Mexico - Fourth Road Project

Document of The World Bank FILE COPY FOR OFFICIAL USE ONLY Report No.2575 PROJECT PERFORMANCE AUDIT REPORT MEXICO FOURTH ROAD PROJECT (LOAN 695-ME) June 29, 1979 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  FOR OFFCIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT MEXICO FOURTH ROAD PROJECT (LOAN 695-ME) Table of Contents Page No. PREFACE 11 PROJECT PERFORMANCE AUDIT BASIC DATA SHEET iii HIGHLIGHTS iv PROJECT PERFORMANCE AUDIT MEMORANDUM 1 - 4 ATTACHMENT: PROJECT COMPLETION REPORT I. Project Identification, Preparation, Appraisal, Negotiations and Loan Confirmation 5 II. Project Implementation 8 III. Cost Estimates, Financing and Disbursements 11 IV. Economic Reevaluation 12 V. Performance of the Borrower 14 VI. Conclusions 15 Annexes 1. List and Cost Estimates of Roads Originally Proposed for Inclusion into the Project 17 2. List and Cost Estimates of Project Roads 18 3. Comparison between the Original and Final Cost Estimates 19 4. Comparison between Forecast and Actual Traffic Volumes (VPD) 20 5. Comparison between Appraisal and Revised Economic Returns 21 Map (IBRD 2822R) Fourth Road Project - Main Highway Network This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  - ii - PROJECT PERFORMANCE AUDIT REPORT MEXICO FOURTH ROAD PROJECT (LOAN 695-ME) Preface This report presents a performance audit of the Mexico Fourth Road Project for which Loan 695-ME in the amount of US$21.8 million was closed, fully disbursed, in June 1977. It consists of an audit memoran- dum prepared by the Operations Evaluation Department (OED) and a Project Completion Report (PCR), prepared by the Bank's Latin America and the Caribbean Regional Office. The audit memorandum presents an abbreviated review of the project based on the Appraisal and President's Reports and on the PCR. An OED mission which visited Mexico during January 1979 in connection with another project had an opportunity to exchange views with the Gov- ernment and clarify certain aspects of the PCR. The views expressed by Government officials on the cost increases, traffic forecasting, the relevance of Bank-required economic data and the performance of local contractors have been taken into account in preparing the audit memoran- dum. The assistance provided by the Government, especially the Secre- tarla de Asentamientos Humanos y Obras Pdblicas (SAHOP), is gratefully acknowledged. A draft of this report was sent to Mexico but no further comments were received. In addition to providing a project summary, the audit memorandum comments on the project cost overrun, the problem of traffic projection, the contracting experience, the desirability of involving borrowers in Bank project work, the inadequate provision of local funds and changes in Government priority with respect to road works.  - iii - PROJECT PERFORMANCE AUDIT BASIC DATA SHEET MEXICO FOURTH ROAD PROJECT (LOAN 695-ME) KEY PROJECT DATA Original Actual or Item Plan Current Estimate Total project cost (US$ million) 56.8 93.a Overrun () - 65 Loan Amount (US$ million) 21.8 21._-- Disbursed ) - 21.8 Repaid ) as of 12/31/78 - 2.0 Outstanding ) - 19.8 Date physical components completed 12/74 9/77 Economic rate of return 10 to 29% 23 to 74% Financial performance - Fair Institutional performance - Good Cumulative Estimated and Actual Disbursements (US$ Million) FY 71 72 73 74 75 76 77 (i) Estimated 1.5 6.7 13.5 19.2 21.8 -- -- (ii) Actual 0.1 1.6 6.3 12.1 16.5 18.9 21.8 % of (ii) to (i) 6 24 47 63 76 87 100 OTHER PROJECT DATA First mention in files or timetable 4/69 - Negotiations 5/13-5/18/70 - Board Approval - 6/16/70 Loan Agreement - 6/26/70 Effectiveness - 10/1/70 Closing date 6/30/75 6/30/77 Borrower - Nacional Financiera S.A. Executing Agency Secretaria de Obras Publicas Fiscal year of Borrower 1/31 - 12/31 Follow-on project name Seventh Highway Project (Loan 968-ME) of 1974 for US$90 million MISSION DATA Month/ No. of No. of Date of Item Year Weeks Persons Manweeks Full Report Identification 3/69 3.3 2 6.6 4/25/69 Appraisal (initial) 11/69 2 n.a. 12/3/69 Appraisal (final) 4/70 .9 1 .9 6/1/70 Supervision 9/70 1.7 1 1.7 10/13/70 Supervision 3/71 1.4 3 4.2 4/15/71 Supervision 11/71 0.4 1 0.4 11/30/71 Supervision 5/72 1.8 2 3.6 5/30/72 Supervision 12/72 0.7 1 0.7 1/03/73 Supervision 2/73 0.4 1 0.4 3/09/73 Supervision 11/74 1.7 1 1.7 12/27/74 Supervision 8/75 1.4 1 1.4 8/18/75 Supervision 7/76 1.4 1 1.4 8/30/76 Supervision 9/77 1.1 1 1.1 10/11/77 COUNTRY EXCHANGE RATES Name of Currency (Abbreviation) Peso (Ps) Appraisal Year Average (1970) Exchange Rate: US$1=Ps 12.50 1971-1975 US$1=Ps 12.50 1976 US$1lPs 15.00 1977 US$1-Ps 22.50 /a Excludes US$3.0 million for interest and other charges. lb Includes US$3.0 million for interest and other charRes. Tc In August 1976, the Peso floated and fluctuated at various rates; in 1977, it stabilized at about US$1.00 = Ps 22.50.  - iv - PROJECT PERFORMANCE AUDIT REPORT MEXICO FOURTH ROAD PROJECT (LOAN 695-ME) Highlights The project, which helped finance construction or improvement of 10 road sections (1,042 km), has been successfully completed. The reestimated rates of return on investment average about 45% compared with the appraisal estimate of 20%. The rapid growth of traffic on these roads is an indication of the project's important developmental impact. Points of special interest are: - rapid inflation and inadequate engineering contributed to cost increases (paras. 3-4 and PCR paras. 3.01 and 4.01); - insufficient local contractor capacity (paras. 3-4, 7 and PCR paras. 2.05-2.06); - the reestimated project return is more than double the ap- praisal estimate, indicating very inaccurate traffic projection (para. 5 and PCR paras. 4.02-4.03); - the need for closer collaboration between the Borrower and the Bank staff during project preparation (para. 6); and - insufficient allocation of local funds to the project resulting from changes in Borrower's investment priorities during project implementation (para. 8 and PCR paras. 6.02-6.03).  PROJECT PERFORMANCE AUDIT MEMORANDUM MEXICO FOURTH ROAD PROJECT (LOAN 695-ME) 1. The Mexico Fourth Road Project was supported by Loan 695-ME of 1970. It was the sixth Bank loan for highways in Mexico, the first five projects having financed improvement of over 11,000 km of federal highways and 844 km of toll roads and purchase of maintenance equipment. Following this project, one other highway loan was made in 1974 and a Highway Sector Project was approved in March 1979. 