Frora-LEXECON 1-312-322-0218 T-223 P.03 F-794 0 Fite No. CT·98J2
THE COMPETITION TRIBUNAL IN THE MATTER OF THE COMPETITION ACT. R.S.C. 1985, c.C· 34, and tne Competition Tribunal RuleS SOR 94-290. as amended; AND IN THE MAlT!R OF an inquuy pursuant to sur>seciion 10(1)(a>) of 'he Competitt0n Actretating ta tne proposecs acqu1s1tion of ICG Propane Inc. Dy Superior Propane Inc.
AND IN THE MATTER OF an application by lhe Commissioner of · Com~tition under secta<>n 92 of the Competition Act.
• BETWEEN: THE COMMISSIONER OF COMPETITION Applicant 0 -and· SUPERIOR PROPANE INC. and ICG PROPANE INC.· Respondents AFFIDAVIT OF DENNIS W. CARLTON AND GUSTAVO E. BAMBERGER
I, DENNIS W. CARLTON. of me City of Glencoe, in me State of Illinois, economist, and I, GUSTAVO E. BAMBERGER. of me City of Chicago, in lhe State of Illinois, economist, MAKE OATH A.ND SAY: TRIBfi~ALMPEDTEITfON TRIBUNAL LA CONCU.RRENCE F I 0 s REGISTRAR • REGISTRAIRE ' ------ i
:~ep-14-99 02:511111 From-LEXECON 1-312-322-0218 T·Z23 P.04 F-794 -2-0 1. We have been askea by counsel to Supenor Propane Inc. ancs ICG Propane Inc. to evaluate me expert repons served tly tne Commissioner of Competition {the "Commissioner; on AtJgust 18, 1999{the ..A ugust18, 1999 Reports•).
2. Attaened hereto and marked as Exnit>it ..A . is a true copy of our report, Yklictl represents tne work we nave done and analyses made with respect to me Commissioner's August 18, 1999 Reports.
Sutiscribed and S;!m to before me, this~day o!,,~Z:.Lts .. 1999. 4:-,z,u.4/) .1 tary . IC 0 My Commission expires: 7- mil ' - ct3 OFFICIAL SEAL MARGARET J KUSH ffDTARY PUBLIC• •Tl•TI OF -..011 MY COMlidllO. . flCtllRES:OJll~ 0
) ) CW.C&rllon ) ) - ~ ~r---) Gustavo E. eamb8f9ef
L . y$ep-t4•H 02:51pm Frat1-LEXECON 0
This Is Exhibit .. A" to the Affidavit of Dennis W. Carlton and Gustavo E. Bamberger ) ) ) ) )
My CommiSSion expires: 0
0
1-312-322-0218 T-223 P.05 F-794
Sep-14-H OZ:51pm From-LEXECON 1·31Z·32Z-0218 T-223 P.06 f·T94 0 REBUTTAL REPORT OF DENNIS W. CARLTON AND GUSTAVO E~ BAMBERGER I. INTRODUCTION. 1. 1. Dennis w. Canton, previously submitted the Report of Dennis w. Canton ("'prior report") in tl'l1s proceeding on oehalf of Supenor Propane Inc. ("Supenon and ICG Pn>pane inc. C-ICG .. ). A copy of my CUrricullJm vitae, wnich aescribes my professional qualifications and prior expenence, is attached to my prior report I, Gustavo E. Bamt>erger, am a Prine.pat and Vice President of Lexecon. A copy of my curriculum vitae is anaehed as Exhibit A. I assisted in tne preparation of ttie prior report. 1 2. In my prior report, 1r eached tne following principal conekls1ons: • Economic evidence aoes not suppon tne daim tnat retail propane prices aepend on tne number of ·nationar suppliers in Canada.
0 • Despite ICG·s relativety nigh share of retail propane sates in Canaaa, there is no systemaoc evidence that ICG substantially consirains Supeno(s prieing. in contrast, tnere is evidence that independent retailers constrain Superior's propane pnces.
• Evidence from a pnor Superior acquisitiOn supports my finding that tne proposed merger will not subStantially lessen competition.
3. I have reviewecs six of the expen reports filed on behalf of the Commiss1oner. 2 on several of the important economic issues in this case, tne Commissioner's experts appear to agree with me and reach conclusions mat are inconsistent w1tl'\ me Commissioner's claims. On otner issues, the Commissioners experts reach conctus1ons that are different from mine. However, these conclusions are reaelled pnmarily on theoretical grounds or based on anecdotal informatiOn; the conclusions are not basea on systematic empirical evidence. In particular, the 1. In some places 1n thiS report, we refer to the Report of Dennis W. Carlton; 10 others, we refer to additional analyses we na"e conducted for mis rebi..mal report. For ease of expositiOn, "I'° refers to Dennis W. Carlton When the earlier report is referenced, ana ·1· refers to Dennis w. Canton and Gustavo e. Bamoerger when we refer to work conCSJJcted for lllis report. 2. In tnis report, I limit my analysis to reports filed on behalf of tne CommiSsioner on the same aay as mine. 1w ill address the reports filed later, 1nciuding Professor ward's. in a later 0
hp-14-H 02:511111 From-LEXECON 1-31Z·322·0218 T·ZZ3 P.OT F·T94 --.2. 0 Commissioner's e~pens· conctusions tl'lat the proposed merger wm teaa to nigner propane prices. and mat t>arrierS to enuy and expansion are hign. are contradicted by empirical evidence. Tne failure of me Commissioners expens U> provide systematic empirical evidence, or to review sud'\ evidence in tneir reports, is a serious flaw in their methoaotogy and undermines tneir conc1usions tnat the proposed merger will substantially iessen competition in tne retail propane industiy. 4. The remainder of my report is organiZed as follows. In Section II, l identify tne unportant areas where the commissioner's experts appear to agree ..vitn me ana diScuss tne $ignificance of mase apparent agreements. In Section Ill. I point out where tne Commissioner's e:>c.perts' conclusions that the proposed merger will lead to higher propane prices is based primarily or only on theoretical gr~unds ana is conuad1ded by empirical evidence. In Sed1on IV, I review wnere the Commissioner's experts' condusions tnat barriers to enw ana expansion are high a1so is basea largely on theory, and aiso is contrac:Jicted by the empirical evidence. 0 U. THE COMMISSIONER'S EXPERTS APPEAR TO AGREE WITH ME ON SEVERAL IMPORTANT ECONOMIC ISSUES.
5. On several important economic isSues in this case, tne Comm1ss1oner's expens appear to agree with my conclusions and contradict me CommiSsi0ners daims. In particular, tne Commissioner's expens agree tnat to evaluate the likely effect of the proposed merger on competition in the retail propane industry, 1t 1s important to determine wnether (1} independent retailers constrain Superior's retail propane prices: and (2) there are significant barriers to entry ana expansion by indepenaent retailers. Funnermcre, they appear to agree that independent retailers ~ constrain Superior's pnces, and ttlat me empirical eviaence shews mat enw by inctependent reta11ers ~taken placi:_ in many local areas.>
( ... continued) report. 3. The Commissioner's experts atso agree that tne empirical evidence suggests that the 0
S611-1'·H 02 :5ZPI Fr oriEXECOlf 1-312-322-0218 T-223 P.08/60 ~T94 ~ -3-0 6. The apparent agreements on these issues are imponant because if Supeno(s prices are constrained by me presence of independents (as tne empmcat evidence reviewed 1n my pnor report snows), and if entry and expansion by olher propane retailers can and does take place, Supenor will continue to face competition that will constrain its prices if its acquisition of 1CG goes forward. A. The Commissioner's Expens Appear to Agree that •ndependent Retailers Could Con$train the Prices of a Merged Firm; Furthermore, They Agree ihat Independent Retailers currently Constrain Prices.
