Optimal user charges and cost recovery for roads in developing countries

Policy, Research, a.:d External Affairs WORKING PAPERS Transport Infrastructure and Urban D)eveiopmeni Department The World Bank October 1991 WPS 780 Optimal User Charges and Cost Recovery for Roads in Developing Countries Ian G. Heggie and Vincy Fon What impact do road user charges have on cost recovery? And when they fail to cover total costs, how should the resulting deficit be financed? ThcPolicx.Rcsearh.hano L xiAm f Alf Crs (\np,ex . otr: N) t I'R I - Work.ng Paperq loGs n c .hc ftrdirn-oc \1vom i or p-gr- a.1 to encourage the exchange of ideas amnong B.nk s-aff and all other, inicicsted in dcslopmrcnl i-sues 1 he-c paper, ca-y .he namnes n the author. refie-it oniv their s ic%cs and shol,d he used and crte acr,rin y T1lhe findings. iin-.eqrela;zons. anid c.c i,ons are the ouihors own Thc) should noL be aitnbu: c to *he Worl!d lBane it, Boa- of lDire,-Cors its manag1-ncnl, or ari of iLs mertnhe counines  Policy, Research, and External Affairs Transport WPS (8U This paper --a product of the 1'ransport Division, Infrastructure and Urban Development Department - is part ol' a larger effort in PRE to understand pricing, cost recovery, and efficient use of resources in transport. Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Pleazsc conitact Pant Cook, room S10-055, extension 33462 (50 pages). October 1991. The optimal charge for road use is equal to 'rhe question is, how should this dericit be variable costs for road maintenance, together financed'? with the cosLs road users impose on other road users and on the rest of society (usually confined On roads carrying heavy volumes of traffic, to the costs of road congestion). it is Tiot economically efficient to bridge the financing gap by cutting back on maintenance. One persistent question raised about such The gap has to be bridged by collecting the charges is what impact they have on cost recov- required revenues through user charges, or b) ery. And when they fail to cover total costs, how mobilizing additional general tax revenues. But the resulting deficit should be financed? the costs of mobilizing additional general tax revenues are high and, given the generally low The theoretical literature argues that if there price elasticity of demand for roads, it is nearlv are constant returns to scale in road construction always more economically efficient to collect the aand in road use. the optimal user charge will required revenues from road users. recover the capital costs of the road network and the total expenditures on .ouad maintenance. It is generally agreed that marginal costs- Empirical estimates for such a system of road corresponding to variable road maintenance user charges in Tunisia similarly suggest that costs - should be the floor below which user they would generate twice the revenues currently charges should never fall. But there is no reason spent on roads. It seems therefore that optimal to stop at marginal costs. An important group of road user charges would not only recover all costs are avoidable, attributable to individual costs but would constribute substantially to groups of users (although not to the individual general fiscal revenues. users tiiciitscives), Cuid ii 3ti.a - si1 grounds of simplicity, equity, and political Heggie and Fon examine these issues from expediency - to charge these costs against the both theoretical and practical perspectives. They appropriate user group. conclude that there arm substantial economies of scale in both road construction and road use. The remaining costs, although also avoid- Also, road maintrnance costs include a number able, are common to all users and, to minimize of fixed costs that do not vary with traffic (up to loss of consumer surplus, should be charged to lhalf of annual expenditures on road maintenance them using the inverse elasticity rule (Ramsey are usually fixed). Moreover, since roads cannot pricing). be smoothly adjusted to traffic, maiginal costs for the entire road network are significantly Heggie and Fon point out that there are lower than avc,rage costs in most developing significant dif'ferences between current user countries, unless capacity is artificiall) con- charges in Tunisia and the user charges calcu- strained by environnmental or other constraints. lated using the avoidable cost methodology Ulnder these (realistic) conditions, optimal user described in this paper. charges result in a substantial financial deficit. The PRE Working Paper Series disseminates the findings of work under vav in thc Banks Policy, RKscatch, and Extemal AffairsComplex. An objective of thescrics is to gettihese findings out quickly, even if presentations are lcss than fully-polished. The findings, interpretations, and conclusions in the'e papers do not necessarily reprcsent official Bank polhcy Produced by the PRE Dissemination Center  i OPTIMAL USER CHARGES AND COST RECOVERY FOR ROADS IN DEVELOPING COUNTRIES TABLE OF CONTENTS PA" ExecutiveSwnmary .................. iv I. INTRODUCTION .............................1.............. I II. OPTIMAL PRICING MODEL .. 