Mexico - Public Enterprise Reform Loan

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 10849 PROGRAM PERFORMANCE AUDIT REPORT MEXICO PUBLIC ENTERPRISE REFORM LOAN (LOAN 3086-ME) JUNE 30, 1992 MICROFICHE COPY Report No.:10849 ME Type: (PPAR) Title: PUBLIC ENTERPRiSE REFORM Author: ANDERSON, J. Ext.:31676 Room:T9 111 Dept.:OEDD2 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. CURRENCY UNIT - PESO (MEX$) On June 7, 1991, the exchange rate in the controlled market was US$1 - Mex$ 2,990.00; the free market exchange rate stood at US$1 - Mex$3,010. ABBREVIATIONS AND ACRONYMS USED AHMSA - Altos Hornos de Mexico, S.A. (Steel Co.) CFE - Comision Federal de Electricidad (Power Co.) CGF - Comision Intersecretarial Gasto-Financiamiento (Interministerial Committee for Expenditure/Financing) CMC - Carbon y Minerales de Coahuila, S.A. (Coal Co.) CONASUPO - Compania Nacional de Subsistencia Populares (Agricultural Products) FERRONALES - Ferrocarriles Nacionales (Railways Co.) FERTIMEX - Fertilizantes Mexi-anos, S.A. (Fertilizer Co.) NAFIN - Nacional Financiera, S.N.C. (Development Bank) PA - Performance Agreement Pacto - Pacto de Solidaridad Economica (Economic Solidarity Pact) PCR - Project Completion Report PE - Public Enterprise PECE - Pacto de Estabilisacion y Crecimiento Economico (Economic Stabilization and Growth Pact) PEMEX - Petroleos Mexicanos (Petroleum Co.) PERL - Public Enterprise Reform Loan PPAR - Program Performance Audit Report QR - Quantitative Restriction SARH - Secretaria de Agricultura y Recursos Hidr6licos (Ministry of Agriculture and Water Resources) SCT - Secretaria de Comunicaciones y Transportes (Ministry of Telecommunications and Transport) SECOFI - Secretaria Je Comercio y Fomento Industrial (Ministry of Trade and Industry) SECOGEF - Secretaria de la Controloria de la Federaci6n (Controller General's Office) SEMTP - Secretaria de Energia, Minas y Industria Paraestatal (Ministry of Energy, Mines aad Public Industry) SHCP - Secretaria de Hacienda y Cr6dito POblico (Ministry of Finance and Public Credit) SIDERMEX - SiderOrgica Mexicana (Steel Holding Co.) SPP - Secretaria de Programaci6n y Presupuesto (Ministry of Planning and Budgeting) TELMEX - Telefonos de Mexico S.A. de C.V. (Telephone Co.) UCD - Unidad de Convenios de Desemperio (Performance Agreement Unit) FISCAL YEAR January 1 - December 31 THE WORLD BANK FOR OFFICIAL US ONLY Washington, D.C. 20433 U.S.A. Ofice of Diftmaweral Opermtint Evakauthm June 30, 1992 HEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Program Performance Audit Report on Mexico Public Enterprise Reform Loan (Loan 3086-ME) Attached, for information, is a copy of a report entitled "Program Performance Audit Report on Mexico - Public Enterprise Reform Loan (Loan 3086- ME)" prepared by the Operations Evaluation Department. Yves Rovani by H. Eberhard K8pp Attachment This document has a rstricted distribution and may be used by recipients only in the performance of their ofcial duties. Its contents may not otherwise be disclosed without World Bank authoriation. FOR OFFICIAL USE ONLY PROGRAM PERFORMANCE AUDIT REPORT MEXICO PUBLIC ENTERPRISE REFORM LOAN (LOAN 3086-ME) TABLE OF CONTENTS PAGE NO. BASIC DATA SHEET . . . . . . . . . . . . . . . . . . . . . . . . . 1n EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . v PROGRAM PERFORMANCE AUDIT REPORT 1. Backud ...... ....................................... Macroeconomic Developments Precedent to the PERL . . . . . . 1 The Process of Structural Reform . . . . . . . . . . . 2 Rationale for Debt Restructuring . . . . . . . . . 3 Terms of the Debt Restructuring Agreement . . . . . . . . . 5 II. The PERL In the Context of PE Reform In Mexico . ...... ............ . 9 (i Divestitureo . . . . . . . . . . . . . . 9 (ii) Enterprise Efficiency Improvement Actions . . . . . . . 10 (iii) Competition Enhancing Measures . . .... . . . . . . 10 PERL Design and Objectives . ... .. ... . . . . . o.... 10 III. Implementation Experience and Results Achle ............ ............. 13 A. Macoeconomic Performance ....... ............................. 13 The Role of the "Brady" Debt Rescheduling . . . . . . . . 14 The Role of the Disengagement Program . . . . . . . . . . 15 Sustainability of Economic Growth and Stability . . . . 16 B. The Disentagement Proram ................................. 17 Overview . . . . . . . . . . * 17 The Divestiture Program: Successes and Failures . . . . 24 Issues in Privatizaon o . .. . . .. . .. .. ... 24 Factors Contributing to the Success of the Divestiture Program * * . * . * * . .. 26 Summary and Lessons Learned . . . . . . . .. .. .. .. 27 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. TABLE OF CONTENTS (Cont'd) C. PubtEnerprise Refoins . .*.............. 28 1. Improvements in the Competitive Environment for Public Enterprise . . . . . . . . . . . . . .*. .. 28 2. Measures to Foster Managerial Autonomy and Accountability . . . . . . . . . . .* 33 3. Measures to Foster Financial Autonomy and Accountability . . .... . . . . . . . . . 39 Conclusions and Lessons . . . . .. . . . . . . . . . 40 IV. Conludift Remarks on the Bank's Role and Loan Impact.. ... ................... 43 ANNEXES 1. Mexicot Privatization of TELMEX . . . . . . . . . . . . 45 2. Mexico: Privatization of Airlines . . . . . . . . . . . 57 PROJECT COMPLETION REPORT . .................................... 65 PAR'. PROJECT REVIEW FROM BANK'S PERSPECTIVE ................... 67 A. PROJECT IDENTITY . . . . . . . . . . . . . . . . . . . . . 67 B. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . 67 Macroeconomic Setting . . . . . . . . . . . . . . . . 9 67 World Bank and IMF's Role in Mexico's Adjustment Process . . . . . . . . ....... . 68 The Public Enterprise Sector and the Government Early Reform Efforts . . . . . . . . . . . 69 C. PROJECT OBJECTIVES AND DESCRIPTION . . . . . . . . . . . 71 The Disengagement Program . . . . . . . . . . . . . . . . 71 Situation Before PERL . . . . . . . . . . . . . . . . . 71 Disengagement under the Project . . . . . . . . . . . . 71 Future Role of Government in the Sector . . . . . . . . 72 Measures to Improve Efficiency of Retained Enterprises . . 72 (a) Measures to Improve the Competitive Environment of PEs . . . . . . . . . . 72 (b) Measures to Foster Managerial Autonomy and Accountability ........ . . . . . 73 (c) Measures to Foster Financial Autonomy and Accountability . . . . ........... 73 D. PROJECT IMPLEMENTATION AND MAJOR RESULTS . . . . . . . . 73 Overall Evaluation . . . . . . . . ........... 73 Disengagement Program . . . . . * *. . .. . .. . . . . .. 74 Sector Improvement Efficiency Program . . . . . . . . . . 75 (a) Measures to Improve the Competitive TABLE OF CONTENTS (Cont'd) Environment of PEs . ......... ...... .. 75 (b) Measures to Improve PE Managerial and Financial Autonomy and Accountability . . . . . . 76 (c) Performar!e Agreements . . . . . . . . . . . . . . 76 (d) Public Investment Evaluation . . . . . . . . . . . 77 E. PROJECT SUSTAINABILITY . . . . . . . . . . . . . . . . . 78 F. BANK PERFORMANCE . . . . . . . . . . . . . . . . . . . . 79 G. BORROWER'S PERFORMANCE . . . . . . . . . . . . . . . . . 80 H. CONCLUSION . . . . . . . . . . . . . . . . . . . . . * 80 Lessons Learned . . . . . . . . . . . . . . . . . 80 PARTI. PROJECT REVIEW FROM BORROWER'S PERSPECTIVE .............. 81 A. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . 81 B. BANK'S PERFORMANCE . . . . . . . . . . . . . . . . . . . 83 (a) As a Catalyst and Coordinator . . . . . . . 83 (b) As a Creditor . . . . . . . . . . . . . . . . . . . 83 (c) As a Supervisor . . . . . . . . . . . . . . . . . . 84 (d) As an Analyst . . . . . . . . . . . . . . . . . . . 84 C. PERFORMANCE OF THE MEXICAN GOVERNMENT . ........ 86 (a) As a Borrover . . . . . . . . .. .. 86 (b) Disengagement and Restructuring of Enterprises . . 86 (bl) The Disengagement Process . . . o . . . . . . . . . 86 (i) Progress . . . . . . . . . . . . . . . . . . 86 (ii) Steel and Mining . . . . . . . . . . . . . . 87 (iii) Tel6fonos de M6xico . . . . . . . . . . . . 87 (iv) Sugar . . . . . . . . . . . . . . . . . . . 88 (b2) Restructuring of Public Enterprises . . . . . . . . 88 (i) Pilot Program of Performance Agreement . . . 88 (ii) Investment Project Evaluation Unit . . . . . 90 (iii) Legal Reforms . . . . . . . . . . . . . . . 91 (iv) Price Policy . . . . . . . . . . . . . . . . 93 D. EXPERIENCE AND LESSONS LEARNED . o . .. . . . . . . 94 PARTII. STATISTICAL INFORMATIO ..................... . .. .. .... 95 ANNEX I Conditions and Compliance Record . . . . . . . . . . . 96 ANNEX I Disengagement Program, 1982 - 91 . . . . . . . . . . . 98 PROGRAM PERFORMANCE AUDIT REPORT MEXICO PUBL C ENTERPRISE REFORM LOAN (LOAN 3086-ME) PREFACE This Program Performance Audit Report (PPAR) reviews the Public Enterprise Reform Loan. The Public Enterprise Reform Loan in the amount of US$500 million vae approved on June 13, 1989, and is expected to be closed on June 30, 1992. The PPAR was prepared by the Operations Evaluation Department (OED) and the Project Completion Report (PCR) was prepared by the Country Operations Division I, Country Department II of the Latin America and the Caribbean Regional Office. The PPAR is based on the PCR, the President's Report, the loan documents, economic and sector reports, staff files, and IMF reports. The PCR provides a satisfactory account and assessment of the loan, and discusses the performance of the Bank and the Borr-wer's executing agenciea adequately. The draft PPAR was sent to the Borrower for comments. None was received by the due date. PROGRAM PERFORMANCE AUDIT REPORT MEXICO PUBLIC ENTERPRISE REFORM LOAN (LOAN 3086-ME) BASIC DATA SHEET (Amounts In US$m) LOAN POSITION As of 1/31/92 Original Disbursed Cancelled Repaid Outstandina Loan No. 3086-HE 500 499.3 - - 499.3 CUMULATIVE LOAN DISBURSEMENT FY90 FY91 (i) Planned 499.0 1.0 (ii) Actual 499.3 0.7 (iii) (ii) as Z of (1) 100.1 70.0 PROJECT DATES Oripi.al Actual Initiating Memorandum 12/15/88 12/15/88 Letter of Development Policy 05/12/89 05/12/89 Negotiatione 03/13/89 03/13/89 Board Approval 06/13/89 06/13/89 Effectiveness 07/11/89 07/11/89 Loan Closing 06/30/91 06/30/92 Actual Completion 06/30/91 06/30/92 - iv - STAFF INPUT (staffweeks) 1989 1990 1991 TOTAL Preappraisal 2.5 -- -- 2.5 Appraisal 89.3 -- -- 89.3 Negotiation 39.6 -- -- 39.6 Supervision -- 38.5 24.0 62.5 TOTAL 131.4 38.5 24.0 193.9 MISSION DATA No. of No. of Date of Month\Year Weeks Persons Staffweeks Report Appraisal 10/88 2.0 19 38.0 05/22/89 Supervision I 09/89 1.- 4 4.0 10/18/89 Supervision II 12/89 0.5 1 0.5 12/12/89 Supervision III 05/90 1.3 2 2.6 08/24/90 Completion 10/90 0.6 2 1.2 04/15,- -v - PROGRAM PERFORMANCE AUDIT REPORT MEXICO PUBLIC ENTERPRISE REFORM LOAN (LOAN 3086-ME) EXECUTIVE SUMMARY Background 3. The PERL was part of a cluster of three adjustment loans of US$500 rail- 1. At the end of 1988, Mexico was in lion each (the others being a Finan- a state of crisis. The price index cial Sector Adjustment Loan and an had risen 159 per cent in 1987 and 52 Industrial Sector Adjustment Loan), per cent in 1988; real GDP, after designed to support the Government's having fallen by 4 per cent in 1986, program. US$125 million from each of had risen only slightly faster than these loans was formally set aside for population in 1987 and 1988; the bud- use by the Goverment in financing get deficit was above 10 per cent of collateral support for the discount GDP. The causes for slow growth were bonds to be issued under t1w debt seen to be the high rate of inflation, rescheduling plan under negctiation. the overhang from external debt with its attendant large-scale capital Loan Objcdies and Design transfers, the huge public sector deficit with its consequence of ex- 4. The PE reforms being supported by tremely high interest rates, and the the PERL were a continuation of the high level of investor insecurity with reform program initiated by the Gov- its consequence for low investment ement in the context of the 1983-88 growth. National Development Plan (NDP), which outlined a new approach for rational- 2. The incoming administration of izing the PE sector. A new Federal President Salinas adopted a comprehen- Law of Public Enterprises (FLPE) was sive program of macroeconomic stabili- enacted (1986) to promote a change in zation as well as sustained efforts to the Government's relationship with the deepen the structural reforms. The PEs from one dominated by ex-ante key elements of this program were: (i) controls to ex-post evaluation. Apart a rescheduling of the external debt to from providing financial backing for reduce the problem of debt overhang; the debt rescheduling effort, the PERL (ii) fiscal austerity incl,-ing a aimed to support continued reform of substantial down-sizing of the public the PE sector. There were two sepa- sector; (iii) tax and financial sector rate components in this effort: (i) reforms to improve resource alloca- continuation of the program of "disen- tion; and (iv) further trade liberal- gagement" that sought to reduce the ization coupled with internal deregu- size of the public sector through lation to enhance competition in d- sale, liquidation, merger or transfer mestic markets. of enterprises; and (ii) a series of - vi - reform measures designed to improve MOSIUM the efficiency of PEs. More specifi- cally, these measures included: (i) 7. The Mexican divestiture program improvements in the competitive envi- began in 1983 as part of the IMF-spon- ronment for PE9, most notably involv- sored stabilization program of that ing price liberalization; (ii) mea- period. The divestiture process can sures to foster managerial autonomy be clearly divided into two stagesq and accountability, including (a) 1983-88, and 1989-the presento corre- fundamental reforms in the nature and sponding to the 1sexenios" of Presi- function of the Boards of Directors, dents de la Madrid and Salinas. While and (b) the design and implementation the earlier period was characterized of performance agreements aimed at by divestment of large numbers of moving to a system of ex-post evalua- small enterprises, with liquidation tion rather than ex-ante control; and the most common method, the latter (iii) measures to foster financial period has seen a big jump in the autonomy and accountability. average size of enterprise, and sale is now the more usual divestiture Impentation and Resulls Achieved option. In the latter period some of the largest companies in the country, 5. Broadly speaking, the PERL was including Telmex, the telephone monop- highly successful in that it achieved oly, all the largest commercial banks, its primary goals: stabilization of Sidermex, the steel conglomerate, and the economy and major positive changes so on have been sold. in the PE sector. 8. Administrative responsibility for Macroeconomic Stabilization selling an enterprise is with a six person privatization unit in the Fi- 6. In terms of the Government's sta- nance Secretariat headed by a Coordin- bilization program, the first point is ador General. Actual management of that, thanks partly to the package of the sales is undertaken by an Agent loans of which the PERL was a part, Bank, appointed and sur-irvised by an the Mexican Government was able to official of the privs ization unit. successfully conclude a rescheduling The basic style of the aale is that of of the bulk of its external commercial a sealed bid first-price auction. debt ("the Brady deal") at terms fa- Bidders are pre-screened to make sure vorable to Mexico. Further, thanks to they are acceptable in terms of their the debt rescheduling and other poli- technical, managerial and financial cies -- including divestiture and capabilities. The Agent Bank examines fiscal austerity -- supported by the and homogenizes the bids, i.e. puts PERL and other Bank loans, the Mexican the bids on the same basis. The plane economy has stabilized to a large of the bidders for the company in extent. Inflation is now down to terms of investment programs and poli- levels around 20 per cent. GDP growth cies towards labor are taken into is now roughly at the 4 per cent lev- account, but the primary choice crite- el, the financial deficit of the Gov- non at this stage is the highest eranent has been reduced from around offered price. The relative clarity 16 per cent to just over 1 per cent of of the choice criterion is a key fac- GDP, real domestic interest rates have tor in assuring transparency and fair- fallen dramatically and private in- ness in the sale process. vestments are up. - vii - 9. An important feature of the Mexi- great success. Probably its greateet can privatization program has been the achievement, and the reason why it is absence of any serious opposition from widely admired, is its magnitude. The labor. Indeed this lack of opposition number of state enterprises have been has been an important necessary condi- cut from 1,155 to around 200 in nine tion to allow the privatization pro- years. Besides this, the privatiza- gram to proceed smoothly and quickly. tion program has generai.ed very large In fact, labor's interests have been revenues for the Government, thereby protected throughout the program. fulfilling a key role in the economic This protection has taken three forms: stabilization program. The sale pro- First, the unions have, by law, a ceeds from PE sales in the two years right of first refusal at each sale. 1990-91 were around $14 billion, which ,ie second form ofis three to five times various esti- labor has been the understanding with mates of the benefits to Mexico from all buyers that worker layoffs are not the debt rescheduling. Thus, in terms to occur, although labor retrenchment of short-term revenue impact, it could did occur in the steel sector before be argued that the privatization pro- privatization. Third, there have been gram has been even more important than instances where employees have been the Brady deal. sold shares in the company in order to encourage their co-operation. For Issues In Poitatdon example, the Telmex union was sold 4.4 per cent of the company for $325 mil- 12. Despite its successes, the Mexican lion; the money for the purchase was divestiture program has rot been with- loaned to the union by a government out some problems. One relates to the bank. regulation of non-competitive indus- tries. The Government has experienced 10. As part of its policy of liberal- some problems in terms of setting up a izing and opening up the economy to regulatory environment. With non- international trade, the Mexican Gov- tradeable goods, a regulatory struc- ernment liberalized rules for foreign ture is essential if there iv any ownership of Mexican companies. Nev- monopoly power in the market. Two ertheless, according to a document of cases have been prominent in Mexico-- the Finance Ministry, 98 per cent of telecommunications and airlines. In 156 enterprises sold from January 1987 the case of telecommunications, a to August 1991 were purchased by Mexi- sophisticated and carefully drafted can investors. Indeed, only 3 out- regulatory structure was put in place right sales to foreign companies took prior to the divestiture of Telmex, place, and they were all relatively and while this mechanism does provide small (this did not include the sale some structure to the market, offi- in January 1992 of the SICARTSA Steel cials in the Mexican communications complex to Caribbean Ispat Ltd., an secretariat are quick to concede that Indian-controlled company based in their regulatory capability at this Trinidad). But there were several time is rather limited. Attempts to other important sales in which for- remedy this sitiiatton are under way. eigners played a minority, but impor- In the case of the airline industry, tant, role. no regulatory changes were made for several years after divestiture, caus- The Dvestiture Progranm: Successes and Failures ing severe disruption of (or at least uncertainty in) the strategic planning 11. There can be little doubt that the of the firms. Deregulation, with Mexican divestiture program has been a clear rules, has now been implemented. - viii - 13. A second problem relates to the financial efficiency. Although prices Government's policy of selling enter- were liberalized in principle, the prises to well-identified buyers rath- implementation of the price adjustment er than to diffuse buyers in the component had a somewhat mixed outcome stock-market. While this style of (para. 27). Electricity prices were sale might be thought desirable as adjusted as envisioned. Steel and giving the buyer all the leeway he fertilizer prices were held constant needs, in particular in order for him as part of the PACTO's anti-inflation to concentrate on maximizing long-term program; they were finally increased ahareholder wealth without worrying as a prelude to privatization. Rail about corporate raiders or other freight prices were adjusted in 1990, short-term coasiderations, it weakens but the railway company'a finances an important characteristic of the remained problematic as it was unable market economy, i.e. the discipline of even to cover its operating cost for the stock market. A third question the first time that year. about the di%estiture program is spe- cial to the case of Telmex. It could 16. The PERL called for a series of be argued that, because a larger pro- other measures to improve the competi- portion of the stock was sold to for- tive environment of PEs, notably the eign investors, too much rent from the injection of private elements in mar- sale dissipated abroad. kets dominated by large PEs. Here, the performance has been satisfactory 14. The problems mentioned, however, on the whole. An important area of are all minor compared to the magni- improvement that was not specifically tude of the successes achieved by the envisioned in the PERL was the privat- divestiture program. A series of ization of the steel and fertilizer factors contributed to the success of sectors. Here the Mexican Government Mexico's divestiture program: (i) the moved further and faster than had been key ingredient was commitment to the anticipated. The steel sector has program at the highest levels of the been privatized and the privatization Mexican Government; (ii) the process of FERTIMEX is proceeding rapidly. of divestiture is clearly defined and transparent. No deviations are made Measures to Foster Manaerial Autonomy and Ac- from the announced set of rules; (iii) countabit administrative control of enterprises in the process of divestiture was 17. The centerpiece of the public generally transferred from the con- enterprise reform component of the trolling secretariat to the privatiza- PERL was a series of measures to im- tion unit at the Finance Secretariat prove PE performance by granting PE preventing interference from sector managers greater autonomy while also ministries with vested interests. strengthening accountability so as to prevent any misuse of the newly ac- Public Enterprise Reform quired autonomy. The program called for a shifting of decision-making Price Reforms and Other Improvements in the Com- responsibility away from the minis- petitive Environment of tries (including the Sector Ministry, Programming and Budget, Finance and 15. The PERL called for price increas- the Contraloria) to the Boards of es for large public sector companies Directors and enterprise management. in order to reduce the budget deficit What was envisioned in the PEEL were: by reducing subsidies to public enter- (a) issuance of regulations to the new prises, and to promote economic and PE law, giving effect to these chang- - ix - es; (b) adoption of measures to through bonus payments tied to the strengthen and enhance the effective- degree to which the objectives are ness of the Boards of Directors; and achieved. Under the PERL, PAs were to (c) implementation of performance be concluded and implemented for 1990 agreements with five major PEs: with five companies: FERTIMEX, FERTIMEX, FERRONALES and three compa- FERRONALES, AHMSA, CMC and SICARTSA. nies within SIDERMEX--the steel hold- This was in fact done. ing company--namely, AHMSA, CMC and SICARTSA. 20. Although some valuable experience has been gained in the process of 18. After a delay of over three years, developing these five PAs, and a cer- the regulations to the PE law were tai level of commitment to and inter- issued in January 1990. Although this est in them has grown in the Mexican was an important legal step forward in Government, the overall performance of improving PE efficiency, major imple- this component of the PERL was disap- mentation problems have prevented the pointing. There were three important law from having its full desired bene- factors that contributed to the disap- ficial impact. Tight Government con- pointing performance: (i) lack of trols over PE budgets remained even adequate consultation between Govern- after the regulations to the PE Law ment and the enterprises; (ii) the were promulgated circumscribing auton- attempt on the part of the Government omy. It was envisioned in the FLPE to use the PAs as an instrument of regulations that PE Boards of Direc- control; (iii) poor design of the PAs tors (BODs) would take over many of themselves. The lack of consultation the functions previously exercised by seems to characterize the process by the Central Government. Because the which PAs are being drafted even now. BODe were typically composed of repre- The perception of the PEs is that PA9 sentatives of various ministries, an are one more way for the Government to essential step to assure their inde- control them. Surprisingly, this is pendence was to include more non-af- the view not only in the enterprises filiated individuals, particularly but also in the controlling agency, business people from the private sec- SPP. tor. This does not seem to have hap- pened; BODs continue to have essen- 21. The technical assistance program, tially the same composition as before. on the whole, was not adequately uti- The continued representation of the lized. Only some $300,000 of the heads of sector ministries on the BODs total of $1 million in TA allocation appears to detract from the envisioned was utilized mainly in the preparation role of the BODs acting as a buffer of the PA blue print, purchase of a between the Ministry and the PE. computer and software, study of FERRON~ALES tariffs, and some training. 19. The principal tool to improve Given that the Government is committed managerial autonomy and accountability to the adoption of the PA instrument in PEs was to be the performance to improve the performance of the PE agreement (PA). In principle, once sector, it is important that the defi- agreement is reached in advance to a ciencies in the present PA framework well-defined objective or set of ob- are corrected. The UCD plans to in- jectives, the enterprise can be grant- troduce PAs for several firms and it ed autonomy. Accountability is as- would be useful if appropriate adviso- sured by an ex-post evaluation of ry inputs were provided to the UCD to performance relative to the objec- both design the PAs and to negotiate tives, and incentives are provided their atreement with the resective PEs. The remaining TA funds could be bring together potential foreign part- utilized for this purpose. ners with domestic investors. Lessons Learnd For tradeable goods, trade liber- alization and deregulation are useful 22. Some of the key lessons are as precursors to divestiture. follows: *For non-tradeable goods in non- * Lending to support debt reduction competitive industries, regulatory can indeed be extremely effective in mechanisms should be designed and helping achieve macroeconomic stabili- implemented prior to divestiture. zation and growth. * For enterprises whose value is * An important ingredient for suc- expected to rise sharply over time, or cess is commitment on the part of the where there do not exist adequate Government; in particular, supporting capital markets to value shares, a a committed, capable Government carry system of selling in tranches can be through its own sound policy program an effective way for Government to is highly worthwhile. Attempts to receive a share of subsequent gains. alter the policy agenda in such situa- tions are likely to be unsuccessful. 0 Labor's cooperation and acceptance can be assured through protection of * Debt reduction and rescheduling by jobs or other ways for labor to par- itself may not be sufficient to bring ticipate in the gains from divesti- about investor confidence. Rather, a ture. comprehensive program of reform and liberalization, combined with large- For companies not viable as sale scale divestiture, may increase the prospects, liquidation can and should chances of success. be used as an alternative to reduce 23. From Mexico's successful divesti- ture program, the following lessons can be drawn: 24. As regards PE reforms, the follow- ingnessns anreadawn * Consolidation of divestiture ac- tivity in a central unit, close to a * Care must be taken to ensure con- power center, which then supervises uistency between different policy management of the sales by qualified elements. Price liberalization was agent banks, can be an effective way not always compatible with the anti- of organizing sales. inflation goals of the stabilization * A clear, transparent, well-defined program. process with clearly specified rules The delineation of authority and and sequences, can ensure orderly responsibilities between the various sales. The scope for negotiations controlling agencies and PEs is a between buyers and the Government complex task. Boards of PEs cannot should be minimized. act as a good buffer if filled by government officials. Until the comn- * Government can serve as an effec- position of the Boards of Directors is tive conduit or clearing-house to changed, little substantive chang in their role can take place. - xi - Complex policy instruments such as employed by the Government. The se- the performance agreements need to be lection of PA criteria has to be con- designed by individuals with the req- ceived in a dynamic setting if PEs are uisite experience in the area. Imple- undergoing restructuring. mentation should be by a specialized unit within the Government. Care has Lack of continuity in loan super- to be taken that the PAs are consis- vision can compromise its effective tent with other instruments of control implementation. PROGRAM PERFORMANCE AUDIT REPORT MEXICO PUBLIC ENTERPRISE REFORM LOAN (LOAN 3086-ME) I. Background Macroeconomle Deveopments Precedent to the PERL 1. Between 1950 and 1974, the Mexican economy enjoyed a prolonged period of strong economic growth with GDP growing at 6.5 percent per year. Conservative fiscal policies kept inflation at single digits until 1972, but subsequently, large increases in public expenditures pushed inflation to double digits and led to the collapse of the Peso exchange rate in 1976. During the next six years fiscal accounts continued to deteriorate. By 1981, the primary deficit reached 8 percent of GDP, compared to 0.4 percent a decade earlier, and inflation reached almost 30 percent. The exchange rate became highly overvalued; imports increased fourfold from 1975 to 1981 and capital flight started. Foreign debt jumped from US$7 billion in 1974 to US$78 million in 1981. 2. In 1982, rising world interest rates and falling oil prices gave the final blow to the unsustainable expansionary policies of the seventies. Foreign creditors refused to roll over Mexico's short-term debt; foreign reserves were exhausted forcing a float of the peso and the temporary suspension of foreign debt service. Inflation skyrocketed to 60 percent, GDP fell by 0.6 percent and the peso suffered a real depreciation of almost 30 percent. In 1983, under a stringent IMF-supported stabilization program,. Mexico was able to turn, in a single year, the primary fiscal balance from a 3.5 percent deficit to a 4.8 percent surplus. The real exchange was further depreciated while inflation was running at 100 percent. The result was a severe recession with GDP falling by more than 4 percent. But the Government was also able to achieve a remarkable improvement in the noninterest current account, from a US$6.3 billion deficit in 1981 to a US$15.2 billion surplus in 1983, and the rebuilding of reserves. The Government persevered with these policies through mid-1985, and the economy responded with GDP growing by about 3 percent and inflation falling to 60 percent. 3. The fragile recovery in 1984 and the first half of 1985 succumbed to two major external shocks. First, the Mexico City earthquake of September 1985 caused major disruptions with damages estimated at 2 percent of GDP. Second, the collapse of oil prices in 1986 led to a shortfall in Government revenues equivalent to 2 percent of GDP. The Government recognized the need for a major reorientation of its economic strategy. A new adjustment program was adopted in July 1986 aimed at integrating the Mexican economy with the rest of the world and mobilizing additional resources from foreign creditors. Exports were aggressively promoted through a 35 percent depreciation of the real exchange rate. The trade regime was liberalized and as part of its commitment to an open economy, Mexico joined the GATT. Although nonoil exports responded by growing - 2 - more than 40 percent, 1986 was not a good year: GDP fell by almost 4 percent and inflation climbed to nearly 90 percent. Economic stagnation and triple digit inflation cortinued into 1987, while the noninterest current account generated a $11.8 billion surplus. The Government's policy of quarterly wage indexation and frequent exchange rate depreciations to maintain export competitiveness introduced a high degree of instability into the system. A stock market plunge in October triggered a run on the peso resulting in a financial crisis. 