7. Accoraing to tne Commissioner, "s1gntficant, overall, sustainea" competition in the retail propane bustness, when 1t takes place, is ·pnmanly" between Superior and tCG.• That is, the Commissioner Claims that Supenor typically does not face "significant" competition from 1nC1ependent retailers. The Commissioner's expens, however, appear to concede that the presence of independent retailers can determine tne effect of the proposed merger on prices. 0 For eJCample, Professors Glollerman and Scnwindt ("Glot>erman and Scnwindr) explain tnat: [ijn geographical markets 1n wl"lich [Supenor} will not account for all (~r vtrtually all} propane sales in tne post-merger period, the competitive effects of the merger could be conditioned by lhe benaviour of other sellers. In particular, tne sut>stantial presence of otner suppliers that are willing and at>le to compete aggressively with [Supencr] might constrain tne merged entity from increasing pnces, even thougn market Share concentration will Increase as a result of the merger. On the other nand .•• the remaining competitors might aecide to matcn any price increases tntroduced by [Superior] in order to maintain me same pnce relationship to [Superior} as exiSted in the pre-merger period (GloDerman ana Schwindt, at 32).
( ... continued) proposed merger snould be evaluated in terms of its effect on ·a11 propane" instead af end-use specific 1oca1 markets. In panicular, Professors Globerman and Scttw1nat con~ude mat tne ·overwhelming majority of recent entrants iaenttfied by [Supenor] serve mora tnan one customer segment wnich further su18orts the appropriateness of evaluating market snare cnanges on an ·au prcpan? basis" 1ooerman and Schwindt, at 51, empnasis added). 4. Notice of Application, "In the Matter of the Competition Act, R.S. 1985, c.C·34, as amenaed, and the Competition Tribunal Rules. SOR/94-290, as amendeo (the •RuleS"); And in the Maner of tne Competition Act relating to the proposed acquisition of ICG Propane Inc. by Superior Propane inc.; And in the Matter of an Ap. plication by the Oirecter cf Investigation and Research for an intenm order pursuant to section 100 of lhe CompetitiOn Act. Between: The Director of Investigation and Research ana Superior Propane, tnc., Petro-Canada inc., Tne Chancellor Holdings Corporation and ICG Propane Inc.·, 1[ 1O . 0
"Sep-14-H OZ:5Zp1 Fra-LEXECON .4. 0
Ttiat 1s, Professors Globerman and Schwindt concec:ie tnat tne "substantial presence" of
independent retailers could constrain a merged Supenor/ICG from raising retail propane prices. Tneir discussion atso snows that they recognize that the extent to wnicn Superior ts constrained by 1naepencients is an issue that cannot De detenninea as a matter of economic tneory. tnstead. 1t must De determined by analyzing tne empirical evidence. 8. In my prior repon. I empirically investigated the extent to whid'\ Superior's pnces are constraine<t by the substantial presene@ in a local area of independent retailers. My empirical analysiS of Superiors pnces showed that independents - but not 1CG - constrained Supenor's prices. I conctueled mat "Supenor's margins are not statisticany significanUy tower in areas where lCG nas a substantial presence - as would De expectea if ICG were a significant constraint on Supenor's pnces - ,~an in areas where it does not. In contrast. Superiors margins are statisticalty significantlY 1ewer in areas where independents, In 1he aggregate, have 0 a suDstantial presence" (Car1ton Repon, '33). 5 9. Professors Glot:>ennan and sc:nwindt do not report any empirieal analysis of the etrect of independents' presence in an area an Supenor·s prices. 1naeea, none of the Commissioners experts appear to have conducted sucl'I analyses, notwrtnstanding the importance oftn1s issue to the Commissioner's daims. Instead, Professors Glot>erman and Schwindt rely, for example, on an 1ntemal ICG document to conclude tnat •[t]he responses [Dy ICG personnel] systematically suggest that competition is greater 1n (markets in wh1cn tnere were several competitors to Supenor ana 1CG] than 1n •.. duopoly {1.e., only Superior ana 1CG] marl<ets. We interpret this as indicating that the eresence of other competitors Cbesicles ICG or [SuperiorJ) ennances competition in the relevant market" (Gtot>erman and Scnwindt, at 34,
5. After I filed my pnor repon, I discovered that a few t>ranches had t>een inadvertently droppea from my analysis and I discovered a few data discrepancies. Adding tnese branches to my analysis and correcting tne discrepancies aoes not subStannally affect my cooctusions -tl'\ere is no systematic evidence tnat ICG substantially constrains Supenor's pricing; in contrast, there is ev1aence that independent retailers constrain Superior's prices. My analysis is attached as Appendix A. The under1ying data and programs used to conch.act the 0
l-312-3ZZ-OZ18 T-223 P.09/60 F-794
t-312-322-0218 T-223 P. 10/60 F-794 -5-0 empnasis added). They also concluae that •[Supenor) and tCG documents iaentifying independent propane suppliers as competitors in tne relevant mal1\et. as well as documents describing price discounting by specific independents are relevant evidence· (Glot>erman and scnwindt. at 34 ).' B. The Commissioner's Experts Appear to Agree that the Ukety Effect of ma Proposed Merger on Propane Prices Depends on the Extent of Barriers to Entry and Expansion, and They Agree that Entry has Taken Place.
10. Professors Gtot>ennan and SchWJndt alsO recognize that evaluating the likely effea of tne proposed merger on retail propane prices requires evaluating the likelihood of entry or expansion by propane suppliers. In particular. tney eJ<Plain that "[t]he likely impacts of the merger on competition are further conditioned by tne potential for propane supply to expand given a price increase. Two sources of aaa1tional supply are passtDle: enuy cf new competitors and expansion of existing competitors· (Globerman and SchWindt, at 41 ). 11. Professors Globerman. and Sc:hwindt explain that ·;1 is impossitlle to be definitive about the extent, nature and speed of rJe novo entry or expansion of incumbents that might follow a significant price increase for propane in the post-merger penod· (Globerman and scnwincit, at 49). Nevertneless, they concede that "'[h)istorical patterns of de novo entry are instructive about tne nature of tne entry process" (Glot:>erman and Scnwindt. at 49). For Ul&S reason, tney investigate the evidence on entry into the retail propane business. They conctude - as do 1-tnat entry into the retatl propane business has taken place. (See Carlton Report., 27). 12. Professors Globerman and Scnw1nat examine the membership list of tne · Propane Gas Association of Canaaa rPGAC") ana fina that "there were 41 new memberships
( ... continued) analysis will be provided on data diskette. 6. Professors Globerman and Sehwindt also cite examples where they t>elieve that independent retailers di<1 not constrain Superior or ICG's prices. Cf
t-312-322-0218 T-223 P. 11/60 F-794 -6-
in the PGACH overtne period 1994 througn May 1999 (Glooerman and Scnwindt. at 50). Tney note that •[t}tlis number bk.ely "nderstates the true numDer of new entrantS,· and they concede tnat "it is dear that new entry has taken plaee" (Globerman and Scnwindt, at 50, empnasis added). They condude tnat ·small-scale entry 1s not an unusual event in propane retailing" (GloDerman and Schwindt. at 52). Professors Gtot>erman and Sdlwindt also report that, accom1ng to an internal ICG assessment of likely enlJY in local areas, about 50 percent of ICG's orancnes face a "medi~m" or "n1gn" threat of entry (Globerman and scnwindt. at 54 ). Presumably, the perceived inreat of entry would be even nigher if tne merger resulted in propane price increases. 13. By way of example, a witness statement filed on Dehart of the Commissioner snows tnat expansiOn by an en~n~ has taken piace. Don Edwards, the President and CEO of EDPRO Energy Group Inc., a rival propane reta11er, explains that EDPRO. wn1ch was formed in CJ June 1997. "nas grown tn me mar1tetp1ace in just over two years of operation. EDP RO has since brancned out to Brampton, Kitcnener-Watenoo, Hamilton ana the greater Toronto area" (Affidavit cf Don EdwaR:is, at 3 ). 7 14. Although Professors Glot>erman ana Sehwindt concede that entry into the retail propane business nas taken p1ace. and that small-scale entry by new firms is feasible. they nevertheless conclude tnat "potent1a1 entry is not a sufficient threat to render unprofitable price increases initiated by (Superior in tne post-merger environment]" (Globerman and Schwindt, at 54 ). But Professors Glot:>erman ana Scnwindt fail to provide systematic evidence to support their condusion tnat the threat of entry would not constrain a post-merger anempt to cncrease retail propane prices. In contrast. as tne evidence discussed in my prior report shows, the aggregate share of independent retailers in an area affects Superiors prices even thougn most of Supenor's independent rivals are small. Thus, a number of small entrants could constrain
