Pollcy, Planning, and Ro arch .WORKING PAPERS Public Seolor Managenont end PrJvate Sector Developmmnt Country Economics Department The Worid Bank October 1988 WPS 118 Contract Plans and Public Enterprise Performance John Nellis Contract plans help clarify goals, increase managerial auton- omy, and open a dialogue between management and govern- ment - but their benefits have been oversold. Ibe Poicy. PLnnmg. an4! Reseuch C«noplex disinbutes PPR Wo*ding Papeua to dsamnate the fndings of wok in progrea and to enouwage the cchmmge of idees among Bank saff and ail othexa intereted in developinint issucs Thcse papers caty the names of the authoms, reflet only iheir views, and should be used and cted accordingy. The findiugs, interpzetationm, and conclusions are ihe authnrs' own. They should nc be attributed to theWorld Bank, iu Boad of Diacors. ita nnagmeen,, oranyofits mnbercountuies. Po oy lning, end Roeeeach PublcSoto Management end Pdivato Sector Dovelopment Roughly 100 public enterprises in developing The problem is, they have been oversold. countries (50 of them with Bank support) are The- are no. he mechanism of choice for using contract plans - ncgotiated performance healing a sick company. They work best with agreements between the govemment/owner and firms that operate commercially and already the enterprises' managers or directors. have decent management and sound financial and reporting procedures. Contract plans have not improved the financial performance of public enterprises In supporting contract plans, the Bank dramatically. The process is probably more im- should keep ambitions modest and emphasize portant than me prcduct. the clarification of goals, increased managerial autonomy, and open negotiating between They do produce clearer goals, open a management and govenmment. Contract plans dialogue between management and the state, and should not be a condition of Bank adjustment offer such benefits as better accounting, audit- operations - at least not until the process has ing, and management information systems, been tested and shown positive results in the including physical and financial performance in- country. dicators and performance targets. This paper is a product of the Public Sector Management and Private Sector Development Division, Country Economics Department. Copies are available free from the World Bank, 1818 H Street NW, Washington DC 20433. Please contact Rose Malcolm, room N9-055, extension 61707. The PPR Working Paper Series disseminates the findings of work under way in the Banks Policy, Planning, and Research Complex. An objective of the series is to get these findings out quickly, even if presentations are less than fully polished. The fmdings. interpretations, and conclusions in these papers do not necessarily represent official policy of the Bank. Copyright © 1988 by the International Bank for Reconstruction and Developmnentirhe World Bank CONTRACT PLANS: AN ASSESSMENT OF TREIR UTILITY AS PERFORMANCE IMPROVEHENT HECHANISMS IN PUBLIC ENTERPRISES Table of Contents Page No. Executive Su-mary ..................................... i - ix I. INTRODUCTION ........................................... 1 A. The Issue .......................................... 1 B. The French Experience .............................. 5 The Nora Report .................................... 6 Implementation - First Phase ....................... 8 Implementation - Second Phase ...................... il Third Phase - 1981 to Present ...................... 15 Conclusions on the French Experience .... ........... 25 Il. The Experience of Developing Countries .... ............ 26 A. Senegal ............................................... 27 The First Phase ...................... .... 30 Fundamental Problems .................................. 33 Evaluation and Revisions in the Process .... ........... 37 The Second Wave ....................................... 39 Detailed Example - The SONATEL CP ..................... 41 Performance Indicators ................................ 43 Senegalese Assessment of CPs .......................... 49 Conclusions on Senegal ................................ 52 Recommendations , .......................... .......... 58 B. Other Developing Countries ............................ 60 Congo . ........................ ......................... 60 Table 1 - CPs in Developing Countries .... ............. 60a Table 2 - CPs in Bank Projects ........................ 60b Morocco . ............................................... 61 Royal Air Maroc ....................................... 61 The Public Enterprise Restructuring Loan - The Moroccan PERL ................ . 64 The Office Nationale d'Electricite .................... 66 Other African Countries ............................... 69 Benin .............................................. 69 Burundi ............................................ 69 Gambia ............................................. 70 Madagascar ......................................... 71 First Efforts in India ................................ 72 III. Conclusions and Recommendations ....................... 74 Summary of the Main Points ............................ 74 Recommendations for Government and World Bank Staff .... 80 CONTRACT-PLANS: THEIR UTILITY AS PERFORMANCE IMPROVENENT MECHANISMS IN PUBLIC ENTERPRISES Executive Summary I. Introduction 1. Contract-plans are negotiated performance agreements between governments, acting as owners of a public enterprise, and the manar,ers or directors of the enterprise itself. In a contract-plan (henceforth CP) the intentions, obligations and responsibilities of the two parties are freely negotiated and then clearly set out. This appears simple; but the fact is that ambiguity of goals and conflicting objectives have long been identified as major obstacles to the effective and efficient performance of public enterprises. CPs define the enterprise's objectives, and state what resources and latitude vill be provided, by government, to enterprise management so that it may accomplish the specified goals. Many CPs set out the physical and financial indicators which will measure enterprise performance. Many establish the principle that government will compensate the enterprise for costs incurred in fulfilling non-commercial objectives, and specify the way in which the compensation vill be made. In theory, the agreement concluded binds both signing parties, in the manner of a formal contract. 2. The mechanism vas invented by the French and applied, in the 1970s and the 1980s, to eight large public enterprises, all monopolies or oligopolies with heavy social service obligations. In 1983, the French government concl"tded thirteen CPs of a slightly different nature with nationalized industrial public enterprises. The concept has been exported; first to francophone countries in Sub-Saharan and North Africa -- with Senegal and to a lesser extent Morocco having the most extensive and lengthy experience with the process -- to several anglophone African countries; then to Argentina, Brazil, and Mexico in Latin America, and most recently, in some experimental efforts in Bangladesh and India. This paper examines in detail the French and Senegalese experiences, reviewing the objectives of the CP process, where these performance agreements have been applied, with what results -- economic, financial and institutional -- and what lessons have been learned. Several other cases of CP use receive a brief review. They are either too new or too few to be analyzed in depth; though the promising Moroccan situation is given some attention. There are roughly one hundred public enterprises in developing ccuntries that have CPs, and another one hundred are in preparation. The World Bank is presently supporting the process in over fifty enterprises in ten different countries. -ii- II. The French Experience 3. The first two CPs -- then called "program-contracts" -- vere signed in 1969-70; with the national electricity company and the national railway. They defined objectives, and quantified financial performance targets, for the firms for the five year period of the agreement. Agreement on objectives was reached through a process of negotiation between government and management. The basic idea was that if goals were both agceed upon and clearly set, then government could lighten its time- consuming a priori controls, provide management with efficiency-enhancing autonomy, and evaluate the firm on the basis of achievement or non- achievement of the set goals. They stipulated significant rate hikes and called for increases in managerial autonomy on personnel and pricing decisions. 4. In the first half of the contracted period, results were positive. Budgetary aid to the two declined by 20 percent in the period 1970-72. This vas encouraging; the budgetary burden issue was the one which most interested the French government of the day (and it remains the issue for LDC governments using the device). The rate increases in electricity were sufficient to allow the utility to cover its variable costs, in the period 1970-73, plus meet two-thirds of the finance charges on new investments. Managers and government supervisors agreed that goals had been clarified, transparency of operations improved, and government reviewers realized that increased management autonomy did not automatically mean increased waste. These were advances. 5. But the first "oil shock" of 1973 led to increased costs of inputs, inflation, a snap political decision to build several nuelear generating plants -- and all the elaborately prepared ratios, financial projections and investment targets of the CPs suddenly vere out of date. The performance-guiding capacity of a CP is tied to the accuracy with which it predicts and projects key financial and economic variables. The greater the discrepancy between assumed and actual circumstances, the less the CP can serve as a detailed guide for management or evaluators. A second serious and recurring problem in France was government's refusal to permit service and staff reductions and price increases necessary to cover costs; even though the right of management to impose such changes was either implied or directly stipulated in the agreements. Faced with a widening gap between the projections of the CPs and actual performance, the contract with the electricity company was not renewed. The railway contract was "extended" for a two year period, but not immediately re-negotiated. 6. No other CPs were signed till 1978-79. Then four public enterprises vere submitted to the process: the national coal company, Air France, the Ceneral Maritime Corporation, and -- the only repeater from phase one -- the railway. In an attempt to deal with the difficulties encountered in the first group, the timetable of the second group of CPs was reduced from 60 to 36 months. Annual revision was built in. The basic goals were repeated: to allow enterprise managment more freedom to set -iii- prices, ieduce the workforce, end unprofitable product lines and close consistently money losing operations. Heightened emphasis was placed on costing out and compensating for socially imposed objectives. Technical innovations, mainly quantified physical and financial indicators to serve as measures of performance in monopoly utilities, were introduced. 7. Once again, implementation got off to a promising start. The coal company broke even in 1980, for the first time since 1951. Railway operations became more transparent through application of the performance indicators. The shipping company was committed to a reduction in its workforce by 25 percent, and a reduction in its fleet by over a third. (It was threatened with liquidation if it did not comply.) Air France had -- and continues to have -- the most positive CP experience. As it operated in basically competitive markets, commercial profitability was a more accurate measurement of performance. Thus, goals were more easily quantified, managerial autonomy more easily accorded, and deviations from commeiical operation more easily demarcated -- and compensated. Targets were set for operational growth rates, financial objectives, and key physical indicators. 8. The Socialist victory in 1981 brought a halt to the implementation of several aspects of the CPs. Prices were not raised for rail services, a decision was made to re-open some unprofitable rail lines which had been closed, the shipping company retained its full workforce, the threat to liquidate the company was dropped. The basic assumptions of the second group of CPs, save for Air France, were simply revised from above. The Socialist government then negotiated some further ntraditional" CPs with utilities or social service providers such as Gaz de France, the electric.ty company (II), Air France (II) and the railway (III). In 1983 the government also negotiated new-style CPs with thirteen industrial and manufacturing SOEs. These were, supposedly, guidance and orienting mechanisms; operational tactics were to be left to boards and managers. These CPs placed less emphasis on productivity and financial targets and more on new product lines, R & D efforts, market shares, and general harmonization of the efforts of these companies with the general goals of technological modernization and enhanced international competitiveness. 