THE COMPETITION 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 by Hillsdown Holdings (Canada) Limited of 56% of Canada Packers Inc. B E T W E E N:
Applicant, - and -HILL.SDOWN HOLDINGS (CANADA) LIMITED, MAPLE LEAF MILLS LIMITED, CANADA PACKERS INC. and ONTARIO RENDERING COMPANY LIMITED R~spondents REPLY AFFIDAVIT OF PROFESSOR D.G. MCFETRIDGE I /I / • ,I J c. ' i I, Profes~or D.G. McFetridge, in the City of Ottawa, in the Province of Ontario, MAKE OATH AND SAY:
l. I have been retained by Maple Leaf Foods Inc. to provide an opinion on whether the efficiency gains offset and are greater than the effects of any lessening of competition resulting from the acquisition of 56% of the common shares of Canada Packers Inc. (now Maple Leaf Foods Inc.) and its subsidiary Ontario Rendering Company Ltd. by Hillsdown Holdings (Canada) Ltd. which, through its subsidiary Maple Leaf Mills (now Maple Leaf Foods Inc.), operated Rothsay Rendering. My background and qualifications are described in my earlier affidavit, which was filed by the Respondents on August 2, 1991.
No. CT-~1/1 I ; I ' ' /A ' c 0
- 2 -2. Attached hereto as Exhibit "A" to this my affidavit is a true copy of my reply to the rebuttal affidavits of Thomas w. Ross and David D. Smith, filed by the Applicant on
August 23, 1991.
Sworn before me at the ) City of Ottawa in the ) ProvilJ.9.! of ont~h1inr: ) this .tt..~day of~ 1991.) A Commissioner, etc. M1~ ~,.J
97000/7-8
- -' ' 9 .. •
This is Exhibit "A" to the Reply Affidavit of Donald G. McFetridge, Sworn before me on the ~ day
of September, 1991
A Commissioner, etc. MJ~~t-J
PROFESSOR D.G. MCFETRIDGE
97000/9
REPLY AFFIDAVIT OF DONALD MCFETRIDGE 1. In paragraphs 18-62 of his rebuttal affidavit Professor Ross argues that the deadweight loss resulting from the Acquisition could be much larger than I have estimated and also much larger than the efficiencies resulting from the Acquisition. Our respective deadweight loss estimates differ for two basic reasons. The first is that Professor Ross defines the market to be larger than I do. The second is that Professor Ross predicts that the merged entity will be able to reduce the price it pays for raw materials by a much larger amount than I have assumed will be the case.
2. With respect to mar~et definition, Professor Ross includes all captive volume plus poultry in the market in the examples underlying Tables 1, 2 and 3 in his affidavit. In the example underlying Table 4 in his affidavit Professor Ross defines the market to exclude captive volume plus poultry. Professor Ross regards the latter definition of the market as "extreme". It is, in fact, the market definition adopted by the Director of Investigation and Research in his Notice of Application (paragraph 23). According to Maple Leaf Foods, the relevant market excludes grease in addition to poultry and captive material. I used this market definition in the deadweight loss calculations reported in my af f id av it sworn August 2, 1991. Whether gr ease is in or out does not make much difference to deadweight loss calculations. 1
Poultry and captive raw material volumes are larger and assuming that a given hypothetical price reduction also applies to them results in a substantial increase in the estimated deadweight loss. 1
3. With respect to the likely decrease in the price paid for raw materials, Professor Ross predicts that, assuming an elasticity of raw material supply of .1, the price paid for raw materials could decrease by between 336% and 594%, depending on how he defines the market and whether he leaves Darling out of it or not. 2 I have assumed 20% and 30% decreases in the pr ice of raw materials. Professor Ross' prediction comes from the application of a dominant firm model to the market as he has defined it. My assumption 1 The deadweight loss formulas in Sections 3 (c) and 4 (c) of the Appendix to my affidavit sworn August 2, 1991 imply that the
deadweight loss increases linearly with the tonnage, Q, to which a given reduction in the price of raw material or increase in the price of rendering services applies (values of F and initial values
of P"' or H remaining constant). resulting from a given reduction in the price of raw material and with given unit values is 64% definition of the market than it is with the Director's definition.
2 Professor Ross makes use of the identity: H ...... H= an ~edicts ble 1) that, as a result of the merger, H will with V unchanged. This implies that Pa is hat is, instead of paying an average of r raw material, renderers now charge an average to ick it u . The implied per centage decline or 94\. Similarly, in line 2, Table 3 Professor • • ic hat, as a result of the merger, H will increase to implying that Pft falls by-or 446\. In line 2, Table diets that, as a result ~merger H, will increase to implying that Pa falls bylillllllllir 336%. Professor Ross cautions that his predictions are~ended to be precise.