7. Mr. Edwards was formeny President and CEO of Superior.
Sep-14-H 02:53pm Fra-LEXECON -7-0 Superiors prices. Indeed, the aggregate snare of independent retailers exceeds ICG's snare in
about two·thirds of the Supenor brancnes. Ill. THE COMMISSIONER'S EXPERTS' CONCLUSION THAT THE PROPOSED MERGER WIL.L LEAD TO HIGHER PROPANE PRICES IS BASED PRIMARILY ON THEORETICAL GROUNDS AND IS CONTRADICTED BY EMPIRICAL EVIDENCE.
15. The Commissioner's expens argue that it 1s "likely tnat competition will t>e reduceCl" in some local areas {Gloaennan and Scnwindt. at 55). Their c:onctusion 1s based largely on ·market structure Changes associated witn tne merger:· That is, Professors Globerman ana Scnwinat predict that because measured concentration 1n several 1ac:a1 markets will increase as a result of the proposed merger. prices in inose markets are likely to nse. 16. If they were mrrec;. one would exped to find that propane prices are currently higher in areas where Superior has a 1arge snare than in areas where Superior ana ICG 0 compete. As 1d iscussect 1n my pnor repon, I nave investigated whether ICG currently constrains Superior's pnces, and find that it does not. 17. Professors Gtot>erman and scnwindt use a1fferent definitions of local geographic markets than I rely on for my analysis.' 1n particular, tney rely en Professor Wesrs delineation of geographic markets for retail propane sales. Although Superior price information iS ava11ao1e for the Superior t>rancnes in tne West-defined markets, Professors Gtoberman and Sehwindt ao not report any empirical analysis of the effect of ICG's presence on Superior's pnces. Their failure to do so is important. because they nave thus failea to empirically test their prect1ction that me proposed merger hkely will leaa to price increases. 18. Because Professors Globerman and Sdlwinat use different tocal market aefinitions tnan I do, 1i nvestigated whether my findings wou1a cnange if I repeated my regression analysis using the West-delineated local markets. I have c:sone so and finci mat my
8. My market definitions were based on information supplied by Superior personnel. 0
1-312-322-0218 T-223 P.12/60 F-794
. ~ep-14-98 02:53pm From-LEXECOPI 1-312-322-0218 T-223 P.13/60 F-784 -··-~8-conclusions are substantially unchanged. It remains the case that tne substantial presence of independent retailers is associated with tower Superior prices, but mere is no systematic evidence that the substantial presence of ICG significantly constrains Superior's prices. 19. Professor West calculated shares for each of his local mancets. I c:ansider has market snare estimates tess reliable tnan information provldeCI to me Dy Superior, so I use sates estimates provided to me Dy Superior to calculate market shares for tne West-csefined geographic markets. J expJain how I calculate snares for these markets, and why Professor Wesrs shares appear to be unreliable, an Appendix B. 20. 1r epeat tne regression analysis in my prior repo" on tne West-csefinea markets using my calculations of market shares. My resultS are summariZed in Tat>1es 1 (for 1998) and 2 (for 1997). Tables 1and2 snow that tne substantial presence of 1CG does not systematically constrain Superior's prices if 'the West-defined loeal markets are used as the unit of obServation in tne analysis. Tnus, ll'lere iS na systematic evidence mat the substantial presence of ICG in 8 the West-defined focal markets significantly constrains Superior's prices in those markets. 21. Professors Glot:>erman and Sehwindt's Claim that the proposed merger will ehm1nate competition anci _lead to higner propane pnces iS basea primarily on "mar1<et structure" cnanges. But they apparently did not examine wnether current propane pl'lces are higher wnere Superior and 1cG ao not compete. My investigation snows that tnere iS no empirical evidence that ICG systematically constrains Superior's prices, Whether geographic mantets are basea on estimates made by Supenor personnel or on Professor Wes'(s definitions.
9. Full regression results are reportea in Appendix C.
S6P-t4-99 02:53P11 Fret1·LEXECON Table 1 Effect of Rival's Presence on Superior Marglna, 1988 (In Cents Per Liter) Substantial Presence Defined as Share Greater Than: 15 Supplier Percent Percent ICG 1.15. 3.48 lnaepenaem ·1.25. -3.78 t·STatisiics are listed Dalow coefficient estimates. • StatiSticany significant at me five percent 18V8t. 0 Eacn Ob&ervation iS me volume weigntaa &V8fllge monthly price within mar1ads as Clefined by West. Market snares are aggregates Of Spring 1999 SPI Template data accot'Qing to West's martcats. · The estimated c::oemcienunows bY how many cents Superiof's overatl margin is affected by me substantial presence of rivals, after adjusting for branch and regaon factors, inclua~ artemative ·tuet prices. For exatnJ>l8. ttte coeff1ciam estimate tor the first row and first column (1.15) mPlies 1llat Superior's ~ins are 1.1 5 cents higner in branc:he& where ICG's Share ~s 15 percent 1han in areas where ICG's snare is less man 15 peroem. If tne coeffieient es?imaie is negative (such as ·1.25 1n 1he first COiumn, second row), the coefficient implies that Superior's margins are 1. 25 cents lower in areas where me indepe:l:ldenlS' snare is Jess than 15 percem. The number under the coefficient estimate (e.g. 3.48) indicates whether the eslimatea coefficient Is stalistically significant. As a general mauer, a t·srmistiC aDOVe iwo in at>soh.ne 'ial\.18 inaicatas stadstical significance at convenuonal testing 18'181&.
t-312-322-0218 T-2?3 P. 14/60 F-7Q4 20 25 30 Percent Percent 1.18. 0.27 1.47" 4.15 0.87 4.59 -1 .71 • ·2.10. -1.71 • -5.69 -7.97 -6.57
Sep-14-99 02:5(pm From-LEXECON Table2 Effect af Rival's Presence on Supertor Margins, 1997 (In cents Per Liter) Substantial Presence Defined as Share Greater Than: 15 Supplier Percent Percent ICG 0.42 1.43 1ncsepenaent -o.eo· ·2.62 1-stallstiCS are listed D&IO'N coefficient estimates. • Statistically Significant at me flVe percent tevel. Each Observation is me VOiume w~ average momnly price within mar1cets as defined by West. Market shares are aggregates at spring 1999 SPI Temprate aara accon:tlng ro Wests mark8tS. The esurnatea coefficient 8h0ws bY how many cents Superior's overall margin i8 a1f4tet8d by 1h• suDStaniial presence of rivals. after adjusting for t>ranch and regtOn factani. inCludinQ allemadw fuaa ~ For example. 1h8 coefficient estimate tor tne first row and fim COiumn (0.42) implies mat ~rior"s margins are 0.42 cems higher in branches wnere ICG's snare exceeds 15 percent Ulan in areas where ICG's share is less than 15 ~rcent. If me coefficient estimate is negative (such as .0.80 in me first row, :second column), the coefficient imPlies that Superds margins are o.ao cenis tawer in areas where ICG's share is greaier tnan 15 P.Srcem. The number under the caeff!Cient estimate (e.g •• 1.43) indicataa Whether tne estimated coaffiCient is statistically significant. As a general maner. a 1«a11stic abOV8 two in at>SOIUt8 -value inaicaies stadstical significanai at conventional te&ting leve!s.
