in the markets could be preserved without the merger. 2

6. This approach is consistent with the policy of the Antitrust Division of the United States Department of Justice: ... where there are continuing returns to scale, and therefore the market can only support a small number of firms, it will usually be preferable to have the firms that will remain in the market determined by growth through competition rather than by merger. This assures consumers the benefit of competition through the shake out period. ("After the Herfindahls are Counted; Assessment of Entry and Efficiencies in Merger Enforcement by the Department of Justice," Remarks at the 29th Annual Antitrust Seminar, Practising Law Institute, 12/1/89, by Judy Whalley, Deputy Assistant Attorney General, Antitrust Division, The United States Department of Justice.)

7. Neither Rothsay nor Orenco claims that it will fail and that its assets will exit the market without this merger. Indeed, both firms are profitable. (Directors' Document 26, 9/10/90 Canada Packers Inc. Presentation to Bureau of Competition Policy, p. 33.) This merger is not an opportunity to achieve unique efficiencies that is about to be lost forever. Indeed, there is evidence that this merger was more accidental than the inevitable result of market forces, since it was part of a larger acquisition. "[T]he acquisition was of Canada Packers and ... [Hillsdown] happened to get a rendering plant in the process." (G. Leslie in Kosalle Transcript, p. 388.) Additionally, Maple Leaf Mills structured no committees "to consider the rendering aspect of the merger." (Answer to Refusal in Kosalle Transcript, pp. 859-860.)

8. Even if the industry were declining, and mergers were thought necessary to prevent assets from exiting the industry, it is not apparent why the two largest firms in the market should be allowed to merge. Since they

3

are both profitable now, they could independently be among the long-term survivors in the market. 9. Furthermore, capacity in the relevant markets may decline due to exogenous reasons. Rothsay's Toronto plant was recently expropriated. Darling's Toronto plant might not have its lease renewed. (Directors Document 37, letter to R. Hughes from J. DeMarco, 11/29/90) These closures would presumably increase the capacity utilization rates of the remaining plants. 10. The forecasts presented by the respondents' experts do not consistently show that the relevant markets will decline. For reasons described in paragraphs 16-20 of my first affidavit, there are two relevant product markets for this merger. The first has red meat by-products and deadstock on the demand side. On the supply side of this market are facilities that render red meat by-products and deadstock, and some that render poultry. The second market has grease on the demand side. Grease, red meat by-products, deadstock, and poultry rendering facilities are on the supply side of this market. The geographic market for each of these product markets is an area in Ontario roughly 200 miles from the Rothsay and Orenco plants. (Smith Affidavit, 'Il21-'Il31.)

11. Table A and Chart A show the Agriculture Canada forecasts of renderable beef, pork, and poultry material for Ontario in the years 1991-1996 that were presented by Dr. Groenewegen. (Groenewegen Affidavit, p. 23.) The supply of beef material in Ontario is expected to decline during this time period, but this is predicted to be roughly offset by the increase in pork material. This table and chart also show predicted increases in the supply 4

of poultry material through 1996. Although data are not presented for grease, qualitative evidence suggests that its supply is not declining. Dr. Bisplinghoff states that "Restaurant grease is the only raw material growth area for independent renderers." (Bisplinghoff Affidavit, 'Il37.) Comparisons t.o the United Kingdom and the United Stat.es 12. The respondents' experts have also compared the rendering industry in Ontario with those in the United Kingdom and the United States. In particular, they have alleged that rendering markets have become more concentrated in these two other countries because of market forces and antitrust authorities approving mergers. They also suggest that this foreign evidence supports their arguments for higher market concentration in Ontario and the approval of this merger. (Trebilcock Affidavit 'Il40, and Bisplinghoff Affidavit 'Il46, 'Il48, 'Il69.) 13. The Application does not claim that every increase 1n concentration in rendering, in any country, is anticompetitive. It does not even claim that every merger among renderers is anticompetitive. It addresses only this one specific merger. There are certainly other possible mergers that would not be anticompetitive.

14. Respondents have presented no analysis to indicate that relevant markets and entry conditions in the U.K. and the U.S. are comparable to those in Ontario.