4. The Government responded to the crisis with the "Economic Solidarity Pact" (Pacto), an agreement between business, labor, and government. This agreement called for accelerated structural reform, further tightening of fiscal and monetary policy, a freeze of minimum wages and of basic public and private sector prices. The cornerstone of the "Pacto," was a freeze of the nominal exchange rate against the US dollar, a heterodox policy response to control inflation. This partial freeze was extended at three-month intervals through the end of 1988, and renewed, with some modifications, by the Salinas Administration under the name of "PECE" (Pact for Stabilization and Growth). The main change introduced under the PECE was a daily adjustment of the exchange rate of one peso against the US dollar. This rate of increase has been slowed down to 0.8 peso per day in mid-1990, to 0.4 peso per day in January 1991, and to only 0.2 peso per day in January 1992. 5. The policy measures had a dramatic success in reducing the rate of inflation, from 159 percent in 1987 to 20 percent in 1989. The key elements contributing to the success of the reform program were fiscal austerity, the opening up of the economy, and a changing role for the Government. This is evidenced in sizeable capital inflows and a sharp reduction in real interest rates. Fiscal adjustment has been particularly dramatic, with the primary surplus since 1983 registering an improvement equivalent to some 16 percentage points of GDP, although the brunt of the adjustment has been on cuts in expenditures. In 1981, noninterest expenditures were almost 35 percent of GDP, but by 1983, expenditures were cut to 28 percent and then gradually reduced to 22 percent by 1989. Capital expenditures were also cut in half. While the cuts brought about increased efficiency of Government operations by eliminating most of the dubious large projects of the late 1970s, expenditures on maintenance and social sectors were also reduced sharply. Total interest payments have represented an important expenditure component. From 1982 to 1989 interest payments ranged from 12 to 20 percent of GDP. While foreign interest payments have been relatively stable at about 4 percent of GDP, real domestic interest payments have fluctuated sharply. After dropping from about 4 percent of GDP in 1983-86 to zero in 1987 they rose to 7.3 percent in 1988-89 when the credibility of the stabilization program was in question. The ltocess of Structual Refom 6. In spite of the uneven macroeconomic performance, Mexico has been able to transform itself into one of the most open economies in the world through extensive trade reforms. Trade liberalization, most of which took place between 1985 and 1988, and was supported by the Bank under TPL-I and II, lowered the percentage of domestic (nonoil) tradeable production covered by import quotas from 100 percent in 1984 to about 20 percent by 1989. Maximum import tariffs were cut by similar magnitudes. Monoil merchandise exports, which represented -3- less than one-third of total exports in 1984, doubled their share by 1989. A major tax reform was initiated in 1987 which expanded the tax base, simplified the system, reduced tax rates and modernized tax administration. 7. Reform of Mexico's inefficient Public Enterprise sector (PE) was another focus of the structural adjustment program. In the early 1980s, the Government started to take steps to trim the public enterprise sector which had expanded to over 1,214 enterprises, employing one million people and accounting for some 12% of GDP. In 1982, the Government initiated a divestiture program by selling small and medium enterprises. By early 1989 this program reduced the number of PEs to some 450 enterprises; however, the divested PEs accounted for less than a quarter of the total asset values of the original universe of PEs. In parallel with the divestiture program, the Government initiated measures starting from mid 1985 to improve the operational efficiency of PEs and reduce financial losses. Subsidies from the Government to PEs amounted to 2.82 and 2.1% of GDP in 1987 and 1988 respectively. A significant part of PES financial problems arose from price controls (e.g. fertilizers) but a great deal of operational inefficiency was due to gross interference by Government in all aspects of PE operation as well as inadequate management autonomy and accountability. In 1986, the Government passed a new PE law to deal with these issues but the law remained inoperative because the detailed regulations supporting the law was yet to be issued. Both the Bank and the Government recognized that PE reform remained a weak link in the chain of structural adjustment measures and needed urgent attention. The PERL was specifically designed to accelerate and deepen the process of PE reform and inter alia help to reduce the drain on the budget (see Chapter II for details). 8. Reform efforts were particularly strengthened in early 1989 following the election of President Carlos Salinas de Gortari in December 1988. The Government took further actions to liberalize the regulatory framework which was supported by the three adjustment loans approved in May 1989. Foreign investment regulations were considerably liberalized and made more transparent. The tax system was modified by bringing marginal tax rates more in line with levels in major industrial countries, encouraging the repatriation of flight capital, and increasing the sanctions for tax evasion. The impact of inflation on the (corporate) tax system was attenuated by eliminating purely inflationary gains from the tax base. Finally, a variety of measures were introduced to increase the number of people actually paying taxes. Measures were introduced to liberalize financial markets. Ceilings on commercial banks' deposit interest rates were abolished; forced allocation of commercial credit towards favored sectors was abolished and credit subsidies through official development banks were reduced. Rationale for Debt Restructurin 9. Despite the far-reaching reforms implamented, Mexico was unable to raise the necessary resources from international c3pital markets to maintain an adequate level of investment financing and thus break out of the low growth rate trap it had fallen into during the 1980s. Continuing high external transfers generated uncertainty about whether the rapidly growing transfer burden could be met. This, in turn, generated increased uncertainty about future exchange rates, taxation, and financial regulation. Thus, to forestall further capital flight, Mexico had to pay unsustainable interest rates on its domestic debt. "Ex post" real interest rates were almost 50 percent in the months before the debt accord was reached. 10. Real interest rates so far above the real growth rate of the economy are detrimental under any circumstance; however, in Mexico there was an additional complication in that the Government was in the middle of a stringent economic stabilization program in which fiscal retrenchment played a central role. High real interest rates were mostly due to credibility problems with the exchange rate component of the "Pacto." Local currency rates were around 30 percentage points above the rates on dollar denominated Government debt, while the exchange rate was fixed against the dollar during 1988. The uncertainty caused by potential transfer problems, because of the perceived unsustainability of the exchange rate regime dramatically pressured upwards domestic real interest rates and thus became a direct threat to the survival of the stabilization program. 11. The Government recognized that in order to reduce the uncertainty arising from the high level of external debt service, some measure of debt relief from external creditors was essential.Y This debt accommodation, however, had to be contracted for a medium-term time horizon in order to reduce uncertainties and improve expectations about future policies on a sustained basis. With a comprehensive solution, the impact of the reduced uncertainties was expected to go much beyond the direct fiscal effects of the lower domestic interest rates. Private investment, particularly foreign direct investment, was expected to respond strongly to the reduced uncertainty as well as to lower real interest rates. 12. The boosting of private investment was thus seen as essential to the resumption of economic growth given that the thrust of the Government's structural reforms since 1985 has been to reduce the role of the public sector. For this purpose, sustained access to international capital markets was essential. Private lenders were, however, holding back because of the perceived lack of credibility of the Government's program and its implications for fiscal and exchange rate policy. In the light of these concerns, the Government and the Bank came to a recognition that a policy financing package had to be developed to meet Mexico's projected resource needs for supporting a higher level of private investment. This policy financing package would need to address the issue of the very high levels of external transfers in order to boost Mexico's creditworthiness. The Bank viewed its proposed lending operations as sending a strong signal to Mexico's external creditors about its confidence in Mexico's reform program and its support for enhanced resource flows to renew economic growth. 13. Both the Bank and the IMF publicly supported Mexico's claim that a reduction in Mexico's public debt was essential if growth was to recover in Mexico. Even before the start of negotiations with its commercial creditors, 1 This was reflected in President Salinas' inaugural address on December 1, 1988 when he asked Mexico's external creditors to contribute to a medium-term debt resolution program. Mexico started negotiations with the IMF and the Bank about major packages to suppor# debt reduction. In February 1989, the IMF provided an Extended Fund Facility for SDR 2.9 million (US$4.1 billion) covering three years and an optional fourth. At the same time, the Bank concluded negotiations on three adjustment loans for US$0.5 billion each and agreed on a lending program of about US$2 billion per year for the period 1990-92. Moreover, both institutions agreed to allow a part of the resources to be used to support debt reductions (Set Asides). Terms of the Debt Restmcturine Azreement 14. The Mexican Government and a Bank Advisory Committee representing some fifteen commercial banks started negotiations on a debt restructuring package in April 1989. These were concluded with the issuance of a term sheet outlining an agreement in principle on July 23, 1989. This agreement which included a 35 percent debt relief option was negotiated and refined further. On September 15, 1989, the Government of Mexico and the Bank Advisory Committee representing the commercial bank creditors reached agreement on a financing package covering the period 1989-92, restructuring approximately US$48.9 billion of Mexico's external debt. The agreement, implemented over the next eight months, offered three choices to creditor banks holding altogether US$48.9 billion debt. First, old debt could be exchanged for new debt carrying LIBOR plus 13/16, like the old debt, but with the principal reduced by 35 percent. Second, old debt could be exchanged for new debt instruments at par, but these instruments would carry a fixed interest rate of 6.25 percent only. Both instruments would be amortized fully in a single bullet payment in 30 years. Their principal would be fully collateralized; 18 months worth of interest payments would also be covered by an escrow account established at the time of the exchange. Third, old debt could be exchanged for new debt at par and market interest rates, but banks choosing this option would have to make additional loans ("new money") over the next three years equal to 25 percent of the amount brought into this option. But, banks opting for this "new money" would not receive any collateral in support of principal or interest payments. Finally, the two debt service reduction instruments would carry a recapture clause: if oil prices rose above US$14 dollar per barrel in real (1989) terms, 30 percent of the extra revenue from crude oil exports would be available for the creditors under this clause, but with two additional restrictions: (a) If less than 100 percent of the creditors chose debt reduction options, the amount available under this clause would be scaled back proportionally; (b) No creditor could receive more than 3 percent of the amount of old debt brought into the debt service reduction options in any given year under this clause. TABLE 1: SUMIARY OF INSTRUMENTS OF THE MEXICO PLAN INSTRUMENT DISTRIBUTION PRINCIPAL INTEREST RATE COLLATERAL GRACE LENGTH OF RECAPTURE BEARER OF OF DISCOUNT FOR PERIOD/ INTEREST CLAUSE REGISTERED INSTRUMENT PRINCIPAL MATURITY ESCROW BONDS ACCOUNT Discount 43Z 352 LIBOR + 13/16 30 years 30 years 18 months Yes ** Registered Bonds U.S. bullet rolling Treasury maturity Zero Coupon Bonds. Par Bonds 471 6.252 fixed 30 years 30 years 18 months Yes ** Registered U.S. bullet rolling Treasury maturity Zero Coupon Bonds. New Money 101 $1.09 bn LIBOR + 13/16 None 7 years/ None No Registered 15 Years . Note:- Funding: Japanese Cofinancing ($2.05 bn), IBD ($2.01 bn), IN ($1.697 bn), Mexican Reserves ($1.243 bn). * These numbers may be subject to marginal changes. * Both bonds Include a recapture clause which maintains that from July 1, 1996, bondholders will be entitled to payments to up to 32 of debt exchanged should the annual average real price exceed $14 per barrel. -7- 15. The structure of the final debt package that was agreed with all foreign and Mexican creditors is shown in Table 1. Par or iscount bonds were the preferred options for most banks. Out of the US$45.8 billion held by non- Mexican creditors, most was brought into one of the debt reduction options, with only about US$4 billion brought into the new money option. Thus, new money commitments from non-Mexican creditors came in at slightly below US$1 billion (25 percent of the amount brought into this option). US$22.8 billion was brought into the par bond option and US$18.8 billion into the discount bond option.11 Since 67 percent of the Mexican creditors (who were holding US$3.1 billion) chose new money, the overall new money option choice was somewhat higher. The remaining 33 percent of the Mexican creditors opted for the discount bond. Thus, with all creditors combined, US$22.8 billion vent into par bonds and US$19.8 billion into the discount bond option (US$1.0 billion of which was brought in by Mexican creditors). 16. The collateralization of principal and eighteen-month interest coverage required an estimated US$7.25 billion in enhancement funds which was to be provided by the IMF, World Bank, Japan and the Mexican Government. The World Bank's contribution was estimated at US$2.01 billion of which US$750 million was, to be provided as set asides from all adjustment loans to be made to Mexico in FY89-91. The balance of US$1.26 billion was to be provided under a separate Interest Support Loan i. FY91. 17. The PERL approved in May 1989 was part of a cluster of three adjustment loans (the others were a Financial Sector Adjustment Loan and an Industrial Sector Adjustment Loan), totalling US$1.5 billion. US$125 million from each of these three loans were set aside for use by the Government in financing collateral support for the discount bonds to be issued under the debt reduction plan. The balance of US$375 million was provided by amending the loan agreements for the Fertilizer Sector Loan (US$150 million), Agricultural Sector Loan (US$175 million) and the Steel Sector Adjustment Loan (US$50 million). The timing and size of all of the three adjustment loans approved in May 1989 were thus designed, primarily, to support the Mexican debt restructuring program. They were brought to the Board in May 1989, just before the Government issued its initial term sheet in July 23, 1989 to its commercial creditors. a There are some residual claims related to old loans that carried partial World Bank guarantees. These received slightly different treatment (ef. van Wijnbergen (1990) for details]. -9- II. The PERL In the Context of PE Reform in Mexico 18. An ex-post assessment of the design and objectives of the PERL can only be undertaken properly within the framework of the overall program of PE reform that has been undertaken by the Government since the early 1980s. A brief historical review of PE reforms in Mexico is thus necessary.!' PE reform was first initiated by the de la Madrid administration in the context of 1983-88 National Development Plan (NDP). The NDP, while recognizing the continuing need for the existence of PEs, outlined a new approach for rationalizing the PE sector. This new approach called for the Government to restrict PE investment to selected activities and to improve the efficiency of public enterprises by granting management greater autonomy and incentives to improve performance. In order to implement the new approach, the Government needed to make two important legal changes. First, amendment to the Mexican constitution (articles 25, 26, 28); these amendments undertaken in 1983, reserved certain sectors exclusively for the state (strategic sectors) while allowing joint participation with the private or social sectors in "priority" sectors. Since the latter category was not explicitly defined, it allowed the Government flexibility to reclassify sectors as priority sectors depending on its interest in entering or withdrawing from the sector. These amendments laid the legal basis for the PE diveciiture program which was initiated after the constitutional amendments were passed. Second, a new Federal Law of Public Enterprises (FLPE) was enacted (1986) which redefined the relative roles and authorities of the various cont -olling agencies with the view to providing greater management and financial autonomy to enterprise management while making them more accountable for performance. The main thrust of the FLPE was to promote a change in the Government's relationship with the PEs from one dominated by ex-ante controls to ex-post evaluation. However, the detailed regulations which were needed to implement the FLPE were not prepared and issued until 1990. This adversely affected its initial implementation. 19. In addition to these legal reforms, the Government took a number of actions to streamline the PE sector and to improve its efficiency, particularly in order to reduce the burden on public finances. These actions can be grouped under the following three heads: () Dyestiur 20. With the constitutional mandate, in 1983 the Government started a program of rationalization of the PE Sector through transfers, liquidation, amalgamation and sale of PEs. By 1988, some 706 mostly small and medium size PES had been divested!' (see PR. paras. 28 and 50 and PCR para. 1.15) but with the 1t A more comprehensive review of PE reform since 1983 is provided in "Public Enterprise Reform in Mexico", Working Paper of the Public Sector Management Division, Technical Department, LAC, 1990. 1 In the industry, mining and energy sectors, the number of PEs was reduced from 398 to 90 in this period. - 10 - advent of the Salinas administration, the Government expanded the program to cover larger enterprises in manufacturing, mining and public utilities. Reflecting this shift, the Government took steps to strengthen the PE sales unit in the Ministry of Finance to enable it to deal with the complex regulatory and valuation issues involved in preparing large PEG for divestiture. At the same time, the entire privatization process was streamlined, and made transparent through issue of guidelines which inter alia clarified the role of various actors in the PE sales process. (II) Enterpdse EfRelenea Improvement Actions 21. Actions taken in this area include: (a) financial restructuring of PEs through debt absorption by the Federal Government. These restructuring agreements (Asuncion de Pasivos y Cambio Estructural or CFR) specify the Government's obligations, mainly financial, to the PE and in turn the obligations of the firm (in terms of financial performance) vis-a-vis the Government. These agreements have led to the Government absorbing a substantial portion of PE long term debt up front, against an undertaking by the PE to improve future operating and financial performance. Until mid-1989, 12 CFRs have been concluded at a total direct cost to the Government of $14.15 billion. Preliminary results indicate that these agreements Lave set targets for improvements too far into the future or too vaguely and thus have not been particularly effective. (b) Deficit-Superavit Agreements (DSA) (Acuerdos de Deficit Superavit) were developed as yet another tool to control PE expenditures and ensure that PE budgets were maintained in conformity with state finances. DSAs are verified every quarter but PEs are required to submit monthly reports to the Finance Ministry. While DSAs have proved very effective in ensuring that PEs adhere to the agreed deficit or surplus targets in periods of high inflation, they represent an extremely tight system of control over PE finances and have constrained PEs to respond to the changing economic situation. (ill) Competidon Enhandar Measures 22. As mentioned earlier (paras. 6-8), since 1985, the Government has taken a variety of measures to enhance competition in domestic markets, particularly through trade reforms. In addition, the Government lifted price controls on a large number of products and even those controlled were close to border prices. However, there still remained in 1988, many areas with large distortions (fertilizer, power and rail tariff, agricultural products) which resulted in significant financial subsidies to PEa. PERL Desin and Oblectives 23. As discussed earlier, the PERL as well as the two other adjustment loans approved as the same date were basically granted in support of restructuring Mexico's commercial debt and helping the economy move out of the low growth trap it had fallen into during the 1980s. The Mexican Government viewed the reform of the PE sector as the centerpiece of its structural adjustment effort since this was not only critical to developing the private sector, but was also extremely important for reducing the fiscal deficit as part of its stabilization policy. The Bank fully shared the Government's aims with - 11 - regard to PE reform. Past Bank loans to individual PEs, however, had failed to address adequately cross-sectoral issues pertaining to the competitive environment for PEs, the role of Government agencies and greater managerial and financial autonomy and accountability. The PERL was thus designed to support in general the Government structural adjustment effort in the area of public enterprise reform but with specific focus on cross-sectoral issues that were not properly addressed under previous loans to individual PEs. The PERL had two components: a component to support the Government's ongoing divestiture program, and the second aimed at inducing PEs to operate on a commercial basis in a more competitive environment with sufficient managerial and financial autonomy to enable them to improve performance but balanced by adequate accountability to provide appropriate rewards (sanctions) to managers. 24. The Bank's contribution to the design of the divestiture component, as described in the President Report (PR) and corroborated in discussions with the Government appears to have been in the nature of advisory and moral support!' per se. Prior to the PERL, the Government took steps to strengthen the PE Sales Unit at SHCP to handle the more complex transactions that were expected to be undertaken in the future. At the Bank's request, the Government prepared an cecelerated program of PE divestiture that called for the sale of 64 PEs, liquidation of 124 PEs and merger or transfer to local governments of 12 PEa, totalling 199 PE9. Significant progress in implementing this program was made a condition of second tranche release. The Government, also clarified its policy of establishing PEs in non-strategic priority areas by confirming that it would establish PEs for the production of goods/services only in cases of significant market failures and where major social welfare gains were involved. Moreover, the Government agreed to tighten the rules for the establishment of new PEs by limiting the authority to the Commission Intersecreterial Gasto-Financiamiento (CGF), also a condition of second tranche release. 25. The PE reform component of the PERL had a more substantive agenda than the divestiture component (paras. 57-85 in the PR). Here the major focus was on increasing managerial and financial autonomy and accountability to spur improvement in PE performance. The instrument chosen for this purpose Was performance agreements (para. 69, PR) which would allow Government supervision of PEs to be converted from ex-ante controls to ex-post monitoring of results. The system would be developed and maintained by a new unit--Unidad de Convenio de Desempeno (UCD)--attached to the CGF. The adoption of this new system required a major revision in the roles of the various government agencies involved in administering PEs, as well as in the authority and responsibilities of the UCD Board of Directors and PE management (para. 66 of PR). These roles and relationships needed to be spelled out in the regulations of the Public Enterprise law of 1986 which had not yet been issued. Issuance of satisfactory regulations to the PE law granting expanded powers to the Board of Directors and management of PEs as discussed in paras. 71-73 of the President's Report was a condition of Second Tranche Release. 26. The PERL also addressed the issue of availability of accurate and 1i This view was confirmed in discussions with officials in the PE Sales Unit of the Ministry of Finance. - 12 - timely information on PEs to support the new performance agreement system by proposing the strengthening of the centralized information system (Sistema Integral de Informacion) of the SPP. 'The PERL called for the implementation of the performance agreements initially for five large enterprises--FERTIMEX, CMC, FERRONALES, SICARTSA and AHMSA. To ensure that the system was designed and implemented properly, the loan provided for technical assistance funding for advisory assistance and training for the staff of the UCD and of managers and board members of PEs to enable them to utilize their new autonomy more effectively. Finally to support the objective of greater financial autonomy for PE managers, the SPP agreed to develop and implement simplified, more flexible budgeting procedures (para. 77 of PR). 27. Another important initiative adopted by the Government under the PERL waF to improve the pricing regime for PE products. Pricing issues in the steel and fertilizer sectors were first addressed under the Steel and Fertilizer Sector Adjustment Loans. While significant liberalization took place, prices of a number of PE products remained strictly controlled e.g. fertilizers, steel, electricity contributing to large fiscal deficits. During negotiations, the Government committed itself to a price reform plan for products of five PES responsible for the largest deficits (CFE, CONASUPO, FERRONALES, FERTIMEX, SICARTSA) as well as to general pricing policy (Letter of Development Policy, para. 8). It also agreed to limit operational and investment transfers to PEs to that specified in the FY89 budget or a real reduction of 13% compared to FY88 (para. 46, PR). 28. The PERL also included a number of miscellaneous conditionalities that were packaged under a Plan of Action whose satisfactory implementation was also a condition of second tranche release. These included: improvements in PE budgeting procedures; studies on railway tariffs and part privatization of secondary activities of FERRONALES (paras. 59 and 61 of PR); measures to improve PE sector planning and corporate programs through guidelines in the NDP; improved guidelines for PE project evaluation and for monitoring project implementation and disbursement performance (para. 80 of PR); and the establishment of a project appraisal unit for PE investments. - 13 - Ill. Implementation Experience and Results Achieved A. MacroeconoMl Performance 29. As the primary goal of the Bank's adjustment lending to Mexico including that of the PERL was to support the Government's macroeconomic stabilization effort, the key test of its success is the macroeconomic results achieved. On the whole, it is safe to conclude that Mexico's macroeconomic strategy has been highly successful. Inflation has been brought down from levels above 100% to below 20%. GDP growth has picked up from being negative to an average level of close to 4 per cent. The financial deficit of the Government was reduced from around 16 per cent to just over 1 per cent of GDP. Transfers to public sector enterprises have fallen from 4.5 per cent of GDP to under 4 per cent. Domestic interest rates have fallen dramatically. Details are in Table 2. Table 2: MEXICO - KEY MACROECONOMIC INDICATORS 1986 1987 1988 1989 1990 1991 Inflationa 105.7 159.2 51.7 19.7 29.9 19.9 Real GDP growth (3.8) 1.5 1.7 3.2 4.4 3.6 Primary Surplus2 2.2 5.0 6.4 8.3 7.9 6.0 Financial deficits 16.0 16.1 12.3 5.7 3.5 1.3 Transfere 4.5 4.3 4.1 4.1 3.8 2.5 Interest Rate4 80.9 94.6 67.6 44.6 37.1 21.8s * All 1991 numbers are CAIE estimates, using data through August or September. December-to-December change in consumer prices. 2 Government revenues minus government expenditures (excluding interest payments), as a percentage of GDP. 4 Interest payments minus the primary surplus, as a percentage of GDP. Average cost of funds, Banco de Mexico. s As of June 30, 1991. 30. Mexico's economic strategy during President Salinas' administration has been based on a continuation and deepening of the economic policies initiated in the early 1980s while simultaneously introducing new policy reform elements. The major planks of the reform program have included tax reform, trade liberalization and divestiture/privatization of PEs. Financial sector reforms including liberalization and deregulation were implemented after the Government had a better handle on public sector deficits which allowed reduced reliance on inflationary financing. The implementation of a major debt reduction exercise (the Brady deal) was the key instrument to reduce net external outflows and restore foreign and local investor confidence in the Mexican economy. 31. Given Mexico's outstanding performance, the questions are: what role, if any, did the PERL play in this? And finally, what problems remain for the - 14 - future? These questions may be answered by considering, (i) the role of the Brady plan, (ii) the disengagement program, and (iii) sustainability of the macroeconomy. The Role of the "Brad" Debt Reschedular 32. The details of the Brady operation are discussed above (paras. 14-16). The immediate effect of the debt rescheduling was a reduction in Mexico's annual interest payments of about $1.3 billion per year, which is equal to nearly 20 per cent of the interest flow prior to the debt buy-down. Total transfers abroad (including amortization payments) were reduced by some $4 billion per year. This is equivalent to about 2% of GDP. Half of this reduction was due to the lengthened maturity of the debt, a quarter due to net interest reductions and a quarter due to new money.Y 33. A key question regarding the Brady Plan is whether Mexico gained from it or not. Bulow and Rogoff (1991)ZI argue that debt buy-backs are generally not a good idea because most (if not all or maybe even more than all) of the benefits accrue to the debt-holders. Their argument is that if the "creditors believe that the buy-back will stimulate growth, then they will demand a higher price in order tL sell". Nevertheless, when they apply their model to the Mexican debt buy-down, they estimate that the welfare impact on Mexico from the buy-back is between a loss of $0.4 billion to a gain of $4.2 billion. Van Wijnbergen' has estimated that the gain to Mexico was between $2.5 billion and $3.5 billion, while the gain to creditors was between zero and $900 million. 34. There are two important points to be made here. First, Bulow and Rogoff's model is that of open-market debt buy-backs which they have then extended to negotiated buy-downs such as the Brady deal. But there is an important difference between the two processes. In open-market buy-backs, the secondary market price of the debt rises as investors realize there is an increased probability of repayment because of the expected growth. Of course if the price rises fully, the growth prospects are jeopardized. But nobody wants to sell at a lower price; there is a free-rider problem. In a negotiated buy-down, however, the Government (and its backers such as the Bank) can make an all-or-nothing offer to break the free-rider problem. And this essentially is what was accomplished in the Mexican debt restructuring. 