1-312-322-0218 T-223 P.15/60 F-794 20 25 30 Percent Percent 0.49 -o.so 0.91. 1.75 ·1.00 3.13 -0.76. -1.71 .. -1.43. -2.55 ·6.27 -5.18
Jep-14-H 02:5.fPI From-LEXECON -9-
IV. THE COMMISSIONER'S EXPERTS" CONCLUSIONS THAT BARRIERS TO ENTRY AND EXPANSION ARE HIGH ALSO IS BASED LARGELY ON THEORY AND IS CONTRADICTED BY THE EMPIRICAL EVIDENCE.
22. Several of the Commissioner's expens arg1Je that barriers to entry to the retail propane business are higtt • Professors Glot>erman ana Schwindt daim mat a variety of factors constitute barriers to entf)l or expansion.
• Mr. Kemp argues tl'\at economies of scale are a bameno entry. • Professor Wh1nston argues that Supenor and 1CG's standard contrad terms constitute a barrier to entry.
• Mr. Matnieson suggests that a Supertor/ICG "supply cost advantage- is a banier to entry.
1. Professors Globerman and scnwindt's Claim that Entry Barriers are High Is not Based on Empirical Evidence.
0 23. Professors G1ooerman and Scnwindt discuss a variety of factors that they contend constitute t>amers to entl)t. After reviewing tnese factors, tney conclude tnat
·conditi0ns surrounding entry do not correspond to me conditiDns desaibing a ·cantestaDle' market. A contestable market is one in wn1cn sunk costS associated wittl entry at an efficient scale are relatively low and entry at an efficient scale can take _place quicklY'" (Globerman and Scnwindt, at 49). Tnat is, tney claim that the tnreat of entry will not constrain pnces in the retail propane industry because sunk costs are relatively high 1n tnis business. 24. Altnough Professors Glot>erman and Schwindt recogni.Ze the importance of sunk costs to an economic analysis of entry. they do not appear to have determined What portion of entry costs are sunk. In particular, they provide no estimates of the sunk costs of entry as compared to tile total costs of entry, or of the expected profits associated with entry. Indeed, they explain that -we [cannot] assess tne typical pattem of financial performance of new entrants (i.e., are they profitable or unprofitable)" (Globerman and Scnwinat. at 51 ). Professors
1-312-322-0218 T•223 P.16/60 F•T94
.. ,Sep-14-H 02:541111 From-LEXECOll 1-312-322-0218 T-223 P. 1 T/60 F-794 • 10 .. r). - " Glot>erman and Schwindt thus provide no empirical oasis for their daim mat sunk costs - and thus enuy barriers - are high in the retail propane Dusiness.
25. As I have discussed. Professors Globennan and Scnwindt concede that numerous firms nave entered the retail propane business 1n the last several years. lndeea, mey conclude tnat ·small scale entry Dy new firms is feas11:>te, especially for applications wnicn require relatively moaest sunk cost investments. e.g., residential heatmg, and to serve local markets that are near sources of propane supply" (GloDerman and Scnwinat. at 54 ). ' 0 Tnis emp1ncal evidence of entry is inconsistent with tnelr daim that entry Darners are h19n. 26. F~rtnermore, the ability of incumbent nvalS to ex.pana sa1es can be even more of a constraint on a merged firm's at>ility to raise prices than me threat of de nova entry. For example, an incumt>ent finn couiq_ incur relatively low, or no, sunk casts if it expandeCI itS sales an incumbent firm already could have established a favorable reputabon in the local area in wh1cn it was trying to expand sales. 0 2. Mr. Kemp's Claim that Economies of Scale are a Banier to Entry is Contradicted by the Empirical Evidence.
27. Mr. Kemp claims that [t}tle Darriers to enuy into this business are formidable. The issue 1s not that an individual may be able to get the appropriate licenses or purchase tne necessary equipment: it is the at>llity to capture and secure ancnor customers, me risk of capital invested, and the ability to maintain a level of threshold business necessary to SuStain a profitable operation. This sustainable retail propane distributor will t>e cnaracterized as a customer who is fully 1nvo1vea in tne Dusiness ana nas achievea a tnreshold \/Olome in eJCcess of 2.0 [million liters per year] from a mixed customer base. Companies Who have not achieved tnis level are likely in the t>usiness as a siaeline either because they atreaay have other related act1vit1es or have ptaced a propane dispenser at a reta11 gas station. Ttlese groups wouto not be Cl'laracterizect as a real measure of competition (Kemp, at 16).
10. Professors Globerman and Sehwinat thus suggest mat entry shouta De feasible only for certain end uses. This suggeStion is contradicted by tne1r finding that "[t]ne overwhelming majority of recent entrants identified by [Superior] serve more than one customer segment (e.g., residential, commercial and industrial)" (GloDerman ana Sehwindt, at 51 ). 0
Sep-14·91 02:55P11 frOll'"LEXECOfl - 11 -
28. Mr. Kemp·s daim tnat a ·sustainable· retail propane d1stnt>utor must nave sales in excess of iwo million liters per year appears to be contradicted Dy 1he empirical e"idence. AS l have diScussed. numerous firms have entered the retail propane industry in the last several years. If Mr. Kemp's daim were correct. entrants that nave survived for a number of years presumaDly sho1.11a currently have sales in excess of two mitlton liters per year. Tne evidence does not support Mr. Kemp's daim. In particular, there are numerous ~mp1es of firms that entered the retail propane business in the last few years that had 1998 sates of substantially 1ess tnan two mi11ton liters per location." See Tao1e 3. 29. Fer example, Canwest has entered in several locations Since 1990. In at least two of mose locations, its annual sales appear to be 1ess tnan two million liters in 1998. Lo Cost nas entered three of Superior's brancn areas since 1990. ana its 1998 sales appear to be less tnan tw0 million liters in each tocation. Similarly. Neufeld Propane also appears to nave opened several locations with 1998 volumes of less than two million liters. 30. Professor West also identifies numerous ·competiti\le Dealers" wilh estimated annual sales volumes of tess than two million liters. See. for example, tabs~. 3, 4, 5, 14. 17, 18, 22. 33, 41, 46, 58, 62, 63 and 67 to nis repon. 31. Mr. Kemp's claim is based on his analysis of tne capital needed for '."a stanup greenfield retail propane operation: Mr. Kemp Dases his analysis on several assumptions that a1so appear to be inconsisteni with the empirical evidence. For example, Mr. Kemp assumes that a startUp propane distributor would need two trucks. However, an !CG assessment of its competitors shows mat many of its rivals have only one truck. Similany, ICG has only one truck at many cf its 1ocatians. 12 11. Even if Mr. Kemp's Claim were correct. several entrants nave achieved volumes substantially in excess of two miHion liters in 1998, as reported in Table 3. Tne success of tnese firms snows that entf'Y nas 1aken ptace even at current retail propane prices and mus demonstrates that firms nave surmounted any daimed tiamers to entry into the retail propane business. 12. Basec:t on a re\11ew of the ICG 'branch templates ... 0
