megaprojects and risks - english

8/10/2019 Megaprojects and Risks - English

1/30


2/30

MegaprojectsandRisk

Megaprojects and Riskprovidesthefirstdetailedexaminationofthephenomenonof

megaprojects.Itisafascinatingaccountofhowthepromotersofmultibillion-dollar

megaprojects systematically and self-servingly misinform parliaments, the public

and the media inorder toget projectsapproved and built. It shows, inunusual

depth, how the formula for approval is an unhealthy cocktail of underestimated

cost,overestimatedrevenues,undervaluedenvironmentalimpactsandovervalued

economicdevelopmenteffects.Thisresultsinprojectsthatareextremelyrisky,but

wheretheriskisconcealedfromMPs,taxpayersandinvestors.Theauthorsnot

onlyexploretheproblemsbutalsosuggestpracticalsolutionsdrawingontheory

andhard,scientificevidencefromtheseveralhundredprojectsintwentynations

thatillustratethebook.Accessiblywritten,itwillbeessentialreadinginitsfieldfor

students, scholars, planners, economists, auditors, politicians, journalists and

interestcitizens.

BENT FLYVBJERGisProfessorintheDepartmentofDevelopmentandPlanningat

Aalborg University, Denmark, and author of the highly successfulMaking Social

Science Matter(Cambridge,2001)andRationality and Power(1998).

NILS BRUZELIUS is Associate Professor at Stockholm University and an

independentconsultantontransportandplanning.

WERNER ROTHENGATTER isHead of Institute ofEconomic Policy Research

andof the Uniton Transportand Communicationat the UniversityofKarlsruhe,

Germany.


3/30

MegaprojectsandRisk

An Anatomy of Ambition

BentFlyvbjerg

NilsBruzelius

WernerRothengatter


4/30

"

Contents

List of figurespage VI

List of tables VII

Acknowledgements VIII

1.

Themegaprojectsparadox1

2.

Acalamitoushistoryofcostoverrun11

3.

Thedemandofmegaprojects

4.

Substanceandspininmegaprojecteconomics

5. Environmentalimpactsandrisks

6. Regionalandeconomicgrowtheffects

7. Dealingwithrisk

8.

Conventionalmegaprojectdevelopment

9.

Lessonsofprivatisation

10.

Fourinstrumentsofaccountability

11.

Accountablemegaprojectdecisionmaking

12.

Beyondthemegaprojectsparadox

Appendix.Riskandaccountabilityatwork:Acasestudy

Notes

Bibliography

Index


5/30

"#

Figures

2.1

ConstructioncostoverrunsfortheChanneltunnel,page

GreatBeltlinkandOresundlink12

2.2 Yearby-yeardevelopmentinforecastsofconstructioncostsandroadtrafficfortheGreatBeltandOresundlinks13

2.3 Inaccuraciesofcostestimatesforrail,fixedlinksandroads

in258projects172.4

Acenturyofcostoverrunin111projects18

3.1ForecastandactualtrafficfortheChanneltunnel

3.2 Inaccuraciesoftrafficforecastin210transportinfrastructureprojects

4.1Eurotunnelshareprices,1987-2001

5.1 ImpactoftheGreatBeltlinkoncommonmusselsbefore,duringandafterconstruction

7.1Theriskmanagerprocess

9.1ThetypicalBOT(build-operate-transfer)approachtomegaprojectdevelopment

10.1Astakeholder-basedapproachtodecisionmaking

10.2Frenchapproachforensuringtransparencyinmegaproject

decisionmaking

10.3Performancespecificationsandthestakeholderapproachincombination


6/30

"##

Tables

II.iExamplesofconstructioncostoverrunsinlargetransport page

Projects14II.iiSpectacularprojectswithspectacularcostoverruns19II.iiiConstructioncostdevelopmentinfourprivatelyownedtransportinfrastructureprojects20

III.iExamplesofprojectsthathaveexperiencedseriousrevenue/benefitproblemsIV.iCostoverrunandridershipfortwelveurbanrailprojectsIV.iiExamplesofprojectsthathaveexperiencedviabilityproblemsVIII.iStepsintheconventionalapproachtoprojectdevelopment

IX.iCostandtrafficdevelopmentinfourprivatelyownedtransportinfrastructureprojectsXI.iAlternative1:theconcessionapproachtoprojectdevelopmentXI.iAlternative2:thestate-ownedenterpriseapproachtoprojectdevelopment


7/30

"###

Acknowledgements

We wish to thank the people and organisations who helped make this bookpossible.SpecialthanksmustbegiventoPatrickPonsolleandJohnNoulton,ofEurotunnel,MogensBundgaard-Nielsen,ofSund&BltHolding,andOleZacchi,oftheDanishMinistryofTransport.Notonlydidtheyandtheirstaffhelpuswithdatafor the book`scase studies, theyalsogavecritical commentsonanearlierversionofthebooksmanuscript.

We also wish to thank Martin Wachs, of the University of California atBerkeley,andDonPickrell,oftheVolpeNationalTransportationSystemsCenteratCambridge, Massachusetts, for their comments onouranalysis of cost overrun.Per Homann Jespersen, of Roskilde University, provided helpful input to ourconsiderationsregardingenvironmentalimpactsandrisks.RogerVickerman,oftheUniversityofKentatCanterbury,gavevaluablecommentsonthechapteraboutregional and economic growth effects. Thanks are due as well to the followingcolleaguesfortheirkindhelpatvariousstagesintheresearchandwritingprocess:Jim Bohman, Irene Christensen, John Dryzek, Raphael Fischler, RalphGakenheimer,MaartenHajer,MetteSkamrisHolm,AndyJamison,BillKeith,Finn

Kjrsdam, Mary Rose Liverani, Kim Lynge Nielsen, Tim Richardson, YvonneRydin,EdSoja,MichaelStorper,AndyThornley,JimThrogmortonandAlanWolfe.Two anonymous Cambridge University Press reviewers provided highly usefulcommentsforpreparingthefinalversionofthetypescript.

Thetransportsectoranditsinstitutionsarehardlyinthevanguardregarding

freedomofinformation.Insomecaseswewereunable,usingtheformalchannelsforinformationgathering,togetthedataandin-depthinformationweneededtowrite the book the way we wanted to write it. We are grateful to those boldindividuals who, when the formal channels dried up, found informal ways offurnishing uswith the information we lacked.Wementionnonames for obviousreasons.