15. The respondents also refer to a report by the U.K. Monopolies and Mergers Commission in 1985, Animal Waste. Professor Trebilcock notes that this report found that PDM, a company with almost 50 percent 5

"market share," "was not operating against the public interest and there was no evidence of monopsony pricing." (Trebilcock Affidavit, 'Ill 7, 'Il40.) The report does not provide enough information, however, to indicate if the market was correctly defined for this "market share" measurement. Additionally, the U.K. report was not prepared in response to a particular merger. When the report mentions the subject of possible future mergers, it says that ... It appears that any further acquisitions by PDM of renderers or collectors could be referred for investigation under the mergers provision of the Fair Trading Act because of PDM's

existing share of the market, and we suggest that if any such

acquisition were to be proposed, particularly acquisition of a renderer, consideration should be given to the question whether it merited investigation because of PDM's already dominant position in the market. C'Il9.62, p. 100.)

This report also indicates that the concentration of rendering in Great Britain is substantially lower than it is in Ontario. The four-firm concentration ratio of renderers in Great Britain is 65.4 percent. (Table 4.3, p. 27.) The four-firm concentration ratio in the relevant markets for this merger is at least 87 percent. (Smith Affidavit, Tables 1 and 2.) Sixty-five percent is the safe harbor four-firm concentration ratio threshold of the Canadian Merger Enforcement Guidelines. (§4.2.1)

16. In general, we cannot know why antitrust authorities in the U.S. permitted certain rendering mergers, or even if they investigated the mergers. Investigations by the U.S. Department of Justice and the Federal Trade Commission are confidential. Outsiders might be able to learn about the questions asked and whether or not there was a lawsuit. But generally little more is revealed. To the extent that this is relevant at all, the best way

6

to determine what the U.S. antitrust authorities' approach to rendering mergers would be is to apply the U.S. Department of Justice Merger Guidelines. My analysis is consistent with those Guidelines. Entry Conditions 17. Dr. Shannon provides evidence that it could take three years to complete the construction of a new rendering facility, including the regulatory approval process, when he explains the time Rothsay would need to expand Moorefield: Typically, where a formal hearing occurs, it now takes a year from the time an application is made until the EAB issues its final recommendation. Any appeals process, such as that

undertaken in 1985, can extend another six to nine months at a minimum. ... Only after a new C of A is issued would Rothsay be justified in ordering the wastewater treatment equipment for its proposed expansion. To obtain and install such equipment would likely take 12 to 18 months, given the usually long delivery times for such equipment, the relatively short construction season in Ontario, and the need to make certain installation procedures coincide with times when plant

production is low. (Shannon Affidavit, 'Ill3.)

Entry should occur in under two years if it is to discipline anticompetitive price increases. This is the standard established by the Canadian Merger Enforcement Guidelines (§4.6.1), and the Merger Guidelines of the U.S. Department of Justice (§3.3).

Geographic Market 18. Dr. Bisplinghoff provides additional information on the shipping distances of renderable material: Normally renderers can only economically pick up raw material within a seventy-five mile radius of the plant. There is a point of diminishing returns due to overtime hours,

7

spoilage of raw material, and insufficient time to maintain trucks. The above conditions have led to building receiving

stations, which can be constructed approximately 125-150 miles from the plants. ... This enables the renderer to service an

area approximately 200 to 250 miles from the plant, but it significantly increases the hauling costs as it adds reload and station costs to the route cost. (Bisplinghoff Affidavit, 'Il14.)

This is consistent with the information on shipping distances on which I had relied when defining relevant geographic markets in my affidavit. (Smith Affidavit, 'Il21.)

Market Concentration 19. In my original affidavit, I presented a market concentration table (Table 1) based upon data provided by the respondents in Schedule D of a letter from J. Kendry to S. Peters dated July 9, 1990. (Directors Document 12.) The Rothsay number for free market material has been revised. (Kosalle Transcript, pp. 873-874.) Accordingly, a revised version of Table 1 from my August 1, 1991 affidavit is attached. The revised number was not referenced in the body of my earlier affidavit.

Rothsay's Position in the Relevant Markets 20. Professor Trebilcock states: As Rothsay (Moorefield) increasingly specializes in poultry rendering, given the general growth in poultry production in Ontario that is projected, its ability to compete in the non-captive red meat material market, with or without the merger, will sharply diminish. Thus, the only question left to be resolved is, does it matter, for competitive purposes, who owns Orenco? [emphasis added.] (Trebilcock Affidavit, 'Il35.)

