tl-16-00 09:50 FROM-STIKEMAN ELLIOTT +14169470866 .. j T-466 P.OZ/09 F-4Z8 I EXPERT REPLY REPOllT OF PROFESSOR GEOllCE A. BAY 1. This supplemental repon responds co the expert repon prepared by Professor Baye on behalf of the Commissioner. In this supplemental repon I am primarily concemed with those points on which there appean to be fundamental disagreement where that disagreement affects our ultimaie conclusions. There is no need to dwell on the points of agreement or on minor d.itlerences th4t do not sign.iticaznly a.tfea the analysis. In this sgpplemmtal repon, I inco:rporare
the implications of recent developments conc~ng the Adams Mine project.
A. GTA Waste 2. There are two signifi.c;:ant differences between the analysis contained in my origin.al expert repon and that conlained in Professor Baye's originad expert repon. first, although the
Application defines the market to include both residential and ICI waste, Professor Baye limits the product ma.deer to ICI waste. In my original repon, ! addressed the product market alleged by
the Commissioner in his Application. However, as will become clear, the substantive problems With Professor Baye's analysis exist regardless of which product market he is addressing. Second, on the supply side, Professor Saye wants to limit the geographic market so to exclude land.fills in the US. J believe this to be rhe :principal error in Professor Baye's analysis and much of this supplemental report is aimed at addressing that error.
3. Professor Baye anemp1s to support his nmow geographic market definition with extensiv~ statistical analysis. However, with respect co his key empirical point, occupying three pages of text and several tables, there is no fundamental disagreement. Indeed, his findings are f'ully compatible with the analysis in my original report. Specifically, I refer 10 his conclusion, based on his regression analysis, that there are added costs associated with utilizing US landfills, such that rational U'ansfer sta.tions (in Southern Ontario) will pay a higher disposal fee to use Southern
Ontario sites. I have no reason to dispute these findings and, indeed, they are amicipatt:d and acknow ledg~ in Paragraph 24 of my origin1d repon. However, if Professor Bay e believes that a locational advantage for landfills located in Southern Ontario is sufficient to support a conclusion that me relevant geograph1c mN"ket c~ be limited to l1U1dfills located in Southern Ontuio. he is gu.ilty of a f\mdamcntal miswiderstanding of the principles of market definition.
4. AS I point out in Paragraphs 19-31 of my original report, the fact that some of the costs of using US landfills (essentially transport costs) arc higher than the corresponding coSts for using Southern Ontario landfills does not preclude US landfills from competing successfully for GTA wasre, since the relevant costs to the transfer sration are the !m!l costs of using the land.fill; i.e., the sum of the transport and the disposal costs. So long as US landfills are willing to c;:harge lower tipping fees (i.e., the disposal ponion of the t&d costs) to offset the higher transpon costs, they will be able to compete effectively against the local suppliers. 1 'It does not maner whether the higher transport costs W"e ~function solely of the additional disiance uaveled or whether there are additional costs (e.g., tolls) of crossing the border. lndeed, the analysis is unchanged even if pan of the disadvantage of shipping to the US is some unspecified "hassle" factor.
tl-16-00 09:51 FROM-STIKEMAN ELLIOTT 5. While regression analysis is not really necessary to establish Uie point, Professor Baye's regression analysis confirms tha[ this is precisely whar has happened. Specifically, US landfills ha\le reduced ripping fees to offset the higher trPnsport costs and. in doing so, have s~ccessfully attracted a substantizd share of the disposal bqsiness for OTA and other Southern Ontario waste. So, instead of supporting his hypothesis that the marker should be limited to South.em Ontario landfills, his analysis is fully consistent with the opposite conclusion - that US landfills can
compete effectively for GTA business and therefore cannot be excluded from the relevant
market. 3 6. Beginning in Paragraph 72 of his report, Professor Baye attempts to set out an altemati\'e scenario in which che geographic market is narrowed and the merger will give CWS marlcet power. (He assens, ··the landscape will change dramatically by the year 2002 ...) In this scenario, all of the proposed expansions for Southern Ontario landfills will have been permitted for the full volwnes apphed for and rhis will result in more than enough capai::ity to meet the deJDands of
the Province. As a rest.lit, in this scenario the competitive price fo:r Southern Ontario lmdfills would have fallen dramatically as the inclividual landfills scramble to use their respec:tive landfills to capacity, .. hut, as a result of the acquisition,. CWS can use its market power (based on a large share of the intra-Ontario capacity) to prevent price from falling. 5 7. Assuming that J have set out Professor Baye's alternative theory more-or-less accurately, our differences with respect to this alternative theory are more empirical than theoretical. That is, I can imagine. pw-ely as a theoreucal matter, a scenmo in which competition from US landfills would no longer be a relevant constraint on the price for di!!!posal of GTA waste and in which a firm controlling most or all of the capacicy in Soi.ithem Ontario would be able to prevent prices
2 In Table: 4 of his repon Professor Baye has indicated that only 56,877 tonnes of GTA IC! waste is disposed of at Arbor Hills. !fowever, the Agreed Statement of Facts states that 34 7 ,8 71 tonnes of GTA waste is disposed of at Arbor Hills. In omitting the GTA residential waste that has been going to Michigan, Professor Saye's tables present a misleading picture of the exteni to which 4isposal sites in Michigan have successfully compered for GTA waste against
the landfills in Southern Ontario.