Lilli Glad expertly transformed our drafts into readable manuscripts. Anni

BuskNielsenprovidedprecioushelpinacquiringtheliteratureonwhichthestudyisbased.TheresearchandthebookweremadepossiblebygenerousgrantsfromtheDanishTransportCouncilandAalborgUniversity.Finally,wewishtothankour


8/30

Acknowledgements

#$

editoratCambridgeUniversityPress,SarahCarowhoprovidedvaluablehelpinseeingthebookthroughtheprintingprocess.BentFlyvbjergwasteamleaderfortheresearchonwhichthebookisbasedandisprincipalauthorofthebook.Weapologisetoanyonewehaveforgottentomentionhere.Responsibilityforerrorsoromissionsinthisbookremainsoursalone.


9/30


10/30

1. The megaprojects paradox

A new animal

Whereverwegointheworld,weareconfrontedwithanewpoliticalandphysicalanimal:themultibillion-dollarmegainfrastructureproject.InEuropewehavetheChanneltunnel,theresundbridgebetweenDenmarkandSweden,theVascodaGamabridgeinPortugal,theGermanMAGLEVtrainbetweenBerlinandHamburg,the creation of an interconnected high-speed rail network for all of Europe,

crossnationalmotorwaysystems,theAlptunnels,thefixedlinkacrosstheBalticSea betweenGermanyandDenmark,plans for airports tobecome gateways toEurope,enormousinvestmentsinnewfreightcontainerharbous,D M200billion

worthoftransportinfrastructureprojectsrelatedtoGermanunificationalone,linksacross the straits of Gibraltar and Messina, the worlds longest road tunnel inNorway,nottospeakofnewandextendedtelecommunicationsnetworks,systemsofcross-borderpipelinesfor transportofoilandgas,andcross-nationalelectricalpower networks tomeet the growing demand in an emerging European energymarket.Itseemsasifeverycountry,andpairofneighbouringcountries,isinthebusinessofpromotingthisnewanimal,themegaproject,ontheEuropeanpolicy-

making scene. And the European Union, with its grand scheme for creating so-called`Trans-EuropeanNetworks,isanardentsupporterandeveninitiatorofsuchprojects,justasitisthedrivingforceincreatingtheregulatory,andde-regulatory,regimesthataremeanttomaketheprojectsviable. 1

Thesituationissimilarinindustrialisedandindustrialisingcountriesinother

partsoftheworld,fromAsiatotheAmericas.Thereis,forexample,HongKongsChekLapKokairport,ChinasQuinlingtunnel,theAkashiKaikyobridgeinJapan,Sydneysharbourtunnel,MalaysiasNorth-SouthExpressway,ThailandsSecondStageEspressway,andproposalforanintegratedEurasiantransportnetwork. In

the Americas there is Bostons `Big Dig , freeways and railways in California,Denversnewinternationalairport,CanadasConfederationbridge,theSoPaulo-BuenosAiresSuperhighway, theBi-OceanichighwayrightacrossSouthAmericafrom the Atlantic to the Pacific, and the Venezuela-Brazil highway. Even aproposedUS$50billionprojecttolinktheUSAandRussiaacrosstheBeringStraitthe`biggestprojectinhistory,accordingto itspromoters itsnotmissinginthemegaprojectschemeofthings2.Outsidethefieldoftransportinfrastructurethereis


11/30

Themegaprojectsparadox

%

theThreeGorgesdaminChina,Russiasnaturalgaspipelines,thePergaudaminMalaysia, flood control in Bangladesh, in Bolivia-Brazil gas pipeline, theVenezuela-Brazilpowerlineand,againandeverywhere,theultimatemegaproject,theInternetwithassociatedinfrastructureandtelecommunicationsprojects.

Zero-fr ict ion society

Megaprojects form part of a remarkably coherent story. Sociologist ZygmuntBaumanperceptivelycallsitthe`GreatWarofIndependencefromSpace,andheseestheresultingnewmobilityasthemostpowerfulandmostcovetedstratifyingfactorincontemporarysociety3.PaulViriliospeaksofthe`endofgeographywhileothers talkof the `death ofdistance.4Bill Gates, founderand chair ofMicrosoftCorporation,hasdubbedthephenomenon`frictionlesscapitalismandseesitasanovelstageincapitalistevolution.5WhenMicrosoftandGatessingleoutaconceptoraproduct,oneiswelladvisedtopayattention.`Frictionlesssocietymaysoundlikeandadvertiserssloganinthecontextofitsusage.Itisnot.Thetermsignifiesaqualitativelydifferentstageofsocialandeconomicdevelopment.

Itthisdevelopment`infrastructurehasbecomeacatchwordonaparwith

`technology.Infrastructurehasrapidlymovedfrombeingasimplepreconditionforproductionandconsumptiontobeingattheverycoreoftheseactivities,withjust-in-timedeliveryandinstantInternetaccessbeingtwospectacularexamplesofthis.Infrastructure is the great space shrinker, and power, wealth and statusincreasinglybelongtothosewhoknowhowtoshrinkspace,orknowhowtobenefitfromspacebeingshrunk.6

Today infrastructureplaysa key role in nothing less than the creation of

whatmanyseeasanewworldorderwherepeople,goods,energy,informationandmoneymoveaboutwithunprecedentedease.Herethepoliticsofdistanceisthe

eliminationofdistance.ThenameofutopiaisZero-FrictionSociety.Andevenifwecan never achieve utopian friction lessness, we may get close, as is currentlyhappening with the spread of the Internet. Modern humans clearly have apreference for independence from space and are consistently undercutting thefriction of distance by building more and improved infrastructure for transport,includingtelecommunicationsandenergy.

Megaprojects form part of a remarkably coherent story, the `Great War of

inde endence from S ace .


12/30

MegaprojectsandRisk

&

Megaprojects are central to the new politics of distance becauseinfrastructure is increasingly being built asmegaprojects. Thus thepast decadehasseenasharpincreaseinthemagnitudeandfrequencyofmajorinfrastructureprojects,supportedbyamixtureofnationalandsupranationalgovernment,privatecapitalanddevelopmentbanks.

