India financing highways

Report No. 30363-IN

India

Financing Highways

Energy and Infrastructure Sector UnitSouth Asia Region

Document of the World Bank

© 2004 The International Bank for Reconstruction and Development/ The World Bank1818 H Street, NWWashington, DC 20433

All rights reserved. No part of this publication may be reproduced ortransmitted, in any form or by any means, without permission. Anyperson who does any unauthorized act in relation to this publicationmay be liable to criminal prosecution and civil claims for damages.

First published, 2004

Produced by :

Colorcom Advertising242, Gulmohar Enclave,New Delhi - 110049Tel. : 26521211, 26521212Telefax : 91-11-51841024

The findings, interpretations, and conclusions expressed herein are those of the author(s) anddo not necessarily reflect the views of the Executive Directors of The World Bank or thegovernments they represent.The World Bank does not guarantee the accuracy of the data included in this work. Theboundaries, colors, denominations, and other information shown on any map in this work donot imply any judgement on the part of The World Bank concerning the legal status of anyterritory or the endorsement or acceptance of such boundaries.

Rights and PermissionsThe material in this work is copyrighted. Copying and/ or transmitting portions or all of thiswork without permission may be a violation of applicable law. The International Bank forReconstruction and Development/ The World Bank encourages dissemination of its work andwill normally grant permission promptly.

All queries on rights and licenses, including subsidiary rights, should be addressed to the Officeof the Publisher, The World Bank, 1818 H Street, NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail: [email protected].

GOVERNMENT FISCAL YEARApril 1 – March 31

CURRENCY EQUIVALENTSCurrency Unit = Indian Rupees (Rs.)USD 1 = Rs. 45Rs.1 = USD 0.02221 Lakh (Lk.) = 100,000I Crore (Cr.) = 10,000,000

ABBREVIATIONS AND ACRONYMSADB Asian Development BankAP Andhra PradeshAPRDCL Andhra Pradesh Road Development CorporationBOLT Built Own Lease and TransferBOOT Build Own Operate TransferBOT Build Operate TransferDBFO Design Build Finance and OperateEPC Engineering Procurement & ConstructionEU European UnionFHWA Federal Highway AdministrationGDP Gross Domestic ProductGOI Government of IndiaGQ Golden QuadrilateralHDM-4 Highway Development and Management ModelIPO Initial Public OfferingKRDCL Karnataka Road Development Corporation Ltd.MORTH Ministry of Road Transport and HighwaysMSRDC Maharashtra State Road Development CorporationNH National HighwaysNHAI National Highways Authority of IndiaNHDP National Highway Development ProgrammeO&M Operations & MaintenancePMGSY Pradhan Mantri Gram Sadak YojanaPSP Private Sector ParticipationPWD Public Works DepartmentRBI Reserve Bank of IndiaSH State HighwaysSPV Special Purpose VehicleUNCITRAL United Nations Commission on International Trade Law

UP Uttar Pradesh

Vice President : Praful C. PatelCountry Director : Michael F. CarterSector Manager : Guang Z. ChenTask Team Leader : Piers A. Vickers

www.highwayfinindia.orgwww.worldbank.org/in

ACKNOWLEDGEMENTS

This report is designed to provide information and advice to the Indian Union and StatesGovernments on the principles and practicalities for establishing a sound and sustainable systemof highway financing. The report also has an associated summary policy note, availableseparately.

Much of the data used to prepare the report was obtained through a baseline study conductedduring 2003 by Consulting Engineering Services Ltd and Ernst and Young. This baseline surveyobtained information from various national entities as well as from six states – Andhra Pradesh,Karnataka, Madhya Pradesh, Maharasthtra, West Bengal and Uttar Pradesh. The baseline alsobenefited from a workshop held in Delhi where representatives of both the public and privatesectors working in this field contributed their views on the preliminary findings. Specific materialprovided by others is gratefully acknowledged in footnotes wherever applicable.

The reports have been prepared by a team of Bank staff, including Abha Joshi Ghani, AlokBansal, Aniruddha Patil, Isabel Chatterton, Raghu Kesavan, Simon Thomas, Supriya Sen, andPiers A. Vickers (Task Leader). Sangeeta Anand and Rajesh Singh provided administrativeassistance throughout. The report has been prepared under the general guidance of Guang Z.Chen, Sector Manager for Transport South Asia Region. Peer reviewers were Kenneth Gwilliam,Stephan von Klaudy and Richard Scurfield.

TABLE OF CONTENTS

EXECUTIVE SUMMARY ............................................................................................. i

INTRODUCTION ........................................................................................................ 1

1. PRINCIPLES OF A USER CHARGE REGIME ............................................................. 3

Objectives for Road User Charging .......................................................................... 3The Costs To Be Covered by Road User Charges ...................................................... 4Suitable Tax and Charging Instruments .................................................................... 7Charging Structures – Effectiveness vs. Efficiency Trade-off ........................................ 9

2. ROAD TAXES AND HIGHWAY EXPENDITURES ..................................................... 11

Road Related Taxes and Charges .......................................................................... 11Government Revenue from the Road Sector ........................................................... 13Government Expenditure on the Road Sector ......................................................... 14Comparison of Aggregate Revenue and Actual Expenditure ..................................... 14

3. THE ROAD USER CHARGE REGIME ..................................................................... 17

Road User Charges in India .................................................................................. 17Road User Charges vs. Road Expenditures .............................................................. 18Road User Charges vs. Full Road Costs .................................................................. 19Assessment of the Road User Charging Regime ...................................................... 21

4. FUTURE FUNDING NEEDS .................................................................................. 23

Highway Maintenance ......................................................................................... 23Highway Investment ............................................................................................ 23The Funding Gap ................................................................................................. 23

5. MAJOR FINANCING ISSUES AND CAUSES ........................................................... 25

Major Financing Issues ......................................................................................... 25Underlying Causes ............................................................................................... 25

6. PRINCIPLES OF PRIVATE FINANCE ...................................................................... 29

Rationale for Private Sector Participation in Indian Highways .................................... 29Enabling Environment for PSP .............................................................................. 31The Public/Private Risk Regime ............................................................................. 31Additional Support by the Public Sector ................................................................. 33Additional Support by the Public Sector ................................................................. 34

7. INTERNATIONAL EXPERIENCE WITH PRIVATE FINANCE ........................................ 35

Overview ........................................................................................................... 35The Enabling Environment ................................................................................... 36The Management of Risks .................................................................................... 38Additional Public Support ..................................................................................... 40

8. INDIAN EXPERIENCE WITH PRIVATE FINANCE ...................................................... 42

Policy and Legal Environment ................................................................................ 42Progress with Private Sector Participation at the National Level ................................. 44Progress with Private Sector Participation at the State level ...................................... 49

India l Financing Highways

Financial Engineering on NHAI and State Projects ................................................... 50Indicators of Remaining Constraints to PSP ............................................................. 52Systemic Constraints To PSP In India ..................................................................... 53

9. RECOMMENDATIONS ........................................................................................ 57

The Foundations for Financial Sustainability in the Sector ......................................... 57Options in Financing the Highway Program ............................................................ 60Recommended Changes to Charging Instruments ................................................... 61Additional Reforms to Help Spur Private Finance ..................................................... 64

ANNEXURES ....................................................................................................... 69

Annex 1 Pump Prices For Super Gasoline And Diesel ................................................ 70Annex 2 Excise Duty Rates, 2003 ...........................................................................70Annex 3 Motor Vehicle Tax in Selected States ......................................................... 71Annex 4 Rates of Sales Tax ................................................................................... 71Annex 5 Assumptions behind the Assessment of Road Sector Revenue ...................... 72Annex 6 Revenue Contribution by Road Network, 2001-2002 ................................... 73Annex 7 Revenue - Cost Comparison at Vehicle Level .............................................. 73Annex 8 Vision 2021 Capital Investment ................................................................ 74Annex 9 Forms of Government Support for Road Concessions ................................... 75Annex 10 Risk Framework of NHAI BOT Models .....................................................76Annex 11 Risk Allocation Framework .................................................................... 77Annex 12 Private Sector Concerns with BOT Projects in India ..................................... 79Annex 13 Practical Limits to Extent of Private Involvement ........................................ 81Annex 14 Examples of Inter-State Cooperation on Tax Instruments ............................ 83Annex 15 Measures to Compensate Non-road Users for Levies on Diesel ................... 84

1 Recently another 7,000 km of lower category roads were reclassified as national highways.2 This report mainly addresses the National and State Highways, “the highway sector”3 Vision 2021

EXECUTIVE SUMMARY

Background

1. India has a very large network of lowstandard roads. The National Highways,totaling only 58,000 km1, carry 45% of totaltraffic but are mostly two lane with hightraffic, low service and slow speeds2. Thereis no network of high quality, high capacityhighways or expressways, and, despite 3.5million km of roads, some 40% of India’svi l lages have no al l weather access.Government expenditure on roads issignificant - representing 12% of capital and3% of total expenditure - but roadmaintenance is grossly under-funded withonly one third of needs being met, resultingin road deterioration, high transport costsand loss of accessibility.

2. Road transport is now the primarytransport system in India and it isincreasingly a major constraint to fastereconomic growth and social development.Union and State Governments recognizethe deficiencies in the road network andplan: (i) over Rs.225,000 crore (US$50billion) on highway improvements in theperiod to 20113; and ( i i ) substantialinvestment, through the Pradhan MantriGram Sadak Yojana (PMGSY), to connectvillages. In addition, annual expenditure ofRs. 7,000 crore is essential to maintain the170,000 km of National and StateHighways and further funding is requiredto maintain the urban networks and districtand rural roads. All these expenditures haveto be financed within a current fiscalenvironment of high government deficits,amounting to 9.5% of GDP. There are majorissues as to who wi l l f inance theseexpenditures and how the financing will bestructured.

Approaches to Highway Financing

3. Traditional. In this approach, roads aretreated as public goods and financed fromgeneral revenue with little connectionbetween the costs of road provision and the

taxes or charges paid by road users (thoughfuel is often heavily taxed for generalrevenue), and there is no attempt at directroad pricing.

4. Commercial. With a commercialapproach, roads are treated as capitalassets, commercial accounting is appliedand users are charged, either directly orindirectly, for road use. Road transportremains a source of general revenue, butsector taxes are designed to minimizedistortions to transport patterns or choices.Road finance is increasingly separated fromgeneral government expenditures and roadusers are increasingly involved in decision-making.

5. Indian. In India, the tradit ionalapproach largely persists, although anational and some state fuel cesses havebeen introduced, tolls are increasinglyapplied and substantial private sectorfinancing is being sought. The approach hascontributed to under-funding of roadmaintenance, a distorted vehicle fleet,perverse incentives for traffic allocationbetween road and rail, and substantialeconomic losses. A coherent structure forhighway financing should have high priority;otherwise the distortions and costs to theeconomy will rise as overall expenditure onroads increases.

6. Motivation for Private SectorFinancing. Private sector participation (PSP)in funding is increasingly perceived as theanswer to highway finance and it has somevery substantial potential benefits.

• Bridging the funding gap. Investmentrequirements are high but thepublic sector already faces a largefiscal deficit. Private finance couldsupplement public funding andpos tpone the cos t o f roadinvestment to the taxpayer and/or road user.

l (i)

Executive Summary


• Increased expenditure and revenueefficiency. The private sector hasstronger incentives to operateeff ic ient ly, minimize revenueleakage, adjust resources tochanging situations, and adopt acomprehensive life cycle approachto road investment andmaintenance.

• Unbundling and reallocating risk.Overall costs may be reduced byallocating individual risks to thoseparties best able to control themthrough some informational orother advantage.

7. PSP depends on a sound framework foroverall sector funding as: (i) most PSPprojects require some public funding; (ii) PSPhighways are part of a network and demandfor their use can be substantially reducedby inadequate connecting links: and, (iii)public acceptance of tolls may be partlydetermined by perceptions of the entire roadcharging regime. PSP cannot replace therole of the public sector nor reduce theimportance of a rational, fair and transparentpublic financing system. Internationally, atmost only 5-10% of highway networks havebeen financed by the private sector.Throughout the world, public fundingprevails for the highway sector and this isvery likely to be the case for India too.

Vehicle-Related Taxation and RoadExpenditure in India

8. In FY2002, the road sector accountedfor slightly less than 3% of total governmentexpenditure and contributed about 15.5%of revenue. The total tax revenue from theroad sector in the same year was Rs.500billion while total expenditures were Rs.210billion. Hence, there was a large revenuesurplus from the sector.

9. However, aggregate numbers are notvery useful for assessing the adequacy ofhighway financing, when: (i) taxation on fuelis a major source of general governmentrevenue; (ii) road maintenance is under-funded and the road asset base is being

eroded; and (iii) substantial investment isrequired to remedy neglect and add capacity.More important is the relationship betweenroad user charges, explicit (e.g. the fuelcesses) and implicit, and the levels of roadexpenditure actually required to meet theneeds of the sector.

Road User Charges in India

10. Vehicles, fuel and related inputs aresubject to a wide range of Union and Statetaxes. Vehicles are taxed on initial purchaseand are then subject to annual fees; fuel istaxed by both Union and States; tolls arebecoming more widespread; and road freightand passengers may be taxed. Some taxes,like the cess, are clearly user charges; butthe definition of other taxes is moreambiguous. Total annual road user chargerevenue, in FY2002, was betweenRs.15,000–20,000 crore4, rather less thantotal government road expenditure. In termsof identified expenditures, there weremarked differences between the varioussegments of the road sector:

• Highways. Road user chargescollected (Rs.10,000 crore) exceedexpenditures (Rs.6,300 crore) butthere would be a broad balance, ifthe highways were fully maintained;

• District and rural roads. Road usercharges (Rs.3,400 crore) aresubstantially below expenditure(Rs.6,400 crore)5;

• Urban roads. Road user charges aresubstantially greater than urbanroad expenditure.

11. While the Rs.10,000 crore collected inuser charges from users of the National andState Highways was sufficient to fundhighway expenditure in FY2002, thestructure of charges is deficient in keyrespects.

• The overall charges are too low.Present revenue is suff ic ientbecause highways are being grosslyunder-maintained. The charges arealso insufficient to fund the costs of

4 The low estimate includes only taxes/charges specific to road transport (like the cess); the high estimate also includes the sales/excise taxes in excess ofthe average rates for such taxes.

5 Rural roads generally carry few vehicles but are essential for access. They are almost public goods and could not be financed from economic road usercharges alone.

(ii) l

road accidents, road safety measures,or environmental damage.

• The structure is distorted andeconomically inefficient. User chargeson heavy trucks do not even covershort-run attributable costs, theabsolute minimum for cost recovery;they need to be substantiallyincreased.

• The structure is inequitable. Buses arethe most heavily charged vehicles,paying over four times the chargeson multi-axle trucks but responsiblefor 60% less costs. As buses are usedby low income groups, the chargesappear grossly inequitable, and shouldbe reduced.

• There is no charge for road space.The present charges take no accountof the occupancy of road space. Thisis important on highways (slow trucksdelay all vehicles) and critical in urbanareas. Congestion charges would beeconomically efficient in major urbanareas, especially if the revenue wereused to fund public transport.

• State charges are inconsistent.Different States have adopteddifferent charges, based on revenueopportunities rather than road costs.There are no nationally consistentpricing method and incentives forvehicle owners.

12. The present structure results in: (i) toomuch road use, especially cars in urban areas;(ii) too little revenue for maintaining andimproving the network; (iii) too many heavilydamaging two axle trucks and little incentivefor re-equipping with modern multi-axle truck/trailers; and (iv) an uneconomic distribution oftraffic between road and rail with road usercharges and rail cross-subsidies combining topromote too much road freight and too manyrail passengers.

The Highway Financing Gap

13. India needs massive investment to providea modern high capacity network; but first

priority has to be given to maintenance of theexisting highways. All the studies, in India andelsewhere, indicate that maintenance,effectively implemented, gives the highesteconomic returns: one Rupee spent well onmaintenance can generate seven Rupees innet benefits. The highway network has an assetreplacement value of roughly Rs.240,000 crore(US$ 53 billion); India simply cannot afford tolose such assets.

14. Over the next ten years, about 32,000 kmof National and 25,000 km of State Highwaysneed to be widened, at a cost of someRs.1,700 billion, and highway maintenance willrequire over Rs.950 billion. This is aconservative estimate and takes no accountof the need to build an expressway system, ata cost of several million US dollars per km,along new alignments on the most heavilytrafficked corridors. The present system of roaduser charges will only generate over the nextten years about Rs.1,600 billion6, leaving a totalfunding gap of some Rs.1,050 billion - or theannual equivalent of more than US$2 billion.

Private Highway Finance: Present Status

15. Government of India (GOI). Less than20% of the National Highway DevelopmentPlan (NHDP) projects are being funded throughprivate sector participation (PSP) which atpresent consists of the following.

• Build Operate Transfer (BOT)concessions. National HighwaysAuthority of India (NHAI) has awardedtolled concessions totaling Rs.3,400crore on highly trafficked road sectionswith relatively low traffic/financialrisks.

• Annuity concessions. NHAI hasawarded 467 km (about 8% percentof the NHDP) of annuity concessionsamounting to Rs.3,000 crore forprojects with generally higher trafficrisks.

• Special Purpose Vehicles (SPVs).The Moradabad Bypass wasconstructed through a SPV andseveral other projects are now

6 Road user charge defined here as (i) Central level – cess on fuel and road tolls (ii) State level – registration fees, cess on fuel, tolls, permits and licenses,fines and penalties, taxes levied specifically on passenger and goods vehicles and the excess percentage over the average state sales tax rate for allcommodities.

l (iii)

Executive Summary


under implementation, including the JaipurBypass Phase II and some portconnectivity road projects.

16. Based on its revenues (Central Road Fundand increasingly toll revenue) and an impliedGOI guarantee, NHAI has itself raisedsubstantial funds from the domestic bondmarket. To date, NHAI has raised Rs 656 Crorefrom two bond issues in 2000-01, Rs. 804 Crorethrough one issue in 2001-02 and Rs. 5,593Crore in 2003.

17. GOI has provided the following fiscalincentives for PSP: (i) full and partial taxholidays for ten years; (ii) tax exemptionsto Infrastructure Capital Funds orInfrastructure Capital Companies providinglong-term finance for infrastructure; (iii)reduced import/excise duty on constructioninputs; ( iv ) reduced stamp duty ondocuments/agreements; and (v) reducedState sales tax on construction inputs.

18. State Governments. Some States aremaking real progress in attracting PSP inthe highway sector. Madhya Pradesh, forexample, has entered into a number of‘maintain and transfer’ concessions. AndhraPradesh, Gujarat, Maharashtra, MadhyaPradesh, Rajasthan and Tamil Nadu haveall signed BOT concessions. But, the majorityof States have little or no PSP experienceand overall progress is very limited. No Statehas a defined policy for the selection andprioritization of PSP projects and, in general,there is no systematic procedure for projectidentification and procurement. RoadDevelopment Corporations (RDCs) havebeen created in Andhra Pradesh, Gujarat,Karnataka, Kerala, Maharashtra, and TamilNadu, and Uttar Pradesh is establishing aState Highways Authority. These agenciesare expected to attract private funds throughcorporate borrowing or leveraging privateproject finance against public grants. Thecorporations have clearer objectives, moreflexible staffing, and more streamlineddecision making processes than the typicalPublic Works Department (PWD). Thisshould improve the quality and timelinessof PSP planning and procurement.However, states need to guard againstcreating hidden contingent liabilities throughthese RDCs without consideration as to

strengthening the revenue stream forrepayment of their debts.

19. L e g i s l a t i v e a n d R e g u l a t o r yEnvironment. GOI and some States havealready made legis lat ive changes tofacilitate PSP generally and for road sectorPSP development. In particular: (i) GOI hasamended the National Highways Act (1956)to permit BOT projects and allow tolls onboth public and private funded roads and(ii) some States have amended the IndianToll Act (1851) to allow the private sectorto levy tolls on State roads and bridges.

Constraints to PSP Development

20. Indicators of constraints. While therehas been progress in creating a betterenabling environment, PSP financingremains limited, PSP projects have taken along time to prepare and award, severalprojects have had to be restructured, andthe “value for money” of Annuity PSPs (andthe additional GOI off-balance sheetliabilities) has been questioned.

21. Absence of Reliable Information. Inmaking PSP decisions, potential investorsmust forecast their cash-flows. In turn thisrequires rel iable data on, inter al ia,construction and maintenance costs, trafficflows, construction history of existing roads/structures in the proposed concessionaire.Yet, reliable information is generally notavailable and investors have to rely on theirown invest igat ions which, howeverexhaustive, provide only limited information.Therefore, investors probably have toinclude a very substantial risk premium intheir bids.

22. Low Willingness to Pay. Indian roadusers’ willingness to pay tolls appears low.Even though toll levels are among the lowestin the world and represent a fraction of theexpected vehicle operating cost savings,their affordability is lower than in many othercountries. Further, as the sector is heavilytaxed, users may be reluctant to pay furthercharges in the absence of substantivedifferences in service quality and greatertransparency in the overall use of taxesraised from the sector.

23. Dispute Resolution. While subordinatelevels of dispute resolution are available,

(iv) l

there remains a high probability that seriousdisputes will end in arbitration, a lengthyand costly process. A swifter, more certaindispute resolution mechanism could reducethe perceived risks for private investors.

24. Additional Constraints at the StateLevel. State governments face even greaterconstraints than GOI: typically lowercreditworthiness and lower traffic on theirhighways. Many states have yet to put thebasic building blocks in place, including notonly the enabling framework but alsoprocurement and contracting capacity anda more assured revenue stream to meetpublic sector contributions. Considering thecurrent status of the institutional andregulatory framework for PSP, the Statesmay need to be realistic in their expectationsfor PSP and probably focus on “rehabilitate,operate and maintain” concessions,bypasses and/or bridges as the most viablePSP projects for the moment.

Foundations for Sustainable HighwayFinancing

25. Improving the Information Flow. Itis very difficult to see how rational decisionsare made with the present level/quality ofinformation on highways and highwayfinance. Comprehensive databases ontraff ic, tender rates, highway assetcondition, etc. would result in betterplanning and more accurate tendering,giving benefits to all parties – the publicagency, the private investor and the roaduser. Many countries, both developed anddeveloping, have started to treat roads as amajor business and moved to commercialaccounting for road agencies7. Suchaccounting clarifies financial flows anddemonstrates whether the asset base (thehighways) is being increased, maintainedor eroded. At the present time, road agencybudgets seem more designed to obfuscatethan reveal. High quality, timely informationis essential for better decision-making andfor creating financial transparency. If NHAIand/or state Road DevelopmentCorporations wish to reduce the cost of theirborrowing, facilitate PSP and meet growingpublic demand for clearer accountability,

adopting a modern accounting system thatshows the economic reality of their businesswill be a critical building block.

26. Establishing National Coordination.There would be many advantages if thepolicies toward road user charges andhighway finance in general could becoordinated between the State and Uniongovernments and between the Statesgovernments. This would ensure thatconsistent pricing signals are given to thetransport industry and an efficient fleetconfiguration developed. A Strategic RoadsAuthority might be the ultimate embodimentof such an approach; it could receive thecess and other designated road user chargesand channel this funding in such a way asto develop a coordinated approach acrossall states. However, at this time, this maybe too radical and a more acceptableapproach may be the creat ion andmaintenance of a Council of TransportMinisters. This would be similar in conceptto the European Conference of Ministers ofTransport which has demonstrated thatthere can be considerable benefit in jointaction to resolve common problems withoutcompromising sovereignty.

27. Generating User Acceptance forIncreased Road User Charges. It is almostinevitable that road users, whether presentor future, will pay for the costs of thehighway network, irrespective of whetherthe investments are made by the public orprivate sectors. Road users already pay hightaxes and many may feel that additionalcharges are simply exploiting an easy targetthat is already overtaxed. Other countriesface the same issues, and many are turningto an approach which combines: (i) greatertransparency in highway financing andgovernance; (ii) greater accountability to theroad users who are financing the network;and (iii) a more business oriented approachto the planning and implementation ofhighway investment and maintenance.

28. Central to the approach is often a RoadFund managed by a Roads Board, outsidethe traditional government structure. TheBoard advises on the level of charges,receives the user charge revenues

7 The physical assets embodied in the highway network dwarf those in perhaps any commercial enterprise in India.

l (v)

Executive Summary


automatically and directly - without therevenue being routed through the Ministryof Finance and Consolidated Fund - andthen buys road maintenance andconstruction from road agencies on the basisof approved work programs. Transparencyis core to the approach; regular financialand technical audits of both the Roads Boardand road agencies are statutorily requiredand a l l aud i t s a re pub l i shed .Representatives of road users and otherprivate stakeholders are represented, andincreasingly these representatives providenot only a majority of the Board but alsothe chairperson. The Ministries of Financeand Works /Roads re ta in in f luence(changes in the fuel levy or other usercharges have to be approved by theMinistry of Finance) but the process iss t rengthened by t ransparency andaccountability introduced by bringing theeffective consumers into the decision-making and monitoring process.

Reform to the User Charge Regime

29. Some of the funding gap may be metby borrowing or private sector finance butthere is still a pressing need to restructureuser charges to remove the economicdistortions and raise revenue for full highwaymaintenance, while continuing (or perhapsincreasing) highway investment. GOI mayconsider increasing the fuel cess to Rs.3/liter,the average level in other developingcountries. States may consider dedicatedfuel levies of perhaps Rs.0.5 – 1.0/liter. Theadditional revenue should be dedicated tohighway maintenance. Adding Rs.2.0/literin higher cesses would initially increase thevariable costs of trucks by 5 – 6%, and theirtotal costs by 3 – 4%. Over time, however,trucking costs would be reduced by theensuing better highway maintenance. Whilethe current high international oil pricecomplicate this action, it does not rule itout.

30. More wide-reaching reforms should bebased on a comprehensive study of roadcosts. The study should establish the mostpractical and economically efficient meansto relate user charges to road costs. Suchmeans might include: (i) imposing a higher

cess on diesel; (ii) restructuring annual fees,especially for trucks; (iii) reducing usercharges on buses; (iv) extending tolling atlow rates but to more of the highwaynetwork; (v) piloting congestion charging;and (vi) assessing the utility of truck weight/distance charging. Road user cost and chargestudies should be repeated periodically andmade public.

Strengthening PSP for HighwayDevelopment

31. Justification for PSP. Many expectprivate finance to meet much of the futurehighway investment; but there has been littleanalysis of what modes of PSP would resultin better-value highways. It is recommendedthat a special Comptroller and AuditorGeneral efficiency/performance audit of the“value for money” from the various PSPmodels be undertaken, pr ior to theconcessioning of the next 10,000 km ofnational highways. Individual projects needdue diligence by the client as well as theinvestor, to ensure that value is obtainedand that consumers are protected. Severalgovernments have introduced detailedPublic Sector Comparator processes; withoutsuch process, a higher probability ofunfavorable outcomes in PSP can beexpected.

32. Project vs. Corporate Approach toPSP. The progress made under the PSPapproach has been achieved at a relativelyhigh risk and cost to the Government. Withthe exception of links having captive traffic,for which the project financing approach issuited, it is worth exploring whethercorporate f inancing, based on thehypothecated cess and uniform tolls, wouldattract the required funds and minimize thehigh traffic risk premium normally chargedunder PSP.

33. With a securitized revenue stream fromthe cess and completed toll roads, NHAIcould raise further debt or enter into nontoll based Design Build Finance and Operate(DBFO) agreements without GOI guarantee8.Alternatively, the private sector could also besuccessfully utilized through maintenanceconcessions or long-term performance basedmaintenance contracts.

8 BOTs could still be used for links with high traffic and limited traffic risk, such as major bridges

(vi) l

34. This corporate approach would: (i)diversify the traffic risk across the wholenetwork; (ii) allow cross-subsidizationbetween l inks; ( i i i ) improve theunderstandability of the toll system for roadusers; (iv) give economies of scale; (v)ensure uniformity of service standards; (vi)allow for the incremental introduction oftolls; and (vii) give greater incentives tomanage the whole network efficiently tominimize costs and maximize revenues. GOIcould provide concessional funding fornecessary but financially unviable projects.NHAI would have several financing options,including (i) issue non GOI guaranteed debton the strength of its future anticipatedrevenues; (ii) enter into non toll based DBFOcontracts, using annuity, lane availability oractive management payments; or (i i i)develop hybrid f inancing using acombination of annuity and incrementalshadow toll payments where only sometraffic risk is passed to the private sector.

35. The Regulatory Framework. If DBFOor toll based concessions are to be a majorpart of the highway sector, GOI shouldestablish a regulator, similar to the Telecom,Insurance and Energy sectors. At the Statelevel, unless the GOI regulator can be used,regulation through contract by a disputeresolution board may be appropriate.

36. Additional State Actions. GOI and afew states have established the broadinstitutional and legal frameworks for PSP;but most States continue within the PWDapproach. They need to make substantialchanges, including (i.) reform or thereplacement of the traditional PWD, (ii)adoption of a corporate approach with RoadDevelopment Corporations/HighwaysAuthorities and dedicated road funds withhypothecated revenues from road usercharges; (ii) adoption of a PSP frameworkand enactment of accompanying laws.

l (vii)

Executive Summary


The Importance of Roads and RoadTransport

1. While the railways remain important forsome bulk commodities and in somepassenger markets, India is increasinglydependent upon road transport. Rail trafficcontinues to grow, but its share of freightand passengers has been falling for manyyears. The growth in road transport has beenaccelerating; during the 1990’s, the nationalvehicle fleet grew from 21.3 million to 48.4million. Faster economic growth, especiallyin non-traditional sectors, and higherpersonal incomes will undoubtedly continuethe growth in demand on the road network.However, unless major reforms as well asinvestment are made, India’s roadinfrastructure will be an impediment toeconomic growth and social development.The Indian Tenth National Plan (2002-2007),projects a GDP growth rate of 8% perannum and an industrial growth of 10% perannum and identified transport infrastructureas a major constraint on accelerated growth.

2 India has 3.5 million km of roads which,by international comparisons, provides arelatively dense network. The major issuesin the sector are not primarily the length ofthe network but its low capacity and poorquality.

• During the 1990’s, the nationalhighway network expanded from33,700 km to 58,100 km9 and,though it constitutes only 2% of thenetwork, it carries about 45% of allroad traffic. Most of the network isstill two lane, providing low servicestandards and slow vehicle speeds.Many NH have simply be reclassifiedfrom lower category roads without anyupgrading.

At the other extreme, about 40% ofvillages are not connected by all-weather roads and have thus limitedaccess to economic and socialinfrastructure and opportunities.

INTRODUCTION

• Road maintenance throughout thenetwork is dismal, contributing to bothpoor pavement condition and the lossof all-weather accessibility.

3. Therefore, India combines both the needto increase very substantially the maintenanceof a very large network and the need toprovide a high quality highway system,sufficient to support the development of arapidly developing economy.

Highway Sector Financing Issues

4. Government expenditure on roads issignificant, presuming 12% of capital and 3%of total expenditure; but road maintenance isgrossly under-funded with only one third ofneeds being met. The Union Government (GOI)recognizes the deficiencies in the roadnetwork. The Tenth National Plan has assigneda high priority to the National HighwayDevelopment Plan (NHDP) for the constructionof a Golden Quadrilateral of high capacity, highquality highways, linking the four major cities,as well as similar highways along North-Southand East-West corridors. Very large investmentsare also envisaged on State highways. Thecapital funding needs are immense:

• over Rs.225,000 crore (US$50 billion)on highway improvements in theperiod to 2011; and10

• substantial investment (about Rs.70,000 crore or US$15.6 billion),through the Pradhan Mantri GramSadak Yojana (PMGSY), to connectvillages

5. In addition, annual expenditure of aboutRs.7,000 crore is essential to maintain the170,000 km of National and State Highwaysand further funding is required to maintain theurban networks and district and rural roads.All these expenditures have to be financedwithin a constrained fiscal environment inwhich the combined GOI and StateGovernment deficits total about 9.5% of GDP.There are major issues as to who will financethese expenditures and how the financing willbe structured.

9 Recently a further 7,000 km have been reclassified.10 Vision 2021

Introduction

l 1


6. It is not only the level of highway fundingwhich is important but also the means by whichit is financed. The financing arrangements forthe highway sector have significant implicationsfor overall government expenditure, the roleof private finance as well as having majorimpacts on the efficiency of the transport sectorand thus indirectly for the efficiency of the entireeconomy.

Approaches to Highway Financing

7. The management and financing of roadsis not a new issue for governments. The RomanEmperors invested heavily in road constructionand maintenance throughout Europe and theMiddle East. However, with the growingtransport dominance of the motor vehicle,roads and highway finance has assumed majorimportance and a number of approaches havebeen adopted.

8. Traditional. In this approach roads aretreated much like public goods and financedfrom general government revenue. There islittle connection between the costs of roadprovision and the taxes or charges paid by roadusers (though fuel is often heavily taxed forgeneral revenue purposed), and no attemptat direct road pricing.

9. Commercial. In the commercialapproach, governments deal with roads as abusiness sector. Roads are treated as capitalassets, commercial accounting is applied andusers are charged, either directly or indirectly,for their use of the roads. Road transportremains a source of general revenue, but taxesare designed to minimize distortions totransport patterns or choices. In some countries,road finance is being separated from generalgovernment expenditures and road users areincreasingly involved in decision-making.

10. Indian. The traditional approach largelypersists in India, although a national and somestate fuel cesses have been introduced, tollsare increasingly applied and substantial privatesector financing is being sought. India may beearly in a transitional stage between thetraditional and commercial approaches. Yet,the present structure of financing contributesto the under-funding of road maintenance, adistorted vehicle fleet, perverse incentives fortraffic allocation between road and rail, and

substantial economic losses. A coherentstructure for highway financing should havehigh priority; otherwise the distortions and coststo the economy will rise as overall expenditureon roads increases.