35. The second point to be made in this context is that the Bulow and Rogoff estimate of the gains to Mexico are probably a serious underestimate. Their model assumes a fixed pool of investible funds which the market allocates between consumption and investment. Using some of this pool for a debt buy-back is 1i See Daniel F. Oks: "Stabilization and Growth Recovery in Mexico," World Bank, Policy Research Working Paper, January 1992. 2i Jeremy Bulow and Kenneth Rogoff: "Sovereign Debt Repurchases: No Cure for Overhand," Quarterly Journal of Economics, November 1991, pp. 1291-1236. 1t Sweder van Wijnbergen, "The Mexican Debt Deal," Economic Policy, April 1991, pp. 13-56. - 15 - welfare-reducing. There are two problems with this argument. Firstly, it assumes that funds used for the buy-back would alternatively have been available for investment. It is fairly certain, however, that the Bank and others would not have been lending quite so much to Mexico had it not been for the Brady Plan. For example, the Bank exceeded its guideline of a 10 per cent cap on lending to any one country, an extraordinary step that could not have been justified under normal circumstances. Secondly, and more importantly, the pool of investible funds may expand because of increased confidence in the economy or because of improved investment opportunities. The latter was central to the Mexican Government's strategy which was aimed at promoting large-scale inflows. Apart from foreign-owne6 capital, they wanted the return of flight capital that had left the country during the crises of the early and mid-1980s. By some estimates, the amount of this flight capital was of the order of $50 billion. The problem was how to encourage capital to flow to Mexico - quickly. It is here that the Mexican Government used the privatization program to great effect. Not content with simply creating conditions conducive to investment, such as the reduction of the debt overhang and liberalization of the economy, the Government offered investors an easy and profitable way to invest their money in Mexico by setting in motion a massive privatization program.. 36. There can, therefore, be little doubt that the debt rescheduling has been beneficial for Mexico .and that it played an important--some would say crucial-- part in the success of the Government's stabilization program. Further, it appears that the bulk of the gain from the rescheduling operation accrued to Mexico, and not to the creditors, as some had feared would happen. This was largely due to the negotiated nature of the agreement, with Mexico making an "all-or-nothinr" offer to break the free rider problem. The Role of the Disengagement Proqram 37. Divestiture of PEs was seen as a way to (a) reduce budgetary pressure by eliminating the fiscal drain caused by PEs, because of their losses--or investment requirements--and (b) improve overall allocative efficiency in the economy. Indeed, during the early years of the Mexican divestiture program, these were arguably the principal factors motivating the program- But by the late 1980s, it appears that an additional major objective of the divestiture program was to encourage capital inflows and, at the same time, raise substantial sums of money for the Government. 38. As can be seen from Table 4, the program was extremely successful, raising around $14 billion in just the two years 1990-91. Thus the short-run impact of the privatization program has been substantially larger than that of the debt rescheduling. Although sales were mostly to Mexicans, the Telmex sale brought 11 Not only was the idea of raising revenue not central to the arguments for divestiture, but many authors specifically felt that this could not be a major source of revenue. For example, in the context of debt-for-equity swaps (which could be substantially equivalent to divestiture combined with debt retirement), it has been said "...unless claims on future oil production are brought in, such schemes will never be big enough to have a major impact": Van Wijnbergen, op. cit., p. 25. - 16 - in $3.2 billion in foreign capital, ' and it is safe to say that much of the rest was probably returning flight capital. Although it is difficult to measure the size of capital inflows, it has been estimated by Bank staff that there were some $2 billion of new portfolio inflows in 1990 and about $9 billion in 1991. Although the Mexican stock market boom and the related liberalized rules on foreign ownership of shares have encouraged capital inflows, all of the capital inflows of 1990-91 can be implicitly accounted for by the divestiture program. 39. Thus there is little doubt that revenues from the divestiture program have played an important role in the macro-stabilization. Large-scale capital inflows have been encouraged. Further, much of the sale proceeds has been used to retire domestic debt. This has had an immediate impact on interest rates and accounts for a large part of the substantial drop in interest rates documented in Table 2. Sustainability of Economic Growth and Stability 40. The key question about the macroeconomic stabilization program is whether it has put in place forces that will make for sustainable economic growth and price stability. The divestiture program, although extremely important in the policy package, is only a one-time revenue source. There is not much discussion of privatization of some of the remaining very large PEs--the railroad, electricity and, of course, the oil company.-L" Thus the divestiture program cannot be expected to provide much more revenue than it already has. Some feel this is not necessary in any case. The debt rescheduling was seen by some as the key element, accompanied by fiscal tightening at home. However, there are some concerns. 41. An issue of some concern is the declining trend in nominal private savings which has fallen by about 10 per cent of GDP between 1983-87 and 1991.Li Based on an estimated $11-$12 billion current account deficit in 1991 compared to a $4 billion surplus in 1987, about half of the deterioration is due to higher investment and the remainder due to lower real private savings since public savings actually improved. Mexico's heterodox anti-inflation policy seems to have been the main cause of the sharp real peso appreciation that has occurred during 1988-91 which has been the major factor behind the deterioration of the non-oil trade balance. The Peso appreciation shifts investment to the non- tradeable sector and the composition of domestic demand towards foreign goods with adverse implication for the trade balance. The sustainability of the current account deficits depends on the availability of foreign capital. While foreign capital inflows overfinanced the current account deficit in 1990-91, the volatile nature of these inflows raises the vulnerability of the economy to LI In addition, Government raised close to $1.5 billion more in May 1992, through a further sale of Telmex stock. -A In fact, current plans call for the privatization unit to be disbanded by the end of the year. L See Stabilization and Growth Recovery in Mexico--Lessons and Dilemmas" by Daniel F. Oka, IBRD WPS 833, January 1992. - 17 - external shocks. The decline in private savings could thus adversely influence Mexico's continued economic growth if the increasing stock of external liabilities is not backed by adequate domestic savings to service these claims. 42. Not withstanding the above concerns there can be no doubt that the country's macroeconomic recovery has been outstanding during the period 1989-91. This experience is a strong endorsement for a policy package including debt rescheduling and fiscal austerity. It also suggests that the PERL successfully fulfilled its primary purpose. It is possible, however, that the contribution of the debt rescheduling may be overestimated. At least some of the credit for the macroeconomic stabilization and recovery can be properly attributed to the disengagement program (also supported by the PERL), which proved to be a major source of revenue during this period. To that extent, other countries that undertake debt restructuring unaccompanied by massive divestiture programs, fiscal reforms, deregulation, and trade and financial sector liberalization may not enjoy the same success as achieved by Mexico. B. The Disengagement Progam Ovenlew 43. The Mexican divestiture program began in 1983 as part of the IMF-sponsored stabilization program of that period. Following a rapid build-up during the 1970s (the Presidencies of Echeverria and Lopez Portillo), the Mexican state- owned sector consisted in 1982 of 1,155 enterprises, producing 12.6 per cent of GDP and accounting for 38 per cent of investment.-' In that year, the budget deficit was nearly 17 per cent of GDP, as rising interest rates and an oil price decline combined to push down public revenues and push up expenditures. The Mexican Government's early response to the crisis was to suspend payments on the external debt and to nationalize the banks. 44. However, in 1983, Government adopted an IMF package of policies aimed at stabilizing the economy. The stabilization program contained several elements aimed at reducing the problem of the budget deficit: (i) reductions in Government expenditures, consisting primarily of investment cuts and the moderation of wage increases; (ii) tax increases; (iii) divestitures; and (iv) reforms (i.e. increases) in public sector prices. Thus the divestiture program was part of a package aimed at reducing the budget deficit. 45. The divestiture program quickly became quite substantial in terms of the number of enterprises "disincorporated," although less so in terms of their importance. Table 3 shows the progress of the program. 13i It is difficult to get precise, comprehensive data on the Mexican PE sector. This is because some series contain data for all PEs, while others include only "directly-controlled" PEs, which leaves out many large PEs such as the telephone monopoly, Telmex, Mexicana Airline, and the commercial banks. The figures of 12.6 and 38 above include only directly-controlled PEs. - 18 - Table 3: PROGRESS OF DISINCORPORATION PROGRAM (1) (2) (3) Number of Other Number of PEs Year PEs Sold Disincorporation at end of year 1982 1,155 1983 4 77 1,074 1984 3 22 1,049 1985 32 76 941 1986 28 176 737 1987 21 99 617 1988 66 139 412 1989 37 0* 379 1990 90 9 280 1991 65 N.A. 215" Total 346 594 940 (Total reduction) Sources Column 1: Tabulated from list of PEs sold, SHCP. Column 2: Obtained as residual from Columns (1) and (3). Column 3: SHCP (Secretariat of Finance and Public Credit). * Not meaningful. The number of sales exceeds the decline in the number of PEs, indicating a data incompatibility problem. The number in Column (3) is simply deduced from Column (1), setting Column (2) to zero. 46. Note that prior to 1989, other forms of disincorporation were more significant than sale. There were four modes of disincorporation: (i) Sale to the private sector; (ii) Liquidation, i.e. sale of the assets and dissolution of the enterprise; (iii) Merger with other public sector enterprises; and (iv) Transfer of ownership to regional or municipal authorities. Of these, the last two are obviously not true divestitures, but rather removals from the Federal Government's list. Up to January 1991, approximately 125 divested enterprises fell into these two categories. Thus the most significant mode of disincorporation has been liquidation. 47. Between sale and liquidation, the latter was more prevalent in the early phases of the divestiture program. At this stage, each sector ministry administering particular state-owned enterprises was responsible for proposing enterprises for divestiture. If an enterprise belonged to a non-priority sector, - 19 - it was a candidate for divestiture. Obviously the first companies to be considered were ones that were non-viable and a drain on the budget. Thus liquidation was the most common outcome. By the end of 1984, 53 companies had been liquidated, only 7 had been sold. As the process continued, the emphasis shifted to more sales. During 1989-91, this was almost the only method of disincorporation being followed. 48. Indeed, the divestiture process could be clearly divided into two stages, 1983-88, and 1989-the present, corresponding to the "sexenios" of Presidents de la Madrid and Salinas. While the earlier period was characterized by divestment of large numbers of small enterprises, with liquidation the most common method, the latter period has seen a big jump in the average size of enterprise, and sale is now the more usual divestiture option. As shown in Table 4, the average price of the enterprise sold (in U.S. dollar terms) tripled in 1989 compared to 1988, doubled again in 1990, and then increased six-fold in 1991. In the latter period, some of the largest companies in the country, including Telmex, the telephone monopoly, all the largest commercial banks, Sidermex, the steel conglomerate, and so on have been sold. This latter phase seems to have been driven by the desire to raise very large sums of money as part of the macro- stabilization program. Table 4: PRIVATIZATION PROGRAM Number of Sales Total Receipts Total Receipts Average Year Nominal True in Mexico Pesos in $ million $ Price 83 4 2 4,847 -- -- 84 3 1 208 1.2 1.2 85 32 10 29,180 113.5 11.4 86 28 14 61,593 100.6 7.2 87 21 16 229,379 167.9 10.5 88 66 51 1,180,750 524.8 10.3 89 37 25 1,798,549 730.8 29.2 90 90 59 9,017,209 3,196.5 54.2 91 65 37 32,656,544 10,813.4 292.3 83-91 Total 346 215 49. The disengagement process has been discussed and described in some detail at various times (for example, PR, Annex IX). There are certain elements of the process, particularly after 1988, that need highlighting. (1) PWosalMehanism. Although formally the sector ministry with administrative responsibility for an enterprise "proposes" it for disengagement, the assumption now is that in principle all enterprises are to be disengaged, and the sector ministry actually has to justify retaining it. This subtle change in underlying attitude is partly responsible for the fact that very large numbers of enterprises have been disengaged. (U) PdadwonUni. In most cases, administrative responsibility for an enterprise slated for sale is transferred to the privatization unit in the Finance Secretariat. This unit, created in 1988, appears to be a model of decentralization and efficiency. It was a newly created unit, therefore none of - 20 - its key officials had any strong vested interests in the status quo. Its head, known as the Coordinador General, is an economist who was a professor at the Instituto Technologico Autonomo de Mexico (ITAM). The privatization unit reports to the Secretary, and the Coordinador General has the rank of a sub-secretary. The Coordinador General is assisted by six officials of the Director-General rank, all highly-trained, motivated and non-bureaucratic professionals. Their job is to manage the specific sales, but they do so in a supervisory capacity as the details are delegated to Agent Banks (see below). Thus the divestitures are carried out in a manner recalling arms-length market transactions. The idea is to use the expertise in business valuation already present in the commercial banks but to do so at arms-length so that competition among the banks for the Government's business assures good performance. The business is very desirable since the Government typically pays the Agent Bank a commission equal to one per cent of the final sale price. The disadvantage in relying on the commercial banks is that their expertise is in valuing companies from the private, not the social, point-of-view. Whether this contributed to the Mexican Government's clear reliance on commercial criteria in their divestiture program in unclear. (iii) Agent Banks As mentioned above, actual management of the sale is undertaken by an Agent Bank, appointed and supervised by an official of the privatization unit. The Agent Bank performs several functions. The first important one is to prepare a prospectus for the sale. Bids are then invited with bidders often pre- screened for financial viability and deposits required before the prospectus is released to them. A second crucial function is to perform a technical valuation of the company in order to arrive at a reservation price below which the enterprise would ostensibly not be sold. The reservation price is not derived from any simple formula; rather it is based on the Agent Bank's judgement after various valuation methods have been employed to estimate the value of the enterprise. The valuation methods used include: (a) Discounted cash flow: the present discounted value of all future cash flows that, it is estimated, the enterprise will generate; (b) Book value: the value of assets minus the value of liabilities as they appear in the balance sheet; (c) Liquidation value: the value of assets if they were to be sold minus the value of liabilities if they were to be discharged; and (d) Market value: the value of the enterprise on the stock market, if it is quoted on the stock exchange. Obviously the reservation price remains a closely guarded secret until after the sale; in fact, it may remain secret even after that. (iv) Selecdon Criterion. The basic style of the sale is that of a sealed bid first- price auction. As mentioned earlier, the bidders are pre-screened to make sure they are acceptable in terms of their technical, managerial and financial capabilities. All bids are then received at the same time in a previously - 21 - specified location in the presence of public notaries (who certify the bids), representatives of the Finance and Comptroller Secretariats, and the media. The Agent Bank examines and homogenizes the bids, i.e. puts the bide on the same basis. The plans of the bidders for the company in terms of investment programs and policies towards labor are taken into account, but the primary choice criterion at this stage is the highest offered price. The relative clarity of the choice criterion (which is a reasonable one given that bidders were pre- screened) is a further factor in assuring transparency and fairness in the sale process. 50. A question arises as to why the Mexican Government did not use the stock market to sell shares and thereby achieve privatization. Several reasons can be advanced for this. First, most PEs were not already quoted on the stock exchange, therefore they did not have established prices. Second, in almost all cases, it was perceived that the management of the enterprises needed to be changed. This would normally not be achieved (at least not immediately) through a stock market placement. Third, and perhaps most important, sales to well- identified investors was likely to yield the highest sales prices, a key goal of the Government. (v) Labor Reladon. An important feature of the Mexican privatization program has been the absence cf any serious opposition from labor. Now it is true that the trade union movement in Mexico has traditionally been closely allied with the political party (PRI) that has held power since the revolution. Nevertheless, if labor felt that its interests were likely to be severely threatened by privatization (as many believe is inevitable), surely some opposition might be expected to develop. Yet it has not, and indeed this lack of opposition has been an important necessary condition to allow the privatization program to proceed smoothly and quickly. 51. In fact, labor's interests have been protected throughout the program. This protection has taken three forms: First, the unions have, by law, a right of first refusal at each sale. That is, once the bids are known, the union can acquire the company by matching the highest bid. Many enterprises have been sold to their respective labor unions under this provision--a total of 16 in the period 1989-91 alone. They have included enterprises from various sectors-- fishing, sugar, food products, fertilizers, mining and auto parts. The largest sale has been the Bajio plant of Fertimex, sold to the union in December 1991 for over $31 million. One key measure to prevent abuse of this labor protection is the prohibition of resale. This prevents an outside bidder making a deal with the union to buy the enterprise after the union has exercised its right, thereby obviating the need to bid in the first place. 52. The second form of protection for labor has been the understanding with all buyers that worker layoffs are not to occur. Thus labor efficiency is expected to improve only gradually as the workforce is reduced through attrition or as output expands to more fully utilize the number of workers. In an early study of the effects of privatization, Perez Escamilla found at most a 5 per cent - 22 - reduction in the number of workers.1 A particular example is provided by Telmex, where, despite a perception of massive labor redundancy (the number of workers per line for Telmrx is twice as high as for American talephone companies), no significant layoffs have occurred; instead, the company plans to retrain its workers and to allow growth to absorb the redundancy. This prohibition against worker layoffs has in some cases worked strongly against sale. For example, as detailed in Annex II, no buyer could be found willing to pay an adequate price for Mexicana Airline at the first two attempted sales of the enterprise because buyers wanted the right to lay off workers or wanted Government to reach a settlement with labor, reducing the workforce. Government did not, however, accommodate these demands. No doubt the bar against layoffs must have reduced the price buyers have been willing to pay for enterprises, although it is difficult to astimate the size of this effect. This may be a small price to pay, however, to assure labor's cooperation with the divestiture process. 53. Third, there have been instances where employees have been sold shares in the company in order to encourage their co-operation. For example, the Telmex union was sold 4.4 per cent of the company for $325 million; the money for the purchase was loaned to the union by a government bank. At today's market price (April 16, 1992) this shareholding is worth $1.37 billion! The gain works out to over $20,000 per worker. It is not surprising that workers have not opposed divestiture in this case. 54. Apart from the protection labor has received, some observers believe that labor's cooperation has been induced partly by the drastic decision by the Government to declare Aeromexico bankrupt (details are in Annex II). In 1988, in response to a labor strike, the Government shut down the airline, fired the workforce (nearly 12,000 workers) and sold the assets. Aeromexico today has about 6,000 workers. A similar bankruptcy-led liquidation characterized the divestiture of the copper company, Minera de Cananea. The fear of job losses may be keeping labor quiescent as divestiture proceeds. (0i) FreinPardichadon. As part of its policy of liberalizing and opening up the economy to international trade, the Mexican Government liberalized rules for foreign ownership of Mexican companies. Minority ownership up to 49 per cent was permissible and even majority (or full) ownership is now possible under certain circumstances. As a result, there has been some foreign participation in the privatization program, although this is less than one might expect when such massive capital flows are involved. 55. According to a document of the Finance MinistryL5 98 per cent of 156 enterprises sold from January 1987 to August 1991 were purchased by Mexican investors. Indeed, only 3 outright sales to foreign companies took place, and L4 Juan Ricardo Perez Escamilla Costas: "El Redimensionamiento del Sector Publico en Mexico, 1982-88," mimeo SHCP, 1988. -W Secretariat of Finance and Public Credit, Office for Privatization: The Divestiture Process in Mexico, August 1991. - 23 - they were all relatively small (this did not include the sale in January 1992 of the SICARTSA Steel complex to Caribbean Ispat Ltd., an Indian-controlled company based in Trinidad). But there were several other important sales in which foreigners played a minority, but important, role. One of these was Telmex, where 25 per cent of the voting stock (10 per cent of total equity) was divided equally between Southwestern Bell and France Telecom, another 5 per cent of the (non-voting) stock was sold to Southwestern Bell, and a further 18.7 per cent of the stock was sold in the international stock markets. Twenty-two per cent of Telmex stock was foreign-owned even before divestiture began; as a result, foreigners now own 55.7 per cent of the company (see Annex I for details). 56. Two other prominent companies where foreign participation has been important are Mexicana Airline and Altos Hornos de Mexico (AHMSA), the steel and coal-mining complex. In Mexicana, the controlling group that acquired 25 per cent of the company is itself 49 per cent foreign. Among the foreign owners is Chase Manhattan Bank, which exchanged debt for 9 per cent of the equity. This is one of the few cases of divestiture in Nexico that involved a debt-equity swap, a practice that has not been encouraged subsequently. 57. In the case of AHMSA, a very interesting situation developed where the buyers--a group of private Mexican investors--had mining expertise (hence their interest in ARMSA's large coal-mining operation) but no experience whatsoever running a steel plant. They appealed to the Privatization Unit for help on this matter. The Privatization Unit acted as match-maker, bringing the buyers together with Dutch steel maker Hoogovens. Eventually, a deal was struck giving Hoogovens a five-year management contract to run their steel plant in exchange for all expenses, fees and a two-and-one-half per cent equity share in the company. Thus foreign technical expertise was acquired in exchange for an equity participation. 58. The need for foreign technical expertise also motivated the desire for foreign participation in Telmex. And once again, the Privatization unit played the match-maker. In the pre-bidding screening, en,uiries were solicited from potential domestic and international bidders. The main qualifying requirement for the foreign bidders was their technical expertise. After all the bidders were screened, the ones who qualified were encouraged to make alliances with local groups so as to have consortia with majority Mexican owners but with significant, technically-competent, foreign participation. 59. It is too early to tell whether the foreign participation has paid dividends to Mexico in terms of technical improvement. It is interesting, however, to contrast the action taken by the buyers of AHMSA, who felt a two-and- a-half per cent equity share was sufficient incentive to motivate Hoogovene to perform well, with that of the Mexican Government, which sold 15 per cent of Telmex to Southwestern Bell and France Telecom. The fact is, foreign investors have earned capital gains totalling well over $10 billion from their investment in Telmex since December 1990. - 24 - The Divestiture Proam: Successes and Failures 60. There can be little doubt that the Mexican divestiture program has been a great success. Probably its greatest achievement, and the reason why it is widely admired, is its magnitude. The number of state enterprises have been cut from 1,155 to around 200 in nine years. Further, the divestiture is not cosmetic. The Government has sold a number of huge companies--the telephone monopoly, the two major airlines, all the large commercial banks, mining companies, the entire steel sector, the fertilizer sector (still in process), and so on. There has been a partial or even complete withdrawal from an array of sectors that arguably should not have had government presence in the first place- -textiles, fishing, sugar, hotels, auto parts, etc. Even 'though the reduction in the assets or value added is much smaller proportionately (Pemex alone constituted about half the PE sector in terms of GDP contribution), this nevertheless constituv s a very significant reduction in Government's presence in the economy. 61. Besides this, the privatization program has generated very large revenues for the Government, thereby fulfilling a key role in the economic stabilization program through reduction of domestic debt and easing pressure on interest rates. As seen in Table 2, the sale proceeds from PE sales in the two years 1990-91 were around $14 billion, which is three to five times various estimates of the benefits to Mexico from the debt rescheduling. Thus, in terms of short-term revenue impact, it could be argued that the privatization program has been even more important than the Brady deal, although some maintain that the revenue- generating capacity of PE sales was greatly increased by the debt reduction that stimulated capital inflows. 62. A third expected benefit from the divestiture program is the increased efficiency of the economy as major firms operate more on market principles. It is still too early to assess the size of this impact. At the micro (i.e. firm) level, there is some mixed evidence. For example, as outlined in Annex II, the two privatized airlines present contrasting outcomes, with Aeromexico performing well and Mexicana performing not so well. However, accurate assessment of this effect even at the firm level cannot be made until more time has elapsed. Isues in Privazatlon 63. Despite its successes, the Mexican divestiture program has not been without some problems. One relates to the regulation of non-competitive industries. The Government has experienced problems in terms of setting up a regulatory environment. In cases of tradeable goods, reliance has been placed on liberalizing imports and, in principle, this measure ought to be adequate. The key sectors in this regard are steel and fertilizers which have only just been privatized, so it is too early to tell whether trade will do the job. The privatization of sugar mills was a problem case because the complicated regulatory structure and low prices hampered the ability of the newly private sugar mills to make adequate profits. The terms of loans the mills had been granted had to be renegotiated. Details on this case are difficult to uncover. 64. With non-tradeable goods, a regulatory structure is essential if there is any monopoly power in the market. Two cases have been prominent in Mexico-- - 25 - telecommunications and airlines. In the case of telecommunications, a regulato-y structure was put in place prior to the divestiture of Telmex, and while this mechanism does provide some structure to the market, officials in the Mexican communications secretariat are quick to concede that their regulatory capability at this time is rather limited. Considering this lack of regulatory capability, the Mexicar Government is to be commended for putting in place a fairly comprehe:ssive regulatory framework prior to the Telmex divestiture. The reguldtions embody quantity targets that were already in Telmex's own strategic plans and a price regulation patterned after Britain's RPI-X formula (see Annex I for details). Unfortunately, this has already been compromised more than once by agreement between company and the Government. 