l lfwe followed Professor Baye's argument about lQcatiomu advantage to its logical conclusion. it would make a mockery of geographic market definition. For any product in which there are non~trivial transponation costs for imponed product (whether imponcd from foreign counoies or simply from neighboring provinces), Professor Baye would apparenlly argu.e for a narrow geographic market even where the imported produce is by far the largest sou.rec of supply .
.. According to the argument, the excess capacity will cause prices to fall below the level at which US landfills are competitive; hence, according to the argument, the only competition that matters is the competition among the Ontario landfills. A corollary of the argument is that all GT A waste not contractually bound to go to the US will remain within Ontario.
s Note that even under this scen.Wo, prices do not increase above current levels; they cannot because competition from US landfills would prevent it.
+14169470866 T-466 P.03/09 F-4ZB 2
~1-16-00 09:51 FROM-STIKEMAN ELLIOTT from falling below CUITent levels. One necessQ.l')' ingredient in this scenario (as Professor Baye indicates) is substantial e3cess capacity in So"Qthern Ontario. However, I think the set of
circumstances wider which this might occur do not equate with any reasonable forecii.St of what is likely to happen in Southern Ontario after 2002.
8. To establish my claim, it is first necessary to emphasise the critical role excess capaciry within Sourhem Ontario plays in Professor Baye's alternative theozy. lf, conttary to bis speculation. there is not enough low-cost capacicy within Southern Oncario to satisfy all of the demand generated for those 14'Jldfills, price for that c:epacity, even in a competitive market, will ha"e to be
at a l~el that renders the ovenall cost of using US landfills comparable. The mechanism by which this happens is straightforward. Assume, hypothetically, that the net cost of using
Southern Ontario landfills is Jess than that of using US landfills. All collectors of waste and mmsfer scations will now want to use Southern Ontario landfills but there will not be enough capacity to satisfy everyone. Very quickly. even in a vigourously competitive market, each Southern Ontario landfill will realize that it can increase price without losing s~les so long as the price js not so high that the overall cost of using an Ontario landfill (transpon costs and disposal fees combined) exceeds the overall cost of using US landfills (Iran.span costs and disposal fees combined). While the Ontario landfills will receive a higher tipping fee than the US landfill, this is simply a locational rent and has nothing to do with market power. Put differently. it would not matter whelhcr there were a dozen independent operators of the local land.fills or they were all owned by one company. The overall c:ost of using the Ontario landfill will not remain below the overall cost of using the US landfills and cannot increase above it.
9. The key question, Uierefore, is whether iris hkely that, effective in 2003, there will be more than enough low-cost capaciiy wirbin Southern Ontario to satisfy overall demand. ln anempting to deal with that question. I have followed a methodology that is generally consistc::nt with that used by Professor Saye. In panicular, at least as PD initial step, l focus on the capac:ity that is potentially available in Southern Ontario to acconunodate GTA waste ('"GTA -permitted landfills"). Hence, I omit from the analysis those m1Ulicipal landfills that are limited to accepting local residential waste. Table 1 below sets ou1 the overall capacity situation in 2003, "Qndcr the assumption that all of the anticipated closures take plaice and that all the scheduled expansions are approved and go forward as planned. The result is a tons.I capacity of 2. 797,000 tonnes. Note that this estimate omits the capacity associated with the Adams Mine as it now appears that project will not go forward.