Performance paradox

There is a paradox here, however. At the same time as many more and much

largerinfrastructureprojectsarebeingproposedandbuiltaroundtheworld,itisbecomingclearthatmanysuchprojectshavestrikinglypoorperformancerecordsintermsofeconomy,environmentandpublicsupport.7Costoverrunsandlower-than-predicted revenues frequently place project viability at risk and redefineprojects that were initiallypromoted aseffective vehicles toeconomic growthaspossible obstacles too such growth. The Channel tunnel, opened in 1994 at aconstructioncostof4.7billion,isacaseinpoint,withseveralnear-bankruptciescausedbyconstructioncostoverrunsof80percent, financingcoststhatare140percenthigherthanthoseforecastandrevenueslessthanhalfofthoseprojected(seeChapters2-4).ThecostoverrunforDenversUS$5billionnewinternational

airport,opened in1995, was close to200per centandpassenger traffic in theopeningyearwasonlyhalfofthatprojected.OperatingproblemswithHongKongsnewUS$20billionChekLap Kok airport, whichopened in1998, initially causedhavocnot only tocosts and revenues at the airport; the problemsspread to theHongKongeconomyas suchwithnegativeeffectson growth ingross domesticproduct. 8 Afterninemonths of operations,TheEconomistdubbed theairporta`fiasco , said tohavecost theHongKongeconomy US$600 million.6The fiascomayhavebeenonlyastart-upproblem,albeitanexpensiveone,butitisthetypeofexpensethatisrarelytakenintoaccountwhenplanningmegaprojects.

Somemayarguethatinthelongtermcostoverrunsdonotreallymatterandthat most monumental projects that excite the worlds imagination had largeoverruns.Thislineofargumentistoofacile,however.Thephysicalandeconomicscaleof todaysmegaprojects issuchthatwholenationsmaybeaffected inboththemedium and long termby the successor failureof just a single project. AsobservedbyEdwardMerrowinaRANDstudyofmegaprojects:

Many projects have strikingly poor performance records in terms of economy,

environment and public support.


13/30


'

Such enormous sums of money ride on the success of megaprojects that company balance

sheets and even government balance-of-payments accounts can be affected for years by the

outcomes. . . The success of these projects is so important to their sponsors that firms and

even governments can collapse when they fail. 10

Even for a large country suchas China, analysts warn that theeconomicramificationsofanindividualmegaprojectsuchastheThreeGorgesdam`couldlikelyhindertheeconomicviabilityofthecountryasawhole. 11Statedinmoregeneral terms, the Oxford-based Major Projects Association, an organisation ofcontractors,consultants,banksandothersinterestedinmegaprojectdevelopment,inarecentpublicationspeaksofthe`calamitoushistoryofpreviouscostoverrunsof very large projects in the public sector . In another study sponsored by theAssociation the conclusion is, `too many projects proceed that should not havedone.12WewouldaddtothisthatregardingcostoverrunsthereisnoindicationthatthecalamityidentifiedbytheMajorProjectsAssociationislimitedtothepublicsector.Privatesectorcostoverrunsarealsocommon.

Forenvironmentalandsocialeffectsofprojects,onesimilarlyfindsthatsuch

effectsoftenhavenotbeentakenintoaccountduringprojectdevelopment,ortheyhavebeenseverelymiscalculated.13InScandinavia,promotersoftheresundandGreatBeltlinksatfirsttriedtoignoreordownplayenvironmentalissues,butwereeventuallyforcedbyenvironmentalgroupsandpublicprotesttoacceptsuchissueson the decision making agenda (see Chapter 5). In Germany, high-speed railprojects have been criticized fornotconsideringenvironmentaldisruption.Dams

areroutinelycriticizedforthesamething.However,environmentalproblemsthatare not taken into account during project preparation tend to surface duringconstruction and operations; and such problems often destabilize habitats,communitiesandmegaprojects themselves, if notdealtwithcarefully.Moreover,positiveregionaldevelopmenteffects,typicallymuchtoutedbyprojectpromoterstogain political acceptance for their projects, repeatedly turn out to be non-measurable,insignificantorevennegative(seeChapter6).

In consequence, the cost-benefit analyses, financial analyses and

environmentalandsocialimpactstatementsthatareroutinelycarriedoutaspartof

megaprojectpreparationarecalledintoquestion,criticisedanddenouncedmoreoftenandmoredramaticallythananalysesinanyotherprofessionalfieldweknow.Megaproject development today is not a field of what has been called `honestnumbers.14Itisafieldwhereyouwillseeonegroupofprofessionalscallingthework of another not only `biased and `seriously flawed but a `graveembarrassment to theprofession.15Andthatiswhenthingshavenotyetturnedunfriendly. In more antagonistic situations the words used in the mud-slinging


14/30

MegaprojectsandRisk

(

accompanyingmanymegaprojectsare`deception,`manipulationandeven`liesand`prostitution.16Whetherwelikeitornot,megaprojectdevelopmentiscurrentlya field where little can be trusted, not even some would say especially not numbersproducedbyanalysts.

Finally, project promoters often avoid andviolate establishedpractices ofgood governance, transparency and participation in political and administrativedecision making,either out of ignorance or because theyseesuchpracticesascounterproductivetogettingprojectsstarted.Civilsocietydoesnothavethesamesayinthisarenaofpubliclifeasitdoesinothers;citizensaretypicallykeptatasubstantial distance from megaproject decision making. In some countries thisstateofaffairsmaybeslowlychanging,butsofarmegaprojectsoftencomedrapedinapoliticsofmistrust.Peoplefearthatthepoliticalinequalityinaccesstodecisionmakingprocesseswillleadtoanunequaldistributionofrisks,burdensandbenefitsfromprojects.17Thegeneralpublicisoftenscepticalornegativetowardsprojects;citizensand interest groupsorchestratehostileprotests; andoccasionally secretundergroundgroupsevenencourageorcarryoutdownrightsabotageonprojects,thoughthisisnotmuchtalkedaboutinpublicforfearofincitingotherstosimilarguerrillaactivities.18Scandinavians,wholikeotherpeoplearoundtheworldhaveexperienced the construction of one megaprojects after another during the pastdecade,havecoinedatermtodescribethelackinmegaprojectdecisionmakingofaccustomedtransparencyandinvolvementofcivilsociety:`democracydeficit.Thefactthataspecialtermhascomeintopopularusagetodescribedwhatisgoingoninmegaprojectdecisionmakingisindicativeoftheextenttowhichlargegroupsin

thepopulationseethecurrentstateofaffairsasunsatisfactory.