65. In the case of the airline industry, no regulatory changes were made for several years after divestiture, causing severe disruption of (or at least uncertainty in) the strategic planning of the firms. Aeromexico was sold in September 1988 and Mexicana in August 1989; yet it was only in July 1991 that the airline market was largely deregulated. In the meantime the airlines had to live with controlled domestic airfares that were too low and had no clear idea what entry regulations were going to be. This cost the industry two years at a time when the market was already unsettled because of the Persian Gulf War. Details are spelled out in Annex II. 66. A second issue relates to the Government's policy or selling enterprises to well-identified buyers rather than to diffuse buyers in the stock-market. Frequently a particular individual can be identified as "the buyer". Thus it may be said that Carlos Slim bought Telmex or that Roberto Hernandez bought Banamex. Typically the buying group would have a dominating controlling share either because they were sold more than 50 per cent of the stock (such as in the case of Banamex) or because of innovations that allow them control even with a minority shareholding (see Annex I for how this was done in the case of Telmex and Annex II for the case of Mexicana Airline). This style of sale might be thought desirable as giving the buyer all the leeway he needs, in particular in order for him to concentrate on maximizing long-term shareholder wealth without worrying about corporate raiders or other short-term considerations (see para. 50 for arguments to support this position). It allows transfer of guaranteed control without requiring the huge capital outlay that purchase of a 50 per cent shareholding might require. But at the same time, this could weaken an important characteristic of the market economy, one that is often cited as one of the key justifications for privatization. Private enterprises are said to operate more efficiently than their state-owned cousins because they are forced to face the discipline of the capital market. Partly this argument has to do with the fact that, if they need capital, private companies must raise it on the capital market and must therefore face the scrutiny of bankers or investors. But partly it also has to do with the assertion that, if the managers of a private company are not performing well, others can take over the company through the stockmarket. This last kind of check is obviously impossible if the manager-owners control more than 50 per cent of the company. An important advantage of private enterprise is lost. 67. A third question about the divestiture program is special to the case of Telmex. It could be argued that, because a large proportion of the stock was sold to foreign investors, too much rent from the sale dissipated abroad. - 26 - Calculations show that whereas foreign investors paid the Mexican Government a total of US$3.35 billion for about 30 per cent of the stock, they have gained about $5.5 billion already through stock appreciation. In fairness, the Government sold at market prices, and therefore this argument could be dismissed as 20-20 hindsight. Nevertheless, the question remains whether Government could have anticipated this because of its intimate knowledge of the case. The point is discussed in greater detail in Annex I. 68. The problems mentioned, however, are all minor compared to the magnitude of the successes achieved by the divestiture program. Thus Mexico provides an example of outstanding success to other countries considering or carrying out programs of divestiture. EactarsQntdWr to the Success of the Divesfiture Progam 69. A series of factors contributed to the success of Mexico's divestiture program: (1) Commitment. The key ingredient for success was commitment to the program at the highest levels of the Mexican Government. This commitment was, at least, only partly ideological. Rather, it was driven primarily by the perception that divestiture could play a crucial role in the macro-stabilization of the economy. (H) Tansarena . The process of divestiture is clearly defined and transparent. No deviations are made from the announced set of rules. To avoid bargaining after a bid is accepted, Government draws up the legal contract of sale before calling for bids and permits potential buyers to question (and even modify) aspects of the contract before bidding commences. The key element for transparency is the system of checks and balances through the inclusion of several actors in the process--sector ministries, the privatization unit and the agent banks. The process of bidding rather than negotiation also creates a more transparent situation for all buyers. (lUi) AdministmiveSeamtdon. As was rentioned above, administrative control of enterprises in the process of divestiture was generally transferred from the controlling secretariat to the privatization unit at the Finance Secretariat. Thus, for example, when the telephone monopoly, Telmex, was being readied for divestiture, control over it was moved away from the Communications Ministry. This policy has had three advantages: (a) actions deemed necessary as precursors to privatization could be taken quickly and without dilution or delay; (b) any potential conflict of interest between privatization and entrer ihed bureaucracies was avoided; and (c) the management of the divestitures was concentrated into a few capable hands at the privatization unit, allowing benefits from learning and from specialization to be realized. - 27 - (&) Seke&n CHtedon. A fourth innovation that probably had an important role to play in the scale of divestiture was the way in which enterprises were selected for sale. In the early years of the program, a more "normal" process was followed, in which different sector ministries were invited to offer enterprises for sale on grounds that they were in non-priority sectors. The approach was to present an argument why particular enterprises were appropriate for divestiture. As the program developed, however, the approach shifted to one where the presumption was that all enterprises would be sold and the sector ministry would have to argue why particular enterprises were not suitable for divestiture if they were to be kept off the list. This change in emphasis may have been strongly instrumental in keeping the divestiture program in high gear even after the obvious candidates had already been sold off. Summa' and Lessons Learned 70. Parformance under the disengagement component of the PERL was extremely good. The Government moved faster than had been envisioned, and carried out a highly successful program. The revenues from PE sales, $14 billion during 1990- 91, played an important role in domestic debt reduction and the stabilization of the economy. Some of the key lessons learned can be summarized below: (a) Consolidation of divestiture activity in a central unit, close to a power center, which then supervises management of the sales by qualified agent banks, can be an effective way of organizing sales. (b) A clear, transparent, well-defined process with clearly specified rules and sequences, can ensure orderly sales. The scope for negotiations between buyers and the Government should be minimized. (c) Government can serve as an effective conduit or clearing-house to bring together potential foreign partners with domestic investors. (d) For tradeable goods, trade liberalization and. deregulation are useful precursors to divestiture. (e) For non-tradeable goods in non-competitive industries, regulatory mechanisms should be designed and implemented prior to divestiture. (f) For enterprises whose value is expected to rise sharply over time, or where there do not exist adequate capital markets to value shares, a system of selling in tranches can be an effective way for Government to receive a share of subsequent gains. (g) Labor's cooperation and acceptance can be assured through protection of jobs or other ways for labor to participate in the gains from divestiture. - 28 - (h) For companies not viable as sale prospects, liquidation can and should be used as an alternative to reduce the fiscal impact. C. Public Enterprise Reform 71. The public enterprise reform component of the PERL was much more specific in its objectives and implementation requirements than the divestiture component. As in the case of the divestiture program, the Mexican Government was pursuing its own vigorous program with respect to public enterprise reform. Where the requirements of the PERL coincided with the policy components of this program, performance of the PERL was quite satisfactory. But where the PERL came into conflict with the programs of the Mexican Government, or where it came up against severe constraints, the goals of the PERL were sacrificed, at least in the short- term. 72. There were three broad elements in the reform component of the PERL. (i) Improvements in the Competitive Environment for Public Enterprises; (ii) Measures to Foster Managerial Autonomy and Accountability; and (iii) Measures to Foster Financial Autonomy and Accountability. These are presented in summary form in Annex V of the President's Report, and are considered in turn below. The responsibilities of the various government agencies responsible for implementing this program are shown in chart 1. (1) Improvements In the Competitive Environment for Public Enterpise (a) Price Reforms 73. The PERL called for a series of price-related reforms, mostly in the form of price increases for large public sector companies. There were two goals in this plan. The first was to help reduce the budget deficit by reducing the extent to which public enterprises were subsidized. The second was to promote economic and financial efficiency. Under the plan (pages 9-10 of the PCR), only electricity prices were raised on a regular schedule. In the case of all other goods--food, fertilizers, steel and rail freight--Government decided that price increases would jeopardize their anti-inflation program, particularly as embodied in their agreement with business and labor, the PACTO. Indeed, the control of inflation was the prime goal of the macroeconomic stabilization that the PERL was supporting. Thus there seems to have been a conflict between the Government's price stabilization objectives and the price liberalization requirements under the PERL. The Government chose to set aside the requirements of the PERL. 74. A further problem arose as the Government sought to meet its deficit reduction target (also required under the PERL) while holding prices constant. Obviously expenditures had to be cut, and public investment fell drastically as a consequence. Non-PEMEX public sector investment fell from 3.4 per cent of GDP in 1988 to 2.9 per cent of GDP in 1989 (down from a peak of 9.1 per cent in 1981 and even 4.8 per cent in 1985 and 1986). The long-term consequences of this dramatic fall in public investment have still to be realized. Chart 1 MEXICO PUBLIC ENTERPRISE REFORM Institutional Responsibilities for Program Implementation SHCP Overa# Progrm Coordineon • Eslabnsh Unidad de Comenlo de Desempeno • Seect PEs fo Pefonance Agrements • Coonae ienee1Pogram • Issu simpHmed PEs integratedS nhormaon system • Caonieleand mSECEae s F :mughPE" •U»n% IssuE ~ 1985 PE Law Reguladons Sudy Govemment mining sector agences' <øle (P sysm proide Be ~ * imfannt agram • Issue Ra* prb TORs fm esW ~ offnani • Prepie .yeramneovalueio flde Bmes e e • SL*on~, s F FERRONALES Law ::=oioe abY PEsi0 olae ec~ømhn Soumw: Presidei Repor .on 3086-ME, Annex Vl. a~sk1613 - 30 - 75. Eventually, however, prices were increased or liberalized in a number of commodities. Steel prices were freed in September 1990 as a prelude to privatization of the steel sector. Fertilizer prices were similarly liberalized in 1991 in anticipation of the sale of FERTIMEX. Trade in both steel and fertilizers is essentially free.!6 Rail prices continued to be a problem, however. A study to examine economic rail prices, as required under the PERL, was completed in November 1990. The study advocated the use of long-run marginal cost pricing, which would involve raising prices an average of 17 per cent with considerable restructuring. In early 1991 prices were raised an average of 25 per cent, although not following the pattern in the study. The consequence of this price increase, however, according to officials at FERRONALES, was a further erosion in demand for rail freight and a transfer to trucks. As Table 5 below shows, the fundamental problem for FERRONALES is the shift in demand to road transport, and this shift is noticeable in every category of freight (see Table 6). Further, rail freight prices have been moving up ahead of inflation throughout the 1980s, between 1983-91, tariffs increased by over 50 per cent in real terms. Thus price increases may not be the solution to the problem of cash flow in FERRONALES. As indicated in the study on railway prices,- the cost per ton-km is 20% higher for FERRONALES than the average for the U.S. railroads (58 pesos per ton-km as against 47.2 pesos in the U.S.), despite considerably lower wages in Mexico. Table 5: DISTRIBUTION OF THE FREIGHT MARKET Million - tons Billion ton - km 1980 1990 1980 1990 Rail 60 51 41.3 36.4 Road 253 311 82.2 107.7 Total 313 362 123.5 144.1 Sourcet FERRONALES U There remained a 10 per cent tariff on the import of phosphate fertilizers. Steel tariffs are down to zero in some cases. L Felipe Ochoa y Asociades: "Evaluacion de los sistemas de costos e integracion de un modelo sistematizado de tarifas," FERRONALES Document, November 1990. - 31 - Table 6: PATTERN OF DEMAND FOR RAIL FREIGHT, 1984-90 Products 1984 1990 Reduction Per cent Reduction Agricultural Products 15,088 11,495 -3,593 -24.0 Coke Barites 1,116 424 -692 -62.0 Iron Ore 305 54 -251 -82.0 Minerals 7,144 4,074 -3,070 -43.0 Combustibles 1,488 723 -765 -51.0 Sulphur 3,493 2,867 -626 -18.0 Gypsum 532 188 -344 -65.0 Cement 570 153 -417 -73.0 Fertilizer 8,020 7,190 -830 -10.0 Billets 3,988 2,240 -1,748 -44.0 Steel Tubes 487 13 -474 -97.0 625 174 -451 -72.0 Subtotal 42,856 29,595 -13,261 -31.0 Others 21,263 21,365 102 - TOTAL 64,119 50L960 -132159 -20.0 Source: FERRONALES. 76. On the whole, therefore, the price adjustment component had a somewhat mixed outcome. Electricity prices were adjusted as envisioned. Steel and fertilizer prices were held constant as part of the PACTO's anti-inflation program; they were finally liberalized as a prelude to privatization. Rail freight prices were adjusted in 1990, but the company's finances remained problematic as it was unable even to cover its operating cost for the first time. This is seen clearly in Table 7 below, where the income before transfers was negative in 1990, unlike the previous three years. - 32 - Table 7: FERRONALES - Income Statement (Billions of pesos) 1987 1988 1989 1990 Sales of Services 849 1603 2002 2313 - Costs of Services 669 1344 1786 2093 - Administrative Costs 63 146 194 247 - Operating Income (Loss) 116 113 22 (26) - Interest expense 68 64 32 42 + Other income (expense) 59 79 37 (32) W Income before transfers 107 128 27 (100) + Transfers from the Government 169 100 143 75 - Exchange losses 276 29 81 48 - Pension expense 182 167 210 300 - Net income (loss) (182) (31) (121) (38) (b) Other Iprovements in the Competitive Environment of PEs 77. The PERL called for a series of other measures to improve the competitive environment of PEs, notably the injection of more private elements in markets now or previously dominated by large PEs. Here, the performance has been satisfactory. New regulations to the Regulatory Law of Article 27 of the Constitution allow private exploitation of certain minerals (phosphorous, sulfur and potassium) previously limited to the state. Another element of this component was aimed at the introduction of some degree of competition in the railway industry by identifying activities that could be separated from the core activities of FERRONALES and opened to competitions. There has been some progress in this regard. Private investment is now allowed in container terminals, repair workshops and even short lines which FERRONALES has decided to close. Further, private companies are able to run unit trains, where the operation is by FERRONALES personnel, but the private company markets the freight service. 78. An important area of improvement that was not specifically envisioned in the PERL was the divestiture of the steel and fertilizer sectors. Here the Mexican Government moved further and faster than had been anticipated. The steel - 33 - sector has been privatized and the privatization of FERTIMEX is proceeding rapidly. (2) Measures to Foster Managerlal Autonomy and Accountablfity 79. The centerpiece of the public enterprise reform component of the PERL was a series of measures to improve PE performance by granting PE managers greater autonomy while also strengthening accountability so as to prevent any misuse of the newly acquired autonomy. The program called for a shifting of decision- making responsibility away from the ministries (including the Sector Ministry, Programming and Budget, Finance and the Contraloria) to the Boards of Directors and enterprise management. What was envisioned in the PERL were: (a) issuance of draft regulations to the new PE law, giving effect to these changes; (b) adoption of measures to strengthen and enhance the effectiveness of the Boards of Directors; and (c) implementation of performance agreements with five major PEs: FERTIMEX, FERRONALES and three companies within SIDERMEX--the steel holding company--namely, AHMSA, CMC and SICARTSA. (a) Reulations to PE Law 80. After a delay of over three years since the passage of the PE law, the regulations to the PE law were issued in January 1990. Although this was an important step forward in improving PE efficiency, there are several major weaknesses that have prevented the law from having its full desired beneficial impact. As outlined in the President's Report (pages 21-22), the key to fostering more managerial autonomy is to move from a system of ex-ante controls to one of ex-post controls. In 1989, it was noted that the "existing managerial/supervisory system for public enterprises...is designed to minimize Government budget deficits and is not conducive to increased productivity and renewed growth" (PR, p. 21). This appears to be true even today. The Government has reserved discretionary powers for itself in all aspects of the budgetary relationship with enterprises, and has continued to exercise these. The supervision report of August 1990 noted these points: "... the individuaL budget of public entities may be readjusted by SPP on a selective basis...." "... the largest 19 PEs...(remained)...under strict controls and prior approval requirements on any changes to the budget." "... quarterly targets (remained) on expenditures for PEs operating under the "Conveitioa Deficit/Superivit" (with subtargets for salaries and other operating expenditures, investment and purchases of equipment)." "... (there has been) resistance, particularly at SPP, to relinquish centralized control and interference on PE's operational decisions..." This last point was echoed in a parallel supervision report on the Fertilizer Sector Adjustment Loan, which noted "the strong resistance at the central - 34 - ministry levels, particularly SPP, to change the well embedded culture of centralized control and daily interference in the management of PE9." 81. The central point is that tight Government controls over the budget remained even after the regulations to the PE Law were promulgated. This was confirmed repeatedly in conversations with officials at enterprises, sector ministries and, surprisingly, even SPP. The perception at SPP is that, because Mexico has been in a state of budgetary crisis, it has been incumbent on them to maintain strict controls over expenditures. They point to the overall success of the program as evidence that they were right. This raises once again the question of compatibility of different goals of the PERL. Clearly, macro- stabilization was the core purpose of the PERL. When other goals--such as improving PE autonomy--came into conflict with this (at least in the perception of the officials who must implement the policy changes), modification to these goals was called for. The Bank implicitly accepted this modification in that the second tranche of the loan was approved despite the fact that this shortcoming in the implementation of the regulations was noted. (b) Chanes to the Role and Functions of the Boards of Directors 82. It was envisioned in the PERL that the Boards of Directors (BODs) would take over many of the functions previously exercised by the Central Government. Because the BOD9 were typically composed of representatives of various ministries, an essential step to assure their independence was to include more non-affiliated individuals, particularly business people from the private sector. This does not seem to have happened; BC1Ds continue to have the same composition as before. For example, the board for FERRONALES consists of senior government officials, two union representatives and two representatives from the business sector.L" This list is typical in that other PEs have BODe with much the same character. Ostensibly, private sector members with requisite skills could not be found, or would not be appropriate because they could be potential competitors or buyers of PEs. For example, as FERTIMEX is now being sold, it is argued that it would have been inappropriate to have placed potential buyers on its board. There is some privatization of operations surrounding FERRONALES so a similar argument is being made there. Although there is some validity to this argument, it nevertheless seems that there is little commitment on the part of the Government to change the composition of the BODs. 1t - Secretary for Communications and Transport - Secretary for Finance (Hacienda) - Secretary for Programming and Budget - Secretary for Energy, Mines and Industry - Secretary for Commerce and Trade - Secretary for Agriculture - Subsecretary for Transport - Representative of the Railway Workers Union - Representative of the Railway Workers Union - Representative of the Federation of Chambers of Commerce - Representative of the Federation of Chambers of Industry - 35 - 83. In addition to changing composition, the BODs were to receive greater resources and incentives to play a more active role in the management of the enterprises. In fact, board members do now receive payment for attendance at meetings, and attendance by the members, rather than their representatives as in the past, has gone up as a result. But since these are some of the highest officials of the Government, it is unrealistic to expect them to really devote much time to the management of the enterprises. Furthermore, since they are representatives of the Government, their active involvement may simply constitute another channel for centralized interference in the day-to-day affairs of the enterprise. Indeed the representation of the heads of the sector ministries on a BOD appears to detract from the envisioned role of the BOD acting as a buffer between the Ministry and the PE. Thus without a substantial change in the composition of the Boards, it is unclear that their active involvement would be beneficial to the PEs anyway. 84. In summary, essentially little progress appears to have been made in enhancing the effectiveness of the BODs in ensuring greater autonomy and accountability in enterprise management. (c) Performance Arreements 85. The principal tool to improve managerial autonomy and accountability in PEs was to be the performance agreement (PA). In principle, once agreement is reached in advance to a well-defined objective or set of objectives, the enterprise can be granted autonomy. Accountability is assured by an ex-post evaluation of performance relative to the objectives, and incentives are provided through bonus payments tied to the degree to which the objectives are achieved. Under the PERL, PAs were to be concluded and implemented for 1990 with five companies: FERTIMEX, FERRONALES, AHMSA, CMC and SICARTSA. This was in fact done. 86. Although some valuable experience has been gained in the process of developing these five PAs, and a certain level of commitment to and interest in them has grown in the Mexican Government, the overall performance of this component of the PERL was disappointing. Perhaps the best evidence of this is that in no case was a follow-up PA signed for 1991;Li and although some draft PAs were in circulation in January 1992 for FERRONALES and some other PEs not included in the first set, none had been signed up to that time for 1992. There were three important factors that contributed to the disappointing performance: (i) Lack of adequate consultation between Government and the enterprises. (ii) The attempt on the part of the Government to use the PAs as an instrument of control. ni Since the PEs under SIDERMEX have been privatized and FERTIMEX is being privatized, new PAs are not relevant for these enterprises. Note, however, that none of these divestitures took place until late in 1991, so PAs for 1991 (which would have been signed in late 1990) could have been appropriate. - 36 - (iii) Poor design of the PAs thempelves. These are discussed in turn. () Lack of consultation 87. A management consulting firm, funded by the PERL, was to work with the sector ministries to design the PAs. But at the end of the process, a decision was made in SPP--and the newly-established PA unit--to develop the PA design in- house as the consultant's design was perceived to have major deficiencies. The outcome of this process, however, was that--at least for FERTIMEX and FERRONALESL1--the final designs had almost no input from the companies themselves. The PAs were simply imposed from above. In both cases, company officials indicated that objections or suggestions they made were simply ignored. In particular, several of the indicators used were found inappropriate since they were dependent on factors outside the company's control. A consequence of this kind of intense disagreement was an unwillingness to truly co-operate to make the program successful. One expression of this was that FERTIMEX set its targets at very low levels so that they would be relatively easy to achieve. Since the only individuals in a position to properly evaluate the appropriateness of the target levels had been by-passed during the process (FERTIMEX and the sector ministry, SEMIP), this subterfuge passed through. 88. The lack of consultation seems to characterize the process by which PAe are being drafted even now. Officials of the PA unit said that the order of events is as follows: a draft PA is developed by one of the seven professionals at the PA unit; it is then sent for comment to various concerned agencies, such as Banco de Mexico, SHCP, SECOGEF (the Controloria), etc.; then finally, after incorporation of the comments, it is sent to the firm. It should be noted that the "concerned" agencies mentioned by the official are all concerned with the budget. A crucial element of a PA system is that they are meant to be agreements, and to be successful, they must have the agreement and co-operation of the enterprises as well as the Government. Full consultation with the enterprise during the design stage is essential to ensuring this co-operation. (II) The Ue of PAs as an Instrument of Control 89. The perception of the PEs is that PAs are one more way for the Government to control them. Surprisingly, this is the view not only in the enterprises but also in the controlling agezicy, SPP. At the enterprise level, the feeling was that this was "just one more of the many forms of control the Government has" and at SPP that the PAs "complement other instruments Government has". It is, therefore, likely that SPP began to view and to attempt to use the PAs as another form of budgetary control, rather than a completely new approach to managing PEs. One manifestation of thir is that the PAs were implemented by officials who retained their nominal positions at SPP, rather than by professionals devoted full-time to the PA system. Until the view of PAs as another control instrument 1t By the time of the OED mission, the steel sector had been completely privatized; the new management was completely unaware of the PAs, so no information on the steel sector PAs could be gathered. - 37 - is abandoned, the PAs cannot perform their function, which is to serve as a mechanism by which PE management can be granted autonomy, given incentives for good performance, and be held accountable for their actions. Indeed, a passage from a Bank memorandum of September 1989, is instructive. SPP officials had stated "that central controls will need to be maintained as long as there are major difficulties in the macroeconomic environment in order to ensure that fiscal and other macroeconomic goals are met. SPP indicated that it does not envisage, in the short-term, substantial reforms in budgeting procedures...Furthermore, it is SPP's view that it would be diffie,-.t to phase out the quarterly "Deficit/Superavit Convenios" by 1990. Because of their restrictive nature, these convenios conflict with the Performance Agreements envisaged under the Project." (U) The DesIgn offih PAs 90. The PAs had many of the characteristics of good design, such as clarity of indicators, consistent weights, well-defined targets and clearly-stated performance levels. They, therefore, form a basis for future development into effective management tools. There are several areas, however, in which they require considerable improvement. 91. The first of these relates to the selection of indicators. As they stand, the PAs indicators lack coherence; they overlap and leave gaps. To illustrate the point, consider the list of indicators in the PA for FERRONALES.L" 11 Since FERTIMEX is being privatized, and SIDERMEX already has been, FERRONALES is the only one of the original five PEs selected for PAs that will continue in the public sector. Nevertheless, the point being made applies also to the other PAs. - 38 - Table 8: INDICATORS FOR FERRONALES PERFORMANCE AGREEMENT, 1990 Indicator ta 1. Operation and Ttaffiel A. .ove Ut of .Inporte cargo 17 .00 8. Paytoeints to waganotoere 2 .00 4. Total traffic 1.00 2a LaComotives and vakonet A. -*ffber of locomotive reconstructed 245 S. Available locomotives 9.75 C. Available care 3.00 3. Track and cotmainicationst A. Completion of double-tracking on Mexico-Queretaro route 3.00 V' Renovation of track- 3.0 C. Rehabilitation of bridget 2.25 P. Amount of stone ballast oTr track 2:25 S. Replacement of eleepers 2.2 P. Lavelling of track 2.25 ..4.:..K <YPersonnel and Administrations A. Completion of.study on human resources 7.50 1. Completion of atudy on otr4ctural changes 7.50 finance:t...... ... A. Level of profit 21.00 $ Completion of study on rail j pricing 14.00 92. Examination of this list of indicators does not suggest any coordinating principle governing their selection. Rather, they are a somewhat considered list of activities that are regarded in some sense as "important" or "problem areas that need extra effort". Indeed this was the way in which they were selected. Such a list would be typical for a management consultant invited to propose a series of measures to improve the efficiency of an enterprise. But they are not appropriate as a list of indicators for a PA. The indicators for a PA should reflect a clearly defined overall objective, they should then be mutually exclusive and collectively exhaustive. The indicators on the list above have neither of these characteristics. For example, there is some overlap in the indicators, which might have the unintended consequence of overweighting the activity or item that is being double-counted. Item 1A is clearly included also in item 2C. Similarly items 2A and 2B are related. The number of locomotives reconstructed clearly influences the number that are available. Also, the various items under 1 and 2 are clearly also reflected in 5A: the level of profit. Thus the indicators violate what has been called "the fundamental principle of performance evaluation," that all benefits and all costs be counted exactly once. 93. A second problem area in the design of the PAs is that the indicators are far too diffuse to give management a clear signal as to priorities. Indicators with very small weights (as low as 1%), inevitable when there are so many indicators, are likely to be ignored anyway. What is needed is a clear objective with just one or two key indicators getting the bulk of the weight. - 39 - 94. Third, explicit recognition needs to be given to factors outside of management's control, so that these do not influence the evaluation of performance. For example, given that rail prices were controlled during 1990, the level of profit is not entirely within the company's control. Modified indicators are required to eliminate the efforts of uncontrollable factors. 95. Thus, although the design of the PAs was based on some desirable elements, there exists a need for fundamental changes. The Bank had expressed reservations to the Government of Mexico on the selection of the consultant who designed the performance agreements. Bank staff made considerable effort through frequent supervision missions to address the problems associated with the design of the PAs. However, the effectiveness of the supervision effort over this aspect of the loan fell off after the departure of the Bank staff member responsible for supervision. Government officials felt they could have benefitted greatly from continued active involvement and interest on the part of Bank staff. The Government came to realize that the PA designs were deficient and took over the drafting of the final agreements from the consultant. However, their own limited experience precluded a thorough re-working of the drafts, and it was at this point that Bank help may have been extremely useful. 96. In summary, the performance agreements fell short of fulfilling their key role in the enhancement of managerial autonomy and accountability. In part, this was due to poor design, lack of enterprise-Government cooperation and the inappropriate use of the PAs as an instrument of control. Given that the Government is committed to continuing with the experiment, the Bank could provide additional technical assistance to deal with the problems discussed above. However, problems of incompatibility between the PAs and other instruments-- notably the Convenios Deficit-Superavit--will have to be resolved before effective performance agreements can be implemented. (3) Measures to Foster Financial Autonomy and Accountability 97. There has already been some discussion, in the context of the discussion on the regulations to the PE law, of attempts within the PERL to encourage a loosening of SPP's budgetary controls over PEs. By and large, these attempts have been thwarted by the conflicting goal of containing the budget deficit. As a result, the overall budgets of PEs have remained under tight control with frequent monitoring. 