2. The Fourth Road Project comprised: (a) construction of five roads totalling 462 km; and (b) improvement of another five roads totalling 580 km (PCR, Annex 2). The Borrower had agreed to complete with its own resources the construction of 120 km of feeder roads to provide access to one of the Bank-financed roads. Also, the toll authority, Caminos y Puentes Federales de Ingresos y Servicios Conexos (CAMINOS), had agreed to undertake a study of the toll rate structure. 3. The project was completed in September 1977, three years later than expected at appraisal. The delay in completing the roads resulted from insufficient contractor financial capacity to cope with cost increases, lead- ing to the need to rebid or negotiate new contracts and to a slowing of work because of insufficient local funds during a period of runaway inflation (PCR, paras. 2.01-2.09 and 4.01). Unanticipated design changes including extra drainage, and inclement weather also contributed to the delay. The toll rate structure study and the feeder roads required were both completed within the appraisal schedule (PCR, paras. 2.09 and 2.11). Other covenants of the loan required that vehicle weight limitations be effectively enforced and that the Control Office, established under the Third Highway Project (Loan 528-ME of 1968), continue to supervise Bank-financed projects. For the most part, the Borrower complied with these undertakings but, in the process, some difficulties were encountered, as explained in the PCR (paras. 2.12-2.13 and 5.04-5.07). 4. The total project cost of US$93.5 million represents an overrun of 65% over the appraisal estimate of US$56.8 million, including contingen- cies but excluding interest charges (PCR, Annex 3). The PCR attributes the overrun to rapid inflation in Mexico during the construction period, 1972- 1976, during which wholesale prices increased by 92% (PCR, para. 4.01). - 2 - While cost increases are normal in an environment characterized by such rapid inflation, at the same time, in another part of the PCR, it is sug- gested that substantial cost increases were caused by the extensive revi- sion of engineering prepared by the Secretarfa de Obras Publicas (SAHOP), which involved increased quantities (PCR, para. 2.05 and 5.03). This lat- ter view was confirmed by OED staff's discussions with Government officials, which indicated that engineering had not been satisfactorily prepared due to shortage of qualified engineering staff to undertake the work. There appears to be a frequent underestimation of costs in the engineering done by SAHOP; for instance, cost overruns in the Mexico Third Highway Project resulted as the engineering was not carefully prepared because of time pressure (see Project Performance Audit Report No. 832, paras. 3.08, 3.10- 3.12). It appears therefore that, while inflation was an important cause for the cost overrun, the engineering underestimation was also a factor. On the basis of information provided in the PCR and by the Government, the audit was unable to determine the magnitude of contribution to cost rise of each of these two factors. 5. The roads have fostered the development of the regions they serve as evidenced by the rapid growth in traffic since their completion (para. 5). The rates of return on the project roads on completion range between 23% and 74% with an average of about 45%, compared with the appraisal expectation of 10% to 29% and an average of 20% (PCR, Annex 5); this better than expected economic result is mainly due to much greater than expected traffic on most of the roads -- on one road the actual traffic was 8-1/2 times the appraisal projection (PCR, Annex 4). This very inaccurate traffic projection is hard to explain. In fact, the PCR did not make an attempt to do so; it merely notes that the Government's assumed average rates of traffic growth applicable to project roads were scaled down by the appraisal and that even the Govern- ment's assumption turned out to be a considerable underestimation. The basis for the scaling down was not explained in the appraisal report, except to say that the assumed rates "may be optimistic for existing highways". Although traffic was generally underestimated, this is not true for all of the roads; on some roads, traffic was, in fact, overestimated. Even for the majority of roads where the underestimation was involved, the rates of underestimation had a wide range of 10% and 860%. These suggest the unreliabliity of the average rates used for projecting traffic for the project roads and the need to investigate each road's traffic potential separately at the feasibility study phase. Although the kind of traffic volume involved (up to 3,500 vehicles a day presently on those roads where the traffic was higher than projected) may not have resulted in a wrong choice of design standards, the inaccurate projection could result in mixing up of the priorities among in- dividual roads both within and without the project. This indicates a need for further improvement in traffic projection techniques. The Borrower informed the audit that it is now attempting to improve traffic forecasting through the development of a traffic projection model. 6. During appraisal, the Borrower requested that the Bank include pre- investment studies in the project. However, during negotiations these studies were deleted because the Borrower considered that the terms of reference pre- - 3 - pared by the Bank were too complicated. In the event, the Borrower secured financing from another international agency for this study. During the au- dit missions's discussions with the Borrower, the Borrower mentioned that it had expressed some doubt on the need for the economic data required in the terms of reference at the time and felt that there was no adequate ex- planation by-the Bank why what appeared to be not very relevant informa- tion had to be prepared. In general, the episode suggests a need for close collaboration between the Bank and its borrowers in project preparation, inviting borrowers' involvement in the Bank's project work and making ef- forts to explain why the data are requested and what can be achieved with them. Bank staff consider that such collaboration has been achieved under the recently approved Highway Sector Loan. 7. During the project's preparation, the Bank noted the absence of foreign contractor participation in the projects it had financed up to that time. It discussed this situation with the Borrower and various foreign contractors, determining that the local contracting industry was strong, well organized and able to carry out works at a lower price, thus making it difficult for foreign firms to compete. Nonetheless, the Bank and Borrower took measures to maximize competition under this project, including packag- ing bids (PCR, para. 1.05). Despite these measures, all contracts were a- warded to local firms and, during construction, a number of these firms ex- perienced financial failure. In the end, some contracts had to be renego- tiated or rebid (PCR, paras. 