-~[)1~··Rb l;.......;;..C_1"o.. ....... ~ -~ V. ~ ~ COMPETmON TRIBUNAl L -, I I TRIBUNAL DE LA CONCURREMC! ~ · W~ f ~ THE COMPETI'f!ON TRleUNAL I Q L M~r 29 1990 ~ ~ I IN THE MATTtR O~ an Application by the T Director of rnv~stigatio1 and Research OTTAWA, ONT. for an order pursuant to section 75 of the GQ~t:.tit.i~Ac..t. R.S. 1985, c. C-34, as amended r~quirin9 that the Respondent accept the Exdos Corporaticn as a custome~ for the supply of a product BETWEEN THE DIR8CTOR OF lNVESTIGAT:ON AND RESEARCH Applicdnt: -and-~~ROX CANADA INC. Respondent
AFFIDAX'II OF D8VID W. GIL.LEN I, DAVID w. GILLEN, of the City of Kitchener ' ' in the Judicial District of Waterl~o, r:-:ake oath and say a.:>
follows;
t.his matter on April 23, 1990. A~.tached ~H:>rF-to ae Exb.tbit "A' is a report which prov ,id@s a r,:.p l:; :: :.: F rofessor Le.:.na..rd Wav~rman's reh''rtal to my report.
' '
MAY zg '90 13:33 CONSL~EP & COPPOPATE LEGAL SERV. F'. 4 l 2. I mak• this affidavit pursuant to Fule 42 of the Competition Tribunal Rules.
SWORN befor~ me at t.he City of Waterloo, in the Judicial District
'~f .W aterloo this ).~o,.day of ~-. ~- ~ 1 , 1990 ... -.. -.--.: -:: -:: ---. -. -~ ~--'-.... --.. t A Commiss
.&
.. ' • !Si •• ' f ~ ··~;·"4 i. ,, '·. J.;. '
F'. 5 1
In the Matter of the Director of Investigation and Research and Xerox Canada Inc. Reply to the Reburtal Evidence of Professor Leonard \Vaverman. . , Dated May 14, 1990 David W. Oilleq, ,.Ph. D. Professor of Economics School of Business & Economics Wilfrid Laurier University Waterloo, Canada for
The Director of Investigation & Research May 28, 1990 f.
.. . .
F'. f, l
2 1. Professor Wavennan 1 s rebuttal of my repoct to the Dirc:cto.r can be di ... i&d into Llu-~c main areas. He fint speaks 10 specific issues ~ing upon my definition of the relevant market Second, he sunes 1 have improperly focused on the pans market and have ignored the truly rd.evatu market which is the final reprographic consumer. Fino.Hy, he discusses eenain issues surrounding venical integration and the potential gains to XCl. In this teply to Professor Wavemla.n 1 will not deal with tl'1e issues of venical integration an<.! efficiency. These are dealt with in thi! repon submitted by Professor Wilson on behalf of the Dircctor.1
2. I organi1.C my reply into [WO pans. First, 1 will deal with hi~ sp~ciflc comments on my . ' mark.et analysis. Second, I discuss Professor Wavenrw.n's comment~ on the relevant
markot and why his approach is not .appropriate in this case.
I. Product and Market Definition. ··. •', 3. Professor Waverman argues that my market dt:finition is incorrect and logi1,;ally lead~ to a series of O\'erly narrow market /product definitions. He would have one ~lieve there are 9& separate markets under my ~'relevant markecu definition. This erroneous conclusion stems from a number of factors that Professor Wave,rman has mi!)underscood or has <:hoscn to overlook.
4. In my report I make exclusive reference to the "product refused", not paper, not copiers but to the- product refused, post 1983 copier pa..tts. This does not poim co an e.\du!\i vc: definition of the market. Rather I am limiting the scope of the parts under consideraLion to those which were refused and which coi~stirute che b:is.e of th~ dtrltlition for th~ m'1Iket.
l Affidavit of Profossor T. A. Wilsen, Report ~ucp;.ir.;4 for 1.h~ D.re.:.wr of 11'1"¢;)1.igt.iuon and R~ch.
MAY 29 '90 13:35 C~~Sl~ER & CORPORATE LEGAL SERV. The markcl which I have identified could easily have ~en tem"led the "Xerox copier pans market" if not so limited.