10. Table 2 shows the demand conditions relevant to OTA-pcnnined landfills for the year 1999. It identities three major waste streams; GTA waste managed by rhe City of Toronto and GTA transfer stations. non-OTA waste that currently goes to GTA -permitted landfills in Southern Ontario, and non-OTA Wii.Ste that currently goes to the US. Total demand exceeds 4.3 million tonnes. Assuming that this panem continues through 2003, it is clear that the relevant demand will exceed the capacity available by a vecy S!lbsta.ntial margin.
11. Of course: the substantial ponion of this capacity deficit is auributable to lhe collapse of the Adams Mine contract. At this writing it is 1.Ulclear how the City intends to deal with the waste
+14169470866 T-466 P.04/09 F-428
l 1-16-00 09: 52 FROM-STIKEMAN ELLIOTT tha.t wa.s to have been shipped to Adams Mine. (Currently, this waste goes to Keele Valley. which will be closed by 2003}. One possibility wo11ld be illl exp1111sio11 of the conn-act with Republic: which permits most or all of that waste to be shlpped to Michigan. If so, a situation in
which the majorit}' of the waste generated in the GTA is shipped to the US would seem to be
tla.tly inconsisrent with any claim that Michigan is nor pan of the relevlUlt geographic marker. Bur in addition. such a contract would not resolve the capacity deficit problem for purposes of Professor Saye's post-2002 scenario unless, contrary to expectations, all of the volume in the connacr is guaranteed by the City. If, as seems likely, rlie contract guaranJees only cenain minimwn volumes, as soon as the Ontario price dropped to where it w~ more attractive than the contractual price (in Professor Baye's scenario). a. racional buyer would. seek to redirect all bu1 the minimum guaranteed quantitY back to Ontario. Assuming that the contract pric~ is at or above current levels. the clear consequence would be excess demand at any pric:e below the current level. H~nce, the marlcet~clearing competitive? price will remain at a level which makes
the US bmdfills a viable competitive altemative. Since it ba.s previously been c::s1ablished that competition from the US landtills will not permit the price to exceed cWTent levels. the restilt is chat there is simply no scope for the exercise of market power. In other words. Professor Baye,s critical conclusion chat, but for the proposed acquisition, prices for the disposal of GTA will fall
significantly once we get past 2002, is simply not a plausible one, based on the anticipated capacity in 2003 and the demand patterns established in 1999.
12. Of course it is possible to argue that the assumptions and projections contained in the preceding paragraphs are subject to error and that demand and supply conditions might result in
excess c"pacity. However, while many lhings are possible, I think that a reasonable observer would have to acknowledge that the mosc reasonable alrerrunive scenarios would result in an even greater capa.cicy deficit.
13. First. rhe demand/capacity analysis ass1Jl'IleS that the proposed expansions will receive reglllatory approval and will be implemented as planned. Failure of only one of the significant
expansion projects -Warwick or Richmond - to go forward as planned would cause a substantial increase in the capacity deficit.
14. Second, the analysis made adjusunents for Green Lane and Essex-Windsor that may be too conservative. Table 1 excluded the capacity of Green Lane and Essex-Windsor and Table 2 excluded from the demand for the cost-effective GTA-perm.itted landfills those volumes ofnon-GTA waste chat have been sent to Green Lane and Essex-Windsor in the p~c. In effect, I assumed that the non-OTA waste that has historically gone to Green Lane and Essex-Windsor (but no more) would continue to go mere. But if Professor Ba.ye is correct tlun these landfills a.re not cost-competitive with the Michigan land.fiJJs, then when, in Professor Ba.ye~s scenario, price falls for disposal in Ontario, those who conuol lhe non-GTA waste that bas gone to Green Lane IU!d Essex-Windsor in the past may seek to redirect some or all of that waste to the cost-etfec1ive Ontario landfills. This could add substantial volume to the demand facing those landfills, making it even less likely that there is enough low-cost Ontario capacity to satisfy cbe entire demand.