9. Despite the CPs, net transfers to all public enterprises increased from 12.8 billion francs in 1975 to 49.5 billion francs in 1982. In 1984, the French government reviewed their experience and offered five main lessons: (a) keep CPs short and simple; (b) keep them flexible; (c) "the existence of the contract-plan is more important than its content" (meaning that the formalized, periodic exchange of views between govern ent and firm were more important than the quantified indicators or performance goals; the process more important than the product); (d) the weaker the performance of the enterprise, the greater the difficulty of negotiating and implementing a CP; and (e) the agreement is not really a contract for the reason that the state cannot be subjected to legal proceedings, by -.he enterprise, in the event that it fails to honor its obligations. -iv- 10. There have been no studies comparing the performance of French SOEs before and after the introduction of CPs; nor has a comparison been made between firms vith and vithout CPs. Assessments have tended to be of a narrative, qualitative nature, polling the attitudes and opinions of managers and government supervisors with regard to CPs; they have concluded that concerned personnel support stiongly the device. The French have invested considerable resources in CPs. In most instances, performance has not improved dramatically (though Air France 1-las performed well under its CP regime). Too much was expected of the process at the outset; and the CPs inability to protect the SOE from the effects of uncertainty has been a disappointment to 4.ts originators. Recent French CPs have been short, general documents, outlining general policy rather than specifying the details of performance. The apparent failure of the mechanism to stem the budgetary outflow has led the managers of the process to scale down their ambitions regarding the utility of the CP device. III. Thé Expérience of Developing Cointries Sene 2al 11. The CP process was introducted to Senegal in 1980, in the framewo*k of a World Bank supported technical assistance project to the parapublic sector. The first five performance agreements were prepared between March of 1981 and January of 1982. In 1984, having reviewed the first experiences, the government decided to renew and expand the process in nine enterprises, and to consider the possibility of extending the device to many other concerns -- mostly utilities and agricultural development agencies, but some financial and industrial enterprises as we]l. As of May of 1988, CPs have been signed in nine different agencies. Three of the nine are now in their second CP, meaning that twelve sets of negotiations have been concluded. Senegal thus has the most experience with CPs of any country other than France. 12. The Senegalese make a distinction between CPs and "lettres de mission" (roughly translatable as "statements of objectives"). The former are used in p -blic utilities and industrial enterprises; the latter have been applied ) rural and regional development agencies, the commercial nature of whi< is more limited. Of the first group of five agreements negotiated in 381-82, three were never implemented due to the lack of promised investment funds (a scarcity heightened by the decision of the Vorld Bank not to release the second tranche of Senegal's SAL I). The first CPs followed the French models, stating the purpose of the enterprise, projected performance, and obligations of the government. But from the outset the Senegalese agreements had to deal with an issue which was not a serious problem for the French; i.e., the question of arrears between the government and the SOEs, and arrears among the SOEs themselves. The first CP for the urban transport company, for example, specified that government would settle its unpaid bills, and previously promised, but partially paid or unpaid subsidies. This issue was and remains a major -v.. problem. In almost every case, the Senegalese governmeat has committed itself in a CP, and then failed to inject promised amouats of equity, or failed to settle its arrears to the firm, or failed to pay off subsidies promised in the past, or failed to allow price increases, etc. 13. The weakness appeared principally to lie in inadequacies of the budgetary review and resource allocation proc-sses. That is, individual CPs specified the steps needed to put a fira i a sound footing, particularly vith regard to financing. The guvernment largely accepted the CPs as drafted, and signed them, but then eventually found that it had nowhere near the resources necessary to cover the aggregate commitments made in the CPs. Resource dispensers then resorted to partial payments, with the sums and recipients being decided by the political process. Thus, the quantified goals set in CPs vere all too often set aside; CPs proved not to be successful as detailed guides on operations at the level of the firm. They did, however, clarify goals and stimulate the dialogue between mar.agement and the state. This was beneficial; and associated with the CP process was a series of reforms -- management information systems, improved accounting and audit systems, the introduction of physical and financial performance indicators -- which lay the base for longer term improvements. 14. The second, or post-1984 generation of Senegalese CPs has been constructed to deal with the high degree of financial uncertainty which prevails in the country. With Vorld Bank support, efforts have been made to strengthen the agency in the Office of the President which is responsible for SOE reform in general and the CP process in particular. CPs have included sections detailing the schedules by which the government would clear past financial commitments, and stipulating what services vould be reduced in case of non-payment. A high-ranking inter-ministerial review body was created to monitor each CP, and lobby for the resolution of financial problems, if and when they should occur. Most of the most recent CPs include statements committing the government to ear-marking sums in the annual budget to cover most if not all of the average annual payments made by government for the good or service provided (water, electricity, telephones, etc.). 15. The CPs have produced a number of positive and beneficial results. Government supervisors and enterprise managers alike strongly approve of and support the CPs, and the open dialogue process which creates them. A late 1986 Senegalese assessment stated that firms with CPs show better personnel management, a superior capacity to estimate investment needs, more success in gaining approva; for rate increases, and improved relations with suppliers of inputs -- resulting, in several cases, in lower prices for these inputs. Another study suggested that enterprises with CPs had higher rates of grovth in turnover and lower labor costs than firms without CPs. There was no difference in profit-making or rate of return on equity invested. (The study had a number of methodological shortcomings; the positive results -- modest as they are -- must be interpreted with caution.) 16. Despite advances, the number and intensity of problems are greater than those of the benefits. For example, CPs are intended to promote managerial autonomy, but in Senegalese practice, managerial latitude remains quite limited. As performance evaluation and stimulation mechanisms, CPs should link an enterprise's results to revards and sanctions for management and the workforce. This has not been done. Above all, the financial problems noted above have not been overcome. The government has, in all too many cases, consistently been unable or unwilling to meet its obligations, despite their supposedly contractual nature, and despite the corrective mechanisma enacted. The fact is that, as in France, the CPs are not true contracts; the enterprises cannot take legal action in case of non-compliance. 17. The state's failure to pay became so widespread that SAL III (1987-88) made the settlement of the state's obligations, as set out in existing CPs, a condition of second tranche release. Government only managed to meet the condition, at the very last moment, by unilaterally decreeing that the sums given in the existing contracts were all revised downwards, thus allowing the state to claim that it met the financing conditions. This tactic negates the basic principle of tne contract-plan process; open negotiation of nles engagements reciproques" (mutual commitments). Most of the CPs produced have been technically competent documents. But despite the CPs and the pressure to honor them coming from the Bank, neither Office of the President reviewers, nor Ministry of Finance or Plan reviewers have been able to aggregate all the resources committed in the individual CPs and match them to overall resource availability. This is not to say that all the problems are the fault of government; it is rather that the CP process highlights government's deficieneies. 18. Suggested ways to improve the situation include: divestiture of part of the sector, allowing the government to concentrate scarce resources on a smaller number of enteprises; overall reduction in the budget of the resources devoted to SOEs, thus forcing a priority ranking of actions (this is being done through the SAL process); involvement of Treasury officials in the CP negotiations; publication by the Ministry of Finance of a formal payment timetable for each CP; procedures whereby the Ministry of Finance at least informs SOEs when payments will be partial or delayed; simplification of paperwork on claims and payments; closer links of SOE payments and the budget cycle; and, finally, the formation of new, high- level committes to monitor the payments process. All of these avenues are being explored. 19. Senegalese CPs might have avoided some implementation problems had there been a stronger supervisory unit. Though perfectly placed at the apex of power in government, the Contract-Plan Cell (or Supervisory Unit) has failed to exert sufficient pressure on budget and disbursement offices to persuade them to make promised payments on time, in the required amounts. One problem has been rapid turnover of its leadership, a second the junior status and lack of experience of its cadre. In 1987, the unit -vii- was reorganized and given expanded powers; however, it still needs internal reinforcement and a clear division of labor with other supervising units, especially in the Ministry of Finance. 20. The primary lesson is clear: contract-plans cannot by themselves overcome the several, serious problems of the Senegalese parapublic sector. Weak budgeting systems (meaning a lack ot` reliable information on either the availability or the liquidity of funds), weak agencies of review in the technical and financial supervising ministries, und weak commitment to the concept of financial discipline -- the CP process highlights these problems, and shows that they are not all the prublem of management; but it does not eliminate them. The CP is meant to be both binding and a plan. But the Senegalese experience reveals the difficulty of vinding the government, and of constructing a plan that is both realistic and imple.-entable. In miniature, the informational shortcomings of comprehensive planning are reproduced in the CP process. 21. Senegal, and other developing countries using the device, cannot easily adopt the French method of solving these problems; i.e., by moving away from comprehensive CPs towards sho,rter, more general statements of intent and broad policy. The French idea is to maximize the benefits of mutual goal-setting and formalized dialogue, which the process provides, vithout becoming bogged down in specific numbers and quantities. Senegal is obliged to retain a greater degree of complexity in its contracts. There the trend has been to use shorter intervals of CP review and modification, and to strengthen the reviewing agencies, in an effort to deal with the prevailing high degree of uncertainty. Irrespective of continuing problems, the World Bank continues to be heavily involved in the CP process. The reasoning is that the CP's emphasis on clear goals and on increased transparency of operations could and should, in the medium term, render important, low-cost assistance in the search for more productive and efficient public enterprises. CPs cannot make governments manage more rationally their resources; but they can provide, and are providing in Senegal, rationalist reformers with the ammunition to structure and legitimate reform. Recommendations for Senegal 22. The CP process has little meaning if the State consistently makes unrealistic financial promises which cannot be kept. Three sets of action should be taken. (i) The cost-reduction and cost-recovery measures already set forth in existing CPs must be enforced. (ii) The links between the resource allocation -- i.e., the budgetary -- system and the CP process should be strengthened. (iii) A strengthened capacity within government, particularly in the Mintstry of Finance, to project total resource availability, and to rank technically priority areas for expenditure, is essential. -viii- Otheg DeveloRint Countries 23. The CP or a related mechanism is in use (or in preparation) in a number of other countries; most, but not all francophone African. Many of these CPs have been required or requested in Bank projects or adjustment operations; most of them have been prepared with technical assistance personnel paid by funds provided by World Bank credits. Benin. Burundi. Congo. Cote d'Ivoire. Gambia. Madagascar. Mal., Morocco. Niger and Tunisia: all have CPs in place or in preparation in Bank-related operations. In none of these cases has the CP process been in place for a sufficient period of time to permit a full assessment. 