The Purpose of the Report

11. This report is designed to provideinformation and advice to the Indian Unionand States Governments on the principles andpracticalities for establishing a sound andsustainable system of highway financing.

• The report reviews the economicprinciples for establishing efficient andequitable road user charges (roadpricing), and examines the potentialmechanisms for charging road users.Present road taxation in India isassessed in the light of theseconsideration and the levels ofhighway funding required to meetgovernment objectives.

• The report reviews the potentialcontribution of private sector financeto the sector and assesses the presentuse of private finance and thealternative possibilities for utilizing theprivate sector in the financing andmanagement of the network.

• The report also examines the needfor an agenda of sector reform whichaddresses both the financial andinstitutional frameworks needed toachieve network sustainability andpublic acceptance of higher usercharges.

12. The report is specifically concerned withthe main highway network (defined as the170,000 km of National and State Highways)which carry the great majority of vehicle-km.There is also a very large network of rural roadswhich carry little motorized traffic but whichprovides basic access for the rural populationand facilitates the administration of the country.These rural roads are crucial to the socialinfrastructure of the country but their financingraises issues outside the scope of this report.These roads generate major benefits but, inview of their low traffic levels, it would beinconceivable to finance their construction andmaintenance from road users alone.

2 l

Objectives for Road User Charging

1. Roads have often been treated as publicgoods, financed from general taxation ratherthan through cost-related charges. In thischapter, the objectives of road user chargesare considered and their implications for thelevels and structure of taxes which aregenerally used as proxies for direct roadcharges11.

2. Efficient allocation of resourcesbetween sectors. Economic efficiencyrequires that the user of resources pays themarginal social costs associated with the useof those resources. If the user is charged lessthan these costs, then the cost of resourcesused will be greater than the benefitsgenerated and the resources could be betterused elsewhere. Conversely, if the user ischarged more than the marginal costs, thendemand will be less than optimum and overallbenefits could be increased by increasedresource use. The use of the resources shouldonly be free in the case of pure public goodsfor which the use by one consumer has noimpact on the availability of the resource forother consumers.

3. For roads and road transport, this meansthat no category of vehicle should pay lessthan the sum of the following.

• The economic cost of the fuel andthe other resources consumed inmaking the trip. These may betermed the private marginal costs ofusing the road network.

• The marginal road maintenance cost:additional traffic, especially heavycommercial traffic, increases roaddeterioration, reduces the pavementlife and increases the cost of roadmaintenance and renewal.

• The marginal environmental cost:increased traffic raises the levels ofvehicle emissions, noise pollution,etc. These costs are not borne by theroad user but by society, mainly those

1. PRINCIPLES OF A USER CHARGE REGIME

people living and working close tothe roads.

• The marginal congestion cost imposedon other vehicles: as vehicle flowsincrease, vehicle speeds decline. Theindividual road user considers only his/her personal time and cost; but theiruse of the road may well increasethe travel time and costs of all usersof that road.

4. There are recurrent costs associated withroad provision and maintenance that are notrelated to use and on which the level ofvehicle flow has no impact: weather and timerelated road maintenance for example. Suchcosts should be financed by the means whichcauses the least economic and equitydistortion.

5. Whether road users should also becharged the capital costs of networkexpansion/enhancement, in response to trafficgrowth and heavier vehicles, what might betermed the long-run marginal cost of thenetwork, raises further issues and is discussedbelow.

6. Efficient use of resources within theroad sector. Another important efficiencydimension in structuring a charging system isto avoid significant distortions within the sector.In the transport sector, it is important to avoidthree distortions.

• Distortion between vehicle classesand their users: charges on differentcategories of vehicle shouldappropriately reflect the differencesin the costs that they impose on thesystem.

• Distortion between modes:inappropriate charging structures ondifferent transport modes, whichcompete closely for the same typesof passenger and freight traffic, canhave a marked negative impact ontraffic allocation and economicefficiency. This is a significantdistortion in India.

11 For a more detailed exposition of this issue see Efficient Transport Taxes and Charges, ECMT, 2000, also available at http://www/oecd.org/cem/

Principles of A User Charge Regime

l 3


• Distortion between locations: this canoccur, if the charging/financingstructure in some states is significantlydifferent to those in others. This maydistort decisions on the location ofeconomic activity or vehicleregistration; but its impact is probablylower than for the first two distortions.

7. Equity. There are several dimensions ofequity that are or may be relevant to thestructuring of a road user charging system.

• Horizontal equity: vehicles within thesame category, imposing the samecosts on society, should pay the samelevel of charges. Fuel tax is equitable,in this regard, as the payment isgenerally proportionate to road use.Different annual vehicle charges,depending upon the place of licensingas in India, may not be.

• Vertical equity: charges paid bydifferent vehicle categories shouldvary in proportion to the costs thatthe categories impose. Fuel tax is notequitable; the increase in fuelconsumed, as vehicle weightincreases, is not proportional to theincrease in road damage imposed.Insofar as heavier vehicles use diesel,a rough degree of vertical equity maybe achieved by imposing relativelyheavier taxes on diesel than ongasoline.

• Distributional equity: this is normallyinterpreted as requiring charges/taxesto be progressive, with higher incomeusers paying higher charges. Thiswould suggest higher taxes ongasoline which, in low and mediumincome countries, is used in the carsowned by the relatively rich. Equityconsiderations may be particularlyimportant to finance those costswhich are not directly attributable toindividual vehicles. Distributionalequity, however, can also beinterpreted as requiring road users tofund all the cost associated with theprovision of roads.

8. Hence, the implications for policy, whenconsidering equity, are complex.

9. Environment. Particularly in urban areas,road transport can be a significant source ofair pollution. Absolute tax levels, anddifferential taxes on specific fuels, can beimportant instruments in supporting policies toreduce total traffic and encourage the use ofless polluting modes of transport, transporttechnologies and/or fuels, such as CompressedNatural gas (CNG).

The Costs To Be Covered by Road UserCharges

10. There is a broad consensus regarding thecosts that should be covered by road usercharges with the exception of whetherinvestment in new or improved roads shouldbe financed exclusively by present road users.

11. Full road maintenance costs. Forefficiency reasons, as discussed above, allvehicles should be required to meet the fullcosts of road maintenance which areattributable to their use of the road network.Such maintenance includes not only the day-to-day routine maintenance (e.g. repairingpotholes) but also the periodic resealing andstrengthening of pavements. Somemaintenance costs are, however, not directlyattributable to vehicle use but are caused bynon traffic related degradation. However, it isnow generally agreed that vehicles, as a whole,should meet the full costs of road maintenance,assuming that all the roads make economicsense.

12. In principle, charges should bedisaggregated by road type or even roadsection but, in practice, this is not possible.The road sections and types of primary roadjointly provide a transport system: it is areasonable approximation to make costrecovery at the level of the aggregate primarynetwork. When considering the costs to berecovered, it is important to remember that itshould be the costs which are imposed on theroads, and not the expenditures on roadmaintenance as these may be inadequate toremedy the level of wear and tear and thequality of the roads may thus be deteriorating.

13. There are some costs of road maintenancewhich are not variable with use and cannotbe strictly attributed to specific vehiclecategories, but costs nevertheless which needto be financed. Two broad approaches might

4 l

be adopted to finance these non-attributablecosts.

(a) Equity distribution approach: Higherincome groups, primarily car owners,should finance the costs throughhigher charges (higher charges ongasoline) as they benefit from theroad network and can afford to payhigher charges.

(b) Economic pricing approach: Fixedcosts should be financed by chargeswhich impose the least distortion onthe use of the road network. Suchcharges could be set by:

• An annual vehicle license fee:once paid, there would be noimpact on the individuals’decisions as to whether to usethe network, the cost of themarginal trip would not bechanged.

• Charges established throughRamsey pricing principles: tominimize the impact on total use,the additional charges necessaryto cover the fixed costs shouldbe set in inverse proportion to thedemand elasticity. Higher chargeswould thus be established forthose vehicle categories with thelowest travel demand elasticity.

14. System administration costs. The costsinvolved in managing road use (traffic police,traffic signaling, etc), in collecting the varioususer charges and in enforcing their payment,should also be met by users. Whereadministrative services (such as licensing,emissions testing, etc) impose costs, these costsat least should be recovered as a minimumdirectly from the users concerned throughvehicle related fees.12

15. Environmental and other externalitycosts. In principle, the monetary cost ofenvironmental impacts should also be includedin the costs which should be recovered fromusers. These externalities include global and

local air pollution, and road accidents. As faras environmental externalities are concerned,the health impact of local air pollution is usuallyconsidered to be the most significant. Thesecan be roughly quantified, using dose/responserelationships and then monetarized usingstated preference evaluation methods forvalues of life, lost output and medical costsfor morbidity.13

16. Where the fuel tax includes an elementfor environmental costs, the revenues whichthis generates should, in principle, bededicated to compensating those who sufferfrom the impacts. In practice, because nodirect compensation mechanisms areavailable, the second best would be to devotethe revenue to reductions in the level ofenvironmental pollution.

17. As far as accidents are concerned, if therewere a well functioning insurance andcompensation system in existence, it wouldbe reasonable to assume that accident costswere being fully covered. However, wheremedical costs are not paid for by the partiesto accidents, and where those are costs borneby the state or by the injured parties, theexcess of total accident costs over insurancepayments should, in principle, also berecovered from road users. These revenuesshould be transferred to the parties bearingthe costs. In practice, these calculations arevery difficult to make, so it is more a matterof political judgment than scientific calculationas to what sums should be transferred.

18. Congestion costs. Road congestionpricing is now being given much widerconsideration; Singapore has had congestionpricing for many years, and London introduceda central area congestion charge in 2003. Inmany countries, congestion pricing is onlyrelevant to urban and suburban areas.However, congestion is also a phenomenonwhich can be experienced on inter-urbanroads. In the UK and USA, for example, severalmotorways now experience congestion. Moreparticularly, in India, the low capacity of theinter-urban highway network means that

12 These charges need not necessarily be restricted to covering merely the administrative costs if they happen to be an effective and efficient instrumentfor allocating the fixed costs between users in a non-distorting way, or to compensate for defects in the precision of charges for the costs that are variablewith use. For example, in the absence of any better way of ensuring that heavy goods vehicles pay their proper share of road maintenance costs, thelicensing duty structure may be set so that, on average, heavy goods vehicles do pay adequately.

13 Guidance on how to do this, and what information is available to assist it are to be found in the World Bank draft “Manual on Air Pollution from MobileSources” to be published later this year.


l 5


14 Efficient Transport Taxes and Charges, OECD, 2000

vehicle speeds are low, service standards poor,and additional vehicles will reduce thesespeeds/standards even further. Much of India’smain road network suffers from a level ofcongestion, and slow moving vehicles imposesignificant costs on other road users and shouldbe charged for such costs.

19. Capital investment costs. Thetreatment of investment costs for new orimproved roads is theoretically more difficult.It is agreed that road users should not becharged for the vast investment that hasalready been made in the road network. It isa more a question of whether road users shouldpay for the investment which is now beingmade in expanding and improving thenetwork; i.e. whether road users should paysimply the short-run marginal costs or a longerrun marginal costs including the capacityexpansion cost.

20. A normal business, expanding too rapidlyto finance capital expansion from revenues,resorts to borrowing. The annual capitalcharge may then be set to service the debton the capital investment. Where the roadsystem is well established, and its size is notgrowing rapidly, current year capitalexpenditure and the appropriate servicingcharge for capital may be approximatelyequal. That is the presumption made in roadcost accounting in some of the industrializedcountries such as the United Kingdom.

21. In periods of very rapid growth of thecapital in the network, as envisaged in the

next decade in India, the annual investmentcosts are likely to exceed the “correct” capitalcharge. Trying to recoup these capital costsfrom current revenue is likely to inhibit thedesired rate of investment as well as imposesubstantial costs on road users. Furthermore,while the efficiency objective requires that allcategories of users pay at least the marginalsocial costs of their use of roads, it does notnecessarily require that the full current yearcosts of investment expenditures be recoveredfrom current users; this would put the burdenof a long-term strategy excessively andunnecessarily on the current generation.Hence, in accordance with normal commercialprinciples, the annual servicing charge on thecapital employed should be recovered fromusers, and not the current year’s capitalexpenditure. However, if the political decisionis made to raise capital finance from vehiclerelated charges, then this revenue should, asfar as possible, be raised from vehicle relatedcharges (based on attributable costs) with azero marginal tax impact.

22. A recent European Conference ofMinisters of Transport report analyzed thiscomplex issue14. Its conclusion was that 100%coverage of infrastructure expenditures bytransport user charges alone is not anappropriate basis for ensuring efficiency. Inthe rail sector, increasing returns to scale meanthat marginal social costs will be belowaverage costs and transfers from generaltaxation will be required to cover total costs.In contrast, in the road sector, marginal social

Figure 1.1: Tax/Charging Instruments Applied to the Road Sector

6 l

costs may vary greatly depending on the levelof congestion and other externalities. Hence,marginal social cost pricing in the road sectormay result in surplus revenues in some urbanareas (of the order of 150%) but underrecovery in rural areas.

Suitable Tax and Charging Instruments

23. A range of instruments has been usedinternationally to tax/charge users for road-related costs. An important definingcharacteristic of these charges is their proximityto the point of use – the “directness” of thetax. The range of instruments, their prevalenceinternationally, and a rough categorization bydirectness is shown in Figure 1.1. In general,for practical and political reasons, mostcountries still use relatively indirect taxinstruments; however, this is changing astechnology has developed and public pressurehas grown to link road charges more directlywith use.

24. As fuel taxes, annual license fees, andtolls are likely to form the backbone of therevenue stream for the highway network inthe future, they are now explored in moredetail. In addition, a brief description is alsogiven of some new user charging initiativesthat have been developed to overcome thedisadvantages of previous charginginstruments.

25. Fuel taxes. Both developed anddeveloping countries rely on fuel taxation asthe major source of taxation to finance roadsector needs. Fuel taxation is also a majorsource of general government revenue. InIndia, gasoline, and less so diesel, are alreadysubject to higher tax rates than othercommodities.

26. The retail fuel prices for motor spirit anddiesel, in different countries, are given inAnnex 1. Countries can be broadly groupedinto four categories with respect to fueltaxation:

i. Oil producing countries withvery low prices or pricesubsidies (Egypt, Indonesia,Iran, Nigeria);

ii. Countries with low overalltaxation rates (e.g. USA, anaverage tax of 10 US centsper liter);

iii. Countries with medium levelof taxes of 10 - 30 US centsper liter (India and manyother developing countries);and

iv. High price countries withtaxes ranging between 60cents to 110 cents per liter(mostly EU and Japan).

27. Fuel tax as an instrument in a wellstructured road tax regime has manyattractions.

• It is fiscally efficient (cheap tocollect with low evasion). It canbe collected at the refinery and/or point of distribution and goodrecords can be maintained toensure transparency.

• Limited impact on demand dueto low price elasticity.

• Relatively progressive as traveldemand is usually income elastic.

• Reasonably good measure fordistance related costs as fuelconsumption is highly correlatedwith the distance traveled; thusreasonably fair for allocatingvariable costs within vehiclecategories.

• Correlated with environmentaldamage; global warming effectsare fairly directly proportional tothe amount of fuel consumed.Where different fuels havedifferent environmental impacts,e.g. emissions of particulatematter, differential levels oftaxation can be levied.

28. However, fuel tax, especially on diesel,has some major limitations as an efficient roaduser charge.

• Fuel consumption does not varyproportionally with vehicleweight. Fuel taxation does notaccurately reflect road damagecosts and heavy vehicles arerelatively under-charged. Thismay adversely affect the choiceof vehicles.


l 7


15 Newbery, Hughes, Peterson and Bennathan, Road Transport Taxation in Developing Countries. The Design of user charges and Taxes for Tunisian, WorldBank Discussion Paper 26, 1988.

16 The annual UK license fee for a 40 ton/5 axle vehicle is currently £3,950: the fee for a 44 ton/6 axle vehicle is £2,95017 Shadow tolls: concessions are awarded to the private sector (e.g. in the UK in the early 1990’s) to build and maintain roads with government payments to the

concessionaire based on the traffic using the road. Such “shadow tolls” combine the advantages of transferring some traffic risk to the concessionaire whileavoiding the deterrence to traffic and other practical and political difficulties and costs in imposing tolls. Shadow tolls should not be confused with real tolls;they generate no additional revenue and hence do not contribute to solving the financing issue. They are only a contractual means of determining paymentsbetween an owner and concessionaire.

• There can be fuel substitution oradulteration15. If kerosene pricesare kept low for social reasons itmay be added to diesel, withadverse environmental and fiscalconsequences. One estimate ofsubstitutability, between fuels,and between fuels and otherinputs, suggests that the deadweight tax loss (i.e. what islos t in excess of what theGovernment receives) mayamount to over 50% of the taxrevenue. Much smaller deadweight losses are estimated fortaxes on vehicles, spare parts andtires. Diesel and close substituteshave usually to be taxed atsimilar rates to avoid widespreadfuel substitution or adulteration.

• About 46% of diesel in India is consumedoutside the transport sector. Non-transport usesof diesel should normally not be charged forhighway use but exemptions are difficult tomonitor and enforce effectively.

29. There are clearly limits to the levels oftax that can be efficiently levied on diesel.

30. Annual vehicle licenses. Many countriesuse annual license fees as both a policing/control measure and as a means tosupplement fuel taxes for road user charging.The great advantage of vehicle licenses is thatthey can discriminate within vehicle categoriesas well as between vehicle categories. Theycan, for example, discriminate within the carcategory by weight or power, and within theheavy commercial vehicle category by weightand/or axle configuration. Vehicle licenses arethus a very flexible instrument for road usercharging.

31. License fees are often used to recover thefixed costs of the network as well as the costsof road damage that heavy vehicles inflict butare not adequately recovered by fuel and othercharges. The fees can thus be set to encouragethe use of larger, multi-axle commercialvehicles which normally cause less road

damage16. License fees are fixed chargeswhich is an advantage for the recovery of fixedcosts as they do not influence the decision asto whether to make particular trips. On theother hand, the fixed nature of the chargemakes annual license fees an imperfect chargefor the recovery of variable costs. The feesdiscriminate against those vehicles which havelow total utilization, often old vehicles makingshort trips, though this may have the benefitof promoting the renewal of the vehicle fleetwith newer, more efficient and less damagingvehicles.

32. While not ideal, annual license fees areused very widely to recover road damagecosts. They are more costly to collect but arealready generally required for registration andvehicle inspection purposes. Very high licensefees may, however, result in evasion and theuse of counterfeit documentation.

33. Road tolls. There are two main economicconstraints to implementing a toll system Firstthe costs of constructing and then operatinga road toll system are quite high and roadswith low traffic are not suitable for tolling.Generally, traffic flows of about 5,000 PCUs/day are required to make tolling financiallyviable, but most of the national highwaynetwork (54,000km) already meet thiscriterion. Second, tolls deter marginal usersand/or encourage the use of non-tolledroutes17. Normally, traffic diversion to nontolled alternate routes should not be morethan about 15%. Any toll, above short-runmarginal cost will generally result in someeconomic cost. However, in aggregate, theimposition of higher toll levels may increasetotal economic welfare in three circumstances.

• Where the toll is necessary to finance,or accelerate, the provision of afacility which would not otherwise beprovided. Benefits of this type areparticularly important in India wheredemand is growing rapidly butgeneral taxation/road user chargesare insufficient to meet the desiredlevel of investment.

8 l

• Where the tolled facility is itselfcongested and the toll secures amore optimal level of utilization(fewer users, moving more quickly).These benefits arise as a consequenceof deterring some trips with a netmarginal value (above privateoperating and time costs) less thanthe toll and hence improve theperformance of the system for thosetrips with higher values.

• Where the whole system, includingthe tolled road and alternativeuntolled routes, is congested. The tolls“sort” the traffic so that the vehicleswith the highest value of time savingsuse the faster tolled route, while thosewith a lower value of time, and hencea lower willingness to pay, use theuntolled route. In such circumstances,often in urban areas, overall usercharges might be increased to reflectthe prevailing congestion.

34. Where a new toll financed route isprovided, which would not have existedwithout tolling, both the traffic on the tolledroute and those users remaining on theuntolled route benefit, when compared withthe situation of only the untolled facility.However, when an existing route is tolled,without any extra capacity or service quality,users with higher values of time will benefitand users with lower value of time will beworse off.

35. When tolls are imposed on existingfacilities, in addition to existing levels oftaxation, they will obviously increase the totalrevenue raised from road users. When theadditional revenue is used to improve thenetwork (additional capacity and/or bettermaintenance) road users may still be betteroff than without the tolls. This impact is oftennot recognized by road users and governmentsneed to ensure that when introducing ageneral toll regime on major links, sufficientattention is given to explaining to road usershow the toll has been set, how revenues willbe used and what benefits will accrue.

36. “What level of toll is acceptable to users”?The answer is often given in terms of the

proportion of the net benefits which arecaptured as tolls. Another answer can besought by considering the motivations forchoice. The users of new tolled facilitiesdemonstrate, by their choice of route, thatthey value their time and other savings morethan the cost of the toll. Conversely, thosewho choose not to use the tolled road aredemonstrating that they value the potentialbenefits less than the cost of the toll (thoughthey may still benefit from lower traffic onthe untolled road and thus increased speedsand higher service standards). Theoretically,social welfare will be maximized when thetoll is set at the level which maximizes thetotal net benefits to both sets of users plusthe profit to the operator.

37. Weight/distance charges. A seriousdeficiency in the road tax structure exists inrelation to heavy commercial vehicles. Theroad damage costs rise more steeply withweight than fuel consumption and thus thetax/charges on the fuel used. Many countriescompensate with annual vehicle fees, butthese are imperfect charges for use-relatedcosts. A more efficient solution is theintroduction of a weight-distance charge forheavy goods vehicles. This type of instrumenthas been used effectively in Switzerland andNew Zealand (see Box 1.1), is to be introducedin Austria in 2004 and the United Kingdom in2006. Germany had hoped to have its GPSbased weight/distance charge systemoperational in 200418. The fact that a weight-distance charge applies only to heavycommercial vehicles makes it less vulnerableto fraud and corruption, especially if associatedwith very heavy penalties.

Charging Structures – Effectiveness vs.Efficiency Trade-off

38. The tradeoffs to be considered in relationto the use of weight/distance chargesrepresent an example of a broader problem:striking a balance between the practical andthe theoretically efficient. This is very importantwhen designing a robust road user chargesystem; simple structures ease administration,reduce administrative costs, reduce taxevasion, and lower the costs of compliance.Complexity encourages evasion, lowers the

18 Its introduction was initially delayed to 2004, by the failure of the system to meet its technical specifications. Notice of contract termination has recentlybeen issued to the contractors. Germany still intends to proceed with some form of weight/distance charge system.


l 9


Box 1.1 Weight/Distance Charge Systems

New Zealand: the low technology route

All vehicles with a gross laden weight in excess of 3.5 tones must pay the distance charge. Distance licensesare purchased in multiples of 1,000 km. Those vehicles paying the distance charge are classified according to:(i) whether the vehicle is powered or un-powered (trailer); (ii) the number of axles on the vehicle, and (iii) thenumber of tires per axle. A license for a two axle truck (six wheels), with a permissible gross laden weight of12 tones, is $161.05/1000 km. Licenses must be carried on the vehicle and displayed on the passenger sidewindscreen. Vehicles with a gross laden weight in excess of 3.5 tones must be fitted with an approved hubodometer (»$35) to record the distance traveled. The revenue goes to the National Roads Fund. The systemis administered by the Land Transport Safety Authority and enforced by the New Zealand Police. Revenue inFY2004 is expected to be $634 million.

Switzerland: the high technology route

In 2001, Switzerland introduced a Heavy Vehicle Fee (HVF) system for all domestic and foreign vehicles. TheHVF has many objectives, including to (i) internalize the external costs from freight transport (ii) finance largescale railway projects and (iii) encourage the transfer of goods from road to rail.

Fees are determined on a tone-km basis and also vary according to the emission category of the vehicle, withthe average rate currently as 1.68 centimes (Rs 0.54) per tone-km. The fee collection is based on the principleof self-declaration and uses Dedicated Short Range Communication technology. For domestic vehicles, installationof an on-board unit (OBUs) is mandatory. For foreign vehicles, installation of OBUs is optional. Foreign vehiclescan also pay using a ticket at self service stations at various entry and exit points to Switzerland. Currently theSwiss OBUs have inter-operability in Austria and future inter-operability is planned in France.

The total investments by the Swiss authorities amounts to 160 million euros. The operating costs for thesystem are around 16 million euros per annum. Total collection costs amount to 4-7% of revenues. Netrevenues in 2002 were 500 million euros. This figure is expected to double to 1 billion euros from 2005, whenincreased tariffs come into force. High emission trucks are being replaced with those conforming to latestpollution norms. The long run trend of a constantly growing number of lorries on the roads has now beenbroken. The effect on consumer prices is negligible.

Source : http://www.are.admin.ch/are/en/verkehr/lsva/ http://www.transfund.govt.nz;http://www.ltsa.govt.nz

probability of detection and reduces transportrevenue collection19. Yet, there are alsosubstantial benefits to charging close to thepoint of use; this allows pricing signals to beperceived more easily and more directly byroad users and enhances the incentives for

19 The Administration of Road User Taxes in Developing Countries, Bahl R, World Bank Working Paper, 1992

rational choice in travel demand. There is thusa tradeoff between imposing charges thatclosely reflect the social marginal cost of use(and are perceived by users as reflecting thecosts they impose) and using instruments thatare cost effective to collect and administer.

10 l

Introduction

1. This chapter reviews the current state ofgovernment finances relating to the highwaysector in India and provides the necessary basisfor assessing the need for restructuring theapproach to the financing of the sector. Thechapter considers:

! the direct and indirect taxes and chargesapplied to vehicles and road transport;

! the total level of government revenuegenerated from the road transport sector;

! the total level of government expenditureon roads and more specifically theNational and State Highway networks;and

! the expenditure responsibilities of differentvehicle categories.

2. The estimates of revenues andexpenditures are based on national data anda sample of six states20.

2. ROAD TAXES AND HIGHWAY EXPENDITURES

Road Related Taxes and Charges

3. India has a federal political andadministrative framework under which theCentral, States and local authorities have welldefined powers for taxation and managementof roads and road transport21. Except for thenational highways, the responsibility for roadsis vested in the State Governments. BothCentral and State Governments impose taxeson vehicle purchase, vehicle ownership andvehicle use, as shown in Table 2.1

4. With the exception of road tolls and roadcesses, revenue from the various taxes/charges are not hypothecated to the roadsector, but form part of general revenue. Thelevels of taxes and charges vary widely acrossthe country.

5. Central Government taxes. Thefollowing taxes are imposed by CentralGovernment.

Customs duty: Transport fuels carry 20% duty

20 Andhra Pradesh, Karnataka, Maharasthtra, Madhya Pradesh, Uttar Pradesh and West Bengal21 See http://parliamentofindia.nic.in/const/const.html, especially Seventh Schedule and associated lists

Table 2.1: Classification of Road Taxes/Charges in India

Central Government State Governments

Vehicle Central customs Sales tax on vehicle/chassis and cab/Purchase Excise duty on motor vehicles body

Central sales tax on inter-statetransactions and shipment of vehicles

Vehicle Motor vehicle tax (annual or lifetime)Ownership Registration fee

Certificate of fitness

Taxes levied on passengers & goodsvehicles

Entry tax(1)

Vehicle Use Excise duty on fuel Sales tax on spares/ lubes/ accessoriesCess on fuel Sales tax on fuelExcise duty on spares/ lubes/ Cess on fuelaccessories Road user tollsRoad user tolls Permits & licenses

Fines & penalties

(1) Applicable to vehicles purchased/ registered in one state and brought into another state

Road Taxes and Highway Expenditures

l 11


+ Rs1.50 per liter. New vehicles attract 60%duty; used vehicles carry 105% custom duty.

Excise duty: Gasoline 30%, high speed diesel(HSD) 14%, other vehicle inputs 16 – 24%(see Annex 2). The general excise rate is 24%.

Fuel Cess: Under the Central Road Fund Act2001, a special cess of Rs.1.50 per litre onHSD and motor spirit has been applied. Thecess supports the development/maintenanceof national and state highways and thedevelopment of rural roads.

Central sales tax: 4% on interstatetransactions and transfer of vehicles bymanufacturer/dealer from one state toanother.

6. State Government taxes. The followingtaxes are imposed at the State level.

Motor vehicle (MV) tax / road tax: Smallprivate vehicles (2-wheelers and cars) aregenerally charged a one time fee at the timeof initial registration. Commercial vehicles arecharged annually on the basis of the numberof seats for buses and weight for trucks (seeAnnex 3). The MV tax for commercial vehiclesis State specific and vehicles, operating in anumber of states, have to pay additional MVtaxes. There are wide variations in MV taxbetween States and between vehicle types.Buses and multi-axle vehicles (MAVs) oftenpay more than private vehicles and smaller

Figure 2.1: Per Ton Incidence Of Motor Vehicle Tax ( In Rs )

goods vehicles. The implicit tax/ton for trucksshows little consistency. Figure 2.1 suggeststhere is no consistent approach to the settingof charges.

Sales tax: There is wide variation in the salestax rates between States (see Annex 4).Generally, fuel is charged at a higher rate thanother vehicle related items: diesel 17.5% –34%; gasoline: 20% - 30.5%; other vehicle-related inputs: 8 – 12%. The median statesales tax is approximately 12%.

Fuel Cess: Some States, such as Uttar Pradesh,levy an additional tax on transport fuel salesfor the development and maintenance of roads

Entry tax: levied by some states oncommodities and vehicles imported into thestate. In Karnataka, an entry tax of 2% islevied on tyres and 5% on fuel and lubricants.

Misc. Fees: fees for vehicle registration,issuing/renewal of driver licenses and vehiclepermits are collected, under the Central MotorVehicle Rules 2001, to meet the cost ofproviding these services and are uniformthroughout the country. Some States collectadditional fees under powers conferred by theirrespective State Motor Vehicle Rules.

7. Road tolls. Road tolls are being collectedby NHAI, some State governments and privatesector concessionaires on specific road sectionsacross the country. However, the total revenue

12 l

22 World Bank assessment23 The main sources were the Directorate of Data Management, Central Excise & Customs for the central taxes and the RBI: States Finance, A Study of

Budgets, 2002-03 for the state taxes. Data from six State Transport Departments and the Sales Tax Commissioners was used to supplement RBI dataand extrapolated for the country as a whole.

from these tolls is still low, perhaps aboutRs.1,500 Crore per year22, and have a marginalimpact on the overall financing of highways.The willingness to pay tolls appears low, andtraffic levels on tolled roads are generallybelow expectations.

8. Local Authority taxes. Local authoritieslevy octroi and terminal taxes on goods,animals and passengers entering a local area.Octroi is not specifically a road-use tax butthe resulting delays have a significant negativeimpact on road transport.

Government Revenue from the RoadSector

9. Total road-related revenue. Estimatesof the total Central and State revenuegenerated from the road sector weredeveloped from a combination of primaryand secondary sources23 and thendisaggregated between different vehiclecategories as shown in Table 2.2. (Annex 5provides the details of the assumptions made).

10. Fuel taxes and cess generate about 55%of total sector revenue. Purchase andownership fees/taxes account for about 40%

of revenue, which is high by internationalstandards and may encourage the intensiveuse of vehicles (high fixed and low marginalcosts). Central and state revenues areapproximately the same; but theoverwhelming majority of roads are theresponsibility of state or local governments.As a result, significant inter-governmenttransfers are required to match revenues withexpenditures.

11. Revenue generated by vehiclecategory. The incidence of taxes on individualvehicles, and equivalent taxes per vehicle-kmare shown in Table 2.3. Among passengervehicles, buses pay the highest taxes on botha vehicle and km basis (though not on a seat-km basis). Buses are taxed substantially morethan any category of freight vehicles, whichis very unusual.

12. Revenue contribution by road type.The levels of revenue generated on the mainroad network (National and State Highways)as well as on rural and urban roads wereestimated, using the distribution of vehicle-km on the network. Traffic on the main roadnetwork is estimated to generate Rs. 254

Table 2.2: Assessment of Total Tax Revenue by Vehicle Type: FY2002 (Rs. billion)

2- Commercial Freight VehiclesTotal wheelers Cars Jeep/taxi Bus

Light Heavy Multi-axleCentral GovernmentExcise on Fuel 150.9 32.8 52.2 4.4 10.1 15.8 34.3 1.3

Excise on Motor Vehicles 31.7 6.6 15.0 5.0 1.1 1.5 2.3 0.1Excise on Tyres 11.2 1.5 1.6 0.4 1.0 1.8 4.2 0.2

Excise on Motor Parts 15.3 1.4 2.5 0.9 1.8 2.7 5.8 0.2

Cess on fuel 28.1 2.8 4.4 1.4 3.2 5.0 10.9 0.4 Total Central Government 237.3 45.1 75.6 12.3 17.5 26.8 57.5 2.3State GovernmentsSales tax on Fuel 87.9 9.4 15.0 4.2 9.7 15.2 33.0 1.3Sales tax on Motor Vehicles 39.0 11.2 14.5 4.8 1.9 2.5 4.0 0.2

Sales tax on Tyres 6.2 0.4 0.3 0.0 1.5 0.6 3.2 0.1

Sales tax on Motor Parts 4.9 2.3 0.9 0.1 0.5 0.3 0.8 0.0Taxes on vehicles** 124.8 7.4 9.4 2.8 81.9 5.2 16.7 1.3

Total State Governments 262.8 30.7 40.2 11.9 95.5 23.8 57.8 2.9 Grand Revenue 500.1 75.8 115.8 24.2 113.1 50.6 115.3 5.2

* not including customs duties which are payable on import/export, a further Rs77 billion in 2001-02** including fees, fines, penalties, passenger and goods taxes


l 13


billion, slightly more than half of total road-related revenue (Table 2.4). Revenue fromfreight vehicles accounts for 50% of the totalrevenue on the main road network, while lightvehicles generate over 80% of the revenueon urban roads. The results are detailed inAnnex 6.