5. Professor Waverman st.ares cha£ I have argued, " a possible market i~ parrs for copiers" (Waverman. pa.ge 2, pmgraph 4) which is quite com~ct. He also staces that I have argued, "within that market there are numerou~ rel~van1 markets", including" new pwts as opposed roused parts"; "parts for independent StNice organizations", 11 p~uts for end users''. etc., (Waverman, page 21 paragraph 4). This 1s incorrect. 6. My repon id .en.t ified two relevant markets. The market for Xerox copier pans and a su~market which is fonned by the (large) subset of parts whic.:h can only be viably sourced ... from 'XCJ. Although these marker.s rnay be viewed as narrow, it is the particular circu.msrances of lh.is case that lead to this conclusion. They include the fac[ that pans a.re not interchangeable between Xerox. copiers and oth~r copjers, mat pans are not available from orb.er sources in adequate supply but only from XCI in its intermediate market, thut parts are nor sold exclusively in a bundled form with other producti; hut arc sold separately, and that they have a separate prit:c and hai.·e been actively purchased separately. It is for these reasons, among others, that it is quite rea.')onable to define a marke' for Xerox copier parts as I indicate in my report.
7. ProfessorWaverman provides a kngl.h.ily quote: from A.reeda and Hovenkamp (p. 8) in an attempt to illustrate hi~ point that prooucts d.ifferendated by bnmd name alone cannot reasonably be placed into separate marke.ts. As a matter of general economic principle J do not disagree with Professor Waverman that there can be i.ruerbrand con1~tition due to 4iubstitution between brands. What Professor \Vaverman has chosen to i"Jjriore is the set of circumsbncts in this ca~e which d(X'.s r~ot :i.tbw the rnif'.: of interbr~md compc::tition
F'. 7 1 3
F'. ::=: 1
4 discussed by Aruda and Hovenkamp because the copier pans are not technii;ally substitutable for one another. This is more than mere differentiation. 8. IE is possible of course to hai;e inrem1ed.iate product markets where my analysis would not find a firm-specific ma.rice[ definable. For example. lumber from brMnded lumber companies (such as McMillan~Blodell) is an intennediatc product but one cannot definl! a market for a particular lumber company because of substirution possibltities. This is quite different than the case at hand. Certain ~opier companies parts, as I underst"1!ld it, may be interchangeable with those of coropetirors but this is not the c~e with Xerox copier pans. It would depend on the panicular fact£ whether other ,;oph:r cornpanit:s' part.s could be defined as a rel~vant marker. Ir may or may not be the case for othtr companies as Profes.sor Waverman suggests.
9. Professor Wavemtan states that I "classify end-users "nd ISOs as distinct markets" (p. 3, paragraph 5). A&ain he is incorrect The iss.1,1e as stated in paragraph l 9 of my repon is
"In order to define the "relevant product " and c.herc(ore the relevant marker, four issues rt'UlSt be investigated: ( l) whether a market can be defined separately for paus used by ISOs as opposed to end u~rs .a,uui.n of a ~mic' gas.;~1e; (underlining added) . .,•. The end user refetTed to was that customer who purchased n0t parts but a service package which is a bundle of pans, labour, reliability and oth~r fai.:tors. I accepted in my report as a given that an end·uSl:!r . w ho purchasers pan:. directly from XCI is in a. )irnil;;u- p0!'>;tion in th!;' . marker a~ an ISO (s¢e, e.g. para. 24). The purpose of thi: ana.lyi;is was to :;how thal 1h~
"service pac~ge" end user ([he dou'llnanc type of end-user in the CC1pier market) h nm in lhe same position as an ISO and tl1&.t an intermediate rrurket for pans C\iJl ix: validly defined. I conclude that they are not and that a separate para markel cxi.sts. As I stated, end-users. are "not in the same busine~s·• and {ignored by Professor W.averman) rhc demand and supply ~:de e"'hibit distinguishing cha.ri:ic:c..ri;;tics. The et1d·\iSers are not in the sAme busine~!1.. and are at different poinLS ir. Uii:: v.Uue <.>r prod>.1ct i,;hain. They choose not to
p. 'C< l
5 identify pans separately as part of the produ~r pa.1.:kage whi(;h i) bundled to~c:ther as rent (copier, parts. labour, etc.) or a service <;ontract (pans, labour, etc.).