24. Still, the Moroccan case offers some encouraging signs. Learning from a trial run in their national airline, which had a CP from 1982-84 (but was never re-newed>, the Moroccans are presently installing CPs in six major SOES, including the national electricity company, the water company and the railroad. Two of the six, including the water company, were signed in December of 1987; the remaining four are scheduled for signature by the end of 1988. Completion of all six is a condition of second tranche release of the World Bank's Public Enterprise Restructuring Loan (PERL) to Morocco. 25. The failure to honor financial obligations was a major problem with the airline CP. The clearance of arrears and the imposition of financial discipline are the goals of the PERL. Attention to financial details thus occupies a central position in the Moroccan CP process. But neither the involved government staff nor enterprise officials regard CPs as ultimate solutions or as rigid and inflexible documents. They are seer. as first steps in a lengthy, experimental process, during which both sides will learn. No consultants, foreign or domestic, have been involved. Moroccan planners feel that it is better to have imperfect documents, the preparation/learning benefits of which accrue to Moroccans, then technically perfect CPs, prepared -- and understood and supported -- only by consultants. 26. Though the first two CPs signed in the framework of PERL have only been in effect for a few months, the state bas cleared the arrears situation in these two firms, for the first time in the 1980s. Enterprise officials attribute this to the CP process, which gives quasi-legal force to the state's promise to pay, and to the firm and persistent action of the supervising agency, which has actively pressured the Treasury to honor contracted obligations. This is a substantial accomplishment. A concern is that neither CP eliminated or drastically revised the powers of the accounting agent or financial controller; officl-'s of the Ministry of Finance who reside permanently in each SOE and *.,se approval is necessary -- and sometimes difficult to obtain -- for expenditures. ': vas thought and hoped that the CPs would lighten their role; this has not yet happened. -ix- 27. In both France and developing countries CPs have been vell crafted. Problems have been more ln implementation than in design. CPs have not proven to be particularly good at restoring basic health to enterprises ln criais, to the point where the French -- and increasingly, the World Bank -- nov recommend that enterprises first be rehabilitated, then submitted to the CP process. Developing country usera of CPs have more often tried to apply the device to ailing firms, vith disappointing results. The CP is best used to regularize and improve the framework for firms operating in basically commercia' situations, endowed with decent management and sound financial and reporting procedures. It is not the mechanism of choice for healing a sick company. 28. The CP process requires a strong, vell-placed supervisory agency to assist in the preparation of the agreement, review implementation, and especially push for the proper fulfillment of the contract. A difficult task is finding the proper balance between empowering the managers through the CPs, and at the same time endowing an oversight agency with sufficient "clout" to enforce the CPs. The hope of the creators of the CP process was that it would lead to a reduction in the powers of government supervising agencies. This has not come about, first, because of a normal reluctance of governments to diminish their preogatives, and second, due to the unforseen need to have a strong voice inside government pushing for the honoring of commitments. The latter is legitimate, the former unwarrented: CPs should replace normal control agents and procedures, not be added to them. 29. Despite the name, CPs have not proven to be legally enforceable contracts. If governments are not legally bound, what is the status of the obligations imposed on the enterprises? Persistent non-compliance with terma of the CPs has discouraged and demotivated enterprise managers. A frequently suggested reform is requesting or requiring governments to treat the terms of the CPs as legally contracted, with either party being able to enforce compliance through the courts. This is unrealistic. Governuents already have the pover to impose their will on enterprises, and none has shown any villingness to allow itself to be sued. Perhaps the term "contract" should be eliminated from the process; the more accurate term would be "performance agreement," a title already used in Latin America and anglophone Africa. 30. Overall, the CP device i of value, but its benefits have been a bit oversold. Government officials and World Bank staff contemplating its introduction should thus exercise caution. Ambitions should be modest. CPs should be introduced in one or two test cases. They should be kept short and simple. Emphasis should be placed on goal clarification, an increase in managerial autonomy, and open negotiating between government and management. CP use should not be a matter of conditionality in Bank adjustment operations; at least until the process has been tested in- country and shown to be workable and beneficial. Contract-Plans: An Assessment of Thoir Utilitv as Performance Imorovement Mechanisms in Public Enterprises I. Introdtiction A. The Issue 1. Contract-plansl are negotiated agreements between governments, acting as owners of a public enterprise, and the enterprise itself. The essential idea of a contract-plan is that the rights and duties of both parties are clearly spelled out. The concept was devised to attack the problems of unclear or shifting objectives, insufficient autonomy of managers, and excessively constraining control systems, problems perceived as major hindrances to public enterprise efficiency and productivity. 2. A contract-plan sets out, on the basis of negotiations between government representatives and enterprise managers, the intentions, obligations and responsibilities of the two parties. Its period of activity extends beyond the budgeting cycle, covering three to five years. A typical contract-plan specifies enterprise objectives, in terms of 1/ Terminology varies from country to country or from period to period within the same country. These agreements are also known as enterprise contracts, program contracts, performance agreements, or action plans. The underlying principle -- mutual specification of rights and responsibilities by government and public enterprise -- is the same in all cases. -2- desired overall socio-economie impact, production goals, and/or quantities and quality of service to be provided during the period of the agreement. It defines poli4ies and parameters vith regard to such items as numbers employed, size and growth of the enterprise's wage bill, and social, or non-commercial activities. Many recent contract-plans stipulate the physical and financial indicators which will measure enterprise performance. What distinguishes the contract-plan from a set of directives imposed by the owners of a firm, or from a corporate plan produced by the firm on its own, is that it also spells out the obligations of, and limitations on, the government. That is, contract-plans normally define what management may do without government's review and approval, what prices the enterprise can charge for its goods or services, and how, and under what conditions, these prices will fluctuate. Many establish the principle that government will compensate the enterprise for costs incurred in fulfilling non-commercial objectives, and specify the means and mechanisms by which the compensation will be made. A typical contract-plan lays out the enterprise's financing and investment program, noting the amount the enterprise must generate internally. the amount to come from government subsidy or equity injections, and the amount to be raised by credit, with or without government guarantee. 3. Historically, elements or precursors of the contract-plan process can be found in a number of countries, but the device first came to prominence and fruition in France in the late 1960s and early 1970s. From 1970 to the present, a group of eight of that country's largest and most important (from several points of view, including degree of burden on the budget) public enterprises have been submitted to varying forms of what the -3- French call "the contractualization process." The test cases, in the energy, transport and utility sub-sectors, have all been monopolies or oligopolies with heavy social service responsibilities. In 1983, following the Socialist electoral victory, thirteen further contract-plans were concluded with industrial public enterprises, a few of which had long been in the public sector, but most of which had been nationalized by the Mitterand government in 1981-82. The results of both of these experiences have been judged as moderately positive, though the criteria used to make this assessment have been more institutional/managerial and qualitative than economic and quantitative in nature. Finally, French analysts are at pains to point out the difficulties of determining the exact relation between shifts in enterprise economic and financial results, and the presence or absence of a contract-plan (henceforth, CP). 4. Given the provenance of the mechanism, it is not surprising that the developing countries which have attempted to borrow or adopt the contract-plan have been, in the overwhelming main, francophone. The first recorded attempt to export the process vas in 1973, when negotiations produced a CP for a Tunisian public textile firm. For unknown reasons, the agreement was never signed or implemented. But, starting in Senegal in 1980, and spreading through most of francophone Africa up to the present, the CP has been adopted, or is in the process of being prepared, in roughly seventy-five public enterprises in ten different francophone Sub-Saharan countries. The process is advanced in Morocco, where six CPs are presently being installed in a group of the most important PEs. The device is again being applied in Tinisia. In many of these instances, World Bank -4- structural adjustment or public enterprise reform programs have supported, and increasingly required, the installation of CPs. 5. Variations on the CP theme are in preparation in three anglophone African countries as well. Outside the African continent, the necessity to improve the performance of public enterprises has combined with the hope that the CP or something like it would effect desperately needed improvements, where so many other procedures and reforms have failed. This has led to the adoption, or consideration of a CP process in Argentina ("program agreements," in negotiation and preparation in the transport sector), Mexico, Bangladesh, India ("memoranda of understanding" in six major enterprises in the energy, industrial and engineering sectors), and inquiries into the prospect of using the CP in a number of other countries. Worldwide, there are over a hundred contract-plans in operation or advanced preparation, and an equal, probably larger number, under active consideration. 