2.01-2.02 and 2.05). The Project Performance Audit Report of the Third Highway Project (paras. 3.02-3.06) also cites that local contractor failure and the need to award new contracts posed a prob- lem during implementation. While, according to the PCR (para. 1.05), the Bank's conclusion that the Mexican contractors were strong and well orga- nized had been based on "research", involving inquiries with six major con- tracting firms in the United States, it is in retrospect unclear whether such a conclusion was warranted. SAHOP now appears to feel that part of the reasons for the contractor failures was that there was no program of support to local contractors, especially during the economic downturn when they experienced financial difficulties. Since, it appears, the promotion of the domestic contracting industry was an equally, if not more, important factor in concluding that all project work would be carried out by local contractors and therefore reducing the Bank financing from 36% to 33% (PCR, para. 1.09), suitable measures to ensure their financial as well as physi- cal capability to execute the work should have been formulated and put into effect. 8. Finally, under this project and the previous one (PPAR No. 832, Paras. 3.08 and 3.18), the Borrower's budgetary allocations were insuffi- cient to meet local costs in a timely way. Although the Borrower was ex- periencing economic difficulties, the inadequate funding for the Bank-as- sisted projects seems to have resulted partly from a change in highway pri- orities effected by a new administration: during the appraisal, the Bor- rower was constructing long sections of roads, whereas the new policy focus- sed on short sections of road near populated, urban areas. To the extent that this new policy correctly reflects true economic priorities, the ques- tion arises as to why these priority investments were not identified during project preparation. Indeed, some such short road sections seem to have priority in terms of the traffic volumes they serve. For exaaple, the PCR shows that such road sections later included in the project carry large traf- fic volumes -- up to 8,000 vehicles a day. - 5 - ATTACHMENT PROJECT COMPLETION REPORT MEXICO - LOAN 695-ME FOURTH ROAD PROJECT I. PROJECT IDENTIFICATION, PREPARATION, APPRAISAL, NEGOTIATIONS AND LOAN CONFIRMATION A. Project Identification and Preparation 1.01 In April 1969, the Bank sent a mission to Mexico to identify a possible Fourth Road Project comprising 13 roads totalling about 1,400 km of primary roads in the federal highway system (list in Annex 1). The prelimi- nary cost estimate of the proposed project was US$75.0 million and the tenta- tive loan amount was US$27.0 million (about 36% of total cost). The economic information and engineering studies on the proposed roads were being prepared by the Secretariat of Public Works (SOP). The preliminary list of roads had not been approved by the "Secretaria de la Presidencia" which is responsible for establishing investment priorities in all sectors of the economy. 1.02 Discussions were held between the Bank and SOP regarding inclusion in the proposed project of one toll road (Annex 1 - Item 8 - IRAPUATO-LEON). The Bank felt that road users might, generally, prefer the nearby free road, making it economically preferable to add new lanes to the existing free road. In the end, the toll road was not included in the project. 1.03 In October 1969, the Bank was informed that the Secretariat of the Presidencia did not approve the inclusion in the project of the following three roads: No. 6 - Guadalajara-Chapala No. 7 - Tepehuanes-Guanacevi No. 9 - Tlapa-Huamuxtitlan-Las Penas The list of roads was, therefore, reduced to 9 roads. Later, at SOP's request, a new road (Ciudad Victoria-Tampico) was included in the proposed project. 1.04 The economic justification for the proposed roads was based mainly on road users savings although for one of the roads (Annex 1 - No. 11, Tuxtepec-Matias Romero), the justification was to be based mainly on the increase of agricultural production. 1.05 At the preparation stage, questions were raised in the Bank on the lack of interest of foreign firms in bidding for highways in Mexico and research was done on this subject. The IDB, the Associated General Contractors - 6 - of America and six major contracting firms were contacted. The main conclu- sions were that (a) the Mexican Construction Industry is strong and well organized and is able to carry out highway works at lower prices than foreign firms and (b) the foreign firms would not be interested in bidding on contracts amounting to, say, US$2.0 to US$3.0 million (which is the size of contract desired by SOP in order to promote the local contracting industry). As a result of the research, SOP and the Bank agreed to the "packaging" of contracts which allowed for substantial increases in the total contract amounts. Under this arrangement, contractors could bid for one or more sections and the contract would be awarded on the basis of the lower of (i) the lowest evaluated bid for the entire road or (ii) the lowest eval- uated combination of bids for its sections. In addition, to maximize compe- tition, the following measures which were in force under Loan 528-ME (Third Road Project) were confirmed: (i) The Government will invite prequalification of contractors annually until such time as all contracts have been awarded; (ii) Invitations to bid for each contract will be issued to eight local and up to eight foreign firms provided a sufficient number of the latter are prequalified; (iii) Contracts will be awarded for complete road sections; and (iv) Contracts will be awarded for the total cost of specific road sections, rather than in annual tranches. 