Risk, democracy and power

Themegaprojectsparadoxconsistsintheironythatmoreandmoremegaprojectsarebuiltdespitethepoorperformancerecordofmanyprojects.Inthisbookwelinktheideaofriskandweidentifythemaincausesofthemegaprojectsparadoxtobeinadequatedeliberationaboutriskandlackofaccountabilityintheprojectdecision-makingprocess.We thenproceed topropose waysout of the paradox.Wewillshowthatintermsofrisks,mostappraisalsofmegaprojectsassume,orpretendtoassume,thatinfrastructurepoliciesandprojectsexistinapredictableNewtonian

Civil society does not have the same say in this arena of public life as it does in

others. Megaprojects often come draped in a politics of mistrust.


15/30


)

worldofcauseandeffectwherethingsgoaccordingtoplan.Inreality,theworldofmegaproject preparation and implementation is highly risky one where thingshappenonlywithacertainprobabilityandrarelyturnoutasoriginallyintended.

SociologistssuchasUlrichBeckandAnthonyGiddenshavearguedthatin

modern society risk has increasingly become central to all aspects of humanaffairs;thatweliveina`risksocietywheredeliberationaboutsocial,economic,political and environmental issues is bound to fail if it does not take risk intoaccount.19Ifthisdiagnosisiscorrectandwewillarguethatformegaprojectsitisthenitisuntenabletocontinuetoactasifriskdoesnotexistortounderestimateriskinafieldascostlyandconsequentialasmegaprojectdevelopment.

The Beck-Giddens approach to risk society is our point of departure for

understandingriskanditsparticularrelevancetomodernsociety.Yetthisapproachdoesnottakeusfarenoughinthedirectionwewanttogo.TheproblemwithBeck,Giddensandrelatedtheoriesisthattheyuseriskmainlyasametaphorformaturemodernity.Wewanttoproceedbeyondthelevelofsymbolandtheorytouseriskas an analytic frame and guide for actual decision making. We will do this bydevelopingasetofideasofhowriskassessmentandriskmanagementmaybeusedasvehiclesforgoverningrisk.20InthewordsofSilvioFuntowiczandJeromeRavetz,wherefactsareuncertain,decision-stakeshighandvaluesindispute,riskassessment must be at the heart of decision making.21 A growing number ofsocietysdecision areasmeet these criteria. Megaprojectdevelopment isoneofthem.

We do not believe risk can be eliminated from risk society. We believe,however, that risk may be acknowledged much more explicitly and managed agreatdeal better,withmore accountability, than is typically thecase today.LikeOrtwin Renn, Thomas Webler and others, we hold that risk assessment andmanagementshould involvecitizens and stakeholders to reflect their experienceand expertise, in addition to including the usual suspects, namely governmentexperts, administrators and politicians.22 We here define stakeholders as keyinstitutional actors, such as NGOs, various levels of government, industrialinterests, scientific and technical expertise and the media. Some of thesestakeholdergroups will claim tobe speaking legitimately onbehalfof the public

goodandsome,butnotallofthem,willbedoingso.Giventhatsuchstakeholdersdo not always adequately represent publics, we recognize the need, on bothdemocraticandpragmaticgrounds,toproperlyinvolvepublicsindecision-making.Suchinvolvementshouldtakeplaceincarefullydesigneddeliberativeprocessesfromthebeginningandthroughoutlarge-scaleprojects. 23LikeRennandWebler,we believe that one should go as far as possible with the participatory and


16/30

MegaprojectsandRisk

*

deliberativeapproachinincludingpublicsandstakeholdersandthattheresultwillbedecisionaboutriskthatarebetterinformedandmoredemocratic.

Wefind,nevertheless,thatdeliberativeapproachestorisk,basedasthey

areoncommunicativerationalityandthegoodwillofparticipants,cantakeusonly

someofthewaytowardsbetterdecisionsandwillfrequentlyfailformegaprojects.24Thisissobecausetheinterestsandpowerrelationsinvolvedinmegaprojectsaretypicallyverystrong,whichiseasytounderstandgiventheenormoussumsofmoneyatstake,themanyjobs,theenvironmentalimpacts,thenationalprestige,andsoon.Communicativeanddeliberativeapproachesworkwellasidealsandevaluativeyardsticksfordecisionmaking,buttheyarequitedefencelessinthefaceofpower.25Andpowerplay,insteadofcommitmenttodeliberativeideals,isoftenwhat characterises megaproject development. In addition to deliberativeprocesses,wealsofocus,therefore,onhowpowerrelationsandoutcomesmaybeinfluencedandbalancedbyreformingtheinstitutionalarrangementsthatformthecontextofmegaprojectdecisionmaking. 26

Basedonthisapproachtorisk,itisanessentialnotionofthebookthatgood

decisionmakingisaquestionnotonlyofbetterandmorerationalinformationandcommunication,butalsoofinstitutionalarrangementsthatpromoteaccountability,andespeciallyaccountabilitytowardsrisk.Weseeaccountabilityasbeingquestionnotjustaboutperiodicelections,butalsoaboutacontinuingdialoguebetweencivilsocietyandpolicymakersandaboutinstitutionsholdingseachotheraccountablethrough appropriate checks and balances. 27 Thus we replace the conventional

decisionistic approach to megaproject development with a more currentinstitutionalistic one centred on the practices and rules that comprise risk andaccountability. 28 We also hold that our approach must be based on actualexperience from concrete projects. The purpose is to ensure a realisticunderstanding of the issues at hand as well as proposals that are practicallydesirableandpossibletoimplement.