6. The purpose of the present paper is to examine the contract-plan experience; to determine what they were designed to achieve, how they have been constructed, where they have been applied, with what results -- economic, financial and institutional -- and what lessons have been learned. The study's major concern is the utility of the device as a way to improve performance in developing country public enterprises. Most of the available information on that subject comes from Sub-Saharan Africa, much of it from Senegal, the country most experienced in its use. The study thus examines in detail the Senegalese experience, and reviews more briefly the CP process in other African countries, Morocco and India. -5- However, the seminal nature of the French experience necessitates that the first order of business be a review of the CP process in that country. B. The French Experience 7. Throughout the less-developed world, governments have relied heavily on public enterprises2 as tools to accomplish their developmental objectives. The use of public enterprises has not been limited to developing countries; most developed countries have had, and retain, substantial PE sectors of great economic and financial importance.3 French PEs, before 1981, employed 12 percent of salaried workers outside agriculture, contributed 13 percent of value added, and accounted for 30 percent of gross fixed capital formation. All three ratios rose substantially after the nationalizations of 1981-82; all three have declined, perhaps to less than 1981 values, following the 1986-87 privatization of thirty PEs -- with a further thirty-five supposedly to follow. Nonethelesse even when (or if) all scheduled privatizations are completed, the French PE sector will still account for a higher percentage of GDP than in most other industrialized countries. Z/ These are also known as state-owned enterprises, parastatals, public corporations, or government-invested corporations. i/ Worldwide, public enterprises account for about 10 percent of GDP; to give an idea of the LDC's heavier use of the device, note that in Sub- Saharan Africa the figure is roughly double the world norm. -6- 8. The French public enterprise sector came into existence following World War I, and undervent surges of expansion in 1936 (the Popular Front), 1945-46 (the reaction to World Var II), and 1981-82 (the nationalizations following the election of the Socialist government). There was no dramatic shift of policy in the period 1947-67, yet the sector expanded steadily in that period as many public enterprises created a large number of subsidiary or affiliated companies. In 1966, the Gaullist governuent, increasingly concerned with the poor financial performance and particularly the large deficits of the monopoly public service enterprises, commissioned a high level panel to review the situation. The result vas the Nora report of 1967 -- named after the chairman of the commission, Simon Nora -- which first suggested the concept of the contract-plan (at that juncture callec the program-contract). The Nora Report 9. The Nora report portrayed the deficiencies of the French public enterprise sector in critical terms and tone familiar to those working on the subject in developing countries today. It deplored the poor state of enterprise accounts and budgeting systems, the enterprises' inability to finance their own investments, the ineffective, time-consuming, dirigiste systems of government control and supervision, the lack of clarity of objectives, the government's failure to award enterprises sufficient or timely tariff increases; and the resulting general transfer of the costs of running these enterprises from the users of the goods or services they produced to the taxpaying public at large. -7- 10. The dual prescription for reform offered by Nora vas equally familiar: it vas necessary to clarify objectives and set commercial profitability as a fundamental aim. Principal sub-objectives vould be to increase management's autonomy, and improve State/enterprise supervisory arrangements. The State should no longer supervise by means of direct ...commands, by a prioLi authoiization, or by taking day to day decisions best left to managers, but rather by the clear setting of criteria and of the rules of the game." It vas recommended that the costs imposed by social service obligations be determined, and compensation mechanisms created. A distinction vas made between public enterprises operating in competitive markets and monopolies; the former should be lightly supervised while the latter needed stricter controls. Only the latter would be subjected to the CP process. But, argued the report, all enterprises, including monopolies, could benefit from increased autonomy. il. The report devised the concept of "engagements reciproques" -- mutual cwmmitments or obligations -- which lies at the heart of the CP process. The report suggested that through negotiations the two parties would reach a binding agreement stating what the enterprise would accomplish and what the State vould do -- and not do -- to guide, support and evaluate the accomplishments. The result would be an appropriate balance between needed enterprise autonomy and government's legitimate right to assure accountability. Contracts were to be used to set appropriate levels of enterprise inputs and outputs, the amounts of capital subvent_ons, borroving limits, tariff parameters and vage bills, and determine the margin of latitude of enterprise managers and directors. The -8- reasoning was that if goals were clearly set, government could lighten its a prlori controls. The social costs of the operation would become more transparent, and this would make easier the task of evaluating the firm. The expectation was that as public awarenesa grew regarding the real costs of the enterprise, then pressure would grow for the reduction of these costs. A higher percentage of the operating costs might then be shifted to users. This, in turn, would create pressure to increase competition between public and private enterprises. The report thus envisaged the evolution of a drastically changed public sector "mentality," where costs -- especially personnel costs -- would be reduced, technology would be modernized, and management methods would change. The contract process was to serve a key role in bringing about this set of improvements. ImDlementation - First Phase 12. Preparatory negotiations for CPs started in 1969. The first two CPs were signed in 1969-70, with the national electricity company, and the national railway (respectively, Electricite de France or EDF, and the Societe National des Chemins de Fer or SNCF). Each was suppcsed to remain in force for five years. The 1970 contract-plan for EDF was considered as the model CP. It spelled out a system of lighter, more flexible government supervision which reduced the powers of the permanent representatives, within the firm, of the ministries of Finance and Energy (representatives of the financial and technical supervising ministries are present in every French PE; their presence slows decision-making and hinders the seizing of opportunities). The CP supported expansion by EDF into nuelear power -9. generation. An early draft called for large rate increases which would allow EDF to cover the heavy investment costs and still turn a profit. Analysts note that it was the Ministry of Finance which insisted that the final agreement hold price increases to less than increases in the CPI (similar to the RPI minus x pricing arrangement used in privatized British monopolies at present). Nonetheless, from 1970 to 1973, the selling price of electricity was sufficient to cover the variable costs of EDF's production, plus meet two-thirds of the finance charges on new investments. 13. But the 1973 first "oil shock" led to vastly increased costs of major inputs, economy-wide inflation, and a snap political decision to increase sharply the capacity to generate electricity by nuclear reactors --- which in turn meant a spate of loans and subsidies to EDF not called for in the CP. In the altered circumstances, it became apparent that many of the elaborately prepared numbers and ratios in the CP suddenly were dated and irrelevant. 14. In the first SNCF CP, signed in 1971, the goals were to reduce the railway system's deficit by cutting staff by 15 percent, and by giving the enterprise considerable latitude on rate setting. The full achievement of all the CP's objectives would have substantially diminished the rate of loss but not turned the SNCF into a profitable company. But even the modest financial goals set in the first CP were not attained. This was partly due -- again -- to unanticipated increases in the costs of major inputs. An important second factor was a lack of political willingness, when the crucial moments came, to tolerate the service and staff reductions and price hikes necessary to cover costs. Third was the railway's -10- inability to compete with truckers in goods transporting, which reduced demand for rail service below projected levels.4 15. The SNCF CP was nonetheless extended for an extra two years, running through to 1977. The rail system's financial difficulties were not resolved, as government repeatedly cited regional development policies or possible social tensions as justifications for inactivity on price increases and cost reduction measures, even though these had been agreed on in the CP. 16. Despite the bleak picture with regard to financial results, assessments of the first CP experience were positive. All those involved, enterprise managers and government supervisors alike, agreed that the institutionalized CP negotiating process did lead to clarified objectives and increased understanding of the constraints operating on the other party. Budgetary aid to the CP'ed enterprises declined by 20 percent in the period 1969-72; though it then shot up to historically high levels following the first oil shock. This was important, for the budgetary burden issue is the one which most interested the French government (and A/ This section on the French experience utilizes the following sources: Michel Durupty, Les Entreprises Publiques - Tome 2 - Gestion/Controle (Paris: Presses Universitaires de France, 1986); J. de Chalander, IGF, Ministere de l'Economie, des Finances et de la Privatisation, "Les Relations entre l'Etat et les entreprises publiques en France," mimeographed, Paris, 1984, and J. P. Anastassopoulos et J. P. Nioche, Entreprises Publiaues: Exoeriences Com»arees (Paris: F.N.E.G.E., 1982, especially Chapitre 1, pp. 19-55. -11- remains lh& issue for most developing country governments presently using the process). Transparency vas somewhat increased, and government controllers recognized that "management autonomy does not inevitably mean vaste."5 These vere substantial achievements. Still, much of the optimism, or idealism, with regard to forward budgeting and financial projections -- key considerations in a process which projected production, costs and investments for five years into the future -- had faded in the light of the unforeseen changes in the world economy starting in 1973. Some observers outside of government thtus called the process into question on the basis that any CP vas only as good as one's ability to prediet accurately the economic future. It was clear that that ability was poorly developed. Others felt that the first experience, especially in the SNCF case, showed that government's power inevitably overwhelmed the enterprise; that is, when there was a dispute or a crisis, it was government's position which prevailed. This led some to believe that the concept was mis-named and perhaps unworkable. None of the observers discussed the time or resources devoted to the preparation of the CPs, which must have been considerable. A noteworthy fact is that neither of the two first phase CPs vzs immediately renewed (though the SNCF CP was, as noted, prolonged, and then went into second phase negotiations). Imolementation - Second Phase 17. The period 1973 to 1977 was one of acute economic crisis and uncertainty during which the CP process was not expanded; but as a modicum l/ Durupty, p. 368. -12- of stability returned the French government again turned its attention to contracts. Two follow,up committees to the Nora body vere formed. They issued reports, judged to be "less imaginative and more pragmatic" than that of Nora, which mnodified the CP process in light of lessons learned. The first change was that of scaling down the time frame of CPs; from 60 to 36 months, with greater provision for annual revision as and when circumstances altered. The second change was that of tightening the foeus of the CPs, which henceforth vere to be based on three principles: - transparence of social transfers; letting the public know whc is paying for what; - transparence of resuli-s of PEs; by publishing and publicizing performance statistics; - transparence on investments, which henceforth had to be justified in cost/benefit terms. The overall goals were repeated: to increase autonomy of management, allow the enterprises more freedom to set prices, reduce the workforce, end unprofitable product lines, and close consistently money losing factories or operations. Heightened emphasis was placed on costing out and compensating for socially imposed operations. To emphasize the shift in focus, these agreements were given a new name: "enterprise contracts." 