3 III. ASSUMPTONS UNDERLYITNG OPTIMAL PRICING . . 9 3.1 Acceptable Simplifying Assumptions .. 9 3.2 Unacceptable Simplifying Assumptions .. 11 IV. RELAX1NG UNACCEPTABLE ASSUMPTIONS AND THE CASE OF TUNISiA ............ ......................... 15 4.1 Relaxing Unacceptable Assumptions . ......................... 15 4.2 The Case of Tunisia ......................... 17 V. FINANCING REMAINING EXPENDITURES ......................... 19 vTr C Ir rt flEVCOWERY ......................... 2A 6.1 Variable and Common Costs ............ ............. 24 6.2 Recovering Collective Common Costs ......................... 26 6.3 Is it Feasible to Charge for Congestion? ......................... 30 VII. APPLYING THE PRICIYiG MODEL -22 VIII. OTHER ISSES AND CONCLUSIONS .37 Annex 1 Calculation of Congestion Costs for Tunisia .39 Annex 2 Recent Estimates of the Price Elasticity of D=m=_d for Transpol-A ......... . . ............... . . . .... 43 References ....... 47 Fig. 1. Relationship Between Vehicle Operating Costs in Tunisia and the Roughness of the Road Pavement . 11  ii Fig. 2. Eonomies of Scale in Constructing Inter-Urban Roads ..... ............ 12 Fig. 3. Rdationship Between Road Capacity and Number of Lanes .... .......... 14 Fig. 4. Effect of Marginal and Average Cost Pricing on Total Surplus .20 Fig. 5. Threshold Value of the Marginal Cost of Public Funds as a Function of the Proportional Fall Price, Given Different Price Elasticities of Demand .21 Fig. 6. Loss of Consumer Surplus Associated with Covering Collective Common Costs . . ........... 27 Fig. 7. Using the Inverse Elasticity Rule to Compute the Mark-ups Needed to Cover Collective Common Costs .... 29 Table 1. Changing Condition of Brazil's Road Network ........ 10 Table 2. Actual Costs and Those Covered by Optimal Pricing Model: Tunisia, 1982 .......... 18 Table 3. Optimal User Charges and Ad Valorem Mark-up Needed to Cover Unfunded Road Expenditures: Tunisia, 1982 .22 Table 4. Analysis of Costs in Terms of their Avoidability: Tunisia, 1982 .25 Table 5. Calculation of User Charges Using Avoidable Cost Principles and the Inverse Elasticity Rule: Tunilsia, 1982 .34 Table 6. User Charges Calculated Using Avoidable Costs Compared with those Used in Tunisia .35  Acknowledgements The report was prepared as part of a project on Pricing, Cost Recovery and Efficient Resource Use in Transport. It was prepared by Ian Heggie and Vincy Fon (Consultant). An early draft of the report was reviewed by an external panel consisting of Professors William J. Baumol, Princeton University, M. Schankerman, London School of Economics, and Kenneth Small, University of California, Irvine. The report was also reviewed by a group of Bank staff including Philip Blackshaw, EMTIN, Jose Carbajo, INUTD, Ciell Harral, ASTIN, and Vinaya Swaroop, CECPE. The author would also like to thank the many other World Bank staff and consultants who commented on specific sections of the text, or supplied some of the empirical data presented in the report. They included Jeffrey Gutman, Asif Faiz, Tim Hau, Roy Knighton, Charles J. Lane, Bill Paterson, Sigfus Sigfusson, Rachel Kranton (Consultant), Graham Smith, Rodrigo Archondo-Callao, Pradeep Mitra, Slobodan Mitric, Richard Barrett, Peter Midgley, John Flora, Dick Podolske, and Margaret Callan (Consultant) The artwork in the report was prepared by Cathy Kocak, GSDAD, and the text was finalized and formatted by Rita Vartanian and Barbara Gregory. Ian G. Heggie is a Principal Economist in the Transport Division of the Worid Bank's Infrastructure and Urban Development Department. Vincy Fon is an Associate Program Director at National Science Foundation and an Assistant Professor of Economics at George Washington University.  iv Executive Swaumary i. Current theories of road pricing argue that net economic benefits will be maximized when prices are set equal to short-run marginal costs (variable road maintenance costs and the costs of externalities, usually confined to road congestion). Several empirical studies have rlaimed to show that such a system of user charges would cover the entire costs of the road network and make a reasonable contribution to overall fiscal revenues. (Paras 1-3). ii. The above conclusions have been formalized into two theorems: (i) if there are constant returns to scale in road construction and in road use, the optimal user charge will recover the capital costs of the road network and total expenditures on road maintenance; and