98. Within the tight budget, there has been some movement toward financial autonomy. Specifically, expenditure limits for individual contracts for civil works or other investment programs not requiring prior approval have been raised substantially. Although these are difficult to compare precisely because of the 86 per cent inflation that has occurred during this period, the expenditure limits appear to have roughly doubled in real terms. Although this falls short of a desirable freeing of management from ex-ante controls, it is clearly a step in the right direction. 99. In addition to improving the budget process, the PERL also called for enhancement of the investment decision-making process through the establishment of an investment-planning unit and the adoption of a multi-year investment budgeting process. The investment planning unit was set up as planned. A write - 40 - up on its activities (paras. 2.56 to 2.58) is presented in the PCR. The mission, however, was unable to discuss the working of this unit with relevant officials and assess its impact. As regards the introduction of a system of multi-year budgeting for large project investments, officials expressed considerable reservations about its feasibility, because Congress, by constitutional requirement, has to approve the budget each year. Although legally this would preclude any multi-year budgetary commitment for a project, certain procedural changes could be developed to ensure that future financial outlays for large investments are noted in advance and taken into account in the budget exercise. Thus future capital expenditures required to complete large projects could be spelled out in an annex as information for Congress to be appropriately taken into account in the budget exercise. 100. As mentior-d earlier, the PERL also included technical assistance to strengthen the existing Systema Integral (SI) or the integrated information system. The major motivation was to develop a simplified information system for monitoring PEs. Technical assistance to the CIGF was included in the PERL for this purpose. The actual results of this effort are modest. Some minor changes were made by SPP to obtain a more detailed breakdown of revenues and expenditures (Form E-01). SPP also made changes to ensure the data are consistent; the system is being computerized. Some YE3 are even sending the base data forms to SPP electronically. 101. The technical assistance program, on the whole, was not adequately utilized. Only some $300,000 of the total of $1 million in TA allocation was utilized m"inly in the preparation of the PA blue print, purchase of a computer and software, study of FERRONALES tariffs, and some training involving visits by UCD staff to countries that have adopted PAs. Two studies for which TA was earmarked-nrole of mining sector agencies and private sector development in PE dominated areas--were not completed, although substantial actions in both mining and other areas (e.g. railways) were undertaken to allow greater private sector participation. (The Mining Sector Study was undertaken subsequently in the context of the Mining Sector Restructuring Project). The experience with the TA program suggests that the Bank should carefully assess the merit of studying specific issues and the need for TA before earmarking funds for these purposes. Often governments are persuaded to accept TA under some pressure but subsequently these funds are not properly utilized. 102. Given that the Government is committed to the adoption of the PA instrument to improve the performance of the PE sector, it is important that the deficiencies in the present PA framework are corrected. The UCD plans to introduce PAs for several firms and it would be useful if the appropriate advisory inputs were provided to the UCD to both design the PAs and to negotiate their agreement with the respective PEs. The remaining TA funds could be utilized for this purpose. Conclusions ad Lessons 103. Notwithstanding Mexico's exceedingly well developed administrative system, experience with the public enterprise reform component of the PERL brings out the formidable problems associated with actually reforming public enterprises. The performance agreement, in concept, is a sound instrument for improving PE - 41- efficiency. In practice, there are many problems associated with its effective implementation not only in selecting appropriate indicators, and setting weights and realistic performance levels, but in actually allowing PEs the necessary autonomy and resources to fulfill their objectives. Development of a satisfactory PA system is thus likely to require several years since it is a system which requires a certain learning period for both the Government and PEs. Another issue is the appropriateness of introducing PAs in a period marked by tight fiscal policies to reduce budgetary deficits. It would appear that there was a substantial incompatibility of objectives between the inflation control requirements of the PACTO and increased financial autonomy required by PAs. A related issue concerns the compatibility of the various instruments employed by the Government in monitoring PEs. Since some of the PEs were engaged in medium term restructuring under the Financial Restructuring Agreements, these requirements should have been addressed in the design of the PAs for these enterprises. These considerations need to be given explicit attention in future PAs. 104. A second set of issues and lessons involve the approach adopted under the PERL with regard to the regulations to the PE law to enhance autonomy for PE management while maintaining adequate accountability. The major thrust of the reform aimed at empowering the Board to take important decisions with the Government limiting its interventions to a few strategic matters including appointing and evaluating performance of Board members and senior management. However, the composition of the Board is a critical factor in accomplishing this objective. If senior government officials, particularly from the same sector ministry are represented in the BOD, it would be impossible for the sector ministry to maintain an arms-length relationship. Thus, "government representation on the board should be kept to a minority and in no case should a high level official chair the board since that destroys its collegial nature and makes it an arm of the ministry". L This change needs to be introduced in PEs concinuing under Government control. LU Bank Lending for State-Owned Enterprise Sector Reform: A Review of Issues and Lessons of Experleuce, IBRD Doc. See M88-1082, September 1988. - 43 - IV. Concluding Remarks on the Bank's Role and Loan Impact 105. The Bank prepared the PERL under great time pressure and thus some technical issues that could and should have been analyzed in greater depth perforce had to be left for follow up at a later stage. In particular this would seem to apply to the regulations supporting the FLPE as well as the framework for the performance agreements. In particular the move from ex-ante to ex-post controls in supervising PEs necessitates a substantial change in the respective roles of the various parties involved viz. SPP, UCD, Sector Ministries, Controlaria, and the Boards of Directors. The fact that after issue of the regulations, there has been little change in the manner in which PBE are supervised/controlled, as well as in the role of the Board or in its composition supports this conclusion. The Bank also could have played a more active role in the formulation of the PA framework, transferring the lessons of experience of other countries.- Active support for both these tasks in the crucial period before release of the second tranche would have perhaps improved their implementation prospects. The implementation experience also highlights the importance of continuity of supervision. 106. By any normal measure, the PERL can be considered a success. Its key goals were to assist in the stabilization of the macroeconomy, restore investor confidence, and reverse net capital outflow. This has been certainly achieved. It supported what is clearly one of the world's most successful divestiture programs. It has, however, fallen short of achieving wide-ranging public sector reforms. This can be rectified with Government support now that the macroeconomic situation is propitious to allow greater autonomy to PEs. 107. The main reason for the success of the PERL is that, to a large extent, it supported a policy reform program to which the Government was totally committed--macroeconomic stabilization through debt rescheduling, fiscal austerity, tax reforms and divestiture; and structural reforms through deregulation, trade liberalization and financial sector reform. Public enterprise reform per se apparently was much lower in the set of policy priorities and suffered as a consequence. This underscores another conclusion: that support for a committed, capable Government in carrying out its sound policy programs has a high probability of success, but in such situations, attempts to amend the policy agenda through fine tuning may be problematic and risky as it involves rearranging Government priorities. - 45 - ANNEX I Page 1 of 11 Mexico: Privatization of TELMEX 1. Telmex, the telephone monopoly, was divested in December 1990. At the time, the sale was the biggest single divestiture transaction seen in Mexico (raising $'.7 billion), and, although this has since been surpassed by the sale of Banamex, the largest commercial bank, Telmex remains the biggest money-maker of the entire divestiture program, having already yielded to the Government an amount in excess of $5 billion. In addition, Government still owns stock in the company, valued at today's prices (April 1992) at around $3 billion, and it has announced plans to sell half its holdings in May. Thus the Telmex divestiture has been at the core of the Mexican Government's strategy of privatization-cum- stabIlization. 2. Telmex entered the parastatal sector in 1972. The company had been formally created (in the private sector) in 1941. At the time, there were two competing telephone systems in Mexico, owned respectively by Ericsson of Sweden and ITT of the United States. The critical feature of this telephone duopoly was that the two networks were not linked; thus a subscriber to one could not call a subscriber to the other. Telmex assumed the concession of Ericsson, and then, in 1950, bought out ITT and began the task of integrating the two systems. Much of the money needed to accomplish this was loaned to the company by the Mexican Government. In 1972, the Government converted its loans to equity and bought some additional stock in the market to give itself a 51 per cent share of the company. At the time of privatization, this share was up to 56 per cent. 3. The announcement that Telmex was to be privatized waa made in September 1989, although apparently the decision had been made a year earlier during the presidential campaign of President Salinas. Following the announcement, administrative control over the company passed from the Secretariat of Communications and Transport to the Privatization Unit at the Finance Ministry. A series of steps were then taken to ensure a successful sale. A major overhaul of the price-cum-tax regime controlling Telmex was made in January 1990 as part of the preparation for sale. An innovative capital restructuring was undertaken in order to enable the Government to transfer control while selling only a part of its holdings. A major regulatory reform was carried out as part of the divestiture, putting into place the regulatory regime to control the private monopoly. Finally, explicit steps were taken to ensure the participation of foreign telecommunications firms in the bidding. The initial sale and transfer of control took place in December 1990--only fifteen months from the time of announcement-- and subsequently the Government has been steadily reducing its shareholding in a series of share sales, expected to be completed in 1993. 4. In what follows, we discuss each of these steps in some detail in order to bring out the remarkably innovative and effective strategy followed by the Mexican Government. Note that the context of the strategy is the motivation on the part of the Government to maximize sale proceeds, yet to move quickly, and at the same time to improve social welfare. - 46 - ANNEX I Page 2 of 11 Price-cum-tax change 5. Prior to 1990, there were heavy indirect (i.e. excise) taxes on Telmex services. As shown in Table A, tax rates had been rising throughout the 1980s and in 1989 were as high as 90 per cent. These high tax rates permitted the Government to extract substantial revenues from the company (equal to more than one-third of total revenues during the 1980s) without having to share with the private shareholders. (Recall that Telmex retained a substantial private shareholding throughout this period of around 45 per cent of the stock). 6. The price-tax reform of January 1990 contained two related steps: (a) all telephone taxes were abolished, leaving only the standard economy-wide VAT in place, but (b) all prices were raised (except international long-distance) by more than the nominal tax reduction, so that the consumer did not feel any relief from the abolition of the tax. The reform, therefore, amounted to a transformation of the indirect tax revenue into company profits (see Table B). In addition, prices of measured local calls were raised dramatically from 16 pesos per minute to 115 pesos per minute (both tax-inclusive) and, the number of free calls per subscriber was reduced; however, international calling prices were reduced. Thus the Government took the opportunity to begin the process of rebalancing rates according to costs, in keeping with the world-wide trend. Also, as is seen in Table C which presents the rates at constant pesos, rental charges and long-distance rates (domestic and international) were lower in real terms in 1990 after the reform than they had been in 1980. Local calls and installation charges, however, were substantially higher. 7. The price-tax reform was a key step in the divestiture of Telmex. Because it was so favorable to the company, it served as an important signal to buyers that the Government was abandoning its policy of "milking" the company, but instead was willing to see a highly profitable Telmex. Further analysis of this point, and of the Government's quid pro quo is contained below. Capital Restructuring 8. A major innovation in the Telmex sale was the capital restructuring of 1990. Because of the size of the company (1990 sales were about $3 billion; also, Telmex accounts for about 41% of the market capitalization of the Mexican stock exchange), there was a question about the ability of the capital market to furnish the requisite capital and absorb such a large sale. It was decided to sell only a part of the Government's holdings. The capital restructuring was carried out so that effective control of the company could be transferred with a minority shareholding. A detailed description follows. 9. Prior to privatization, Telmex had two classes of stock: AA shares were reserved for government ownership, A shares were unrestricted. Government first reduced the number of AA shares (from 55.9% to 51%) by converting some into A shares. Next, the company declared a stock dividend of 1.5 newly-created "L" - 47 - ANNEX I Page 3 of 11 sharesl, for each A or AA share held. The L shares had essentially no voting rights. Thus the AA and A shares (40 percent of the stock) collectively controlled the company. The divestiture of December 1990 then took place when Government sold all its AA shares (51% of the voting stock, 20.4% of the total stock) to the buying group. 10. The partial sale of Government's shareholding not only reduced the strain on the capital market, it also allowed some more time to elapse for investors to gain confidence in the Mexican economy and in the credibility of the Government's commitment to privatization. To the extent that the sale price of Telmex in 1990 was artificially low because investors were still trying to judge the sincerity of the Government, the partial sale allowed the Government to transfer control while allowing it to retain the majority of its shareholding for future sale at higher prices. Regulation 11. Because Telmex is a monopoly, its rates must be regulated in order to protect consumers. Government took the wise step of working out a detailed regulatory framework in advance of divestiture, so that potential bidders would be fully informed on the "rules of the game". Such a step obviously contributed to raising the potential sale price, since bidders were faced with less uncertainty about what they are buying. 12. The fundamental regulatory mechanism for Telmex follows the one introduced by the U.K. Government when British Telecom was sold--the so-called "RPI-X" formula or price-cap regulation. This places a cap on the amount by which the average price of a basket of services (including installation and rental charges and local and long-distance calls) can rise each year. The cap is the rate of price inflation (RPI) minus a factor "X" to reflect technological progress in the telecommunications industry. For the period 1991-96, X has been set at zero, and for 1997-98 at 3 per cent. Services where competition is present or permitted, such as cellular telephony, yellow pages and private circuits are unregulated. Also, the long-distance market will be opened to competition after August 1996, when the Telmex monopoly on that segment is set to expire. Telmex will retain a monopoly on local service. 13. In addition to the price regulation, the Telmex concession requires the company to meet certain system expansion and quality targets. These include: (a) The number of lines in service must expand at a minimum rate of 122 per year until 1994. ' The "L" is for "Limited" reflecting the fact that the L shares do not vote on normal matters, but have voting rights only on the following specific matterst (a) changes in the campaign's line of business; (b) merger with another company; and (c) withdrawal of the company from the Mexican stock exchange. Items b and c are particularly interesting; they present a barrier to the acquisition of Telmex by another company or by a private leveraged buyout. - 48 - ANNEX I Page 4 of 11 (b) All towns with population in excess of 500 must have telephone service by the end of 1994. (c) The number of public telephones must be increased from 0.8 per 1,000 population to 2 per 1,000 in 1994 and 5 per 1,000 in 1998. (d) In towns with automatic exchanges, the maximum waiting time for a new connection must be six months by 1995 and one month by 2000. (e) The speed of repairs and other quality indicators are to be improved at a specified rate. 14. As is apparent from the above list, the quantity targets set for Telmex are fairly ambitious, and require massive investment. The price-cap regulation, with its value of X set at zero until 1996, combined with the generous price-tax- reform of January 1990, provide the company with the potential to make huge profits with which to pay for the investments. It is estimated that the company will be able to finance 75 to 80 per cent of its investment requirements through internally-generated cash flow. Who therefore, pays the price? Obviously it is the consumer. Estimates are that consumers will pay an additional $16 billion in present value!/ as a result of the privatization price-tax changes. Sale Sequence 15. The divestiture of Telmex has been a complex transaction and, indeed, is not yet entirely over. Table D provides a chronology of the sales so far. The key transaction was of course the first one, in which the controlling shareholding was sold. In its call for bids, the Government explicitly invited Mexican investor groups and foreign companies with relevant expertise to bid. Bidders were pre-secreened to ensure (i) on the side of the Mexican bidders, that they had the financial strength and managerial expertise necessary, and (ii) on the side of the foreign bidders, that they had the relevant technical expertise. Thus, for example, foreign financial institutions were ruled ineligible to participate. At this stage the Mexican and foreign groups had not formed consortia. The formation of alliances was brokered by the privatization unit of the Government. The winning bid of $1.76 billion came from a consortium consisting of 51 per cent Mexican investors (led by Mr. Carlos Slim Helu through his companies Grupo Carso and Seguros), 24.5 per cent Southwestern Bell International Holdings, and 24.5 per cent France Telecom (the French Government- owned telephone monopoly). The bid was for all the AA shares (20.4 per cent of the stock) plus options to buy a further 5.1 per cent "L" shares. By looking at exactly how much each party paid and how many shares and options they acquired, it is possible to infer exactly the implicit share price and option price set by it See Pankaj Tandon with Manuel Abdala: "Case Study: Telmex," Chapter 16 of Ahmed Galal,... (see p. 55 for ref.). - 49 - ANNEX I Page 5 of 11 the buyers. The share price turns out to have been 76.69766 cents per share.! The implicit value to the whole of Telmex was therefore $8.132 billion. 16. Simultaneously with the divestiture sale, Government sold 467 million shares (4.4 per cent of the total stock) for $325 million (about 70 cents per share) to the workers' union. The purchase was funded by a loan from a Government bank, Nafin. At today's prices, these shares are worth about $1.35 billion! The profit works out to more than $20,000 per worker. Thus the sale of shares to the workers has worked greatly to their favor. Whether it motivates workers to put out greater effort will be seen over time. 17. The third major tranche sale took place in May 1991, when the Government decided to use the international capital market very explicitly by making a public offering of Telmex stock in all the major stock markets throughout the world. Eight hundred million shares (7.5 per cent of the company) were sold on the New York Stock Exchange, another 577 million in other countries and 138 million were sold in Mexico. A total of just under 16 per cent of the company's stock was sold for a total of $2.37 billion. The implicit value of Telmex was therefore $14.45 billion, a jump of 78 per cent in five monthsI The widening of exposure of the company to the entire world's investing public may have had a lot to do with the rapid rise in the value of the stock. 18. In September - October 1991, members of the original controlling group exercised their options to buy a further 5.1 per cent of the stock for $475 million. At this point, Government had reduced its holding from 56 per cent of the company to 10 per cent, and had raised about $5 billion in the process. Sale of an additional 5 per cent is planned for May 1992, with an estimated market value of nearly $1.5 billion. The market value of the company today is over $30 billion. 19. The key feature of the sale sequence has been the gradual sell-off of the Government's holdings. The sale in tranches has had the benefit of allowing the sale to proceed smoothly without in any way swamping the capital market, and has had the additional benefit of permitting the Government to participate in the spectacular rise in the market value of the firm. Since privatization, this value has increased almost four-fold. Much of this increase can be attributed to the greater confidence investors now have in Mexico; during the same period, the Bolsa index (of the Mexican stock market) has gone up two-and-a-half times. This is a reflection of the success of the Government's stabilization program. Thus the pure Telmex stock price increase is more of the order of 50 per cent. This in turn could be explained by greater confidence among investors in the sincerity and permanence of Mexico's privatization program, in other words, by a reduction in the subjective uncertainty in the investment community. Since this uncertainty relates to the question of the Government's commitment, it is rational for the Government to expect it to be resolved over time, and hence to anticipate the stock price increase. t This contrasts with the usually quoted figure of 81.256 cents per share, which incorrectly assigns a value of zero to the options. In fact the options were valued at 13.86936 cents per share. - 50 - ANNEX I Page 6 of 11 20. There remain two questions about the sale process. First, could the Government have obtained a better price for the company? According to calculations made elsewhere,!i the value of Telmex (assuming a very conservative 15 per cent discount rate) prior to divestiture was about $9.7 billion. The reason the market valued it so much lower was that Government was not permitting it to operate in the shareholder interest. This suggests that the price obtained for the company was a backward-looking price that was roughly correct, assuming no changes were made to the company. But the price-tax reform of January 1990 greatly increased the profit potential of Telmex; the increase in value is estimated to be about $15 billion from this change alone. From this point of view, the price obtained was too low. Although this argument can be criticized as being based on hindsight, the price-tax reform and its attendant revenue changes were so transparent that Government ought to have been able to perform this analysis. 21. Even had the Government known that a fair price for Telmex was $25 billion and not $8 billion, it is not necessarily obvious that it would be able to extract this price from the market. A first-price sealed bid auction (the sale procedure used) is supposed to yield at least the second-highest estimate of the perceived value. If investors perceived the sale as risky because of Government's past behavior toward the company, perhaps they could not be persuaded to pay anything closo to full price. Thus it could be argued that Government adopted the right strategy by selling in tranches so that the market could have time to resolve the uncertainties and Government could get close to full value for at least part of its holding. 22. The second question arises as a direct consequence of this argument. If Government's ultimate goal is social-welfare maximization, and if the welfare of all nationals is included in social welfare, then any gains that accrue to Mexican buyers from the purchase of Telmex would be included in social welfare. If these gains are very large rather than small, society is in some sense no worse off. However, if the gains accrue to foreigners, Mexico is worse off, because rents have dissipated abroad. In the case of Telmex, nearly 30 per cent of the company was sold to foreigners (out of a total of 46 per cent that the Government has sold). At present market prices, this block of shares is now worth some $8.9 billion, while the Government received $3.4 billion. Foreign investors have enjoyed a bonanza of $5.5 billion in a period of a little over a year. Had the Government confined the sale to Mexican investors, or sold only the first tranche of 10 per cent of the stock to the foreign members of the controlling group, the "leakage" of rents abroad could have been significantly reduced. The argument is made that the sale to foreigners, particularly of the second tranche, was absolutely essential to the process of confidence building and that the big run-up in the value of Telmex would not have occurred otherwise. Of course there is no way to settle the validity of these speculative arguments. The sale of the commercial banks was restricted to Mexican investors, and the Government appears to have realized extremely favorable prices, thus suggesting the sale of Telmex shares abroad was not essential. To the extent that the fact that Telmex was such a profitable investment for early investors may have 11 See Pankaj Tandon, op. cit. - 51 - ANNEX I Page 7 of 11 contributed to the high prices paid for the banks, the low price for Telmex could be seen as a "loss-leader" strategy. Summary and Lessons 23. The divestiture of Telmex has, on the whole, been very successful. Most important, the Government was able to smoothly sell the third-largest company in the country without any undue disruption of the capital market. It has put into place a competent management and established links with two technically sophisticated foreign partners. A comprehensive regulatory environment has been created. A total of about $5 billion were raised, thereby making a substantial contribution to the Government's stabilization program. Some of the keys to the successful sale are as follows: (a) The sale was made in tranches to minimize capital market disruption and to allow Government to benefit from a rising stock price. (b) A capital stock restructuring was carried out to enable transfer of control with only a minority shareholding. (c) Workers were sold shares at a slightly discounted price to encourage their participation and co-operation. (d) A comprehensive regulatory regime was put in place and took effect simultaneously with divestiture. (e) During the sale, administrative control of the company was transferred to the privatization unit, in order to ensure efficiency of decision-making and to pre-empt the possibility of conflict between the concerned secretariats. There were some negatives: (a) Because of the desire to accomplish the sale quickly, Government may have accepted a rather low price. The subsequent quadrupling of the stock price in just over a year supports this argument. (b) A large fraction of the sale was to foreign investors. By itself, this is not undesirable, but because the price was low, large rents have dissipated abroad in the process. (c) Consumers will bear the big costs from divestiture because of the large price increases sanctioned just prior to divestiture. Although the price increases brought most prices only to around international levels, the installation charges are extremely high (at around $300 per line). Over time, it is expected consumers will get additional benefits from improved quality and a more rapid expansion of the network. - 52 - ANNEX I Page 8 of 11 24. On balance, the benefits to Mexico are tempered by the large dissipation of rents abroad. To put it into perspective, Government has so far rpoeived less in total (about $5 billion) than the new foreign shareholders have gained in value (about $5.5 billion). This may have been a rather steep price to pay for the ostensible gains in improved confidence in the Mexican economy. TABLE A: INDIRECT TAX RATES ON TELMEX SERVICES (in per cent) Year Telephone Tax VAT Total Tax Local National Inter- Local National Inter- Service Long national Service Long national Distance Distance 1980 54.22 30.90 30.90 10.00 69.64 43.99 43.99 1981 54.22 30.90 30.90 10.00 69.64 43.99 43.99 1982 54.22 30.90 30.90 10.00 69.64 43.99 43.99 1983 54.22 30.90 30.90 15.00 77.35 50.53 50.53 1984 54.22 30.90 30.90 15.00 77.35 50.53 50.33 1985 54.22 30.90 30.90 15.00 77.35 50.53 50.33 1986 65.63 37.23 30.90 15.00 90.48 57.82 50.53 1987 65.63 37.23 30.90 15.00 90.48 57.82 50.53 1988 65.63 37.23 22.00 15.00 90.48 57.82 40.30 1989 65.63 37.23 22.00 15.00 90.48 57.82 40.30 1990 0.00 0.00 0.00 15.00 15.00 15.00 15.00 Source: Telmex ANNEX I Page 9 of 11 TABLE B: TELIIEX OUTPUT PRICES (AT CURRENT PESOS) 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Rent (Annual) Residential 576 696 780 1392 1392 3132 4872 13428 50640 50640 85080 Tax 369 446 499 995 995 2238 4092 11280 42538 42538 12762 Total 945 1142 1279 2387 2387 5370 8964 24708 83178 93178 97842 Business 1044 1248 1500 2676 2676 6012 9372 25824 97356 146040 251276 Tax 793 948 1140 2248 2248 5050 9166 25256 95214 142827 37691 Total 1837 2196 2640 4924 4924 11062 18538 51080 192570 288867 288967 Ieasured Service (per minute) Residential 0.12 0.12 0.17 0.29 0.54 0.70 1.13 3.04 8.69 8.69 100.00 Tax 0.08 0.08 0.11 0.21 0.39 0.50 0.95 2.55 7.30 7.30 15.00 Total 0.20 0.20 0.28 0.50 0.93 1.20 2.08 5.59 15.99 15.99 115.00 Business 0.12 0.12 0.17 0.29 0.54 0.70 1.13 8.69 8.69 13.04 100.00 Tax 0.09 0.09 0.3 0.24 0.45 0.59 1.11 8.50 8.50 12.75 15.00 Total 0.21 0.21 0.30 0.53 0.99 1.29 2.24 17.19 17.19 25.79 115.00 Installation charges Residential 1251 5027 6489 13578 13578 27844 43374 147488 889094 889094 900000 Business 1558 7587 8438 17658 17658 36832 57269 195195 1459028 1459028 1560000 Domestic Long Distance (per minute) Residential 5 6 8 14 21 27 46 125 275 420 841 Tax 2 2 3 6 10 12 24 65 142 216 126 Total 7 8 11 20 31 39 70 190 417 638 967 Business 5 6 8 14 21 27 46 125 275 420 841 Tax 2 3 4 8 12 15 29 79 174 266 126 Total 7 9 12 22 33 42 75 204 449 686 967 International Long Distance (U.S. outbound), net of settlement Residential 15 19 44 93 133 200 466 1053 1804 2017 1781 Tax 6 7 17 42 60 91 211 478 727 813 267 Total 21 26 61 135 193 291 677 1531 2531 2830 2048 Business 15 19 44 93 133 200 466 1053 1804 2017 1781 Tax 7 9 21 51 73 111 257 582 727 813 267 Total 22 28 65 144 206 311 723 1635 2531 2830 2048 ANNEX I Page 10 of 11 TABLE C: TELMEX OUTPUT PRICES (AT CURRENT PESOS) 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Rent (Annual) Residential 14284 13411 7560 7462 4687 6443 4872 5181 12879 10759 14267 Tax 9143 8584 4839 5333 3350 4604 4092 4352 10818 9038 2140 Total 23427 21995 12399 12795 8037 11047 8964 9532 23697 19797 16407 Business 25889 24047 14538 14345 9011 12367 9372 9963 24760 31028 42137 Tax 19676 18275 11049 22050 7569 10388 9166 9744 24215 30346 6320 Total A5565 42322 25588 26396 16580 22755 18538 19707 48974 G1374 48457 Measured Service (per minute) Residential 2.98 2.31 1.65 1.55 1.82 1.44 1.13 1.17 2.21 1.85 16.71 Tax 1.90 1.48 1.05 1.11 1.30 1.03 0.95 0.99 1.86 1.55 2.52 Total 4.88 3.79 2.70 2.67 3.12 2.47 2.08 2.16 4.07 3.40 19.28 Business 2.98 2.31 1.65 1.55 1.82 1.44 1.13 1.17 2.21 2.77 16.77 Tax 2.26 1.76 1.25 1.31 1.53 1.21 1.11 1.15 2.16 2.71 2.52 Total 5.24 4.07 2.90 2.86 3.35 2.55 2.24 2.32 4.37 5.48 19.28 Installation charges Residential 31022 96861 62893 72788 45721 57275 43374 56901 226114 188900 150922 Business 38636 146188 81783 94660 59460 75763 57269 75307 371059 309991 261597 Domestic Long Distance (per minute) Residential 124 116 78 75 71 56 46 48 70 89 141 Tax 48 45 30 34 32 25 24 25 36 46 21 Total 172 161 108 109 103 81 70 73 106 135 162 Business 124 116 78 75 71 56 46 48 70 89 141 Tax 60 56 38 41 39 31 29 31 44 56 21 Total 184 172 115 117 110 86 75 79 114 146 162 International Long Distance (U.S. outbound), net of settlement Residential 372 366 426 499 448 411 466 406 459 429 299 Tax 145 143 166 226 203 167 211 184 185 173 45 Total 517 509 593 725 651 598 677 591 644 601 343 Business 372 366 426 499 448 41 466 406 459 429 299 Tax IO 178 207 275 247 227 257 224 185 173 45 Total 552 544 633 774 695 639 723 31 644 601 343 - 55 - ANNEX I Page 11 of 11 TABLE D: TELMEX SHARE SALES CHRONOLOGY Date # of Shares Sold To Whom Sold Receipts in US$ 12/20190 1,103,150,896 Grupo Carso, et. al. 859,112,200* 1,059,890,076 SW Bell, France Tel. 898,489,580* 12/20/90 466,000,000- Employees 324,953,222 12/20/90 7,104,105 Mex. investors-options 6,266,536 533,656,138 SW Bell, options 465,773,414 France Tel., options 3,099,818 5/15/91 1,377,000,000 Foreign oft ing 1,876,162,500 138,000,000 Mexican offering 150,000,000 Telmex 1,876,162,500 6/16/91 80,000,000 Foreign offer 109,000,000 overallotments Total to date 4,914,801,215 4,925,744,810k * Includes price paid for options. This number is approximate. * This total is in nominal dollars. - 57 - ANNEX II Page 1 of 7 Mexico: Privatization of Airlines 1. Mexico has divested both its previously parastatal airlines, and although only one sale (that of Mexicana Airlines) was carried out during the PERL, it is instructive to examine both as they provide a study in contrasting divestiture process and outcome. Mexicana Airlines: Background and Sale Process 2. Mexicana Airlines, formed in 1921, is the oldest airline in Latin America and the fourth oldest in the world. For much of its history (1926-67) it was owned by Pan American, and it was acquired by the Mexican Government only in 1.992 during the Mexican economic crisis when it was unable to meet its debt obligations. As in the case of Telmex, the Government did not acquire all the shares of the company, but only 51%, leaving the remainder for the investing public. 