18. Four of the supposedly new style (in reality, they did not differ dramatically from the earlier agreements) CPs were negotiated in the period 1977-79; all were signed in 1978 or 1979. The companies vere: the national coal company, Air France, the General Maritime Corporation and -- the only repeater from phase one -- the SNCF. The coal company, Charbonnages de France, vas given a freer hand in pricing. This was -13- combined with a strong demand for coal as oil prices maintained high levels; these factors led to the company breaking even in 1980, after adding in government's subsidy (since 1951, the accounts had not been in balance even with the subsidy). Company managers hoped that the CP would "put an end to the arbitrary regime" under which they had labored. The expectation was that U ks_t evaluation of results would henceforth replace detailed « ante reviews and approvals. Management wished to gain the power to run the operation as a commercial firm; government's role would be to provide the company with access to sufficient investment capital to allow it to modernize key facilities. But with the election of the socialists in 1981 the company's top management was changed. The new managers simply ignored the conditions of the CP negotiated by the previous regime, and the new government reversed previous decisions to close several old, unproductive, costly mines. In the last year of the CP, government subsidies to the coal company grew rapidly. 19. The second CP negotiated with the SNCF gave limited power of price setting to enterprise management. The major innovation of this particular document was its use of quantified performance indicators, physical and financial, to be used in the evaluation of results and the rewarding of management. This feature has been widely adopted in subsequent CPs, in France and elsewhere. But management's latitude in price setting was marginal, and in a most important cost area, that of personnel, management was powerless. (French CPs do not normally Lackle head on numbers or pay rates of personnel; except in the case discussed in the following paragraph these matters are handled by national conventions. The failure to deal directly with personnel issues has been a deficiency of the program.) In -14- the first two years of the CP several uneconomic passenger lines were closed, which led to an improvement in the financial picture. Here as well, however, the victory of the socialists led to a government order to freeze fares and to reopen most of the closed lines. Levels of subsidization grew; the financial situation worsenad. 20. The shipping company CP was more of a restructuring plan than a contract; that is, it vas a statement of directives, along with a warning of what vould happen if performance did not improve. In essence, the company vas viewed by government in 1979 as having too many advantages and workers, and the CP's goal was to suppress the former and -- in a rare move -- reduce the number of the latter. A capital sibsidy was specified and the compensation policy for the money-losing Corsican service was established; but the main components of the CP were the directives to reduce the workforce by 25 percent, and reduce the number of ships from 98 to 36. The CP stipulated that if the goals were not clearly met in three years, the government could exercise a right to sell or liquidate the firm. Yet again, the 1981 election changed government policy; the CP was essentially abandoned and the company allowed to survive, apparently with the same number of workers. 21. The CP process in Air France was -- and continues to be -- a much more positive experience. This vas the first CP in an enterprise operating in basically competitive markets. Largely because of this factor, the negotiated CP gave the airline a greater degree of managerial autonomy than was accorded other firms. Since commercial profitability vas a fairly accurate measure of Air France's success, the goals of the CP could be -15- specified with greater precision. The deviations from commercial operation vere also easily demarcated: the use of inefficient Caravelles and the Concorde, the splitting of Parisian service between Orly and Charles de Gaulle airports, the maintenance of certain lines with uneconomic load factors -- the agreed-upon compensation for these non-commercial activities imposed on Air France could be and was precisely calculated. 22. The targets set in the CP for growth in the level of operations, financial objectives, and for the key physical indicator of number of flight hours per year, by aircraft type, were met or exceeded in 1978 and 1979, and vere just short of targets in 1980. The result vas that the firm turned a profit in 1980; "one of the very few European airline companies to present positive financial results" in that difficult year. It is not clear from available documentation how the government's "capital endowment" to Air France figures in the calculation of profitability. Nor can one gauge the role of factors outside the control of management, such as the reduction in the price of fuel. Nonetheless, what is clear is that fitancial targets set in the CP were met or exceeded in each of the three years covered in the period. This was enough to term the process a success. Air France has had its CP renewed every three years since. Third Phase: 1981 to the Present 23. After 1981, CPs of a more or less traditional type continued to be installed in several of the monopoly or oligopoly public enterprises with heavy social service obligations: Gaz de France, EDF, Air France, and -- -16- for the third time, following lengthy and difficult negotiations -- the SNCF. But in the first half of 1983 the Socialist government also installed what it insisted vas a quite different type of CP in thirteen industrial and manufacturing public enterprises, several of which had been nationalized in 1981-82.6 To show that the new contracts in the industries differed from past activities in the utilities, the agreements were, finally, given the name by which they are today generally known, "contract- plans." In these CPs, enterprises were to be given only guidance and policy orientation of a long-term, strategic nature; operational tactics were, in principle, left to managers and boards. (But similar promises had been made concerning the earlier processes.) The documents drafted did place much less emphasis on productivity and financial targets; they stressed instead the creation of new product lines, areas in which the firms would concentrate their R&D efforts, the maintenance or regaining of a set amount of market share, and, in general, harmonization of the efforts of this group of state enterprises working in basically competitive markets with the overall goals of the technologl.cal modernization and enhanced international competitiveness of French industry. The Socialists' conception was that CPs should link long-term performance of the enterprises to social aspirations outlined in the national plan. i/ The thirteen were: CGE, Thomson, Bull, Renault, Usinor-Sacilor, CdF- Chimie, EMC, St-Cobain, Pechiney, Aerospatiale, SNECMA, Matra, and S.N.P.E. The last two firms produce armaments and are under the control of the Ministry of Defense; the first eleven are supervised by the Ministry of Industry and Research. -17- 24. The CPs vere negotiated vith parent corporations; subsidiaries and affiliatca were not dealt with directly. The CPs were designed to be all- encompassing; i.e., to replace all other forms and procedures of control. They were also supposed to be supervised by a single tutelle agency, the appropriate ministry of technical supervision. No other ministry or body was to play a direct role in the supervisory process. (Even the spokesmen for the system admit the difficulties of maintaining this position of each firm dealing with a single offic,.al interlocuteur.) Another difference in the post-1981 process was that, for the first time, workers' representatives were conculted and involved in the CP negotiations. 25. For several reasons, the impact of this last group of CPs is particularly difficult to judge. First, the details of these CPs were, at the insistence of the enterprises, kept confidential. There was a fear that compbtitors would reap unfair advantage by studying the details of the agreements. Thus, one cannot compare post-CP performance to the projections in the agreement. True, the essential features of the CPs vere published in official documents and press releases, but the fact remains that the full texts are unavailable. Second, eveni if they were they would be of limited value as mechanisms to evaluate performance. As noted, this group of CPs concentrated on long-term and general objectives, aot quantified short-term targets. It is only now, as one approaches the fifth anniversary of their signing, that one might be in a position to assess the achievements. But in the interim the circumstances have -- once more -- changed; first, with the return to parliamentary power of a center-right government, and second, in 1988, the re-election of Mr. Mitterand and the re-installation of a Socialist cabinet. The result has been the - 18,- reprivatization of several of these industrial units, the diminution of the importance of plannîng in the French systen, and a general reduction in importance of the contractualization approach. 26. Following the wave of nationalizations and the drafting of the new CPs, the government took general stock of the public enterprise situation in the utility sector. The conclusions of its review vere disquieting. Although the CP process had been brought into being to combat the poor financial performance of public enterprises, especially their persistent tendency to drain budgetary resources, net transfers from government to public enterprises had increased from 12.8 billion F in 1975 to 49.5 billion F in 1982. It appeared that these increases took place despite, not because of, the CP; it was argued that unforeseen shifts in economic and political circumstance had brought about the higher level of funding. But this excuse only seemed to illustrate the marginal nature of the CP. The implication was that CPs could do little to predict, much less offset, market conditions. This raised the question of whether the CP process has been worth the massive administrative and supervisory investment made by the French. At the core of the CP process is the capacity to project forward quantified statements concerning production, costs, investments, borrowing, vage bills, capital structure, etc. If unpredictable changes in polltical and economic circumstances invariably lead to targets, inticators, ratios and particularly commitments becoming irrelevant and untenable, then what is the utility of the exercise? What was the French goverrment and public gaining from the exercise? -19- 27. It vas to combat this problem of coping with uncertainty that the second and third phase CPs had been reduced in length from 60 to 36 or at most to 48 months. Secondly, the Socialist governuent, in line with the philosophy shown in their industrial sector CPs, altered the objectives of the process. In essence, the government increased the social as opposed to technical content. The modified objectives of the CP process became: (i) assure coordination between the medium term objectives of the enterprise and the policy of the State; (ii) associate the enterprise in the "realization of objectives of national interest;: (iii) mobilize the personnel of the enterprise to support the defined objectives; (iv) allow autonomy of management; and (v) define and specify the financial relations of the enterprise and the State. These goals were more general, less technical in nature, than those put forward by the ambitious Nora Report; and as such they were less likely to be called into question by shifts in economic conditions.7 28. The goals had been redefined in light of the four major lessons which had emerged from previous experience. These lessons were: a. Keep contract-Rlans short and simple. Initial attempts were too ambitious, long and complicated; they attempted to specify too minutely the economic future and the actions and 1/ The lists of objectives (above) and lessons (below) are found in J. de Chalendar's paper, cited above in footnote 4. -20- reactions of the enterprise. The preferred CP became shorter, and concentrated on grand aims and principles. b. Keep CPs flexible. Every CP should be treated as a "sliding" or rolling plan; every aspect is subject to renegotiation for change if the numerous assumptions concerning future conditions prove inaccurate. CPs now have built-in provisions for review and modification on an annual basis; even more frequent review is possible. c. "The existence of the contract-plan is more important than its content." That is, the utility of the CP is less in its quantified performance indicators or goals and more in the formalized periodic exchange of views between government owners/supervisors and managers. It is the very process of being exposed to the o.her eide's point of view while negotiating an agreement which is the major and justifying accomplishment of the contract-plan, even though the specific numbers prod-ced in the exercise are sometimes misleading or outrightly incorrect. This is a very important finding; the activity is seen as beneficial in and of itself, not necessarily because of material outputs. d. Fourthly, in general, but particularly in periods of economic crisis, the weaker the performance of an enterRrise. the greater the difficultv of negotiating and imnlementing a contract-Rlan. Negotiations with the SNCF for its third CP took close to two years, and produced one of the thicker and -21- more complicated of documents (31 pages of text, 44 articles, seven annexes). Negotiations with the national steel company took place over several years starting in 1979, but never led to the signing of a CP. Both were cases of generally weak past performance, combined with particularly high levels of uncertainty with regard to future demand. It would seem that there must be a modicum level of both economic stability and optimism concerning the future of the firm before enterprise officials can envisage committing themselves to a concrete set of short-term goals and obligations. 