1·312-322-0218 T-223 P.18/60 F-794
0 0. o. .~,~ · ~ I Table3 -.c..:.t. Esflmated 1998 Volume of Independent Retailer Recent Entrants -Years of Entry 1990-1998 'I -n Less ij Than .-m Two ~ Superior . Miion ~ Vearol Branch Superior Loca11o1'.' Name ol Ennnt 1998 Volume Uters? Entry Number 19IO 1 CONCORD CAMCARB 360,00D Yea s FENELON FAU.S BUD~. PROPANE 1,033,209 Yes OTIAWA LEVAC. 11,849,851 23 Vea 42 SUDBURY/SAULT SlE MARIE BOC OASES 342.328 BROWN PROPANE 99,67<1 Vea LIQUID AIR 1,026,984 Yes PAAXMR 342,328 Vea WE LOCO 308,09& Yes 8,595,562 202 Sl -CONST ANT SONIC PROPANE 10,973,017 235 JOLIETTE SONIC PROPANE -• 241 VIMONT lVAL DAVID) 241 CARON PROPANE 904,693 Yes . ~ .... I PCQ-PROPANE CONSIBNE 1,602,375 Yes -.......... PROPANE PLUS 1,173,150 Yea C•t ~ 24& lliETfOHO MINES Bo-GAR 1,132,12.8 Yes '408,837 Vea -2-48 LESCEDRES BUDGET PRO?ANE 250 RIVffiRE·DU·LOUP SONIC PROPANE 1,161,970 Yea $94,020 Yes 443 SLAVE lAKEIREO EAAlH AMP .. CANWE81 9,208,900 -. .. .. . . .. . -469 RED DEERIROCKV MOUNTAIN flOUSE QASEX 9,200,0DO LOOOST 978,000 Ye& --v MOUNTAIN VIEW COOP '490,000 Yes -~ 15,477,137 a 801 COQUIT.t.AM SUPERSAVE -"..'I." . MURRAY PATION 2,861J,000 .... 1991 1 CONCORD ':~"C""-' ' - ' ~--~ - ;-~-· ·~--·~ _, -·~ ~-,re-~ ·~ ~~ ~ •"~ -~ '"<' ·~- - •"~~·'" "'.'; ~~-- -~ ,
0 o; Table3 Ea11mated 1998 Volume of Independent Relaller Recent Entrants Years of Entry 1990-1998 Superior Yearot Branch Entry Number Superior Location f» fENELON FAU.S 13 PETERBOROUGH 39 CflATHAM 420 PROVOST 444 GRANDE PRAIRIE 1992 5 FENELON FAU.S 39 CHATHAM 445 PRINCE GEORGE BC 1993 6 FENElON FAU.S 13 PETERBOROUGH 202 Sl-OONSlANT 243 ST-ROMUALO 401 swtn CURRENT 413 KELOWNA 444 GRANDE PRAIRIE 845 PASADEHA 1994 5 FENELON FALLS I 7 GUEtPH -420 MAIDSTONE 431 FOM ST JOHN BO
() • ( .. - 'J ~ ! - cs U ""I UI 1j , .. _ ... Lesa Than m Two i"'1 Miiiion SI Name o1 Entnlnt 1990 Volume Lilers? HIGHLAND'S 400,205 Yes HtGHLAND'S 88,676 Yes UPI 304,584 Yes CAI.GAS 6,928,708 NEUFELD PROPANE 8,907,720 NORTHWOOD 200.103 Yes AABCO 6,179,909 DEL'S PROPANE 4,70S,610 BRITISH EMPIRE "400,205 Yes ~ 8Al11SH £MARE 1,&19,318 Yea "~" BUDGET PROPANE 9,286,847 ""J"". PROPANE CHAlEVOIX 3,0IM,067 ~ CANWEST 4.s28,800 -l~STAR 8,.451,()43 NOMHERN PROPANE 3,343,180 IRVING PROPAME 5,949,102 -I' " " " " VALUE PROPANE 2/m,717 -0 CAL EDON 2,626,886 "c"s PRIM AX 1.527,567 Yes -...... ct CAtfNEST 1,647,222 Yea ..... I NEUFELD PROPANE 1,626,643 Yea -.... .....
0 0, 0 " ..,A. · . --: I Table3 .c..:.t Eadmated 1998 Volume of lnclependent Retaller Recent Entrants , e e,n n Years of Entry 1990-1998 · • ...... . Less f.... f.. . Than . Two ii1 Superior MiUion SI Yearot Branch 1998Volume Liters? - Entry Number Superior loca11on Name ot ETI!ranl a,D20,1J11 807 Cl£1WYNDBC JOY PROPANE LODGE PROPANE 76,298 Yea . :• 1995 23 OTIAWA BUDGET PROPANE 2,466,028 41 NEW LISKEARD NASCO 6S41ne Vn Yes 42 SUDBURY/SAULT STE MARIE KING PROPANE 16S,11"1 BUDGET PROPANE 4/i07~73 233 GATINEAU KELOWNA PERFORMANCE PROPANE CAGENT) .400,962 Yes 413 CANWEST -4,081,963 <420 ELK POINT COAST PROPANE 327,322 Ye• 806 TERRACE I NI I 1,.440,000 Yes ... 1998 1 CONCOflD CHEMWElD NI NI I 5 FENELON FALLS CSE 200,103 Yea c:t ~ SPARLING PROPANE 2,oa1.m t HUNTSVUE LEGGAT PROPANE 375,'479 Yes 13 PElERBOROUGH SPARLING PROPAT« 3,335,703 NASO() 7-42,189 Yes 30 HEAAST -t 118,328 Ye1 I 39 CHAT""'-' EDPRO "..".. AESHU 1,040,3&1 Yes 100 WINNIPEG -1011,133 Yes -a S43 DARTMOUTH PRAXAIR . COLUMBIA fUELS 770,288 Yee -... -421 CAMPBELL RIVER ~ SUPERSAVE t,'180,663 Yes Cl Yea 459 RED DEER/ROCKY MOUNTAIN HOUSE TESKEY 1,380,000 ..".. I' . " . 203,10 Yes . 707 RADIUM SATEWTE LO COST •
0 0, 0, ~~-1 -- I Table3 Cl Estimated 1998 Volume of Independent Retailer Recent Entrants -VI Years of Entry 1999-1998 1- ...... Lesa ;,,_ Than lwo 'i't"1 Superior Vearot BR1t1ch MIHlon ~ Entry Number Superior Location Name of En11ant 1998Volume Lhers? 1997 1 CONCORD EDPRO,. 11,.480,000 SMS . 360,0DD Yes SUNRISE 17,160,000 WELDERS SOURCE 720,000 Yes 5 BANCROFT KELL Y'S FUELS 918,745 Yes 13 PET£RBOROUG+i KEUY'SfUELS 6,011,683 20 WHITBY SPARLING PROPANE 2,793,074 SUNRISE 2,tn,102 VALUE PROPANE 3,SOB,918 ~ - 22 PEMBROKE BUDGEl PflOPANE 57,611 Yes .... OTTAWA EXPERT PROPANE 300,516 Yes 23 -I 202 ST-CONSTANT . PROPANE SUROJT 6,873,7. . 1 --I 233 GATINEAU PROP EXPERT 134~8 Ve1 a; LESC£DRE$ PROPANE SUAOIT 179,408 Yes 248 260 RIVIERE-DU·LOUP MIUfARD 18S,714 Yes 343 DARTMOUlH ACTION PROPANE 82,354 Yes COWBOY PROPANE 78,.ilOO Vea -.. 704 CALGARY 709 'EOMONTONICAMROSEIWESTI.OCK BO-LI 290,783 Yes -HUGHES 2,001,87!> --0 ~ .ALGONQUIN PROPllNE 662,701 Ye& Cl 1998 6 BANCROFl NORTHMORE 200,103 Yes --n I 6 BARRIE BUDGET PROPANE 16,~3.810 -
0 0 0, . ! • ,, ' Table3 l - .C..». Estimated 1998 Volume of Independent Retailer Recent Entmnts .e.n. Years of Entry 1990-1998 I . ,. .... . Less _ lhan s Superior Two Mllllon Vear o1 Branch !ii Entry Number Superiot Location Name of Entrant 1998Volume llteJ&? CO-OP 4,204,762 SPARLING PROPANE 15,916,667 SUNRtS.lf 3~.768 ..... 23 OTIAWA Yes 248 SEPT-ILES SOUDER 59,965 KENTVILtE IRVING PROPANE 7,287,"4&1 346 350 SYDNEY SCOllA PROPANE 332,206 Yes REGINA. WEYBURN, YORKTON LO COST 119,172 Yes 401 LOW COST 103,908 Yes "°6 SASKATOON CANWEST 2' ,490.230 423 EDSON NEUFELD PROPANE 1,614,664 Yea Yes 431 FORT Sl JOHN BC C"NWEST 1,235.804 ~ Si.AVE LAK£JRE0 EARlH CM.GAS &78,700 Yes .... I ~ -.N. .. NEUFELD PROPANE 8,191,080 I C» 4'44 GRANDE PRAIRIE CANWEST -4,673,073 ~ RED DEERIROCtCY MOUNTAIN HOUSE CALGAS 6,708,957 --469 KEE BEE 316,000 Ye• 107 RADIUM SATELLITE GASPlUS 474,771 Yes NANAIMOIVJCTORIA COLUMBIA FUELS 16,680,967 --t .. I . 802 N . -. '"O . . . . . . ..... ct . .. . . ....!.,..