15. Third, the analysis assumes that the amount of the non-GT A waste that currently goes to municip~l landfills will remain constant in the fu.ture. If the Ontario pnce were to drop
+141694TOB66 T-466 P.05/09 F-4ZB
ll-16-DD D9:5Z FROM-STI KEMAN ELL IO TT temporarily below current levels. it would become amactive to seek to divert more of this to the lower cost GTA -pemrined landfills. This would exacerbate the pressure on c~acity causing the: Ontario prices ro rerum to fonner levels.
16. Of course, there are some factors that might act in the opposite direction. Additional landfills might be approved and implemented, although Professor Baye has asserted (Paragriq>hs 94-100) char this is highly unlikely. Alternatively. rhe amounr of diversion might increase, l:hus reducing overall demand although this could easily be offset by general popul~ion growth. (ln any e\'enI, a cemporary decline in prices would have the effect of reducing diversion.) Overall, however, the contingencies that would increase the overall capacitY deficit seem more likely than those contingencies thac would reduce ir. The epd result is that. regardless of who owns the Ontario capacity, a price significandy below current levels is simply not sustainable. rendering Professor Baye's concerns moot.
B. Chatham-Kem 17. The portion of the Agreed Statement of Facts dealing with Chatham-Kent had not been completed at the time my original expcn report was prepared. I bad anticipated that it might be available in order for me co address che Chaiham-.Kent issue in this supplemental report, but I am told that the document has still not been finalised. However. since Professor Saye made reference to Chatham-Kent, I will make a few comments based on infonn<ion available, lest my silence be interpreted as reflecting complete agreement with Professor Baye on this issue.
18. It is my understanding that the following can be established: (a) Based on 1999 data, the total annual volume of waste generated in Chatham-Kem is approximately- tonnt!s annually. Ofthal Jotal ....t onnes is resjdenrial waste.
(b} - tonnes of the residential waste annually is disposed of at the Ridge under a Host Community Agreement at a fixed price of $36.06 per tonne, subject ro wljustments for overall
volwnc:s and inflation.
( c) Of the remaining ...t onnes of ICI waste, approximacely - toMes goes to the Ridge, - tolllles to Gore .Road, tlm' tonnes to Essex-Windsor, and '119' tonnes to CWS Samia (under favourable terms pursuant to an earlier divestiture order).
(d} The sowces of the~ tonnes ofICI waste are: (i) CWS collection operations - -tonnes disposed of at Gore; (ii) BFI collection operations - ..t ormes disposed of at Ridge; (iii) CER collection operations - ..ro nnes disposed. of aJ Lasalle; (iv} Erie Environm~tal collec:tion operarions-appro~imately .tonnes of which. was disposed of at Oore and .tonnes we:re disposed of at Essex-Windsor facilities;
+14169470866 T-466 P.06/09 F·4ZB
rl-16-00 09:53 FRDM-STIKEMAN ELLIOTT - 6 -( v) Jomac Disposal collection operations - -tonnes of which -tonnt:!s were disposed of at Ridge and ltonnes were disposed of at Gore;
(Yi) Wa.stewood Disposal collection operations and transfer starion-~onne:s disposed of at Gore;
(vii) appn>xim~tely 65 other ICI direct haJJJ customers. Approximately 20 of these customer disposed of a total of-tonnes at the Ridge. Approximately 4.5 of these customers disposc:d of ( ] Lonn.es at Gore_
(e) excluding the CWS volumes the amount of ICJ waste QriginatiDg in Chatham-Kent is thus approximately -tonnes;
(f) The Wastcwood Disposal facility is located a short distance nonh of Dresden. in Chatham-Kent. WWD operates a ttansfer station which it uses for waste collected by its roll off business as well as wast~ (primarily residential) that WWD uansports to the Gore lQndfill for Chatluun-Kent. WWD
has a permit for a MRF. Faced with a price increqse at the Gore Road landtill, it is my undi=rstiUlding chat che owner of WWD will testify that he would likely build and open the MRF to process waste anc:J would also likely transfer waste to Michigan and seek other IC! waste volumes for the transfer station. I also underscand that WWD has a permit far a landfill th.at is not operinional, and which could be made operational again. However, I have made no independent assessment of the feasibility or likelihood of such responses or the extent to which the)' wo~ld be
significant enough to render a price increase by CWS unprofitable.