29. Finally, the reviewers recognized a f-ndamental problem with CPs, a problem that has proven even more intractable iii the developing countries which have attempted to use the device: the agreement is not really a contract for the simRle reason that the state cannot be sublected to legal Rroceedings. bv the enterDrise. in t-he event that it fails to honor its obligations. Repeatedly in France, this has proven to be the common case and the evident difficulty. The state could agree in the abstract to rate or fare increases, and even commit itself on paper, in a document modeled on the language of a legally binding contract, to grant firm X a tariff increase of Y percent at time Z. But all too often, when the moment came to honor the commitment, the political costs appeared to outweigh the financial and economic benefits. Thus, fare increases tended to be reduced or postponed, several uneconomic factories or lines slated for closure were kept open, the workforce in various firms was maintained at excessively high levels, etc. Enterprise management could negotiate, complain, bring to bear what external pressure they could muster, and threaten to resign to -22- force the point; but they could not invoke the povers of the judicial system. The CP is thus not really a contract; it in an agreement which is only as good as the degree of government commitment to it. 30. French analysts have tended to judge the utility of the CP process on the basis of whether the set targets -- physical, financial and social -- have been met, and whether the actors involved in the process, from both the government and enterprise side, have assessed the CP as being worthwhile. These are important issues, but they do not adequately address two fundamental questions: What has been the comnarative performance of publie enterorises before and after the introduction of CPs; and what bas been the comparative oerfortance of enterprises with and without the CP? On both of these critical counts, information is, at best, sketchy. 31. Before and after: As indicated, performance of a number -- though by no means all -- of the French public enterprises submitted to the CPs, at various points in the process, showed financial improvements. It seems reasonable to think that some responsibility for the improvements could be attributed to the CP process, for example, in Air France, EDF and Charbonnages de France. But in all but Air France, improvement was short- lived. Either renewed or unanticipated external shocks, or government failure to take or sustain cost-cutting or revenue enhancing measures, or both, usually intervened. This may indicate that CPs can only guide and positively promote enterprise activity during periods in which the initial set of assumptions holds tiue. It might further be argued that the period in which the CP process was introduced -- 1970 through 1984 -- was an extraordinarily unsettled period, economically and politically, and that it -23- vas simply unlucky timing which accounts for the difficulty in comparing firm performance before and after the introduction of the CP. But this is not a particularly strong argument since all periods are unsettled to those living through them; politicians always have short-run survival interests which conflict with long-run economic gains; this is simply a constant. An ideal mechanism would deal with this uncertainty; though, admittedly, that is asking a great deal. A reasonable conclusion is that the CP seems to have been capable of influencing but not determining enterprise performance. It vas of significant but secondary importance, and its impact was overshadowed by macro-economic changes and shifts in government policy and commitment. As is the case vith most management improvement mechanisms which attempt to substitute goal specificity, set by administrators or by negotiation, for market mechanisms, the information problem quickly becomes overwhelming. That is, the beneficial impact is related to the degree of realism of the assumptions and projections contained in the CP. As external conditions shift, the assumptions no longer depict accurately the real economic world, and the directing force of the CP weakens. 32. As for with and without, the discussion is even more problematic. For example, to what does one compare EDF? The performance of other European electricity corporations? Or to some set of synthetically derived performance norms? Increasingly, it has been the latter measurement which is regarded as promising. However, it has been the CP process itself which provided some of the first attempts, in what remains a new field, to construct performance indicators of this type. Only in the case of Air France, which could use many straight-foryard market signals as a guide to -24- good performance, can one trace out consistently positive financial results in `Lne with the goals set in the CP. (An economic assessment would be much more difficult.) But this raises another question, i.e., to what extent can positive financial results be attributed to the existence of the CP? It could be argued that the prime factor which led to Air France's good performance in the early 1980s was the decline in energy prices, and that this price decline was external and had nothing to do with the airline's management or internal procedures. But other airlines also benefitted from falling oil prices, and most failed to translate this gain into a demonstrably improved financial position. In brief: Air France did fairly well when most European airlines, 3perating in roughly the same economic environment, did not. During the period of good performance Air France was operating with a CP when the other airlines were not. The CP was thus associated with comparatively good performance. To what extent was ':t responsible for it? 33. Available data do not allow one to say. The issue of assessing rigorougly the economic and financial impact of the CP process is complex. The episodic experience of Air France, or the other seven enterprises, scattered over a seventeen year period, in a variety of activities, and operating in rapidly shifting policy and economic environments, does not add up to a comprehensive or rigorous test of the concept. Thus, an emphasis on whether physical and financial targets have been met, and the subjective assessments of whether or not participants approve of the process, is about all French observers reasonably could be expected to do. -25- Conclusions on the French Exoerience 34. The conservative government in power in France from 1986 to 1988 concentrated its energies on the privatization of public enterprises. They started the process with manufacturing and financial sector firms, operating in competitive markets, many of which had been rationalized by the socialists in 1981-82. Privatization was that government's economic preoccupation in the public enterprise sector; reform of PEs retained in State hands in general and concerns of the CP process in particular slipped to secondary importance.8 However, all of the sixty-five firms -- the thirty privatized and the thirty-five additional firms which were to be privatized before 1990 -- are in competitive sectors. Unlike Great Britain, the French have no announced plans to divest the natural monopolies in the utility sector. This suggests that CPs, or perhaps yet another modified version of the CP, will continue to be used to guide, orient and evaluate the performance of major French public enterprises; a possibility increased by the results of the 1988 elections. 35. The French have invested a fair amount of administrative and intellectual resources in the CP process (though the marginal difference in administrative costs may be slight, since PEs were being extensively supervised before the introduction of the concept). In most instances, performance has not dramatically improved -- though note must be made of ./ A summary of the privatization experience, to the end of September, 1987, was vritten by the then Minister of Finance, Edouard Balladur: Je crois en l'homme plus qu'en l'Etat (Paris: Flammarion, 1987). -26- Air France's consistently respectable performance under the CP regime. Failure to improve performance may not be the "fault" of the CP; but the process's inability to protect the contracted public enterprises from the vicissitudes of uncertainty has proven disappointing to the concept's originators. In several French instances where performance has improved, either temporarily or enduringly, it is difficult to do more than correlate the presence of the CP with the improvements; one cannot say what percentage of the positive (or negative) variance is attributable to the CP. At no point in the process have the French attempted to link management or worker incentives with fulfillment of a CP's objectives. In France, this issue seems to have been generally ignored. Still, the actors in the process, managers and government supervisors alike, support CPs because they enhance the clarity of goals, the transparency of operations and achievements, and the ease of evaluation. These are advances. But, partly because the costs of creating the CP have not been calculated, one cannot tell if benefits justify the investment. Il. The Experience of Developing Countries 36. As noted, the first recorded attempt to export the CP process was in 1973, in a single Tunisian textile public enterprise. It never proceeded to the implementation stage. The first concerted effort to replicate the CP process in a developing country took place in 1980-82, in Senegal. -27- A.Senegal 37. The Senegalese case illustrates the thesis that there are similarities in the problema faced by public enterprises in industrialized countries and in developing countries. The distinctions are mainly a matter of degree rather than of kind. Ambiguous or contradictory goals, politically determined deviations from profit-making or profit- maximization, efficiency-reducing bureaucratie control and supervision systems, financ'ng and investment decisions based on inadequate or inappropriate analysis -- these and other deficiencies are seen in Europe as well as Africa, in North America as well as Asia. However, some aspects of these problems in developing countries are particularly intense. For example, rarely if ever would one find a European public enterprise failing completely to produce accounts, or being incapable of submitting accounts to auditors for several years running, or having no internal financial control systems, or not receiving payments, for its goods or services, from the government itself, to the tune of millions of dollars of arrears. Yet these difficulties are common in LDC public enterprises. It vas to attack grave deficiencies of this nature, and to begin the process of resolving the large scale, systemic, economie and financial problems these deficiencies led to, that the World Bank and the Government of Senegal, in 1976, launched the preparations which vould lead to the signing of a Technical Assistance Project to the Parapublie Sector. It vas the first of its kind in Africa, and in the Bank.9 k 91 The details of the project, and an assessment of its accomplishments and shortcomings, can be found in the "Project Completion Report - Senegal - First Technical Assistance Project to the Parapublic Sector," World Bank, January, 1985. -28- 38. A 1977 Bank review of the Senegalese PE sector had stated: Supervision needs. to be carried out on a contractual basis for enterprises to be able to manage themselves. Management by objectives could thus meaningfully be introduced under which each enterprise would negotiate with Government annually the targets to be set for it and vould then be judged accorded to the results it achieved- This idea vas not pursued in the early stages of the ensuing project, which entered into effectiveness in 1978. It was not until 1980, and partly in response to the project-assisted improvement in the flow and accuracy of information on the poor performance of public enterprises, that the government decided to: - adopt CPs, - create a CP "Cell" in the Prime Minister's Office, and - ask the Bank for assistance with this aspect of PE reform. The objectives vere the same as those which had inspired the French CP process: to clarify goals; to coordinate the various government supervisory bodies, which either offered conflicting directives or operated according to different schedules; and to commit both the government and the enterprise to a clear course of action which would result in higher productivity and lower costs. 