3. The Government decided to divest the company in 1986, but had trouble receiving adequate bids. Bidders cited the need for an overhaul of the regulatory structure with deregulation of most fares, for agreement with the labor unions on work force reductions, and for negotiations on the restruc- turing of the company's debt. Some bidders even wanted prior approval for a merger with Aeromexico, which would have essentially created a monopoly in the domestic airline market. Government did not yield on any of these issues. 4. Finally, in 1989, a new approach to divestiture was tried in which investors were to be given an equity stake in the company for exchange for a capital infusion. Thus, the Government itself received no immediate cash during the divestiture of Mexicana. Rather, the money paid by the buyers went into the company's treasury to help ease its cash flow problems. As in the case of Telmex, the winning bid came from a consortium including both Mexican and foreign investors. As constrained by law, the foreign bidders owned 49% of the consortium (35.7% Chase Manhattan Bank, 7.1% Drexel Burnham Lambert, 6.1% Sir James Goldsmith), while the Mexican investors (led by Israel and Pablo Brener and Carlos Abedrop) owned the remaining 51%. The consortium, called Grupo Falcon, paid $140 million for 20% of the equity of the company, thus placing an implicit value of $700 million on the whole company. As a result of this transaction, Government's owneship of the company was diluted from 51% to just over 40%, and the private shareholders suffered a similar dilution of their stock. The deal was completed in September 1989. 5. The divestiture of Mexicana had a few additional features worth noting. First, although Falcon received only 20% of the shares in return for its equity infusion, it was required under the agreement to acquire an additional 5% of the company from the stock market. Then, matching Falcon's 25% share, Government placed another 25% of the company's stock in a trust over which Falcon had voting control, thereby giving them assured control over the company. Second, Falcon had an option to buy the shares in the trust for about $115 million (based on the same price per share -- $2.47 -- as used for the capital infusion) plus interest any time before September 1992. Third, Falcon promised an ambitious J- . tment program to modernize and expand Mexicana's - 58 - ANNEX II Page 2 of 7 fleet. All these features played an important role in the outcome of the Mexicana divestiture, as we will see below. Aeromexico: Background and Sale Process 6. Aeromexico presents a background similar in some respects to that of Mexicana. It is a relatively old airline, founded in 1934, and also was part- owned for a long time (1940-59) by Pan American. It, too, was acquired by the Government at a time of crisis. The year, however, was earlier -- 1959 -- and the cause of the crisis was internal, not external, being triggered by a pilots' strike. In a sharp contrast to the Mexicana case, the Government acquired all of the shares of the company. Ironically, it was another employee strike that led to the company's divestiture. 7. Traditionally, Aeromexico was more oriented towards the domestic market, as compared to Mexicans's more international stance. In 1987, for example, Aeromexico earned 52% of its revenues from domestic and 40% from international passengers; the corresponding numbers for Mexicans were 33% domestic and 4b. international. Because domestic flights were far less profitable (because of the regulatory environment -- see below), Aeromexico performed more poorly. It was a consistent loss-maker; details can be seen in Table F..1 Thus, the company was a drain on the budget (receiving subsidies of 35 billion pesos in 1986 and 96 billion pesos in 1987) and did not have the financial strength to surmount major dislocation. 8. In April 1988, Aeromexico's workers went on strike, demanding higher pay and better working conditions. The Government decided to seize the opportunity to divest the company. Three days afte- the strike began, they put the company into bankruptcy. Under Mexican law, workers' jobs are protected, but this protection ceases once their company is bankrupt. All the workers were fired and paid their statutory compensation of pay equal to three months plus twelve days pay for each year of service. The airline was temporarily grounded but resumed skeleton service shortly thereafter. 9. Technically, the corporate entity that owned Aeromexico -- officially known as Aeronaves de Mexico, S.A. de C.V. -- has never been divested, but rather is still in bankruptcy proceedings. Instead, Government created a new corporation -- called Aerovias de Mexico, S.A. de C.V. -- and placed substan- tially all the assets of Aeronaves into it. Note that substantially all the liabilities of Aeronaves were retained, and are s,ill in the process of being liquidated. The new corporation, therefore, began operations with a clean balance sheet and, perhaps more important, no labor contracts. It was free to hire only the workers it needed. 10. Aerovias was sold in November 1988. The purchase price for the assets was 656 billion pesos ($287 million). The buying group received 75% of the company's stock. This would imply a total value of the company of around SIt might be speculated that the fact that Aeromexico was fully state-owned while Mexicana was still 49% privately owned (some private investors sat on the Mexicana Board) may also partly account for Mexicana's slightly better profit performance. - 59 - ANNEX II Page 3 of 7 $383 million. The remaining 25% of the shares were owned by ASPA, the pilot.' union.!' The buying group consists of a consortium of Mexican investors, led by one Mr. Gerardo de Prevoisin Legorreta, and including a substantial investment by Bancomer, a commercial bank. Note that, as in the case of Mexicana, the Government has not received any of the proceeds of the divesti- ture. Rather, the money has been paid to the trustees of the bankrupt Aeronaves de Mexico, and it is estimated that all of it will be used up in the process of liquidation of the liabilities of that company. Regulatory Framework in the Mexican Airline Industry 11. A key feature of the divestiture of Mexico's airlines is the regulatory environment they face. Fares on international flights are governed, as they are throughout the world, by bilateral and international agreements. Thus, Mexico does not have much flexibility on this front. But domestic fares are not subject to these agreements. In Mexico, domestic fares were regulated until 1991, well after the divestiture of the airlines. 12. Prior to divestiture, all domestic air operations were tightly regulated, including frequency of flights, schedules and fares. Domestic fares were kept extremely low, although the purpose of this is not entirely clear. Table E shows a sample of 1989 Mexican and U.S. fares, illustrating how much lower Mexican fares were. Although it could be argued that Mexican costs are lower (indeed they are), a principal cost component -- jet fuel -- is no cheaper in Mexico. Fares were simply not adequate to cover costs, and this accounts at least partly for the poor performance of the airlines, particularly Aeromexico. Table E: SAMPLE FARES, U.S. AND MEXICO, 1989 Route Fare Distance Fare per Mile (US$) (miles) (US$) Boston-Newark 178 206 864 Mexico City-Acapulco 52 191 273 New York-Cleveland 296 417 710 Guadalajara-Monterrey 75 410 182 Seattle-Los Angeles 472 1,133 417 Mexicali-Guadalajara 142 1,113 128 11 However, the transaction was fairly complicated. Tandon (1992) has estimated that the actual price paid by the buyers was about 340 billion pesos, implying a value for the firm of about 450 billion pesos ($200 million). - 60 - ANNEX II Page 4 of 7 Source: Pankaj Tandon with Inder Ruprah: "Case Studies: Aeromexico and Mexicana de Aviacion*9 chs. 17 and 18 of Ahmed Galal, Leroy Jones, Pankaj Tandon and Ingo Vogelsang: Welfare Consequences of Selling Public Enterprises, World Bank, 1992. 13. Government made no regulatory changes prior to or simultaneous with divestiture. This is due in part to the suddenness of the Aeromexico sale; there was no long preparatory period. However, Government had decided to sell Mexicana already and therefore it is not unreasonable to argue that the authorities ought to have been in the process of creating a post-divestiture regulatory framework. 14. Regulatory change came in July 1991. There were two forms of liberalization -- in entry conditions and in fares. As regards entry, new airlines are permitted to enter, and -- more important -- airlines are permitted to add new routes or flights freely. As regards fares, they are free on any routes where more than one airline operates. Only on routes where there is only one carried are fares regulated. Thus, this has been a substantial deregulation. Performance of the Airlines Since Divestiture 15. Both airlines have struggled since divestiture. Income statements are presented in Tables F and G. They show that Aeromexico turned in two break-even years in 1989 and 1990, but went into the red in 1991. Mexicana has been unprofitable in both 1990 and 1991. The poor performance is reflected in the stock market valuations of the two companies. Mexicana is presently valued by the market at about $260 million, considerably lower than the implicit $700 million value at the time of divestiture. Aeromexico has done better, being valued now at about $285 million. There are therefore two key questions: (i) Why have the airlines done so poorly, and (ii) Why has Mexicana done so much more poorly than Aeromexico? - 61 - ANNEX II Page 5 of 7 Table P: AEROMEXICO - SU0RY PROFIT AND LOSS STATEMET (Billions of Pesos) 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 "et Sales 1? 19 48 74.60 119.61 246 640 491.8 981.1 1,624.1 2,065.1 Cost of Sales 8 15 33 34.60 103.37 209 520 296.8 831.3 1.333.0 1,984.4 Oross Profit 4 4 15 20.01 16.25 37.02 120 195.0 149.8 291.1 80.7 ådmin./Selling/ 3.36 6.11 14.27 20.96 34.14 72.00 193.08 234.7 154.1 296.6 232.4 Diet. Expense Financial Expenes 0.39 0.91 2.16 2.37 1.64 0.81 19.75 18.4 -12.ý 4.7 36.5 Other Oprating -0.36 0.60 0.40 1.78 0.70 3.12 4 5.4 - - - Exenss Nerating Profit 0 .4 -1.59 -5.10 -20.23 -38.91 .97 -63.5 8.1 -10.2 -188.3 Other Incowe -164.5 2.2 23.4 57.4 Subsidies . . . . - 8 0.0 - - . :on-Operating . - - - - • - 0.2 - - . Exense. Profit Befors Tax 0.22 -3.60 .1.59 -5.10 -20.23 -38.91 -96.99 -228.2 10.3 13.2 .130.8 Provisions for . • • - - - 0.0 1.5 4.2 11.0 0.0 Taxes Profit After Tax 0.22 .3.60 -1.59 .5.10 .20.23 -38.91 -96.99 .229.7 6.1 2.1 -130.8 Sourcet Company reporte (1991 stimated on basis of 8-month company ddta). Table G& CORPOEACION HEXICANA DE AVIACION * SR0(ARY PROFIT AND LOSS STATEHT (Billion@ of Pson) 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 Net Sales 15 22 61 96 162 334 867 1,703 1,881 2,5.5 2,669 Coet of Sales 13 24 53 83 157 349 855 1,538 1.856 2,714 2,941 Groas Profit 2 .1 8 13 5 -15 12 165 26 -189 -272 Aduin. ISellingIDist. - - - - . . . . . . . Expense-- - - - - - -- Financial Expencce 1 2 7.49 9 12 26.89 94.50 136.50 27.29 58.21 120.99 Other Operating - - - - - . . . . - . Expenses Operating Profit 1 .3 0.78 4.05 -6.75 -41 -82 28.17 -1.72 .247.0 .392.8 Other Incon. 0.92 0.21 0.02 0.00 0.04 1.31 0.47 20.26 21.84 -7.90 -4.91 Subsidies . . . . - . . . Non-Operating - - - 0.05 0.15 - - - - Exense Profit Befors Tax 1.50 -2.67 0.80 4.00 .6.87 -40.13 -81.59 48.43 20.12 -254.9 -397.7 Provision* for Taes 0.00 0.00 0.00 0.00 0.04 0.00 2.21 91.05 29.01 8.20 10.01 Profit After Tax 1.50 .2.67 0.80 4.00 -6.91 -40.13 -83.81 -42.63 -.89 -263.1 -407.7 Source Company reports. - 62 - ANNEX II Page 6 of 7 16. Part of the explanation for the poor performance lies in the world- wide weakness in the air travel industry. In 1989, after the Iraqi invasion of Kuwait, oil prices jumped. Then in 1990, although oil prices eventually came down, air travel suffered a severe slowdown in connection with the Persian Gulf War. Airlines throughout the world have had some difficult times. 17. The situation was perhaps exacerbated in Mexico by regulation. As argued earlier, domestic fares were held at low levels. It might have been expected that, once they were deregulated, the prices would rise to more economic levels. Surprisingly, however, this has not happened. Rather, the industry has entered into a period of intense competition, with each airline "invading" the other's major routes and fares being slashed. Thus, the process seems to mirror the experience of the U.S. air market after deregulation. There was initially a period of interse competition as airlines sought to expand aggressively or simply tc meet the competition by cutting fares. That situation, however, was not sustainable. It was followed by a period of consolidation, characterized by bankruptcies, mergers, and eventually higher fares. Thus, we might expect history to repeat itself in Mexico. It is not clear that the Mexican market can sustain two airlines, and it could perhaps be argued that a more careful approach to coordinated divestiture and deregulation might have prevented the turbulence that we now see. Ly itself this process is not necessarily undesirable, but if one of the airlines were to go bankrupt, there may be negative repercussions in the Mexican capital market. 18. As regards the differential performance of Mexicana and Aeromexico, there are two sets of factors that explain this. The first set of factors has to do with the method of divestiture. Aeromexico was declared bankrupt and liquidated. As a result, the new airline was able to re-establish itself on a sounder footing. The most important change that became possible was the sharp cutback in staff. From 11,500 employees prior to divestiture, Aeromexico has cut back to under 6,000 at the end of 1991, while output is at about the same level. Labor productivity in 1991 was 91% higher than in 1987, and total factor productivity was 11% higher. By contrast, Mexicana inherited the entire work force of its parastatal forebear and has been unable to realize economics of the same magnitude. Labor productivity has risen 21% during 1988-91, but total factor productivity is in fact 2% lower than it was in 1988. The drastic restructuring of Aeromexico obviously paid off for the company. 19. The second set cf factors accounting for the differential performance relate to strategic decisions within the firms. There are two important differences in the strategies followed by the two companies. First, while Mexicana decided to concentrate its efforts on th) tourist trade (perhaps partly because the majority shareholders have prior interests in resort hotels), Aeromexico focused on business travellers. As Mexico has boomed, while tourism has declined worldwide, this difference clearly favored Aeromexico. Second, while Mexicana embarked on a major fleet modernization through acquisition of 22 new Airbus jetliners, Aeromexico decided to lease its fleet and presently owns only two aircraft. With the worldwide slump in the air industry, this again has been very advantageous for Aeromexico as they have obtained favorable terms on their leases, while Mexicans has been unable to sell aircraft that it is replacing by its new ones. In addition to these two sets of factors, observers of the two airlines point to many minor examples of - 63 - ANNEX II Page 7 of 7 managerial decisions that worked in the same direction -- poor outcomes at Mexicana and good outcomes at Aeromexico. 20. Now it is natural to expect that some divested companies will perform well and some not so well, just as is the case with private companies. However, poor managerial performance in the private sector can sometimes be pre-empted by takeover. One of the important ostensible advantages of private enterprises is that the capital market controls them through the threat of takeover, and that poorly run companies can actually be taken over to improve performance. In the case of Mexicana, however, the capital market '&as been unable to fulfill this function because the present management has solid voting control over the company, through its 25% ownership and voting authority over 25% more of Government's stock. Thus, the strategy of giving the buying group a dominant control is a double-edged sword: while this gives management leeway and a strong incentive to perform well, it also prevents the capital market irom performing its control function in the event that managerial performance is inadequate. It should be mentioned here that the Mexican Government placed a time limit on its award of voting rights to the buying group. By the end of this year, it will be able to reassess its sale strategy for Mexicana. Summary and Lessons 21. Airline divestiture in Mexico has been something of a mixed bag. Although both airlines have successfully re-entered the private sector, their performance since divestiture has not been outstanding. This reflects partly the difficulties airlines worldwide are facing, but also partly the lack of a well thought out, coordinated plan of deregulation and divestiture. There are several important lessons to be learned from this experience. (1) Deregulation and divestiture, particularly in non-tradeable goods industries, require coordination and careful planning. (2) Ideally, the regulatory environment should be created and put in place prior to, or simultaneously with, divestiture. (3) Major overhaul of an enterprise prior to divestiture can be very beneficial. Particularly, resolution of an over-staffing problem is best achieved before an enterprise is sold. (4) Granting of majority control to a buying group can prevent the capital market from exercising its proper control function on enterprises. - 65 - PROJECT COMPLETION REPORT MEXICO PUBLIC ENTERPRISE REFORM LOAN (Loan 3086-ME) June 12, 1991 Country Operations Division I Country Department II Latin America and Caribbean Regional office - 67 - PROJECT COMPLETION REPORT K01llCO PUBLIC ENTERPRISE REFORM LOAN (Loan 3086-ME) PART I PROJECT REVIEW FROM THE BANK'S PERSPECTIVE A. PROJECT IDENTITY Project Name: Public Enterprise Reform Loan (PERL) Loan Number : 3086-ME RVP Unit Latin America and Caribbean Country Mexico Sector Public Enterprises B. BACKGROUND Macroeconomic Setting 1.01 Following three decades of high growth, low inflation and moderate internal debt accumulation, Mexico went through a period of high spending in the late 1970s in an attempt to stimulate growth through an expansion of public involvement in the economy and the encouragement of high-cost, capital-intensive investments. This stvategy was made possible by sharply rising oil export earnings coupled with large foreign borrowing, but failed to sustain sound economic growth. It was put to an end in 1982 when world interest rates rose substantially, while oil prices fell. The resulting shortage of foreign exchange to meet debt service obligations and the refusal of external creditors to rollover Mexico's short-term debt left no option but fiscal retrenchment. 1.02 An IMF-supported stabilization program was launched in 1983 and succeeded in arresting the deteriorating trend. These pisitive results were, hoever, reversed by some relaxation of fiscal and monetary discipline in 1984-85. Moreover, the earthquake in late 1985 and major terms of trade deterioration in 1986 due to falling oil prices worsened Mexico's prospects. In response, the Government adopted a deeper and growth-oriented stabilization program designed to eliminate structural rigidities in the economy. The most significant elements included a major reorientation toward fuller integration with the world economy, the disengagement of the state and redefinition of the role of the public sector in the economy. In return for renewed monetary and fiscal austerity, the program also called for a concerted financial effort on the part of Mexico's creditors. 1.03 Performance under the program has been impressive in many areas. The trade reforms were far-reaching and have gone beyond the Government's initial plans, commitments under the GATT and targets of - 68 - Bank-supported trade policy loans (Ln 2745-ME in FY87 and Ln 2882-ME in FY88). Fiscal austerity was strictly maintained in the face of adverse terms of trade developments and Mexico's continued difficulties in accessing the international capital markets. However, there was no real growth, and hence a severe decline in per capita income from 1982 to 1987. Also, inflation, rather than slowing down, in fact accelerated, partially in response to sharp nominal devaluations in 1986 and 1987. 1.04 In December 1987, the novernment responded with the Economic Solidarity Pact (PACTO), a concerted effort between Government, labor and business to bring down inflation that was running well into triple digits by end-1987. The program consisted of further tightening of fiscal and monetary policy and renewed structural efforts in the areas of trade liberalization, credit and public enterprise divestiture. These measures were supplemented by a freeze of minimum wages, public sector prices and tariffs, and what was a corner stone of the PACTO: the freezing of the nominal exchange rate against the US dollar. 1.05 On almost every aspect, performance under the PACTO has been exemplary and often went beyond what was initially planned. The fiscal deficit was brought closely in line with the low inflation targets embedded in the program. This effort was all the more noteworthy, given the negative budgetary impact of further drops in oil prices and increasingly high real interest rates on foreign debt towards the end of 1988. The policy of a fixed Peso-Dollar rate, however, came under increasing pressure and, by the end of 1988, was abandoned. 1.06 In December 1988, the Mexican Government and the various sectors of society agreed on a renewal of the PACTO to consolidate price stability and further improve the budget situation. Under this renewed PACTO -- now called Economic StaDilization and Growth Pact (PECE) -- public sector prices and tariffs, as well as minimum wages, would remain controlled after initial adjustments, but the peso would be depreciated daily, according to a pre-announced schedule of one peso per day. The PECE has been extended several times and is still in force. World Bank and IMP's Roles in Mexico's Adiustment Process 1.07 once it became evident that Mexico's economic crisis would not be transitory and that the Government began looking for ways to address its structural development problems, the role of the Bank in Mexico was bound to change. In 1984, the Bank proposed to the Mexican Government a change in the direction of its lending strategy. A major discussion took place within the Mexican administration on the direction and importance of the Bank role in Mexico's adjustment process. In the end, the authorities agreed that they wanted the Bank to provide quick disbursing policy-based loans to Mexico and to assist them in their negotiations with the commercial banks to obtain rescheduling and new money. The Bank formulated a program consisting of a series of sector loans, which taken together included the major elements of a SAL program. The two Export Development Loans (FY83 and FY87 for USS 250 million each) and-the two Trade Policy Loans (FY87 and FY88 for US$ 500 - 69 - million each) were the first such operations and were followed by adjustment loans in the agriculture (FY88; US$ 400 million), steel (FY88; US$ 400 million), fertilizer (FY88; US$ 265 million) and roads and telecommunications (FY90; US$ 300 million) sectors. The project described in this report is part of a second series of adjustment loans designed to address important policy issues and the pervasive problem of continued low growth of the economy. It was prepared in parallel with the Financial Sector Adjustment Loan (FY89; US$ 500 million) and the Industrial Sector Policy Loan (FY89; US$ 500 million). Together, they form a policy package aimed at providing the framework for an increase in private investment necessary for resumed growth. 1.08 Since the start of its adjustment lending in Mexico, the Bank has collaborated closely with the IMF. Most of the Bank's sectoral adjustment lending was carried out concurrently with IMF programs. Since 1982, Mexico has benefited from an extended arrangement for SOR 3.4 billion, a special emergency drawing for SDR 291 million after the 1985 earthquake, a stand-by for SDR 1.4 billion in 1986 and, in 1989, a drawing of SDR 453 million under the compensatory facility, and another extended arrangement of SDR 2.8 billion through 1992. Mexico's performance under the on-going EFF arrangement has been excellent. The Public Enterprise Sector and the Government Early Reform Efforts 1.09 During the 1970s, the PE sector grew rapidly without a clear definition of the roles of the Government and the private sector. In the early 1980s, the sector had become very large, contributing about 12% of GDP and employing about one million people. Public enterprises controlled the most important branches of basic industry and infrastructure. There were 1,214 PEs in 1982, while only 491 existed in 1970. A major source of growth was the acquisition by the Government of bankrupt private firms to avoid their closure and maintain employment. 1.10 The economic and financial performance of PEs was generally poor. Efficient administration was hampered by the fact that prevalent PE legislation at that time focused on processes and controls that relied heavily on the administrative capacity of the Government and was not suited for industrial and commercial enterprises. PE efficiency was also affected by Government price controls which kept PE prices and tariffs generally below their economic opportunity costs. 1.11 In 1982, as part of its structural adjustment strategy, the Mexican Government began implementing a major disengagement program for most of the small and medium PEs and for some of the larger ones. By February 1989, (i.e., just before negotiations of the PERL and finalization of the reform program to be supported under that operation), this process left only 449 entities in the public sector, of which 90 were in the industrial, energy and mining sectors. Excluding the oil and electricity companies, which, according to the Mexican Constitution, must remain with the Government, the enterprises disengaged since 1982 in the industrial, energy and mining sectors represented 77% of the number of enterprises, 66% of labor, and 41% of - 70 - sales, but only 23% of the asset value of the initial universe of PEs in these sectors in 1982. 1.12 While implementing its disengagement program, the Government also took measures to improve the competitive environment for industrial and commercial activities. It thus carried out a major liberalization of the trade regime, the most important elements of which were the following: (i) elimination of most QRs and other NTBs on imports; (ii) reduction in the level of import tariffs and in the number of rates from a range of 0-100% with 12 different rates to a range of 10%-20% with only 5 rates; and (iii) elimination of export taxes and abolition of export licences for 332 tariff positions, i.e., more than half the tariff positions under control in 1980. Mexico's trade reforms have been far reaching and have been supported by Trade Policy Loans I and II. Under the Industrial Sector Policy Loan prepared in parallel with this project, the Government also undertook the deregulation of the automotive, pharmaceutical, petrochemicals and micro-computer sectors, the easing of bureaucratic procedures and barriers of entry, increasing of factor mobility (e.g., direct foreign investment, technology transfer, research and development, and labor), and the improvement of trucking services. Finally, the Government reduced substantially the scope of price controls, making them more flexible for most products. Still, a significant number of goods and services produced by large PEs remained under price controls. 1.13 By early 1989, it was clear that although the disengagement program had helped reduce the cost of PEs to the budget, many enterprises were still facing low capacity utilization, high operating costs and substantial financial losses. They continued to receive large subsidies from the Government totalling 2.8% and 2.1% of GDP in 1987 and 1988, respectively. Despite the improvements in the environment introduced by the Government since the early 1980s, the performance of the public enterprise sector remained adversely affected by two sets of issues. First, pricing policies for a few large PEs remained inadequate, forcing them to operate at a loss. In fertilizers, the controlled domestic price was on average about 60% of the import price, while in steel, controlled prices of about 100 products were 20% to 30% lower than international prices. This situation impeded import competition, while exerting a great financial burden on the companies, making them dependent on input subsidies and budgetary transfers from government. Second, the managerial and financial autonomy of the PEs was still very limited due to the over.entralized structure of budgeting and control. Major budgeting, personnel, salary, investment, procurement and even some operational decisions of PEs were made at the Ministerial level, often with a short-term cost-containment focus and without sufficient attention to the companies particular investment and operating requirements. In 1986, as part of the program, the Government enacted the Law of Parastatal Entities, a major step forward to create the legal conditions for a more appropriate balance between PE's autonomy and accountability. However, the detailed interpretative regulations to this law, which are essential for its implementation, were not issued. Without these regulations, the PE Law was not - 71 - operative and PEs remained subject to the web of legal, regulatory and administrative systems that deprived them of managerial and financial autonomy and negated their acco-antability. Reform of PEs had also been hindered by the inertia characteristrE of large public organizations and the political sensitivity of many of the issues. As a result of these constraints, substantive reform of the PE sector during the period 1982- early 89 had been largely confined to the disengagement program. C. PROJECT OBJECTIVES AND DESCRIPTION 1.14. The project was part of a program of Bank support to the Mexican Government's efforts to stabilize the economy and resume growth. It aimed at improving the efficiency of public enterprises whose poor performance accounted for much of Mexico's financial problems and at reducing the heavy burden that they imposed on the Government budget. The project supported a PE reform program which included: (i) reforms in the policy and institutional environment for PEs to improve the efficiency of the enterprises to be retained by the Government; and (ii) a disengagement component to support the sale, liquidation and merger of PEs. The project obtained cofinancing from the EXIMBANK of Japan in an amount of US$300 million equivalent. The Disengagement Proaram 1.15 Situation Before PERL. Since 1982, the Government had disengaged itself from 706 enterprises (61% of the sector in number) through liquidation, sale, merger, or transfer to municipalities, so that by early 1989, the PE sector only included 449 enterprises. The disengagement program had been particularly important in the industrial, energy and mining sectors where the number of PEs was reduced from 398 to 90. During this first phase, most disengaged enterprises were small and medium sized PES, but valuable experience and knowledge have been accumulated that were very useful in subsequent, more complex transactions. The disengagement process, which was a loosely defined system, decentralized among sectors ministries, was tightened up and standardized procedures established. The Disincorporation Unit at SHCP was also strengthened and allocated budgetary resources for administering and monitoring the PE sales program. In complex cases or large transactions, the Unit had the authority to establish an ad hoe committee that could draw as needed on other key Ministries that were to be involved in the decision-making process. 