Government Expenditure on the RoadSector

13. Expenditure on the overall roadnetwork. Total recurrent and capitalexpenditures on the road network aresignificant for both Central and StateGovernments, totaling almost 3% of thecombined recurrent and capital expenditure.A broad breakdown of total governmentexpenditure, for recent years, is provided inTable 2.5.

14. Roads account for a little less than 2% oftotal government recurrent expenditures andabout 12% of total capital expenditures.

15. Capital expenditure on roads has beenincreasing and this is expected to continue,up from Rs. 85 billion in FY2000, reflectingthe focus on connecting rural villages andimproving the major National Highway system.

Road maintenance, on the other hand, hasshown little increase in recent years and mayactually be falling in real terms. BetweenFY2000 and FY2002, expenditure on roadmaintenance only increased from Rs. 50 toRs. 53 billion24.

16. Expenditure on the main highwaynetwork. All road expenditures made by theCentral Government can be attributed to theNational Highways. For State expenditure, itwas assumed that funds were allocated toState Highways, District Roads and Rural Roadspro rata to their total lengths and the unitcosts of construction/maintenance for eachcategory of road. This may understate totalexpenditure on highways which have beenreceiving priority in the allocation of availablemaintenance funds. The estimates of totalexpenditure on the main highway network aregiven in Table 2.6.

Comparison of Aggregate Revenue andActual Expenditure

17. For FY2002, the total tax revenues derivedfrom road users were very substantially higherthan total expenditures on the road sector,(see Table 2.7).

Table 2.3: Total Tax Revenue per Vehicle FY2002 (Rs.)

2-wheeler Cars Jeep/ Taxi Bus Freight Vehicles

Light Heavy Multi-axle

Average tax/vehicle:

Purchase 680 8,087 9,087 8,863 4,428 7,488 10,818

Ownership 266 2,487 2,487 210,340 5,515 17,628 33,663

Road Use 1,778 19,947 10,026 71,228 43,460 96,499 91,071

Total annual tax/vehicle 2,725 30,521 21,600 290,431 53,403 121,615 135,551

Total tax/vehicle-km 0.44 2.39 1.03 5.69 1.48 2.03 2.51

24 Recurrent expenditure in Table 2.5 also includes administrative costs for the sector and are thus significantly greater than maintenance expenditure.

Table 2.4: Road Network Distribution of Road Tax Revenues FY2002, Rs Billion

Total Revenue Light Vehicles Buses Freight Vehicles

National/State Highways 253.7 100% 52.4 21% 73.5 29% 127.8 50%

District and village roads 77.6 100% 21.8 28% 28.3 36% 27.5 35%

Urban roads 168.7 100% 141.5 84% 11.3 7% 15.9 9%

Total 500.1 100% 215.8 43% 113.1 23% 171.2 34%

14 l

18. Overall, total tax revenues from road usersare two and a half times total expenditure onthe road network. Users of the primary roadnetwork are paying about four times the levelof expenditure, and the revenue/expendituredifference is probably much greater in theurban areas. Only for district and rural roads

(with low traffic flows) is there a broad balancebetween total revenue and expenditure.

19. This may suggest that, in terms ofhighway financing and user charges, littlechange is necessary. However, Table 7presents a very incomplete representation ofreality in the road sector.

Table 2.6: Expenditure on National and State Highways FY2002 (Rs. billion)

Item National Highways State Highways Total Network

Highway Construction/Improvement 26.17 14.01 40.18

Highway Maintenance:

Routine 1.85 3.80 5.65

Periodic 5.55 11.39 16.94

Sub -total 7.41 15.18 22.60

Total Highway Expenditure 33.58 29.19 62.78

Source MORTH Directorate of Transport and Planning Commission, RBI State Finances

Table: 2.5 Government Recurrent and Capital Expenditure (Rs Billion)

Central Government State Governments All Governments*

FY2002 FY2003 FY2002 FY2003 FY2002 FY2003

Total Recurrent Expenditure 3,016 3,405 3,314 3,552 6,330 6,957

Development 823 974 1861 1971 2684 2945

Social Services 185 193 1174 1207 1359 1400

Economic Services 638 781 687 704 1325 1485

(Roads/Bridges) (68) (68) (48) (48) (116) (116)

(Other Transport) (20) (17) (17) (18) (37) (35)

Non-Development 1,746 1,926 1,399 1,544 3,145 3,470

Total Capital Expenditure 444 521 383 437 827 958

Development 84 172 365 415 449 587

Social Services -34 11 85 94 51 105

Economic Services 118 161 283 322 401 483

(Roads/Bridges) (29) (34) (66) (82) (95) (116)

(Other Transport) (55) (57) (10) (10) (65) (67)

Non-Development 178 232 18 22 196 254

Other capital/loans 182 118 182 118

Roads/Bridges as % of:

Recurrent expenditure 2.3% 2.0% 1.4% 1.4% 1.8% 1.7%

Capital expenditure 6.5% 6.5% 17.2% 18.8% 11.5% 12.1%

Recurrent + capital 2.8% 2.6% 3.1% 3.3% 2.9% 2.9%

* Estimates are marginally higher than other sources


l 15


(a) Taxes on road use (vehicles, fuel, etc)are used almost universally togenerate general governmentrevenue as well as to finance the roadsector. Taxes on road use are notnecessarily equivalent to road usercharges.

(b) Actua l expend i ture on roadmaintenance may not be sufficient

to maintain the network, which maybe deteriorating, and thus does notreflect the real road damage costsimposed by vehicles.

(c) Present expendi ture on roadconstruction and improvement maynot reflect the needs of the sectornor provide a realistic reflection oflikely future financing requirements.

Table 2.7: Total Revenues and Expenditure in the Road Sector FY2002 (Rs. billion)

Tax/ChargeRevenue Road Expenditure Expenditure as% of Revenue

National/State highways 254 63 25%

District and Rural Roads 77 64 83%

Urban Areas (1) 169

Total 500 211(2) 42%

(1) Relatively small expenditures included with District roads(2) Including administrative costs

16 l

Road User Charges in India

1. In some countries, there are specific anddesignated charges for road use (the roadcess, for example). But most taxes on vehiclesand operating inputs (fuel, tires, etc) are partof the general tax structure; they generaterevenue for general expenditure as well asimplicitly performing a road user charging role.A distinction needs to be made between these“normal taxation” and “road user charging”roles in order to try and ensure that both areset at appropriate levels. However, the roaduser charging function is rarely recognizedexplicitly in tax setting and thus it is necessaryto determine an implicit separation betweenthe two functions. There is a spectrum ofpossible approaches to making the separation.

(a.) At one end of the spectrum: alltaxes and charges paid by roadusers (i.e. total road user generatedrevenue), from their ownership anduse of vehicles, could be consideredas road user charges. This wouldimply that consumption of roadservices is exempt from the generalapplication of indirect taxation; thisis implausible.

(b.) At the other end of the spectrum:only those specific charges whichare dedicated to road financing(such as the road cess and roadtolls), together with vehicle anddriver registration fees might bedefined as user charges, and allother revenues considered asgeneral taxation. This would implythat very high rates of indirect taxeson road-use related items containedno implicit element of road usecharge; this seems unlikely.

2. Often, it is assumed that all commoditiescarry the same rate of indirect tax for raisinggeneral revenue and that any excess over thisgeneral level can be considered as a road usercharge. This approach can certainly be applied

3. THE ROAD USER CHARGE REGIME

in India at the state level, where there is ageneral standard sales tax and higher rateson transport fuels. It is more difficult to applyto Central Government’s customs and exciseduties, which vary widely. In thesecircumstances, the excess over the averagerate can be used, but it is less logical; adistinction can also be made between averagetaxes on consumer goods (and applied to lightvehicles) and average taxes on productiongoods and raw materials (and applied to freightvehicles).

3. There are conceptual problems with usingthe “excess over average tax” to define roaduser charges. A general principle of indirecttaxation is to minimize economic distortions.Efficient taxes should thus be structured ininverse proportion to the price elasticity ofdemand for the products. The demand fortransport fuel is relatively price inelastic so,for macro-economic efficiency, the generalrevenue tax on fuel should be higher than theaverage general revenue tax. For economicequity or redistribution reasons, general taxesmay be above average on those goodsconsumed disproportionately by the rich.Gasoline, used in private cars, may thus bemore heavily taxed than the average.

4. Without an institutional separation of roaduser fees from general taxation, there will beambiguity in the estimation of road usercharges. It is useful, however, to have abaseline upon which to judge the adequacyof user charges, and to explore theimplications for increasing road expenditures.Despite the methodological issues, the “excessover average indirect taxation” has been usedas the basis for one estimate of implicit roaduser charges25. A second estimate of road usercharges has been made on a rather narrowdefinition, which excludes the “excess overaverage indirect taxation” element.

5. Based on the above assumptions, totalroad sector revenues have been divided intogeneral taxation and road user charges inTable 3.1.

25 For excise duties, the road user charge is the excess rate above the average excise rate of 24%, For state sales tax, the road user charge is the excessrate above the average rate of 12 percent.

The Road User Charge Regime

l17


6. Depending on the assumptions adoptedregarding indirect taxes, such as excise andsales tax, total road user charges (bothexplicit and implicit) in India are between30-40% of the total revenues generated fromthe road sector.

7. The levels of implicit road user charges(RUC) paid by vehicle category are outlinedin Table 3.2. The level of road user chargesare very substantially lower than total taxrevenues for all vehicle categories, with thepartial exception of the bus category. Themost marked impact is on two-wheelers andcars.

8. The distribution of total user chargesbetween the different road categories israther different to the distribution of totalroad-user related tax revenues, reflectingthe very high proportion of light vehicles inthe traffic flows on urban road networks(Table 3.3).

9. The level of road user charge revenuegenerated in urban areas is much lower thanits charge of total road-user related tax

revenues. The proportion of total road usercharges generated by buses on the non-urbanroads is very substantial.

Road User Charges vs. Road Expenditures

10. In terms of the total Centre and StateGovernment expenditure, reported in Table5, there is a small shortfall when comparedto the revenue raised using the broaddefinition of user charges, and a rather largershortfall on the narrow definition, (see Table3.4).

11. At a network level, there is sufficientfunding from highway users to cover identifiedactual expenditures on national and statehighways. There is a significant road-userfunding deficit for district and rural roads; thisis perhaps not unexpected as a large part ofthis network carries little traffic but is requiredfor basic access. However, as indicatedpreviously, actual road expenditures may notreflect the level of road maintenance actuallyrequired to keep the network in a goodcondition and remedy the damage inflictedby road users.

Table 3.1: Total Road User Charges, Rs. Billion FY2002

Item Rs billion % of total revenue

Broad Definition of Road User Charges*

General Taxation 300 60

Road User Charges 200 40

Total 500 100

Narrow Definition of Road User Charges

General Taxation 347 69

Road User Charges 153 31

Total 500 100

* including excess over average taxes for excise and sales taxes on fuel

Table 3.2: Road User Charges per Vehicle FY2002 (Rs.)Vehicle Type 2- wheelers Cars Jeep/taxi Bus Freight Vehicles

LCV HCV MAV

Total Tax RevenuesPer vehicle 2,725 30,521 21,600 290,431 53,403 121,615 135,551

Per vehicle Km 0.44 2.39 1.03 5.69 1.48 2.03 2.51Broad Definition of Road UserCharges

Per vehicle 747 8,098 4,902 226,424 15,866 40,055 54,910Per vehicle Km 0.12 0.63 0.23 4.44 0.44 0.67 1.02

Narrow Definition of Road userCharges

Per vehicle 365 3,642 3,723 218,574 10,814 29,109 44,548Per vehicle Km 0.06 0.28 0.17 4.29 0.30 0.49 0.83

18 l

Road User Charges vs. Full Road Costs

12. Full road maintenance. Actualexpenditure on road maintenance is often nota good approximation to road damage costs.Like many countries, India under-maintains itsroads and investment crowds out operationsand maintenance. New roads are beingconstructed while the overall road network is

deteriorating through lack of maintenance. Theexpenditure required to maintain fully theNational and State Highways was estimated,using GOI’s own norms26. The full maintenancerequirement for the highway network is threetimes the level of actual maintenanceexpenditure, (see Table 3.5).

26 An analysis of maintenance needs was also made using the World Bank’s highway design model (HDM4). This analysis confirmed the magnitude ofmaintenance needs as determined by the norms, with estimated requirements within 20% for the National Highways and 10% for State Highways.

Table 3.4: Total Road Expenditures – Road User Charges FY2002 (Rs billion)

Road User Charges

Road Expenditure Broad Definition Narrow Definition

National and state highways 63 112 92

District and rural roads 64 36 31

Urban N/a 52 30

Total 211 200 153

Table 3.3: Distribution of Road User Charges by Road Category, FY 2002

Total Total User Light Vehicles Buses Freight Vehicles

Revenue Charges

Rs. billion Rs. billion % Rs. billion % Rs. billion % Rs. billion %

Broad Definition of RUC

National/State Highways 257 112 56% 14 12% 57 51% 41 37%

District and Village Roads 77 36 18% 6 15% 22 61% 9 24%

Urban Roads 169 52 26% 38 73% 9 17% 5 10%

Total 500 200 100% 57 28% 88 44% 55 28%

Narrow Definition of RUC

National/State Highways 257 92 60% 7 8% 55 60% 30 32%

District and Village Roads 77 31 20% 3 10% 21 69% 6 20%

Urban Roads 169 30 20% 18 60% 9 29% 3 11%

Total 500 153 100% 28 18% 85 56% 40 26%

Table 3.5: National and State Highway Maintenance Needs FY2002*(Rs. billion)

Item Highway Maintenance Needs

National State Total

Maintenance Requirement 33.76 37.17 70.93

Routine 7.68 10.98 18.67

Periodic 26.07 26.18 52.26

Actual Expenditure 7.41 15.18 22.60

Actual/Requirement 22% 40% 32%

* Assumes the standard unit rates for Zone IV given in the norms for Road Maintenance in India published by MORTH


l19


13. In aggregate, road user charges wouldmeet the full maintenance costs on the mainhighways: Rs. 112 billion (Rs. 92 billion, usingthe narrow definition) compared withmaintenance needs of Rs. 71 bill ion.However, not all vehicle categories may covertheir road costs.

14. The road maintenance costs, both routineand periodic, incurred to maintain a roadnetwork in a good and stable condition canbe broadly divided into two categories.

(a) Vehicle attributable costs: These arevariable costs that increase with the levelof traffic flow. The costs can be specificallyrelated to individual vehicle categories,according to either the damage thevehicles cause or the road space that thevehicles occupy.

• Road Damage: road deterioration andthe required strength of pavements isrelated primarily to the number andweight of axle-loads and is measuredby total Equivalent Standard Axles(ESA). Light vehicles, such as cars,impose very little damage to pavedroads, heavy commercial vehiclesimpose substantial damage, especiallyif they are heavily overloaded.

• Road Space: the use of road space,important in determining bothcongestion and the required width fornew road construction, is a functionof the size and speed of vehicles andis normally related to total equivalentPassenger Car Units (PCUs).

(b) Fixed costs: These costs cannot bedirectly attributable to any particularvehicle category and are usually causedby the passage of time or the effects ofweather. The costs are sometimesdisregarded for the purposes ofdetermining appropriate road user chargesand financed from general revenue, or areallocated on an ‘equitable basis’ to thedifferent vehicle categories, normally onthe basis of total PCUs.

15. The full road cost requirements for thehighway network, both capital andmaintenance, have been allocated to

individual vehicle categories on the basis oftheir cost contributions. Estimates have beenmade for both full maintenance costs, withthe fixed costs allocated according to theirPCUs, and variable costs which can bespecifically attributed to vehicle categories,(see Table 3.6 and Annex 7).

16. If full maintenance was to be undertakenon the network, few vehicle categories wouldfund their combined capital and maintenancecosts; under the narrow definition of usercharges, only buses cover their full share ofroad costs.

17. Much more seriously, the heavycommercial vehicle categories (Heavy andMulti-axle) only just cover their roadmaintenance costs, even under the broaderdefinition of user charges. Under the narrowdefinition of user charges, the heavycommercial vehicle group (excluding multi-axlevehicles) fails even to cover their attributableroad maintenance costs, which is the absoluteminimum requirement for a road user chargingstrategy.

18. The user charges on buses, however, aregreatly in excess of their share of road costs,irrespective of the definition of cost or thedefinition of road user charges. Quite clearly,buses, and consequently bus passengers, aresubsidizing the system and implicitly the heavycommercial vehicles, in particular.

19. R e q u i r e m e n t s f o r a c c i d e n texternalities and network operations. Inaddition to under-maintenance, India allocatesinsufficient resources to ensure that roads areoperated efficiently and safely. There are over70,000 road deaths annually, including manypedestrians. A recent study estimated thatroad traffic accidents cost India Rs. 190 billionin FY2001, or about 1% of GNP. Improvedengineering, education and enforcement couldsubstantially lower these costs27 but only Rs.350 million is presently invested in road safetyby central government, as well as an unknownbut probably very small amount by state andcity authorities. NHAI is now startingretroactively to provide adequate safetyinfrastructure on completed four lane highways,at a cost of Rs. 2 – 2.5 million lakh/km28.

27 Definition of Road Safety Policy and Action Plan for India, Span Consultants/DRD for MORTH, September 200328 Detailed Project Reports for Minor Improvements Works, Third National Highway Project, NHAI, 2003

20 l

20. It has not been possible to estimate thefunding provided for operating (excludingroad safety) the highway network. However,as very few sections of highways have anytraffic information, emergency facilities orother road services, the funding must below.

21. Congestion Costs. Very heavily loadedtrucks are driven at low speeds. These slowtruck speeds, together with the limited roadwidths, road side development and non-motorized traffic result in low averagespeeds for all vehicles on most of India’shighway network. Light vehicles, whichcould travel at much faster speeds, aredelayed by the heavy trucks, i.e. even onthe inter-urban highway network there is alevel of congestion. Ideally, those responsiblefor the slower speeds should be charged forthe delays that they cause to other roadusers. This would significantly increase thetotal road costs attributable to all freightvehicle categories.

Assessment of the Road User ChargingRegime

22. Approximately Rs. 100 bi l l ion iscollected, as explicit and implicit road usercharges, from vehicles using the nationaland state highway networks. While this issufficient to fund present public investmentand maintenance on the network, thepresent charging regime is deficient in anumber of major respects.

23. The total level of road user chargesis insufficient. The present level of usercharges is sufficient for present expendituresonly because the highway network is beinggrossly under-maintained, poorly operatedand with little attention to road safety. Asizeable proportion of the investment beingmade in the network is, in effect, capitalizedmaintenance, a very inefficient way ofmaintaining roads. Moreover, it is apparentthat the present levels of user charges cannotinclude any significant element for theimportant externalities of congestion, roadsafety or environmental damage.

24. The structure of road user chargesis economically inefficient. The presentcharges on heavy commercial vehicles cover

Table 3.6: Road User Charges : Full Road Costs FY2002National and State Highways (Rs. Per km)

Vehicle:- 2- wheeler Cars Jeep/taxi Bus Freight Vehicles

Item Light Heavy Multi-axle

Broad Definition of Road User Charges

User Charges/vehicle 0.12 0.63 0.23 4.44 0.44 0.67 1.02

Road Cost/vehicle

Total cost 0.17 0.33 0.33 1.14 0.55 2.03 2.81

Capital cost 0.13 0.25 0.25 0.84 0.41 1.34 1.88

Maintenance cost (total) 0.04 0.08 0.08 0.30 0.14 0.69 0.93

Maintenance cost (variable) 0.02 0.04 0.04 0.19 0.09 0.58 0.78

User Charge : Cost Ratio

Total cost 0.7 1.9 0.7 3.9 0.8 0.3 0.4



Narrow Definition of Road User Charges

User Charge/vehicle 0.06 0.28 0.17 4.29 0.30 0.49 0.83

User Charge : Cost Ratio

Total cost 0.4 0.8 0.5 3.8 0.5 0.2 0.3




l 21


only 80% of their attributable damage costs(variable road maintenance), which is theabsolute minimum level for user charges.In addition, trucks impose substantial delayson other road users through their slowspeeds, even on the highway network. Roaduser charges on heavy commercial (two,three and multi-axle) vehicles need to besubstantially increased.

25. The present charging structure takes noaccount of the road space occupied by eachvehicle. This is important for the highwaynetwork and critical on urban road networks,where charges related to congestion costswould be economically efficient and yieldsubstantial revenues for the improvementof public transport services. The presentcongestion charge in Central London hassubstantially reduced congestion andincreased service standards and has beenaccompanied by a significant shift ofpassengers to the bus service29.

26. The structure of road user chargesappears inequitable. Buses are the mostheavily taxed/charged category of vehicles;they are very heavily taxed in most statesthrough a heavy vehic le tax and in

Maharashtra by a very heavy tax based onpassengers carried. The implicit road usercharges on buses are more than four timesthe charges on multi-axle freight vehicles,despite their much lower road damage costs.Buses are also charged much more heavilythan cars, in relation to their road damagecosts, even though they are far moreefficient than cars in terms of road space.There is no economic reason for such highcharges and, as buses are used by the lowerincome groups, they appear inequitable.

27. The structure of road user chargespromotes an uneconomic distribution oftraffic. Road freight vehic les areundercharged. Road freight rates arebelow their economic level encouragingthe shift of freight from rail (rail freightgenerates profits for Indian Railways).Buses are overcharged, bus fares aremuch higher than their full economic costand thus more people decide to travel byrail (Indian Railways lose money on mostpassenger services, especial ly thosecompetitive with bus). Both the road andrail sectors, as well as the overall economy,lose as a consequence of the present leveland structure of road user charges.

29 See http://ecaweb.worldbank.org:8080/Transport.nsf/ECADocByUnid/D1B8523C31854F3685256D4700536A4E?Opendocument

22 l

Highway Maintenance

1. Maintaining India’s present highwaynetwork to full maintenance standards willrequire annual funding of about Rs. 70 billion,three times the current level of expenditure.Some may argue that India cannot afford tomaintain fully its roads and priority should begiven to expanding the network. All theevidence suggests that India cannot afford notto maintain its highway network – valued atroughly Rs.240,000 crore (US$53 billion). Arecent study analyzed the economic impactof inadequate road maintenance30 and foundthat:

• The economic road user costs are23% higher on roads in poorcondition than on good roads, and55% higher, if the roads are in verypoor condition;

• The cost of surface dressing (for roadsin good condition) is 66% lower thanresurfacing or strengthening (for roadsin fair condition); and only 25% ofthe reconstruction cost;

• The annual maintenance backlogsrange from 2.5 – 4 times the requiredsteady state expenditures; and

• In some states, for every one Rupeespent on maintaining the network,there are net benefits (NPV) in excessof Rs 7.

2. The study makes a strong case forsubstantial increases to the level ofmaintenance expenditure, if necessary byreallocating from capital expenditure.Extrapolating the results to India gives amaintenance backlog in the order of Rs. 130billion. Cutting back on road maintenanceneither makes economic sense nor long-termfiscal sense as the future costs for roadreconstruction will be much higher.

Highway Investment

3. The Central and State Governments have

4. FUTURE FUNDING NEEDS

realized the importance of improving India’sroad system, both in terms of providingwider accessibility to rural areas, and addingtraffic capacity and improving service levelson the primary highway network. Theseimprovements are essentia l i f theGovernments are to meet their objectivesof achieving high economic growth ratesand reducing poverty.

4. Plans have been prepared for roaddevelopment by the MORTH and the PlanningCommission. The GOI’s Vision 2021 assessedthe demand for road transport, based on thedesired future annual economic growth of the6-8%, and estimated the need for roaddevelopment in the country for the next 20years. Vision 2021 sets out physical andfinancial targets for highway development. Inbroad terms, the investment needs of theExpressways, National Highways, and StateHighways, in the ten years 2001 – 2011, areestimated as Rs.300, Rs.1,200 and Rs.750billion respectively (1999 prices), over Rs.2trillion in all (the estimates are summarized inAnnex 8). In addition, the PMGSY programwill require substantial funding, of the orderof Rs.70,000 crore over the period to 2010, toconnect every village with all weather roadaccess.

5. So far, significant progress has beenmade by central government in theimplementation of the National HighwayDevelopment Program and PMGSY, and stategovernments have made improvements toabout 20,000 km of the State Highwaynetwork.

The Funding Gap

6. This section estimates the overall financialresources needed, over the next 10 years, todevelop and maintain the primary highwaynetwork, assesses the likely revenue from roaduser charges, and compares the likelyexpenditures and road user charge revenues.The estimates are based on network wideanalysis using the HDM4 model.

30 Costs of Deferred Maintenance in India, World Bank, 2003. http://www.worldbank.org The study undertook a network strategic evaluation, using HDM-4, for the core highway networks in Tamil Nadu, Gujarat, and Karnataka.

Future Funding Needs

l 23


• Road user charge revenues areestimated on the traffic growthrates for different vehicles types,based on projected economicgrowth, and the present charges foreach vehicle category. No allowanceis made for the widespread impositionof tolls.

• The maintenance costs are basedon the norms recommended in the‘Report of the Committee on Normsfor Maintenance of Roads in India -2000’.

• Capital investment (four/two laningexisting two/intermediate) lanehighways was assumed necessarywhen traff ic levels reachedcategory C as per IRC norms.

7. The analysis estimated the need towiden 15,000 km of national highways fromtwo to four lane, and a further 16,500 kmfrom intermediate to two lane. The totalcost would be about Rs.1,098 billion (2003prices), very close to the estimates in Vision2021. The analysis suggested that about25,000 km of state highways will needwidening to two lanes, at a cost of aboutRs.623 billion (2003 prices). This is a veryconservative estimate. No allowance ismade for addressing the maintenancebacklog, which may be considerable.

However, much of the backlog would becovered in the widening works, which wouldalso include rehabilitation. Nor is anyal lowance made for establ ishing anexpressway which at roughly Rs. 15-20 croreper km would add considerably to thefinancial requirement (Vision 2021 estimatesa further Rs. 300 billion for expressways from2001-2011).

8. Revenue from road user charges willmore than cover highway maintenance, ifthe funding is dedicated to highways.However, even with this conservativeestimate of needs, the required capitalinvestment cannot be fully funded by roaduser charges, (Figure 4.1).

9. With the revenue from the defined roaduser charges, the cumulative fundingshortfall over the 10 year period is estimatedat Rs. 1,048 billion, 39% of the totalrequirement. Highway maintenancerepresents about 35% of the total projectednetwork cost, considerably above the actual22% allocation in FY2002. The funding gapassumes that all the road user chargesgenerated on the highways are returned tothe highway sector. If the current proportionof road-user charge revenue is returned(56%), then available funding for highwayswould be only Rs.912 billion, less than themaintenance needs, and the funding gapwould rise to Rs.1,760 billion.

Figure 4.1 Revenue from Sector Taxes/RUCs vs. Highway Requirements, 2002-2011

24 l

Major Financing Issues

1. The major issues, requiring urgentattention, include the following.

Highway Financing

• Inadequate maintenance funding.This applies to both national andstate highways, and is partly due tothe low political profile of roadmaintenance (particularly at thestate level).

• Investment financing gap. Thisresults from the very rapid increasein expected road investment, fromNHDP and PMGSY, and theinadequacy of budget funds in theearly years of the programs.

Road User Taxation

• Proliferation and overlapping oftaxes. This arises from the allocationof expenditure responsibilities andtaxing powers to the states, togetherwith the lack of standardizedapproaches. Significant differencesbecome entrenched and are verydifficult to eradicate.

• Regional disparities. The states havethe constitutional prerogative to levycharges on transport and there areneither national guidelines nor aconsultative forum to help ensureconsistency.

Road User Charges

• Undercharging of heavy goodsvehicles. This may result from thedesire to keep freight rates low butit is also a consequence of relianceon fuel taxes which are inadequateto reflect the costs imposed byheavy vehicles. The costs are notfully recovered by fixed annual feeswhich are a state responsibility.

• Overcharging of buses. Both vehicleand passenger taxation are statefunctions and bus transport is aneasy revenue source. The levels of

charges appear neither equitablenor efficient.

• Urban congestion. There is neitherurban road charging nor extensivetraffic management to controlcongest ion. Cost ly addit ionalinfrastructure (e.g. the Mumbaioverpasses) is constructed, withsignificant financial implications.

• Lack of Direct Charging. As mostcharges are not applied at the pointof use, consumers have noincentive to manage their demandfor road transport.

Inter-modal Transport Policy

• Lack of coordinated inter-modalpolicies. There is limited coordinationbetween the policies of IndianRai lways and the rest of thetransport sector. Railway pricing androad user charges encourage toomuch road freight and too many railpassengers, result ing in aninefficient distribution of traffic.

2. The National Five-Year Plan has a clearvision of an improved/expanded highwaynetwork as an essential foundation for fastereconomic growth. This vision may befrustrated unless solutions are found to:

• the inadequate provision for roadmaintenance;

• the insufficient budgetary resourcesfor the capital program; and

• the fragmented and inconsistentroad tax and charging systems.

3. Solutions have to be found within apolicy environment in which there is heavyreliance on road-user taxes for generalrevenue, a fragmented decision makingstructure and inadequate information.

Underlying Causes

4. Road users as general tax revenuegenerators. In India, only a third of roaduser taxes are returned to the road sector

5. MAJOR FINANCING ISSUES AND CAUSES

Major Financing Issues and Causes

l 25


as investment or maintenance; this iscomparable to Western Europe but muchless than the USA (+90%, most road taxesare hypothecated)31 or Australia (50 – 60%)32.

While road-related tax revenue in India, as ashare of GDP, is similar to other countries, road-related tax revenue as a share of total taxrevenue is much higher, (see Table 5.1).

31 See http://www.fhwa.dot.gov/ohim/hwytaxes/2001/32 Australian Bureau of Statistics and Dept of Transport33 Highway Efficiency Study, World Bank, 2003

5. Total tax revenue in India is only about18% of GDP, and both the Government ofIndia and the State governments rely heavilyon the road sector for general revenue; roadrelated taxes are generally cheap and easy tocollect. A broadening of India’s tax base maybe desirable but perhaps not achievable inthe short term. Increasing expenditure in theroad sector may thus require additionalcharges on road users. These additionalcharges may be more acceptable to road usersif, in conjunction with the higher charges, thereis greater earmarking of road user taxes/charges to the road sector. Some countrieshave accompanied higher road-relatedtaxation with giving road users some controlover how the funds are spent.

6. The high level of tax on the sector mayhave a profound effect on public acceptabilityof tolls. Tolls are being set generally at a smallfraction of the operating cost savings expectedfor passenger vehicles and nominal toll ratesin India are some of the lowest in the world,(Table 5.2). However, in relation to average

incomes (the affordability index) the tolls maybe considered as relatively high, althoughprivate car users have much higher incomesthan the average. The Indian truck/car tollratio also appears high but, on the other hand,two/three axle trucks are generally much moreheavily loaded than elsewhere and thusimpose much greater costs on the roadnetwork.

7. A study, commissioned by the Bank,concluded that the actual savings to trucksmay be much lower than is often used in tollstudies. The study estimated time costs at onlyabout Rs.70-80/hour (perhaps slightly morethan Rs. 2/km33), and doubling speeds wouldonly save truck owners about Rs.1/km. Thislimits the potential for high tolls on freightvehicles unless there are also appreciabledistance savings, as generally the speed ofthe trucks establishes the speeds on theuntolled network. Tolls for trucks are oftenhigher than this, even in India. Overall, thesurvey suggested that the perceived userbenefit of high quality toll roads is relatively

Table 5.1: Road Sector Tax as Share of Total Tax and GDP

Country Road Taxes as % of Total Taxes Road Sector Tax Revenue as % of GDP

Australia 4.7 2.0

Germany 5.6Z 2.1

Italy 4.5 1.9

United Kingdom 6.7 2.4

Denmark 5.9 3.0

Finland 7.0 3.2

Greece 10.0 3.3

Ireland 10.2 3.2

Netherlands 6.8 2.7

USA 3.6 1.1

India 15.5 2.2

Note: Data for EU countries is for year 1999, and for India relates to year 2001-2002Sources: Study on Vehicle Taxation in the Member States of the European Union, January, 2002 and Data for USA is for 1996-97, Source: International Road Federation, World Road Statistics, 1999

26l

low at present, especially for freightvehicles34.

8. So while general tax levels are so highand benefits, to commercial user at least,from upgraded roads less certain, willingnessto pay is low.

9. Fragmented tax decision-makingstructure. Highway financing is complicatedby the levels and types of sector taxationand expenditure which are established byseveral different agencies and layers ofgovernment. The Constitution of Indiaincorporates detailed provisions relating tothe enactment of laws and the principles ofdevolution of taxation powers between theCentral and the State Governments.

(a) Central Government is responsiblefor customs duty, excise duty andcentral sales tax on inter-statetrade.

(b) Both the Central and the StateGovernments are empowered tolegislate on mechanically propelledvehicles including the principles onwhich taxes on such vehicles are tobe levied.

(c) State Governments have the rightto levy taxes on motor vehicles(road tax), on goods and passengerscarried by road, tolls and octroi andentry tax. Regulatory control isexercised under State specific rules,within the broad framework ofMotor Vehicles Act 198835.

10. There is no national road pricing/usercharging policy nor any procedures toharmonize the type and level of road taxes.Consequently:

• there is multiplicity of taxes, dutiesand fees, levied at var iousadministrative levels;

• motor vehicle taxes vary substantiallybetween states without anyapparent rationale for the levels anddifferentia ls between vehic lecategories; and

• tax rates appear to be fixed in anad hoc, arbitrary manner.