10. l-Iaving misconceived I.he issue, Professor WavcrmM sels up (p.4) what h~ claims is a . a':~. clear example of how my reasoning would lead to hamb 1Jtger me:af sold to Mc Donalds and l.oblaws as being in separate markets. 1 would a~ci:pl thi:it t.he:se pun.·hases are in the same market. In fac" rus analogy provides an excclltr:t illustration of preci!>ely the disrim:tion I make between intermediate markets and final mark.:ts. A consumer has the choice of gQing to McDonalds and purchasing a bundled product in th~ form of location, speedy service, uniform service quality, hamburger and con.dim.ems. 111is consum~ ii in the fin1'-l consumer market. The same individual could also go to Loblaws and purchase hamburger, take it home and prepare it himself. l!l this latter case he demands an input whereas in th~ fonnet case he demands an· output. In just the same way the end·user can purchase a bundled service contract from Xerox in which <;asc: he 'is hoc pan of the pans market. Alternatively, he can purchase ch~ parts ano repwr the copier him.~tf. In th.it> case he is pan of the Xerox copier parts market This is made quile clc::ar in parasraph 24 of my report,
11 ••• copier pans are an input into the supply of copier servicing by exdo~ and other ISOs. ISOs nod soroe CUS1Qll1'I'S 'kb~ their own servicir~ ~w1ire on11 the. 12aos and not the ~erykt terhnk.ian from I. Parts, tht-n:foro, repr~sent a separate market from ~-rvice.'' (i::rophasi$ apded} . ' Professor Wavennan is clearly on the wrong turf.
Section Il. Professor Waverman':; Characterization of the R~levant !\tarket 11. Professor Wavennan ~rates. "The relevant marke( should Lllclude that package of i;crvice commonly p;.1rcha:;ed by trn;h:sers; ,iiJlQ;. X..C! ;.:QU1I1:1i:~ n:ilh...Q.il.l.cr..£iJ~~ UbQI~~l'.i~;erviGes. r...'lt rcky21v ilJMl .. ,1 jj DJ;.{i X~.n;s ~fiiOd ,p~ms, Qllt the •pw:.lq:t.in
F'. 1c 1 1
6 'tYbkb XCI cornpet~f (p. S, paragraph 6). This is Professor Waverma.n's rhesis and it suggcscs mat relevant competition caru101 exisl in an intem1edi"te market of a vertically integrated industry He is sugges.ting you cannot have separate suppliers, d~manders and pri¢es established in such an intermediate marker. Based on this logic thete is not only no relevant pans market bm no market for toner and presumably no markcl for 3.llY such intennediate input Do we ignore rhe mlil'ket despite the evidence of separate prices and !>-peciaH2ed suppliers? Because General Motors and Ford compete 'in the fin:il market for .automobiles does this mean there i:; no dtfmable market for captive pans for Ford cars? Professor Waverrnan would have one believe that the interaction between demanders and suppliers or irt~ate goods is of no consequence, the prices established are irrelevant. Does this mean that ror purposes of rhe Competition Act one can ignore the concentration of upstream asset$ in the hands of supplier such as an oil e¢mpany who faces retail competition downstream? Surely not. In all of the examples cited the intermt.diaie market r. ..,., is a real market, there is interaction between suppliers and demandtrs, r:ransactions occur
and prices are established.
12. Despite the position taken ir1 his reply as quoted in paragraph 11 above, Professor ·. Waverman, in his original report; statCld ''This application by the Director involves one portion of a c:hain of delivering product from a vondor to the final consumer." ( P. 2. para 4) Ye: he then proceeds to develop a Jefinition.of.t·he rdtva:11 rnar~:e! which completely ignoreit this real Md important fact The value chain from <:opier manufacturer to <,;onsumer ot" copier services is more complex than Professor Wavcrrnan acknowledges.
13. If it were the c"se tha.t Xerox controllw the- l!nti.rt: dlstributi.on process from manufacture to final con:!:umer and there was therefore only Me transaction, Professor Wavem1m's definition of che relevant mAfkc:t might be correct. There would be no intermediMe markets becaus< there are no dernarider;-; and ~uppliers, thtn:: ll'C no prices
' ·~. t "
F'. 11 l
7 established for parts except for the "administered" prices for intra·firm tranlktctions. The;;e are not market prices.
14. Howevert the face of the man.er ili XCI enoouraged 1he dc:.velopmcm of an ' ...' inrem1ediary. Exdos. Without this diro:l encouragement other ISOs also entered that m&keJ and obtained pans from XCI. ISOs found thac there was an economic opportunity available by servicing Xerox brand copiers by offering S¢rvice contra<.:ts and pw-is. There is a transaction whi<:h takes place between Xerox and a number of cuscomers for pans alone. There are demanders, suppliers and transaction' taking place. TI1ere are prices ~stablished. in a market. There is a. market for pans which lies between the oopfor ' ' manufacturer and the final ompuc of copier services. I therefore reject Profe~sor
Wavennan's notion that one can ignore the Xerox copier pan~ marker~. ll ignores the reality of active intennedlate markets. ~.I .-. 15. Professor Wavennan states, "Xerox as a provider of photoCOpying services faces competition from many suppliers" (p. 7; paragraph 9) and "Professor Gillen does not hold chat XCI has any market powerin the market for photocopying scNices. XCI men has little market power in any market for parts" (p. 10, paragraph 11) and "In economks language the e.lasticity of derived demand (for pansl depends on the elasticity of demand in the end-us.er marktt...",(p. 10, paragraph 11). fie argues here !h:'it XCI h<1.s litrli: mJik~t power in the market for phorocopying services and therefore cannot have ma:l'ket power in any upstream markets.