1.16 Disengagement under the Proiect. To continue with the disengagement program, during preparation of this operation, all sectoral ministries submitted (i) a list of PEs to be disengaged under the loan and (ii) t%e justification for the permanence of the remaining PEs in the public sector. Based on this preparatory work, a disengagement program involving 199 PEs was established. Of these, sixty-three were to be prepared for eventual sale, 124 for liquidation and 12 for either merger or transfer to local governments. Satisfactory - 72 - progress in the overall implementation of this program was a condition for second tranche release of this operation. 1.17 Future Bole of Government inTheSector. To clarify the .roles of -he private and public sectors, the Government had stated that it would limit its involvement in the economy to strategic or priority sectors (Letter of Development Policy dated May 12, 1989). However, while the Mexican Constitution clearly defines strategic sectors (to include oil, hydrocarbons, radioactive materials, electricity, basic petrochemicals, mail, satellite telecommunications, and railways), it does not define priority sectors and this term has been widely interpreted. Thus, under this project, the Government undertook to reconfirm that, in non- strategic areas, it would not get involved in enterprises -- or establish new parastatals or PE subsidiaries, or acquire existing private sector companies -- which are of a commercial nature or are involved in the production of tradeable goods, except in cases of significant market failures involving the produLAon of goods/services with major social welfare impact. Furthermore, as a condition of second tranche release, the Government would empower the Comisi6n Interse:retarial Gasto-Financiamiento (CGF) with the authority to give final recommendation on a case-by-case basis on the creation of PES (and subsidiaries), or the acquisition of companies by Government agencies -- consistent with a redefined priority concept. Measures to Improve Efficiency of Retained Enterprises. 1.18 Following experience elsewhere; the Government believed that a valid strategy to encourage remaining PEs to improve their efficiency was to force them to operate (i) in a competitive environment that subjected them to market forces; (ii) with sufficient managerial autonomy and incentives to encourage them to seek agreed upon performance goals and be rewarded (sanctioned) if the goals are (are not) achieved; and (iii) with sufficient financial autonomy and - accountability. Measures agreed in the context of this operation and related to the above three aspects are briefly discussed below. (a) Measures to Improve the Competitive Environment of PEs. 1.19 Since major reforms to liberalize the economy and increase its exposure to market forces had already been undertaken by the Government, this project only addressed market imperfections affecting the operation of PEs. More specifically, the project i-cluded (I) cross references to conditions under other sector adjustment or investment operations to liberalize the prices of products of five large PEs: SIDERMEX (steel), FERTIMEX (fertilizers), CONASUPO (marketing of agricultural products), CFE (electric power) and FERRONALES (tariffs); (ii) limits on the aggregate transfers to PEs in 1989 which were set at 13% below 1988 levels in real terms; and (iii) issuance of regulations to the Procurement Law, making it clear that there is no discrimination between local and foreign supplies. - 73 - (b) Measures to Foster Managerial Autonomy and Accountahility. 1.20 These essentially included (i) issuance of interpretative regulations to the Federal Law of Public Enterprises redefining the roles of different Government agencies involved in PE guidance, performance and evaluation, and clarifying the roles of PE Boards of Directors and PE management; (ii) preparation and implementation of Performance Agreements for five large PEs: FERTIMEX (fertilizers), CMC (coal), FERRONALES (railways), SICARTSA (steel), and AHMSA (steel); (iii) measures to make the Performance Agreement Unit (Unidad de Convenio de Desempeno - UCD) in CGF operational; and (iv) strengthening of PE reporting, financial management and rzcounting systems. Under the project, training was also foreseen for PE managers, Board appointees, and UCD officials. (c) Measures to Foster Financial Autonomy and Accountability. 1.21 These essentially consisted in (i) simplifying PE's budgeting procedures and practices which gave PE managers little financial autonomy to operate efficiently; and (ii) rationalizing public involvement processes. Thus, under the project, the Government was requested to furnish to the Bank an Action Plan to simplify, deregulate, and make more flexible the budgetary process and its implementation for PEs operating under Performance Agreements. With regard to rationalization of investment processes, concrete measures included issuance of guidelines for economic evaluation of public investment projects, introduction of multi-plan budgeting and establishment of a public investment evaluation unit. D* PROJECT IMPLEMENTATION AND MAJOR RESULTS Overall Evaluation 1.22 When assessing the PERL in terms of the disengagement of the Government, reduction of the size of the PE sector, and policy and institutional cnanges to improve the working environment for retained PEs, there is little doubt about the project's success. The conditionality of the loan was satisfactorily met with only small delays (see Annex I which summarizes the effectiveness and second tranche release conditions under the PERL and the compliance record). The project became effective (and the first tranche released) about one month after loan signing (which followed Board presentation by 2 days). 2lais was due to the intense discussions and important policy actions taken during project processing. Indeed, all the effectiveness conditions, except one were already met at the time of Board presentation, following two rounds of lengthy negotiations in March and April 1989. 1.23 The release of the second tranche was dplayed somewnat, by two and a half months. In retrospect, it appeared that the target date (end - 74 - November 1989, i.e., 4 months after effectiveness) was too light, given that some of the conditions were not only politically sensitive, but also required substantial preparatory_work which was time consuming. This was particularly true for the preparation of Performance Agreements for the five selected large PEs and the issuance of regulations to the Procurement Law and the PE Law. Eventually, all the conditions of the second tranche release were met and the project generally achieved its dual objectives of reducing the size of the PE sector and improving its efficiency. The achievements of the project (and difficulties encountered during implementation) are discussed further below. Disengacement Proqram 1.24 The implementation of t-i distangagement program has been fully satisfactory. From January 1989 to January 1991, 216 PEs have been divested, and 195 were at different stages in the process. Liquidation accounted for 42% of the enterprises, mergers and transfers for 16% and privatization for the remaining 42%. Most of the divested PEs were small to medium-sized enterprises, but there were a few very large ones, such as Telmex and Mexicana airlines. Last year, the Government announced its decision to privatize the steel sector, and to allow again private foreign ana Mexican capital participaticin in the banking system. FERTIMEX is also being restrtzctured for privatization. This is much beyond what was agreed with the Bank under the project and demonstrated beyond any doubt the determinition of the Mexican Government to disengage itself from the economy and to reduce the size of the PE sector. 1,25 Gross proceeds from total PE sales are estimated at about US$5 billion, or 2.6% of 1990 GDP. A survey of the divested enterprises reveals the following characteristics: (i) they are predominantly small to medium-sized firms, for the most part operating in competitive markets; (ii) the buyers have been mainly local investors. Unions were involved only in the purchase of text:I- companies, while foreigners participated in a few large cases (at-to part companies, Mexicana airlines, TELMZX); (Lii) physical restructuring of the enterprises prior to sale was kept to a minimum. r,w investments for physical rehabilitation and employment adjustments have been mostly left to the private investors. Most of the divested enterprises were of a modest size. Only for larger PEs, a certain amount of restructuring has or is taking place in pr- ration for sale. In the case of FERTIMEX, for example, investmen, are being made to rehabilitate some of the fertilizer plants. 1.26 The divestiture has achieved the desired objectives, a the process has worked well with relatively few implementation prc as. However, there has been a worsening of the already skewed and concentrated pattern of ownership distribution in the economy a- an increase in vertical integration. Only a small grnup of local conglomerates have been involved in purchasing PEs. To a large extent, this outcome reflects current ownership pattern in Mexico, and not the effects of divestiture Per se. Mexican policy makers have not yet, as - 75 - in some other countries, used divestiture as a major tool to avoid further asset concentration and distribute ownership more widely in the economy. Their main objectives were to reduce the size of the'PE sector and maximize the financial gains as quickly as poseLble. Achievement of these objectives have led the Government to sell PEa to qualified investors capable of improving performance, rather than to a large number of small and widely dispersed shareholders. Preparing the companies for sale to individual investors through the stock exchange was considered time-consuming and premature, given the large number of companies to be sold. Most of the divested PEs were also small and relatively unknown and did not fit the criteria necessary for public floatations. 1.27 Lessons Learned. The strategic choice of starting small, gradually expanding to larger firms, and the importance given to speed paid off handsomely. Small firms operating in competitive markets were easier to sell, did not need muct, prior restructuring, and would only have marginal impact on allocative efficiency if mistakes were made, given their competitive reture. More ver, early and quick sales helped demonstrate commitment, provided the i,ecessary learning curve and tested the institutional procedures necessary for managing more complex privatizations. 1.28 Other factors of success included the strong institutional capabilities for handling the disengagement program and the transparency of the process. Initially, sector ministries were responsible for the sale of their PEs. Since 1985, however, this responsibil.ty has been vested with the Disincorporation Unit in SHCP, which is staffed with motivated and qualified personnel. This has helped reduce inconsi.stencies and confusion and curtain the power of sector ministries to block reforms. Currently, the involvement of sector ministries is limited to the identification of candidates for privatization, with no other decision-making powers. Sector Improvement Efficacy Program (a) Measures to Improve the Competitive Environment of PEs 1.29 So far, the policy reforms undertaken under this project as well as under parallel operations has only started to have an impact on the efficiency of most PEs. While the trade opening and industrial deregulation have had a positive impact on a number of small and medium- sized public enterprises which began to react to foreign competition, the performance of large public enterprises has remained poor, mainly as a result of the price controls on their products. Although the Government agreed under several different sector loans to a program of price adjustments on goods and services produced or marketed by five large PEs (SIDERMEX, FERTIMEX, CONASUPO, CFE and FERRUNALES), their Implementation has been postponed several times because of the need to fight inflation under the PECE. Prices of goodo and services provided by these enterprises have thus remained for a large part controlled during the implementation of this project, with the notable exception of - 76 - electricity tariffs for which the price adjustment program agreed with the Bank was satisfactorily implemented. On the other hand, the Government implemented the transfer reduction program agreed under this loan. Caught between significantly scaled down Government transfers and the inability to raise prices of their products, the financial situation of many of these enterprises deteriorated. In the case of FERTIMEX, the difficult financial situation and the resulting liquidity problems are complicating the on-going restructuring process as well as the privatization efforts. In early 1990, the Government allowed steel companies to set their prices at parity with international prices and, as part of its decision to privatize SIDERMEX, it completely liberalized steel prices iq September 19q0. It also agreed to fully implement the fertilizer price adjustment program agreed with the Bank, which had experienced temporary delays. Consideration is also being given to a full price deregulation of FERTIMEX products, going beyond the commitments under the Fertilizer Sector Adjustment Loan. Finally, in March 1991, the Government allowed FERRONALES to set its own tariffs as an incentive to private participation in tais PE. (b) Measures to improve PE Managerial and Financial Autonomy and Accountability 1.30 The Government introduced the agreed institutional and legal reforms designed to enhance the autonomy and arcountability of PEs. The Regulations to the PE Law -- intended to elim: .ate contradictions among different laws and decrees dealing with budgeting requirements of PEo -- were issued as agreed, while the approved FY90 budget decree contained many measures going in the same direction of greater PE autonomy, such as the autorization given to PE's Boards of Directors to create new job positions and to modify the personnel structure provided that the changes do not imply increases in the personnel budget. The centralized Information System, which should allow Pe to report information to Government agencies in a timely and efficient manner was simplified, while the amount of information to be reported to the core economic ministries (SPP, SHCP, SECOGEP) and Banco de Mexico was reduced as agreed. Although it may be too early to judge the impact of these reforms on the performance and efficiency of public enterprises, some PE Boards of Directors have taken advantage of their new greater legal autonomy and have been more active in the management and decision-making of the companies. The problem of Government agencies' interference in PE management and operation, which was identified by Bank supervision missions as a major project issue has eased somewhat. While some companies still report tight Government control over their budget and other operational matters throughout the year, the issue seems les acute in others, particularly those implementing well designed Performance Agreements. (c) Performance Agreements 1.31 The establishment and development of a Performance Agreement system in Mexico has been progressing relatively well. As agreed under the project, PAs were developed for five major PEs: Altos Hornos de - 77 - Mexico-(AHMSA), Carbon y Minerales Coahuila (CMC), Siderurgica Lazaro Cardenas - Las Truchas (SICARTSA), Ferrocarriles Nacionales de Mexico (FNM), and Fertilizantes Mexicanos (FERTIMEX). While it may be too early to assess the results of these first PAs, it appears that they are already having a positive impact on some of the enterprises. At FERTIMEX the system has forced the company to prepare for the first time an internal plan which outlined its strategy for reaching efficiency and restructuring goals and set detailed internal targets and responsibilities. Furthermore, in order to monitor performance and prepare information for thidad de Convenios de Desempeno (UCD), FERTIMEX has complemented its cost accounting system with the design of systems permitting accurate measurement and consolidation of costs, margins and volumes, and has revised its accounting practices to better reflect values of inventories and assets. Although part of its management still perceive the PA system as an.additional and time consuming way of Government control, the system has forced FERTIMEX to better analyze results from its production, distrib,ution and financial operations and has given it a good basis for a meaningful discussion on its operations and needs with the Government. FERRONALES also reported that the PA has had a beneficial impact by helping focus management's attention on important targets and has led to the development of a good MIS at the company. 1.32 It is more difficult to judge the impact of performance agreements on the three other companies (AHMSA, CMC and SICARTSA, all of which are part of the SIDERMEX complex), as the introduction of PAs virtually coircided with the decision of the Mexican Government to privatize the steel sector. Thus, while these companies said they folt less Government interference and that their Boards of Directors have been taking a more active role in the management of the companies' operations, they did not know whether thib greater managerial authority was due to the PA system or resulted from the privatization decision. 1.33 Based on the experience gained in 1990 under the project, the coverage of the PA program is being expanded in 1991 to cover the following entities: Productora y Importadora de Papel; Instituto Mexicano del Caf6; Azufrera Panamdrica; Compaftia Exploradora del Itsmo; Comisi6n Nacional del Agua; and Fondo de Fomento al Turismo y Aeropuertos y Servicios Auxiliares. (d) Public Investment Evaluation 1.34 Following a slow start, progress in this area is now satisfactory. Delays in commissioning the agreed study on public investment evaluation and multi-year investment programming procedures, and unforeseen complexities in the process have delayed the original scheduling of this exercise by four months. The final studies are now completed and a plan of action for implementing the new system has been discussed with the Bank. The proposed centralized investment evaluation unit for public sector investment proposals has been established and startad operating in early 1991. - 78 - Technical Asgistance 1.35 The PERL included a US$ 1 million (0.2% of the loan proceeds) technical assistance component to support overall implementation of the PE reform program. Specifically, the TA component included: (a) Studies to assist the Government in the following areas: (i) economic pricing in FERRONALES; (ii) role of mining sector agencies; and (iii) private sector development in PE- denominated areas; (b) Specialized support to SHCP, SPP, SEMIP and CGF for (i) improving existing financial management, accounting and reporting systems; (ii) developing a simplified integrated PE information and monitoring.system; and (iii) preparing PAs for the five selected PEs; and (c) Training to UCD staff and PEe' senior management 1.36 So far, only USS 300,000 (30% of the component) have been used to prepare PAs for the five selected PEs and Lo improve existing reporting systems. Although the carrying out of the key activities under the TA program was a condition of second tranche release, the Bank showed flexibility by releasing the second tranche before full implementation of the TA program. These studies and actions were judged noq-essential and there was sufficient evidence of satisfactory progress on the overall PE reform program. The Mexican Government had requested and the Bank had agreed to extend the Loan's closing date by a year to allow completion of the TA program. E. PROJECT SUSTAINABILITY 1.37 The loan documents identified three major risks concerning the sustainability of the PE reforms: (i) disruption of the macro-economic adjustment program; (ii) possible opposition to some of the reforms by the strong public sector labor unions and/or political parties; and (iii) resistance to reforms by existing vested interests within government institutions and affected PEs. 1.38 The first major risk identified refers to a possible return to expansionary policies as a result of the high domestic costs of stabilization or insufficient foreign financing given the requirerants of the macro-economic adjustment program and growth targets. Were ihis to occur, a reversal of PE reform might take place. The President'. Report for this operation commented that this risk was mitigated by -he continued efforts made by the Government in maintaining budgetary discipline and in controlling inflation. Indeed, in both areas, the Government performance has been exemplary and it can be said that in Mexico the budgetary situation and inflation are now largely under control. With regard to external financing, the debt reduction deal and continued large flows of external assistance have, so far, assured adequate financing for the implementation of the macro adjustment and - 79 - economic recovery program. Under these circumstances, the risk of a derailment of the reform process appears small indeed. 1.39 The i.econd risk is of a political nature and refers to possible opposition to the reforms by labor unions and the political parties. The President's Report noted that this risk was very much mitigated by Government's good and open dialogue with the private sector, labor unions, and political parties, as evidenced by the formulation process of the PACTO and PECE, and by the clarity of the reforms' objectives which were shared by these various economic and political groups. Indeed, implementation of the reforms did not meet strong opposition as the privatization and enterprise liquidations did not result in drastic reductions of personnel. In addition, actions have been taken to minimize the short-term social cost. These included severance payments, early retirement plans, special retaining programs. 1.40 The third risk is mainly institutional and concerns existing vested interests within governmental institutions which may resist the measures aimed at reforming the situation, whereby various Ministries interfere with PE management and exercise ex-ante controls over critical PE activity. 1.41 The reluctance of some Government agencies to relinquish centralized control and interference in PE management have been identified as the main issues during implementation of the project. They have slowed down the transition from a centralized ex-ante system of control to one of ex-post evaluation, based on the assessment of PE management and performance. However, the problem seems to have eased recently with many enterprises being subjected to less interference and enjoying more flexible control from Government agencies. Continued progress in this area of PE managerial autonomy is essential to the sustainability of reforms achieved so far in the public enterprise sector. F. BANK PERFORMANCE 1.42 In late 1988, when the Mexican Government asked the Bank to put together a package of reforms and financing that will permit Mexico to resume growth after six years of stagnation, the Bank's response was quick and thorough. On very short notice, some 50 professionals visited Mexico to prepare and appraise simultaneously three large sector adjustment operations (The Financial Sector Adjustment Loan, the Industrial Sector Adjustment Loan referred to in para. 1.07 and the PERL discussed in this report), which together amounted to US$1.5 billion and were expected to help secure other sources of financing to trigger the increased investment necessary for renewed growth. The ability to respond quickly was made possible in part by the intense program of economic, sector and operational work in Mexico over the preceeding years. It had, however, necessitated some modifications to the Bank's - 80 - work program in Mexico with more emphasis given to policy-based lending and economic and sector work linked to structural adjustment. 1.43 The Bank's performance during project implementation was also fully satisfactory. The continuity of the Bank team working on the project and the good relationship they have succeeded in establishing with their Mexican counterparts facilitated communications, but could not totally eliminate the impression felt by some officials in Mexico of receiving at times contradictory comments from various Bank missions coming to Mexico under different terms of reference during 1989-90, a period of extensive economic dialogue and simultaneous implementation of six major adjustment operations. Supervision was intensive with the task manager or a full supervision mission visiting Mexico every quarter. Mexican representatives also came to Washington to discuss project related matters. Implementation issues were thus detected at a very early stage and c(:rrective actions taken rapidly. The delays in the issuance of the regulations to the PE Law (a sec.nd tranche release condition), far example, whre brought to the attention of the highest authorities and the issue was quickly resolved. G. BORROWER'S PERFORMANCE 1.44 There cannot be any doubt about the commitment of the Mexican Government to its PE reform program and this very much explains the success of the PERL. In many instances, the Government went much beyond what was agreed under the project. In fact, the Bank in essence played a supportive role in the Mexico PE reform process, which started in 1982 and, after some interruptions, picked up steam in the second half of the 1980s. The Bank-financed PEPL fitted into that program and helped strengthen the position of t.e reform-minded elements in the Government. The Bank's assistance proved particularly useful in developing the PA system in Mexico and in starting the process of enhancing PE managerial and financial autonomy. 1.45 The Borrower's performance in carrying out its operational responsibilities under the project was also very good. There were some initial project ccordination problems at SHCP, but they were resolved once NAFIN assumed -.rrject coordination responsibility jointly with the SNCP team. The project accounts were regularly audited as mandated by the Loan Agreement and the audits did not reveal any particular problem. H. CONCLUSION 1.46 Progress under the public enterprise reform program in Mexico has been impressive. The Government fully achieved what it set out to do in this phase of its PE reform program and in the Bank-supported PERL operation. Now that an adequate legal and institutional framework is in place, the next phase of the reform is likely to involve the privatization and/or restructuring of large PEs such as FERTIMEX, SIDERMEX, PEMEX petrochemical, CONASUPO, and FERRONALES, while - 81 - PROJECT COMPLETION REPORT MEXICO PUBLIC ENTERPRISE REFORM LOAN (Loan 3086-ME) PART II: PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVEJ A. INTRODUCTION 2.01 This document supplements the version prepared by the World Bank on the results of Loan 3086-ME for public enterprise reform. 2.02 It provides the perspective of the Mexican Government on the project and its course, in relation to three criteria: (1) performance of the World Bank, (2) performance of the Mexican Government, and (3) experience and lessons learned. 2.03 In the first instance, comments are given on the World Bank from various standpoints (catalyst of efforts, supervisor of activities, expert on the economic, financial and institutional environment in the public sector, among others), with a view to evaluating its performance. 2.04 The performance of the Mexican Government is assessed via a brief evaluation of public enterprise disengagement, price policy, increase in efficiency and productivity for certain PEs and, lastly, legal reformi. 2.05 Brief comments are also made on the positive aspects in general of the experience with the project and lessons learned under the loan for public enterprise reform. ?.06 The following members of a working group assisted in the preparation of this paper: CIGF (Comision Intersecretarial Gasto- Financiamiento) technical secretariat, umbrella and sectoral coordinating departments, participating entities and the Sociedad Nacional de Cr6dito, a financial agent of the Federal Government. l/ Translated from Spanish. The original text in Spanish appears as Annex III. - 82 - improvements in the working environment for retained PEs continue to be made. This is a difficult and complex undertaking as it involves the companies in Mexico which count. It is, however, essential to the success of the PE reform process and-Eo carry through this difficult endeavour, Mexico needs and deserves continuing strong donor support. 1.47 Lessons Learned. The principal lessons learned, or again confirmed, with implications to other projects, as a result of the experience during project implementation and the preparation of the PCR are summarized below: (a) The key factor behind the success of the PERL has been the firm commitment of the Mexican Government to the adjustment process in general and to the PE reform in particular. Given this commitment, it was possible to implement a comprehensive reform program quickly and effectively; (b) The reform program was well designed and the reform objectives were shared by the various affected groups in Mexico. In restrospect, it would appear that the target date for the second tranche release (four months after effectiveness) was too optimistic, given the conditions to be fulfilled. Overall, however, the loan's conditionality was well gauged: Tough on a few essential conditions, which were worded in such a way as to give little room for interpretation, while remaining flexible on less essential measures in order to avoid locking in rigid targets. Thus, all the studies and secondary actions did not appear as dated covenants in the loan's legal documents, but were included in an Action Plan whose implementation should be satisfactory to the Bank. This allowed the Bank to pass judgement on the overall progress without blocking the project if some non-essential actions had not been implemented; and (c) Supervision was intensive and benefitted from the continuity of key staff on both the Government and the Bank's sides. However, no amount of supervision would have made a bad project into a good one. The Mexico PERL had all the ingredients of success. Good supervision helped make its implementation smoother. The lesson is clear: The Bank should be prepared to support committed Governments with both financial resou.-ces and high-quality analysis and project work. - 83 - B. BANK'S PERFORMANCE (a) As a catalyst and coordinator 2.07 An analysis of the Bank's performance in generating standard criteria for use in resolving issues and problems shows up a number of differences regarding the approaches used by the officials of that financial institution. 2.08 Because of that dichotomy, the comments made by the various Bank missions were not always consistent and caused a lack of consensus on important issues. 