The evidence relied upon by Professor Trebilcock, however, is not relevant to the question of whether Rothsay planned to reduce its rendering of red meat 8

by-products if there were no merger. In paragraph 34, Professor Trebilcock refers to a table, "Projection of Free Market Share by Renderer Over Five Years," as evidence that Rothsay's Moorefield plant would process only poultry by 1995, even without the merger. This table was originally prepared as part of the September 10, 1990 Canada Packers Inc. Presentation to Bureau of Competition Policy. (Directors Document 26, p. 14.) On page 12 of this document it says: In summary we see the following scenario in Ontario in the next five years:

... 5. Canada Packers/Maple Leaf Mills operating two rendering facilities in the Province: (a) a plant processing poultry by-product material...(b) a plant processing red meat material located outside of Toronto ....

Thus, the table to which Professor Trebilcock refers represents the industry five years from now assuming that this merger occurs. It does not contain evidence suggesting that Rothsay will focus on poultry processing in the future absent the merger.

21. Mr. Kosalle has stated that Rothsay had wanted to continue processing red meat material apart from the acquisition. In fact, prior to the merger, Rothsay was actively taking steps to ensure that it would be able to continue to process red meat material. (Kosalle Transcript, pp. 831-834.)

22. Even if the Moorefield plant were devoted to processing poultry material, for purposes of merger analysis it remains in the relevant markets. This plant has been used to render red meat by-products as well as poultry material. To switch this plant from red meat to poultry, or back

9

again, it is necessary to "change a few settings [on the equipment] but that's it." (Kosalle Transcript, p. 707 .) This capacity might also be used to render grease in response to a significant and nontransitory price increase. The respondents say that this would be inefficient, but we do not know in what sense. (Kosalle Transcript, pp. 268, 685-686.) It may be profitable to use this capacity to render grease if prices are higher. Because of this high degree of supply-side substitutability, the Moorefield plant could be used to process red meat by-products, deadstock, and perhaps grease in response to an anticompetitive price increase. The ownership of the Orenco plant is, as a result, highly relevant. The Rothsay and Orenco plants are competitors in the relevant markets.

Trebilcock's Railroad Example 23. Professor Trebilcock also uses a "limiting example" of the railroad industry to explain why "there is clearly a danger in focusing excessively on the competitive state of the input market and abstracting from the competitive state of the output market." (Trebilcock, CfI41.) In this example, the only two railway companies in a particular geographic market are allowed to merge to be able to compete more effectively with other forms of transportation. In the input market, after the merger, specialized suppliers of inputs have fewer purchasers of their inputs available to them. Nevertheless, it would seem to make no sense to hold up the merger, which by hypothesis is socially desirable from an output (and consumer welfare) perspective, in order to

preserve a greater degree of competition in the input markets.

This example is correctly classified as "limiting'' or extreme. It appears to differ from the relevant markets in the present case in at least two 10

important ways. First, the railroads in the example were assumed unable to be "viable" competitors with other forms of transportation services before the merger. Orenco and Rothsay are now competing in their output markets well enough to be profitable. Second, in the railroad example, Professor Trebilcock hypothesizes that "If such a merger were to occur, this may enhance rather than reduce competition in the output markets," i.e., other transportation services. Professor Trebilcock has already stated, however, that the renderers' output markets already "are as perfectly competitive markets as one is likely to find." (Trebilcock Affidavit, 'Il18.)

Efficiencies 24. My comments on the appropriate way to define efficiencies for the pnce increase/efficiencies tradeoff in merger analysis still stand. (Smith Affidavit, 'Il58-'Il61.) Because of recently received information, however, I now elaborate on my earlier analysis of one particular efficiency alleged by the respondents.

25. This efficiency refers to savings in manufacturing costs. (Response 'II71.) The respondents say that as a result of the acquisition, they no longer need to purchase beef tallow from a U.S. renderer, Taylor By-Products, at a premium over what they would have paid for the product in Canada. (Kosalle Transcript, pp. 796-797, 803, 958-960, and document RC-83.) It is not clear, however, why these results could only be achieved through this merger. Since the merger, beef material from Better Beef in Guelph has been shifted from Moorefield to Dundas. (Kosalle Transcript, p. 805.) As a result, the Dundas plant has increased its rendering of beef material while reducing its rendering of pork and other material. The 11

Moorefield plant has increased its rendering of pork and other material while reducing its rendering of beef material. (Kosalle Transcript, p. 965.) Because of this reconfiguration, the Dundas plant now produces the beef tallow that had previously been bought from Taylor. (Kosalle Transcript, pp. 796-797.) But this array of production could have been achieved in other ways. If Orenco required additional beef material to produce the desired grade of tallow, it could have bid for the beef material from Better Beef or elsewhere even without the merger.

26. I may file a Supplemental Affidavit on August 30, 1991, as allowed in the Competition Tribunal's order dated July 26, 1991. There are still answers to undertakings outstanding from the August 15 examination for discovery of the respondents with respect to their efficiency claims.