1.06 During the preparation state, the Bank also expressed to the Govern- ment, its concern over the difficult financial situation of the toll authority "Caminos y Puentes Federales de Ingresos y Servicios Conexos" (CAMINOS) to which the Bank had made two loans (No. 317-ME - US$30.5 million - June 1962 and No. 401-ME - US$32.0 million - February 1965). CAMINOS was also in heavy debt to the "Banco Nacional de Obras y Servicios Publicas" (BANOSP). Conse- quently, the Loan Agreement included (Section 7.01) as a condition of effec- tiveness that the debt from "CAMINOS" to "BANOSP" should be rescheduled under favorable conditions to be determined. The Guarantee Agreement also included a provision that the Guarantor, together with CAMINOS, should carry out a study of the toll rate structure of CAMINOS - operated facilities. B. Project Appraisal 1.07 Late in October 1969, the Bank received for review, the economic and technical information prepared by SOP for the appraisal of the proposed project. The appraisal mission was in Mexico from November 10 to 29, 1969 and comprised a civil engineer, a transport economist and an agriculturist. The mission was not able to complete the appraisal of the project because (a) the engineering for roads and bridges was not completed; (b) there were doubts on the economic justification of two roads; and (c) the route for one of the roads required further studies. - 7 - 1.08 However, drafting of the appraisal report was started and in mid- January 1970, the necessary additional information was received and the Yellow Cover Report was issued on February 26, 1970. The total cost of the project was estimated at US$58.1 million out of which the Bank would finance 36% (assuming a 20% foreign contractors participation) or about US$21.0 million. Because of the favorable experience from previous projects, the Bank agreed to SOP's proposal to carry out supervision of construction with its own staff. 1.09 By mid-March 1970, and while the Yellow Cover Report was still under review, it became evident that foreign contractors were not interested in bidding on any of the project roads. Accordingly, the Bank proposed to reduce its participation to 33% but to include in the loan appropriate amounts for interest and other charges during construction. The revised total project cost became US$56.0 million and the amount of Bank participation, US$18.5 million. 1.10 In late April 1970, the Bank sent a mission to Mexico to check the detailed engineering and final cost estimates. At this time, the Mexicans proposed the addition of about US$1.0 million to finance preinvestment studies but the proposal was dropped during negotiations. C. Negotiations and Loan Confirmation 1.11 Negotiations took place in Washington between May 13 and 18, 1970 and the Gray Cover Appraisal Report was prepared on the basis of the revised cost estimates obtained during the April 1970 mission. The total project cost was US$59.8 million and the loan amount was US$21.8 million, including US$3.0 million for financing interest and other charges during construction. The loan was approved by the Bank on June 16, 1970, and signed on June 26, 1970. The project comprised the improvement of five roads totalling about 580 km (Annex 2 and IBRD Map 2822). Details of project costs are shown in the following table. US$ Million Local Foreign Total 1. Highways (a) to be improved (5) 13.6 6.7 20.3 (b) to be constructed (5) 18.5 9.1 27.6 Subtotal 32.1 15.8 47.9 2. Right-of-Way 0.2 - 0.2 3. Contingencies (on Item 1) Physical 10% 3.2 1.6 4.8 Escalation, about 8% 2.5 1.4 3.9 Subtotal 38.0 18.8 56.8 4. Interest and Other Charges During Construction - 3.0 3.0 Total T679 1Tgr 3T7 - 8 - 1.12 The project, including the construction of about 100 km of agricul- tural feeder roads complementary to the Tuxtepec-Matias Romero road (not Bank financed), was expected to be completed by December 31, 1974. 1.13 With letter of September 24, 1970, Nacional Financiera submitted to the Bank a copy of the contract signed between BANOSP and Hacienda reschedul- ing CAMINO's debt, thus complying with the only special conditions for effec- tiveness. The loan became effective on October 1, 1970. 1.14 This Completion Report is based on information obtained from the Bank's Record Center, from LAC Files, Appraisal and Supervision Reports on Loan 695-ME and on information provided by SOP. II. PROJECT IMPLEMENTATION A. Improvement and Construction Works 2.01 In mid-August 1970, SOP invited interested contracting firms to prequalify in accordance with the agreed procedures. Four foreign firms (all US), three foreign embassies (France, Korea and China-Taiwan) and 42 local firms requested prequalification documents but at the deadline date of October 6, 1970, only the 42 local firms submitted applications and 34 were prequalified. In late 1970, two additional foreign firms (both from China-Taiwan) and six additional local firms submitted applications and were prequalified. During implementation about 18 additional local firms were prequalified. 2.02 In early October 1970, the first invitation to bid was issued for drainage and paving works on the Puerto Vallarta-Barra de Navidad Road. Six local firms submitted bids and contract was awarded in December 1970 to "Constructora el Aguila," the lowest bidder for the "package" of two lots. 2.03 By May 1972, SOP had invited bids and awarded contracts for all project roads except for the Zapotlanejo-Lagos Road for which SOP proposed to substitute two access roads to Guadalajara where traffic is substantial (about 8,000 v.p.d.) and to improve the Zapotlanejo-Lagos Road with its own funds. The cost estimate of the two access roads was about the same as the original project road. After confirming the economic evaluation of the two access roads (the ER was over 50%) and the detailed engineering, the Bank agreed to the substitution. In mid-April 1973, the project description was amended to delete the Zapotanejo-Lagos Road and substitute the Guadalajara- Ameca Junction and Guadalajara-Acatlan Junction access road. Construction contracts for those two sections were awarded in mid-1974. 2.04 Although the bidding and contract awards were made following agreed procedures, there were three cases where SOP proposed the award of the con- tract to other than the lowest bidder on the basis that the lowest proposal was not responsive. The Bank, after asking for additional information, agreed -9- to the awards in two cases but not for the third case (works on the Cardel- Veracruz Road) where the difference between the lowest bid (Ps 21.7 million) and the second lowest bid (Ps 23.8 million) was only 9%. The initial case made by SOP was not convincing and a special mission was sent to Mexico to discuss the matter with SOP authorities. The mission found that, among other reasons, the lowest bidder, was fully committed on other highway construction works contracted after prequalification and did not have enough capacity for undertaking the project road. The Bank subsequently agreed to the contract being awarded to the second lowest bidder. 