A brief overview

Webuildourcaseforanewapproachtomegaprojectdecisionmakingintwomainsteps.Inthefirsthalfofthebook,weidentifytheweaknessesoftheconventionalapproach to megaproject development. By so doing we argue that a differentapproachisneeded.Ourcritiqueoftheconventionalapproachisproactive;fromthecritiqueweteaseouttheproblems,namelyproblemsthatneedtobeembracedbyanalternativeapproach.Inthesecondhalfofthebook,weargueempiricallyand theoretically how the weaknesses of the conventional approach can be


17/30


+

overcomebyemphasizingrisk,institutionalissuesandaccountability.Finally,anexampleforreaderswithapracticalbentisincludedintheappendix,whichshowshow our approach tomegaproject decision making was employed for a specificprojectwithwhichwehavebeeninvolvedasadviserstotheDanishgovernment,namely the proposed Baltic Sea link connecting Germany and Denmark across

FehmarnBelt,oneofthelargestcrossnationalinfrastructureprojectsintheworld.Throughout the book illustrate majorpoints on the basisof in-depth case

studies of three recent megaprojects that form part of the so-called Trans-European Transport Network sponsored by the European Union and nationalgovernments:

1) TheChanneltunnel,alsoknowasthe`Chunnel,betweenFranceandthe

UK, which opened in 1994 and is the longest underwater rail tunnel inEurope;

2) The Great Belt link, opened in 1997-98, connecting East Denmark withcontinentalEurope,andincludingthe longestsuspensionbridgeinEuropeplusthesecondlongestunderwaterrailtunnel;andfinally

3) TheresundlinkbetweenSwedenandDenmark,whichopenedin2000,andwhichconnectstherestofScandinaviawithcontinentalEurope.

Thethreecasestudiesaresupplementedbymaterialfromalargenumberof

othermajorprojects,mainlyfromthefieldoftransportinfrastructure,butalsofrom

otherfieldssuchasinformationtechnology,powerplants,waterprojects,oilandgas extraction projects and aerospace projects. The economics and politics ofbuilding abridgeor an airportare surelydifferenton manypoints from those ofspaceexploration,watermanagement,orprovidingglobalaccessto theInternet.But despite such differences, our data show that there are also importantsimilarities,forinstanceregardingcostoverrunandfinancialrisk,wherewefindaremarkably similar pattern across different project types. We argue that themeasuresofaccountabilitythatarenecessaryfordetectingandcurbingsystematiccost underestimation, benefit overestimation and other risks are quite similaracrossprojects.Thus,eventhoughthemainfocusofthebookisthedevelopment

ofmega transport infrastructureprojects, theapproachdeveloped is relevant forothertypesofmegaprojectaswell.

Our case studies and other data cover both public and private sector

projects. We argue that for megaprojects there is no simple formula for thegovernment-businessdivide.Megaprojectsaresocomplicatedthatbynaturetheyareessentiallyhybrid.Thisisthecaseevenforprojectsthatareconsideredfully


18/30

MegaprojectsandRisk

,

private, for instance the Channel tunnel, because the sheer complexity andpotentialimpactsofamegaprojectdictatedeeppublic-sectorinvolvementformanyissues, for instance regarding safety and environment. Thus public-privatecollaboration is crucial, even for private-sector projects. The question is notwhether such collaboration should takeplacebut how. InChapter 9 and 10 we

addressthisquestionandredrawtheborderlinesofpublicandprivateinvolvementinmegaprojectdevelopmentwithaviewtoimprovinggovernanceofrisk.

Bylinkingtheideaofmegaprojectswiththeideaofriskwehopetobroaden

thescopeoftheriskliteratureandtoattractattentiontothistopic.Asfarasweareaware,noother studydoes this today. Inwriting thebook,wehaveaimedataninterdisciplinary audience of students and scholars in the social and decision-makingscienceswithaninterestinrisk,publicpolicyandplanning,rangingfromsociology and social policy to political science and public policy to publicadministration, management and planning. Policy makers, administrators andplanners are also an important target group for the book, as are consultants,auditors and other practitioners working with megaproject development. Wemaintain that governments and developers who continue to ignore the type ofknowledgeandproposalspresentedheredosoattheirownperil.Megaprojectsareincreasinglybecominghighlypublicandintenselypoliticisedventuresdrawingsubstantial international attention with much potential for generating negativepublicity.

TheThreeGorgesdammentionedaboveisacaseinpoint.Soisthe650km

Myanmar-Thailand natural gas pipeline and maintenance road, built throughpristine natural forests and habitats. Lonely Planet, the worlds leading travelguidebookpublisher,decidedtoprint,upfrontinitsbest-sellingThailandguide,ahighlyvisibleprotestagainstthepipelinewhichcalledtheactionsofboththeThaigovernmentandnamedtransnationalcompanies,suchasAmericanUnocalandFrench Total, a `scam, `shameful and a `fait accompli. 29 Lonely Planetencouragesthereadertojointheprotestsagainsttheprojectandliststhoroughasalwaysthreeaddressesandtelephoneandfaxnumberswherethatispossible.ThisishardlyhowtheTouristAuthorityofThailandwouldhavepreferredtopresentthe country to its visitors, nor is it the type of publicity that transnational

corporationsoptfor,iftheyhaveachoice.Ourpointistheydo:thereisanotherwaytodealwithmegaprojectsandthisbookexplainswhatthisis.

Finally,thoughwedidnotwritethebookwithlayreadersprimarilyinmind,

we hope that individuals, communities, activists, media and the general publicinterestedinandaffectedbymegaprojectdevelopmentwillfindusefulinsightsinthebook,forinstanceregardingthedeceptionsandpowergamestheyarelikelyto


19/30


-.

encounter if they get involved withmegaprojects.Understanding theanatomyofmegaprojects isnecessarytobeaneffectiveplayerin projectdevelopment.And,as mentioned, we see stronger involvement by civil society and stakeholdersgroups inmegaprojects decision makingasaprerequisite fordecisions that arebetterinformedandmoredemocratic.

Theoristsofrisksocietyanddemocracyhaverecentlybeguntocontemplate

the typeofpractical policy and planning needed for dealing with risk in real-lifepublic deliberation and decision making. `In risk society, one study concludes,`publicpolicyrequireslong-termplanningforuncertainty,withinaclearframeworkofprinciplesandevidencetosupportdevolvedandflexibledecisionmaking.This,inturn,requirestheinvolvementofinformedandactivecitizens,enjoyingamature,adult-to-adultrelationshipwithexpertsandwithpoliticians.Ahigh-trustdemocracy:theonlywaytofaceariskyfuture.30Inorderforthisapproachtowork,thetrustin`high-trust democracy must be based on, not feet-in-the-air idealism about themerits of democracy, but hard-nosed considerations about risk and democraticaccountability.Lifewillneverberiskfree,wearehappyto report.Butriskcanbefacedinwaysmuchmoreintelligentthanthosecurrentlyseen.Weofferthisbookasanattemptatfleshingoutinpracticethetypeofdecisionmakinganddemocracycalled for bytheorists ofriskand democracy for aspecific domain of increasingsocial, economic and political importance, namely that of megaprojectdevelopment.