11. The overriding motivation for changing taxlevels appears to be generally to increase taxrevenue, with little regard to economicefficiency, equity or other public policyobjectives.

34 Road freight services in India are primarily bulk, low cost and relatively low quality in which high speed delivery commands relatively little premium35 The Motor Vehicles Act 1988, a GOI Act, covers the regulation, control and operation of transport vehicles, including the licensing of drivers and

conductors, registration of motor vehicles, control of traffic, insurance and other related matters. This Act, supplemented by the Central Motor Vehicle(CMV) Rules 1989, define the powers of the Central and State Governments with regard to the framing of rules, practices and procedures. It is the maininstrument through which motor transport is regulated in the country by State Governments.

Table 5.2: International Road Toll Rates

Road Toll US¢/km

Cars Trucks Affordability Index*

Country 2/3 axles 4+ axles Car 2/3 axle truck

Australia 6.0 15.3 15.3 0.284 0.621

Italy 5.3 6.7 12.1 0.285 0.273

Portugal 4.0 8.0 10.0 0.395 0.452

Spain 6.5 14.9 17.4 0.352 0.750

USA 2.6 5.6 9.0 0.204 0.163

Brazil 3.8 8.9 16.2 0.990 1.259

South Africa 3.2 8.7 13.0 0.642 0.797

Average toll 4.5 9.7 13.3

Truck/Car Index 1.0 2.2 3.0

India 1.0 3.5 2.482 1.241

* toll/per capita income at purchasing power parity

Major Financing Issues and Causes

l 27


12. Lack of information for policyformulation. There is a lack of reliable,complete and timely information on thecurrent levels of road user taxes andcharges, and their allocation to roads. Therehas been no study of road user charges/costssince 1988-8936.

• In some cases, the primary sourcesare not held in a way that allowsanalysis without making far-reachingassumptions. For example, mostroad agencies do not record thedivision of expenditure betweenroad categories; many states do notreport sales tax on motor spirit andlubricants separately, only totalsales tax revenue.

• In other cases, information ispresented in such an opaquemanner that interpretation is almostimpossible. The state budget inKarnataka, for example, has dozensof budget heads/subheads coveringexpenditures by the Public WorksDepartment on different roadcategories. Many of the budgethead t i t les are not readi lyunderstood, outside a select few inGovernment, and relate to projectactivities no longer active.

13. This lack of reliable and/or coherentinformation affects both policy makers andthe road user who pay the taxes and charges.It is difficult to see how policy makers canmake coherent policies and expendituredecisions without accurate data on the leveland distribution of taxation and expenditure.Even if a road user charging policy existed,a sound information base would benecessary to monitor its impact and providethe basis for corrective changes. The lackof information also makes it difficult for roadusers to hold anyone accountable for themore effective or efficient use of taxescollected from the road sector. The consumerof publicly provided water and sewerservices receives regular detailed bills, andaudited annual accounts are publiclyavailable. With such information, there canbe accountability for both service standardsand the level of user charges. There isnothing comparable in the road sector and,while it may not be possible to provide thesame level of detail as for a water utility,there should be accurate, comprehensibleand timely information on the chargesraised from and the expenditures madeto the sector. Indeed, th i s leve l ofdisclosure is now mandated through theRight to Information Act.

36 The World Bank commissioned a study on Vehicle Fleet Modernization and Road User Charges

28 l

Introduction

1. There is an ongoing public policy debatein India on how to fund the necessary newinvestment as well as operations andmaintenance on the growing national andstate highway network. The GOI and manystate governments are interested inbroadening the role of the private sector inhighway development with a view tostrengthening and expanding private financingof highways. There are several keydeterminants of the viability of privatelyfinanced road programs including the countryregulatory and legal environment and theresulting nature of the public/private riskregime. Compared with other infrastructureprojects, there are several financing difficultiesinherent in road projects – e.g. the acquisitionof long-segments of right of way andassociated resettlement issues as well asunforeseen geological and weather conditions.In addition, there are substantive risksassociated with the unpredictability ofrevenues and toll receipts due to competingroutes, unexpected revision in toll rates,adverse local reaction and availability ofconnecting roads. This section briefly describeswhy and how governments can facilitateprivate sector participation (PSP).

Rationale for Private Sector Participationin Indian Highways

2. Bridging the funding gap. Privatefunding first and foremost is often seen as anincreasingly important means to bridge thefunding gap between the requirements of thesector and public resources available. Likeother public services, the sector faces aconstrained fiscal environment - the generalfiscal deficit has returned to the 9-10% of GDPrange during the Ninth Plan period (1997/98-2001/02)37. It has been shown elsewhere thatthe funding gap is about Rs.1,048 billion overthe next ten years. Into this gap, private fundshave already started – approximately Rs.6,500 Crore on NHs and a further fewthousand crore of Rupees for SHs.

6. PRINCIPLES OF PRIVATE FINANCE

Furthermore, expectations are high thatconsiderable additional private funds willcome in the future. While the Tenth Planindicates a further 10,000 km of NH to befour laned, GOI budget allocations are onlyproviding for 40%; hence; roughly Rs. 24,000Crore as balance is being expected to comeform the private sector through toll based BOTs.

3. However, it is critical to remember thatthe flow of private capital only helps postponethe capital costs to the tax payer and/or roaduser until future years. Insofar as directpayment of tolls to a private operator detractsfrom the Government’s capacity to charge,indirectly or directly, for road use, whethernew construction is funded by the public orprivates sectors does not in of itself make anydifference from an economic perspective.

4. Achieving efficiency gains. A secondcommon argument in favor of private fundingof roads is to achieve efficiency gains. Thisargument contends that gains can beattributed to the following.

• A system of incentives andsanctions: motivation at companylevel to earn a good return andfear of bankruptcy are passed onto individuals within the f irmthrough wage increases andcareer development opportunitiesencouraging them to work harder andsmarter.

• Flexibility: the private sector hasgreater flexibility in adjusting itsresources (personnel, equipment andmaterials) to a constantly changingsituation.

• Comprehensive approach: whenentrusted with a long-term contractand a wider scope of work, privatefirms have an incentive to balanceexpenditure over a project’s life andmake effective trade-offs betweeninvestment, maintenance andope ra t i on co s t s sub j e c t to

37 India. Why Fiscal Adjustment Now, World Bank, Forthcoming

Principles of Private Finance

l 29


environmental, social and economicconsiderations.

• Access to technology: large firmshave greater incentive to invest inresearch and development toimprove the quality and efficiency ofconstruction techniques, processesand equipment.

5. Note however, that these potentialefficiency gains do not require the privatesector to assume traffic risk through theirrevenues being dependent on real or shadowtolls.

6. Broadening the revenue base wherethere is direct tolling. Common sensesuggests that private investors will have anincentive to maximize toll receipts when theyrely on toll receipts in whole or in part forreturns. There are several reasons why thismay be the case. First, a normal profitmaximizing firm will seek to reduce leakageand ensure efficient collection. Second, thereis an incentive on an agency that relies ontolls for its revenue to set its prices at a levelthat maximizes revenue – i.e. to manage thedemand and supply curves in a more rigorousway than the public sector that might haveother objectives in mind. Third, given thatneither users nor the government has primafacie any incentive to raise toll levels tofinancially sustainable levels, private financierswill be the only party clamoring for upwardrevision of tolls. Highways have in the pastsuffered from inadequate funding due to freeride users and politicians. Having an interestedparty argue for a sustainable approach is betterthan all parties colluding in an unsustainablesystem that postpones too much payment tofuture generations. Already, about 1,400 kmand 500 km of national and state highwaysrespectively are tolled raising about Rs. 1500Crore per year38.

7. Unbundling and Reallocating Risk.More generally, the potential benefits notedabove can be described as unbundling andshifting the risk and rewards of a newinvestment from the public to the privatesectors with the expectation that this will loweroverall costs to society. Like any humanactivity, building a road is inherently risky not

just because estimated construction costs mayescalate but more significantly also due to theuncertainty of its use in the future. In a normalEngineering Procurement Construction (EPC)cash contract, road agencies share some ofthe construction risk with the private sectorbut retain all the risk associated withinsufficient use or misuse in future (forexample overloading leading to highmaintenance costs). There is nothinginherently wrong in this traditional allocationof risk – after all, demand for a facilityproviding potential benefits far into the futureis at least significantly dependent on factorswell outside the control of the private sector;for example, country and regional economicgrowth, inter-modal competition and inflation.Even governments struggle to manage suchfactors. Moreover, on the one hand, if a roadbuilt at public expense exceeds the forecastsof traffic made prior to construction, thensociety in general is likely to be rewarded moreby that risk having been assumed by thepublic sector than in the case of private fundingwhere shareholders may corner the majorityof the benefit. On the other hand, if the publicsector does not assume a risk then society ingeneral may not reap the full rewards.

8. However, road agencies may determinethat they prefer to manage the overall risk ofthe investment by asking other parties to takeon some risks, at a price of course, whilekeeping some risks under their own control.Other parties may be better placed to controlrisks due to the prevailing incentive system orsome informational advantage. Overall costsof the investment to society may therefore belowered. The choice to reallocate risks should,however, be based on a sound analysis ofalternatives.

9. As will be argued below, the debatetoday on private funding of highways doesnot often go back to the first principles ofwhether it makes sense to use private fundsto manage a particular risk for society’s overallbenefit on a highway investment. Thedecision is driven purely by the mistaken beliefthat only private funding can make up thefunding gap.

38 World Bank estimates

30 l

Enabling Environment for PSP

10. International best practice shows that asound legal, institutional and procurementframework form part of the necessaryfoundation to enable and encourage privateinvestment in the sector. Indeed, providingan enabling environment and improving theexisting framework within which PSP canbe implemented is a key factor in mobilizingprivate participation and leveraging publicfunds to meet the financing needs of theroad sector.

11. Legislative Environment. A fair andimpartial legislative regime is an essentialprerequisite for sustainable development ofhighways projects with private sectorparticipation. Existing laws and regulationscan impose constraints for execution of suchprojects. A fair, impart ial , and non-discriminatory legal regime helps allayinvestors’ apprehensions about discriminatorytreatment under law. Investors wish to bereassured that the legal provisions in forceand applicable in relevant spheres like land,contract, property, corporate functions,taxation, labor relations will be applied in afair, equitable and objective manner andwould allow, inter-alia:

• Grant of concession i.e. transfer byGovernment of the r ight ofconstruction and maintenance ofroads to a private party;

• Levy and collection of tolls for useof roads and facilities by privateinvestors;

• Statutory backing for theConcession Agreement specifyingthe rights, duties and obligations ofthe parties involved;

• Repatriation of investment and netprofits after expiry of the Concessionperiod; and

• Safeguarding of investors’ interestsin case of change of Government/Government policy.

12. Regulatory Environment. Coupledwith a sound legal environment is the needfor fair and independent regulation. Thisplays a significant role in dispute resolution,fixation and revision of tolls and monitoring

of construction and maintenance contracts.Broadly, a regulator needs to strike a finebalance between investors’ legitimateconcerns in getting a reasonable return, thegenuine needs and requirements of usersand governmental concerns of getting valuefor money. Regulation can be achievedeither through “regulation by contract”where each contract specifies the roles andobligations of the parties involved, and/orby setting up a regulator to regulate sectorwide activities. Where a large privatef inance program is planned, theestablishment of a regulator would be apreferable option to maintain a consistentregulatory regime and ensure a level playingfield for all stakeholders. An autonomousand independent regulatory authority is thepreferred option where a Government isser ious about faci l i tat ing PSP. To beeffective, such a Regulatory Authority wouldideally be a statutory, autonomous andimpartial body with appropriate powers andfunctions.

13. Institutional arrangements forfacilitating PSP. Fragmented jurisdiction,multiple authorities, complex proceduralrequirements and lack of capacity are sitedas amongst the key deterrents for facilitatingPSP in infrastructure. Cumbersome approvalprocesses leave many important and routinedecisions to administrative authorities.Responsibilities should be clearly delegatedbetween central and local approvalauthorities and rules and regulations forapproval process should be madetransparent. Standardized processes such asinquiry and submission forms, concessioncontracts and O&M contracts facilitate PSPby introducing efficiency and transparencyin procedures. “One- stop- shops” areusually successful if there is support fromthe highest levels of governments. Providingall relevant approvals under a single statuteis one way to simplify the approval process.

The Public/Private Risk Regime

14. Efficient private participation andsuccessful private-public funding of a projectdepend upon an appropriate identification,measurement, distribution and managementof all risks associated with a project.Theoretically, risk should be properly


l31


quantified and allocated to the party that isbest able to either control or bear that risk.However, the difficulty in predicting andsubsequent management of risks makesnegotiation and contracting of privatelyfunded highways complicated, intense, andchallenging. For both the private investor andthe Government, this represents a balancingact of risks and rewards. A typical broadallocation of risk is indicated in Table 6.1.

15. The Risk-Return Relationship Agovernment can reduce overall project costsby directly assuming the risks than cannotbe assumed by the private sector at a lowerpremium. This may still result in residualcontingent liabilities for the government. Agovernment should be clear about the typesand quantum of risks it wishes to transfer tothe private sector. To the extent possible, itis in the hands of the government to: (i) use

Table 6.1: Indicative Allocation of Risks in a Road Concession Project

Type of Risk Government Private Sector

Political Risks

Expropriation of the company ✓

General modifications of the law and tax system ✓

Specific modifications of the laws and tax system ✓

Political events ✓

Termination of the contract by the government ✓

Limitation of currency convertibility ✓

Materially adverse foreign action ✓

Construction Risks

Land acquisition ✓

Cost overrun (excluding change of project) ✓

Cost overrun (change of project) ✓

Increase of financial costs ✓

Risk on schedule and quality of works ✓ ✓

Risk on administrative procedures delay time ✓ ✓

Damages incurred by the works ✓

Bankruptcy of the private company ✓

Operation Risks

Impact on the environment ✓

Force majeure ✓ ✓

Technology risk ✓

Cost overrun ✓

Change in specifications ✓

Commercial Risks

Traffic shortfall (to reference case) ✓ ✓

Price control policy (tolls) ✓

Other revenues ✓

Construction of competing facilities ✓ ✓

Financial Risks

Inflation ✓ ✓

Interest rate ✓ ✓

Exchange rate ✓

Legal Risks

Permits and licenses ✓ ✓

Litigation ✓ ✓

Source: PIARC Committee on Financing and Economic Evaluation, Financing of Road Infrastructures – Guide for New Methods ofFinancing and Public/Private Partnerships, 1999.

32 l

financial engineering techniques to reduceproject f inancing costs; ( i i ) reducesystematic risks39 with a view to minimizingassociated contingent liabilities; and (iii)manage all the non-systematic risks thatcannot be avoided through fair allocationof these risks between the public and theprivate sector. This exercise should be donewith a view to minimizing the per capitacost to society. Many non-systematic riskscan be controlled in theory by the privatesector and systematic risks can be partiallymanaged by the government. In theory, onlythose risks where the private sector has aninformational or other advantage over thepublic sector should be transferred –otherwise society may end up paying more.From an investor’s perspective, non-systematic risk must be adequately rewardedand systematic risk in its investment mustbe reasonably circumscribed.

16. Managing Highways’ Main Risk -Traffic Risk. One risk bears special mentionand will be returned to in later sections –volatility in toll sales, where they form asignificant portion of total concessionairerevenue. This is frequently the mostsignificant risk for tolled road projects, dueto the price-elasticity of demand (the higherthe toll the fewer the users and vice versa)impacting traffic levels. Unfortunately,forecasts of current and future demand areseldom exact and are often incorrect byorders of magnitude. Natural biases and aprincipal/agent problem arise in forecastingwhereby the host government could bemore optimistic than private operators; andwith the protection of limited liability,investors could be more optimistic thanlenders in their expectations. Moreover, theavailability of alternative roads and potentialconstruction of competing routes by otherpart ies, as wel l as deter iorat ion ofconnecting roads, are also major risksaffecting revenue forecast for privatelyfunded toll road projects. Although toll roadprojects frequently assume that one or bothparties to the contract can significantlycontrol this risk, this assumption may prove

a fatal flaw in the logic of many projectsunless there is a captive market such as fora bypass, tunnel or bridge.

Additional Support by the Public Sector

17. Road concessions often require varyingforms of government support, if only asinter im f inancing unti l the project’sperformance has been demonstrated toincrease the possibility of raising long-termcapital (see Annex 9 for a list of examples).This support contributes to the managementof risks. One tool that is available togovernments to adjust and fine tune thedistribution of risks between the parties isthe payment mechanism. This may takemany forms of revenue support, revenuesharing, subsidies/ grants, and subordinatedloans. Further public support to the financingof PSP programs could include r iskidentification and allocation in a project.This could include risk mitigation instrumentssuch as sovereign risk guarantees. PartialRisk Guarantees provided by multilateralbanks, for example, could facilitate privateproject financing, by covering debt servicedefault due to non-performance ofcontractual obligations undertaken by thegovernment or their agencies in roadprojects. These guarantees would catalyzefurther private finance by assisting inallocating the risks of a project to the partiesbest able to bear those risks.

18. The Risk-Return Relationship. Agovernment can reduce overall project costsby directly assuming the risks than cannotbe assumed by the private sector at a lowerpremium. This may still result in residualcontingent liabilities for the Government. AGovernment should be clear about the typesand quantum of risks it wishes to transfer tothe private sector. To the extent possible, itis in the hands of the government to: (i) usefinancial engineering techniques to reduceproject f inancing costs; ( i i ) reducesystematic risks40 with a view to minimizingassociated contingent liabilities; and (iii)manage all the non-systematic risks thatcannot be avoided through fair allocation

39 Systemic risks are exogenous factors outside the control of the investor that reflect the sensitivity of the expected return of the project in relation tothe market and overall economy. These risks are measured by Beta factors (covariance between the return of the project and the overall market, dividedby the variance of the overall market). Non-systematic risks are endogenous and specific factors to the project.

40 Systemic risks are exogenous factors outside the control of the investor that reflect the sensitivity of the expected return of the project in relation tothe market and overall economy. These risks are measured by Beta factors (covariance between the return of the project and the overall market, dividedby the variance of the overall market). Non-systematic risks are endogenous and specific factors to the project.


l 33


of these risks between the public and theprivate sector. This exercise should be donewith a view to minimizing the per capitacost to society. Many non-systematic riskscan be controlled in theory by the privatesector and systematic risks can be partiallymanaged by the government. In theory, onlythose risks where the private sector has aninformational or other advantage over thepublic sector should be transferred –otherwise society may end up paying more.From an investor’s perspective, non-systematic risk must be adequately rewardedand systematic risk in its investment mustbe reasonably circumscribed.

19. Managing Highways’ Main Risk -Traffic Risk. One risk bears special mentionand will be returned to in later sections –volatility in toll sales, where they form asignificant portion of total concessionairerevenue. This is frequently the mostsignificant risk for tolled road projects, dueto the price-elasticity of demand (the higherthe toll the fewer the users and vice versa)impacting traffic levels. Unfortunately,forecasts of current and future demand areseldom exact and are often incorrect byorders of magnitude. Natural biases and aprincipal/agent problem arise in forecastingwhereby the host government could bemore optimistic than private operators; andwith the protection of limited liability,investors could be more optimistic thanlenders in their expectations. Moreover, theavailability of alternative roads and potentialconstruction of competing routes by otherpart ies, as wel l as deter iorat ion ofconnecting roads, are also major risks

affecting revenue forecast for privatelyfunded toll road projects. Although toll roadprojects frequently assume that one or bothparties to the contract can significantlycontrol this risk, this assumption may provea fatal flaw in the logic of many projectsunless there is a captive market such as fora bypass, tunnel or bridge.

Additional Support by the Public Sector

20. Road concessions often require varyingforms of government support, if only asinter im f inancing unti l the project’sperformance has been demonstrated toincrease the possibility of raising long-termcapital (see Annex 9 for a consolidated listof examples). This support contributes to themanagement of risks. One tool that isavailable to Governments to adjust and finetune the distribution of risks between theparties is the payment mechanism. This maytake many forms of revenue support,revenue sharing, subsidies/ grants, andsubordinated loans. Further public supportto the financing of PSP programs couldinclude risk identification and allocation ina project. This could include risk mitigationinstruments such as sovereign r iskguarantees. Partial Risk Guarantees, forexample, could facilitate private projectfinancing, by covering debt service defaultdue to non-performance of contractualobligations undertaken by the governmentor their agencies in road projects. Theseguarantees would catalyze further privatefinance by assisting in allocating the risksof a project to the parties best able to bearthose risks.

34 l

Introduction

1. Internationally, there is a spectrum of risktransfer to the private sector with constructioncontracts transferring the least risk, and tollbased Design-Build-Finance-Operate (DBFO)transferring the most risk (see Figure 7.1).Although the overall rationale for choosing aparticular arrangement is to promote efficiency(including reducing the overall costs to societyof management of the inherent risk in roadconstruction), fairness and accountability,different arrangements address differentneeds. A project’s specific objectives maydetermine to a large extent the type ofarrangement to use. Specific objectives caninclude, for example: (i) to reduce public sectormaintenance responsibility; (ii) to reduce coststo the public sector through use of innovativeconstruction or design techniques and tighterrisk control; and (iii) to improve incentives forrevenue collection and reduce revenueleakage. This chapter seeks briefly to identifyhow some governments have applied thetheory noted above, most especially in thoseareas where it is believed that India can mostbenefit from such experience. It is notintended as a full account of this experiencewhich can be accessed elsewhere41.

7. INTERNATIONAL EXPERIENCE WITH PRIVATE FINANCE

Overview

2. Privately funded road projects typicallyform only a small percentage of the totalhighway network of a country and overallprivate funding flows have been modest incomparison with total flows to the sector.From 1992 to 2003, globally privateinvestment in highways had a median valueof $4.2 billion/year representing a smallfraction of total expenditure on the World’shighway network42. Except for Argentina andChile, the countries with the most active PSProad programs, most countries have entrustedless than one-tenth of their main road networkto the private sector43. Indeed, Latin Americancountries have had the highest share of theirnational roadway funded and operated by theprivate sector – the region accounted for 53%of private highway investment from 1992 to200244. For example, two-fifths of the mainroads in Chile, and about a third in Argentinaare toll roads with private participation. China,Indonesia, Mexico and Malaysia are otherdeveloping economies where there has beensignificant private investment in highways.Elsewhere, private investment has been moremodest. Nevertheless a large number of

41 Far wider coverage of this issue is available athttp://rru.worldbank.org/Toolkits/highways/documents/start.HTMhttp://www.worldbank.org/html/fpd/guarantees/assets/images/120.pdfAsian Toll Road Development Program Assessment Report, Ministry of Construction Japan and World Bank, 2001

42 http://rru.worldbank.org/ppi/43 Toll Roads - Public Policy for the Private Sector, No 224, World Bank, 200044 Source, ibid

International Experience with Private Finance

Figure 7.1

Public-Private Partnerships and the Risk-transfer Continuum

l35


countries have experimented with contractsthat allow different degrees of privateparticipation in the road sector, transferringdifferent risks and to different degrees,particularly for maintenance.

3. A review of recent experience in some ofthe Asian countries e.g. Indonesia, Thailand,Philippines, Hong Kong, and India is helpful inunderstanding how a government can promotePSP in the road sector and identifying some ofthe favorable/inhibiting factors. These factorsare explored below.

4. In the context of significant inhibitingfactors, as described above, the Governmentsneed to be sure that the motivation for usingprivate finance is properly considered andjustified. PSP projects require considerable duediligence on the client’s side as well as theinvestor’s side – both in order to establishwhether value for money is being obtainedwhen providing public support and to protectconsumers from monopolistic pricing over whatis generally a very long time horizon. Somegovernments, such as Canada, the State ofVictoria in Australia and the UK, haveprescribed a detailed process to undertake thisdue diligence (see Box 7.1). In the absence ofsuch a process, governments can expect ahigher probability of failure in PSP.

The Enabling Environment

5. Legis lat ive and RegulatoryEnvironment. Where governments havebeen serious about facilitating PSP in the sector,

they have been willing to provide a newenabling legal framework as necessary. Wherethe legal framework has been weak, PSPprograms have often faltered. For example, inMexico the 6,000 km long privately financedtoll road program has come unstuck. Part ofthe reason for this is the lack of a clear legaland regulatory institutional arrangement whichdiscouraged lenders and investors fromrespecting agreements. The independentregulatory authority also lacked sufficientcapacity so disputes were resolved in localcourts which discouraged foreign investorswary of local biases. The conflicting dual roleof Secretariat for Communications andTransport as part regulator part concessionpartner sent conflicting signals toconcessionaires45. In comparison, Box 7.2shows an example of the key legal actionstaken by Chile.

6. Publ ic Sector Planning andImplementing Capacity. One crit icalsuccess factor identified in a review of anumber of concession projects in sevencountries was the establishment of adedicated project team on the client sidemade up of experienced professionals in theareas of engineering, financing, marketanalysis, revenue forecasting and legalmatters46. The team would need to have theski l ls and autonomy to prepare andsupervise PSP projects and should not relyheavily on external technical assistance. Thisapproach is used in some Australian states

45 “Asian Toll Road Development Program Assessment Report”, Ministry of Construction Japan and World Bank, 200146 Bidding for Private Concessions. RMC Discussion Paper 120, 2001, World Bank

Favorable factors Inhibiting factors

Firm political will and commitment to the PSP Lack of an integrated institutional policy andapproach framework for project identification, feasibility

studies, project approval, etc.

Sound and growing economy, as well as a low Lack of coordination between various governmentand stable exchange rate agencies involved with road sector projects

Developed/developing capital market, capable of Absence of a statutory, autonomous, regulatoryhandling project financing through innovative authority (at arms’ length from government) forprocedures. dispute resolution, toll fixation/revision and for

ensuring a level playing field for all participants

Delays in decisions regarding government support-both in kind (land for example) and/or investment –delay in land acquisition is a bottleneck in manycountries

Absence of transparency and competition inprocurement and award processes-competitivebidding procedures not followed in many cases.

36 l

Box 7.2 Legal and Regulatory Provisions: Chile·

• Key legal provisions put in place through Decree in 1991 and then laws of 1993 and 1996,together with associated regulations in 1991 and 1997.

• Legislation created system of competitive bidding based on flexible arrangements for awardingconcessions, establishing mutual rights and setting up conflict resolution mechanism.

• Bi-lateral sole source agreements not permitted though unsolicited bids acceptable providedthere is competition in actual procurement.

• Criteria for awarding bids explicitly stated in law and regulations. Bid documents to specifyevaluation criteria.

• Three person conciliation commission to settle all disputes originating from the interpretationand implementation of the concession contract consisting of one government nominee, oneconcessionaire nominee and one agreeable to both parties who acts as chairman. Thecommission can also act as an arbitration tribunal in the event that conciliation is not successfulwithin 30 days

Source: Toll Road Concessions, the Chilean Experience, PFG Discussion Paper 124, World Bank, 2003

Box 7.1 The Public Sector Comparator

In early 1990’s, the UK government enacted legislation to guide, facilitate and finance the developmentof Compulsory Competitive Tendering, Private Finance Initiatives (PFIs) and Public-Private Partnerships(PPP). Under this legislation, the UK Treasury ensures that any initiative contemplating a PFI or a PPPdevelops a Public Sector Comparator (PSC) based on a highly prescribed process. The motivation forenforcing the use of a PSC was poor performance in earlier PFI/PPP projects, union pressure and a strongTreasury commitment to accountability and transparency in public spending.

The PSC is a method used to calculate the “in-house” cost of delivering a project and/ or service andcomparing this with a private finance option. It serves the following purposes: to determine if the projectis affordable to government by ensuring full life cycle costing at an early stage; to test whether a PPP isviable and demonstrates value for money; to communicate with partners on such key aspects as outputspecifications and risk allocation; and to encourage broader competition by creating greater confidencein the bidding process.

A typical PSC includes:·

• An estimation of the basic costing including capital and operating costs;·

• The approach taken in relation to third party revenues;·

• The approach taken on asset values on transfer, disposal and termination of the contract;·

• A risk matrix, showing the various sources of risk, their costs, the likelihood of their occurrence,and the consequences for the project;·

• A discounted cash flow forecast; and·

• Sensitivity analyses, showing the consequences of varying key assumptions.

The PSC commences as a high level document to gauge internal acceptance and then evolves into adetailed assessment. The responsibility of designing and implementing the PSC remains with the publicsector agency involved in the particular project, although private sector advisors can provide support.The signing authority for a PPP has to satisfy him/herself that the PSC has been properly conducted andthe PPP option found preferable.

Source : http://www.ogc.gov.uk/sdtoolkit/reference/ogc_library/PFI/series_3/technote5/5tech_contents.html

47 Designing Toll Road Concessions Lesson from Argentina, Public Policy for the Private Sector, World Bank, 1996

and was used in some early transport projectsin the UK. The Philippines established a “One–Stop-Action Center” to help investorsunderstand the process of obtaining consentsand permits. One analysis of the mixed successin Argentina stated that a key lesson learnedwas that institution building must be taken

seriously if poor and slow decision making wasnot to plague the process47.

7. Project Procurement. Experience hasshown that to maximize interest from qualifiedbidders, governments need to adopt a strategydesigned to reduce the costs of tendering andrestrict the number of bidders in the final round


l 37


48 Source. Bidding for Private Concessions. RMC Discussion Paper 120, 2001, World Bank

to only three or four. The tender procedurealso needs to be undertaken openly andswiftly and bidders should have to respondto a c lear set of requirements andspecifications covering the commercial,financial and technical aspects of theproject48. A number of different criteria havebeen used to evaluate proposals includinglowest toll rates, shortest concession period,lowest subsidy to be provided by the state,highest payment to be made by theconcessionaire and least present value ofrevenue levels. Experience demonstratesthat evaluation of bids should ideally be onthe basis of a single criterion to providetransparency and prevent subject ivetradeoffs during evaluation.

The Management of Risks

8. Volatility of Toll Sales. In formulatingthe procurement system and associatedpreferred risk allocation, the risk associatedwith traffic demand is the most difficult todeal with. Some countries introduced tollsfor publicly financed roads first, paving theway for a toll paying culture amongst users,in order to facilitate the later developmentof private sector funded toll roads with fulltransfer of traffic risk (see Box 7.3 on Korea).Revenue risks could be mitigated for privatefinanciers of tolled projects by the hostgovernment through the use of minimumrevenue guarantees (which may be coupledwith upside revenue sharing).

9. Pr ivate f inanciers can be moreeffectively shielded from public resistance

to tolling yet made accountable for theiroperational performance through the use of“shadow tolls”. In earlier UK DBFO projectsand in countries such as Holland, Norway,Portugal, Poland and the Czech Republic,payments have been made by the state toconcessionaires based on traffic levels(shadow tolls) or availability of the road.These payment mechanisms reduce thedemand risk that otherwise exists on realtoll roads. Shadow tolls may be set in“bands” so that at low traffic levels the toll(government payment) per vehicle is higher,and the average toll (government payment)decreases as the number of vehicles increase(one version with a fixed minimum paymentis shown in Figure 7.2). This enables biddersto develop expectations of revenue whichare more robust compared with real tollroads for a number of reasons. First revenuesare not subject directly to elasticity ofdemand. As the government is making thepayment per vehicle rather than the user,there is no overt disincentive for the usernot to use the facility (they are not paying atoll at the point of use). Second, downsiderisk, where traffic volumes are low, ispartially mitigated by banding of rates withhigher rates at lower traffic levels. Thus, ashadow toll mechanism is more attractiveto lenders as this reduces the volatilitycaused by real tolls.

10. Protection from competing facilities andinadequate connect ing roads var iesbetween countries and road programs, butit is considered equitable for this risk to be

Figure 7.2: Two Tier Shadow Toll Variations

38 l

principally borne by the state and onlyincidentally by the private sector. Thecontracting authority may not have directresponsibility for the competing route orconnecting road (often a local authorityasset), but clearly it has more influence overthe local authority than the concessionaireand therefore is best placed to manage thisrisk. The outcome of transferring this risk tothe private sector is, at best, higher financingcosts and/or lower concess ion feesreceivable by the government, and, atworst, a deal that is not bankable.

11. For real tolled facilities, financiers need toanticipate the revenue volatility anduncertainty caused by real tolls and base theirfinancial projections on some of the worstdownside characteristics. Experience showsfinanciers are not always adept at managingthe demand risk, as evidenced by theconclusions of a recent review of cancelledprivate infrastructure projects between 1990and 2001 from around the globe49. About 6%of toll roads (mostly in Mexico and Hungary)were cancelled during this period representingabout 16% by value of the private investmentin the sector. The one factor that contributedmost significantly to these cancellations of tollroad projects was that roads could not attractenough users to meet optimistic trafficforecasts. Further, the authors conjectured that

where governments were willing to assumesome or all traffic risks, investors due diligenceon demand forecasts may have been lessthorough.

12. Overall, it is debatable whether it makessense to transfer traffic risk substantially to theprivate sector apart from in cases such asbridges, tunnels and bypasses where trafficdemand can be better assessed and managed.The Public Accounts Committee that scrutinizesall public expenditure in the UK had this tosay about four DBFO projects procured duringthe early 1990’s where shadow tolls were thepayment mechanism.

“Departments should consider carefully theimplications of basing payments to operatorson volumes of activity over which neither thepublic sector nor the operators have anyeffective control. In the case of these fourcontracts, payments to operators are basedprimarily on traffic volumes which are,however, notoriously difficult to forecast. Inother words, the Agency have created a riskwhich is borne by the operators and whichcan be expected to have increased their costs.PFI can deliver better value than traditionalmethods of procurement if risks are transferredto the parties best able to handle them. But itis a mistake to confuse risk transfer with riskcreation, which is simply likely to increase coststo the taxpayer.”

49 Infrastructure Projects – A Review of Cancelled Private Projects, Public Policy for the Private Sector, Note No 252, World Bank, January 200350 Of course, any revenues accruing to Government from indirect road user charges will add to the total burden on road users.