2.05 Problems with contractors began in early 1973, when project works were scheduled to be well under way (original project completion date was December 1974). Generally, because of the need for increased financial capacity to cope with cost increases, some contractors failed and, after considering each particular case, SOP and the Bank agreed either to rebid or to award a contract for completion of the works to the second lowest bidder. Substantial cost increases were mainly caused by (a) changes of design standards (which implied increased quantities), (b) the need for additional drainage structures, and (c) the sharp escalation of prices which started in October 1973. At final completion (September 1977), cost increases had reached 65% overall (para 3.01). 2.06 Because of the contractor problems outlined in the preceding para- graph, budgetary allocations for project works inadequate to cover cost in- creases (para 3.02) and unexpected heavy rainy seasons, project works were not completed until September 1977, about three years behind the scheduled completion date of December 31, 1974. 2.07 Because of the shortening of distance provided by two of the proj- ect roads (Cardel-Veracruz and Compostela-Chapalilla), SOP unilaterally decided to operate them as toll facilities and toll booths were installed at strategic locations. 2.08 Paragraph 9 of the Procurement Schedule in the Loan stated that contracts for traffic signs, for bridge works over 15 m span and other minor or specialized works, not exceeding in the aggregate Ps 75.0 million could be awarded to local contractors on a competitive basis. During the implementa- tion period, SOP obtained Bank concurrence on its proposals for accomplishing such minor works and these were successfully completed. 2.09 The construction of agricultural feeder roads to the Tuxtepec-Matias Romero road was completed in late 1973. The total program included 120 km of roads with an estimated total cost of US$0. 5million. These feeder roads were included in the project but were not financed by the Bank. B. Supervision of Construction 2.10 Supervision of construction was successfully carried out by SOP staff. At the time of negotiations, it was agreed to include in the loan an amount of 6% of construction costs to finance supervision of construction. - 10 - In the cost table of the Appraisal Report (para 4.04), the 6% is included in the construction costs but because the final costs shown in Annex 3 are the contracted amounts, the 6% is shown separately. The quality of the construc- tion works performed is good. C. The Toll Rate Structure Study 2.11 The study, to be done in accordance with Section 3.11 of the Guar- antee Agreement, was carried out by "CAMINOS" in coordination with the Direc- torate of Tariffs, Secretariat of Transport and Communications (SCT) follow- ing the Outline of Objectives stated in NAFINSA's supplementary letter dated June 26, 1970. A traffic survey was prepared on free roads paralleling toll roads; historical trends and indexes of traffic growth were determined for free and toll facilities; new locations for some of the toll booths as well as some rate adjustments were proposed. The study was completed on schedule in late 1971 and some of its recommendations were implemented. D. The Control Office 2.12 Section 3.10 of the Guarantee Agreement states that the Guarantor shall continue to maintain within SOP a Control Office with the function, among others, of ensuring that the project is carried out in accordance with the detailed engineering and cost estimates approved by SOP. In January 1973, when SOP was reorganized, the Control Office, then under the authority of the technical adviser to the Sub-Secretary for Public Works, was transferred to the Directorate General of Planning and Programming. Between January 1973 and January 1976, the Control Office had a very limited function and the Bank sent various letters to SOP asking for compliance with Section 3.10. Finally, in January 1976, a "Department of International Credits" within the Planning and Programming Directorate replaced the Control Office. The new department has three sections (credit negotiations, credit management and credit imple- mentation) and is staffed with professional and support personnel. However, because of the widespread task of trying to control too many international loans and the bureaucratic structure of the office, its impact in improving the implementation of project works is rather limited. E. Enforcement of Vehicle Weights and Dimensions 2.13 Section 3.09(ii) of the Guarantee Agreement requires the Government to take reasonable steps (including the appropriate applications of regula- tions on axle loading and vehicle size) to ensure the integrity of its roads and bridges. Initially, the enforcement was carried out through fixed weigh- ing stations located at strategic points along the main trucking routes. Such enforcement proved generally ineffective because it was practically impossible to ensure sufficient reliability in the performance of the scale operators. In 1974, the Secretariat of Transport and Communications (SCT), in charge of the enforcement, took two actions to correct the situation: (i) elimination of the "Pirate" truckers (individual truckers who drove along any route without license or control), and (ii) the creation of the so-called "Certificados de Porte" which is a kind of registration card to be issued at terminal freight stations with all the information on freight, weight, route and tariff charge specified. So - 11 - far, four terminal stations are operative (Nuevo Laredo, Reynosa, Veracruz and Monterrey); another is starting operations in Guadalajara and others are under construction. The stations are built and operated jointly by SCT and the Transporters' Associations. SCT expects to maintain effective control of weights and dimensions of vehicles by means of the "Certificados de Porte" and the terminal freight stations. From the "Certificados de Porte," SCT is gathering data on axle loads, origin and destination, types of vehicle, freight transported, etc. These data will provide a basis for revising the regulations on axle loads and dimensions of vehicles, which are obsolescent. SCT also carries out weight spot checks through mobile units. The system is still at the experimental stage and it still has to prove its effectiveness. III. COST ESTIMATES, FINANCING AND DISBURSEMENTS 3.01 The outstanding characteristic of the loan was the sharp increases of final project costs over initial estimates. These were, however, mainly nominal cost increases although some real cost increases in constant currency units also occurred on some of the project roads. Total costs of the project increased 65% over initial cost estimates (para 2.05). This took place in an unprecedented period of inflation (1972-76) during which wholesale prices in Mexico increased 92% (see paragraph 4.01). An updated economic evaluation in 1974, however, indicated that benefits were inflating at least as fast as costs and the projects remained economically feasible. The comparison between appraisal estimates and final nominal costs on an annual basis is given below: Year Appraisal Final ---- Ps million ---- 1971 99.0 104.8 1972 240.9 171.8 1973 202.4 181.9 1974 166.4 198.3 1975 - 214.0 1976 - 201.3 1977 - 97.5 708.7 1,169.6 (+ 65%) (See Annex 3 for the investments road-by-road). 