2
Thus, the deadweight loss larger with Professor Ross'
regarding the possible decrease in the price of raw materials is based on a statement by Mr. Steve Peters of the Bureau of Competition Policy that market participants interviewed by the Bureau anticipated price decreases of 20% or more.
4. It is desirable, in principle, to predict the effect of a change in market structure on market prices using a model of the market concerned as Professor Ross has attempted to do. The problem is that, while they may seem complex to lion-economists, textbook dominant firm and cartel models are actually rather naive. They do not reflect the full range of considerations entering into the pr i c i n g de c is i on ( s ) o f do mi nan t f i :r ms or o f carte 1 s . F o :r th i s :reason, the consensus of market participants :regarding the likely effect of a merger on market prices is often an attractive alternative source of information.
5. The dominant firm model used by Professor Ross is quite stark in its specification. In the version of his model underlying Tables 1 and 2 of his rebuttal affidavit, Darling, which, according to Dr. Bisplinghoff, is the world's largest renderer, simply vanishes. In the version of his model underlying Tables 3 and 4 in his :rebuttal affidavit, Darling remains in the market but, like the balance of the so-called "fringe", never poses any threat to expand beyond its current rated capacity. Raw material suppliers are assumed to behave in an equally passive manner. The suppliers in Professor Ross' model do not excercise any bargaining power they may have. 3
Operating in ~he ~bsolute certainty that its actions will never evoke a competitive response of any kind, the dominant renderer in Professor Ross' model not only ceases to pay for raw materials but also imposes a hefty pick-up charge in addition.
6. Among the factors that might constrain dominant renderer in a more fully specified model are: - a perceived threat of Ontario "fringe" rendering capacity expansion; - a perceived threat that renderers outside Ontario will seek business in Ontario, perhaps with the encouragement of large raw material suppliers; - a perceived threat that large raw material suppliers will engage in some forward integration; - a perceived threat that several large raw material suppliers will guarantee their business to a new entrant or to an expanding member of the Ontario "fringe".
7. The ability of large raw material suppliers, operating in conjunction with Ontario "fringe" renderers, renderers based outside Ontario and other potential entrants to credibly threaten a dorni nant render er with a significant loss of business if it attempts to reduce the price of raw material too far has led Dr. Groenewagen to characterize the raw material supply schedule as being kinked (see Figure 2.5 of his affidavit sworn August 2, 1991). At prices above the kink, raw material supply is highly 4
inelastic. At prices below the kink, raw material supply is much more elastic. It will normally not be profitable to reduce the raw material price below the price at which the kink occurs. There is an effective price floor at this point. By similar reasoning, the demand schedule for rendering services is also kinked and there is a price ceiling at this point. The crucial question is whether this raw material price floor/rendering services price ceiling is such as to p:recl ude the kind of pr ice decreases/increases that Professor Ross forecasts.
8. According to Dr. Groenewagen's reply affidavit, raw material suppliers possess considerable bargaining power. Dr. Groenewegen concludes that the threat of forward integration is credible. He also concludes that the threat to turn to an out-of-province renderer could limit the price decrease that a dominant renderer is able to impose on beef material suppliers to under 10% (in the case of potential competition from a Buffalo-based renderer).
9. I calculate in my affidavit sworn August 2, 1991 (p.13) that price pa id by Rothsay' s Tor onto plant for raw material captive, red meat, including grease) averaged of pick-up charges in 1990. According to Dr. Groenewegen's reply affidavit, the transportation cost advantage of the Dundas and Moorefield plants over Buffalo ranges to an advantage of location of the supplier. 'The maximum advantage over Buffalo
(non-net from a disadvantage of depending on the 5
amounts to under - of the average net pr ice paid by the 'I'oronto plant for all raw material in 1990. Dr. Groenewegen also finds that the transportation cost advantage of the Dundas and Moorefield plants over Detroit ranges from to 11111111111' These per centages are reduced if grease is excluded from
the calculation of average raw material cost. Moreover, threats to add to Ontario rendering capacity may become credible before the full transportation cost advantage of a dominant Ontario renderer over out-of-province renderers is exploited. In the light of these considerations, the rates of raw material price decrease I assumed in my deadweight loss calculation do not appear unreasonable. Experiments with greater rates of decrease in raw material prices may, nevertheless, be worthwhile. For example, the pre~ent value of the net efficiency gain over the period 1992-1995 assuming a raw material price decrease of 40% and a raw material supply elasticity of . 1 is
3 All other assumptions are the same as those made in my affidavit sworn August 2, 1991. Net efficiency gains assuming 20% and 30% rates of reduction in raw material prices are reported on page 25 of that affidavit.
6