Box 7.3 Toll Pooling

Japan: Since 1972, tolls from expressways and other regional highways have been pooled through theJapan Highway Construction Corporation. The tolls are set according to the following principles: (i)redemption - toll roads must pay off their debts; (ii) benefit - toll rates must not exceed the benefit theusers receive; and (iii) just-and-fair principle - toll rates charged must be set at ‘just and fair’ levels.

Korea: The Korea Highway Corporation, 82% Government owned, was established in 1969 and isresponsible for the development, maintenance and operation, including tolling, of about 2,100 km of theKorean highway network. The roads constructed by the Corporation were funded initially by Governmentsupport and debt on the Corporation’s balance sheet. Tolls were earlier set to cover the anticipated O&Mcosts. Now toll rates are determined on the basis of the benefit to user principle. In 2000, the Corporationraised $1.43 billion in tolls. Over the period from 1995 to 2006, it is expected that tolls will raise about31% of total agency revenues, loans/bonds about 42%, Government support 22% and other revenues5%. Of this 36%, 22% and 42% will go towards debt repayment, maintenance and new constructionrespectively. This means that toll receipts are covering not only O&M but also construction, either by wayof repayment of existing debt or in actual new investments50. About 25% of its staff are devoted to tollcollection. More recently, Korea has started to concession some more heavily trafficked existing and newhighways.

Source:Kengo Mizuno, “Lessons Learned: Public-Private Partnerships on Japan’s Mega-projects,” PW Financing,January 1996, pp. 1-7http://www.freeway.co.kr/eng/html/Corporation/sub01_01.html


l 39


51 http://www.macquarie.com/uk/about_macquarie/media/20030113.htm

13. In more recent DBFO projects, the UK nolonger uses shadow tolls as the paymentmechanism, but rather pays according to laneavailability and active congestion/safetymanagement.

14. A Possible Solution - Toll Pooling. Oneoption that has been practiced in a number ofcountries with some success is theestablishment of a uniform publicly managedtoll system (see Box 7.3). The pooling of tollrevenues allows for cross-subsidization oflower trafficked routes, efficiencies in tollcollection, a more gradual increase in thelevel of tolls to build user acceptance anddiversification of traffic risks. It also removesa large portion of the political risk ofinterference in a toll rate that had earlierbeen agreed between a government and aconcessionaire. The government is then freeto choose between a more or less directuser charging regime to meet public policyobjectives at the same time as managepublic opinion. Many European countriesseem to moving towards this concept. Forexample, Switzerland and Austria haverecently implemented additional networkwide weight distance toll systems recoveringnew funds from the trucking industry. The tollreceipts are pooled and then hypothecatedtowards highway improvements andmaintenance. A Global Positioning System(GPS) based weight/distance toll system inGermany is also to be made operational nextyear. This is designed to raise dedicatedrevenue from truckers for six-laning themajority four-lane motorway network.

15. A Portfolio Approach for RiskDiversification. Similar to the publiclymanaged toll pooling approach mentionedabove, the private sector can also diversify risksthrough adoption of a portfolio investmentapproach adopted. An example of thisapproach is Road King Infrastructure Limited(RKI) of Hong Kong (RKI). As of 2002, RKI hada portfolio of 22 toll road projects in mainlandChina covering about 1,000 km, mainlyoperating with joint venture partners for specificprojects. RKI leverages the creditworthiness andthe track record of its portfolio of highwayprojects in China, thereby diversifying its

investment portfolio into various regions (highgrowth centers) and risk profiles. A similarapproach is being applied by Australia’sMacquarie Bank and the Korean lenderShinhan Financial Group, joint managers ofthe Korean Road Infrastructure Fund (KRIF).KRIF was set up as a 10-year closed fund byMacquarie and Shinhan in January 2003,raising 247 billion won (about $200 million)from Korean institutional investors. It has sinceraised a further 120 billion won. It made itsfirst purchase in March 2003, spending 177.8billion won on a toll road in the southern cityof Kwangju and is planning to invest in anothersix toll roads in Korea. In future, the intentionis to list the KRIF and thereby tap the retailinvestor market51.

16. Because the useful lives of most roadassets are quite long, the most appropriatelong-term capital is provided by insurancecompanies, pension funds and similarinstitutional investors. Listed companies thatinvest in a portfolio of highways with proventraffic trends can provide the required interimfinancing for highway projects until theoperating performance of these highways havebeen sufficiently demonstrated for the projectto receive an investment grade rating and toobtain long-term capital.

Additional Public Support

17. Loans are the most common form of publicfunding. Generally, these are low-cost funds,raised by the issuance of government-guaranteed bonds or long-term, low interestor interest-free loans supplied from general orearmarked accounts. In the US, public toll roadauthorities issue debt instruments known as‘revenue bonds’, which are serviced entirelyby the toll revenues collected by the issuingauthority. The tax-exempt status of theseinstruments has been a key consideration forenhancing investor interest in such instruments.Initially, the respective US state governmentsguaranteed toll road bonds. However, externalcredit enhancement measures are generallynot required, as the credit worthiness of theissuing authorities is now established.

18. In France, limited access, grade-separatedinter-urban expressways, known as‘Autoroutes’, have been developed over the

40 l

last 30 years. Such expressways account foronly 4% of the national road network, andcarry 40% of the road traffic. As of 1997, thetoll road network in France was around 6,700km, involving an investment of about $28billion. 72% of the total motorway network inFrance is tolled. In the initial stages, theGovernment extended concessional finance upto 70% of the project costs to privateconcessionaires; 52% from the CentralGovernment and 18% from local governments.The balance was met through equitycontribution of private concessionaires. In thesecond phase of development, governmentassistance was less than the 1960’s, withconstruction advances (interest-free loans)being offered only for segments that theGovernment deemed unprofitable52.

52 Source: a. Caisse Nationale des Autoroutes, Annual Report of the Board of Directors, 2001.b.http://www.unescap.org/tctd/pubs/files/econreg_chII.pdf53 Kengo Mizuno, “Lessons Learned: Public-Private Partnerships on Japan’s Mega-projects,” PW Financing, January 1996, pp. 1-7

19. Japan has one of the longest toll-roadnetworks in the world (9,200 km). Nearly two-thirds of the 12,700 km of trunk roads andexpressways (of the total road network of morethan 1 million km) are tolled. Roads wereinitially funded through public meansdependent on a Road Improvement SpecialAccount. This drew resources from tolls, centraland local taxes on fuel and freight transport.Later, highway projects were promoted by largeconstruction companies with subsidized debt.Such companies inject the equity, usually about10% of the funding requirement. Senior debthas been provided by Japanese commercialbanks, regional banks, or trust companies,with some participation by the DevelopmentBank of Japan. Debt has been underwrittenby Government. There is no sign yet of ‘pure’financial investors or of bond issues53


l 41


Policy and Legal Environment

1. Government Policy for PSP. Theoverall objectives for the development ofroads in India are laid down in the TenthPlan (Box 8.1). This plan recognizes that

8. INDIAN EXPERIENCE WITH PRIVATE FINANCE

to achieve the projected GDP growth, theindustrial sector needs to grow at morethan 10% per annum with a resultantincreased demand for transport services.An increased role for the private sector inmeeting this demand is explicitly sought.

54 Amended in 1977, 1992, 1995 and 1997.

Box 8.1 Transport Policy Goals in the 10th Plan

The following goals are made in the Tenth Plan:·

• Meeting the transport demand generated by higher rate of GDP growth;

• Ensuring transport development is balanced, with special attention to remote regions such as theNorth-East;

• Capacity augmentation, quality and productivity improvements through technology upgradation;

• Higher maintenance standards to reduce the need for frequent reconstruction;

• Higher generation of internal resources and increased private sector participation in providingtransport services;

• Increase efficiency by bringing competition in the provision and maintenance of infrastructure andservices wherever possible;

• Higher emphasis on safety, energy efficiency, environmental conservation and social impacts; and

• Developing an optimal inter-modal mix where each mode operates efficiently, according to itscomparative advantage, and complements services provided by other modes.

Source: Para 8.3.18 of Draft 10th FYP, Planning Commission

2. National Legislative Changes. The1956 National Highways Act has beenamended54 to permit private entrepreneursto undertake NH projects on a BOT basisand recover their investments throughto l l s . Under th i s Act , a s imp l i f i edprocedure was prescribed for acquisitionof land for the building, maintenance,management or operation of a NationalHighway, and separate provisions weremade for the levy and collection of feesin respect of both public and privatefunded projects. The GOI has framed asupporting policy (Box 8.2).

3. State Legal Framework for PSP.The 1851 Indian Toll Act makes it possiblefor State Governments (SG) to levy and

collect tolls on any road or bridge whichhas been made or repaired at the expenseof the central government or any SG.However, the Act needs to be amendedby respective SGs to allow the privatesector to levy and collect tolls on Stateroads and bridges. Some SGs have indeedamended the Act – for example UttarP radesh and Madhya P radesh - o rotherwise taken legal steps in order topromote PSP (see Table 8.1).

4. In addition to amending the IndianToll Act, another avenue being followedby some States is to enact a uniform lawfor infrastructure development. Enactedlaws include the 2001 Andhra PradeshInfrastructure Development Enabling Act

42 l

Indian Experience with Private Finance

Box 8.2 Main Features of GOI Policy for PSP in the Roads Sector

• Foreign Direct Investments up to 100% is permitted.

• NHAI can participate with up to 30% of the total equity of a company floated to develop a road project.

• NHAI can provide capital grants to the developers of a road project on a case by case basis.

• Suitable traffic support/guarantee is provided on a case to case basis.

• Real estate development can be made an integral part of BOT projects to enhance their financialviability.

• Entrepreneurs to be protected against force majeure situations including political, non-political andlegislative changes.

• Dispute resolution of arbitration would be under the 1996 Indian Arbitration and Conciliation Act,incorporating United Nations Commission on International Trade Law (UNCITRAL) provisions.

• The ownership of the land for highway construction and roadside facilities continues to vest inGovernment. Mortgaging of such land is not allowed. However, land can be leased to entrepreneurs.

• Unified check barriers at the inter-state borders to be provided. Such barriers would be located outsidethe right-of-way with proper entry/exist layout.

Source: Road Development Report Vision 2021, MORTH, Government of India (as amended)

55 See http://www.aproads.com/html/salient.htm for more details56 See http://www.gsrdc.com/htmls/index3.htm for an extract of the full Act

(IDEA) and the 1999 Gujarat InfrastructureDevelopment Act (GIDA)56. The latter is agood example of how SGs are seeking tocircumscribe risks for private parties. TheAct is very detailed and provides for several

private sector participation options. The Actalso provides for the use of an opencompetitive bidding process based on pre-qualification, a technical evaluation, and afinancial evaluation, and lays down the criteria

Table 8.1: Legal Provisions to Encourage PSP in Six Selected States

State Legal Provisions

Andhra Pradesh The SG’s publication ‘Private Investment in Road projects in Andhra Pradesh’, proposesto amend the AP State Motor Vehicles Act along lines similar to the National HighwaysAct to enable private parties to levy tolls and regulate traffic on the facility constructedby them.55

Issued a policy framework for private participation in Roads Sector through aGovernment Order in September 1997.

Karnataka Not defined

Madhya Pradesh 1992 Amendment to Indian Toll Act allows the SG to lease the levy tolls on roads orbridges, at prescribed rates, by public auction or private contract for up to 15 years.

Maharashtra Amended Section 20 of the 1958 State Motor Vehicle Tax Act to enable privateparties to levy tolls on any motor vehicle on bridges and tunnels through a BOTagreement.Section 17(4) of the 1975 MMRDA Act provides for charging tolls for theuse of amenities provided by MMRDA.

Uttar Pradesh 1974 Indian Toll Amendment Act transfers State responsibilities to the U.P. StateBridge Corporation Ltd., a State Government Company under the Companies Actwith the right to manage and collect the tolls levied on roads and bridges. Section 2of the same Act permits the SG and the U.P. State Bridge Corporation Ltd. to leasethe right to collect tolls to any third party.

West Bengal Legal provisions relating to levy of tolls: (i) on heavy trucks and buses under the 1993West Bengal Municipal Act, as amended in 1997; (ii) on roads and vehicles under the1980 Calcutta Municipal Corporation Act; (iii) on the Howrah Bridge under the 1926Howrah Bridge Act; and (iv) on all types of vehicles and animals passing over orthrough any bridge constructed under the 1969 Hooghly River Bridge Act.

l 43


for financial evaluation. It allows: (i) the projectdeveloper to charge fees as specified in theconcession agreement which can be revised,based on criteria specified in the agreement;and (ii) inflation and variation in foreignexchange rates to be taken into account forsuch revision of fees. The Act identifies severalways for the State Government to provideassistance for a project as well as a frameworkfor termination of a concession.

5. Securitization. Infrastructure financing,being highly capital intensive, does not oftenpermit the possibility of a security structure thatallows lenders full recourse to the assets ofthe promoters. Hence, a ‘non-recourse’ or‘limited recourse’ arrangement is more likelyto be adopted, involving the creation of alegally independent project company financedwith debt that is secured by the assets of thatcompany (e.g. land, industrial assets, licenses,and contractual rights). In India, the immovableassets (e.g. land, machinery, buildings) of theproject company are usually secured in favorof the lender by way of a mortgage while themovable assets are ‘hypothecated’. Thecreation of security by leasing out land for roadprojects is not possible as its ownership restswith the Government. In cases of default byborrowers, the 2002 Securitization Act57

enables a bank or a financial institutionwhich is a secured creditor under a securityagreement with a borrower to enforce anysecured interest created in its favor withoutthe intervention of any court or tribunal. Thisrepresents an important step forward whichallows lenders to avoid lengthy and timeconsuming litigation thereby reducing creditrisk for lenders. Similarly, a provisionincorporated into the standard ConcessionAgreement used in India allows the lender,in the event of a default, to continue theimplementation of a project.

6. Environmental Clearances. Actionshave also been taken to reduce the risk ofdelay from environmental clearances. Underthe 1986 Environment (Protection) Rules, theMinistry of Environment and Forestry(MOEF), via Notification from January 1994,states that highway projects shall not beundertaken anywhere in India unlessenvironmental clearance has been granted

by the Central Government. In view of thepronouncements by the Supreme Court incertain public interest litigation cases involvingenvironmental issues, the NationalEnvironment Appellate Authority was set upunder the 1997 National EnvironmentAppellate Authority Act. This serves as anindependent body to deal with petitions,complaints, representations and appealsagainst the decisions of competent authoritiesgranting environmental clearances todevelopment projects under the provisions ofthe 1986 Environment (Protection) Act. Byproviding for a focus of independent decisionmaking on the issue, the creation of theAuthority represents an important step forwardin facilitating private participation in roaddevelopment.

7. Arbitration and Conciliation. The 1996Arbitration and Conciliation Act brought aboutmajor changes in Indian Arbitration law byincreasing investor confidence in the systemof commercial dispute resolution andenforcement of foreign awards in India. Thiswas done by removing obstacles to commercialdispute resolution outside of a court andcreating an environment with minimum courtinterference.

8. Fiscal Incentives. The Government ofIndia announced various tax exemptions andfiscal incentives to facilitate PSP in severalsectors including roads and highways (Box8.3).

Progress with Private Sector Participationat the National Level

9. Nearly 80% of all the projects executedunder the NHDP are still EPC cash contractswhereby the contractor builds the road andhands it over to NHAI. The remainder are beingfunded through a mixture of PSP models. Thedifferent models demonstrate a sliding scaleof risk management by the NHAI as shown inFigure 8.1 below. In general terms, high-densitystretches with more assured returns arenormally awarded under toll concessions, whileSPVs and annuity concessions are used forprojects that have higher traffic risks.Approximately 8% of the length58 of the NHDPmay be implemented through annuityconcessions. In tandem,NHAI has awarded

57 2002 Securitization and Reconstruction of Financial Assets and Enforcement of Security Interest Act (previously Ordinance).58 The actual stretch of NHDP Phase I & II commissioned under annuity scheme is 8%.

44 l


around Rs. 3,400 Crore worth of projects undertolled concessions.

10. BOT Toll Road Concessions. BOT TollRoad Concessions awarded by NHAI are basedon a model concession agreement withmodifications for each specific case59. Thebroad risk allocation framework is, therefore,common for NHAI toll road concessions, withsome slight differences reflecting increased andmore equitable risk sharing between NHAI andthe private investors (see Annex 10). Generally,there has been an increasing amount of riskand responsibility sharing with theConcessionaire (see Table 8.2 below).

11. On a pilot basis, NHAI has also proposedto let a concession agreement which combinesthe construction of a greenfield project withthe rehabilitation and upgrading of an existingstretch60. This would address the problem thatthe new stretches have low traffic volumesmaking them commercially unviable. Existing

stretches could generate enough toll revenuebased on existing traffic patterns to improvethe cashflows of the concessionaire, especiallyduring the construction stage. In addition,overall risks would be mitigated to some extentbecause of pooling of projects at various stagesof risk profile.

12. The typical grant-based BOT model usedin India on National and State highwaysinvolves: partial funding of project cost througha non-refundable grant disbursed by thegovernment/ road agency to theConcessionaire during construction; a fixedconcession period and toll rate structure;“grant” required by bidders constitutes the bidvariable; grant contribution by the governmentnormally does not exceed the amount of equitycontribution by the private investor; and theremainder of project cost is typically metthrough non-recourse debt.

13. NHAI is currently in the process of

59 See http://www.nhai.org/concessionagreement.htm for copy of the standard form60 The proposed project is part of the Porbander-Deesa segment of the East-West Corridor. New construction refers to the Jetpur-Gondal section (26 Km

of four lane highway) and the Rajkot Bypass section (10 km), and the rehabilitation component refers to a four-laned stretch of the 32 Km Gondal-Rajkotsection.

Box 8.3 Summary of Fiscal Incentives for Private Sector Participation in Roads·

• Exemption is given from import duty on identified high quality construction plant and equipment.·

• A ten year tax holiday is available and this concession can be accessed in any 10 consecutive years duringthe first 20 years of operation (Section 80 IA (12) (ca) of the 1961 Income Tax Act).·

• A maximum of 40% of the income derived by financial institutions from the finance they provide forinfrastructure projects may be excluded from Income Tax.·

• Exemption from Income Tax on the income from dividend and interest on long term capital gain derivedfrom investment in the form of shares or long term finance to any enterprise set up to develop, maintainand operate an infrastructure facility.·

• Subscriptions to equity shares and debentures are eligible for deduction under Section 99 of the IncomeTax Act.

Source: Road Development Report Vision 2021, MORTH, Government of India (as amended)

Figure 8.1: Risk Allocation in Different NHAI Models

l 45


concessioning the four-laning of a further10,000 km of national highways outside ofthe GQ and NS/EW networks. Initially theseroads had been under the charge of theMORTH but were transferred during late2003 to NHAI. The program is still underpreparation and the details of how this willwork are not yet known. However, currentthinking is that the majority of approximately48 projects would be procured through atoll based BOT mechanism with less viableroutes being procured through conventionalcash contracts or annuities. The process ofempanelment of suitable concessionaireshas begun. Public support will be capped at40% of project cost, 25% duringconstruction and 15% during the O&Mperiod of approximately 15 years.

14. Annuity Concessions. The Annuityscheme is a variant of a BOT Toll Concessionscheme whereby the concessionaire is not left

to recover its costs through collection of tollsbut is assured a fixed periodic payment by theNHAI. The Annuity scheme offers theadvantage as a means of “borrowing” fromthe private sector for road developmentthrough bundling together construction andlong term O&M responsibilities onto a singleparty. To date, NHAI has awarded a total of467 km of annuity concessions amounting toabout Rs.3,000 Crore - five stretches (with atotal length of 373 km) in the GoldenQuadrilateral and three other stretches (81Km), with payments partly related to shadowtol l ing (Durg Bypass, Nandigama-Vijayawada Phase II and Delhi-GurgaonExpressway) – see Table 8.3.

15. As almost all projects structured as annuityconcessions have reached financial closure,this would suggest that this method is popularwith private investors in India.

Table 8.2: Progression in Risk Sharing Arrangements in NHAI Toll Road Concessions over Time

Construction Related Delay Risk Progressively has been made the entire responsibi l ity ofConcessionaire. Concessionaire has choice of re-allocation toEPC contractor, but remains primarily responsible to NHAIthrough liquidated damages. Liquidated damages have replacedearlier corporate guarantee of promoters for timely completion,making the model more “non-recourse” in nature.

Compulsory surveys and tests In latest models, left to the discretion of Concessionaire

Independent Consultant Concept introduced in later concession arrangements, withselection process becoming more participative (i.e. involvingConcessionaire as well)

Traffic Risks Compulsory routing of traffic through project road not observedin latest concession, with traffic retaining full choice.At thesame time, protection against development of alternativeprojects is provided, with assurance of levying higher tolls onalternative competing facilities.

Promoter lock-in Greater flexibility over time, with latest model allowing higherlevels of promoter divestment.

Lender Protection Slight reduction in amount of debt protection in case ofconcessionaire events of default, probably with the steadyincreasing lender comfort and appetite related to these projects.

Change in Law In later models, compensation for change in law to be triggeredonly in case of adverse effect exceeding a certain floor amount;this is an incremental improvement which avoids complicatedadjustment procedures for minor impacts, although it impliesthat Concessionaire has to bear change in law risk for smallamounts.

Termination Payments for Increasingly refined formula which takes into account effectsForce Majeure of inflation and passage of time.

NHAI support in case of poor Originally, revenue shortfall loan envisaged to enable projectproject performance subsistence in case of revenue shortfall for any reason, however,

in subsequent models, such loan is provided only in case ofsevere shortfalls due to political force majeure.

46 l


61 The weighted average cost of debt for NHAI is estimated as 11.35%, using NHAI’s debt structure for 2002-03 (55,930 million in market borrowings at anaverage cost of 8.5%; Rs. 78,620 million as external assistance of which 20% was at the cost of 13%; the rest 80% as a grant; and Rs. 23,540 million ofAnnuity Projects at a cost of 17%).

62 The sponsor’s total exposure would be a combination of equity invested and contingent liability to the amount of guarantee issued.

16. Private Investment Premium forAnnuity Contract. The effective cost ofannuity borrowing for NHAI is high (effectiverate of 17-18% p.a., compared to NHAI’s costof funds which are around 11%. p.a. 61).Individual factors leading to higher costs arediscussed below.

• Risk Premia for Risks passed on toPrivate Investor. Under the annuityscheme, the private investor isrequired to assume various risks thatwould normally remain with NHAI,many of which can significantly affectthe investor’s cashflows vis-à-visoriginal projections. Hence, investorsusually factor in premia to price theserisks in their financial proposals.These risks include: (i) risks relatedto price and quantity variations; (ii)changes in financing rates; (iii) long-term inflation; and (iv) constructiondelays.

• Cost of Debt. This is higher for non-recourse lending to project companiesthan the cost of a loan to NHAI, dueto its current creditworthiness. Inaddition, the prevailing practice offinancial intermediaries in India is toinsist upon the creation of a “DebtService Reserve” (DSR) for additionalcomfort during the tenure of their non-recourse loans. Adding the DSR to

total project cost increases the capitalservicing requirements therebyadding considerably to the annuitypricing. In essence, the borrowingcompany is required to build upan investment reserve, usuallyequivalent to the next six months’worth of lender’s dues, in an accounton which the lender has a lien. Incase of default on debt obligationsfor any period, the lender would beentitled to recover dues for that periodfrom the debt service reserve. DSRacts as a buffer to ensure smoothpayments of interest and principaleven during brief periods of cashshortfall within the project company.DSR can be created: (i) out of totalproject cost, in which case it is fundedthrough a combination of debt andequity, raising the total project cost;or (ii) through a guarantee providedby the bankers to the projectpromoters, increasing the sponsor’soverall exposure to the project62.

• Cost of private equity investment inthe project is naturally higher thanthe NHAI’s cost of capital estimatedat around 11%.

17. Regulatory Inefficiencies in AnnuityContracts. Although a project may showhealthy cash-flows, net return in the hands of

Table 8.3: Annuity Concessions Structured by NHAI

Road Annuity Amount, Rs. Million

Golden Quadrilateral

Panagarh – Palsit 5,550

Palsit - Dankuni (Durgapur Expressway) 3,998

Maharastra Border-Belgaum 5,050

Ankapalli - Tuni 2,948

Tuni – Dharmavaram 2,791

Dharmavaram – Rajamundry 2,962

Nellore Bypass 1,296

Other Projects

Tambaram - Tindivanam 4,186

Total 28,781

l 47


equity holders of the project company is oftenmuch lower than the profitability of thecompany may suggest, owing to acombination of statutory reserves, deferred taxliability and taxation inefficiencies.

• Statutory Reserves. According toCompany law, corporate entities arerequired to set aside a sum of up to10% of book profits every year, intoa statutory reserve which is notaccessible for distribution to share orequity holders. This results inconsiderable cash sitting idle for a longtime.

• Taxation Inefficiencies. From theGovernment’s point of view, incometaxes and taxes on dividendsdistributed raise the effective cost ofthe annuity concession, because it hasthe effect of increasing the effectivecost of annuity “borrowing” vis-à-visa conventional loan. This is due to: (i)a limited period of tax holiday forinfrastructure companies underSection 80 IA of the Income Tax Act(10 years during operations); (ii)distribution tax on dividends paid byproject companies with effect fromApril 2003; and (iii) minimumAlternate Tax which forces companiesto pay tax even during the tax holidayperiods. Taxes paid by the projectcompany represent monies that aresourced from the Central and StateGovernments and paid back to theCentral Government.

18. Impact of Annuity Concessions onRisk and Project Costs. For the privateoperator annuity concessions reduce theuncertainty and risk associated with BOTsin relation to revenue generation, but theassociated regulatory requirements notedabove raise the effective cost of an annuityfor NHAI. As all annuity concessions are stillunder operation, it is not possible to inferthe actual cost of these concessions.Nevertheless, using a financial model of atypical annuity concession, it was concludedthat the l ikely cost of the annuity“borrowing” from the government’sperspective due to the factors noted abovewould be around 17% per annum.

19. Special Purpose Vehicles (SPV). The SPVroute essentially involves setting up a jointventure company between two or morepromoters to execute the project. These aregenerally toll roads implemented inindependent corporate entities owned by NHAIitself and are commonly used for portconnectivity where other user entities cancontribute to project funding. Appropriate long-term capital is usually provided by insurancecompanies, pension funds and similarinstitutional investors. These investors usuallyrequire the investment to have an investmentgrade rating from at least one credit ratingagency. Participation by state governments isalso significant in these projects. Moradabadexpressway was developed through this route.Other projects under implementation are JaipurBypass Phase II, Ahmedabad-Vadodaraexpressway and 10 port connectivity roadprojects.

20. NHAI Debt. Based on its revenues fromthe Central Road Fund and increasingly fromtolls as well as on an implied GOI guarantee,NHAI has itself raised substantial amounts offunds from the domestic debt markets with anominal interest rate between 7-9%. Thecredit rating of NHAI is AAA and stable but issusceptible to the following factors – fungibilityof cess inflows and their continuation,prioritization of debt servicing, retention of tollsand support from GOI63. The credit rating statesits assumption of stable revenues from theCentral Road Fund to NHAI.

21. To date, NHAI has raised Rs 656 Crorefrom two bond issues in 2000-01, Rs. 804 Crorethrough one issue in 2001-02 and Rs. 5,593Crore in 2003. Investments in NHAI bonds havebeen exempted from capital gains tax underthe Income Tax Act Section 54EC and havebeen oversubscribed. In addition, NHAI hasapplied for a Rs. 6,000 Crore line of creditfrom the Life Insurance Corporation. Theagreement envisages securitization of cess onpetrol and diesel, put aside in the non-lapsableCentral Road Fund. The interest on the LifeInsurance Corporation (LIC) loan has beenproposed at a premium of 1% on 18 yearGovernment Securities (currently at 7.35%) fora loan duration of 25 years with repaymentsstarting after 5 years. This is the only NHAI

63 Source - http://www.crisil.com

48 l


debt instrument that is proposed to be explicitlybacked by the GOI at a fee of 0.25% perannum.

Progress with Private Sector Participationat the State level

22. There is a wide and growing dividebetween some States that are making realprogress in attracting PSP to the sector, whilethe majority of States have little or noexperience and seem unlikely to do so forthe foreseeable future (see Table 8.4).

23. Madhya Pradesh, for example, hasmade good progress concerning PSP in roadshaving entered into a number of ‘Maintainand Transfer’ concessions with the privatesector. The States of Andhra Pradesh,Gujarat, Maharashtra, Madhya Pradesh,Rajasthan and Tamil Nadu have alreadyentered into BOT concession arrangements,each using a fairly standardized model forBOT projects (see Annex 11 for the riskallocation framework for small road projectsin Andhra Pradesh). As a typical exampleof how these work at the state level takeMadhya Pradesh. The MP Br idgeCorporation uses a grant-based BOT modelin which up to 50% of a BOT project’s costis provided by the Corporation in the form of agrant to enhance commercial viability. Thegrant amounts required for implementation of14 key state highway stretches on this basishave been raised through borrowings largelyfrom HUDCO. The amounts thus raised areaccordingly leveraged through individualproject companies.

24. Project Development. Some Stateshave set up Project Development Funds, suchas the Andhra Pradesh InfrastructureInitiative Fund to finance the preparation ofroad projects for private involvement. Whilethe concept is good and should help facilitatePSP, such funds have not yet played asignificant role.

25. Identification and Award of PotentialPSP Projects at the State Level. Some statesidentify road projects for development andmaintenance through Strategic OptionsStudies commissioned by the respective roadagencies. However, states are yet to have apolicy setting the basis for selection andprioritization of projects for implementation

through PSP and in general, states have notyet adopted a systematic procedure for projectidentification and investment facilitation. Inpractice, routes with higher traffic density andof general strategic importance are the onesprioritized for development by privateinvestors. States that have attracted BOTs haveadopted a two-part bidding procedure,involving invitation of technical and financialproposals from bidders. After evaluation ofbidders’ technical proposals, parties meetingthe minimum technical requirements are “pre-qualified”. The project is eventually awardedpurely on the basis of the most attractivefinancial proposal, as is the norm forcontracting by the Central and StateGovernments in India. Depending on the PSPmodel being adopted by the StateGovernment, the most attractive financialproposal may imply the lowest grantrequirement (such as in Madhya Pradesh), orthe shortest concession period (such as forsmall projects in Andhra Pradesh). As the tollsfor each State are pre-determined, there isno case for financial proposals beingexpressed in terms of the required tollstructure.

26. Implementing Vehicles at the StateLevel. At the state level, Road DevelopmentCorporations have been created in AP,Gujarat, Karnataka, Kerala, Maharashtra, andTamil Nadu. Part of the objective of these newagencies is to bring private sector funds tothe sector either through corporate borrowing(AP, Karnataka MP, Maharashtra,) or throughleveraging public subsidies against privateproject finance (Tamil Nadu, Gujarat). RoadDevelopment Corporations generally enjoymore flexible human resource managementarrangements, clearer objectives as well asmore streamlined decision making structuresthan typical Indian road agencies. Thesecharacteristics are potentially important interms of improving the quality and timelinessof the PSP project planning and procurement,thereby reducing some risks to potentialsponsors.

27. Aggregate Status of States’ EnablingEnvironment for PSP. Many states haveyet to establ ish the basic enabl ingenvironment (see Table 8.5 above).Considering the current status of the

l 49


institutional and regulatory framework forPSP, States may need to be realistic in theirexpectations for PSP and probably focus on“rehabil itate, operate and maintain”concessions, bypasses and/or bridges as themost viable PSP projects for the moment.

Financial Engineering on NHAI and StateProjects

28. Grant Supported BOTs. To date,typically grants have been provided duringconstruction only. There is now some debateas to whether moving towards partiallydisbursing grants during operations may notmake sense. However, ignoring any taxeffects, there is less difference between thetwo approaches than may first appear. Thesize of the payment made by thegovernment will merely adjust for the timevalue of money. The main difference wouldbe on who carries the risk of default on theconcession contract terms. If the grant is alldisbursed up front, the government riskshaving no remaining financial mechanismto enforce proper O&M – i.e. default onbehalf of concessionaire in relation to hisservice obligations. If the grant is disbursedover time, then the concessionaire risks the

government not being willing or able to paythe promised amount in the future – i.e.default on behalf of the government.

29. Least Present Value of Revenue(LPVR) based Bidding. The three roadconcessions being implemented byInfrastructure Leasing & Financial ServicesLtd. (IL&FS) in India are examples of “FixedIRR” or “Assured Return” models. Thesemodels provide a guaranteed level of netreturn on equity, taking the time value ofmoney into account. The concession periodlasts until the concessionaire recovers itsinvestment, thus a concession is terminatedonly upon the Concessionaire achieving apredetermined level of IRR. However, sincethe Fixed IRR model takes into account thereturns over and above the costs of the tollroad operator, there is no incentive for costefficiencies. All cost overruns, constructionor operational, are compensated for by theregulatory authority.