3.02 While construction prices were increasing rapidly, the public sector budget was not keeping pace as macro economic inflation control mea- sures were taken. The road budget was cut in real (though not in nominal) terms to reduce the impact of the public sector deficit on inflation. The domestic component for completing the projects on schedule was not available and Bank funds could not be disbursed. The Bank insisted unsuccessfully that budget allocations be increased for the projects and loan disburse- ments lagged far behind appraisal estimates as shown below: - 12 - Accumulated Disbursements in US$ Millions Fiscal Years Appraisal Estimates Actual % 1971 1.5 0.1 1 1972 6.7 1.6 24 1973 13.5 6.3 46 1974 19.2 12.1 63 1975 21.8 16.5 76 1976 - 18.9 87 1977 21.8 100 3.03 The loan disbursement percentage was reduced from 33% to 18% to adjust for the effects of price inflation and still maintain the Bank partici- pation in the total original program. Loan funds were fully disbursed in February 1977. The closing date was postponed three times: from 6/30/75 to 12/31/76 to 6/30/77. 3.04 The allocation of proceeds of the loan was revised two times as follows: Category Original August 1973 April 1974 I (improved works) 6,700,000 5,300,000 5,660,000 II (construction works) 9,100,000 12,500,000 13,140,000 III (unallocated) 3,000,000 1,000,000 - IV (interest) 3,000,000 3,000,000 3,000,000 Totals 21,800,000 21,800,000 21,800,000 IV. ECONOMIC REEVALUATION A. Performance of the Economy 4.01 During the period leading up to the appraisal in late 1969, the Mexican economy had performed well. GDP grew in excess of 6% per year between 1965 and 1969. During this period, prices were increasing at an annual rate of about 2% per year. Mexican GDP and Wholesale Price Changes 1965-1976 in Percent 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 Real Growth GDP +6.8 +6.3 +8.2 +6.3 +6.9 +3.5 +7.2 + 7.6 + 5.9 + 4.3 + 2.0 Wholesale Prices +1.0 +2.9 +1.9 +2.5 +5.9 +3.7 +2.9 +15.7 +22.5 +10.6 +22.2 Source: Special study of the Mexican economy IBRD, December 2, 1977. - 13 - This generally good performance continued through 1972 by which time only 24% of the loan was disbursed. In 1973, Mexico experienced the first year of a strong and continuing inflation. Wholesale prices increased 16% in 1973, 23% in 1974, 11% in 1975, and 22% in 1976, the period during which the bulk of the loan was being disbursed (para 3.03). Between 1972 and 1976, wholesale prices increased approximately 92%, and this is one of the main reasons why the project costs increased 65% over the appraisal estimate, and, in fact, for some of the project roads, they did not increase in real terms. The original cost estimate included a price contingency of only 8% over the project period, reasonable in view of the pre-1970 inflation rate but completely inadequate for the post-1972 inflation. Traffic 4.02 The basic approach to traffic projections was to separate the proj- ects into two categories, improvement projects that already had substantial traffic, and new projects that would be shortcuts and could expect substantial diversion traffic. Project roads 1 through 4 and 10 and 11 in Annex 4 (the latter two added after appraisal) were improvements, while projects 5 through 9 were short-cuts. Base year traffic (1969) was then estimated on the improve- ment projects presumably based on actual earlier traffic counts although this is not clear. For the shortcuts diverted traffic was to be estimated based on 0 and D surveys whose year was not specified. Given the 1969 base year traffic, the Mexicans assumed a profile of growth from the 1969 estimated base traffic of about 7% p.a. to the opening of the new facility at which time growth rates would increase to 15% p.a. for ten years and then decline again to approximately 8% p.a. for the last ten years of project life. These growth rates were reduced in the appraisal report not to exceed 10% p.a. at any time during the life of the project. 4.03 Annex 4 presents the initial 1969 base estimates together with the Bank appraisal projections for 1975, the original SOP projections for 1976, and finally the actual 1976 traffic volumes based on SOP traffic counts. In general, the actual traffic volumes were far in excess of the SAHOP 1/ projec- tions and even more in excess of the more conservative Bank appraisal projec- tions. Only in the two projects added after appraisal was there a serious overestimate of 1976 traffic volumes. The largest discrepancy was for the Ciudad Victoria-Tampico project in which the Bank projected 1975 traffic volumes of 360 vpd; SAHOP predicted 1976 volumes of 486 vpd, and the actual 1976 counts were 3,450 vpd. This project was to shorten the distance between Ciudad Victoria and Tampico by 45 km. The analysis of diverted traffic beyond these two cities was complicated, and the origin-destination information was apparently inadequate. Also during this period the Port of Tampico was becoming increasingly important. (Traffic volumes at the port more than doubled between 1971 and 1976). A somewhat similar situation existed for the project Cardel-Veracruz. Otherwise, the SAHOP estimates were reasonably accurate, the Bank estimates less so. B. Revised Rates of Return (ROR) 4.04 In 1969, SAHOP prepared an economic analysis of each project based on preliminary cost estimates and the traffic projections indicated above. 