20/30

2 A calamitous history of cost overrun

Inthisandthefollowingchapterswereviewtheexperiencefromalargenumberofmegaprojects, including the Channel tunnel and the links acrossGreat Belt andresund.allthreemultibillion-dollarprojects.Althoughwesubjecttheseprojectstocritical scrutiny, our objective is not to criticise them, even where they haveunderperformed,buttolearnconstructivelyfromexperiencebyidentifyinglessons

thatmayproveusefulinimprovingfuturedecisionsregardingmegaprojects.Giventhelargeamountsofmoneyspentonmajortransportinfrastructureprojects,itisremarkablehowlittledataandresearchareavailablethatwouldhelpanswerthetwobasicquestions:(i)whethersuchprojectshavetheintendedeffects:and(ii)howtheactualviabilityofsuchprojectscomparestoprojectedviability.Therefore,in addition to transport projects, wehave found in pertinent to review dataandresearchfromothertypesofinfrastructureprojectandtocompareexperiencefromthese projectswithexperience from the transportsector. Inthismanner,wewillreviewdatafromseveralhundredlargeprojects.Inthischapterwefocusonthecosts of megaprojects. In Chapters 3 and 4 we consider the demand for, and

viabilityof,suchprojects.

The problem of cost overrun

Cost overruns in major transport infrastructure projects are widespread. Thedifferencebetweenactualandestimatedinvestmentcostisoften50-100percent,andformanyprojectscostoverrunsendupthreateningprojectviability.Afirststepinreducingcostoverrunistoacknowledgethatasubstantialriskofoverrunexistsandcannotbe

A first step in reducing cost overrun is to acknowledge that a substantial risk for

overrun exists and cannot be completely eliminated; but it can be moderated.


21/30

Acalamitoushistoryofcostoverrun

-%

Figure2.1ConstructioncostoverrunsfortheChanneltunnel,GreatBeltlinkandresundlink(constant prices)

completelyeliminated;butitcanbemoderated.Anextstepistoallocatetheriskof

overrun to those best able tomanage it. Both stepswill be considered inwhatfollows.

Amaincauseofoverrunsisalackofrealismininitialcostestimates.Thelength and cost of delays are underestimated, contingencies are set too low,changes in project specifications and designs are not sufficiently taken intoaccount, changes in exchange rates between currencies are underestimated orignored,soisgeologicalrisk,andquantityandpricechangesareundervaluedasare expropriation costs and safety and environmental demands. Many majorprojects also contain a large element of technological innovationwith high risk.

Such risk tends to translate into cost increases,which often are not adequatelyaccountedforininitialcostestimates.Letuslookatcostdevelopmentinanumberofactualprojects.


22/30

MegaprojectsandRisk

-&

Channeltunnel,GreatBeltandresundTheChanneltunnel,alsoknownasthe`Chunnel,isthelongestunderwaterrailtunnelinEurope.Itopenedin1994andconnectsFranceandtheUK.WhentheChannelTunnelTreatywasratifiedbytheFrenchandBritishparliamentsin1987,total investmentcostsforthisprivatelyfinancedprojectwereestimatedat2,600million(1985prices).Uponcompletingtheprojectin1994actualcoststurnedouttobe4,650million(1985prices)resultinginacostoverrunof80percent.Actualfinancingcoststurnedouttobeasteep140percenthigherthanforecast. 1

Figure 2.2 Year by-year development in forecasts of construction costs(constant prices)androadtrafficfortheGreatBeltand resundlinks.

TheGreatBeltlinkopenedin1997-98;the resundlinkin2000Source: Mette K. Skamris and Bent Flyvbjerg, Inaccuracy of TrafficForecastsandCostEstimatesonLargeTransportProjects,Transport Policy,vol.4,no.3,1997,pp141-6

TheGreatBeltlinkopenedtorailtrafficin1997androadtrafficin1998.Itconnects East Denmark with continental Europe and comprises the longestsuspensionbridgeinEuropeplusthesecondlongestunderwaterrailtunnel.Whenthe construction law for Great Belt linkwas ratified by the Danish Parliament in

1987, total investment costs were estimatedat D K K 13.9 billion (1988 prices).Whenallconstructionwasfinishedin1999,costshadincreasedby54percentinrealtermstoD K K21.4billion.2

The resund link between Sweden andDenmarkopened in2000and is

one of the largestcross-national infrastructure projects in the world. It connectsSwedenandNorwaywithcontinentalEurope.Whentheconstructionlawforthe


23/30


-'

resundlinkwasratifiedbytheDanishParliamentin1991,totalinvestmentcostswereestimatedatD K K 11.7billion(1990prices)forthecoast-to-coastprojectand

D K K3.2billionforAccesslinksontheDanishside.3WhentheDanishaccess

linkswerecompletedin1998,actualcostsforthesewereD K K5.4billion(1990

prices),whichtranslatesintoacostoverrunof68percent. 4Whenthecoast-to-coastlinkopenedtwoyearslater,in2000,costsforthisfacilityhadincreasedby26percenttoD K K14.8billion(1990prices).5TheSwedishaccesslinkshavenotyet

beenfinished,butthecompletedpartshaveanoverallcostoverrunof16percent.The largestextra costfor the Swedishaccess linkswillbeanS E K7billionrail

tunnelinMalm, the so-calledCitytunnel.