30. In the ‘Least Present Value of Revenue’auction (LPVR), the bidding variable is thepresent value of revenue throughout the lifeof the concession that firms are willing toaccept to undertake the project. The private

64 World Bank estimates

Table 8.4: Overview of Achievements in PSP at State Level

State Extent of Private Participation Agency and PSP Approx Value ofMechanism Lenghth SH Private

Constructed FundingKm64 Rs Cr

Andhra Some, but in very small size Andhra Pradesh Road Bridges and 188

Pradesh projects Development Corporation ROBs plus(APRDCL)Debt to APRDC and BOTs 36 km

Gujarat Limited although many are under Gujarat Road Development 110preparation. Corporation SPVs and BOT

Karnataka Limited in new construction but Karnataka Road Nil Nilmaintenance of most of core Development Corporation Ltd.network outsourced on term basis (KRDCL) Debt to KRDCL

Madhya Significant, but only a few of the MP Bridge Corporation Debt 2000Pradesh projects envisaged have to MPBC and BOTs (Re-habilitation

progressed to financial only)closure & construction

Maharashtra Many toll based projects being Maharashtra State Roadimplemented but limited private Development Corporationinvestment (MSRDC) Debt to MSRDC

and BOT

Tamil Nadu Significant but few projects have Tamil Nadu Road Development 180 300reached financial closure Corporation TNRDCL

West Bengal None Under discussion Nil Nil

50 l


operator that bids the lowest present valueof revenue wins. The duration of theconcession is then flexible and depends onthe effective traffic levels encountered. Bothfixed IRR or LPVR models offer someadvantages over alternative BOT schemesincluding the fact that: (i) tolls can beadjusted without negotiation with theconcessionaire, as long as the expectedinvestor return is maintained - this flexibilityis important in urban road concessionswhere it is difficult to determine the optimaltariff ex-ante, especially during congestionperiods; (ii) these methods reduce theuncertainty of demand, as they transferpolitical and demand-related risks to the userin the form of an endogenous concessionperiod; and (iii) calculation of compensationpayments on concession termination isstraightforward at any point in time during

65 World Bank assessment66 Debt for which the borrower is not personally liable.

the concession period. In addition to theseadvantages, the LPVR method of BOTproject implementation manages certainrisks better than the IRR fixed return model.These are summarized in Table 8.6.

31. National highway PSP projects in India aretypically funded through a mix of conventionalnon-recourse debt66 and equity. Bankersnormally accept a maximum debt-equity ratioof 70:30 for toll-based projects. The annuityBOT projects awarded by NHAI have beenfunded with debt-equity ratios of 75:25 owingto the much higher perceived certainty ofproject cashflows. Typical loan duration variesfrom 7 to 13 years, although only a fewprojects are able to secure debt of more than10 years. Normally, there is no obligation forrepayments during the construction period andthese only commence immediately after

Table 8.5: Comparison of State Fundamentals/Enabling Environment for PSP in Highways65

Fundamental Factor Andhra Kerala Karna- Gujarat Tamil Madhya Mahara- NHAIPradesh taka Nadu Pradesh shtra

Legal Framework ✓ ✓ X ✓ X ✓ ✓ ✓

Regulatory Framework ✓ X X ✓ X X X ✓

PPP Policy ✓ ✓ X ✓ ✓ ✓ ✓ ✓

Contracting Transparency ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓

Implementation Capacity X X X ✓ ✓ ✓ ✓ ✓

Semi autonomous road agency ✓ ✓ ✓ ✓ ✓ ✓ ✓ ✓

History of tolling ✓ X X ✓ ✓ ✓ ✓ ✓

Assured revenue stream for X X X X X X X ✓public contribution

Relative level of private finance Low Nil Nil High Medium Medium Low High

Table 8.6: Advantages of LPVR over Fixed Return Model

Item Conventional “Fixed IRR” Model LPVR Model

Return Assurance Investment on private equity pre- Investment on private equity pre-determined and assured determined and assured

Capital Cost Efficiencies Does not encourage concessionaire to Encourages optimization of capitaloptimize project capital cost, as benefits costs as concessionaire retains thedo not accrue to him, even though the benefits of capital cost efficienciesfinancial risk is still borne by the sponsors

Operating Cost Does not encourage operating cost Encourages operating costEfficiencies efficiencies, as costs are essentially efficiencies as higher costs

“passed on” to government agency/ reduce concessionaire’s returns andusers lower costs increase its returns

Monitoring Requirements More difficult, as it requires the employer More straightforward as thefor assessment of return to know both actual revenues and costs employer needs to know only actualat any point in time of the concessionaire concessionaire’s revenues

l 51


project commissioning. Loan repayment istypically on a quarterly basis with equalamortized repayments of principal andinterest.

32. The client/demand side is driving PSP,and so is the supply side, with market forcesencouraging the development of a numberof new financial instruments to help bringin private financing for roads. For example,the Infrastructure Leasing and FinancingServices Ltd (IL&FS) created a deep discountbond instrument (redeemable at the end of17 years) for use in its two BOT projects inGujarat, combined with a takeoutarrangement. IL&FS also pioneered the useof subordinated debt in financing two BOTprojects in Gujarat State (Vadodara-Haloland Ahmedabad- Mehasana highways).Bankers in India accept a partial investmentin subordinated debt by the projectpromoters as “equity” for the purposes ofcomputing debt equity ratios. The use ofsubordinated debt improves the effectivedebt coverage available to senior lenders,and helps in overcoming statutory limitationson dividend payment, improving sponsorreturns. Similarly, the GMR Group isimplementing two NHAI annuity BOTprojects in Southern India through pinpointequity. The group has made its equityinvestments in the project companiesthrough a mix of over 90% preferenceequity, with the balance investment asordinary equity. Preference equity is easilyredeemable under existing company law inIndia, while extinguishing of ordinary equityinvestment through a capital write downrequires approval from the High Court. Thepromoters of the Second VivekanandaBridge Project on an NH in West Bengal areproposing to place a convertible instrumentwith financial investors which is originallyraised as subordinated debt and iseventually convertible into equity. Theinstrument creates a convenient mechanismto offer lower risk with upside potential forfinancial investors.

33. Overal l , there appears to be areasonable amount of interest amongpotential lenders in taking debt exposure inBOT road projects being undertaken by acreditworthy sponsor. This may partly be dueto the general high availability of funds inthe banking system and a relative shortageof attractive lending opportunities. Inparticular, lenders are: (i) willing to considerattractive repayment terms, includingstepped up repayments and post-construction moratoria for projects perceivedas generally bankable; and (ii) offeringattractive interest rates for such projects, ona non-recourse basis. This opens up furtheropportunities for the government to lowerrisks and overall project costs and it is inline with the current trend to adopt moreinnovative financing instruments in India.

Indicators of Remaining Constraints toPSP

34. Limited Private Funds Raised. Oneindicator of weakness in the system is that thelevel of private investment is still relatively lowin comparison with other countries that areactively pursuing private funding of roads.While a number of states are using RoadDevelopment Corporations to access privatefunding, given the implicit Governmentguarantees provided, this amounts to little morethan off budget borrowing. The number ofStates that have actually seen BOT or othermodels where more risks are transferred tothe private sector are very few – in fact justseven states67. More substantive programshave not materialized, for example, inRajasthan68. Moreover, the majority of theseare short sections of bypasses or bridgesamounting to just a few 100 km. NHAI hasbeen rather more successful – about 450 kmworth Rs.3,400 Crore under BOT and a further470 km worth Rs.3,000 Crore through Annuityschemes. However, in comparison with someother countries this is low – in Chile, $3.3 billionwas raised from the private market over the1990’s to upgrade the 2,000 km highwaynetwork69.

67 AP, Gujarat, MP, Maharashtra, Rajasthan, UP, Tamil Nadu68

“Rajasthan government’s ambitious Rs 747-crore project for development of road network in the state under build operate and transfer basis has runaground with investors having slapped a “not feasible, no interest” notice in the first tranche of bids for such projects” Reuters, 6 June2002.

69 Toll Road Concessions – the Chilean Experience, PFG Discussion Paper No 124, World Bank, 2003. See also the PSP database for global information at http://rru.worldbank.org/ppi/

52 l


70 Presentation by CEO of Noida Toll Bridge Co, November 200271 http://www.thehindubusinessline.com/bline/2003/01/28/stories/2003012800960600.htm

35. Long gestation of PSP Projects. Asecond indication that there are constraints inthe system is that PSP projects are generallytake a long time to prepare and procure –typically at least 24 months. A good exampleof this is the Bangalore Mysore InfrastructureCorridor Project. While the Government ofKarnataka and the proposed concessionairesigned a Memorandum of Understanding in1995, the project has still to come to financialclosure. The Delhi Noida took nearly nine yearsbetween the signing of the MOU andcommencement of operations70. However,given more recent events and additionalreforms, it is not expected that these lengthsof delays will be repeated in the future. Theconcessioning of a further 10,000 km of NHis ongoing and the actual timeframe of thisprocess will provide an indicator of whetherthings are improving.

36. High Rate of Restructuring. Anotherindicator is that quite a few PSP projects havehad to be restructured or renegotiated. Forexample, the Durg bypass on the NH6 inMadhya Pradesh now requires renegotiation.This 18km long bypass has suffered from adouble blow of escalated costs due to achange in the scope of work imposed aftercontract award by NHAI, for whichcompensation has not been paid yet, togetherwith traffic volume being about 40% belowprojections. Competing routes that had notbeen adequately factored by either party hada role to play in reducing the traffic. The DelhiNoida BOT has also had to be restructureddue to lower than expected commercialtraffic, itself partly the result of the failure ofthe government to provide adequateconnecting routes71. The restructuring of theHubli Dharwad, Vadodara Halol BOT projectsare also understood to be currently undernegotiation.

37. Questionable Value for Money. A finalindicator that there are improvements to beachieved is that there is skepticism over thevalue for money that has been obtained insome PSP models in India. In particular, theAnnuity model has come under scrutiny as towhether the eight agreements (474km)

reached was a good use of taxpayers’ money.This model provides for a concessionaire todesign, build, finance and operate a projectroad for a predetermined time. Aftercommissioning NHAI makes semi annualpayments until the end of the concessionperiod as the only revenue for theconcessionaire. NHAI retains the right tocharge tolls on the road. In an analysis donefor this study, the cashflow profile of the eightconcluded annuity projects have been shownto be equivalent to NHAI borrowing atbetween 17-18% (derived as the IRR of thecashflow stream from NHAI’s perspective).This compares with a weighted average costof capital for NHAI of about 11.3%. Althoughsome design/survey change, operation andmaintenance risks have indeed beentransferred to the private sector, and thereare some regulatory inefficiencies (statutoryreserve requirements and taxes) this may notexplain the spread between the two rates orwhether the premium being paid isappropriate for the risks transferred. It is notclear what level of due diligence wasundertaken prior to entering these contracts– however, with hindsight the annuity modelas implemented appear to providequestionable value for money.

Systemic Constraints To PSP In India

38. Absence of Reliable Information. Inmaking a decision to buy or sell, lend of borrowany asset, people use information to determinethe relative price of the asset and how thiscompares with the expected future cashflowsfrom the transaction. The cashflows can eitherbe inferred by comparison with current marketprices if there is a market for that asset orthey can be estimated through projectingforward historical data of relevant factors. Inthe case of highway projects that have theirown individual characteristics, potentialinvestors must use a forecasting method todetermine the cashflows – data on constructionunit costs, traffic, and the history ofconstruction of any existing road or structurethat is to be incorporated in theconcessionaire’s highway. The problemhighway investors and creditors face in India

l 53


is that such information generally is notavailable even within the road agencies, letalone in a usable format or on a public basis.Where there is data available, there areconcerns over the accuracy of the data,especially in relation to traffic counts (seeAnnex 12 on private sector concerns).Hence, investors rely on their own limitedinvestigations which are unlikely to capturea long history or wide breadth of experience.Without this, no probability distribution ofcosts and benefits can be constructed onwhich sensitivity or scenario analysis rely.The resulting analysis is more likely to beinadequate, subject to bias and ultimatelythe conclusions may not hold. The perceptionwithin the private sector is that the estimatesof construction costs indicated to bidders atthe time of bidding may not entirely reflectcosts of construction in practice, as certainallowances for price escalations may needto be made (the t ime gap betweenconstruction cost estimates and the time ofactual bidding has been 1-2 years).

39. High Price Elasticity of Demand forTolled Services. In many countries, privatefunding for roads depends heavily on tollcollection to provide for revenue. In India,both Central and State Governments areseeking to increase toll collection both forpublicly and privately funded roads. Yet, the

systemic problem faced across the countryis that, largely due to high levels of existingtaxation on road users, willingness to payfor tolls is rather low, despite the fact thatIndian toll levels are on average some ofthe cheapest in the World – roughly 15-25%of the savings in vehicle operating costs forcars and 50-70% for commercial vehicles.Where tolls are being imposed, expectedtraffic is not materializing due to diversionundermining the viability of existing projectsand deterring private investment in futureprojects. This risk is genuine and canthreaten the financial sustainability of aproject. The Mumbai-Pune expressway, forexample, is not able to attract the estimatedtraffic levels due the presence of alternateroutes and the user’s reluctance to pay theexisting toll charges (see Box 8.4). Thetrucking industry has already raised itsobjection to the so called “double dipping”of increased cess and toll72.

40. A revealed preference surveyundertaken for this study looked at userattitudes and preferences on the tolledMumbai Pune Expressway and the parallelexisting un-tolled route along the NH4. Thesurvey was taken from 600 private andcommercial users each on both routes. Thesurvey also collected traffic data from thetwo routes - the free route has three to four

Box 8.4 The Experience with the Mumbai-Pune Expressway in MaharashtraThe first phase of the project between Khandala and Lonavala was opened to traffic in May 2001 and theentire expressway was ready by mid-February 2002. The expressway reduces travel time from Kalamboli toDelhi Road to about one and a quarter hours as against 5 hours on existing NH-4. Despite its inherent appeal,the road has failed to attract the estimated traffic levels, and hence raise the expected revenues, due to thepresence of alternate routes and the users’ reluctance to pay the existing toll charges. After a year ofoperation, revenues fell short of its first year toll collection target of Rs. 1.2 billion. Currently MaharashtraState Road Development Corporation (MSRDC) is collecting an average of about Rs. 37 million a month as tollcharges from an average of 12,000 vehicles per day plying on the expressway. This, coupled with higher thanexpected construction costs due to delays in execution (over Rs. 16 billion versus Rs. 14 billion budgeted), isthreatening the sustainability of the agreement. MSRDC is optimistic that there will be an upswing in revenuesonce the entire stretch is complete and properly operational as two underpasses and a 10 km long PanvelBypass were completed recently.

To get out of this debt trap, MSRDC has recently handed over the expressway along with the parallel Mumbai– Pune section of the NH-4, on a 15 year lease. The concessionaire has been granted the sole rights to collecttolls on both these roads in lieu of an upfront payment. Under the deal, the concessionaire is required to (i)upgrade the NH4 to 4 lanes by March 2006, after which it can start collecting tolls on the NH4 and (ii) maintainthe expressway, (however structural repairs on the expressway will continue to be the responsibility ofMSRDC). Opposition to this move has already begun with local residents demanding that at least one roadshould be toll free.

72 “There should not be any imposition of tolls on existing roads after their conversion from single lane to two-lanes and from two-lanes to four-lanes. If theGovernment wants tolls, then it should stop collecting cess on diesel “, the All India Motor Transport General Secretary, Mr. J.M. Saxena, as reported inThe Hindu, September 23, 2002

54 l


73 Take the example of the East-Coast Highway (100km), the Government of Tamil Nadu’s contribution of Rs.10 crore leveraged an investment of Rs.51 Crfrom the private sector. It is expected to save the Government in excess of Rs.100 Cr by way of savings on future maintenance expenses and ensuresustained O&M for the entire concession period of 30 years.

times as much traffic as the tolled route.The survey sought user’s views on five mainfactors – travel time, riding comfort, cost,reliability and traffic interference. While thetolled route is preferred by cars, taxis, busesand vans due to time savings, heavy freightusers prefer the NH4, overwhelmingly dueto cost savings. 50% of respondents usingthe NH4 stated that they were not usingthe expressway due to the toll, despite therate being set at about 25% of VOCs savingsfor cars and 60% for trucks – i.e. every useris better off using the expressway. Overall,the survey suggests that the perception ofthe value of benefits to higher class butdirectly tolled roads is relatively low atpresent.

41. Furthermore, private sector perceptionsare that total traffic estimates in the biddingdocuments do not necessarily the reflect thetotal “tollable” traffic on these roads, whichis somewhat lower than total traffic. Ingeneral, road projects have often shown lowtollable traffic, warranting, in some cases,grants as high as 60-70% of constructioncost in order to enthuse private investors.The case for adopting a BOT at such highlevels of subsidy becomes very hard tomake.

42. Dispute Resolution. At present, thestandard concession agreement in use byNHAI, provides for the following levels ofdispute resolut ion ( i ) mediat ion byindependent consultant appointed by NHAIfrom panel of three firms identified byconcessionaire which are to come frompanel of five firms chosen by NHAI (ii)Chairman of NHAI and Chairman ofconcessionaire and (iii) arbitration underRules of the Arbitration Act. Those BOTs thathave been let at state level have generallyfollowed this same mechanism as they havebased the concession document on theNHAI’s. While this mechanism is a step inthe right direction, for more serious disputesthe probability of ending up in arbitration isstill high. The time and costs of arbitration

are, however, often high and a more certainand swift dispute resolution mechanism mayreduce perceived risks for private investors.

43. Indebtedness Levels. It is an acceptedpractice in India to leverage infrastructureprojects with a large debt component.Private sector participation and innovativefinancial structures have sometimes beenimplemented by increasing the indebtednessof the public sector to levels beyond thoserecommended for India. These high gearinglevels, when compounded with the need toaccess long term debt of a maturity andrepayment profile suitable to the project’scashflows, subject the project to significantfinancial risk. For example, the Mumbai-Pune Expressway project costing Rs. 12,55billion is highly leveraged with a debt equityratio of 4.5:1. The entire debt amount ofRs. 10,30 billion was raised in the domesticcapital market through bonds issued by theMaharashtra State Road DevelopmentCorporation (MSRDC). Even the equityinvestment of Rs. 2,25 billion was receivedfrom the Mumbai Metropolitan RegionalDevelopment Authority (MMRDA) in theform of subordinated debt. As the Indiancapital markets are still underdeveloped, adebt to equity ratio in the range of 2.2:1and 1.5:1 are usually accepted by financialinstitutions and banks for road infrastructureprojects, significantly below that for theMumbai-Pune Expressway. Guidelines toassure gearing levels are kept at reasonablelevels are required.

44. Additional Weaknesses at StateLevel. State governments face a moredifficult task than the central governmentin securing PSP, although some States haveproved it can be done on certain projects73.The former not only have lowercreditworthiness when concessionaires areconsidering the timeliness and completenessof any public subsidies, but also networksgenerally with lower traffic. This puts apremium on state governments improvingthe enabling environment and their own

l55


capacity to implement PSP projects. Manystates have yet to put the basic buildingblocks in place, including not only theenabling framework but also procurementand contracting capacity and certainty ofrevenue stream from public sector.

45. Summary. While it has been arguedabove that international experience hasdemonstrated that private funding is notlikely to prove a panacea for the sector, andthat there are still many constraints in Indiato facilitating more private investment,actions should still be taken. The privatesector can bring funds and new techniquesto improve the network, especially its busiestl inks. The GOI is c learly banking onsubstantial PSP in the next phase of NationalHighway development with 10,000 km ofroads being expected to be fundedsubstantially through BOT. Moreover, there

is also opportunity at the state level. Ananalysis that was carried out for this studysought to determine the likely scope of PSPby way of toll based BOT projects in six states(see Annex 13)74. This concluded that of the37,000 km state highway network in thesestates, roughly 13,000 km or 36% waspotentially viable as a BOT project, eitheras a rehabilitation or upgrading projectdepending on traff ic levels. The keythreshold for viability was around 5,000 PCU,below which more than 70% public supportwould be required and a BOT was thereforedeemed unsuitable. The range of publicsupport required for these 13,000 km of BOT,assumed at 40% disbursed over theconstruction period, was estimated as beingfrom Rs.6 to Rs.72 Crore per annum, sumsthat are quite feasible even for fiscallyconstrained states.

74 AP, Karnataka, MP, Maharashtra, UP and West Bengal

56 l

T h e F o u n d a t i o n s f o r F i n a n c i a lSustainability in the Sector

1. Previous chapters of this report haveoutlined the large financing gap in the highwaysector, and the likelihood that increasedfunding will have to be financed from thehighway users themselves, irrespective ofwhether investment is financed from publicor private sources. It is clear that theinstitutional arrangements of the sector needto change in order to improve the quality ofdecision making, enhance the transparencyof financing arrangements and obtain theacceptance and support of the road usercommunity for the proposed changes in thelevel and structure of road charging. Thereforms suggested are designed to help tacklethe underlying problems of inadequateinformation, fragmented decision making, andlack of transparency.

2. The reliance on the road sector for generaltax revenue is a systemic problem whichaffects many developing countries. Abroadening of India’s tax base may be verydesirable but may not be achievable in theshort-term. Consequently, road users are likelyto have to pay more for better roads75. Severalcountries have found that, in order to makeadditional user charges acceptable to roadusers, rather different institutional governancestructures are desirable; structures in whichrepresentatives of road users and otherstakeholders are brought into the decision-making process and given some influence onhow their user charges are spent.

3. Create and Disseminate BetterInformation. The first step in strengtheningthe decision-making process is improvedinformation and information dissemination.Many governments regularly re-assess thestructure/level of road user charges throughRoad Cost Allocation and Road Cost RecoveryStudies76. Such studies may cover:

9. RECOMMENDATIONS

• the road costs imposed by differentvehicles on each road category in thenetwork;

• the total of taxes and fees paid byeach vehicle type and for differentcategories of roads; and

• the level of road user charges paidby each vehicle type on each roadcategory.

4. In India, a road user charge studywas undertaken in 1988-89, but itsrecommendations were given l i t t leconsideration and the study has not beenrepeated. When road users are expected tofund the road network, such studies areessential. The allocation of costs betweenvehicles needs to be kept under continuousreview to ensure that no vehicle group ispaying less than its road damage costs. Roadcost allocation reviews should be undertakenbiannually and, as taxes and charges are raisedby both GOI and the States, they should becommissioned by a central agency with theStates providing whatever data on costs andtaxes are necessary. These cost allocationstudies should be made public and form thebasis for determining user charges.

5. Furthermore, the nature and quality of thefinancial information collected and reportedby road agencies should be greatly improved,especially if the road agencies are hoping toraise funding from the private sector. Globally,road agencies are adopting a business orientedapproach to managing the road network andshifting from traditional governmentaccounting to quasi-commercial accounts inwhich:

• accrual rather than cash accountingis employed77;

• road/bridge assets are capitalized anddepreciated; and

• the cost of maintenance is reportedas an expense78.

75 Of course, the benefits of the better roads to road users should more than offset the additional charges, but may not be immediately perceived.76 See for example the results of such a studies in the USA at http://www.fhwa.dot.gov/ohim/hwytaxes/2001/77 See “Transition to the Accrual Basis of Accounting: Guidance for Government and Government Entities (2nd Edn.)” IFAC, December 2003, available from

http://ifac.org78 See Asset Management for the Road Sector, OECD, 2001

Recommendations

l 57


6. This follows the accounting conventionsused by the private sector firm, showing fixedassets on its balance sheet and accountingfor depreciation through the incomestatement. Road agencies are thusdemonstrating more coherent and transparentaccounting for the very considerable assetsthey manage: NHAI, itself, would be high inany ranking of Indian companies by balancesheet size if road assets were properly valued79.In the USA for example, the GovernmentAccounting Standards Board promulgatedStatement 34 in 1999 requiring all governmentbodies to calculate, audit and disclose the costof infrastructure and the associated annualdepreciation/maintenance expense. Theintroduction of the requirement being phasedin by 2006 is likely to have profound effectson the bond market which road authoritiesrely on for a considerable portion of theirfinancing80.

7. Audited financial statements (balancesheet and income and cash-flow statements),that reflect the reality of the sector, wouldsend a clear message to taxpayers, bondholders and underwriters of the road agencies’stewardship of the road network and theirfinancial capacities to repay future liabilities.If NHAI and/or the state Road DevelopmentCorporations want to reduce the cost of theirborrowing; facilitate private funding of thesector; and meet the growing public demandfor clearer accountability, adopting a moderncommercial accounting system will be a criticalbuilding block.

8. If public and private finance is to be usedcost –effectively in the improvement of India’shighway network, there is also a pressingneed for improved road and traffic data.Potential investors in the network mustforecast their cash flows and need data onconstruction unit costs, traffic, and theengineering history of those roads/structures.Unfortunately, such information is generallynot available even within the road agencies,let alone in a usable format or on a publicbasis. The likely consequence is that investorsincorporate much higher risk premiums intotheir bids than might be justified. Contractingout, using more automated methods andindependently auditing data collection may

help inspire greater confidence in the qualityof the data. By having a longer and widerhistory of traffic, tendered rates for BOQ items,asset condition and traffic, clients can expectprivate parties to be more accurate in theirtendering with benefits accruing to all sides.

9. Create a Coordinating Mechanism toHarmonize User Charges. While theconstitutional responsibilities for setting andcollecting taxes and charges need not bemodified, there would be major advantagesin a more coordinated approach to setting roaduser charges. At present, there is littlecoherence in either the methods or levels ofcharges and, in some instances, conflictingincentives are being given. There are twobroad options for establishing a harmonizedroad user charge regime: a strategic roadauthority or a looser affiliation of the centerand state governments.

10. A Strategic Roads Authority (SRA) wouldbe responsible for the planning, coordinatingand financing of the overall highwaysprogram. The Central Road Fund alreadyoffers a source of finance for national andstate highways, as well as local roads; butit cannot influence the allocation of fundsbetween activities (set by statute), and hasno powers over the road-user taxationpolicies adopted by the individual states. TheSRA could receive the cess, and all otherdesignated national road user charges andthen channel this funding to develop acoordinated approach to highway fundingacross all the states. It would be chargedwith undertaking studies of road costs andrevenues and developing, with the relevantstate agencies, an integrated approach toeffective road user charging.

11. The SRA should be given theresponsibility for designing an approach tothe f inancing of capital investment,consistent with budget limitations and thecapacity to raise new sources of revenues,over the next two Five Year Plan periods. Itis perhaps inevitable that the present heavyfinancial burden of developing a modernhighway network will be spread more evenlyover time. The SRA would, therefore, needpowers to raise medium term (5-10 years)

79 Australia, UK, USA and Canada are all moving in this direction80 Primer GASB 34, US Federal Highway Administration, 2000

58 l

debt, to be held on its balance sheet, to berepaid by a secure revenue flow from the fuelcess and highway tolls.

12. If the SRA is to facilitate co-ordinationbetween central and state functions, therewould need to be a governing councilcomprising representatives from the central andstate governments. In addition, to helpgenerate support for increased charges, thegoverning council should also includerepresentatives from road users and other roadsector stakeholders. The SRA would need tobe staffed by engineers and, more particularly,by experts in economics and finance.

13. A powerful SRA may be too radical. Inthe short term, it may be more feasible todevelop a council of transport ministers or somesimilar body whose mandate would be toencourage dialogue between the states to helpharmonize policies toward the road sector,including road user charging and expenditureallocations. Such a body could perhaps becreated as a sub-committee of the Council ofChief Ministers. It might well be theappropriate body to commission the proposedstudies on Road Costs and Cost Allocation;there would then be common ownership ofthe resulting product by all the relevantgovernments.

14. Internationally, two similar institutionsalready exist: the European Conference ofMinisters of Transport (ECMT), and theInternational Fuel Tax Association in the USAand Canada. (see Annex 14). The ECMT hasa much broader mandate than theInternational Fuel Tax Association which isdesigned purely to reduce and improveadministration of fuel taxation. Both examplesdemonstrate that there can be considerablebenefit in joint action on common issues,without compromising sovereignty. To establisha similar organization for India, with substantialpotential benefit, only requires political will andsome limited secretariat support.

15. Create and Maintain National andState Road Funds. Differentiating implicitroad user charges from general taxes alwaysrequires assumptions and consequently theconclusions are speculative; there are nounambiguous criteria for what constitutes a

“normal” tax. However, if responsibility forsetting road user charges were separated fromgeneral revenue taxes, and the user chargerevenue accrued to a commercially managedRoad Fund, then there would be clearer linesof responsibility and accountability. Severalsuch road funds have now been establishedand are working satisfactorily. The maincharacteristics of such a road fund include thefollowing81.

• A separate, quasi-autonomous RoadFund, should managed by a RoadsBoard.

• The Roads Board inc ludesrepresentatives from private road-users and other stakeholders as wellas the Government. Preferably, thereshould be a majority of private sectormembers with a private sectorchairperson.

• Road user charges are specific andseparated from general taxation.

• Revenues are paid automatically anddirectly to the Road Fund, rather thanthrough the Consolidated Fund.

• Strict procedures are established forthe allocation of funds betweenactivities and the approval of workprograms submitted by the variousroad agencies.

• Independent financial and technicalaudits are undertaken and all reportsare published.

16. India already has a national Road Fund,financed by the road cess. The allocations ofthe receipts are statutorily predetermined(national, state, rural, and urban), with 50%allocated to rural road development. A fewstates have also established their owndedicated road funds to increase road sectorfunding. The Indian Central Road Fund differsfrom the recommended “second-generation”road funds82 in two principal aspects:

• the emphasis on road developmentrather than road maintenance withthe revenue from the fuel cess islargely used for new roadinvestment; and

81 See Commercial Management and Financing of Roads, World Bank, 1998 http://www.worldbank.org/afr/ssatp/techpaper/tp409.pdf82 “First generation” road funds were simply revenue raising from dedicated taxes. In many countries, such funds developed a very poor reputation with

high levels of corruption and the diversion of the funds to other uses.

Recommendations

l 59


• the complete dominance of thepublic sector in the management/governance of the Fund83.

17. Rational management of the road sectorrequires a coordinated and integratedapproach to investment and maintenance.Unfortunately, politicians often take a veryshortsighted view of road maintenance,when faced by resource constraints, and givepriority to investment. In India, this is a majorproblem and there is a huge shortfall onroad maintenance. This could be addressedby legislating that the Central Road Fund,or a successor Strategic Roads Authority, andState Road Funds must adequately maintainthe core national and state highwaynetworks, and that maintenance should bethe first charge on resources.

18. The continued involvement by theMinistry of Finance and Roads has notundermined the effectiveness of road fundselsewhere. In India, it is perhaps inevitablethat road funds will have mixed public/private boards, and considerable influencewil l be retained by the Ministr ies/departments of Finance. Such involvementwould probably include MoF agreement tothe level of any cess or other specific roaduser charge. This, however, should not bean impediment to the development of abusiness oriented and user responsiveprocess for the raising of revenues, andallocation and implementation of roadexpenditure.

Options in Financing the HighwayProgram

19. Assuming that the highway network isfully maintained and improved to meet theneeds of a rapidly growing economy, there islikely to be a cumulative funding gap, duringthe next ten years, in excess of Rs 1,000 billion,or more than US$2 billion per year. Indeed,the funding gap would be much greater (Rs1,760 billion) if the present proportion ofhighway user charges continues to be used inother sectors. The Government of India andthe State Governments have a number ofpossible options to fill, bridge, or reduce thefinancing gap:

i) Increase the proportion of roadrelated revenues returned to thesector, i.e. reduce theimportance of road users as asource of general tax revenue;

ii) Raise the levels of existing taxesand user charges;

iii) Introduce more direct roadpricing, such as road tolls orweight-distance charges;

iv) R e d u c e p u b l i c s e c t o rcommitments by concessioninghighways to the private sector forconstruction and/or maintenanceand then require concessionairesto recover some/all of their coststhrough tolls;

v) Bridge the immediate financinggap by increased borrowing, thusspreading the financing burden;and

vi) Reduce investment expenditureby greater prioritization of worksand lowering unit costs bysystematic value analysis at everystage of the project cycle.

20. Reducing the role of the road sector as asources of general revenue (option i), maysimply not be feasible where there are highdemands for public finance and limitedopportunities for efficiently increasing taxrevenues. Concessioning highways, with theprivate sector assuming the traffic/revenuerisks, (option iv), may have some role but anetwork wide approach, with a public roadagency assuming the traffic risks and settinguniform tolls, may provide greater overallbenefits.

21. Resolving the financing issue thus appearsto revolve around either borrowing (option v)or increasing road user charges, raising existingcharges and/or introducing new charges(options ii and iii). Prioritization of works andobtaining better value from expenditure(option vi) should apply to all publicinvestment, as it does in the private sector. Acombination of borrowing and increasedcharges seems almost inevitable, the balance

83 The public sector dominance of the governance of road funds is also found at the state level, except in UP.

60 l

between them is essentially a policy decisionfor the central and State Governments.

22. The very high financing needs in the roadsector reflects not only growing traffic levelsbut also the relative neglect of the sector overan extended period. It can argued that thepresent road investment needs are muchhigher than the longer term equilibrium level,once the back-log of rehabilitation andcapacity expansion has been rectified. To fundthe proposed expenditure from present usersalone, will either require a very substantialincrease in user charges or the allocation ofmore general tax revenue to the road sectorwhich may be difficult to achieve.

23. The benefits of the proposed investmentwill extend to future road users and thus itwould be appropriate to recognize that this isessentially a long term investment issue andspread the capital financing burden of thehighway network expansion/improvement tofuture road users, through borrowing. Suchborrowing could be through the public sectoror through various forms of private sectorfinancing. The extent of generational costsharing is essentially a political decision, andone which needs to be made in order to definethe funding required from existing road users.

Recommended Changes to ChargingInstruments

24. The levels and structure of road taxes haveevolved without any underlying economicrationale and/or specific objectives other thanrevenue generation. The present tax structureprovides neither economic efficiency norequity. Any efficiency or equity achieved hasbeen incidental and not the result of deliberatepolicy decisions. The absence of a welldefined, rational and coherent road usercharge policy at the National and the Statelevels has contributed to the multiplicity oftaxes and the pricing distortions.

25. Therefore, it is strongly recommendedthat a national taxation policy be developed,for the entire land transport sector, that would:

• promote efficient competitionbetween alternative modes;

• encourage market oriented solutionsthat reduce subsidies;

• help develop optimal pricing policiesfor different land transport modes;

• acknowledge the desire to shifttowards more direct charging for roaduse; and

• differentiate road user charges fromgeneral taxes and define the sharesfor road construction and roadmaintenance.