Sep-14-99 02:5611111 From-LEXECON 1-312-322-0218 T-223 P.24/60 F-794 -12-0 32. Mr. Kemp also !lases his analysis on tne assumption that the startup propane retailer would ha'lle a storage tank witn a capaeity of 30,000 u.S. water gallons ('"USWG"). However, ICG believes Ulat a substantial numt>er of its rivals nave storage tanks smaller tnan 30,000 USWG (e.g., 18,000 USWG).':i 33. Mr. Kemp·s analysis also assumes that a startup propane distributor would l'lave a particular customer mix. "Tl'le customer makeup for a small operator would largely be residential, commercial, agricultural and industr1ai- {Kemp, at 8). In particular, ne assumes a customer mix of 40 percent residential; 17.5 percent commercial; 20.0 percent agricultural; 17.5 percent industrial: and 5,0 percent otner {Kemp, Tat>le 3.2). However, the empirical evidence is not consistent with Mr. Kemp's assumptions. For example. Superior believes tnat the business mix for several recent entrants is ~ubstantially different than the one Mr. Kemp assumes. For example, only 20 of tne 106 entrants (since 1990) identified by Superior had a a.istomer mi>l wl'\ere residential sales accounted for Detween 30 and 50 percent (i.e., about 40 percent) of 0 sale. 14 See Tat>le 4. 34. Finally, Mr. Kemp provides no empirical SL1pport for his claim inat entrants that se111ess than two million liters per year are not a ·rear source of competition. As I have discussed, my analysis in my poor report showed that the aggregate presence of independents. not necessarily individual indepenaent retailers, constrains SL1penor's prices. 3. Professor VVhinston's Claim that Standard Contrac:t Terms Reduce the Threat of Entry is Based on Economic Theory Only and is unsupported by
Empirical Evidence.
35. Professor Whinston Claims that contractual previsions in Superior and 1CG·s standarQ contracts ·have the effed of making competitive entry t>y new firms, as well as expansiOn t:>y existing ri'lals. less aurac:tive. This reducea threat of entry Dy new firms and 13. Based on a review of the tCG 'branch templates.· 14. CL1stomer bL1S1ness miX fer entrants is taken from the Superior-supplied 'l>rancn template·-·
0 0. 0 .C,j) • Table4 --f Eatlmatad 1998 Bualnus •x of Independent Retailer Entrants co "" Years of Entry 1990-1998 UI I- .... ~II' ef S.181 lnEndU.. ........ 0 8-lllDf R8•'"30 •I YNl«ll CW>!iO .-E.ll1ry Number Superior Lemllon Namt ol £nll'IRI .,.. Ag11cd1111ll Aull) Com--' lndultnll Re~ Pea . . . , -, "~' tllO 1 COllCOllO CAMCMlll uo am ooe Gill 19000 Dot ., !i 6 fEIElON IAUB llJD!JEl l'ftOPANE 12• Ht U 2 .., 3 OIUWA l.EYAO ..... ...... ,,.I, .46.6..1 , DY 8101 •• •1.01 v. . IUDllURY~\l.181E MAftlE IOCG.\SEB OOD O.DO OAMt ... 10UG I.DO v. . llAOWNmDPMIE OOD om ODO at.211 84114 ...., UOUIDAIR !JOO om Oot ooa l8000 oo• .,. . - f'RA)CAIR DOO DOCJ 000 ODO 10000 GOO y. . wa..oco 0.00 . OCG 000 DOD 10UO GOO v. . I 81--00NST AHi IOMCPROl'Mft '"" 44.ll9 ,.., 14.11 2'• as .IOUETlE SONICPAOPAtfE 1U ., ll 10.IM .ID.1.9 .... ... v.. •• VII '41 "IMONl (Y.-1. DAVID) 241 CMOHPAOl'Mff o.oc. O.ot om ant PCQ.f'AOMNE COH9tOIH£ ·o-w om Got ,. ..~ MOO •• v. . f'ROPNU PlUI DOI> om DOI 1 ., 1 14 •• 1>00 Ya "'~ 1MF1FORI> MINH ll0-9llR 011 ID.l6 000 .. ..... 13.10 .•.•, Ya "' IHCWRfB 9Ul>llFl PAOl'ANE •47 Oll 9Jlt ez.M Y• 260 IWIElll· OIK.oUP 90NIC l'flOl"Mll 043 1816 •.oa 80.fl' 2904 O.GCt \'• . 4 . 43 . SIAVf lM!UAED EAl\1H MP ... D..00 .. ltAI e&tt t.at V416 CAMWEtn ... 0.00 1.a1 1 tt ... tM Ye• ftE"OOHMtOOICVMOOHTAllt HOUBE GASlJ( •• . I . U . IO . \DOOlt OOI ODO ow Yn I ~ lOCXJlll OGO 47.Ga H6 0..80 aue I .,, t.IOUN'IAIN VIEW COOP O .D..D J041 0.00 000 O.OG ,. ... Y• -"" COWMM BUl'fl\M'tl GDO euo IUO SM •• "" "" ,. .. "I" IN ' CONCORD MUAMY PATIOff DOD 0.00 1DD.DO tN 0.00 OGI Yel --• FPID.ON fALLI HIOHLAND'I 1U7 0.00 e.o ••S 000 •• ti 1'f"1E IUIOROUOH tt191UND'8 tJIO uo 0.111 000 DAO tto.00 Vii • CHAlfWI u,., uo ..,. O.oD 000 •• .au1 ,. 4'0 PllO'IOST CAI.GAi o• , .,. . .. o.tt 0.47 •.r1 ,.. YM 444 GRi'Mll PIWRIE •Uftl.D PADPME 009 au• 'BA1 1109 tD.11 ..... ..... • IN l'fNE I.ON Fill.I.I NORfllWOOD tUY II.OD 84' ..... 0.80 »DI ..•..•. c;K\"TlMM M1ICO 140 121 1H1 .tall •• .,. YM -PllllOE GEOAliE 8C DEl'lflROl'ME . uo HD I041 111.IO 14AS ,. ... v. . -0 "U"I tllD -$ fDIElONFM.LS 8Alll9M EUPIRE , 1s . . J , 1 •.oo ..., 000 ~ 11 PnEABOAOUGH ...l llSlt EMl'IN: . ..... tu• ·64~20 nu •• •..•.. 8'.cDN81Atfl e110Gn MDPAME , U ,. 4 . R ,. U .. D . OAIO SUl8 ..., ..,. . co .... M ST-llOMIWn PROPANE QIW.£VOllC tO.t.I 6119 o.oc> "• «>t 8WIF1 CUN'.EM CANWI.81 000 ooe 000 De? et.W 'H"t "V"• -. I . . . . .. .