TableII.iExamples of construction cost overruns in large transport projects Constantprices. For patronage figures, see Chapters 3 and 4

Project Cost overrun ( )

Bostonsartery/tunnelproject 196Humberbridge,UK 175BostonWashingtonNewYorkrail,USA 130GreatBeltrailtunnel,Denmark 110A6MotorwayChapelen-le-Frith/Whaleybypass,UK 100ShinkansenJoetsurailline,Japan 100Washingtonmetro,USA 85Channeltunnel,UK,France 80Karlsruhe-Brettenlightrail,Germany 80

resundaccesslinks,Denmark 70MexicoCitymetroline 60Paris-Auber-Nanterrerailline 60TyneandWearmetro,UK 55GreatBeltlink,Denmark 54resundcoast-to-coastlink 26

Thetunnelwasnotincludedintheoriginalschemeforthe resundlink,butitissaidtobenecessarytolinkMalm effectivelywiththeresundrailconnectionsto

Copenhagen and Denmark. Delays and cost overruns of 36 per cent havehappenedevenbeforeconstructionhasstarted.AnotherextracostontheSwedishsideisthenewten-mileringmotorwayaroundMalm , whichiseffectivelyservingasanaccesslinktotheresundbridgeandwhichwasbuiltabouttenyearsearliertanitwouldhavebeenwithoutthefixedlinktoDenmarkCostdevelopmentfortheGreatBeltandresundprojectsareshowninFigure2.2. 6


24/30

MegaprojectsandRisk

-(

Other transport infrastructure projects

Itisoftenarguedthatnomajorprojectsarealike,and,therefore,thatsuchprojectscannotbecompared.Indeed,theChannel,GreatBeltandresund projectsaredifferentinmanyways.Nevertheless,withrespecttocostdevelopment,thereisastriking similarity between these and other major projects: there is a tendencytowardsasignificantunderestimationofcostsduringprojectappraisal.Thisisalsothe conclusion we draw when the review data from a large number of majortransportinfrastructureprojects,andfromothertypesofprojectaswell.InTableII.icost overruns for the Channel tunnel, Great Belt link and resund links arecomparedtocostoverrunsforanumberofothertransportprojects.AsweshowinChapters 3and4,theproblemwithcostoverrun isexacerbatedbythe fact thatoftenthisproblemcomeshandinhandwithlower-than-estimatedrevenues.The

consequenceisprojectsthatarefinanciallyriskytotheseconddegree.Onlyafewstudiesexistthatrigorouslycompareforecastwithactualcosts

formorethanafewtransportinfrastructureprojects.Wehaveidentifiedfoursuchstudies. The first study was carriedout by the Auditor General ofSweden andcovered fifteen roadand rail projects witha totalvalue of S E K13billion(1994

prices).7Theaveragecapitalcostoverrunfortheeightroadprojectsinthestudywas 86 per cent, ranging from 2 per cent to 182 per cent, while thee averageoverrunforthesevenrailprojectswas17percent,rangingfrom14to+74percent.ItshouldbenotedabouttheSwedishstudy,however,thattwo-thirdsofthe

projectsin thestudywerestillunderconstructionwhenthestudywascarriedout.FinalcostsfortheseprojectsmaythereforeturnouttobehigherthantheexpectedfinalcostsquotedbytheSwedishAuditor-General.

ThesecondstudywascarriedoutbytheUSDepartmentofTransportation

and covered ten US rail transit projects with a total value of US$22-165 million(1987prices).9Forthirteenofthemetros,capitalcostoverrunscouldbeestimated.Sixmetroshadoverrunsabove50percent.Threemetroshadoverrunsinthe20to50percentrange,andtheremainingfourinthe-10to+20percentrange.

ThefourthandfinalstudywascarriedoutatAalborgUniversity,Denmark.Thisisthemostcomprehensiveofthe4studiesandwithasampleof258projectsworth approximately US$90 billion (1995 prices) it is the first study to allowstatisticallysignificantconclusionsregardingcostoverrunintransportinfrastructureprojects.10Theprojecttypesarebridges,tunnels,highways,freeways,high-speedrail, urban railandconventional (inter-urban) rail. The projects are located in20countrieson5continents,includingbothdevelopedanddevelopingnations.The


25/30


-)

projectswerecompletedbetween1927and1998.Olderprojectswereincludedinthesampleinordertotestwhethertheaccuracyofestimatedcostsimprovesovertime,thatiswhetherlearningregardingcostestimationtakesplace.MainfindingsfromtheAalborgstudyare(allhighlysignificantandmostlikelyconservative):

Innineoutoftentransportinfrastructureprojectscostsareunderestimated,resultingincostoverrun;

Forrail,actualcostsareontheaverage45percenthigherthanestimated

costs(standarddeviation,sd=38); Forfixedlinks(tunnelsandbridges),actualcostsareontheaverage34per

centhigherthanestimatedcosts(sd=62); Forroads,actualcostsareontheaverage20percenthigherthanestimated

costs(sd=30); Forall projecttypes,actualcostsare,on theaverage,28percenthigher

thanestimatedcosts(sd=39); Costunderestimationandoverrunexistacross20nationsand5continents;

itappearstobeaglobalphenomenon; Cost underestimation and overrun appear to be more pronounced in

developingnationsthaninNorthAmericaandEurope(dataforrailonly); Costunderestimationandoverrunhavenotdecreasedoverthepastseventy

years.Nolearningseemstotakeplace;

Costunderestimationandoverruncannotbeexplainedbyerrorandseemtobebestexplainedbystrategicmisrepresentation,namelylying,withaviewtogettingprojectsstarted.

Figure 2.3 shows the inaccuracy of cost estimates in the 258 projects

dividedintorail,fixedlinksandroads.Thisandotherstudiesofcostdevelopmentin major transport infrastructure projects reveal the same overall pattern: costoverrunsabove40percentinfixedpricesarecommon,especiallyforrailprojects,andoverrunsabove80percentarenotuncommon.Therefore,costdevelopmentsfortheChunnelandtheGreatBeltand resundlinksarenottheaberrationstheymayappeartobeatfirstsight:theyarecommonplace.

Figure2.4showsaplotofcostoverrunagainstyearofdecisiontobuildfor

the111 projects in the sample for which thesedataare available. Thediagramdoes not seem to indicate an effect from time on cost overrun. Statistical testcorroborate this impression. A test using year of completion instead of year ofdecision,withdata for246projects,givesasimilarresult.Wethereforeconcludethatcostoverrunhasnotdecreasedovertime.Costoverruntodayisin thesameorderofmagnitudeasitwasten,thirtyorseventyyearsago.Iftechniquesandskills for estimating costs and avoiding cost overrun in transport infrastructure


26/30

MegaprojectsandRisk

-*

projects have improved over time, this does not show in the data. No learningseemstotakeplaceinthisimportantandhighlycostlysectorofpublicandprivatedecision making.This seems strangeand invites speculation that the persistentexistenceovertimeandspaceandprojecttypeofsignificantandwidespreadcostoverrunisasignthatanequilibriumhasbeenreached:strongincentivesandweak

disincentivesforcostunderestimationandthus forcostoverrunmayhavetaughtproject promoters what there is to learn, namely that cost underestimation andoverrunpayoff.Ifthisisthecase,costoverrunmustbeexpectedanditmustbeexpectedtobeintentional.