26. Road transport and road infrastructure arevital for accelerating the rate of economicgrowth in India, and a well-defined andcoherent financing structure for the roadnetwork should have very high priority.

27. Increasing the general fuel cess. Theuse of highways differs from the purchase ofmost other services as there is rarely a directprice. There is, however, a strong case forusing a “proxy charge” for the use of the roadnetwork. Such a charge is often levied onautomotive fuels84. Many countries, facing thesame funding constraints as India, haveimposed a fuel levy/cess, with the revenuededicated to the road sector (often specificallyroad maintenance). While dedicated taxes aregenerally anathema to finance ministries andthe IMF, there is growing acceptance of thebenefits of a reliable source of roadmaintenance funding, within a well structuredand accountable road fund.

28. In 1998, the GOI introduced a new fuelcess, dedicated to expenditures in the roadsector, which was subsequently formalizedthrough the Central Road Fund Act, 2001. Thecess was initially set at Rs.1/liter, for bothgasoline and diesel, and then increased toRs.1.50/liter (»US¢3/liter) in 2003. This hassubstantially increased the resources for thehighway sector: Rs 20 billion/year for nationalhighways and about Rs 11 billion for statehighways. Some states have introducedadditional fuel cesses; Uttar Pradesh, forexample, imposed an additional sales tax of4% on diesel, and 6% on gasoline, raisingabout Rs. 2.5 billion/year for the road sector.

29. Financing the funding gap from the fuelcess would require a cess rate of about Rs.7.5/liter; such a massive increase is unlikelyto be politically acceptable. The fuel cess levyin India is relatively low in comparison with

84 Charges on vehicle tires might perhaps give a greater correlation to road costs but high charges on tires could have serious road safety issues.

Recommendations

l 61


similar levies in other countries (they rangefrom US¢4.5 – 10/liter but most are US¢ 6 –7/liter) and some increase would probably beacceptable, especially if accompanied byinstitutional reforms to the management ofthe Central Road Fund. It is recommendedthat the GOI considers doubling the fuel cessto Rs 3/liter, and dedicating the additionalrevenue entirely to road maintenance. It isalso recommended that the Stategovernments establish dedicated cesses tofund the maintenance of the state highways;this might require fuel cesses of Rs. 0.5-1.0/litre on both diesel and petrol. The currenthigh international oil prices complicate thiscourse of action somewhat but do not makeit impossible.

30. Higher diesel cess . There is a strongeconomic case for a higher cess on diesel asthe variable costs imposed by heavycommercial vehicles are about 7-10 timeshigher than those by passenger cars85. Heavytrucks are undercharged for their road damagecosts and raising the cess on diesel would beone way of recovering these costs. The dangerof a higher cess on diesel is that it wouldincrease the price differential with keroseneand encourage fuel adulteration.

31. Increasing the diesel cess could havenegative impacts on non-road users of diesel,who consume about 46% of diesel86. It isrelatively easy to exempt large individual users(railways and power), but much more difficultto give exemptions to small users in agricultureand industry. The measures that governmentshave used to address this issue are outlined inAnnex 15, but some governments have simplyaccepted the inequity and provided noexceptions. It is recommended that GOIcommission a detailed study to review thecase for higher diesel cess and the likelyimpacts on other sectors.

32. Purchase and ownership taxes.Purchase and ownership taxes on vehicles(fixed taxes and charges) can and often areused as part of a road user charge strategy.While it is generally preferable to relatecharges to the use of the vehicles, this is notalways practicable. There are also goodarguments for having a fixed charge to cover

the fixed costs of providing and maintainingthe network. The annual fixed charge shouldbe structured to reflect not only the non-attributable costs of the network but also thoseattributable costs which are not adequatelyrecovered by use related charges. For freightvehicles, this would suggest annual licensefees based on a combination of gross vehicleweight and the number of axles. The annuallicense fee can thus provide incentives to roadusers to employ less damaging vehicles.

33. Most states already charge annual feeson motor vehicles but it is clear that there isno standard approach to the level or structureof these fees. It is recommended that,following a study of road user costs, all stategovernments should be encouraged to adopta uniform approach to the establishment ofannual fees for freight vehicles in order toensure that the road transport industry isprovided with consistent incentives to adoptthe most economically efficient vehicle fleetconfiguration.

34. Tolling the National and StateHighways. Road tolls generate a smallfraction of the total revenue from road users,possibly around one percent. However, bothin India and internationally, there is growinginterest in directly pricing for road use both toraise revenue and to relate user charges moredirectly to user costs (including congestion).

35. In the light of the present low willingnessto pay tolls and the limited value of timesavings for freight vehicles, the appropriatepolicy is to keep tolls low but apply them morewidely. This is NHAI’s intention, once theGolden Quadrilateral and North/South East/West corridors are completed. The NHAI draftCorporate Plan indicates that tolls on thisnetwork will generate Rs. 20 billion annually,by 2007, sufficient for corridor managementand maintenance plus »50% of interest onoutstanding debt, excluding annuity schemesand SPVs/BOTs.

36. It seems probable that India will relyincreasingly on tolls for highway finance andthe real issue is the speed of their introduction.The NHDP is an ideal opportunity to implementa systematic toll policy for major interstate

85 Where differential fuel levies are used, it is often gasoline which is a higher rate86 Agriculture: 20%; power generation: 7%; industry: 8%; railways: 4%; and other: 7%. Source: Ministry of Petroleum and Natural Gas, 2000-01

62 l

highways. Except on high cost bridges, wherehigher tolls may be both necessary andfeasible, it would seem sensible to have anational toll schedule, designed to financeconstruction and maintenance over a 20 - 25year period. This could be introduced over 3- 7 years. It is recommended that the GOIdesignates the new routes on the GoldenQuadrilateral and on the North-South and East-West corridors as a national toll network. Thetoll network would be controlled by, and thetoll receipts would accrue to, NHAI as generalincome rather than specifically to individuallinks, except where existing toll basedconcessions have already been let. The actualoperation of the links could be outsourced toprivate operators. This initial toll network couldgrow as the remaining 44,000 km of nationalhighways are upgraded.

37. It is also recommended that Stategovernments implement a uniform tollingsystem on their core highway networks, onceindividual links are improved to a reasonabletwo lane standard with traffic levels sufficientto justify the cost of tolling. The tolls shouldbe set to recover, at least, the cost of operatingand maintaining the roads and should becommensurate with the national system. Inthe longer term, the levels of tolls may beincreased to cover the costs of strengtheningor rehabilitating further links on the corehighway network.

38. A study, undertaken in preparing thisreport, assessed the potential for toll basedBOTs on state highways in six states with atotal state highway network of 37,000km;such schemes may be feasible on about 35%of the network, assuming relatively low levelsof investment. Moreover, new technologiesare reducing tolling costs and the levels ofevasion and revenue leakage. Even adoptinga cautious approach, it may be feasible to tollabout one quarter of the state highwaynetwork.

39. Weight/distance charges. One of themost serious deficiencies of the road usercharging regime in India is the under-chargingof heavy commercial vehicles for their roaddamage costs. This undercharging can bepartly compensated by annual charges butthese are a rather blunt instrument and donot discriminate between vehicles with

different levels of utilization. Moreover, asimple charge, based solely on gross vehicleweight, can give distorted signals with respectto vehicle choice.

40. In India, a weight/distance charge couldbe introduced for national/state highwaysthrough the methods used elsewhere, or bystopping vehicles at toll plazas for manualcollection of tolls related to the vehicle weight/axle category. An Expert Group should beestablished to elicit the views of the truckingindustry and then consider the proposal indetail.

41. Urban congestion charges. It isrecommended that urban road pricing shouldbe considered to address the increasingcongestion problems in major urban centers.However, this should be viewed as a separateexercise to general road user charging andshould be initiated and managed by the citiesas part of integrated urban transportstrategies. This will require a definition ofurban boundaries for the purposes of inter-urban toll and charging purposes; it issuggested that these are established early inthe development of the national and statehighway programs.

42. Enhancing Equity. Moving towardsgreater reliance on differentiated fuel cessesand road tolls will, in the longer term, help toallocate more directly the costs of differentvehicle categories and make the chargingstructure more equitable between users. Inthe short term, it would be prudent also torationalize the current tax regime. In a countryof sub-continental size like India, efforts shouldbe made to harmonize the methods ofcalculating levels of user charges so that thecurrent wide variation between states andbetween different users can be reduced. Thistype of exercise is currently being undertakenby the 25 countries of the EU.

43. The existing tax structure should besimplified first, by minimizing the ambiguitiesin vehicle classification and the basis ofcharges. Certain taxes, which are difficult toadminister, like passenger tax and goods taxmight be merged with road tax (the principaltax levied by the State Governments) and asingle consolidated charge levied. Rajasthanand Andhra Pradesh have alreadyamalgamated various taxes without loss of

Recommendations

l63


revenue. Another form of simplification, nowin use, is an one-time tax on private personaltransport, i.e. private cars and 2-wheelers. Thiswould reduce administrative as well ascompliance costs. However, some form ofannual registration may still be desirable forvehicle control purposes, such as for vehicleroad-worthiness.

44. States also need to review whether thecurrent charging of buses and commercialtrucks is equitable or whether it should berevised. The overcharging of buses andundercharging of freight vehicles has negativeimplications for both transport efficiency(leading to inter-modal traffic distortions) andequity. It appears to be the result of the heavyfinancial dependence of states on bus andpassenger taxation. The solution to this mayneed to be partly institutional with somecompensation, possibly through a re-allocationof fuel tax and other central funding, if thestates forgo the excessive taxation on buses.

Additional Reforms to Help Spur PrivateFinance

45. Establishment of a HighwayRegulatory Authority. While this aspect isalready fairly well developed at the nationallevel, the GOI can consider the followingoptions (i) do nothing; (ii) incrementallystrengthen the standard concession documentas is already being done; or (iii) establish anindependent regulatory authority under law.The third option may be sensible in the longerrun to help standardize process therebyproviding greater predictability to investors andallow for shorter concession contracts thateasier to develop and conclude. Authoritieshave become functional in the Telecom,Insurance and Energy sectors in India and areplaying an important role in opening up thesesectors to competition. Establishment of aHighway Regulatory Authority on similar linesfor the road sector would certainly provide animpetus to private sector participation invarious modalities in road projects, thusquickening the pace of growth by removing/mitigating infrastructural hurdles. Typically,such an Authority should consist of not lessthan three members having a minimum of 20years’ experience each drawn from the:

Engineering profession – preferablywith expertise in road construction/traffic engineering;

Legal profession – one who has beenor is eligible for being a High CourtJudge; and

Industry and/or financial analyst cumeconomist – familiar with the sectorand road freight industry.

46. From amongst the members, theGovernment could nominate one as theChairman and another as the Vice chairman.The tenure of a Member would be three years,renewable for one more term of three years,but not beyond the age of say sixty five. Theyshould be empowered to act in anautonomous, independent manner through alegislative enactment which would alsoprovide for its composition, qualification ofmembers, powers and functions and otherrelevant details.

47. At the State level, putting in place a fulltime regulator may not make sense until asufficiently large program of PSP is in place inwhich case regulation through the contractby a dispute resolution board may beappropriate. However, such a mechanismwould need to have sufficient powers to ensurethat any ruling was implemented quickly.

48. Institutional Framework. Reform ofsector institutions from old fashioned PWDsystems to a more effective and efficientsystem is a key building block. Alternatively,establishing a State Road DevelopmentCorporation with improved governance andmore nimble procedures may be givenresponsibility to act as a nodal agency forinduction of private investment in the sectorby commissioning project preparation andprocurement activities on behalf of states.

49. Adoption a PSP framework andenactment of accompanying laws.Formulating a comprehensive roaddevelopment policy and articulating anassociated PSP framework for the state,defining public contribution/participation, risksharing, and regulation of contracts. Theexistence of a legal framework or specificguidelines for inducting private investment

64 l

help provide comfort to investors and lay downprocesses for adoption by the PWD/RoadDevelopment Corporation. The existence ofan established policy and guidelines assumesspecific importance in ensuring consistencyand transparency.

50. Increase Incentives for PSP throughReduction of Existing Inefficiencies.Despite the incentives in place for PSP, moreremains to be done concerning whether: (i)the provisions regarding income deductionsfor tax purposes in the 1961 Income Tax Actcould be applied to all bank financing, notjust specialized financial institutions; (ii) theprovisions regarding the deduction of 20%,for tax purposes, of equity shares subscribedor debentures issued, could be extended toall financing parties to a PSP project and notjust the infrastructure company; (iii) CompanyLaw’s statutory reserve requirements of 10%could be reduced in line with internationalexperience applicable in India; (iv) thedepreciation method to estimate deferred taxliabilities could be left unspecified; and (v)other taxation inefficiencies could beaddressed. These include, for example: (i) thelimited period of tax holiday for infrastructurecompanies under Section 80 IA of the IncomeTax Act (10 years during operations); (ii) thedistribution tax on dividends paid by projectcompanies with effect from April 2003; and(iii) the minimum Alternate Tax which forcescompanies to pay tax even during the taxholiday periods.

51. Improve Project Planning,Prioritization and Justification. Bystrengthening the process of planning,prioritizing and procuring PSP projects, allroad agencies can help promote moreprivate funding to the sector. This requiresthe preparation and implementation of clearguidelines to identify the range of PSPmodels to be applied and the circumstancesunder which each would be used. Furtherprojects should be prioritized using welldefined selection criteria such as type ofvehicular traffic, traffic volumes, connectivitywith major roads, present road conditions,and decide on the form of private sectorengagement. Finally, all PSP projectsinvolving the outlay of public funds or thegranting of rights to collect tolls in amonopolistic manner for an extended period

should be subject to detailed value formoney assessment by the road agency. Theonus should be on that agency todemonstrate objectively the advantages tobe achieved from the proposed PSP. Thisassessment should be reviewed andapproved by a suitable Government bodyunconnected with the promoting agency –the Public Investment Board in the case ofNHs and the relevant unit with theDepartment of Finance for SHs. The PublicSector Comparator system used in the UKand Australia as described in Chapter 7 mayprovide a suitable model for adaptation tothe Indian environment. Applying a morerigorous project planning and preparationprocedure will also entail building skillswithin road agencies and approving bodiesto undertake such analysis.

52. Network Finance and ProjectFinance Approach. While putting in placea stronger regulatory framework, improvingthe level of information and strengtheningthe planning and procurement process willfacilitate PSP, this paper also exploreswhether for the NH network, taking aproject by project approach as well as tryingto shift a significant portion of the trafficrisk to the private sector brings the bestresults. This may result in concentratingtraff ic r isk result ing in the need torestructure projects, increasing costs tosociety as a whole and leaving more remotesections of the network behind. In addition,the progress made under the PSP approach,to br idge the f inancing gap for thedevelopment and maintenance of roads,has been achieved at a relatively high riskand cost to the Government.

53. With the exception of links havingcaptive traffic, for which the projectfinancing approach is suited, it is worthexploring whether corporate financing,based on statutory hypothecated cess anduniform tolls (established for each categoryof vehicle on a per kilometer basis), wouldattract the required funds and minimize thehigh traffic risk premium normally chargedunder PSP. Corporate financing couldprovide funds for less trafficked sections thatare not attractive to the private sector. Witha securitized revenue stream from the cessand completed toll roads, NHAI could raise

Recommendations

l 65


87 BOTs could still be used for links with high traffic and limited traffic risk, such as major bridges88 Latter assumes that NHAI was to have more financial autonomy and concomitant financial accountability.89 Recent line of credit that was proposed but yet to reach closure from Life Insurance Corporation offered at 100bp over 18 year treasury bond rate

further debt or enter into non toll basedDesign Build Finance and Operate (DBFO)agreements without GOI guarantee87. Thetotal revenues accruing through fuel cess andtolls and the relative weighting between thetwo can be subject to periodic review by theNHAI Board, subject to the obligation to justifyproposed changes in front of an independentreview panel of interested parties establishedby the GOI. The justification would clearlydemonstrate the operational and financialimpact of failure to adopt the proposed revisedrates. The panel would have the right to vetoproposed changes in the charging regime.Alternatively, the private sector could also besuccessfully utilized through maintenanceconcessions or long-term performance basedmaintenance contracts.

54. The advantages of this approach asopposed to a project by project concessionapproach are as follows: (i) diversification oftraffic risk to whole of network; (ii) capacity tocross subsidize between links on the network;(iii) greater transparency for road users on roadcharges, revenues and expenditure andtherefore increased likelihood of useracceptance of system; (iv) economies of scalein establishing national highway tolling system;(v) greater uniformity of service standards; (vi)allow for more incremental introduction oftolling; and (vii) greater incentive on NHAI tomanage efficiently all items of work on thewhole of the network so as to minimize totalcosts and maximize total revenues88. GOI maystill be free to provide concessional or grantfunding for specific projects that are clearlydemonstrated not to be financially viable butwhich the GOI wish to see completed to meetother public policy objectives such as regionalintegration or strategic requirements.Moreover, for specific links where traffic canbe more accurately forecast - bridges andbypasses – BOTs may still be employed onstandalone basis using real tolls.

55. In this structure, where current cess/tollrevenues fall short of funding for O&M, debtservicing and desired new investment, NHAIwould have two options. First, it could issuenon GOI guaranteed debt on the strength of

its future anticipated revenues alone. Such debtwould be expected to be a raised at asignificant but not large premium to GOI debtof equivalent maturity89, reflecting investorperceptions of the operational risks of costoverruns and revenue shortfalls. NHAI canmanage its financial risks through furtherdeveloping its Treasury function and using thederivatives market to hedge unwanted risks.Second, NHAI could also enter into non tollbased DBFO contracts, using either straightannuity, lane availability or active managementpayments. If NHAI still considered that aconcessionaire has some but limited controlof the levels of traffic, a hybrid option of usingboth annuity and incremental shadow toll canbe adopted - a base payment which is assuredregardless of actual traffic combined with anadditional payment per vehicle that actuallyuses the road. Implementing either financingoption would be subject to the approval of theFinance Ministry. Further strengthening theO&M implementation system, accounting andthe corporate governance arrangements wouldall be expected to have positive impacts onreducing the cost of NHAI borrowing, eitherthrough debt or concessions, by reducingcreditor perceptions of the likelihood andseverity of operational risks.

56. In fact the choice between using corporatedebt on NHAI’s balance sheet or procuring longterm DBFO contracts to elicit private fundingmay not be so relevant as it first appears. In aworld without tax, without transaction costsand with equal skills in the private and publicsectors, the economic reality of the two optionsis substantially the same – i.e. a liabilityrequiring a periodic entry in the incomestatement either as an interest payment to acreditor and/or depreciation of an asset. Inaccounting terms also, the two may also beequivalent when the DBFO contract is viewedas a lease by NHAI from the DBFO partner.International Accounting Standards 17stipulates that a lease should be classified asa capital lease and held/amortized as a liabilityon the lessee’s balance sheet if any one of thefollowing conditions prevail: (i) when the termof the lease is the major portion of the asset’slife; (ii) the present value of the lease payments

66 l

is equal or greater than the fair value of theasset; (iii) where the asset ownership revertsto the lessee at the expiry of the lease; or (iv)where the lease contains a bargain purchaseoption90. These conditions would likely prevailin a DBFO contract in India, as has been thecase with Annuity contracts. NHAI is requiredto operate and account along commercial linesand is incrementally adopting ICAI/IASstandards.

57. With these observations in mind, it isrecommended that a broad public debate beundertaken - to be informed by a specialComptroller and Auditor General efficiency/performance appraisal of the value for moneythat has been obtained through the variousPSP models in use by NHAI - prior to finalizingthe ongoing procurement process for theconcessioning of 10,000 km of nationalhighways. Of special interest in this review,should be the impact that NHAI and the states’current creditworthiness has had on the costof private sector financing and overall on thesuccess of PSP projects and the suitability ofthe institutional, legal and regulatoryframework of the Central Road Fund forsecuritization.

58. A corporate approach to PSP could alsobe explored at the state level. Corporatefinancing could be implemented through theestablishment of semi autonomous RoadDevelopment Corporations. These would be

coupled with dedicated road funds using roaduser charges, largely fuel cess, and publiclyimposed uniform tolls to assure fundingstreams into the future against which privatecapital markets can provide debt or debt/equity through DBFO contracts. States shouldnevertheless guard against creating newcontingent liabilities without due regard tostrengthening associated revenue capacity tofund new debts (see box 9.1).

59. Incorporating road users into roadcapital structures. Certain national and statehighway stretches directly serve individualproject beneficiaries in urban agglomerations.The involvement of direct project beneficiariesas part financiers of the capital structure ofsuch BOT projects may benefit such projectsthrough: (i) facilitation of greater publicacceptability of such projects, with possiblebenefits of increased acceptance of the “userpays” principle; (ii) creation of a more directlinkage between the costs and benefits of suchprojects; and (iii) creation of an additionalsource of strategic road equity for such projects.For example, road users may be included intothe capital structure of a project through sale(during project construction period) of long termpasses91 entitling the user to an unlimitednumber of trips on the project road for a limitedperiod of time. As opposed to the existingpractice of issuing monthly passes to frequentusers (after commissioning of the project), thealternative measure would sell larger value

90 The Analysis and Use of Financial Statements, White et al 1997. The same definition is used by the Accounting Standard 19 of Council of CharteredAccountants of India

91 If may be noted that various BOT roads already offer a “monthly” pass system with discounted pricing for frequent users

Box 9.1 Managing Fiscal Risks of Off Budget LiabilitiesThe following voluntary standards have been recommended for governments to better manage off budget liabilities.

• Before it accepts a new contingent liability, a government should assess the risk to its fiscal condition,including the probability of future payouts. The assessment should be conducted by an independent entity.

• The government should periodically compile an inventory of outstanding contingent liabilities and report onthe volume of these liabilities, their legal basis, and the probability of losses.

• Government fiscal analysis published in the annual budget or other documents should discuss the major riskfactors affecting revenues and expenditures for the next fiscal year or beyond.

• The government should establish a risk management strategy to guide public organizations when they takeactions that expose them to financial liability.

• The government should promote cost-and risk-sharing to discourage moral hazard, ensure the economicviability of guarantees, and reduce the probability and amount of loss.

• The budget should set aside funds , within an overall fiscal constraint, for expected losses during the year

• The budget should limit the amount of guarantee and other contingent liabilities to be tendered during theyear, as well as the total amount that each institution authorized to issue guarantees may have outstanding.

Source Government at Risk - Contingent Liabilities and Fiscal Risk, H. Polackova Brixi and A. Schick, World Bank, 2002

Recommendations

l 67


passes to such users during the constructionperiod, thereby raising part of the capitalrequired for project implementation. Theeffective cost of this instrument would be thetotal value of toll unpaid during the operationsperiod by a portion of frequent users holdingsuch passes. The choice and responsibility forissue of such road users “equity” lies with theprivate sponsors of BOT projects. However,NHAI or state governments could play apromotional role in realizing such objectivesby: (i) stipulating that a certain portion ofspecific projects should attempt to tap thispotential market on a “pilot” basis; and (ii)educating road users on the attractiveness andrationale of such schemes.

60. Corporatization/Securitization.Securitization techniques can be used tostructure securities designed to maximize theirinvestor appeal and the benefits of portfoliodiversification and secondary liquidity. Thisentails corporatization and public offerings of

road packages. In India, significant portions ofroads to be tolled are being implementedthrough contracts on a cash payment basis.The responsibility for operating and maintainingthese stretches lies with the government. Asan alternative, government agencies canconsider transfer of long-term operation,maintenance, capacity augmentation and usertoll levying rights on these roads to newlycreated and wholly owned corporate entities(SPV’s).

61. Indian government agencies haveaccessed the retail equity market to a limitedextent. The development of a vibrant marketin road equities could result in two key benefitsfor the sector in India: (i) possibility of accessingadditional funding sources for the sector; and(ii) creation of investment exit possibilities forthe existing investors, who are primarilyconstruction contractors, thereby enabling suchstrategic investors to channel existing “locked”investments into new projects.

68 l

Annexures

ANNEXURES

l 69


Annex 1 Pump Prices For Super Gasoline And Diesel

(In US Cents Per Liter) 1991-2000

Country Super Gasoline Diesel

1993 1995 1998 2000* 2002 1991 1993 1995 1998 2000

Australia 46 57 50 45 57

Bangladesh 36 47 46 52 31 26 29

Belgium 118 112 96 104 82 85 78

China 27 28 40 42 24 25 45

France 117 111 99 105 78 77 82

India * 48 56 60 66 19 21 39

Italy 118 119 97 105 86 93 83

New Zealand 61 64 48 55 32 39 34

Pakistan 47 46 53 52 20 19 27

Singapore 72 84 85 36 38

Sri Lanka 75 84 66 54 23 30 27

Uganda 79 98 86 86 83 55 71 85 68 75

United Kingdom 92 111 117 118 84 111 122

Zambia 72 60 53 72 24 66 57 49

South Africa 52 51 43 50 43 52 46 39 50

Brazil 53 63 80 92 55 38 39 34 34

Canada 47 45 41 58 51 39 36 39 47

United St. (average) 32 34 32 47 40 28 33 27 48

México 39 32 36 61 62 28 25 28 45

Argentina 79 60 94 107 30 29 28 42 52

* 2000 Price for Regular GasolineSource: Fuel Prices and Vehicle Taxation with comparative tables for more than 160 countries; Pricing Policies for Diesel Fuel,Gasoline, and Vehicle Taxation in Developing Countries by Gerhard P. Metschies

Annex 2 Excise Duty Rates, 2003

Item Excise Duty (%)

Motor Spirit 30%

High Speed Diesel (HSD) Oil 14%

Motor Vehicles*-

2-wheelers 16%-Cars 24% (including 8 percent special excise duty)

-Bus & Freight Vehicles 16%

Tyres & Tubes 24%

Parts and accessories 16%

* In addition, motor vehicles attract a cess of 0.125%

70 l

Annexures

Annex 3 Motor Vehicle Tax in Selected StatesBasis of Tax and Periodicity of Payment

Vehicle Type Andhra Karnataka Madhya Maharashtra West UttarPradesh Pradesh Bengal Pradesh

For differentCC categories Fixed or % of For different Fixed-lifefixed or % of cost % of cost of % of cost of CC categories, time

Two-Wheeler cost whichever is vehicle - life vehicle-life fixed-lifewhichever is higher-life time time timehigher-life time time

Related to Fixed or % of For differentCar weight, fixed cost %of cost of % of cost of weights, % of cost-

or % of cost whichever is vehicle-life vechile-life fixed-life time life timewhichever is higher-life time time for 5 yearhigher time

Seating Seating Seatingcapacity and Seating capacity for Seating Seating capacity

Bus distance capacity - different capacity- capacity- and Routecovered in a quarterly route km annual annual Categoryday - quarterly range - monthly quarterly

RLW - annualGoods Vehicles Laden weight- Laden weight RLW - or 7 times GVW- GVW-

quarterly quarterly quarterly annual rate for annual quarterlyone time tax

Note: In Uttar Pradesh, additional tax on passenger vehicles is charged quarterly on the basis of km run andadditional tax on goods vehicles is charged quarterly based on GVW

RLW = Registered Laden Weight; GVW = Gross Vehicle Weight

Annex 4 Rates of Sales TaxItem Andhra Karnataka Madhya Maharashtra West Uttar

Pradesh Pradesh Bengal PradeshMotor 12% 12% 12% + 15% surcharge 12% 12% 12%Vehicles = 13.8%Spare 12% 12% 12% + 15% surcharge 12% = 10% sales 8% 8%parts = 13.8% tax + 1% turnover

tax +1% surchargeTyres & 10% 8% 8% + 15% surcharge 12% = 10% sales tax 8% 8%Tubes =9.2% + 1% turnover tax

1% surchargePetrol 30.55% 28% 25% + 15% surcharge 30% plus cess @ 20.00%

= 28.75% Re 1 per litre plus cess @ 20%*Re. 1 per

litreDiesel 19.33% 17.50% 25% + 15% surcharge 34.% plus cess @ 12.55% plus 20%*

= 28.75% Re 1 per litre cess @ Re 1per litre

Lubricants 16% 15% 12% + 15% surcharge 13% 15% 20%= 13.8%

* including additional sales tax of 4% on diesel and 6% on petrol, the proceeds of which are credited to State Road Fund

Annual Incidence of Motor Vehicle Tax(in Rs per annum)

Vehicle Andhra Karnataka Madhya Maharashtra Uttar West

Pradesh Pradesh Pradesh Bengal*Two-wheeler (100 CC) 210 208 150 210 125 80Car (Santro) 2042 2625 1458 1167 729 1172Mini Bus (35 Seater) 101500 70000 48300 2485 68580* 72555Bus (55 Seater) 159500 110000 75900 3905 132360* 113355LCV (4.5 ton GVW) 2796 4800 3728 2190 3510* 3365HCV (16.2 tonne GVW) 9675 10240 12100 9360 12636* 11160MAV (28 Tonne GVW) 19304 18800 20108 18790 21840* 25610

* including additional motor vehicle tax.Note: 1. In case of life-time tax, annual tax is worked out assuming a life of 12 years. In case of West Bengal, life-

time tax is collected for 5 years.2. In case of Maharashtra in addition to motor vehicle tax, stage carriages are charged passenger tax

l 71


Annex 5 Assumptions behind the Assessment of Road Sector Revenueby Vehicle type (2001-2002)

Category Method of estimation

Central Government

Excise on motor spirit An assessment of vehicle wise consumption of fuel is made by usingfuel consumption norms (km/litre), annual vehicle utilization (in km)and number of vehicles on road. The number of vehicles on road isestimated by adjusting rexgistered vehicle fleet in light of the vehicle-wise production figures over a period of time.

Source: Indian Petroleum and Natural Gas Statistics published bythe Ministry of Petroleum and Natural Gas

Excise on motor vehicles Distributed amongst vehicle type on the basis of its assessedshare in excise collection related to vehicle production in theyear 2001-02

Excise on tyres – vehicle purchase Distributed amongst vehicle type in proportion to the vehicle wiseproduction data

Excise on tyres – use “Use related” excise is distributed on the basis of fuel consumptionshares

Excise on motor parts – vehicle 10 % of total excise collections is assigned for vehicle purchase andpurchase balance 90% for use

Excise on motor parts – vehicle use The vehicle purchase related component of excise collection isdistributed amongst vehicle types on the basis on assessed share inexcise related to motor vehicle purchaseThe use related componentis distributed amongst different vehicles using share in fuelconsumption

Cess on fuel Distributed amongst vehicle types using the share in fuelconsumption. (Diesel is adjusted for road sector consumption)

State Government

Sales tax on motor spirit Sales tax was estimated using fuel consumption by vehicle type andSales tax on diesel average sales tax rate.

An average all-India sales tax rate was calculated using the salestax rates in the different states and union territories. Similarprocedure was followed for the fuel consumption.

Sales tax on motor vehicles Distributed amongst vehicle types on the basis on production figuresof each vehicle type

Sales tax on tyres Estimated using sales tax amount per tyre and the number of tyresbeing used.

Sales tax on motor parts Estimated by applying the ratio of spare parts cost and tyre cost invehicle operation cost of each vehicle type to the sales tax collectionof tyres

Taxes on vehicles (incl. fees, fines, Total revenues under these head for all the states is obtained frompenalties, passenger and goods taxes) the RBI Study on State Finances. As vehicles in Union territories

amount to 9.7 % of motor vehicles in States, an additional 9.7% ofrevenue on motor vehicle tax revenue is added for unionterritories.This total figure is distributed amongst different vehiclesbased on the assessed share of revenue resulting from vehicles onroad and average charge per vehicle type obtained from six studystates.