2.09 However, the role of the Bank as a catalyst in the v. '--1s aspects related to loan effectiveness is considered favorable. The presence of Bank missions did in fact accelerate the action taken by various entities and departments in implementing the measures agreed jointly. (b) As a creditor 2.10 The financial conditions and terms of the loan were considered favorable, as there was a repayment period of 17 years, grace of 5 years, an interest rate of 7.6% adjustable every six months, and a commitment fee of 0.75% per year. 2.11 Loan approval and signature were also smooth. Disbursements were made on the dates agreed, and 99.9% of the proceeds was spent within 10 months because compliance with the conditions negotiated with the Bank were satisfactorily met. Only a small share of the technical assistance component was not spent, and its deadline runs until June 30, 1991. 2.12 In this regard, the technical assistance component was designed to provide comprehensive support to project implementation, and the commissioning of certain studies included in the component was a condition for the second tranche release. Nevertheless, the Bank remained flexible as regards loan supervision and even though that condition was not met in its entirety, the funds under the second tranche were not at any time held back. This also made it possible to disburse US$12S million to support the efforts of the Federal Government in renegotiating the external debt with the commercial banks. 2.13 The foregoing involved a major effort by the World Bank, the Mexican Government and the various entities that participated in the project to coordinate the package in its various phases: preparation, negotiations, supervision and the attainment of the agreed targets within the planned timeframes for proper loan execution. - 84 - (c) As a supervisor 2.14 Supervision was constant throughout the disbursement periods and found to be satisfactory with regard to the monitoring of the pilot program of performance agreements and the investment project evaluation unit. In some cases, the Bank's involvement entailed making suggestions and approving the CVs of the participating consulting firms, in addition to considering various ideas and criteria related to those matters. 2.15 Nevertheless, the supervision occasionally provided partial analyses, depending on the specific sector or area, and did not always have a big picture of the problem. Thus, from the microeconomic standpoint, it was frequently suggested that various types of control be eliminated from the public enterprises, which were nevertheless accepted by other Bank staff because they have shown their merit in regulating public expenditure and thereby helping achieve the macroeconomic targets behind the stabilization policy. (d) As an analyst 2.16 The Bank document (p. iii of the summary and p. 11 of the text) mentions three risks that can affect the continuity of the reforms to improve the efficiency of the PEs. 2.17 The first of these refers to the institutional risk, described as the vested interests within government institutions that could block the measures aimed at increasing the administrative and financial autonomy of the public enterprises. 2.18 This interpretation by the Bank may come from an unawareness that the Federal Government's economic policy is basically designed to accelerate public enterprise disengagement, the ultimate goal being to have a basic core of strategic enterprises. In this context, the comments by Bank staff lose their significance. They seem to be more concerned with marg4nal questions of supervision and control than with the structural change that is being carried out rapidly by the Mexican authorities to scale down the parastatal public sector. 2.19 From the microeconomic standpoint of the monitoring and supervision mechanisms, these arrangements are not fully understood. It is not known that the deficit/surplus agreements are part of an overall planning process in which a basic part is the annual budget approved by the Chamber of Deputies, which is negotiated intensively with the entities involved. 2.20 The physical and financial objectives indicated in the deficit/surplus agreements -- emanating from the budget process -- are the result of an intense exchange of opinions over various months between the enterprises and authorities. In other words, the procedure is similar to that used for the performance agreements, an instrument which does enjoy the =ecommendation and favorable opinion of the Bank. In this regard, the two types of agreements are considered complementary - 85 - and to-have common features as regards negotiation, and should not be abandoned until the macroeconomic problem is totally resolved and/or the impact and significance of the parastatal sector has been substantially pared down. 2.21 The deficit/surplus agreements for budget monitoring consider only overall ceilings and not specific headings. They have the advantage of allowing for timely preventive or corrective measures in enterprises that have major financial imbalances, with their consequent unfavorable impact on important economic variables. 2.22 The second risk, identified as a disruption to the macroeconomic adjustment program, refers to the possibility of returning to a policy of expansionary spending. 2.23 In this regard, there was some inconsistency by a "microeconomic" evaluation mission, when the "macroeconomic" missions from the Bank itself were fully aware that the Mexican Government was complying strictly with the negotiated and agreed measures regarding balancing the budget. 2.24 At best, there is the impression that there was not due communication and exchange of information within the Bank itself. There was palpable evidence and clear statements made by the Federal Government to reaffirm the goal of maintaining discipline in all aspects so as not to undermine the efforts and sacrifices made for years. (The Bank) clearly needs to be better informed in these areas. 2.25 This situation cannot be separated from effectiveness in containing PE expenditure. Had such supervision not existed, public spending could have grown beyond the limits agreed under the economic stabilization policy. The economic vision of the Mexican Government is comprehensive and aligns the macroeconomic targets with those of a microeconomic nature. 2.26 The third risk identified by the Bank relates to opposition to the program by labor unions and political parties. In this regard, the reforms have not run into strong opposition. On the contrary, there is fertile ground for consensus, and even private initiative has supported the process of disengagement and greater efficiency in the public sector as there is keen awareness that the country needs to become part of the international context, changing its modalities and approaches. All of this is rooted, among other factors, in the PECE and the opening up process, which have simultaneously prompted a change in the standard for national economic growth and consequently the attitudes of the private sector and labor unions. - 86 - C. PerformancM of the Maxica_n Government (a) An a borrower 2.27 In general, compliance with the conditions and commitments established by the Bank for public enterprise reform was done on schedule as agreed in the terms for disbursement under tranches I and II of the Loan, regarding the disengagement program, improvement in the competitive environment, and measures to strengthen managerial and financial autonomy. (b) Disengaaement and restructuring of enteiRrises 2.28 In improving the performance of public enterprises, the Federal Government embarked upon two simultaneous processes. First, through disengagement, the number of entities under public control was reduced and, second, those enterprises retained in the public sector have increasingly clear guidelines as to economic rate of return and competitiveness. 2.29 Likewise, anong th-- agreements negotiated with the Bank, with regard to institutional aspects, the Federal Government took appropriate measures to strengthen the disengagement process. It thus strengthened the institutional capacity of the Disengagement Unit of the Secretariat of Finance and Public Credit and the SECOGEF established a new Direcci6n General to oversee that process. Both the selection of public enterprises targeted for disengagement as well as justification for retaining the others could thus be examined, evaluated and an appropriate decision reached by the economic cabinet and/or the Intersectoral Expenditure/Financing Commission. 2.30 As a result, the parastatal sector could be streamlined faster than anticipated in accordance with the premises of budget control and disengagement from nonpriority enterprises. (bl) The disengagement process 2.31 Progress with the disengagement program shows that whereas on December 1, 1988 there were a total of 618 PEs, of which 204 were subject to disengagement, on January 31, 1991 there were 417, disengagement of which was under way for 140. 2.32 Thus, 216 PEs had been disengaged and 10 new ones established, of which 4 are involved in priority activitieg and 6 set up to facilitate disengagement of the steel subsector. In addition, five PRO that were being disengaged were retained. 2.33 Of the total disengaged, 42% was done through liquidation or shutdown, 6% through mergers or transfers, 42% through sale and the - 87 - remaining 10% ceased to be public, as they did not meet the requirements set by the federal PE law. 2.34 The social sector participiEed in the disengagement process, through various labor groups acquiring four sugar mills, two fishing operations and one freezing plant, a processing plant for agricu'ltural inputs and equipment from another, a textile operation and two plants in the food industry. 2.35 Revenue from the sale of PEs, amounting to some Mex$10 billion in nominal terms, has been a noninflationary source of income for the public coffers. Beginning in 1990 it has been used as a contingency reserve, which became extensive this year. 2.36 The disengagement of the large enterprises transferred or in the process of being soid to the private sector is briefly described below. (ii) Steel and Mining 2.37 The scope of some measures went beyond what was originally negotiated with the World Bank. This was the case with the AHMSA and SICALRTSA steel companies and their subsidiaries, the divestiture of which had to be accelerated to ensure their modernization, without necessitating more funding from the Federal Government. 2.38 This made it possible to start their restructuring, and a plan was formalized for ARMSA and significant headway made with SICARTSA. These efforts reduced the administrative payroll by 22% at both companies. Furthermore, regulations were formulated for procurement, leases and the provision of services, contracts and public works at both companies. 2.39 In the mining sector, Carbon y Minerales de Coahuila was merged with three coal producers (Carbon y Cok, Hullera Saltillito and Minera de Guadalupe), and Real Del Monte and Pachuca, and Minera de Cananea were sold. Studies were begun on the disengagement of the Rio Escondido coal-mining company, which supplies the fuel needed for the CFE's coal- burning power station. 2.40 As part of the disengagement process, large-scale programs were carried out to contain costs and increase the productivity of the mining companies, with greatest efforts being made in the most productive areas and less profitable operations being closed down, in particular the unproductive and risky mines in the SIDERMEX raw materials group. (iii) Tel6fonos de M6xico 2.41 By the size of its assets (Mex$24 billion), this service company is one of the largest PEs in Mexico. To improve telephone service radically, ensuring a steady expansion of service and promoting scientific and technological research in telecommunications, the Federal - 88 - Government decided to disengage the company from the public sector, through transfer of its holdings. 2.42 The first phase in the sale to the private sector was-carried out in late 1990, when all preferred shares held by the State were sold. In early 1991 the new administration took over operations at the company and the Government will put the rest of its shares up for sale to complete privatization in the fist six months of this period. (iv) Sugar 2.43 In 1990 the State completed its withdrawal from sugar production, with the sale to the private and social sector of 22 industrial units. The State was left with 5 sugar processing plants, down from the 54 that it owned previously. 2.44 Likewise, efforts to deregulate marketing continued. The way was opened for participation by the private sector in the marketing of sugar, which uncil 1989 was the private preserve of Azucar, owned by the Federal Government. A timetable was set up for a reduction in personnel and restructuring of all its aspects, linking this action to progress in disengagement. The objective of the program is to pare personnel by the end of 1991 to 1,000 positions, from 2,391 at the end of 1989. (b2) Restructuring of public enterprises 2.45 For the retained public enterprises, in accordance with the criteria indicated in the preceding section, the Federal Government has introduced a series of instruments, including the pilot program for performance agreements and the investment project evaluation unit, which ensures full linkages between overall and sectoral objectives and between the short and medium run. The ultimate objective is to channel funds in the most optimal way so that those enterprises become models of efficiency and achieve the objectives set for them. Using the methodology of the performance agreements to improve PE operations and productivity, SEMIP and SCP promoted the conclusion of another type of agreement for the monitoring and supervision of the entities they coordinate. Thus, together with the other agreements already established by the Federal Government (deficit/surplus and financial rehabilitation agreements), the basic core of instruments for raising productivity and efficiency in the parastatal sector was strengthened. (i) Pilot program of Performance agreements 2.46 Performance agreements are negotiated by the Federal Government with the enterprises to improve their operations within a framework of greater operational autonomy. They make possible a systematic analysis of the behavior of the enterprises in accordance with targets set in the various areas of their operation. In addition, the attainment of the targets is associated with the granting of incentives to the directors. - 89 - 2.47 The pilot performance agreement program overseen by the Performance Agreement Unit, composed of SECCGEF, SHCP, SPP, BANXICO, and the Technical Secretariat of STPS and CIGF, was carried out in three phases: Phase 1 2.48 In November 1989 the AHMSA, SICARTSA, CMC, FERTIMEX and FERRONALES agreements were signed, establishing a set of targets for each enterprise for production, marketing, finance, use of inputs, and modernization or restructuring programs. 2.49 In each enterprise the targets were expressed in terms of specific indicators, each of which has a relative weight. Five ratings were adopted to rank the achievement of the targets agreed with each enterprise. Three of the ratings (1-A optimal, 2-B excellent and 3-C outstanding) were above the norm, in accordance with the budget. Level 4-D is the enterprise's level of operations on a daily basis. Lastly, level 5-E indicates deficient performance. 2.50 At present, the indicators and their weighting differ from enterprise to enterprise, as a result of the negotiations with each of them and their specific problems. In the future the agreements should contain more standardized indicators, giving due attention to the specific problems of the entity in question. Phase 2 2.51 Phase 2 consisted of the development of an information system that made it possible to do monthly, quarterly, six-monthly and annual monitoring and evaluation. The some 90 variables were organized into a set of formats and guidelines that form part of the Federal Government's centralized information system coordinated by the CIGF Technical Information Committee. Phase 3 2.52 The pilot program concludes with an annual evaluation of the performance of the targets set for each enterprise and based on that evaluation, distribution of incentives to the directors whose ratings so merit. Evaluation 2.53 The following comments derive from the results of the pilot program: - The selection of targets should be improved. It appears that some irrelevant indicators were used, and others that could be key, such as quality control (CMC, FERTIMEX and FERRONALES) or market participation (AHMSA), were not included. The indicators could also be changed when - 90 - external factors totally beyond the control of the enterprises are present. - The quantitative targets must be more precise; for;the five enterprises, some indicators far exceeded the optimal target and need to be adjusted to real levels. This was particularly true for AHMSA, CMC and FERTIMEX. - The weightings assigned to each target could be refined, depending on the structure of the entity's operations, so as to better reflect its main problems over time. - The analysis of the performance agreements should be strengthened with the results of the deficit/surplus agreements and other instruments, with a view to gaining a more complete picture of the operations and real problems. 2.54 In summary, the performance agreements were a satisfactory means for administering, evaluating and monitoring the behavior of the public enterprises, not only for the Federal Government, but above all for the director general of the entity himself. There is no question that there is room for improvement, given that this is a medium-term task. Nevertheless, the course embarked upon appears appropriate. Main activities in 1991 2.55 In 1991, previous experience will be used to develop the performance agreement program and expand its coverage, to bring in, beyond FERTIMEX and FERRONALES, the following entities, which include both producers of goods as well as services: Productora e Importadora de Papel; Mexican Coffee Institute; Azufrera Panamericana; Compalia Exploradora del Itsmo; National Water Commission; and the National Tourism Promotion, Airports and Auxiliary Services Fund. A document with the basic criteria to be used in those agreements will also be prepared and disseminated, so that the criteria are consistent and uniform. (ii) Investment Proiect Evaluation Unit 2.56 In 1989 the Investment Project Evaluation Unit was established to increase the efficiency of investments and appraise them properly prior to their inclusion in the federal budget, with a view to maintaining at all times a suitable balance between the entities' own funds, tax revenue and external resources. The basic functions of the Unit are thus: - to seek consistency among institutional, sectoral and overall objectives based on the macroeconomic environment surrounding the national development plan; - serve as liaison in the annual programming and budgeting process, with medium-term planning; - 91 - - help impletient the economic policy as regards public investment. 2.57 The unit was set up to opeiate through an Executive Committee, on which would sit a representative of the SPP Secretariat and the Under-secretariats for Development Planning and Budget Control, and Programming and Budget. It maintains direct relations with the Directorates General of Programming and Budget and serves as advisor to the Undersecretariat of Programming and Budget. 2.58 Initially, it served as adviser and gradually became a decisionmaking body. It has two types of linkages with the rest of the government: - formal, which begins with the sending of information from the enterprise through the sectoral coordinating bodies and SPP Directorates General of Programming and Budget, routed through to the Unit; - informal, through which it has the flexibility to act to establish contact direct with the planning and coordinating units in question. This communication is exceedingly important in proceeding efficiently with the analysis of investment projects. 2.59 In this regard, the Unit has specific functions in terms of the monitoring and evaluation of public investment projects on a selective basis. It provides institutional coordination to strengthen, support and serve as liaison with the planning units within the entity and the sectoral coordinating body. It establishes and disseminates criteria for the formulation and appraisal of investment projects, including those that have a multi-year structure and multisectoral impacts. It also trains government workers in the formulation and appraisal of investment projects. 2.60 In accordance with the powers assigned to the SPP, this support unit reports to the Undersecretariat for Programming and Budget. 2.61 The Unit thus began analyzing, among other projects, three strategic PEMEX operations: ecological, liquid petroleum gas and telecommunications, in addition to the "Solidarity" satellite, the Mexican airport system and roads program. (iii) Legal reforms 2.62 The measures negotiated with the World Bank to improve the competitive environment have been fully carried out and the managerial and financial autonomy of the PEs strengthened. 2.63 A fundamental aspect of the public enterprise reform project was to improve the legal framework to give the enterprises greater managerial and financial autonomy and improve the competitive - 92 - environment in which they operate. In this regard, the regulations governing the federal public enterprise law were issued, and the regulations governing the law on procurement, leases and provision of services related to movable assets were updated. The regulations governing the public works law, bases and guidelines for the competitive bidding on works and services financed by the World Bank and regulations governi-, the regula-ory law of Article 27 of the Constitution regarding mining were amended or expanded, and the law on sugar taxes repealed. 1. Requlations of the federal law on Public enterprises 2.64 The regulations were issued on January 26, 1990 and stressed the following actions regarding the powers of government bodies: - to issue the criteria and policies by which the entities must abide, reflecting their financial situation and the objectives and targets; - to establish criteria for budget rationality, austerity and discipline; - to modify prices and rates for all goods and services; - to receive assistance from nongovernmental sectors, except where prohibited by law; - to determine the criteria for selecting the members of government bodies; - to support the monitoring of special programs, implementation of modernization processes, and approval of equipment and facilities through technical committees or commissions. 2. Requlations governing the Law on Procurement, Leases and Provision of Services Related to Movable Assets 2.65 The regulations were issued on February 13, 1990. They allow bidders from member countries of the World Bank to participate as suppliers in national projects, provided they comply with Mexico's other legal requirements. The regulations also expand the functions aud powers of government bodies in this regard. 3. Requlations governing the Public Works Law 2.66 The amendments and additions to these regulations published on January 9, 1990 establish that, for public works with external financing granted to the Federal Government or with its backing, the bases, guidelines and requirements.for registration are to be established in each case by the Secretariat of Programming and Budget, in accordance with the conditions, circumstances, amounts and complexity of the works. This led to an agreement with the World Bank to establish the bases and - 93 - guidelines for competitive bidding on works and goods financed under loans and credits from that international institution. 4. Requlations governing the Requlatory Law of Article 27 of the Constitution regarding Mining 2.67 The new Mining Law regulations provide a new legislative and fiscal framework that tends to stimulate private investment in the sector by allowing exploitation of products such as phosphorus, sulfur and potassium, previously reserved for the State. 5. Law on Sugar Taxes 2.68 This law was repealed, so that individuals and bodies corporate that acquire the product for the first time do not pay taxes, making private participation in the marketing of sugar possible. (iv) Price Policy 2.69 The measures to improve the competitive environment of the PEs include negotiated and gradual price adjustments for entities with the largest budget transfers in the 1989-90 period. This resulted in satisfactory progress in execution of the program, despite the fact that it was held back by macroeconomic considerations such as economic stabilization and containment of inflation, in addition to the faster opening up of trade to external markets, which selectively permitted greater deregulation and flexibility regarding price levels. 1. Steel 2.70 With the opening up of the borders, the steel industry has been gradually deregulating steel prices to levels consistent with those for imported products, promoting the need for efficiency in the operations of sectoral enterprises, through modernization and streamlining programs that enable them to achieve a competitive position on the market. 2. Fertilizers 2.71 As FERTIMEX is increasingly able to cover its operating expenditures and amortize its liabilities, the Federal Government decided on a price policy that would allow for an annual increase in real terms of 12.3% during 1989-92, with a commitment by the entity to improve its operations, administration and financial management. To date, those prices have changed as follows: - On July 31, 1989 an average increase of 18.8% was authorized, which although not consistent with that stipulated, made it possible to recover the price level in force in December 1987, when it was seen as the point of departure for implementing the strategy of annual increases of 12.3% in real terms. - 94 - - In May and September 1990 average increases of 20% on were authorized. Additionally, in November a new average increase of 12.7% was granted. With those adjustments, the Federal Government carried-through on its commitment to grant an increase of 12.3% in real terms, as the increase in nominal terms was 14%. D. Experience and Lessons Learned 2.72 The Bank's technical and financial support for the Public Enterprise Reform Project (Loan 3086-ME) was provided within a context of an overall adjustment program begun by the Mexican Government several years previously that stressed trade opening, a paring of the public sector, and a strategy of growth with price stability. This contributed to the success of the loan, as the frame of reference was a new model of economic growth supported by the Federal Government. 2.73 The process of disengagement and greater efficiency on the part of the public sector went beyond what was agreed and at times faster than planned because it was based on the terms negotiated among the groups involved and transparency in the management of information through the various programs (PLANADE, PSE and PECE) established by the last two administrations that created the fertile ground for the participatory negotiations. As a result, private initiative and labor unions could participate actively in the process. 2.74 The World Bank evaluation points out risks that could have disrupted the public enterprise reform process: change in the macroeconomic adjustment program; probable opposition from labor unions and/or political parties; and the existence of vested interests within the government institutions. The Bank may have sensed these risks because the various supervision and analysis missions it sent did not share a common pool of information. In this regard, the Bank team must be fully aware of Mexico's financial problems through greater dialogue within the Bank and more interaction with Mexican officials. This would further enhance the favorable results of the loan. Nevertheless, the loan has brought greater soundness and security to the development process, improving prospects for long-term planning. 2.75 In summary, the Mexican Go%arnment considers the experience of the negotiations with the World Bank for implementing the public enterprise reform loan to be positive. - 95 - PROJECT COMPLETION REPORT MEXICO PUBLIC ENTERPRISE REFORM LOAN (Loan 3086-ME) PART III: STATISTICAL INFORMATION (Amonats (US$ million) Original Disbursed Cancelled Repaid Outstanding Loan 3086-ME 500.0 499.3 - - 499.3 Original Loan Dates AcGtua3 or Re-estimated Initiating Memorandum December 15, 1988 Dee-cr 15, 1988 Negotiations March 13-19, 1989 March 13-19, 1989 April 25-29, 1989 Letter of Development Policy May 12, 1989 May 12, 1989 Board Approval June 13, 1989 June 13, 1989 Effectiveness July 1989 July 11, 1989 Loan closing June 30, 1991 June 30, 1992 Actual completion June 30, 1991 June 30, 1992 --- ----------------------------------------------------------------------- cuIO4uLATIVE LOAN DISBURSEMENT "90 FY91 Mi Planned 499.0 1.0 (ii) Actual 499.3 0.7 (iii) (ii) as % of (i) 100.1% 70.0% --- ---------------------------------------------------------------------- MISSION DATA No. of No. of Staff Date of Month. Year Weeks Persons Weeks Reort Appraisal Oct./Nov.S8 2.0 19 38.0 may 22,1989 Supervision I Sept. 89 1.0 4 4.0 Oct.18,1989 Supervision II Dec. 89 0.5 1 0.5 Dec.12,1989 Supervision III May/June 90 1.3 2 2.6 Aug.24,1990 Completion October 90 0.6 2 1.2 Apr.15,1991 FOLLOw-ON ADUSTMENT OPERATIONS NONE - 96 - ANNEX 1 CONDITIONS AND COMPLIANCE RECORD A. EFFECTIVENESS: July 11, 1989 Conditions Compliance (i) level of operational and satisfactorily met investment transfers to PEs in 1989 to be 13% below the level in 1988 in real terms (ii) Preparation of a price adjustment satisfactorily met plan for goods and services produced by 5 PEs (iii) Selection of five large PEs for satisfactorily met the implementation of the PA Plan and 199 additional PEs for the disengagement program (iv) Statement by the Government of its satisfactorily met role and policy in the production of goods and services, defining clearly the concept of priority area for public sector participation (v) Strengthening of the PE-Sales unit satisfactorily met at SHCP and issuance of clear guidelines on PE sale procedures (vi) Establishment of the UCD (Unidad satisfactorily met de Convenios de Desempleo), appointment of the Unit Coordinator and agreement on an action plan for the development and application of PAs (vii) Substantial progress in obtaining satisfactorily mct adequate financing for its requirements for FY89/90 in the context of the adjustment program (viii) Issuance of draft regulation to satisfactorily met the 1986 PE law (ix) Submission by NAFINSA of satisfactorily met subsidiary agreements to the Bank - 97 - B. SECOND TRANCHE RELEASE: February 22, 1990 The release of the second tranche was conditional on the Batk being furnished with evidence satisfactory to it that: (i) the macroeconomic policy framework was consistent withthe PE reform program; and (ii) continuous progress has been made in obtaining adequate financing for Mexico's requirements for FY89/90 and over the medium term in the context of the adjustment program. Both conditions were found to be satisfactorily met. In addition, the Bank needed to receive sati.sfactory evidence on the following: Conditions Compl1iance (i) CGF has been formally empowered to An interministerial give final recommendation on a case-by- resolution dated July case basis on the creation of PEs, their 10, 1990, formally subsidiaries and acquisition of companies empowered CGF to by public sector entities approve or r-,act, on ac .se-by-case basis, the creation of PEs (ii) Continued progress in restricting satisfactorily met transfers to PEs in a manner consistent with the approved FY89 Federal Budget (iii) Issuance of regulations to the Regulations issued on 1986 PE law defining: (a) Government January 26, 1990 agencies' role in supervising PEs; (b) role and composition of PE's Boards of Directors; and (c) charters of PEs enhancing the role of Boards' technical committees (iv) Satisfactory progress in the satisfactorily met. implementation of the Disengagement PAs for the 5 program selected PEs were signed in December 1989 (v) Issuance of the regulations to the Regulations issued on Procurement Law February 13, 1990 (vi) Satisfactory operation of UCD satisfactorily met (vii) Satisfactory guidelines for the satisfactorily met elaboration of sector plans and corporate programs provided in the National Development Plan (viii) Satisfactory progress in carrying satisfactorily met out the Implementation Program. ANNEX II Disengagement Program, 1982-91 (Number of PEs) Disengagement Proposa a Oversight Ministry/ Situati.on in Situation in initiated new processes Situation in Sector pec, t in Dec._88 before Dec. 88 89-90 Total Jan. 91 SG (Interior) 74 17 2 3 5 12 SRE (External Relations) 1 1 -- SHCP (Finance) 183 168 33 70 103 65 SEDENA (Defense) 7 7 1 1 2 5 SAR (Agriculture and 94 92 41 37 78 14 Natural Resources) SCT (Communications and 76 41 16 5 21 20 Transports) SZCOFI (Trade and indus- 50 36 2 1 3 33 trial development) SEP (Education) 56 48 3 8 11 37 SSA (Health) 23 18 2 1 3 15 SEPESCA (Fisheries) 40 15 11 2 13 2 0 SEMIP (Energy, Mines 400 93 68 5 73 20 and Industry) SEDUE (Urban Develop- 80 28 17 4 21 7 ment.and Environment) SECTUR (Tourism) 51 21 6 8 14 7 SPP (Plan and Budget) 23 17 1 - 1 16 DDF (Federal District) 17 14 1 2 3 11 other 39 17 1 4 5 12 Total 633 205 151 356 277 Source: SECOGEF, Mexico