39. Between March of 1981 and May of 1988, contract-plans vere signed in nine different public enterprises. Since three of the nine are now in their second CP, a total of twelve sets of negotiations have been concluded. The first three Senegalese CPs (rural development agencies receive a more preliminary agreement, called a "lettre de mission," roughly -29- translatable as "statement of objectives") vere prepared between March of 1981 and January of 1982, in SOTRAC, SAED and SODEFITEX. Respectively, these firms operated in the fields of transport, agriculture and rural/regional development. In 1984, the government decided to renew the process in SOTRAC, SAED and SODEFITEX, and to apply CPs to other major enterprises. CPs or "lettres de mission" vere subsequently instelled in SONATEL (telecommunications), SONEES (water), SENELEC (electricity), SOMIVAC, SODEVA and SODAGRI (agriculture/rural development). CPs are presently being negotiated in OHLM (housing), OCPE (post and savings), RCFS (railway), and PAD (the port of Dakar). Thus, in terms of number and range of types of PEs affected, Senegal now has the largest CP program in the world. 40. Thf 1977 diagnosis of the sector had revealed fundamental problems in a number of important PEs. In some cases -- SOTRAC and SAED for example -- it was thought that the installation of a CP would be the main improvement and restructuring method in and of itself. In other cases, most notably that of the telecommunications/post office, the OPT, it was decided that problems vere s0 severe that reorganization should precede the application of the CP. In the case of the OPT, the result was a splitting of the firm into two enterprises, both of which (SONATEL and OPCE) were then subjeci.ed to the CP process. This raised a question which is not yet resolved: should CPs be used on the most severely troubled enterprises, or should their use be reserved for non- or mildly-troubled companies? Developing countries tend to use CPs as rescue devices for poorly performing firms, with questionable results. The Air France experience -30- indicates that their utility as improvement devices in basically sound, quasi-competitive firms has been insufficiently examined. 41. The first set of Senegalese CPs followed closely the French conception and practice. In line with the lessons learned in France in the 70s, they were fixed for three year periods. Good examples of the first group of CPs, illustrating the range of bodies to which the process was applied, and of subsequent implementation and modification problems, vere the cases of SOTRAC, the Dakar region bus company, and SAED, the Senega]. River and Delta Development Authority. (It must be noted that of the first three, only the SOTRAC CP was carried through to conclusion, while the SAED agreement came close; the third, and others whieh had been considered but never negotiated, were all dependent, for the funding of their investment programs, on resources to be made available through the first Structural Adjustment Credit, negotiated with the World Bank in 1981. Due to a dispute over agricultural prices, only the first tranche of this credit Was .eleased. Given the unavailability of funds, two of the first five CPs were never negotiated, much less implemented.) The First Phase 42. The first Senegalese SOTRAC CP was a short document of eight articles, stating the purpose of the enterprise, the assets to be deployed during the period of the CP, projected performance, obligations of the government, and means of implementing and modifying the CP. SOTRAC's future position, the agreement stated, could only be secured by dealing with three critical issues: renewed investment in vehicles, a paying off -31- of the substantial arrears (1.1 billion francs CFA) in goveriument subsidies, and government permission to increase prices. The agreement specified the three year investment program of SOTRAC, and projected a stabilization of losses, as opposed to substantially increased losses which might normally be expected during a period of massive reinvestment. Improvements were projected in the cash flow posicion. 43. The government's obligations were clearly spelled out. Gover.nent zgreed to a precise timetable to pay off the arrears on previously promised subsidies, and to pay future subsidies according to a set schedule. SOTRAC was declared exempt from normal duties and taxes on the materials it imported. The national government promised to "do its utmost" to persuade the communal government of Dakar to bear some of SOTRAC's subsidy, maintain the regional road network, and assist in the construction and maintenance of SOTRAC's terminals. 44. A technical innovation of the SOTRAC CP was the use of three different pricing hypotheses. That is, during negotiations the government committed itself to alloving SOTRAC a price increase during the CF period, but government representatives vere unable to specify the amount by which prices would be authorized to rise. The CP posited three different scenarios, which traced out differing financial results according to the amount of tariff increase allowed. 45. The CP covered the main issues of enterprise purpose, prices, investment capital, financial results and CP monitoring. It did not cover performance indicators other than net financial results. This was an -32- inadequacy; commercial profitability is an imperfect indicator of the performance of a monopoly, heavily subsidized public enterprise, especially one receiving the indirect subsidies mentioned in para. 43. (This shortcoming vas corrected in the second SOTRAC CP.) 46. In contrast to SOTRAC, the CP -- termed a 'lettre de mission" -- with SAED (the Senegal River and Delta Development .:athority) was a long document, produced after complicated negotiations. Covering the period 1981-84, the agreement was more a restructuring device than a contract; its purpose was to resolve fundamental operational difficulties of an extremely expensive and inefficient rural development agency. This was the first time, anywhere, that the contract process had been applied to an enterprise of this type (though it is not clear whether the targets were negotiated, as they would and should be in a CP, or whether the goals were simply imposed by government directives). The document laid out a variety of quantified production goals for SAED: number of undeveloped hectares to be turned into farmland, number of existing farms to be expanded, and production and marketing targets for paddy, tomatoes and maize. The agreement dwelt on the reorganization of SAED, emphasizing the need for management information and budgetary control systems, rational personnel management and improved accounting and expenditure control, the details of which were spelled out. Government obligations vere stated more vaguely: for example, the CP noted that the State "would be required to provide (SAED) with adequate working capital;" ther-. was an admission that only half of the minimally essential investment funds were available, and that "Government now has to find the remaining 50 percent, i.e., CFAF 5.435 billion;" and on the crucial matter of farmgate pricing, the agreement was -33- less than legally precise: "Government will maintain producer prices at levels sufficiently remunerative not to interfere with debt servicing and sufficiently attractive to stimulate production.' 47. The CP set out the schedule for government's quarterly subsidy to the enterprise, necessitated by "the difference between the billing rate set by the Government and the actual cost of production and services rendered;" and to cover those operations imposed on SAED "for which there is no billing provision." A second table presented the estimated payment dates for the clearing up of previously promised subsidies, now substantially in arrears. The document concluded with a statement that it was "binding on the two parties." It was signed by the Director Gerneral of SAED, the Chairman of SAED's Board, and -- for the government -- the Prime Minister, the Minister of Finance and the Minister of Rural Development. The letter of understanding had been conceived, prepared, negotiated and signed in a period of seven months. (This appeared long at the time; subsequently, preparations for most CPs have taken much longer.) The overall result was a document of some thirty pages, not including annexes, but with most of the specificity on the side of the enterprise's goals; the government's obligations remained vague. Fundamental Problems 48. The initial experience with CPs was judged as only moderately positive. Management autonomy did not increase to the anticipated degree. The lack of physical performance indicators in the SOTRAC CP proved to be a problem, as did the shortfall in specificity of government obligations and -34- undertaklngs in the SAED agreement. Hovever, from the outset and continuing to the present, the two more important problems with the CP concept in Senegal vere: (a) the government's inability to keep its financial promises, and (b) the enterprises' -- and the supervising agencies' -- inability to force, pressure or cajole the government into honoring those commitments. In SOTRAC and SAED, and later on in almost every enterprise submitted to the CP process, the government committed itself in the CP and then either failed to settle its arrears to the firm, or failed to honor an agreement to pay off subsidy sums promised in the past, or failed to meet timetables on current subsidy payments, or failed to inject promised amounts of equity, or failed to give final approval to previously agreed upon investments, or failed to allow price or rate increases which vould have increased the self-financing capacity of the enterprises. In most cases, a combination of these defaults vas present. 49. Neither Government's lack of resources nor its lack of financial discipline vas caused by the CPs; rather, the CPs rendered transparent and quantified government's shortcomings -- and it can be argued that these were necessary and progressive steps. Drawing attention to a problem is a step in overcoming it, but despite repeated flagging of the issue it remains unresolved to the present. The Government recognizes the problem; in its 1988 review of the "State of Application of the Reform of the Parapublic Sector," the chief PE supervising agency in the Office of the President stated: "Non-respect of the State's financial commitments remains the central problem with lettres de mission and certain contract- plans... The result is that the credibility of the CPs is veakened..." -35- 50. The general lines of the problem would have been familiar to the French; but in the Senegalese case the difficulty vas of a greater magnitude and intensity, in terms of the number of CP'ed enterprises in such difficulties, and the size of the gap between what government had said it would do, and what it actually did. Several factors contribute to this problem. First, the entire Senegalese planning tradition and experience has consistently tended to produce extensive and ambitious lists of desirable activities, vithout sufficient specificity on where the money vill come to pay for them.10 CPs have proven to be but another example of this well-entrenched process. A second factor contributing to the financial unreality of the CPs has been that they were all produced, in first draft form at least, by consultants who naturally attempted to produce agreements which vould improve the position of their main client, the firm to be CP'ed. Moreover, and again naturally, they were produced in isolation one from the other; no drafter of one CP knew the budgetary implications of any other CP. Thus, the draft CPs tended to specify financial conditions which would allow the firm to carry out what management and the drafting consultants saw as the enterprise's mandate. The viewpoint was that of adding all resources necessary to rebuild the enterprise; the notion of what the State could afford was secondary, if taken into consideration at all. 1Q/ See Pierre Guislain, "Les Contrats-Plan dan les pays en developpement: Un moyen d'organiser les relations entre le gouvernement et les entreprises publiques en vue d'ameliorer les performances desdites entreprises,' World Bank, LEGWA, May, 1987. -36- 51. The first level of review for the draft CPs is, theoretically, in the technical ministries. Officials and managers interviewed state that technical tutelle ministries almost always approve, without question or deletion, whatever proposals are put forward by the consultants/enterprises. No global consideration of what Government can afford takes place at this level. Indeed, one could argue that it is unrealistic to expect the technical ministries to restrain the growth or operations of "their" enterprises; that they are quite content to see their PEs expand. This means that responsibility for global consideration of the aggregate financial impact of the CPs rests with the central agencies, the Delegation for the Reform of the Parapublic Sector in the Office of the President, and the Parapublic Sector Cell in the Ministry of Finance. 