1/ In early 1977, the "Secretaria de Obras Publicas" (SOP) enlarged its functions and became the "Secretaria de Asientos Humanos y Obras Publicas" (SAHOP). - 14 - These analyses were modified in the Bank appraisal report to reflect the conservative traffic projections developed by Bank analysts. In 1977, SAHOP prepared a revised set of analyses reflecting actual 1976 traffic volumes. The results of these three efforts are presented in Annex 5. The appraisal ROR's range from a low of 10% to a high of 29%, with an average of about 20%. SAHOP intial estimates ranged between 13% and 34% with an average of about 25%, reflecting the higher SAHOP traffic projections. The much higher actual traffic volumes, together with the fact that real cost increases for most project roads were negligible, then yield considerably higher expected ROR's for the projects. It is expected now that the ROR's will range between 23% and 74% with an average of about 45%. The only project in which the ROR is expected to be much less than that originally forecast by SAHOP is the Guadalajara Periferico project added after the appraisal. Instead of an initial estimate of 60%, it is now more likely to be 26%. In an environment of such rapid growth in traffic, it would be difficult to construct a project that did not yield an acceptable rate of return, and the projects all appear to have been exceptionally good investments. V. PERFORMANCE OF THE BORROWER 5.01 The performance of the Borrower in implementing the project had mixed aspects as follows: A. Financial Performance 5.02 The project ran into difficulties when, because of cost increases, financial requirements were substantially higher than originally envisaged. The additional allocations were not available on time, thus delaying the completion of project works. Under these circumstances, the financial per- formance of the Borrower fell short of that on which the Loan Agreement was based. B. Engineering Preparation 5.03 The engineering preparation of project roads was not satisfactory and, especially on three roads (Santa Rosa-La Barca, Ciudad Aleman-Sayula and Tuxtepec-Matias Romero), extensive design changes were required; real cost increases occurred on works on these roads and there is no doubt that a more careful engineering preparation would have prevented the above situation. The performance of the Borrower on this matter, therefore, was not satisfac- tory, particularly as the engineering was fully under the Borrower's control. C. The Control Office 5.04 During a three-year period (January 1973-January 1976), the Control Office, for all practical purposes, had very little impact on project works. The situation improved in early 1976 when a "Department of International Credits" was created (para 2.13) but it had limited effect in correcting the problems in project execution. It may well be that the Department was created - 15 - too late in the project implementation period. At this stage, most items of financing, changes of design, and delays were already in force and not respon- sive to the new Department's corrective efforts. It is expected that, with strong support of the Directorate under which it operates, the Department will perform a more relevant role in the implementation of future highway projects. D. Enforcement of Regulations on Weight and Dimensions of Vehicles 5.05 The system for the control of weights and dimensions was changed from fixed scales stations to freight terminal stations (paragraph 2.14) and is now implemented only in the areas where such terminal stations are in operation. So far, the control of the freight arriving and departing from the main consumption and production center, which is Mexico City, is not carried out. 5.06 Some ten years ago, a freight terminal was built in Mexico City to control the trucks moving to the northern part of the country. However, prob- lems with the truckers caused the terminal not to operate as such and it is being used primarily as warehousing space for trucking firms. SCT expects to establish weight control to the incoming and outgoing traffic from Mexico City through other terminals to be built. Since the construction of terminals is, in general, expensive and time-consuming, the enforcement of regulations on weight and dimensions of vehicles nationwide will not be carried out for some time to come. 5.07 It should be realized that the truckers are normally opposed to any weight control; the truckers' unions are strong and that Mexico has compli- cated traffic regulations. It is likely that these factors militate against the setting up of a system of weight control and have caused the control to be improperly established. As a result, the performance of the Borrower on this element has not been up to expectations. VI. CONCLUSIONS 6.01 The total investment in roads in Mexico (including maintenance) during the period 1971-76 was about US$2,700 million. The Bank financed project cost about US$93.5 million of which the Bank loan accounted for US$21.8 million. In the overall budget of SAHOP, Bank participation was very small, less than 1%. 6.02 SAHOP tends to view the Bank participation in roads as something imposed on them by the Ministry of Finance which is concerned with financing the entire Federal budget. SAHOP then attempts to turn the Bank participa- tion to their advantage by getting formal commitments to projects which require a long completion period and to use the projects to obtain larger budget allocation than would otherwise be the case in times of fiscal austerity, citing the loan agreement, and Bank letters urging the provision of adequate domestic allocation to match the Bank commitment. - 16 - 6.03 For other than Bank and BID projects, SAHOP does not make formal commitments to projects of long duration but tends to operate on a year-to-year basis with many small segments of larger projects designed to keep the const- ruction industry stable, provide an equitable spatial distribution of invest- ment, and stay within the current budget constraint. With Bank funds tied to a few large projects, failure to allocate domestic funds for these projects for reasons of budget limitations or changes in priorities after signing the loan agreement leads to a slow disbursing project about which there is little the Bank can do. Designing a loan to finance larger numbers of small segments and allowing for changing priorities over time would be desirable if the Bank wishes to disburse larger loans over shorter periods and increase its relevance in the subsector. 