Costoverrunforrail,%(N=58)

Costoverrunforfixedlinks,%(N=33)


27/30


-+

Costoverrunforroads,%(N=167)

Figure 2.3 Inaccuracies of cost estimates for rail, fixed links and roads in 258projects.Thefigureshowsthepercentagedistributionofprojectswithrespecttocostoverrun(constant prices)

Figure2.4Acenturyofcostoverrunin111projects( constant prices)


28/30

MegaprojectsandRisk

-,

Other major projects

In addition to cost data for transport infrastructure projects, we have

examined cost data for several hundred other projects, including power plants,dams, water projects, oil and gas extraction projects, information technologysystems, aerospace projects and weapon systems. 11The data show that othertypesofmajorprojectareatleastas, ifnotmore,prone tocostoverrunsasaremajortransportinfrastructureprojects.

Among the more spectacular examples of cost overruns are the Sydney

Opera House with actual cost approximately fifteen times higher than thoseprojectedandtheConcordesupersonicaeroplanewithtwelvetimeshighercosts.12

The data also indicate that cost overruns for other major projects have neither

increasednor decreasedhistorically and thatoverruns are common inboth firstand third wouldcountries.When the SuezCanalwas completed in1869actualconstructioncostsweretwentytimeshigherthantheearliestestimatedcostsandthreetimeshigherthanthecostestimatefortheyearbeforeconstructionbegan.ThePanamaCanal,whichwascompletedin1914,hadcostoverrunsintherangefrom70percentto200percent. 13

TableII.iiSpectacular projects with spectacular cost overruns


SuezCanal 1,900SydneyOperaHouse 1,400Concordesupersonicaeroplane 1,100PanamaCanal 200BrooklynBridge 100

Sources: PeterHall,`GreatPlanningDisastersRevisited,p.3;RobertSummers, `Cost Estimates as Predictors of Actual Costs: AStatisticalStudyofMilitaryDevelopments,inThomasMarschacketal., eds., Strategy for R&D: Studies in the Microeconomics of Development(Berlin: Springer-Verlag, 1967), p.148; and Mette K. Skamris,conomic Appraisal of Large-Scale Transport InfrastructureInvestment,Ph.Ddissertation(Aalborg:AalborgUniversity,2000).

Insum,thephenomenonofcostoverrunappearstobecharacteristicnotonlyoftransportationprojectsbutofprojectsinotherfieldsaswell(seeTableII.ii).


29/30


%.

The `whatand `thatof cost overrun

AsmentionedinChapter1,theMajorProjectsAssociation ina recentpublicationspeaksof`thecalamitoushistoryofpreviouscostoverrunsofverylargeprojectsinthe public sector.14 With the privately owned, privately financed and privatelyoperatedChanneltunnelinmind,onemustaddthatthereisnoclearindicationthatcostoverrunsformajorprojectsareasingularpublic-sectorphenomenonorthatoverrunswillbeeliminatedbysimplyplacingprojectsintheprivatesector,evenifthismayhelpinducedisciplineandaccountabilityinmajorprojectsasweshallseebelow(seeTableII.iii).

Wherethepatternofcostoverrunisoftenstrikinglysimilarbetweenprojects,the causes of overrun typically differ. For the Channel tunnel, changed safetyrequirementswereamaincauseofoverrun.FortheGreatBeltlink,environmental

concerns and accidents with flooding and a devastating fire made the Budgetballoon.Fortheresundlink,itprovedmorecostlythanestimatedtocarvemajornewtransportinfrastructureintodenselypopulatedCopenhagen,andsoon.What,exactly, causes cost overrun in major infrastructure projects is more difficult topredictthanthefactthatcostoverrunislikelytohauntprojects.

Table II.iii Construction cost development in four privately owned

transport infrastructure projects. The table includes private projects for

which data were available. Constant prices


Channeltunnel,UK,France 80ThirdDartfordCrossing,UK 20SecondSevernCrossing,UK 20PontdeNormandie,France 15

Sources: Mette K. Skamris, `Economic Appraisal of Large-ScaleTransportInfrastructureInvestment.

But knowledge of the latter fact the `thatof cost overrun is the appropriate,necessary and sufficient point of departure for the type of risk analysis andmanagement that we advocate in this book and that is sorely lacking in theplanningofmostmajorinfrastructureprojectstoday.Therefore,atthisstagewearemoreconcernedwiththe`thatofcostoverrunthanwiththe`what.


30/30

MegaprojectsandRisk %-

Conclusion: dont trust cost est imates

Onthebasisoftheevidencepresentedin thischapter,weconcludethatthecostestimates used in public debates, media coverage and decision making fortransport infrastructure development are highly, systematically and significantlydeceptive.Soarethecost-benefitanalysesintowhichcostestimatesareroutinelyfedtocalculatetheviabilityandrankingofprojects.Themisrepresentationofcostsislikelytoleadtothemisallocationofscarceresources,which,inturn,willproducelosersamongthosefinancingandusinginfrastructure,betheytaxpayersorprivateinvestors.

An important policy implication for this highly expensive and highlyconsequential field of public policy is for the public, politicians, administrators,

bankers and media not to trust the cost estimates presented by infrastructurepromotersandforecasters.Anotherimportantimplicationisthatinstitutionalchecksand balances including financial, professional or even criminal penalties forconsistentor foreseeable estimation errors shouldbedeveloped toensure theproductionoflessdeceptivecostestimates.Inlaterchapterswewilseewhatsuchchecksandbalancemightlooklike.

Thatlargecostoverrunsarecommoninmajorprojectsdoesnotmeanthat

examplesofgoodpracticedonotexistregardingcostestimationandmanagement.For transport infrastructureprojects,suchpractice ismostcommonforroads.But

evenforrail,examplesofgoodpracticeexist,forinstancetheconstructionofParisSouthEastandParisAtlantiquehigh-speedraillines(TGV)inFrancewhichhadonly small cost overruns, as had the Toronto Danforth rail extension and theCologne Metro.15 For the average major project, however, there is substantialscope for improvement in cost-estimation procedures and in the institutionalarrangementstocontrolcosts.

megaprojects and risks - english

Documents