., ' .c,;, ! -.. . . Table4 I c:::t Estimated 1918 Bualnea Mix of Independent Retailer Entrants Years of Entry 1990-1998 -.e.n. 11 .... ... .....l lllg9ot S.i.. ti £llcl UM .... s.trto1 '*-•30 If Yu1oll lllMdl Ol>fiO .-Gn1ly h\llnbel 8upMll loodall Nlllnl or f'l1hnl l'JnOIAnl AUIO ~ lndue11111 Rollllfen1flll . fl'llrotnl? "~' 413 ICt'l.DWNA TRIBIAA 000 DIJ 'M• WO . H . CI 1318 .,. . ~ 44-f GIWClE PMVllE NORTHrANf'ROPMIE , 00 . . 0 . II.II 40• U'l 11. .. ... IU1 en _ PASADENA IRVltfB PAOPA!tE 0..00 HI to81 aoo 'Yll 111M 6 ff1EQN 'AU.I \IAWE PROPNl1 1119 1.» U3 474' ODO 11.86 1 GUEi.PH CAL.EDON GOO ... llUI llti 5Ut t.00 'I'• MIAAX eoo HO 0.00 •• 073 •ne . C . 'O IMIDllONE GAHWE&T ·uo •• ... 1 DI ...... U8 'l'•I 431 , POI" St JOHN ec NlurlLD PACIPME .., C1€TYNH>BC ¥JYPAOPANE ,. .. eOD 1 a U1' 14.GI OJIO DAiii ., .. Oft t• 1ua t.GD 3011 .. LOOGI l'llOPANE o.eo 000 Got •011.110 o.eo 0..00 'Y• 2J Ol1AW1' llUDG£T PROl'AME ODO :sr.u 000 41111 DGO ttn 'I'•• 41 HEWU8K£NC> IMBCO ODO O..IO ODD ... 710t ... v. . 41 IUDllUR'flMUl.1 8fE MAftlE ICING PROPANE too O.IO OJID ... ao-n 4401 :1133 MTllU.AU auoan PIAOPANE ... 4M OJID 11.a 1in1 1316 Y• . 41 . $ ICELOMIA PEMORIMlllC£ f'ftOPME ~ 000 D.10 0.00 4U1 HD MOI v. . uo EUCPOan CAN'Anf ... tW 1• Mt •21 •14 Y• I lllFIMCE Oo.\81 PROPANE UD •.oo .... OAIO •• 11.oe v. . ... . ~ ... , I IXINCOllO GH£ t.twr:AJ) 41.W 000 o.oD O.GO UIOA» ... v. . ' .... • R!N[LON rAtU C81 1 t 1 r .1 • 7 OOD eu ... 11'.0I .-... I llWllNG PFIOPAME u• ••• .•.". . c . a . M16 ........ • HlM'&VllU 1.EGICMT PROPAME GOD O.OD 000 ow 009 100.00 Y• 13 PE1tAllOAOUOlt 91'MUJND f'NJPMIE U1 19.M ..... tetlt Yel -IO HEAMt NASCO .·.~. D.00 1145 ar.ee ... 40.N • OMTHW IDPAD Otlt lt.00 .0..0.0 D.O• . toO . O . D DQO \'• WHil•to AHtlU HD 0.00 tlo.7' 140. 1111 100 \'• 0.00 0.00 . 14 . 3 DMTNOUTH •RAJWA ,, t. . 0 .. 0 . IUD ... "I• 411 CM1P8W. fW£A ~ruua ". . •...•.. .. 11• 1470 0.00 ti.It V• .... 8UPERSAVE - llll4 ~Ill O.OC> aw ... 'tel llEll DEf IWAOCl(Y MOlllflA IH ffOUll 1'91CEY •o•• •• 1110t0 O.GO Oat O.llO -llllCllUU MTU.IJlt tOOOSl OOt o• O.GO Oot 1llOAIO "Y'•• .... 101 CClHOOIJI) - I .... ll>MD 0.00 000 .,. ..J ODI» •ua om 11111 Y• -' .~... O.ot 000 1100 0111» 1DOID 000 'Y• iUfRlllE OCIG O..IJO tlOOO ODO DOD 000 Y• .... W!.LOIRI 80U"l:l 000 o.ao D110 ..0..0.0 10000 000 Yea ..I. . 6 MHCROf1 KEIU'8fVEU DOI OJID 0.00 0.00 1361 Y• .. ~ ..
.. 0 o. 0 ·..,... ! Table4 l Estimated 1998 Business Mix of Independent Relaller Entrants ... "'" Yeara of Entry 1990-1998 . en i ,_,.nlageid S.lel tn End U• ... "81..<30 V•ol ~ . B i EnlPy .. r . a .. ·n .. c . h .. or>50 811P1tfol Loollioll Nllll1e ot rmnt .,.; ~lrlOlllU'lf "'*' Co11 . 1 . llM .. ld lll lndullrllll ......... Pe"*111 11 PmABOROUGtt KW' l'&fUUS •• . . . . . . . . . 1 . 1. .. uo 'Mtlm' Sf'MUNGf'IO'Mtf 11.a •• ..... m IO 0.00 IUI U1 IUI v. . 8UNI& 2116 uo 7JA5 000 0.81) om Yel VA1.W PROPllNE 0.1111 4.l& »llfl 2HJ 14'11 8'11 21 l'fMIAOICE BUOGnl'IUMtE ... 0.00 0.81) 11141 000 116.11 ~ 01lA~ EXl'f RI PAOPllNE 000 0.00 o• DJIO IGOOD CAMI .,. . m ST .allf81jl,Nl MOP~EBUROll ,~. ... 0.00 '1M ... IUI 'Y• . 2 . 3 . 3 . WtllNMI PAOPDPEM 000 DOO OOG IU1 ..... DUO 'Y• .. LUI OE.D11E S PftOf'.v.E 8Uft01f .... D'M OOll Ul:I 7UO ... Y• 2!IO , RIVKRHJIH.OUP MIUJATG> uo 0..00 0.00 GG• lH4 n.ae 'V• ~ AC'llONPRDPANE HO 0..00 o.DO GOD OGO ICJO.OD Y• .1,0.4. CALAWt'I' COWBOY~ft HO I.ID 0.00 llCl.00 000 009 .,. . '°"°"10NICAMAOSUNES'll..OOI\ llOU 000 s• - HUQHU 011 o• • 0 0 .0 1 0 DAO .0.0 •• .0 •o• v. 000 .,. .. • MNCftOfT ALOONOUINPAOl'.-Nf 114 OAID 1M7 .... 000 M711 \'• NORlfflQIE. 18.11 GOO ... • ••• o~ ... IWRE 9001K'T PROP/lftt Ua \ . I. . a . ODO 000 HI ID.PA Y• COOP UC> 91.18 o.ao . oa . o 009 IHt .... .-.I . ,,. SPAAl..tlG f'IQl'ME U4 oec. ICU1 ODO 7611 .., OTlAWA SUNIU9£ 0 . .o . l 1.00 1111109 , 0 11 , > 0.llO ,a. . .G .. O , . . ., ~ - •• •• . .. .=::: ,M.6. llEPT.US IOUDER 0 , . ol 000 .... - llDmlW IAYINGPllOPAll! m11 to.r4 JM.ti ••• ,, .. v.. 6 ~ 8YllNEY ICO'l~PADPNli Oot oot 0.00 1114 o• ••• Ve• RE~ WEYBUllf. VORK'IDN UIC08l 000 OOI I.OD 000 000 100.GI v. . -.... LOWCOBf ODO 0.00 •• o.ao 000 1DOOI Yei MSKATODH CMMIST . • . • . 0.00 '1.11 O.llO .,om 1tl Y• 413 mlOM NEVFUDMOl'Nll! uo DJJO llltl ,,., HAt Y• .4.1.1, •• FOm 8l IOliN IC ~t\MMST D.10 tea.DO •• 0.QI) O.DD Y• 81.AVi IMEmED fARnt C"1.GA8 •• . • . • . llDTI HO .71.2.1 0.111> V• i - NEUFEUIPR>PMU I.DI o.m 141 t081 tn Y• ti <M4 mw.OEPMRI CNMUll 000 71.114 1Ht U4 409 Y• llEO CU IVAOCIC'f llOUlff" 1N ffOUll CN.GAI 008 .... WA toe 13"8 1U.t V• "'O llCEOEf DOD DOI uo· t112 om IUI YM ... 707 RADIUM IAlllU'lt G.\IPWS, 000 D.00 uo .. 0.00 11"44 f.YI Y• :::! g IO'I fWWMIWICTI)Alo\ ~f\1£1.& ooa OAIS •ta• ,,.. 1 DI '1•1 .... .... ~ 4flo.