6.04 The project roads, although completed far behind schedule and costing substantially more (in nominal terms) than estimated at appraisal are serving important economic areas of the country. This is reflected in the traffic growth (paragraphs 4.02 and 4.03) and the increased rates of return (paragraph 4.04). Under these circumstances, the project can be considered successful in promoting the development of the areas served by the roads. ANNEX 1 - 17 - PROJECT COMPLETION REPORT MEXICO - LOAN 695-ME FOURTH ROAD PROJECT List and Cost Estimates of Roads Originally Proposed for Inclusion in the Project Preliminary total cost Length estimate 3/ Highways Km Type 1/ Surface 2/ Ps Million Highways to be Completed 1. Chilpancingo-Tlapa 180 C ST 59.4 2. Puerto Vallarta-Barra de Navidad 220 C ST 87.1 3. Apatzingan-Tepalcatepec 70 C ST 47.5 Sub-Totals 470 194.0 Highways to be Reconstructed 4. Zapotlanejo-Lagos 160 A Rev. AC 84.5 5. Santa Rosa-La Barca 80 A Rev. AC 42.2 6. Guadalajara-Chapala 48 Special AC 66.0 Sub-Totals 288 192.7 Highways to be Constructed 7. Tepahuanes-Guanacevi 89 D ST 35.6 8. Irapuato-Leon 65 Special AC 118.8 9. Tlapa-Huamuxtitlan-Las Penas 104 D ST 58.1 10. Compostela-Chapalilla 37 B ST 29.0 11. Tuxtepec-Matias Romero 180 C ST 142.6 12. Ciudad Aleman-Sayula 136 B AC 132.0 13. Cardel-Veracruz 31 A Rev. AC 34.3 Sub-Totals 642 550.4 Totals 1,400 937.1 US$75.0 milli 1/ Design standards according to the classification of the road. 2/ ST surface treatment, AC asphaltic concrete. 3/ Including 10% for engineering and 20% for contingencies. January 1978 ANNEX 2 - 18 - PROJECT COMPLETION REPORT MEXICO - LOAN 695-ME FOURTH ROAD PROJECT List and Cost Estimates of Project Roads Cost Estimate Length (Including 18% Contingencies) Highways Km Ps$ Million (a) To be Improved 1. Chilpancingo-Tlapa 56 17.8 2. Puerto Vallarta-Barra de Navidad 221 71.3 3. Apatzingan-Tepalcatepec 67 38.9 4. Zapotlanejo-Lagos 158 96.8 5. Santa Rosa-La Barca 78 57.9 Sub-Total 580 282.7 (b) To be Constructed 6. Compostela-Chapalilla 37 27.2 7. Tuxtepec-Matias Romero 169 145.0 8. Ciudad Victoria-Tampico 90 55.2 9. Ciudad Aleman-Sayula 135 122.7 10. Cardel-Veracruz 31 33.2 Sub-Total 462 383.3 Totals 1,042 666.0 + 6% for supervision 42.0 Total Ps 708.7 US$ Equivalent 56.5 million Source: Appraisal Estimate January 1978 ANNEX 3 - 19 - PROJECT COMPLETION REPORT MEXICO - LOAN 695-ME FOURTH ROAD PROJECT Comparison Between the Original and Final Cost Estimates (Mex$ Million) Cost Estimates Length (Including Contingencies) Increases Roads Km Original Final Amount % To be Improved 1. Chilpancingo-Chilapa 56 17.8 31.0 13.2 74 2. Pto. Vallarta-B. de Navidad 221 71.3 91.1 19.8 36 3. Apatzingan-Tepalcatepec 67 38.9 39.5 0.6. -- 4. Zapotlanejo-Lagos 1/ (158) 96.8 0.3 -- -- 5. Santa Rosa-La Barca 78 57.9 132.2 74.3 128 Sub-Total 422 To be Constructed 6. Compostela-Chapalilla 37 27.2 27.2 3/ -- -- 7. Tuxtepec-Matias Romero 169 145.0 280.3 135.3 93 8. Ciudad Victoria-Tampico (Llera-Estacion Gonzalez) 90 55.2 52.1 (-3.1) -- 9. Ciudad Aleman-Sayula 135 122.7 282.7 160.0 130 10. Cardel-Veracruz 31 33.2 37.0 3/ 3.8 12 11. Guadalajara-Ameca Junction 2/ 18 -- 45.3 12. Guadalajara-Acatlan Junction 2/ 27 -- 84.7 Sub-Total 507 666.0 1,103.4 Total Km 929 + about 6% for supervision 42.7 66.2 Total 708.7 1,169.6 65 (US$56.5 (US$93.5 Million)4/ Million)4/ 1/ Deleted from the project after minor works started. 2/ Included in the project. 3/ Additional investments on these two roads for works not included in the original project (such as over and underpasses, fencing, toll booths, etc.), for operating the roads as toll facilities, have not been taken into account for cost comparisons. 4/ Does not include US$3.0 million for interest and other charges during construction. January 1978 ANNEX 4 - 20- PROJECT COMPLETION REPORT MEXICO - LOAN 695-ME FOURTH ROAD PROJECT Comparison Between Forecast and Actual Traffic Volumes (VPD) Base 1969 Bank 1975 SOP 1976 1/ 1/ 2/ 3/ Road Estimate- Estimate- 1976 Estimate-- Actual- Chilpancingo-Chilapa 220 390 535 750 Puerto Vallarta-Barra de Navidad 150 270 404 750 Apatzingan-Tepalcatepec 440 780 946 2,401 Santa Rosa-La Barca 1,670 2,640 3,475 3,467 Compostela-Chapalilla 400 720 875 791 Tuxtepec-Matias Romero 250 350 461 958 Ciudad Victoria-Tampico 250 360 486 3,450 Ciudad Aleman-Sayula 770 1,230 1,546 1,145 Cardel-Veracruz 1,000 1,590 3,179 3,504 Periferico Guadalajara-Ameca Junction4/ 14,404 7,856 Periferico Guadalajara-Acatlan Junction4/ 15,443 9,375 1/ Appraisal Report 1969 2/ Estimated by SOP in 1969 Study 3/ Based on actual 1976 traffic counts obtained from SAHOP in 1977 4/ Not included in appraisal - 21- ANNEX 5 PROJECT COMPLETION REPORT MEXICO - LOAN 695-ME FOURTH ROAD PROJECT Comparison Between Appraisal and Revised Economic Returns 1/2/ 3/ SpO4/ Initial- Final- Appraisal/ SOP SAHOP Cost Cost Rate Initial Revised Year Estimate Estimate of Rate Rate of Million Million Return of of Project Completion Pesos Pesos Return Return Chilpancingo-Chilapa 1976 18.3 42.4 23 21 33 Pue-,o Vallarta-Barra de Navidad 1974 73.0 140.2 14 22 44 Apatzingan-Tepalcatepec 1974 39.9 68.5 23 23 64 Santa Rosa-La Barca 1976 59.8 142.6 17 25 34 Compostela-Chapalilla 1974 26.5 47.6 29 30 43 Tuxtepec-Matias Romero 1976 154.8 374.7 10 14 36 Ciudad Victoria-Tampico 1974 56.5 85.5 11 13 74 Ciudad Aleman-Sayula 1976 126.3 301.3 23 29 23 Cardel-Veracruz 1974 34.0 64.8 18 34 68 Periferico Guadalajara-Ameca JunctionS/ 1976 29.7 45.4 60 26 Periferico Guadalajara-Acatlan Junction5/ 1976 53.6 81.7 N.A. 41 1/ Economic costs in 1969 pesos including engineering, physical contingencies and right of way. Source: Appraisal report. 2/ Economic costs in constant 1975 pesos. One peso in 1969 prices is approximately equal to 2.24 pesos in 1976 prices. 3/ Appraisal report estimate with lower traffic volumes. 4/ SAHOP estimate 1977 with actual 1976 traffic volume. 5/ Not included in appraisal.  PCP IB R D 2822R ----i -~¯g 9óW MARCH f978 ewich C~r.. ,C J 're P.,' g o ton,., Eloren M EXICO cT FOURTH ROAD PROAJECT ø MAIN HIGHWAY~ÑETWORK t 3 , I Pfear , ighway contructed or under construction C4 Negrtur finaned by P-iots BRO lianr Empl. rim Other main highway, Cd. Ob,egón NORTHERN PL4TEA Nuevo 99La,,d. 27-- BANK FINANCED PROJECT g -eeeeess ~s Higtwayinmproed Río •Reynce. prEondd PM r Rmad N4 Z raotamejo-Lagoi was døbited from project v cs Topoooo Po Torreón o and two short aes ona ids to Guadtljira were included. i Po o - 0 00 200 300 400 500K. op 00 10 200 2 4W LD a- Cd~GU VFtro orog OF MWEX/CO wun ztion CENrRAlTmpc AreQ coveredle N. , I Prag pour Ur,d, PrATEAa×ttanta Ppa R.. 8 c B y of CmpecheI San Bios Tepie PUrCorur Chapalulo~~~1 LaoC-utilar.lePo La Vra ahal* jolostoltion 1 ócnm pAá ixll n Aue ENNSULA adatena G d4TpttMaulo Sn Mas4 esDba s l m. .,-- Amec Va ur órea R- 5 Lplr- A AutánG d Ac-pulo.Ar-- LoRee -, sl o / 9.c. Techildnt Urupa Pho Eson,d eu/0 c Cone BiaC.lim. Buenovist" Hulti, G UAT E MA LA HONDURA, d,"NLvld.d Mzm|o Tplape tr P A C / F I C O C E A N 3 0 '050'p. 9 96. 93. g