16. I first would only note that txml.ination of the data set out in Professor Wavemum'!i original report sugi;ests that it is by no means cleat that Xerox: ha~ no market power in the copier markt:t. Jn certain !".egmc:nis of t.ht! .:-n~kc:t, particularly the high end, XCl does not face competition from many supplit:r'S us Prnfessyr Waverrna.n claim.;. Thert:feirc riot only
.,..
F'. 12 1
8 can one question Profe~sor Waverman's lack of rccogrution of valid intermddiate mark.els but even accepting his thesis one can also question rh3r end use <:ompetition affects 1.he relevance of intermediate markets.
17. Second, to say lhe 'elasticity or deri\led demand (for p:is..'"tS) ~~nds.JlQ the elasticity of demand in the end· user market' (emphasis added) as Professor Wavennan does, implies that the end-user trtarket is the determinant of pric~s in the in1ernledil:lte market. This suggests there is one transaction becw~n copier manufacturc:r and fmal consumer and this is incorrect. His assertions ignore the existence of a Xerox copier pans market, an intermediate matk¢t to ptoouce a service package thar involves competition betwoon xa and ISOs and chat parts are generally bundled with labour and other products in the end use market. There are catainly linkages between demand in the end-user market and the demand for pans but these linkages are in my view insufficient lO conclude thal the murkets ........ art indis.dnct. Some consumers of McDonalds hamburgers mny chose to buy the
hamburger meat for home consumption . lf the price of fast food burgers rises and there is a weak impact on home consumption ir does not necessarily place fast food and raw hamburger in the same market
18. I agree with ProfessorWavennan when he states, "A relevant market musl be consid~red within the purpose~ of the Competi;:itin A~t''. I do not however agree tl1'1t the market defined by Professor Wavenn3.n is the releva.rit one for the variC>llS ¢lcmems of Section 75 as applied to this Cllse. Tho objectives of the Act are to ". .. maintai.r1 and encourage C(lmpetition in Canada in order to promote the efficiency Md adapt:ibiliry of the Can~ economy ... in order to ensure that small and rrK:dium sized enterprises have an " equitable opporturnty to particip::He in !!1e C;m~dbn e.conomy end in order to provid¢ customers with competitive prices and product cho'i\:es" (Com,Petilio.n A;;t, 1.1 ).
F'. 1 :::
9 19. To satisfy the elements of :section 75 of the Competition Act dbes noc require that compelition has been effected. The section focus.eh on the analysis or me. effects on an individual business. one that may exist at any s.tag!! in the production process and include intermediate markets. The purpose of this section of rhe Act on this basis appears to be to promote consumer choice and efficiency through the promotion of cornpetitors in the marketplace. In the case at hand the refu~al to supply Exdos. will remove a competitor from the marketplace.
20. The impact on competition goes beyond the immediate effect of lack of choke for users of Xero~ copi~rs in their selection of a service represeritative. h also eliminates any competition in the buy-back market. This has two effects: it reduces the resale \lalue to cunent copier owners and thereby raises the net cost to them, and ic could penn1t Xerox ro remove machines from the market more cheaply. Wilhout an effe.ctive market for uSed machines. there will be less competition in the price sensitive mflrket segMents such as are currently served by Exdo$.
21. The objectives of consumer choice and equity are .-:erve.d by II. market definition whkh ascribes importance to th~ inren:nediak: rnarket An order to supply Exi.lo:s would permir the objectives of the Act to be realiw.i. Exdos would be,ruaintairied as a competitor in che downstream mlil'ket fqr service contracts; 3 tn$"k~t in whkh it l:ompec,s with XCI. Ir would also promote consumer choice in lhc downstream market as well ::is m.:.eting the small 3l1d medium size finn opportunity goal. For the ri:asons de"-:ribed in Professor Wilson's report competition in the intt:rme<liatc product market may also be important for efficiency reasoos.
22. 'The fo:egoi.ng considerations suggest that the ma.rkc:t d~finitiort I have proposed is indeed the more relevant one to the apparent objectives in thi~ section of the A~t.