Jep-14-81 02:5TPll From-LEXECON -13-0
expansion by existing ones all reduce tne competitille constraints tnat might otherwise ne1p limit
a merged [Superior]/ICG's exercise of market power" (Whinston, at 3). 36. Prcfessor Wh1nston's Claim iS Dased on tneoretical groundS, ana in any event aoes not establish tnat Supenors standard contract tenns make entry more difficult. Indeed, witness statements Dy nval propane aistributors filed on behalf of tne Commissioner suggest tnat the type of contract terms a1scussed by Professor Whinston may mal<e entry easier. For example, according to Sob Bush, Vice President of Cal-Gas Inc.. 1a] five year contract 1s typical for a large account. Time is needed to cover the cost of tne equipment, installation of the eQuipment, pipelines from tankers to customer buildings, etc.'" (Statement of Bot> Busn. 1J 27.) Similarly, MacDonaia·s Propane nas five year contracts with customers (Statement of Rod MacDonald, at 1.) 37. Professor Whinston·s claim appears to t>e based on the theory tnat IOng-term 0 contracts ma1te enuy more difficult because tney "forectose" sufficient sales volume so that an entrant will not t>e abte to enter at an efficient scale. However, as I nave discussed, many independent retailers appear to operate at small volumes (1.e., less than two. million liters per year). Thus. Professor Wh1nston has not estaDlished that "foreclosure· is a concern in tne retail propane t:>usiness. 38. Even if Professor Whinston were correct mat mese standard contract rerms reduce competition, the unClertakings that Superior has made, 1f implemented, resolve tne anticompetitive concerns ra1sea Dy Professor Wl'l1nston. indeed, if these contract terms reduce competition as Professor Whinston Claims, then the unaertakings that Superior nas made are a procompetit1ve result of tne merger. That is, allSent tne merger, tnese a11eg~1y anticompetitive stanaard contract terms could remain in place.
{ ... continued) data.
1-312-322-0218 T-223 P.28/BD F-704
Sep-J4-H OZ:57pm Froa-LEXECON 1-312-322-0218 T-223 P.29/60 F-794 -14-0 39. Furtnermore, if Superior were unable to enter into these types of contracts, 1t could be disadvantaged relative to rivals who were not constrainecf in the same way. In this case, disallowing effioent contracts between Supenor and potential customers could rectuce competition (between, for example, Superior and Cal-Gas). Tnus, if the Commissioner is incorrect that Supenor's contracts with its customers are anticompet1t1ve, disallowing such contracts may narm consumers. ' 5 4. Mr. Mathiason•s Suggestion that a SuperiorncG Supply Cost Advantage is a Barrier to Entry is a Fundamental Economic Error. 40. Mr. Math1eson·s analysis ·examines the barriers to entry tor potential new paniopants and specifically how high me bar has been raised with tne formation of me mergea company" (Mathieson, at ii). Mr. Matn1eson concludes mat tne proposed merger will give tne mergea firm a cost advantage over potential entrants. For example, he concludes tnat ·any 0 small new entrants wilt be at a disaavantage to the merged company in regard to tne cost of supply acquis1t1on" (Matnieson, at 87). He atso condudes that •it appears likely that the merged company would have significant transportation cost advantages over new entrants• (Mathieson, at 87). 41 Mr. Math1eson·s Claim that such merger-related "cost ad\lantages" are a earner to entry is a fundamental economic error. Merger-related reductions in costs are efficienoes, wh1cn are procompet1ti\le. This is a fundamental tenet of merger analysis recogn1zea in tne Comm1ss1oner's own Merger Enforcement Guidelines. Indeed, any merger-specific efficiency will give merging parties a cost advantage over rivals. 42. None of the CommiSsioner·s expens has aasputecs Superior's daim that there are substantial effioencies associated With me proposea merger. Indeed, Mr. Mathieson agrees that the merger will anow tne merged firm to reduce its costs. However, none of tne 15. This is a patt1cu1ar concern if the merged firm will be an especially efficient supplier. as Mr. 0
.. Sep-1'·19 02:57Pll From-LEXECON 1·31Z-3ZZ·D218 T·ZZ3 P.30/60 F-794 0 Commissioners experts has investigated, or even discussed.· the hkely effect of these efficiencies on retail propane prices. This is a key flaw in the Commissioner's experts' reports wttnout sucn an analysis, the Commissioner's expens cannot condude tnat the merger will hkely lead to price increases. As I discussed in my pnor report. ·even a monopolist has an incentive to pass along some portion of cost savings Decause 1t increases its sales (and profits) by ao1ng so" (Cartton Report. 125). Thus, if a merger generated sufficient cost savings, even a ·merger to monopoly" cculc:S 1eaa to~. not nigher prices. 43. Merger-related cost efficiencies will give the merging parties an advantage over rivals only if the merged firm uses its 1ower costs to reduce its prices. If me merged firm attempted tc increase its price after a merger, rivals wouia find it easier, not narder, to compete. Thus, rivals will De disadvantaged by a merger only if the mergea firm reduces itS costs as a result of the merger and passes along some or all of those cost savings to its customers. 44. I note that several of Supenor's rivals have filed witness statements on behalf of 0 tne Commissioner. Several of these rivals' statements suggest tnat the merger will aead to lower retail propane prices. For example, a representative of Autogas Propane argues tnat tne merger could allow Superior to "k.eep prices depressea" (Statement of Bot> Good. at 2). Similarly, a representative of Mutual Propane claims that •[t)he combined entity wou1d have the means to create a spot price war" (Statement of Jack Osland. at 1). 16 This eviClence iS consistent witn Superior's claim that tne merger will make it a more effect1tJe competitor, and contradicts tne Commissioner's da1m tnat the proposed merger will result in a subStant1a1 lessening of competition.
(. .. cont1nuecl) Mathieson (ana otner witnesses for me Commissioner) suggests. 16. Tne Commissioner nas not daimed that tne merger will lead to preaatory pricing by the mergea firm. That is. the Commissioner has argued tnat Ule merger will leaa to nigher, not lower, retail propane prices. - -0