(g) Another possible alternative for waste from Chatham-Kem is the Tilbury transfer station. All permits are in place but the faciliiy is not being used to transfer more than minimal volwnes of !Cl waste at the present time. Therefore, I have no information from which to determine whether this is a realiscic alternative for buyers in the event CWS anempted to increase prices after i:he acquisition.
19- Based. on the forgoing, it is possible that th~re are alternatives for Chatham-Kent customers if CWS were to attempt to increase price following the acquisition of the Ridge. However. even assuming no other realistic alternatives are available, the problem. created by the acquisition is small in terms of the amown of waste that is affected and will persist only so long as Gore would otherwise continue to operate. (Once Gore ceases to operate, whoever owns Ridge might still enjoy market power, but the degree of market power would be wiatfected by the acquisition.)
+14169470866 T-466 P.07/09 F-4ZB I
t l-16-00 09: 53 FROM-STIKEMAN ELLIOTT Tahir ] ; Capariry of Co51-Effectiyr GIA SN.HW-fermittcd J~y.dQlls in OntArio in 2003'
facility Annual Capaciiy - ToDDes Walker 617,000 Ridge 680,000 Warwick 750,000.l Richlnond/Napanee 750,000J Tola) 2,797,000 1 Green Lane and Essell.•Wmdsor itrc nor 1ru::liided :is potc:nr1al landfills for GTA Wit.Ste on rne basis rhaT 1hc;- c1;Ln nor compc:Jc wnh Michigan landfills accorc:hng to Ptqfessor Ba)'c's analysis. 2 .Regul~tory appraval tor an c:11.pans1on from it currem pcrmmcd annual capacny of 56,000 t0nncs, lUld for an i:-J1.p3ns1on of the service arell to include the CTA area, hllli been llllwght but not obt1tine4 For the purpose of analysis only, a possible e..:~s1on, 1n rhc full :MTtOunt applied for ha.; bc:cn assumed ~ Rc:gularory approval for 1J1 n~s1on from ll c~rrcnr pcrm1tte4 11nnulll ca~•ty of 125 ,000 t'mncs has bc:cn sough• but no1 QPt;unocd. for the purpo:;i;: of ;in:liysis only, a po$$1blc cii.pansion, in the full amai.ni applied for, h:is been usumed
+14169470866 T-466 P.08/09 F-4Z8
tl-16-00 09:54 FROM-STIKEMAN ELLIOTT Table 2: Demand in 1999 for lagdfills that !U vermined to receiye GIA sww=. adiu.strd tP reflect closure of Jrir9gnia in 2002/2003~ Source Waste managed by the Cicy of Toronto, CWS 3,020,319° transfer stations an<! independent transfer stations
Non-GTA waste received at Walkers, CWS -IS Samia. Riclunond. Ridge and Warwick 7 Non-GTA Ontario wasie received al 189,930" Michigan facilities
Non-OTA Ontario waste received ar Ref- ..,u Fi.iel
Non-GTA Ontuio waste received at Bf! .., , Niagara Falls landfiU
Britannia landfill Total 4 Int'omi&r1Qn on om.no w;,.s1e rti.J w.s d1spo:.c:d of iu th1; M0C1e01 Disposal l11ndtill in New Yori!; 1:> not av1&1lilbl~ at this time , Pew.and has nor berm 1ndudarl for voh1mc,; rhlll currently are rec-ei11ed 1u the tl11hon and Calcdon landfills illld the Peel
1nc1nerator on rhe iiss"'mpnon ihat the~ ~111 not close 11nd '¥ill nor lo~ 11ol11mi: to other f_.:1h11ca. To the ~ii;renr they lose \'Plume, ck~ "'01.1lc1 111cri;11SC. "Table 3 - pwagraph 75 of the Agr=d Swement of Facrs. 7 Non-GTA dem~d for Green Lane and Essu Windsor is not mdqdcd on i:hc assi.1mpr1on lhar 1t will contin~e to be senr 10 those hindfills and rhCTcforc wall not aaa to the demand on the c•pacuy 1dcn,1ficd m Table I. 1 Total ca 1ty lc:>s v om CTA. - A S e ni of Fw:cs, Contidenual AppendP. B, Table JU.
+14169470866 T-466 P.09/09 F-4ZB I Tonnes
197 619 1 z ' 4,353,466