Excise on diesel

72 l

Annex 7 Revenue - Cost Comparison at Vehicle Level

Item 2- wheelers Cars Jeep/taxi Bus LCV HCV MAV

All Roads excluding Urban Roads

a) Revenue and cost in Rs. per vehicle-km

Revenue 0.44 2.39 1.03 5.69 1.48 2.03 2.51

Capital Cost 0.13 0.25 0.25 0.84 0.41 1.34 1.88

Maintenance Cost (total) 0.04 0.08 0.08 0.30 0.14 0.69 0.93

Maintenance Cost (variable) 0.02 0.04 0.04 0.19 0.09 0.58 0.78

b) Revenue-Cost Ratio

With total cost 2.7 7.2 3.1 5.0 2.7 1.0 0.9

With maintenance cost 11.3 30.8 13.3 19.2 10.4 3.0 2.7

With variable maintenance cost 20.5 55.8 24.0 29.5 16.5 3.5 3.2

c) Revenue-Cost ratio with revenue contribution and cost share

With total cost 1.4 3.7 1.6 2.6 1.4 0.5 0.5



NH & SH Network

a) Revenue and cost in Rs. per vehicle-km

Revenue 0.44 2.39 1.03 5.69 1.48 2.03 2.51

Capital Cost 0.06 0.11 0.11 0.36 0.18 0.57 0.80

Maintenance Cost (total) 0.02 0.04 0.04 0.16 0.08 0.36 0.48

Maintenance Cost (variable) 0.01 0.02 0.02 0.10 0.05 0.30 0.40

b) Revenue-Cost Ratio

With total cost 5.7 15.5 6.7 10.9 5.7 2.2 2.0



c) Revenue-Cost ratio with revenue contribution and cost share

With total cost 1.4 3.8 1.7 2.7 1.4 0.5 0.5



Annex 6 Revenue Contribution by Road Network, 2001-2002 (Rs million)

Road Category Total 2- wheelers Cars Jeep/taxi Bus LCV HCV MAV

National Highways & 253680 11369 28951 12113 73494 30377 92236 5140State Highways

All other state roads 77641 7580 5790 8479 28267 10126 17294 105(District and Villageroads)

Urban Roads 168740 56847 81062 3634 11307 10126 5765 0

Total 500061 75796 115803 24226 113068 50628 115295 5245

Road User Charges Contribution by Road Network, 2001-2002(Rs million)

Road Category Total 2- wheelers Cars Jeep/taxi Bus LCV HCV MAV

National Highways & 112332 3119 7681 2749 57297 9025 30379 2082State Highways

All other state roads 36324 2079 1536 1924 22037 3008 5696 42(MDR, ODR & VR)

Urban roads 51648 15594 21507 825 8815 3008 1899 0

Total 200303 20792 30724 5498 88149 15041 37974 2125

Annexures

l 73


Annex 8 Vision 2021 Capital Investmenton Expressway, National and State Highway Network, 2001-2021 (at 1999 prices)

Scheme Period 2001-2011 Period 2011-2021

Length (km) Amount Length (km) Amount

(Rs. million) (Rs. million)

A. Expressways 3000 300,000 7000 700,000

B. National Highways

i) Four Laning/Six Laning 16,000 640,000 19,000 760.000

ii) Two-Laning with hard shoulders 15,000 187,500 7,000 87,500

iii) Strengthening Weak Pavements 20,000 150,000 24,000 180,000

iv) Bypasses, bridges, over bridges, safety Lump sum 72,500 Lump sum 92,500

and drainage measures

v) Expansion of NH System 10,000 150,000 12,000 180,000

Total for National Highways 1,200,000 1,300,000

C. State Highways

i) Four Laning/Six Laning 3,000 100,000 7,000 250,000

ii) Two-Laning with hard shoulders 35,000 280,000 60,000 500,000

iii) Strengthening Weak Pavements. 30,000 220,000 40,000 300,000

iv) Bypasses, bridges, over bridges, safety Lump sum 100,000 Lump sum 100,000and drainage measures

v) Expansion of SH System 10,000 50,000 20,000 100,000

Total of State Highways 750,000 1,250,000

74 l

Annex 9 Forms of Government Support for Road Concessions

Support Measures DescriptionComfort Letter The Chinese government commonly issues a legally non-binding letter to give

support to certain actions not clearly stated in contractual agreement such asperformance of a public corporation as a grantor of concession. These letterscan provide financiers and sponsors a minimum level of assurance when noimplicit government support is attainable. However, the disadvantage is thatthe letter is not legally binding.

Land Acquisition Expropriation of right of way for toll road construction. Cost of land acquiredmaybe borne either by the government or the concessionaire. This is commonin China, Thailand and United States and is helpful to the concessionaire becausethe right of expropriation usually resides with the government. This supportusually improves “project economics” to a great extent when implemented atno cost to the sponsors as land acquisition measures tend to cause delays in theprojects.

Extension of Concession The Indonesian government takes measures to provide compensation for thePeriod loss of profit due to circumstances caused by the government. Although this

results in improved project economics, its effect on current cash flow is negligible.Construction of Related The British and the Thai governments commonly provide for the construction ofFacilities connecting roads, access ramp, etc. This contributes significantly to the project

since connecting roads and other facil it ies are critical elements forcommencement of operation. However, construction delays may critically impairthe commencement of operation.

Revenue Support Revenue support is usually done with a minimum threshold for compensationpaid by the governments in Malaysia and China (including Hong Kong SAR). Thisprovides facilitation for the financial closing and project completion. Weakdesign may impose a large contingent liability on the government.

Revenue Sharing with In Malaysia, Thailand and the United Kingdom, deriving revenue from an existingExisting Facilities toll road facility can take the form of taking over the complete facility including

employees and assets as well as debts. Hence, there is a possibility of mitigationof revenue shortfall risk in the startup years. On the other hand, the revenuesharing formula requires careful design.

Shadow Toll The Dutch and Argentine government pay toll to the concessionaires accordingto the vehicle- kilometers of the traffic counted automatically. This provides fora means of introducing private financing without stimulating resistance to tolling.Possible financial burden/ fiscal inflexibility in later years may hinder transition toreal tolling.

Provision of Development This measure involves the transfer of right of commercial development alongRights and Third-Party the toll road to supplement project economics. The advantage is that thisRevenue enhances project economics but excessive dependence on this measure may

have just the reverse effect.Subordinated Loan A type of loan for which repayment is subordinated to the senior loan (ordinary

loan). Government, parent company and, in some cases, institutional investorsare providers of the loan. The interest rate is higher than that for a senior debt.This is a common measure in Malaysia for facilitation of financial closing becauseit is treated as equity. It could also be used as a stand-by facility to mitigate riskssuch as cost overrun and revenue shortfall. The disadvantage associated withthis form of government support is possible deterioration of project economicsdue to higher interest cost.

Foreign Exchange In Indonesia, the Philippines, and Spain, government provides compensation forGuarantees the impact of the devaluation of local currency. Such a provision is built into the

tariff formula, and is instrumental in facilitating financial closure. Thedisadvantages are:

• Possible large contingent liability on the government;• Moral hazard for concessionaire and the lenders.

Loan (Bond) Guarantees The Chinese government gives a guarantee on repayment of loan and onredemption of bond for the facilitation of project proposals and implementation.This results in a large contingent liability for the government.

Equity Guarantees Guarantee of equity investment may be given for facilitation of financial closingin foreign currency when country risk in this respect is high. Large contingentliability for the government in the event of large currency devaluation is possible.

Annexures

l75


Commercial Risk

Liquidated Damages are a part of the Concession Agreementand the responsibility for timely completion rests entirely onthe Concessionaire. Damages are linked to achieving eachProject Milestone (as against Completing entire Project). Thedamages will be paid by the Concessionaire at the rate of Rs.1 lakh per day for each project milestone and an amountcalculated at the rate of 0.01percent of the total Project Costper week for the entire Project.No form of Corporate Guarantee provided.Concessionaire has the right to commence construction at itsown risk.Independent Consultant to be appointed in consultation withthe Concessionaire.

No such notification to be issued.

- do -

Consortium Members to maintain minimum 26percent equitystake in the SPV during the entire concession period.In case of Concessionaire event of default 90percent ofoutstanding dues of the lenders would be met by NHAI.

Terms of CA would be modified by NHAI to bring theConcessionaire in substantially the same commercial andfinancial position as it was prior to the change in law in casethe effect of the said change in law is greater than Rs. 10million. In case NHAI does not do so it would be termed asNHAI event of default and in terms of the CA, all lendersoutstanding dues would be paid by NHAI.

All insurable Force Majeure Events would be covered throughsuitable insurance policies.In case of other FM events, the compensation provides for a)senior lenders dues b) 1.2 times dues of subordinated lendersc) Equity compensation to protect nominal value of 1.5 timesequity (but the actual return depends on inflation) d) negativegrant amount paid by the Concessionaire.NHAI to provide Revenue Shortfall Loan to the Concessionairein case the shortfall is due to an Indirect/Direct Political Eventof Force Majeure.

- do -

- do -

- do -Information not yet available.

- do -

Annex 10 Risk Framework of NHAI BOT Models

MORADABAD BYPASS DELHI GURGAON EXPRESSWAYProject Completion Risk

Liquidated Damages are a part of the Concession Agreement,which stipulates that in case the delay is on account of theConstruction Contractor, the LDs shall be equal to the amountreceived from the Construction Contractor, else the damageswill be paid by the Concessionaire at the rate of Rs. 1 lakh perday up to a delay of 90 days and Rs. 2 lakhs per day for adelay beyond 90 days, subject to maximum of 5percent of theCapital Cost.No form of Corporate Guarantee provided.Concessionaire has access to carry out the tests and surveysrequired.Independent Consultant to be appointed by NHAI from ashortlist of consultants.

Issue of a notification restricting the plying of CommercialTraffic/Heavy Vehicles on the existing stretch of NH-24 (throughMoradabad City) during the entire life of CA.Development of any alternate road to the Bypass restricted

Main Promoter (NHAI) to maintain minimum 51percent equitystake in the SPV during entire concession period.In case of Concessionaire event of default 90percent ofoutstanding dues of the lenders would be met by NHAI.

Terms of CA would be modified by NHAI to bring theConcessionaire in substantially the same commercial andfinancial position as it was prior to the change in law in casethe effect of the said change in law is greater than Rs. 10million. In case NHAI does not do so it would be termed asNHAI event of default and in terms of the CA, all lendersoutstanding dues would be paid by NHAI.

All insurable Force Majeure Events would be covered throughsuitable insurance policies.In case of other FM events, the compensation provides forlenders dues and equity. Equity compensation essentially aimsto protect nominal value of 1.5 times equity. But the actualreturn depends on inflation.

NHAI to provide Revenue Shortfall Loan to the Concessionairein case the shortfall is due to an Indirect/Direct Political Eventof Force Majeure.

In case of termination due to Concessionaire’s default, all hisrights and obligations will automatically stand assigned toNHAI (who may assign this to another Concessionaire).For the purpose of financing, the concessionaire has the rightto transfer and assign its rights and interest in the Project andto create security interest in the CA for the benefit of thelenders.No form of Revenue shortfall loan provided.NHAI has given an undertaking to provide subordinated loanto the extent of Rs. 5 crores to meet the shortfall in debtservicing to build debt service of 6 months principal andinterest.

If CA is terminated due to NHAI event of default or due toForce Majeure Event the lender shall have the first charge onsuch sums payable by NHAI to the Project.

Market Risk

Operating Risk

Legal Risk (Change in Law)

Force Majeure Risks

Political Risks

76 l

Annex 11 Risk Allocation Frameworkin “Model” Concession Agreement for Small Road Projects in Andhra Pradesh State

Key Risks Treatment

Delays related to • Concession Period is reckoned form the “commencement date”, which is theLand Acquisition date on which physical possession of project site is delivered to the Concessionaire;

all milestones and obligations of Concessionaire are reckoned from this date(including, for example, financial close deadline), thereby avoiding placing thisrisk on the concessionaire

Delays in • Allocated to Concessionaire, with liquidated damages imposed for each dayConstruction of delay

• Delays beyond 180 days from scheduled completion day liable to result inconcession termination

Delays in • Period for financial close is subsumed within the concession period and as suchFinancial Close financial close is not a milestone or trigger event for any other event under the

concession·

• Delays in financial close do not affect concessionaires obligations under theagreement, and concessionaire would therefore in any case be required tocomplete the project utilizing any other funding at its disposal

Delays in Approval • Independent Engineer who is appointed for review of drawings & design isof Designs & selected through a fair transparent process, with the Concessionaire havingDrawings some element of choice in the appointment process

• Independent Engineer is required to submit and comments or objections tosubmissions by the Concessionaire within a stipulated time period, failing whichthe Concessionaire can proceed with its design proposals, thereby eliminatingrisk of delays in such approvals beyond the control of the Concessionaire

Risks Related to • Responsibility for arrangement of clearances and approvals from the MinistryPermits of Railways lies with the Concessionaire, although GoAP commits to providing

“necessary assistance” to the concessionaire in this matter

• GOAP commits to awarding, on a timely basis, such approvals or permits withinits own jurisdiction

Changes in Scope • Concessionaire required to bear risks of change in project scope up to a predefinedlimit (typically 5 percent of estimated project cost”) at the behest of GoAP

• Changes in scope having higher impact to be compensated through appropriateextensions in concession agreement (calculated in such a manner as to maintainthe same project internal rate of return)

Force Majeure • Distinctions made between “Non-political”, “Indirect-political” & “Direct-political”FM events, as is NHAI’s practice

• Delays caused by FM events during construction result in extension of allmilestones by the period of delay

• Cost sharing (as is standard NHAI practice):

1. Each party bears its own costs in case of non-political FM

2. Costs are shared 50:50 in case of indirect political FM

3. Costs borne by GoAP in case of direct political FM

Termination in • FM subsistence for 120+ days can result in termination of concession at eithercase of FM events party’s option

• Compensation payments by GoAP in case of Termination as follows:

1. Non-political FM event: compensation limited to 90 percent of debt due

2. Indirect political FM event: total debt due plus 110 percent of equity subscribed in cash with adjustments for inflation and passage of time

3. Direct political FM event: total debt due plus 150 percent of equity subscribed in cash with adjustments for inflation and passage of time

• In all above, computations, compensation is reduced to the extent of pendinginsurance claims; however 80 percent of claims not admitted will be payable ascompensation by GoAP

Annexures

l 77


Termination owing • Various events of default definedto Concessionaire • Concessionaire given 60 day “cure period” to rectify defaults, failing whichEvents of Default concession may be terminated by GoAP·

• Termination payment by GoAP to Concessionaire: 90 percent of debt lessinsurance claims pending (80 percent of claims not admitted will be payable ascompensation by GoAP)

Termination owing • Various events of default definedto GoAP Events of • Concessionaire given 60 day “cure period” to rectify defaults, failing whichDefault concession may be terminated by GoAP

• Termination payment by GoAP to Concessionaire: entire debt due plus 150percent of equity subscribed in cash with adjustments for inflation and passageof time less insurance claims pending (80 percent of claims not admitted will bepayable as compensation by GoAP)

Traffic & Revenue • Allocated entirely to ConcessionaireRisks • No support by GoAP even in case of chronic revenue shortfalls

Competing Facility • GoAP commits that any competing facility developed during the tenure of theRisk Concession period will necessarily be a tolled facility with vehicle users on the

competing facility being charged user tolls of at least 133 percent the toll levelson the Project

Change in Law • Changes in Law resulting in “aggregate financial effect” exceeding apredetermined amount shall result in adjustments to the Concession provisionsto as to put the Concessionaire in the same financial position as before theChange in Law

• Such adjustments to take the form of extension of Concession Period

Lender Protection • Lenders protected through a Substitution Agreement (Direct Agreement) withGOAP & the Concessionaire, whereby events of default by Concessionaireunder the Financing Documents are recognized as also constituting events ofdefault under the main concession agreement, and Lenders have rights tosubstitute the Concessionaire with another concessionaire of their choice insuch cases

78 l

An

nex

12

Priv

ate

Sect

or

Co

nce

rns

wit

h B

OT

Pro

ject

s in

Ind

ia

No

Issu

eK

ey C

on

cern

sO

bse

rvat

ion

s

1.C

ount

ry R

isk

Perc

epti

on•

Neg

ativ

e pe

rcep

tion

of

risk

s as

soci

ated

wit

hPe

rcei

ved

lack

of

cred

itwor

thin

ess

of t

he r

oad

agen

cies

and

gov

ernm

ent

inve

stm

ents

in

Indi

a; p

erce

ptio

n of

exc

essi

ve r

isk

•Po

tent

ial

inve

stor

s in

dev

elop

ed w

este

rn n

atio

ns a

re e

ssen

tially

not

leve

ls l

arge

ly c

ente

rs a

roun

d:aw

are

of d

evel

opm

ents

and

inv

estm

ent

pote

ntia

l in

Ind

iao

Perc

eive

d ri

sk o

f st

ate/

cen

tral

gov

ernm

ents

•A

t pr

esen

t, t

he t

oll

road

mar

ket

in E

urop

ean

coun

trie

s is

per

ceiv

edde

faul

ting

on o

blig

atio

nsto

off

er a

mpl

e in

vest

men

t op

port

uniti

es f

or i

nves

tors

oC

once

rns

on f

unda

men

tal

viab

ility

of

toll

road

s,ba

sed

on a

ctua

l pe

rfor

man

ce o

f pr

ojec

ts o

ver

the

last

few

yea

rso

2.Pe

rcei

ved

inad

equa

cy o

f•

Inve

stor

iss

ues

in r

elat

ion

to B

OT

proj

ect

prep

arat

ion

•In

pra

ctic

e, in

vest

ors

perc

eive

tha

t th

ey c

anno

t re

ly m

uch

larg

ely

onp

roje

ct p

rep

arat

ion

incl

ude:

proj

ect

stud

ies

com

mis

sion

ed b

y ro

ad a

genc

ies,

and

req

uire

oPe

rcei

ved

inac

cura

cies

in

tech

nica

l st

udie

s, w

ith

unde

rtak

ing

deta

iled

eval

uati

on o

f th

eir

own

befo

re m

akin

gco

nseq

uent

im

pact

on

proj

ect

cost

sin

vest

men

t de

cisi

ons.

oC

once

rns

over

acc

urac

y of

his

tori

cal

traf

fic

coun

ts•

How

ever

, it

is n

ot f

easi

ble

for

an i

nves

tor

to m

ake

a th

orou

gho

Rela

tive

lack

of

flexi

bilit

y on

the

par

t of

roa

d a

proj

ect

eval

uatio

n to

the

pre

cise

deg

ree

of a

ccur

acy

requ

ired

with

inge

ncie

s w

ith

resp

ect

to k

ey p

roje

ct p

aram

eter

sth

e ty

pica

l tim

efra

me

avai

labl

e du

ring

the

proj

ect

prop

osal

pro

cess

.e.

g. D

PRs

are

base

d on

fro

zen

proj

ect

spec

ifica

tions

Ther

efor

e, i

nves

tors

per

forc

e ne

ed t

o re

ly a

t le

ast

in p

art

on t

hean

d co

sts

are

calc

ulat

ed t

here

on;

how

ever

, in

road

age

ncy’

s te

chni

cal s

tudi

es. A

ccor

ding

ly, i

nves

tors

rou

tinel

y m

ake

prac

tice

act

ual

proj

ect

spec

ific

atio

ns a

re l

ikel

y to

gene

rous

allo

wan

ces

for

quan

tity

var

iati

ons

and

unfo

rese

end

iffe

rci

rcum

stan

ces

in t

heir

proj

ect

cost

est

imat

es•

It is

com

mon

to

find

in p

ract

ice

that

DPR

est

imat

es o

f pr

ojec

t co

star

e lo

wer

tha

n bi

dder

’s es

timat

es o

r ac

tual

pro

ject

cos

ts.

The

DPR

estim

ate

of p

roje

ct c

ost

is n

orm

ally

inc

orpo

rate

d in

the

con

cess

ion

agre

emen

t, a

nd t

erm

inat

ion

com

pens

atio

n pa

ymen

ts a

re l

imit

edby

thi

s co

st e

stim

ate

•Th

ere

does

not

app

ear

to b

e a

form

aliz

ed p

roce

ss f

or a

ppro

val

ofch

ange

s in

pro

ject

con

figur

atio

n (w

ithin

rea

sona

ble

boun

darie

s)af

ter

awar

d of

a p

roje

ct t

o a

pref

erre

d bi

dder

Annexures

l 79


No

Issu

eK

ey C

on

cern

sO

bse

rvat

ion

s

3.Re

gula

tory

Iss

ues/

Ris

k re

late

d•

At

all

times

dur

ing

the

proj

ect

life,

the

Con

cess

iona

ire•

It is

obs

erve

d th

at in

vest

ors

perc

eive

diff

eren

t pr

ojec

ts b

eing

off

ered

to p

erfo

rman

ce o

fre

quir

es a

ctio

n or

int

erve

ntio

n by

Gov

ernm

ent

by t

he s

ame

inst

itutio

n (N

HA

I) as

con

side

rabl

y di

ffer

ent

in t

erm

s of

Gov

ernm

ent

Obl

igat

ions

agen

cies

in

vario

us a

spec

ts,

incl

udin

g:ris

k pr

ofile

. Th

is i

s in

spi

te o

f th

e fa

ct t

hat

traf

fic o

rigin

atio

n fo

ro

Tim

ely

Toll

notif

icat

ions

– i

n pr

actic

e no

tific

atio

nsna

tiona

l hi

ghw

ays

is n

ot s

olel

y de

pend

ent

on t

he p

artic

ular

Sta

tere

late

d to

tol

l es

cala

tion

are

ofte

n de

laye

d an

dth

roug

h w

hich

the

hig

hway

pas

ses.

The

diff

eren

ce in

ris

k pe

rcep

tion

ther

e is

no

com

pens

atio

n to

a c

once

ssio

naire

for

is a

ttrib

uted

par

tly t

o pe

rcei

ved

diff

eren

ces

in s

uppo

rt l

ikel

y t

belo

ss i

n pr

ofit

durin

g no

tific

atio

n de

lay

perio

dre

ceiv

ed f

rom

dif

fere

nt

Stat

e G

ove

rnm

ents

oTa

x no

tifi

cati

on–n

otif

icat

ions

fro

m C

BDT

rela

ted

to e

xem

ptio

ns u

nder

10

23 G

may

be

dela

yed

inp

ract

ice

oPr

acti

cal

impl

emen

tati

on o

f St

ate

Supp

ort

– In

prac

tice

, co

nces

sion

aire

s ha

ve f

aced

pro

blem

sre

late

d ei

ther

to

timel

y si

gnin

g of

sta

te s

uppo

rtag

reem

ents

wit

h r

esp

ecti

ve s

tate

go

vern

men

ts,

or a

ctua

l re

ceip

t of

all

nece

ssar

y su

ppor

t as

envi

sage

d in

the

se a

gree

men

ts,

part

icul

arly

at

the

loca

l le

vel

4.Fi

nanc

ing

Mar

ket

Dep

th•

Gen

eral

ly,

Cre

dibl

e In

vest

ors

do n

ot a

ppea

r to

•Th

e Pr

imar

y In

vest

ors

in B

OT

road

pro

ject

s ar

e in

varia

bly

cont

ract

ing

perc

eive

a s

igni

fican

t la

ck o

f de

bt f

inan

cing

opt

ions

com

pani

es o

r su

bsid

iarie

s of

gro

ups

with

spe

cific

foc

us o

n ro

adin

the

cou

ntry

cont

ract

ing.

Suc

h in

vest

ors

have

lim

ited

inve

stib

le s

urpl

uses

for

BO

T•

How

ever

, th

ere

is a

cle

ar l

ack

of f

inan

cial

inv

esto

rspr

ojec

ts;

ther

efor

e th

e pr

esen

ce o

f fi

nanc

ial

inve

stor

s w

ould

for

brid

ging

equ

ity g

aps

in s

uch

proj

ects

eff

ectiv

ely

exte

nd t

he a

bilit

y of

the

con

trac

ting

com

mun

ity i

n In

dia

to f

und

mor

e BO

T pr

ojec

ts

5.In

vest

or E

xit

•Th

ere

are

no r

eal

exit

optio

ns f

or i

nves

tors

in

road

•Th

e la

ck o

f fin

anci

al i

nves

tors

in

such

pro

ject

s is

lar

gely

att

ribut

able

proj

ects

tod

ay.

No

proj

ect

has

atte

mpt

ed a

pub

licto

the

lac

k of

exi

t op

tions

aft

er a

rea

sona

ble

inve

stm

ent

teno

r.of

ferin

g of

equ

ity,

nor

are

ther

e an

y se

cond

ary

deal

sFi

nanc

ial

inve

stor

s lo

ok f

or i

nves

tmen

t m

atur

ities

of

not

mor

e th

anin

roa

d se

curit

ies

ten

year

s, w

hile

mos

t to

ll ro

ad c

once

ssio

ns a

re m

uch

long

er t

han

this

in

dura

tion,

and

equ

ity r

etur

ns w

ithin

the

firs

t te

n ye

ars

are

very

low

•Fi

nanc

ial

inve

stor

s, t

here

fore

, ne

ed t

o pa

rtia

lly “

unlo

ck”

thei

rin

vest

men

ts i

n pr

ojec

t sh

ortly

aft

er c

onst

ruct

ion

in o

rder

to

be a

ble

to m

ake

inve

stm

ents

in

fres

h pr

ojec

ts.

•Th

e la

ck o

f ex

it op

tions

is

larg

ely

attr

ibut

ed t

o th

e na

scen

t st

ate

ofth

e se

ctor

in

Indi

a. T

he o

ldes

t BO

T ro

ad p

roje

ct a

war

ded

thus

far

has

bare

ly r

ecor

ded

2-3

year

s of

ope

ratio

ns,

whi

ch,

base

d on

glo

bal

expe

rienc

e, i

s to

o ea

rly f

or a

pro

ject

to

stab

ilize

in

term

s of

cas

hfl

ows,

and

att

ract

sec

onda

ry i

nves

tors

80 l

Annex 13 Practical Limits to Extent of Private InvolvementTable 1 provides a breakup of state highway length within the States under consideration, classified on thebasis of lanes and traffic density. This breakup does not include minor and village roads which are not under thejurisdiction of the respective State Public Works Departments.

• Roads with vehicular traffic of 15,000 PCUs and above will need capacity augmentation to four lanes,whereas, the rest can be rehabilitated and strengthened.

• An analysis has been conducted to broadly assess the extent of private sector participation potentialin these States, using the following assumptions:

o The States adopt a PSP model based on fee recovery through user tolls, with a partial grantdisbursed by the state to the concessionaire during the construction period to enhancereturns on equity and increase investment viability

o Projects with a traffic of 15,000+ PCU are taken up for capacity augmentation (4-laning)while projects with traffic lower than 15,000 PCU are rehabilitated.

o In both cases, the concessionaire takes up long term O&M responsibility for the duration ofthe concession period. For 4-laning projects, the total concession period is taken as 25 yearsand the concession period for other projects is taken as 17 years (the rehabilitation modelassumed is similar to the model used in Madhya Pradesh)

o Construction period in both these cases is taken as 2 years; concession period includes thetime required for construction

o Capital costs for 4-laning/ rehabilitation are assumed as follows:

- Cost for 4 laning: Rs.35 mn per Km

- Cost of rehabilitation: Rs. 7.5 mn per Km

- Cap on Grant of 40 percent of the Construction Cost

Table 1: Length of State Highways

(in km)

Traffic in PCUs

State <5000 5000-10000 10000-15000 15000-20000 >20000

Lane Type: 1.5

Andhra Pradesh 70.0 32.6 137.0 0.0 0.0

Karnataka 4028.5 2658.6 1211.2 6.5 0.0

Madhya Pradesh 1987.6 21.0 0.0 0.0 0.0

Maharashtra 9967.6 2742.1 1193.7 130.8 26.3

Uttar Pradesh 4222.7 216.0 109.5 0.0 0.0

West Bengal 1415.7 431.7 51.7 0.0 0.0

Lane Type: 2.0

Andhra Pradesh 234.2 1347.6 184.6 38.0 0.0

Karnataka 60.3 20.2 0.0 0.0 0.0

Madhya Pradesh 217.0 39.0 0.0 0.0 0.0

Maharashtra 1105.7 1507.9 370.1 43.1 32.8

Uttar Pradesh 215.5 693.5 138.7 0.0 0.0

West Bengal 0.0 0.0 105.4 0.0 0.0

Annexures

l 81


o Other Assumptions:

Traffic Growth 5 percent per annum

Toll rate per PCU 0.50 Rs. per PCU (for 4 laned roads)

Toll rate per PCU 0.25 Rs. per PCU (for 2 & 1.5 lane roads)

Escalation 3 percent per annum

Interest Rate 14 percent per annum

Routine O&M 1 percent of original capex per annum

Periodic O&M 5 percent of original capex every five years

Inflation 3 percent p.a.

Tax rate 35 percent

For the purpose of this analysis, the threshold return requirement for a project to be taken up by a privateinvestor has been assumed as 20 percent Equity IRR.The main observations are as follows:

• Under the model assumptions indicated above, the grant requirement from government to enable theprivate investor to achieve the required returns on equity investment in roads with traffic density below5,000 PCU is well in excess of 70 percent of the project construction cost. This leaves little room for fundleveraging from commercial lenders. It is suggested that for this reason, this category of roads may not beconsidered for PSP under a BOT model

• The grant requirement for development of the remaining road network under a PSP framework isindicated in the Table 2 (grants are expressed in terms of Rs. mn per annum; disbursement is required atthis level per year for two years i.e. over the construction period of the projects)

The success of the PSP model considered here is contingent in part upon the confidence of private investors aswell as bankers/lenders in receiving the required grant disbursements over the construction period from theState Government. In order to mitigate such concerns, the State Governments may consider ring-fencing anidentified revenue stream and disbursing the required grant amounts from this identified stream.

Table 2: Grant per Annum for Rehabilitation of State Highways (Rs. million)

Traffic in PCUs

State 5000-10000 10000-15000 15000-20000 >20000

Lane Type:1.5

Andhra Pradesh Grant per annum 4.3 2.6 0.0 0.0

Karnataka Grant per annum 348.9 22.7 4.6 0.0

Madhya Pradesh Grant per annum 2.8 0.0 0.0 0.0

Maharashtra Grant per annum 359.9 22.4 91.6 8.3

Uttar Pradesh Grant per annum 28.4 2.1 0.0 0.0

West Bengal Grant per annum 56.7 1.0 0.0 0.0

Lane Type: 2

Andhra Pradesh Grant per annum 176.9 3.5 26.6 0.0

Karnataka Grant per annum 2.7 0.0 0.0 0.0

Madhya Pradesh Grant per annum 5.1 0.0 0.0 0.0

Maharashtra Grant per annum 197.9 6.9 30.2 11.3

Uttar Pradesh Grant per annum 91.0 2.6 0.0 0.0

West Bengal Grant per annum 0.0 2.0 0.0 0.0

82 l

EUROPEAN CONFERENCE OF MINISTERS OFTRANSPORT (ECMT)92

History: The European Conference of Ministers ofTransport (ECMT) is an intergovernmental organizationestablished by a protocol signed in Brussels on 17October 1953 by fifteen European countries. It is aforum in which Ministers responsible for transport, andmore specifically inland transport, can co-operate onpolicy. Within this forum, Ministers can openly discusscurrent problems and agree upon joint approaches toimprove the utilization and rational development ofEuropean transport systems of internationalimportance. Today the ECMT comprises of 43 fullmember countries, 7 associate countries and 1 observercountry.

Mission: the ECMT’s role primarily consists of:

• Helping to create an integrated transportsystem throughout the enlarged Europe thatis economically and technically efficient,meets the highest possible safety andenvironmental standards and takes fullaccount of the social dimension.

• Helping to build a bridge between theEuropean Union and the rest of the continentat a political level.

• Providing a forum for analysis and discussionon forward looking transport policy issues.

Structure: The Council of the Conference comprisesthe Ministers of Transport, the main body of theConference. A Chairman is appointed annually fromthe Council and is assisted by two Vice-Chairmen. Anannual Ministerial Session of the Conference is hostedby the country holding the chairmanship. The mainformal decisions of Ministers are contained inResolutions, Recommendations and other acts agreedby the Council. A Committee of Deputies, composed ofsenior civil servants, prepares proposals for considerationby the Council of Ministers. The Committee is assistedby working groups, each of which has a specificmandate. The Secretary General heads the Secretariatin its role of assisting the statutory bodies. TheSecretariat consists of three units: Transport Policy,Economic Research and Statistics, and Communicationsand Administration.

Recent Initiatives:

At present, the ECMT has working groups for thefollowing areas: Access and Inclusion; Combating crime

and terrorism in transport; combined transport;economic research; fiscal and financial aspects oftransport; integration of new member states;railways; road safety; road transport; statistics;sustainable urban travel ; t ransport andenvironment; trends in traffic; and infrastructureinvestment.

INTERNATIONAL FUEL TAX AGREEMENT (IFTA)93

The International Fuel Tax Agreement (IFTA) is anuniform system across all US states (except Alaska& Hawaii) and Canadian provinces (except theNorthwest Territories, Nunavut and Yukon) foradministrating and collecting fuel consumption taxesfrom inter-jurisdictional motor carriers (IJCs)

History: In 1982 representatives of the States ofArizona, Iowa and Washington expanded thediscussions of uniformity for motor carrier fuel usetax reporting to an experimental agreement knownas the International Fuel Tax Agreement (IFTA). In1986, 1987, and the early part of 1988 the statesof Oklahoma, Minnesota, Idaho, and South Dakotajoined IFTA. During this period of time there was asubcommittee of the working group developing arecommended “Model Base State Fuel Use TaxReporting Agreement.” In 1990 it was decided thatIFTA should become an independent organizationhaving legal status. In 1991 the corporation wascreated as the International Fuel Tax Association,Inc. (Association). Congress passed the ISTEA,mandating membership in IFTA not later thanSeptember 30, 1996. Today all 48 continental statesand 10 Canadian provinces are members of IFTA.

Benefits: The purpose of this agreement is toestablish and maintain the concept of one fuellicense and administering base jurisdiction for eachlicensee and to provide that a l icensee’s basejurisdiction will have the primary responsibility foradministering the International Fuel Tax Agreementand executing all its provisions with respect to sucha licensee.

IFTA reporting significantly reduces the paperworkand compliance burdens for fuel tax reporting.Carriers file a single quarterly return with a singlepayment to their base jurisdiction that covers alltheir travel in other IFTA member jurisdictions. Thebase jurisdiction processes the IFTA tax return andforwards funds to, or requests funds from, eachjurisdiction for net fuel taxes.

Annex 14 Examples of Inter-State Cooperation on Tax Instruments

92 http://www1.oecd.org/cem/93 http://www.iftach.org/index4.htm

Annexures

l 83


Annex 15 Measures to Compensate Non-road Users for Levies on Diesel

Measures Countries Accepted Ways Adopted Remarks

Weight-Distance New Zealand, Iceland Licenses issued for a certa For diesel vehicles onlycharges in distance to be traveled

and graduated according tothe axle configuration andgross vehicle weight

Diesel exemption Central African Republic, Users to file an exemption Evasion & leakagecertificate Chad, Sierra Leone, US certificate, certifying that unavoidable

diesel will not be used topower road vehicles

Colouring untaxed Finland France, Pakistan, Vehicles prohibited from having Penalties imposed indiesel US, UK dyed diesel in their fuel tanks case of non-compliance

Reimbursement of New Zealand, US A tax exemption card or Costlytaxes paid to non complete documentation withroad users request for refund

Compensating Non Latvia, Mozambique Non road user groups In Mozambique, 20Road users Ex-post compensated based on their percent of diesel

outputs and average diesel levy goes into a specialconsumption per unit of fund for supportingoutput farmers

Special Zambia Concerns of farmers looked No exemption as sucharrangements into – farmers as activefor farmers decision makers in fixing the

level of fuel levy and howproceeds are to be spent

84 l