52. These bodies have not done their job. These central reviewers have tended to: - accept uncritically the draft CPs. At first, they often failed to appreciate the financial implications of the agreements, individually much less than in the aggregate. The resource implications surfaced later, and either delayed the signature or implementation of the CP, or only became apparent after signature, leading to problems discussed below. - More recently, Ministry of Finance reviewers of CPs have recognized that financial commitments called for in some CPs vere unrealistic or excessive. They have made note of their concern, and called the attention of higher officials to their view. However, they have been told that they must approve and sign the CPs, even though they know the promised resources will not be available. The justification has been that the political situation requires the signature of the obviously unrealistic -37- document, and that the resources might surface (from some unspecified source) during the course of the CP. 53. In short, central officials have regularly approved expenditure forecasts which, when sumed, far exceeded government's resources. They then applied what one might call traditional budgeting techniques, which can be summarized as: use whatever resources are on hand to deal with the most acute crisis, as determined by the pol.itical process or leadership; ignore previously determined budgets, agreements, commitments, contract- plans, etc., and give the squeakiest wheels the smallest amount of oil which will quiet them for the moment. The CPs were supposed to solve this problem but they did not, due to the simple fact that the CPs, taken together, committed the government to spend far more money than it had available. Evaluation and Revisions in the Process 54. Despite the overall financial problems, can one perceive differences in micro-performance between enterprises with CPs and those without? Only one study has been undertaken which attempted to compare results. Performance of the first six PEs with CPs was contrasted to the performance of a group of eighteen non-CP'ed public enterprises. The analysis reviewed the period 1980-83. Both groups ran losses in these three years; and both groups reduced their rate of loss by about the same percentage at the end of the period. Perceptible differences occurred in only two areas: the contracted PEs had a growth in turnover of 93 percent in the period, as opposed to 66 percent in the non-contracted firms; and -38- personnel costs rose 13 percent in the contracted firms, versus 45 percent increases in the non-contracted enterprises. Levels of debt reduction vere about the same in both groups. For several reasons, these moderately positive results have to be interpreted cautiously. First, the indicators showing variance a:e of secontdary importance. The major performance indicators of profitability, self-financing ratios and return on capital invested were the same (and poor) for both sets of firms. Second, the differences seen are of questionable size and significance. Third, the period reviewed was quite brief. Fourth, the enterprises compared differed greatly in structure, purpose, market position, etc. It is difficult to argue that like vas being compared to like. Nonetheless, despite the drawbacks of and qualifications to this study, it deserves praise as the only attempt to &ssess quantitatively the impact of the CP. More and better studies of this nature are required. It would be very useful to mount a detailed comparison of performance of firms with CPs as opposed to those vithout, as well as a comparison of performance of enterprises before and after the introduction of CPs. 55. Regardless of the lack of major improvements in performance, and as was the case in France, Senegalese managers and government supervisors working with CPs have almost universally concluded that they are beneficial devices, the use of which should be continued. Interviews with eighteen Senegalese officials and managers, concluded in January of 1986, indicated strong support for the CP process. Managers approved of the heightened clarity of goals, the specification of government's financial responsibilities, and the concept of government compensation for non- commercial activities imposed on the enterprise. Officials saw the major -39- advantage as the increase in the autonomy and responsibility of management. (One could argue that both groups liked those features which h "-ighted the responsibilities of the other party.) Indeed, all involved .arties, including project officers from the World Baik, have found the CP irresistibly appealing. In theory, it clarifies objectives, determines priorities, allocates resources, guides and stimulates productive behavior and eases evaluation. 56. The CP appears to be the perfectly logical response to an intractable set of problems and needs. Though its application has failed, at least in the short run, to produce dramatically improved performance, it is nonetheless regarded as a worthwhile innovation because of the constructive dialogue it creates between managers and owners. The problem is that CPs do not force financial discipline on the government. In response, the tendency has been to shift the criteria by which its utility is judged (as in France), or to hope that modifications vill increase its impact (the Senegalese approach). The Second Wave 57. The workings of the second group of CPs (see para. 39) showed a continuation of the problems. (In the case of SENELEC, the CP vas not signed until 1987.) The SONEES example, while perhaps something of an extreme case, reveals the workings of the process. The main features of the agreement vere increases in the tariff structure, and a rehabilitation/investment program, to be funded through settlement of the governsment's outstanding debts to the company. The functioning of the CP -40- was subjected to a review at the end of 1985. Major difficulties were apparent. In the review period, despite the enactment of the price increase, revenues of SONEES were 16 percent lower than anticipated; costs 5 percent higher. But the critical problem was the usual one: "neither the State nor the local governments have respected their commitments" to settle outstanding accounts. The result was a deficit five times greater than anticipated, and a reduction by half in the essential rehabilitation/investment program. 58. As of June, 1985, the total public agency debt to SONEES was 5.5 billion F CFA. The company's financial situation was desperate. Operating costs were being financed by short-term bank overdrafts, at high interest rates. Suppliers vere demanding higher and higher advance payments, or adding substantial "waiting premiums" to their bids, or -- most commonly -- simply refusing to do business vith the company. SONEES stopped paying its bills, with the result that it was threatened with stoppage of electricity services, among others. "These financial difficulties have led us to juggle our creditors, and put them off as long as possible." The company cancelled all training programs, ignored scheduled maintenance procedures, and froze emplcyment in an attempt to cut costs. The negative long term implications of these measures -- save perhaps the last -- is evident. 59. In accord with the provisions of the CP, SONEES approached the Ministry of Finance. Despite repeated resubmissions of their case, they received no reply of any sort during the whole of 1985. The simple fact was that "...despite the provisions of the contract-plan, no deposit has -41- been made to our account since 1 January 1985." The suggested solutions were obvious: government and the other public agencies must pay at once their bills; and steps must be taken to see that the arrears do not simply start again. SONEES officials suggested that at the start of each fiscal year the government should deposit in SONEES' account 65 percent of the previous year's budgeted payments. At the very least, pleaded the enterprise, government should pay its outstanding debts. A Detailed Example: The SONATEL CP 60. Despite the difficulties with enforcing the provieions of the SONEES' and other CPs, "contractualization" was made a key feature of the government's "New Policy Towards the Parapublic Sector," announced in mid- 1985. CPs were cited as one of the principal means by which enterprise performance would be improved. The policy stated that each year five enterprises would be submitted to the process. (The schedule has not been met.) Money-losing enterprises with a heavy social service obligation were to be given first priority. The policy implied that enterprises without a clear social service mission would be prime candidates for divestiture. 61. A positive example of the more recent crop of Senegalese CPs, showing how they have been modified in an attempt to deal with the prevailing high level of budgetary and economic uncertainty, can be found in the contract negotiated vith SONATEL. The CP was signed in 1986, after close to two years of negotiations. (This enterprise is the telecommunications parastatal which was created after the break-up of the -42- OPT.) The standard features of a CP were retained: statement of enterprise objectives, commitments of the government, and mechanisms of monitoring the agreement. To these were added several other features: a detailed set of physical, financial and quality of service performance indicators, several means tonpressure government ln case of failure to honor obligations, and a section on how the plan could be modified, quickly, ln case its financial and operational assumptions proved inaccurate. A review of the sections of this carefully prepared CP is revealing. 62. The preamble noted the high level of unsatisfied demand for telephone services in Senegal. It recalled government's decis'on to create a new, commercially-run and above all autonomous body to deal exclusively with this service (in line with the "New Policy..." the slogan of which was "manage less but manage better"). Section 1 (of seven) summarized the laws and decrees creating and regulating the company, emphasizing that the supervising ministries, technical and financial, "...carry out normally their roles, which consists of overseeing the strict application of the texts regulating the public service, but that they do not interfere, in either their advice or theié actions, in the daily management of the enterprise." Section 2 offered the objectives of the CP exercise: to determine SONATEL's performance criteria; to determine the strategy and the actions the enterprise would take to attain its aims; and to define the obligations of the State to the firm. The third Section set out the firm's objectives in the contract period: -43- to improve the quality of service; to complete on schedule several major investment and supply projects; to better utilize the firm's humen resources; to determine a schedule for giving to every citizen "easy access to a telephone;" and to improve the image of the national telephone service. Performance Indicators 63. Section four dealt with performance criteria. The enterprise committed itself in the contract period to hold constant the number of employees, and to limit the annual rate of growth of its vage bill to 10 percent. The main measures of physical production were "number of requests filled for new principal telephone and telex lines;" and growth of the total number of telephones in use. Specific, quantified targets vere fixed for each activity for each of the three years of the CP. The enterprise contracted to improve service quality, as measured by reducing the length of repair time, reducing the delay in obtaining a dia'. tone, and reducing the number of misdirected and incorrect calls dup to faulty equipment or switching. Again, precise numerical mc.asures of each of these indicators vas stated in the CP (for example, repair of a telephone within eight days of request would rise from 66 percent of cases in 1985 to 80 percent at the end of the