12

I declare under penalty of perjury hat the foregoing is true and correct.

DISTRICT OF COLUMBIA Subscribed to and sworn before me this 21st day of August, 1991.

DAWN J. HIGGINS A Notary Public o{ n;str:'t d Crhmbi<l J,-!y Co111•'·;~ ',:•- .~:~.: ~;·:\ J·ic·e :~u, t·};,.::~ 13

Table A Inedible Renderable Material Ontario Predicted 1991-1996 (000 KG)

Year ~ fQrk Meat Materials 1991 144,685 159,400 1992 143,900 165,640 1993 143,176 158,720 1994 142,517 153,160 1995 141,930 156,280 1996 141,421 164,120 Source: Groenewegen Affidavit, p. 23.

Total Red Poultry Qffa1 Tu.ta! 304,085 67,151 371,236 309,540 69,629 379,169 301,896 71,996 373,892 295,677 74,577 370,254 298,210 77,087 375,297 305,541 79,884 385,425

Chart A Inedible Renderable Material Ontario Predicted 1991-1996

Revised Table 1 Rendering Concentration Estimated Raw Material Volumes (OOO's oflbs per week)

Red Meat By- By-Products, Red Meat Low Grade Products and Deadstock, Name ofComuanv ~oducts.t ~ad§t!l!:k l.21111.d:i~k £2l!ltn'. 11.nd fgyltr:x: ~ 11.lld Gmas11 Orenco 2,795 1,032 3,827 Rothsay & Fearmans 4,035 104 4,139 4,460 Banner 1,500 1,500 Darling - Toronto 1,700 160 1,860 440 J.M. Schneider 900 900 225 Maple Lodge Farms 50 Ray Bowering 25 25 Quebec Renderers (Couture)tt 550 295 845 30 Oxford Deadstockttt 1,000 1,000 Totals 11,480 2,616 14,096 5,205 t Total of categories "Bon ... & Fat" and "Packinghouse" as reported in source. tt Ontario sourced material only. lnclud'"' material delivered to Couture by Phil's Rendering. ttt Future capacity, not production. Oxford is not yet licensed to produce. Source: Peters Transcript, p. 38. • Source: J. Kendry letter to S. Peters, 719/90, Schedule D, and Koealle Transcript, pp. 873-874. Rothsay number for current material has been revised.

Red Meat Red Meat By-Products Low Grade Deadstock, Future Salvage Poultry, Total Free Market Free Market .B.lQQd Material Material• Material"' 3,827 675 4,502 566 5,068 3,547 3,961 8,599 295 8,894 860 9,754 4,665 2,903 1,500 1,500 1,500 1,500 1,671 2,300 115 2,415 205 2,620 2,180 2,437 1,125 1,125 95 1,220 145 159 50 50 50 25 25 25 25 25 875 150 1,025 10 1,035 1,005 1,120 1,000 1,000 1,000 19,301 1,235 20,536 1,736 22,272 13,067 12,276

Revised Table 1 (continued) Rendering Concentration Estimated Raw Material Volumes (OOO's oflbs per week)

Share of Share of Share of Red Meat By- By-Products, Red Meat Low Grade Products and Share of Deadstock, Naine ofComnanv Bv-Products.t ~1ul~~k D!l!!!l.§W ~ Wld. f2Ylt0'. ~ l!ll!l. Q[!ll!§!l Orenco 24% 39% 27% Rothsay & Fearmans 35% 4% 29% 86% Banner 13% 11% Darling - Toronto 15% 6% 13% 8% J.M. Schneider 8% 6% 4% Maple Lodge Farms 1% Ray Bowering 1% 0% Quebec Renderers (Couture)tt 5% 11% 6% 1% Oxford Deadstockttt 38% 7% Totals 100% 100% 100% 100% Pre-Merger HHI 2303 3199 2014 7434 Post-Merger HHI 4014 3512 3608 7434 Change in HHI due to Merger 1711 314 1594 Post-Merger CR4 95 99 87 99 t Total of categories "Bones & Fat• and "Packinghouae• as reported in source. tt Ontario sourced material only. Includes material delivered to Couture by Phil's Rendering. ttt Future capacity, not production. Oxford is not yet licensed to produce. Source: Peters Transcript, p. 38. • Source: J. Kendry letter to S. Peters, 7/9/90, Schedule D, and Koealle Transcript, pp. 873-874. Rothsay number for current material has been revised.